Disrupted glucocorticoid receptor cell signalling causes a ciliogenesis defect in the fetal mouse renal tubule.

Reagents and tools table Experimental models Reference or source Identifier or catalogue number GRcdKO, C57BL6/J ( M. musculus ) This study N/A GR-null, C57BL6/J ( M. musculus ) Cole et al, 1995 N/A mIMCD3 cells ( M. musculus ) American Type Culture Collection CRL-2123 Induce pluripotent stem cell kidney organoid, line 522.3 ( H. sapiens ) GM04522 NIGMS Human Genetic Cell Repository Primary antibodies Reference or source Identifier or catalogue number Akt rabbit polyclonal Cell Signaling Technology 9272 AMPKα rabbit monoclonal Cell Signaling Technology 5832 Ki67 rabbit polyclonal Abcam Ab15580 Pericentrin rabbit polyclonal Abcam Ab4448 CEP290 rabbit polyclonal Novus Biologicals NB100-86991 Dolichos Biflorus Agglutinin (DBA), biotinylated Vector Laboratories B-1035 Glucocorticoid receptor rabbit monoclonal Cell Signaling Technology 12041 Hoechst 33342, trihydrochloride trihydrate Invitrogen H1399 Nephrin sheep polyclonal R&D Systems AF4269 IFT88 rabbit polyclonal Proteintech 13967-1-AP JNK1 (2C6) mouse monoclonal Cell Signaling Technology 3708 KIF3A rabbit polyclonal GeneTex GTX134434 Lotus Tetragonolobus Lectin (LTL), biotinylated Vector Laboratories B-1325 MEIS 1/2/3 mouse monoclonal Active Motif 39795 Acetylated tubulin mouse monoclonal Sigma-Aldrich T6793 β-actin mouse monoclonal Sigma-Aldrich A5441 p44/42 MAPK (ERK1/2) rabbit monoclonal Cell Signaling Technology 4695 P-Akt (S473) rabbit monoclonal Cell Signaling Technology 4060 Phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) rabbit monoclonal Cell Signaling Technology 9101 Phospho-AMPKα (Thr172) rabbit monoclonal Cell Signaling Technology 50081 Phospho-S6 Ribosomal Protein (Ser235/236) rabbit polyclonal Cell Signaling Technology 2211 Phospho-β-Catenin (Ser675) rabbit monoclonal Cell Signaling Technology 4176 IAK1 (Aurka) mouse monoclonal BD Biosciences 610939 Gamma Tubulin rabbit monoclonal Abcam Ab179503 S6 Ribosomal Protein mouse monoclonal Cell Signaling Technology 2317 SUFU rabbit monoclonal Cell Signaling Technology 2522 β-Catenin rabbit monoclonal Cell Signaling Technology 8480 Secondary antibodies Reference or source Identifier or catalogue number Anti-mouse, HRP-linked Antibody Cell Signaling Technology 7076 Anti-rabbit, HRP-linked Antibody Cell Signaling Technology 7074 Donkey anti-Goat, Alexa Fluor 647 Invitrogen A32849 Donkey anti-Mouse, Alexa Fluor 555 Invitrogen A31570 Donkey anti-Mouse, Alexa Fluor Plus 647 Invitrogen A32787 Donkey anti-Rabbit, Alexa Fluor 488 Invitrogen A21206 Donkey anti-Rabbit, Alexa Fluor 568 Invitrogen A10042 Donkey anti-Rabbit, Alexa Fluor 647 Invitrogen A31573 Donkey anti-Sheep, Alexa Fluor 647 Invitrogen A21448 Streptavidin, Alexa Fluor 488 Conjugate Thermo Fisher Scientific S32354 qRT-PCR primers Gene Forward primer 5’ Reverse primer 5’ Atat1 TGTTACAGAAAGAGCGAGTGGA TGGTCTCCAGGTTGTAGTGC Aurka TCTCGGGTTGAATTCACTTTCC CCTTTGGCTTGCGTTGTGTT Ccp110 CCGTGGCAAAAGGGTTTC CTCGCAACTGAGCTAACAC Cep97 AGTCTGAAGGCTGAGTGG ACTCTTCATCTTGGCTCTGG Cep290 AATGAAGATGAAAGCCCAGGA TGCTTCTCAAGTTGGCGAAT Crisp1 TGGTCTTCTGCAATCCAAGG TGAGTTCCAAACAACCTGAGT Gbp7 CAGCTGCACTATGTCACAG CACGTCCCTCTAACTTCAGC Hba-a1 AAGCCCTGGAAAGGATGTTT CTCAGGAGCTTGAAGTTGAC Ifi203 CAGTGGTGGTTTATGGACGA CTCAGGAGGCACACATCTTT Ifit1 GCTGTCCGGTTAAATCCAGA AAGTAGCCAGAGGAAGGTGA Igf2 TCTCATCTCTTTGGCCTTCG CAAACTAAGCGTGTCAACA Kap TTAAGCACTGACTCTGAGCA ACTGTGATGTCTGTGTTCTCA Kif3a CGTGGATGAAATGAGGGGAA CCCATTGTAGCCTTCCAGAA Mndal TCGTCAAGATCAAGGTCACC ACGTGATAGTCTGGCATCTC Nedd9 GCGAGGAATCTTATGGCAAGG TCAAGCCCTCCTGTGTTCTG Oas2 AGTCTACTCCGCCTGATCAA TGTCAAAGTCATCTGTGCCA Oasl2 GATTAAGGTGGTGAAGGGAGG CCTGCTCTTCGAAACTGGAA Rab8a GCGAAGACCTACGATTACCT CCGTGTCCCATATCTGCA Rpgr CATCCGCTGCTCTTACTGA GCTCCCCATCCATTGTTACA Rps29 GGACATAGGCTTCATTAAGTTGG TCAGTCGAATCCATTCAAGGT S100g GAGCTGGATAAGAATGGCGA TTCAGGATTGGAGAGCGTG Chemical, enzymes and other reagents Reference or source Identifier or catalogue number Accutase Stemcell Technologies 07920 Charcoal, Dextran Coated Sigma-Aldrich C6241 CHIR99021 Stemcell Technologies 100-1042 Clarity Western ECL Substrate Bio-Rad 1705061 cOmplete™, EDTA-free Protease Inhibitor Cocktail Roche 04693132001 Dexamethasone Sigma-Aldrich D4902 Donkey serum Sigma-Aldrich D9663 Essential 8™ Medium Thermo Fisher Scientific A1517001 E6 medium Thermo Fisher Scientific A1516401 Fetal Bovine Serum, qualified, New Zealand Gibco A3160902 FGF9 In Vitro Technologies RDS233FB025 Fujifilm Medical X-ray Film Blue Sensitive Super RX-N Fujifilm 47410 19289 GlutaMAX™ Supplement Thermo Fisher Scientific 35050061 Heparin Sigma-Aldrich H3149 Immobilon®-P PVDF Membrane Merck IPVH00010 Matrigel Corning FAL35427 Mifepristone Sigma-Aldrich M8046 N,N,N′,N′-Tetramethylethylenediamine Sigma-Aldrich T7024 Penicillin-Streptomycin (5000 U/mL) Thermo Fisher Scientific 15070063 PhosSTOP™ Roche 4906845001 ProLong™ Gold Antifade Mountant Invitrogen P36930 SP Bel-Art Flowmi 70 Micron Cell Strainers Bel-Art Products H13680-0070 TGX Stain-Free™ FastCast™ Acrylamide Kit, 10% Bio-Rad 1610183 Triton™ X-100 Sigma-Aldrich X100 TRIzol™ Reagent Invitrogen 15596026 TrypLE Select Enzyme Thermo Fisher Scientific 12605028 Y-27632 (Dihydrochloride) Stemcell Technologies 72302 Software and packages Reference or source BIORAD 384w rtPCR software CFX manager Bio-Rad CIBERSORTx Steen et al, 2020 Cyclone Scialdone et al, 2015 DAVID Sherman et al, 2022 edgeR R package Robinson et al, 2010 FIJI https://imagej.net/software/fiji/ GraphPad Prism https://www.graphpad.com/features Imaris https://imaris.oxinst.com/ Scrublet Wolock et al, 2019 Seurat (v3.1.4) Butler et al, 2018 ; Stuart et al, 2019 STAR (v2.5.1b) Dobin et al, 2013 Microscopes Leica SP8 confocal microscopy Nova NanoSEM 450 scanning electron microscope ZEISS LSM 980 with Airyscan 2 Kits Reference or source Identifier or catalogue number 10x Chromium v3 kits Millennium Science PN-1000699 QuantiNova SYBR Green PCR Kit Qiagen 208052 QuantiTect Reverse Transcription Kit Qiagen 205311 Reagents and tools table Use of mice was approved by the MARP-2 Animal Ethics Committee at Monash University. Global GR-null (Bird et al, 2014 ; Cole et al, 1995 ) and collecting duct-specific GR-null (GRcdKO) mice were all of an isogenic C57BL/6J genetic background. GRcdKO mice were generated with HoxB7 promoter-Cre mice crossed with GR-floxed allele mouse (Yu et al, 2002 ). Fetal kidneys at E18.5 were dissected from embryos and either snap frozen in liquid N 2 or fixed in 4% paraformaldehyde. Tail snips were collected for genotyping by qPCR (Short et al, 2020 ). The E18.5 kidney-SC data was generated as previously described and is available at GEO ( GSE108291 ) (Combes et al, 2019 ; Data ref: Combes et al, 2019 ). SC RNA sequencing at E13.5 and E15.5 utilised wild type embryonic kidneys. The kidneys were dissociated in 500 µL Accutase (Stemcell Technologies) at 37 °C for 6–8 min, gently agitated every 2 min then washed with cold PBS 0.05% bovine serum, pelleted by centrifugation (400 ×  g , 5 min), and stored on ice. Samples were filtered with Flowmi Cell Strainers (70 µm, Bel-Art Products) and stained with DAPI before removal of dead cells by FACS (100 µm nozzle). Cell concentration was determined using a hemocytometer and adjusted prior to the generation of single cell libraries using 10x Chromium v3 kits. Sequencing data was processed using Cell Ranger (10x Genomics, v1.3.1,) and aligned to mm10 with STAR (v2.5.1b) (Dobin et al, 2013 ). Subsequent analysis was performed in the R statistical programming language using Seurat (v3.1.4) (Butler et al, 2018 ; Stuart et al, 2019 ). Quality control for the E13.5 and E15.5 datasets involved removing cells with 20% mitochondrial gene content (E13.5); 8% mitochondrial gene content (E15.5). Doublets were identified and filtered out using Scrublet (Wolock et al, 2019 ) or with HTODemux function in Seurat. Cell cycle phase was predicted using either Cyclone (Scialdone et al, 2015 ) or Seurat’s CellCycleScoring function. Cell cycle effects were regressed out and gene expression data normalised using SCTransform with default parameters. Following all quality control steps, the E13.5 dataset consisted of 19,252 genes and 4176 cells and the E15.5 dataset of 18,549 genes and 3294 cells. Cluster identity was determined by referencing top cluster marker genes ( FindAllMarkers ) to Combes et al previous analysis (Combes et al, 2019 ). The new E13.5 and E15.5 datasets reported in this study are available upon request. Total RNA was isolated from embryonic kidneys and IMCD3 cells using TRIzol TM reagent (Invitrogen, USA) according to the manufacturer’s instructions. Total RNA was analysed using a Bioanalyzer 2100 (Agilent Technologies, USA) and Next generation RNA sequencing (NGS RNA-seq) was performed by Genewiz Biotechnology, Suzhou, China. RNA sequencing (20 million reads) was performed on the Illumina Hiseq platform, in a 2 ×150 bp paired-end format. The gene expression count matrix underwent preprocessing and differential expression analysis with the edgeR R package (Robinson et al, 2010 ). Genes with low expression levels were filtered out, and the resulting log-transformed counts per million (CPM) values were utilized to create heatmaps. Quasi-Likelihood was used for statistical test. Sample N3 (PCA plot) was added as the extra covariate to the design matrix additionally to control versus knockout groups. Differentially expressed genes (DEGs) with a false discovery rate (FDR) less than 0.05 were identified and employed for generating volcano plots. Additionally, enrichment analysis was conducted using the DAVID online tool ( P value < 0.05) (Sherman et al, 2022 ). RNA-seq datasets reported in this study are available at GEO ( GSE290962 ). A reference count gene expression matrix was created by selecting the top 5000 most variable genes and randomly choosing 50 cells for each cell type from E18 developing mouse kidney single-cell RNA-seq data. This reference matrix was then provided to an online tool called CIBERSORTx (Steen et al, 2020 ). Additionally, the complete bulk RNA-seq count matrix was also provided to the software in order to generate the signature matrix and produce deconvolution results. These results display the percentage of each cell type in all bulk samples in a tabular format. cDNA was synthesised with a QuantiTect RT kit (Qiagen) according to the manufacturer’s instructions from the same fetal kidney RNA samples used for RNA sequencing. mRNA levels of Atat1 , Aurka , Ccp110, Cep97, Cep290, Crisp1, Gbp7, Hba-a1, Ifi203, Ifit1, Igf2, Kap, Kif3a, Mndal, Nedd9, Oas2, Oasl2, Rab8a, Rpgr and S100g were determined by qRT-PCR using QuantiNova® SYBR® green master mix (Qiagen) on a CFX384 Touch Real-Time PCR Detection System (Bio-Rad). Relative mRNA levels were normalised to the housekeeping gene Ribosomal protein 29 ( Rps29 ) using the ∆∆Ct method (Pfaffl, 2001 ). PCR products for each primer set (Reagents and tools table) were verified by a PCR melt-curve analysis and DNA sequencing. Differentially expressed ciliary genes from the RNA-seq data of GR-null mice were matched from the CiliaCarta (van Dam et al, 2019 ) compendium and selected for qRT-PCR analysis. Differentiation of human induced pluripotent stem cell (iPSC) derived kidney organoids were developed as previously described (Takasato et al, 2015 ) with minor changes. 80,000 iPSCs were seeded in Matrigel on a 6-well plate supplemented with Essential 8 medium (Thermo Fisher Scientific), 10 μM ROCK inhibitor Y-27632 (In Vitro Technologies) and 1% penicillin-streptomycin (pen-strep) (Thermo Fisher Scientific) on day 0. Media was changed to Essential 6 (E6) medium, supplemented with 4 μM CHIR99021 (In Vitro Technologies) and 1% pen-strep the following day. In total, 200 ng/mL of FGF9 (In Vitro Technologies) and 1 µg/mL of Heparin (Sigma-Aldrich) were added to the media on day 4. Fresh media was changed every 2 days. iPSCs were dissociated with TrypLE Select Enzyme (1×) (Thermo Fisher Scientific) and 150,000 cells were used to generate each 3D kidney organoid on day 7. Kidney organoids were treated with 4 μM CHIR99021 in E6 medium for 1 h and changed to E6 medium, supplemented with 200 ng/mL of FGF9, 1 µg/mL of Heparin and 1% pen-strep. Fresh media was changed every 2 days until day 13, where media was changed to E6 medium supplemented with 1% pen-strep. Kidney organoids were collected at day 20 for 48 h of vehicle (ethanol) 10 −6  M or dexamethasone 10 −6  M in serum free media to induce ciliogenesis. Mouse inner medullary collecting duct (IMCD3) cells were maintained in DMEM: Nutrient Mixture F-12 (DMEM/F12), supplemented with 10% FBS (Gibco), 1% l -glutamine (Thermo Fisher Scientific) and 1% pen-strep (Thermo Fisher Scientific) in 5% CO 2 at 37 °C. Cells were incubated in media containing charcoal-stripped FBS for 16 h, and then treated with either vehicle (ethanol), dexamethasone 10 −6  M, vehicle + RU486 10 −6  M, or dexamethasone 10 −6  M + RU486 10 −6  M for 48 h. Cells were incubated in serum free media to induce ciliogenesis for an additional 48 h with either vehicle, dexamethasone, vehicle + RU486 or dexamethasone + RU486. Fetal kidneys fixed in 4% paraformaldehyde were embedded in paraffin and cut at either 4 µm for periodic acid-Schiff (PAS) staining or 10 µm sections for immunofluorescence staining. PAS staining was performed on a Leica ST5010 Autostainer and CV5030 coverslipper and scanned with an Aperio Scanscope AT turbo. Immunofluorescence was performed following a standard protocol (Seow et al, 2019 ). Antibodies and stains used are listed in Reagents and tools table. Sections were imaged using a Zeiss LSM 980 confocal microscope with a ×63 objective. In total, 25–30 z-slices with an interval of 0.5 µm were imaged to ensure the entire primary cilia were captured. Four images (~100 cilia) per animal were captured, for n  = 3–4 animals per experimental group. IMCD3 cells grown on coverslips were fixed with 4% paraformaldehyde for 10 min at room temperature and washed three times with DPBS for 3 min each. Fixed cells were permeabilised with 0.1% triton X-100 (Sigma-Aldrich) diluted in DPBS for 10 min and washed three times with DPBS. Permeabilised cells were incubated with block buffer (5% donkey serum in PBST) at room temperature for 30 min. Cells were incubated with primary antibodies (Reagents and tools table) at room temperature for 1 h, then washed three times with DPBS for 5 min each and incubated with secondary antibodies (Reagents and tools table) at room temperature in the dark for 1 h. Cells were washed with DPBS for 5 min each and mounted with ProLong gold antifade moutant (Thermo Fisher Scientific). Cells were imaged with a z-stack using a Leica SP8 confocal microscope with a 63x objective. Overall, 10–15 z-slices with an interval of 0.5 µm were imaged to ensure the entire cilium was captured. Four images (~100 cilia) per biological replicates were taken, where n  = three biological replicates per treatment. Confocal immunofluorescence z-stack images were used for primary cilia analysis. Primary cilia were measured manually with the polygon measurement tool in Imaris imaging software (version 9.8.1). Primary cilia per nucleus ratio were counted with FIJI. Protein was extracted from embryonic kidneys and IMCD3 cells using Radioimmunoprecipitation (RIPA) buffer (150 mM NaCl, 50 mM Tris-HCL pH 8.0, 1% IGEPAL, 0.5% sodium deoxycholate, 0.1% SDS, protease inhibitor and phosphatase inhibitor) and protein lysates (10–20 μg) analysed by western blot as previously described (Short et al, 2020 ). All antibodies are listed in the reagents and tools table. Membranes were also probed with a β-actin antibody (1:50,000; Sigma-Aldrich A5316) to control protein loading. Blots were incubated with ECL and imaged using X-ray films. The films were scanned and analysed using FIJI imaging software. Whole kidneys were fixed in 2.5% (v/v) glutaraldehyde and 2% (v/v) paraformaldehyde in 0.1 M sodium cacodylate buffer overnight at 4 °C. Kidneys were washed with three times with 0.1 M sodium cacodylate buffer for 30 min each with rotation at room temperature. Samples were post fixed in 1% (v/v) osmium tetroxide in 0.1 M sodium cacodylate buffer at room temperature for 2 h and washed three times with milliQ water for 30 min each. Following washes, kidneys were bisected into longitudinal halves with scalpel blade. Fixed kidneys were dehydrated in increasing concentrations of ethanol: once in 30%, 50%, 70%, 90% and twice in 100% ethanol for 20 min each. Dehydrated kidneys were dried in an EM CPD300 Critical Point Dryer (Lecia Microsystems) and mounted on 12 mm diameter aluminium scanning electron microscopy stubs using stick carbon tabs. Mounted samples were gold coated with EM Ace600 sputter coater (Lecia Microsystems) and kidneys were imaged with Nova NanoSEM 450 scanning electron microscope (Thermo Fisher Scientific) at a voltage of 2 kV and a spot size of 2. All statistical analysis was performed using GraphPad Prism statistical analysis software, with statistical significance set at P  < 0.05 and all error bars as standard error of the mean (SEM). Two groups were compared using two-tailed unpaired t test with unequal variance, and multiple groups were compared by a one-way ANOVA with a Tukey’s post hoc test.

While loss of GR is linked to hypertension and changes in solute transporter gene expression, the cellular expression profile and role of GR during kidney development remains unclear. The expression of GR was investigated using single-cell RNA sequencing datasets from the developing mouse kidney at embryonic day (E)13.5, E15.5, and E18.5 (Combes et al, 2019 ). Expression of the GR gene (official symbol Nr3c1) was observed within early nephron, ureteric epithelium, stromal and immune cell clusters at E13.5, with highest expression within stromal clusters 3 and 4 which are marked by collagen type III alpha 1 ( Cola1 ), Decorin ( Dcn ), Delta like non-canonical Notch ligand 1 ( Dlk1 ) and Periostin osteoblast specific factor ( Postn ) (Figs.  1A–C and  EV1A ). GR expression was detected in the proximal tubule, distal tubule/loop of Henle, ureteric epithelium/collecting duct, immune cells and vasculature at E15.5 and E18.5 (Figs.  1D–G and EV1B,C ). As such, GR signalling has the potential to impact most stromal and epithelial cell types in the developing kidney. Figure 1 Glucocorticoid receptor gene expression in the fetal kidney during embryonic development. ( A ) Average expression of GR across the cell clusters at E13.5, E15.5 and E18.5. ( B ) tSNE plot showing cell clusters within the kidney at E13.5. ( C ) tSNE plot showing the expression of GR, purple within the cell clusters at E13.5. ( D ) tSNE plot showing cell clusters within the kidney at E15.5. ( E ) tSNE plot showing the expression of GR, purple within the cell clusters at E15.5. ( F ) tSNE plot showing cell clusters within the kidney at E18.5. ( G ) tSNE plot showing the expression of GR, purple within the cell clusters at E18.5. ( H ) Immunofluorescence of glucocorticoid receptor (GR) in the fetal kidney during embryonic development at E18.5. Sections were stained Hoechst (blue, nucleus), Dolichos Biflorus Agglutinin (DBA) (green, collecting duct), Lotus Tetragonolobus Lectin (LTL) (green, proximal tubule), nephrin (NPHS1) (green, podocyte), MEIS123 (green, stroma) and GR (red, GR). Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 μm. All images are representative of n  = 4 animals per experimental group.  Source data are available online for this figure . ( A ) Average expression of GR across the cell clusters at E13.5, E15.5 and E18.5. ( B ) tSNE plot showing cell clusters within the kidney at E13.5. ( C ) tSNE plot showing the expression of GR, purple within the cell clusters at E13.5. ( D ) tSNE plot showing cell clusters within the kidney at E15.5. ( E ) tSNE plot showing the expression of GR, purple within the cell clusters at E15.5. ( F ) tSNE plot showing cell clusters within the kidney at E18.5. ( G ) tSNE plot showing the expression of GR, purple within the cell clusters at E18.5. ( H ) Immunofluorescence of glucocorticoid receptor (GR) in the fetal kidney during embryonic development at E18.5. Sections were stained Hoechst (blue, nucleus), Dolichos Biflorus Agglutinin (DBA) (green, collecting duct), Lotus Tetragonolobus Lectin (LTL) (green, proximal tubule), nephrin (NPHS1) (green, podocyte), MEIS123 (green, stroma) and GR (red, GR). Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 μm. All images are representative of n  = 4 animals per experimental group.  Source data are available online for this figure . To compare GR localisation at the protein level at different stages of fetal kidney development, we performed immunofluorescence using a GR antibody with markers of the proximal tubules (lotus tetragonolobus lectin, LTL), collecting duct (dolichos biflorus agglutinin, DBA), podocytes (nephrin, NPHS1) and stroma (MEIS123). At E14.5, GR was strongly localised to the developing collecting duct, limited in the proximal tubule (Davidson, 2009 ; Short and Smyth, 2016 ) compartments consistent with the single cell data from E13.5 to E15.5 (Fig.  1A ; Appendix Fig. S 1 ). At E16.5 and E18.5 GR is widely localised in the kidney with positive staining in the proximal tubules, collecting ducts, podocytes and stroma (Fig.  1H ; Appendix Figs. S 1 and S 2 ), suggesting GR signalling having important roles in these structures during kidney development. The role of GR-mediated signalling in the developing fetal kidney was analysed in GR-null mice at E18.5. Loss of GR was confirmed by immunohistochemistry and western blot analysis (Cole et al, 1995 ) (Fig.  EV2 ). Histological analysis by periodic acid Schiff staining of control and GR-null littermate kidney showed no major abnormalities in renal size or structures at E18.5 (Appendix Fig. S 3 ). As the GR is a steroid ligand-activated transcriptional regulator we then analysed the effect of loss of GR expression on the fetal kidney transcriptome at E18.5 by RNA sequencing. Total RNA was extracted from control ( n  = 4) and GR-null ( n  = 3) kidneys and analysed by NGS RNA-seq. Global changes in mRNA levels were displayed as a heatmap for genes with a fold change greater than 1 and a false discovery rate (FDR) of less than 0.05 (Fig.  2A ) and principal component plot showing separation of control and GR-null kidney at E18.5 (Fig.  2B ). Loss of GR expression resulted in 2473 differentially expressed genes (FDR 1 & FDR < 0.05 (Fig.  2C ; Table EV 1 ), which identified 16 upregulated and 25 downregulated ciliary genes (FDR < 0.05) (Fig.  2D ). To confirm gene expression changes detected by RNA sequencing, qRT-PCR was performed to quantify mRNA levels of twenty selected differently expressed protein coding genes. There was a significant decrease in GR-null fetal kidney mRNA levels at E18.5 in downregulated genes compared to controls (Fig.  2E–G ). scRNA-seq data was used to identify the localisation of the fifteen selected differentially expressed genes within the E18.5 embryonic kidney. Igf2 was expressed across the kidney while other genes had a more restricted expression (Fig.  2H ). Gene set enrichment analysis identified increased signatures of AKT (Nishimura et al, 2021 ), ERK/MAPK (Kuonen et al, 2019 ) and mTOR (Lai and Jiang, 2020 ) pathways which have been linked to regulate or be regulated by primary ciliogenesis and function (Fig.  EV3A ). Figure 2 Transcriptome analysis of E18.5 fetal kidney RNA from GR-null and control mice. ( A ) NGS RNA-seq was performed on total RNA isolated from control and GR-null mouse kidneys at E18.5. Heatmaps generated from differentially expressed genes (LogFC >1 and FDR < 0.05) log2-CPM values. The number of suppressed genes is four times more than augmented genes in GR-null compared to control. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( B ) Principal component plot showing separation of control (red) and GR-null (blue) kidney at E18.5. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( C ) Volcano plot of top differentially expressed genes with a significant FDR 1.5 and < −2. Red dots represent gene mRNA levels that are significantly increased, and blue dots represent genes with significantly decreased. Dashed lines show the LogFC 0.5 and FDR 0.05 cut-offs. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( D ) Volcano plot of differentially expressed ciliary genes with a significant FDR < 0.05. Red dots represent gene mRNA levels that are significantly increased, and blue dots represent genes with significantly decreased. Dashed lines show the LogFC 0.5 and FDR 0.05 cut-offs. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( E ) mRNA levels of eleven target genes identified from NGS RNA-seq in fetal GR-null mouse kidney at E18.5. Crisp1 (fold −9.6, P  = 0.00030), Gbp7 (fold −3.7, P  = 0.012), Ifi203 (fold −5.6, P  = 0.00073), Ifit1 (fold −8.6, P  = 0.024), Kap (fold −9.2, P  = 0.0024), Mndal (fold −3.8, P  = 0.00046), Oas2 (fold −2.6, P  = 0.016), Oasl2 (fold −4.6, P  = 0.0062) and S100g (fold −2.7, P  = 0.00094) were downregulated. Consistent with RNA sequencing Hba-A1 (fold 3.9, P  = 0.049) and Igf2 (fold 1.7, P  = 0.071) were upregulated. The mRNA levels are expressed relative to mRNA levels of the housekeeping gene Rps29 . All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null samples, n  = 3 animals per experimental group for each gene. ( F ) mRNA levels of six primary cilia structural genes in fetal GR-null mouse kidney at E18.5. Ccp110 (fold −2.2, P  = 0.011), Cep97 (fold −1.8, P  = 0.048), Cep290 (fold −2.9, P  = 0.012), Kif3a (fold −1.8, P  = 0.026), Rab8a (fold 1.1, P  = −0.80) and Rpgr (fold −1.9, P  = 0.026). The mRNA levels are expressed relative to mRNA levels of the housekeeping gene Rps29 . All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control ( n  = 3–4) and GR-null ( n  = 3) animals for each gene. ( G ) mRNA level of three primary cilia regulatory genes in fetal GR-null mouse kidney at E18.5. Atat1 (fold −1.19, P  = 0.57), Aurka (fold 1.32, P  = 0.19) and Nedd9 (fold 1.16, P  = 0.79). The mRNA levels are expressed relative to mRNA levels of the housekeeping gene Rps29 . All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null, n  = 3 animals for each gene. ( H ) Expression of target genes within different cell types of the kidney at E18.5. Scale represents average expression, generated from single cell dataset.  Source data are available online for this figure . ( A ) NGS RNA-seq was performed on total RNA isolated from control and GR-null mouse kidneys at E18.5. Heatmaps generated from differentially expressed genes (LogFC >1 and FDR < 0.05) log2-CPM values. The number of suppressed genes is four times more than augmented genes in GR-null compared to control. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( B ) Principal component plot showing separation of control (red) and GR-null (blue) kidney at E18.5. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( C ) Volcano plot of top differentially expressed genes with a significant FDR 1.5 and < −2. Red dots represent gene mRNA levels that are significantly increased, and blue dots represent genes with significantly decreased. Dashed lines show the LogFC 0.5 and FDR 0.05 cut-offs. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( D ) Volcano plot of differentially expressed ciliary genes with a significant FDR < 0.05. Red dots represent gene mRNA levels that are significantly increased, and blue dots represent genes with significantly decreased. Dashed lines show the LogFC 0.5 and FDR 0.05 cut-offs. Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( E ) mRNA levels of eleven target genes identified from NGS RNA-seq in fetal GR-null mouse kidney at E18.5. Crisp1 (fold −9.6, P  = 0.00030), Gbp7 (fold −3.7, P  = 0.012), Ifi203 (fold −5.6, P  = 0.00073), Ifit1 (fold −8.6, P  = 0.024), Kap (fold −9.2, P  = 0.0024), Mndal (fold −3.8, P  = 0.00046), Oas2 (fold −2.6, P  = 0.016), Oasl2 (fold −4.6, P  = 0.0062) and S100g (fold −2.7, P  = 0.00094) were downregulated. Consistent with RNA sequencing Hba-A1 (fold 3.9, P  = 0.049) and Igf2 (fold 1.7, P  = 0.071) were upregulated. The mRNA levels are expressed relative to mRNA levels of the housekeeping gene Rps29 . All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null samples, n  = 3 animals per experimental group for each gene. ( F ) mRNA levels of six primary cilia structural genes in fetal GR-null mouse kidney at E18.5. Ccp110 (fold −2.2, P  = 0.011), Cep97 (fold −1.8, P  = 0.048), Cep290 (fold −2.9, P  = 0.012), Kif3a (fold −1.8, P  = 0.026), Rab8a (fold 1.1, P  = −0.80) and Rpgr (fold −1.9, P  = 0.026). The mRNA levels are expressed relative to mRNA levels of the housekeeping gene Rps29 . All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control ( n  = 3–4) and GR-null ( n  = 3) animals for each gene. ( G ) mRNA level of three primary cilia regulatory genes in fetal GR-null mouse kidney at E18.5. Atat1 (fold −1.19, P  = 0.57), Aurka (fold 1.32, P  = 0.19) and Nedd9 (fold 1.16, P  = 0.79). The mRNA levels are expressed relative to mRNA levels of the housekeeping gene Rps29 . All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null, n  = 3 animals for each gene. ( H ) Expression of target genes within different cell types of the kidney at E18.5. Scale represents average expression, generated from single cell dataset.  Source data are available online for this figure . We next employed a deconvolution approach, where scRNA-seq data was used to estimate changes in cell type signatures or cell proportions within the bulk RNA-seq dataset (Fig.  EV3B,C ). Small but s ignificant reductions in the estimated proportions of early proximal tubule and medullary stromal populations (which express higher levels of GR ), were identified in GR -null kidneys, suggesting subtle changes in the development of GR -expressing cell types (Fig.  EV3B,C ). Transcriptome sequencing of whole GR-null fetal kidney at E18.5 showed a reduction in a number of key ciliogenesis genes, including Ccp110 , Cep97 , Cep290, Kif3a and Rpgr which indicated a potential deficit in primary ciliogenesis. qRT-PCR analysis confirmed downregulation of these cilia-associated genes in GR-null mouse kidney with significant reductions in mRNA levels of Ccp110 (fold −2.17, P  = 0.011), Cep97 (fold −1.79, P  = 0.047), Cep290 (−2.9-fold, P  = 0.012), Kif3a (fold −1.82, P  = 0.026) and Rpgr (fold −1.87, P  = 0.025) (Fig.  2F ), but no changes Rab8a (fold 0.921, P  = 0.805) or in ciliary genes known to regulate primary cilia; Atat1 (fold −1.19, P  = 0.573), Aurka (fold 0.778, P  = 0.148) and Nedd9 (fold 0.879, P  = 0.802) (Fig.  2G ). To investigate further, the primary cilium was visualised in the proximal tubule of the GR-null fetal kidney by immunofluorescence (Fig.  3A ). Acetylated tubulin and ARL13B were stained to confirm co-localisation of the cilia axoneme in GR-null fetal kidney (Fig.  EV4A ). Kidneys were stained with acetylated tubulin to mark microtubules of the cilia axoneme, LTL to mark proximal tubules and pericentrin to mark the basal body. Compared to control mice, primary cilia located on kidney proximal tubule cells were stunted and abnormal in the GR-null mouse (Fig.  3A ). Primary cilia length on GR-null kidney proximal tubule cells (5.10 ± 0.11 µm) was significantly decreased compared to control mice (6.20 ± 0.15 µm) (Fig.  3B ). Additionally, the percentage of primary cilia longer than 5 µm was significantly lower on GR-null proximal tubule cells (38.47 ± 4.21%) when compared to controls (56.38 ± 4.45%) (Fig.  3C ). There was no significant difference in the percentage of ciliated proximal tubule cells between GR-null mice and controls (Fig.  3D ). The presence of abnormally shaped cilia in the renal tubule of GR-null mice, compared to controls, was further confirmed with scanning electron microscopy images that showed abnormal primary cilia morphology with many cilia displaying bulging areas and an abnormal shape (Fig.  3E ). Primary cilia structure in other renal tubule segments such as the collecting ducts could not be assessed at E18.5 because the intralumenal space was too constricted at this stage of development (Fig.  EV4B ). Finally, in contrast to the fetal lung of GR-null mice, immunohistochemistry staining for Ki67, a marker of cell proliferation, showed no change in cell proliferation in the fetal kidney at E18.5, both in the mesenchymal compartment and in the proximal tubule (Bird et al, 2014 ; Bird et al, 2007 ) (Fig.  EV5A–C ). Figure 3 Glucocorticoid regulation of primary cilia length on GR-null fetal kidney proximal tubule cells at E18.5. ( A ) Immunofluorescence of primary cilia morphology in control and GR-null proximal tubules at E18.5. Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules), pericentrin (PCNT) (red, basal body) and Lotus Tetragonolobus Lectin (LTL) (grey, proximal tubule). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( B ) Proximal tubule primary cilia length was measured using Imaris software. Four images were taken per an animal, 574 primary cilia were measured in the control and 712 primary cilia were measured in the GR-null mouse kidneys. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.0001), n  = 4 animals per experimental group. ( C ) Percentage of proximal tubule primary cilia greater than 5 µm in control and GR-null proximal tubules. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.027), n  = 4 animals per experimental group. ( D ) Percentage of ciliated proximal tubule cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.11), n  = 4 animals per experimental group. ( E ) Scanning electron microscopy of primary cilia morphology in control and GR-null kidneys at E18.5. Kidneys were imaged with Nova NanoSEM 450 scanning electron microscope (Thermo Fisher Scientific) at a voltage of 2 kV and a spot size of 2. Black arrows indicate primary cilia. Scale bar represents 4 µm (control) and 5 µm (GR-null). All images are representative of n  = 5 animals per experimental group.  Source data are available online for this figure . ( A ) Immunofluorescence of primary cilia morphology in control and GR-null proximal tubules at E18.5. Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules), pericentrin (PCNT) (red, basal body) and Lotus Tetragonolobus Lectin (LTL) (grey, proximal tubule). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( B ) Proximal tubule primary cilia length was measured using Imaris software. Four images were taken per an animal, 574 primary cilia were measured in the control and 712 primary cilia were measured in the GR-null mouse kidneys. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.0001), n  = 4 animals per experimental group. ( C ) Percentage of proximal tubule primary cilia greater than 5 µm in control and GR-null proximal tubules. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.027), n  = 4 animals per experimental group. ( D ) Percentage of ciliated proximal tubule cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.11), n  = 4 animals per experimental group. ( E ) Scanning electron microscopy of primary cilia morphology in control and GR-null kidneys at E18.5. Kidneys were imaged with Nova NanoSEM 450 scanning electron microscope (Thermo Fisher Scientific) at a voltage of 2 kV and a spot size of 2. Black arrows indicate primary cilia. Scale bar represents 4 µm (control) and 5 µm (GR-null). All images are representative of n  = 5 animals per experimental group.  Source data are available online for this figure . The global GR-null mouse dies at birth due to respiratory failure, therefore further analysis of a cilia phenotype postnatally in the renal tubule was not possible. To allow further analysis postnatally, conditional deletion of the GR in collecting ducts of the kidney was achieved using a HoxB7-Cre recombinase transgenic mouse crossed to a GR-loxP/loxP mouse strain. The effect of GR loss on collecting duct primary cilia morphology was analysed using a conditional collecting duct GR deletion at postnatal day 11 (Fig.  4A ). Postnatal day 11 was selected for primary analysis because mouse models of PKD showed prominent cysts by 2 weeks in age (Rachel et al, 2015 ). Kidneys were stained with acetylated tubulin to mark microtubules, DBA to stain collecting ducts, and pericentrin to visualise the basal body of cilia. Primary cilia length on GRcdKO kidney collecting ducts cells were significantly decreased (3.30 ± 0.06 μm) when compared to control (3.63 ± 0.05 μm) (Fig.  4B ). There was no significant difference in the percentage of primary cilia longer than 5 μm in GRcdKO kidney collecting ducts (18.09 ± 1.24%) when compared to controls (22.21 ± 8.48%) (Fig.  4C ). In addition, there was no significant difference in the percentage of ciliated collecting duct cells between GRcdKO kidney collecting ducts (66.38 ± 4.56%) and control (65.88 ± 3.58%) (Fig.  4D ). Figure 4 Glucocorticoid regulation of primary cilia length on GRcdKO postnatal kidney collecting duct cells at P11. ( A ) Immunofluorescence of primary cilia morphology in control and GRcdKO collecting ducts at P11. Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules), pericentrin (PCNT) (red, basal body) and Dolichos Biflorus Agglutinin (DBA) (grey, collecting duct). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( B ) Collecting duct primary cilia length was measured using Imaris software. Four images were taken per an animal, 1048 primary cilia were measured in the control and 957 primary cilia were measured in the GRcdKO mouse kidneys. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control vs GRcdKO ( P  = 0.0001), n  = 3 animals per experimental group. ( C ) Percentage of collecting duct primary cilia greater than 5 µm in the control and GRcdKO mouse kidney at P11. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GRcdKO ( P  = 0.66), n  = 3 animals per experimental group. ( D ) Percentage of ciliated collecting duct cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GRcdKO ( P  = 0.94), n  = 3 animals per experimental group.  Source data are available online for this figure . ( A ) Immunofluorescence of primary cilia morphology in control and GRcdKO collecting ducts at P11. Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules), pericentrin (PCNT) (red, basal body) and Dolichos Biflorus Agglutinin (DBA) (grey, collecting duct). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( B ) Collecting duct primary cilia length was measured using Imaris software. Four images were taken per an animal, 1048 primary cilia were measured in the control and 957 primary cilia were measured in the GRcdKO mouse kidneys. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control vs GRcdKO ( P  = 0.0001), n  = 3 animals per experimental group. ( C ) Percentage of collecting duct primary cilia greater than 5 µm in the control and GRcdKO mouse kidney at P11. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GRcdKO ( P  = 0.66), n  = 3 animals per experimental group. ( D ) Percentage of ciliated collecting duct cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GRcdKO ( P  = 0.94), n  = 3 animals per experimental group.  Source data are available online for this figure . Immunofluorescence was used to assess the morphology of primary cilia on GR-null kidney podocytes. Kidneys were stained with acetylated tubulin and nephrin which stains the podocytes. Primary cilia on GR-null kidney podocytes also appeared stunted when compared to controls (Fig.  5A ). Primary cilia length on GR-null kidney podocytes was significantly decreased (2.43 ± 0.09 μm) when compared to controls (2.73 ± 0.10 μm) (Fig.  5B ). There was no significant difference in the percentage of primary cilia longer than 5 μm in GR-null kidney podocytes (7.25 ± 2.04%) compared to controls (12.2 ± 3.48%) (Fig.  5C ) and no significant difference in the percentage of ciliated podocyte cells between GR-null kidney podocytes (25.64 ± 3.48%) and controls (24.15 ± 2.06%) (Fig.  5D ). Figure 5 Glucocorticoid regulation of primary cilia length on GR-null fetal kidney podocyte cells at E18.5 and induced pluripotent stem cell kidney organoid podocyte cells after dexamethasone treatment. ( A ) Immunofluorescence of primary cilia morphology in control and GR-null podocytes at E18.5. Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules) and Nephrin (NPHS1) (grey, podocyte). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( B ) GR-null podocyte primary cilia length was measured using Imaris software. Four images were taken per animal, 235 primary cilia were measured in control and 212 primary cilia were measured in GR-null. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.030). Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( C ) Percentage of primary cilia greater than 5 μm in control and GR-null podocyte cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.32). Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( D ) Percentage of ciliated podocyte cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.71). Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( E ) Immunofluorescence of primary cilia morphology in induced pluripotent stem cell (iPSC) kidney organoids. iPSC kidney organoids were treated with vehicle (veh) or dexamethasone (dex). Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules), nephrin (NPHS1) (red, podocyte) and Lotus Tetragonolobus Lectin (LTL) (grey, proximal tubule). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( F ) Podocyte primary cilia length was measured using Imaris software. Four images were taken per animal, 1685 primary cilia were measured in vehicle (veh) treated organoids and 1638 primary cilia were measured in dexamethasone (dex) treated organoids. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between veh and dex treated cells ( P  = 0.0001), n  = 3 biological replicates per experimental group. ( G ) Percentage of primary cilia greater than 4 μm in veh and dex treated organoids. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between veh and dex treated cells ( P  = 0.0093), n  = 3 biological replicates per experimental group. ( H ) Percentage of ciliated podocyte cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between veh and dex treated cells ( P  = 0.99), n  = 3 biological replicates per experimental group.  Source data are available online for this figure . ( A ) Immunofluorescence of primary cilia morphology in control and GR-null podocytes at E18.5. Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules) and Nephrin (NPHS1) (grey, podocyte). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( B ) GR-null podocyte primary cilia length was measured using Imaris software. Four images were taken per animal, 235 primary cilia were measured in control and 212 primary cilia were measured in GR-null. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.030). Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( C ) Percentage of primary cilia greater than 5 μm in control and GR-null podocyte cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.32). Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( D ) Percentage of ciliated podocyte cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null ( P  = 0.71). Data from control ( n  = 4) and GR-null ( n  = 3) animals per experimental group. ( E ) Immunofluorescence of primary cilia morphology in induced pluripotent stem cell (iPSC) kidney organoids. iPSC kidney organoids were treated with vehicle (veh) or dexamethasone (dex). Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules), nephrin (NPHS1) (red, podocyte) and Lotus Tetragonolobus Lectin (LTL) (grey, proximal tubule). White arrows indicate primary cilia. Slides were imaged with a Zeiss LSM 980 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images are representative of n  = 4 animals per experimental group. ( F ) Podocyte primary cilia length was measured using Imaris software. Four images were taken per animal, 1685 primary cilia were measured in vehicle (veh) treated organoids and 1638 primary cilia were measured in dexamethasone (dex) treated organoids. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between veh and dex treated cells ( P  = 0.0001), n  = 3 biological replicates per experimental group. ( G ) Percentage of primary cilia greater than 4 μm in veh and dex treated organoids. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between veh and dex treated cells ( P  = 0.0093), n  = 3 biological replicates per experimental group. ( H ) Percentage of ciliated podocyte cells. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between veh and dex treated cells ( P  = 0.99), n  = 3 biological replicates per experimental group.  Source data are available online for this figure . To further explore glucocorticoid regulation of primary ciliogenesis in a human cell-based model, induced pluripotent stem cell-derived kidney organoids were treated for 48 h with the synthetic glucocorticoid dexamethasone. Primary cilia were visualised by immunofluorescence as above (Fig.  5E ). After dexamethasone treatment primary cilia were significantly longer (1.77 ± 0.02 μm) compared to controls (1.59 ± 0.02 μm) (Fig.  5F ). Additionally, the percentage of primary cilia longer than 4 µm was significantly greater in podocytes for dexamethasone-treated organoids (2.80 ± 0.32%) compared to controls (0.9 ± 0.24%) (Fig.  5G ). There were no significant differences in the percentage of ciliated cells in human kidney organoids treated with dexamethasone (76.62 ± 4.99%) versus controls (76.51 ± 2.75%) (Fig.  5H ). The effect of dexamethasone was next investigated with cultured mouse IMCD3 cells. Prior to steroid treatment IMCD3 cells were grown in charcoal-stripped media to reduce endogenous steroid hormones followed by serum starved media to induce ciliogenesis (Fig.  EV5D ). Primary cilia morphology was visualised with immunofluorescence as above after 96 h of vehicle or dexamethasone treatment (Fig.  6A ). Primary cilia length in IMCD3 cells treated with dexamethasone (2.89 ± 0.04 µm) was significantly longer than vehicle (2.46 ± 0.03 µm) (Fig.  6B ). Furthermore, the percentage of primary cilia longer than 4 µm was significantly higher in dexamethasone treated cells (54.56 ± 1.80%) versus controls (43.85 ± 1.30%) (Fig.  6C ), with no significant difference in the percentage of ciliated collecting duct cells in cells following either treatment (Fig.  6D ). To determine whether primary cilia elongation was in response to dexamethasone acting via the GR, IMCD3 cells were treated with 10 −6 M RU486, a high affinity GR antagonist. Primary cilia morphology in IMCD3 cells after RU486 and then 96 h of dexamethasone treatment was visualised as above. Primary cilia length on IMCD3 cells treated with dexamethasone (2.89 ± 0.04 µm) was significantly longer than vehicle (2.46 ± 0.03 µm) and dexamethasone + RU486 (2.00 ± 0.02 µm), indicating that RU486 blocked the effect of dexamethasone acting via the GR (Fig.  6B ). In addition, the percentage of primary cilia longer than 4 µm was significantly higher in cells treated with dexamethasone (21.78 ± 1.68%), compared to all other treatments (vehicle: 11.17 ± 1.70%, vehicle + RU486: 5.50 ± 0.10% and dexamethasone + RU486: 6.12 ± 0.88%) (Fig.  6C ). There were no significant differences in the percentage of ciliated collecting duct cells, except between dexamethasone (65.77 ± 2.87%) and dexamethasone + RU486 (77.43 ± 2.29%) (Fig.  6D ). Figure 6 Glucocorticoid regulation of primary cilia in IMCD3 cells after dexamethasone treatment. ( A ) Immunofluorescence images of primary cilia morphology in IMCD3 cells. IMCD3 cells were treated with vehicle (veh), dexamethasone (dex), vehicle + RU486 (veh + RU486) or dexamethasone + RU486 (dex + RU486). Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules) and gamma tubulin (γ-Tub) (red, basal body). White arrows indicate primary cilia. Slides were imaged with a Lecia SP8 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images represent n  = 3 biological replicates per experimental group. ( B ) Primary cilia length was measured using Imaris software. Four images were taken per biological replicate, the number of individual cilia measured per a treatment group was 1698 vehicle, 1514 dexamethasone, 1966 vehicle + RU486 and 2143 dexamethasone + RU486. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by one-way ANOVA with multiple comparisons indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant. Veh vs dex ( P  = 0.0001), veh vs veh + RU486 ( P  = 0.0001), veh vs dex + RU486 ( P  = 0.0001), dex vs veh + RU486 ( P  = 0.0001), dex vs dex + RU486 ( P  = 0.0001), veh + RU486 vs dex + RU486 ( P  = 0.28). Data from n  = 3 biological replicates per experimental group. ( C ) Percentage of primary cilia greater than 4 µm. All data presented as mean ± SEM, significant differences were analysed by one-way ANOVA with multiple comparisons indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant. Veh vs dex ( P  = 0.0026), veh vs veh + RU486 ( P  = 0.073), veh + dex + RU486 ( P  = 0.11), dex vs veh RU486 ( P  = 0.0001), dex vs dex + RU486 ( P  = 0.0002), veh + RU486 vs dex + RU486 ( P  = 0.99). Data from n  = 3 biological replicates per experimental group. ( D ) Percentage of ciliated collecting duct cells. All data presented as mean ± SEM, significant differences were analysed by one-way ANOVA with multiple comparisons indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant. Veh vs dex ( P  = 0.59), veh vs veh + RU486 ( P  = 0.29), veh vs dex + RU486 ( P  = 0.046), dex vs veh + RU486 ( P  = 0.049), dex vs dex + RU486 ( P  = 0.0081), veh + RU486 vs dex + RU486 ( P  = 0.57). Data from n  = 3 biological replicates per experimental group.  Source data are available online for this figure . ( A ) Immunofluorescence images of primary cilia morphology in IMCD3 cells. IMCD3 cells were treated with vehicle (veh), dexamethasone (dex), vehicle + RU486 (veh + RU486) or dexamethasone + RU486 (dex + RU486). Sections were stained with Hoechst (blue, nucleus), acetylated tubulin (AceTub) (green, microtubules) and gamma tubulin (γ-Tub) (red, basal body). White arrows indicate primary cilia. Slides were imaged with a Lecia SP8 confocal microscope (×63 objective, 2× digital zoom), scale bar represents 20 µm. All images represent n  = 3 biological replicates per experimental group. ( B ) Primary cilia length was measured using Imaris software. Four images were taken per biological replicate, the number of individual cilia measured per a treatment group was 1698 vehicle, 1514 dexamethasone, 1966 vehicle + RU486 and 2143 dexamethasone + RU486. Lines represent median and quartiles. All data presented as mean ± SEM, significant differences were analysed by one-way ANOVA with multiple comparisons indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant. Veh vs dex ( P  = 0.0001), veh vs veh + RU486 ( P  = 0.0001), veh vs dex + RU486 ( P  = 0.0001), dex vs veh + RU486 ( P  = 0.0001), dex vs dex + RU486 ( P  = 0.0001), veh + RU486 vs dex + RU486 ( P  = 0.28). Data from n  = 3 biological replicates per experimental group. ( C ) Percentage of primary cilia greater than 4 µm. All data presented as mean ± SEM, significant differences were analysed by one-way ANOVA with multiple comparisons indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant. Veh vs dex ( P  = 0.0026), veh vs veh + RU486 ( P  = 0.073), veh + dex + RU486 ( P  = 0.11), dex vs veh RU486 ( P  = 0.0001), dex vs dex + RU486 ( P  = 0.0002), veh + RU486 vs dex + RU486 ( P  = 0.99). Data from n  = 3 biological replicates per experimental group. ( D ) Percentage of ciliated collecting duct cells. All data presented as mean ± SEM, significant differences were analysed by one-way ANOVA with multiple comparisons indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant. Veh vs dex ( P  = 0.59), veh vs veh + RU486 ( P  = 0.29), veh vs dex + RU486 ( P  = 0.046), dex vs veh + RU486 ( P  = 0.049), dex vs dex + RU486 ( P  = 0.0081), veh + RU486 vs dex + RU486 ( P  = 0.57). Data from n  = 3 biological replicates per experimental group.  Source data are available online for this figure . Gene set enrichment analysis identified several signaling pathway genes that were differentially expressed in GR-null fetal kidney when compared to control (Fig.  EV3A ). It has been demonstrated that several WNT signaling proteins are upregulated in the GR-null lung (Bridges et al, 2020 ), suggesting that primary ciliogenesis which normally occurs during the G 0 quiescent phase of the cell cycle, could potentially be controlled by developmental and proliferation pathways. To explore the mechanism of glucocorticoid regulation of primary cilia, western blot analysis was used to quantify several signaling pathways and ciliary proteins (Fig.  7A–L ). There was no significant difference in protein levels and phosphorylation of AKT (Fig.  7A ), AMPKα (Fig.  7B ), β-catenin (Fig.  7C ), JNK (Fig.  7E ) and S6 kinase (Fig.  7F ) signalling proteins between control and GR-null mice. P-ERK and total ERK protein levels were significantly decreased in GR-null mice (0.53 ± 0.04) compared to controls (0.81 ± 0.09) (Fig.  7D ). Suppressor of Fused (SUFU) protein levels were significantly reduced in GR-null mice (0.81 ± 0.06) compared to controls (0.95 ± 0.02) (Fig.  7G ). There were no significant differences in the levels of ciliary structural proteins acetylated tubulin (Fig.  7I ), CEP290 (Fig.  7J ), IFT88 (Fig.  7K ) and KIF3A (Fig.  7L ) between control and GR-null mice. Aurora kinase A (AURKA) has been shown to promote cilia resorption (Pugacheva et al, 2007 ) and so we next examined whether AURKA may be regulated by the GR. IMCD3 cells were treated with dexamethasone and reduced levels of AURKA protein were observed (0.43 ± 0.03) compared to control cells (0.75 ± 0.09), thereby confirming AURKA expression is negatively regulated by GR activity (Fig.  7H ). Figure 7 Analysis of cell signalling pathways and cilia structural proteins by western blot analysis as targets for glucocorticoid regulation of primary ciliogenesis. Western blot analysis of signaling pathway proteins ( A ) P-AKT and AKT ( P  = 0.067), ( B ) P-AMPKα and AMPKα ( P  = 0.22), ( C ) P-β-catenin and β-catenin ( P  = 0.96), ( D ) P-ERK and ERK ( P  = 0.022), ( E ) JNK ( P  = 0.06), ( F ) P-S6 and S6 ( P  = 0.64), and ( G ) SUFU ( P  = 0.049) in control and GR-null fetal kidney at E18.5. Western blot analysis of ( H ) Aurora kinase A (AURKA) ( P  = 0.017) in IMCD3 cells treated with vehicle (veh) or dexamethasone (dex). Western blot analysis of cilia structural proteins ( I ) acetylated tubulin ( P  = 0.090), ( J ) CEP290 ( P  = 0.52), ( K ) IFT88 ( P  = 0.15), ( L ) KIF3A ( P  = 0.34) in control and GR-null fetal kidney at E18.5. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null, n  = 4 animals per experimental group or veh and dex, n  = 3 biological replicates per experimental group. Western blots from P-AKT and AKT ( A ) were re-probed with IFT88 ( K ) and acetylated tubulin ( I ), respectively, and share the same β-actin blot. Western blots from P-ERK and ERK ( D ) was re-probed with P-S6 and S6 ( F ), respectively, and share the same β-actin blot. KIF3A ( L ) was re-probed with CEP290 ( J ) and share the same β-actin blot. The same β-actin panels were included in each figure for ease of comparison.  Source data are available online for this figure . Western blot analysis of signaling pathway proteins ( A ) P-AKT and AKT ( P  = 0.067), ( B ) P-AMPKα and AMPKα ( P  = 0.22), ( C ) P-β-catenin and β-catenin ( P  = 0.96), ( D ) P-ERK and ERK ( P  = 0.022), ( E ) JNK ( P  = 0.06), ( F ) P-S6 and S6 ( P  = 0.64), and ( G ) SUFU ( P  = 0.049) in control and GR-null fetal kidney at E18.5. Western blot analysis of ( H ) Aurora kinase A (AURKA) ( P  = 0.017) in IMCD3 cells treated with vehicle (veh) or dexamethasone (dex). Western blot analysis of cilia structural proteins ( I ) acetylated tubulin ( P  = 0.090), ( J ) CEP290 ( P  = 0.52), ( K ) IFT88 ( P  = 0.15), ( L ) KIF3A ( P  = 0.34) in control and GR-null fetal kidney at E18.5. All data presented as mean ± SEM, significant differences were analysed by unpaired T tests indicated by * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001, ns=not significant, between control and GR-null, n  = 4 animals per experimental group or veh and dex, n  = 3 biological replicates per experimental group. Western blots from P-AKT and AKT ( A ) were re-probed with IFT88 ( K ) and acetylated tubulin ( I ), respectively, and share the same β-actin blot. Western blots from P-ERK and ERK ( D ) was re-probed with P-S6 and S6 ( F ), respectively, and share the same β-actin blot. KIF3A ( L ) was re-probed with CEP290 ( J ) and share the same β-actin blot. The same β-actin panels were included in each figure for ease of comparison.  Source data are available online for this figure .

This study has investigated GC/GR signalling in the mouse fetal kidney and demonstrates a key regulatory role for GC signalling in primary ciliogenesis. We first showed that the GR is expressed at E13.5 within nephron progenitor, collecting duct, stromal and immune cell populations, and expands into almost all cell types within the fetal mouse kidney where it has the capacity to respond to the rising levels of late gestational endogenous corticosterone from E15-E16 prior to birth (Fowden and Forhead, 2015 ). To further investigate the role of GC signalling in the developing kidney prior to birth we utilised mice deficient in the GR. Loss of GR in the fetal kidney had a substantial effect on the renal transcriptome, with 288 genes differently expressed more than 1-fold at E18.5. The majority of these genes were significantly decreased in expression including the kidney tubule markers Kap and S100g . Kap is one of the most abundant genes expressed in the kidney proximal tubule cells and is regulated by androgens however the role and function of Kap in proximal tubule cells is unknown (Teixido et al, 2006 ). One study by de Quixano et al showed that overexpression of Kap protected mice from metabolic syndrome induced by a high fat diet including associated hypertension (de Quixano et al, 2017 ). S100g , is a calcium-binding protein localised to the distal tubule at E18.5. Decreased S100g may indicate that GC signalling is involved in calcium transport within the kidney. However, S100g knockout mice are indistinguishable from control mice (Kutuzova et al, 2006 ; Lee et al, 2007 ). This may be a result of compensation from other calcium transporter proteins. Of the ciliogenesis-related genes, Cep290 was strongly decreased in the mouse kidney of GR-null fetal mice at E18.5. Although we detect similar levels of Cep290 protein in total fetal kidney extracts, further studies are required with isolated primary kidney epithelial cells to specifically assess Cep290 protein levels in ciliated renal tubular cells. Other structural cilia proteins, such as Ccp110 and Cep97 , also need to be assessed. The CEP290 protein is localised to the transition zone of primary cilia where it is essential for assembly of microtubules and for primary cilia formation. Cep290 genetic mutations are found in a number of human ciliopathies including Joubert syndrome, characterised by brain abnormalities and in some instance’s polycystic kidneys, as well as Meckel syndrome which is characterised by cystic dysplastic kidneys, brain malformations, hepatic fibrosis and proliferation of the bile ducts. This commonly leads to end-stage renal disease by the age of 30 and is characterised by cystic kidneys, and finally Senior-Loken syndrome characterised by kidney cysts, inflammation and scarring leading to end stage kidney disease (Baala et al, 2007 ; Helou et al, 2007 ; Hildebrandt et al, 2011 ; Valente et al, 2006 ). Furthermore, mice deficient in Cep290 develop symptoms consistent with ciliopathies including reduced numbers of primary cilia on renal epithelial cells and the development of kidney cysts (Rachel et al, 2015 ). Development of kidney cysts in Cep290 knockout mice is progressive and cysts become more prominent at 2 weeks of age (Rachel et al, 2015 ). Unfortunately, due to the early lethality of GR-null mice we are unable to determine if global GR-null mice develop kidney cysts. Five other cilia-associated genes including Ccp110, Cep97, Kif3a and Rpgr were also significantly downregulated in the GR-null mouse kidney at E18.5. Finally, we also assessed protein levels of two other cilia-related structural proteins, IFT88 and acetylated tubulin with no significant change in their protein level detected by western blot analysis in the E18.5 fetal kidney of GR-null mice. This likely either reflects no direct regulation of their expression by GR-mediated cell signalling or no direct role in the mechanism causing the cilia defect in these mice. Future analysis in isolated primary kidney tubule epithelial cells could be used to rule out a mechanistic role here for IFT88 and acetylated tubulin, and also other cilia-related structural proteins. Analysis of cilia structure at E18.5 in proximal tubule cells by both fluorescent confocal microscopy and scanning electron microscopy showed shortened and abnormally shaped primary cilia in renal tubules from GR-null mice. Strikingly, this phenotype was replicated in renal collecting ducts postnatally in collecting duct-specific-GR-null mice. In contrast, treatment of the renal IMCD3 collecting duct cell line with dexamethasone, a synthetic GC agonist, drove increased primary cilia growth that was inhibited by pre-treatment with the GR antagonist RU486. Dexamethasone stimulation of cilia length was also replicated in iPSC-derived kidney organoids supporting a conserved role for GC/GR signalling in human kidney cell types. Further studies using human kidney organoids co-labelling specific renal tubular cells will clarify if these regulatory effects on cilia length occur in all human tubule cells or are restricted to specific tubular segments such as the proximal or distal tubules. Analysis of key ciliogenesis signalling pathways showed reduced ERK and SUFU activity in GR-null fetal kidney extracts and repression by dexamethasone of an important promoter of cilia disassembly, AURKA, in the IMCD3 cell line (Pugacheva et al, 2007 ). However, cystogenesis is not as simple as stabilising or destabilising the primary cilia. While Kif3a (Lin et al, 2003 ) and Cep164 (Airik et al, 2019 ) deletion can cause cystogenesis due to a lack of primary cilia in neonatal contexts, in adult models of autosomal dominant PKD, Ma et al have demonstrated a cilia-dependent cyst activation pathway (CDCA) where a structurally intact cilia transmits an aberrant growth signal which cilia removal via Kif3a co-deletion prevents (Ma et al, 2017 ). Hence, there are likely multiple cystogenic pathways associated with cilia, such as the major and minor pathways proposed by Hwang et al (Hwang et al, 2019 ). Apart from the single case report that described enhanced survival of a preterm infant with advanced PKD following treatment with dexamethasone (Katz et al, 1979 ), no clinical studies have been reported directly using glucocorticoid steroids to treat PKD (Forcioli-Conti et al, 2015 ; Katz et al, 1979 ). In summary, these results clearly define glucocorticoid signalling as an important regulator and driver of primary cilia formation in renal proximal epithelial cells. The exact mechanism of cilia shortening is not clear and may involve a defect in formation or maintenance of correct cilium length or perhaps increased shedding of cilia. Scanning electron microscopy images of fetal kidney cells from GR-null mice (Fig.  3 ) show the presence of abnormally shaped bulging cilia that could contribute to cilia shortening, but this requires further more detailed analysis. Our results complement the report showing that another steroid hormone aldosterone, acting via the mineralocorticoid receptor, also controls primary cilia growth and length in collecting duct cells (Komarynets et al, 2020 ). Glucocorticoid signalling via genomic actions positively regulate an important subset of ciliogenesis genes, such as Cep290 , Cep110 and Kif3a to promote normal cilia growth and structure. Non-genomic effects where the GR may antagonise other intracellular signalling pathways may also contribute to cilia formation under high endogenous glucocorticoid actions or delivery of exogenous synthetic agonists. Non-genomic cytosolic interactions of the GR with a number of signalling kinases has been described and include RAS, JNK and PI3K where the interaction by GR contributes to anti-inflammatory or tumorigenic responses (Caelles et al, 1997 ; Caratti et al, 2022 ; Limbourg et al, 2002 ). A key regulator of ciliogenesis is AURKA, a kinase that promotes cell proliferation and negatively regulates ciliogenesis in the kidney. Deletion of AURKA in mouse models of autosomal dominant PKD and Joubert Syndrome reduces cyst formation and the disease severity of PKD, in part via modulating AKT signalling in the kidney (Tham et al, 2024 ). We observed reduced AURKA protein levels in dexamethasone treated renal collecting duct cells, indicating the glucocorticoid signalling via the GR regulates AURKA during kidney development, which could be either direct or indirect. We did not see mRNA changes for Aurka in GR-null mice (Fig.  2F ), so we predict this may be post-translational mechanism that involves a direct interaction between the two proteins. In fact in a recent study, Qiao et al demonstrated that the bone osteoclast lineage, AURKA interacts with the GR to prevent glucocorticoid-induced bone loss (Qiao et al, 2023 ). Compellingly in another paper by Sun et al studying mechanisms in ovarian endometriosis, AURKA has also been shown to interact with a related steroid hormone receptor, ERβ, where AURKA promotes a state of cell proliferation and invasion (Sun et al, 2024 ). Further studies are required to confirm direct interaction of the GR with AURKA in ciliated cells and the effect this has on signalling cascades. Finally, this study provides a rationale for the potential use of synthetic glucocorticoid agonists, such as dexamethasone or prednisolone, as a treatment option of human ciliopathies. Primary cilia are cellular and environment sensing organelles that protrude from cell membranes and have important roles during embryogenesis and tissue homeostasis. Defects in cilia cause many human ciliopathies including nephronophthisis and polycystic kidney disease. We show that glucocorticoid steroids action in vivo and in vitro via the cellular glucocorticoid receptor regulate and enhance normal ciliogenesis and cilia growth. Our data demonstrate that glucocorticoid signalling is required for normal primary ciliogenesis.

Primary cilia are microtubule-based organelles found protruding from the surface of most mammalian cells (Satir et al, 2010 ; Zhao et al, 2023 ). The primary cilium mediates diverse developmental signalling pathways and senses extracellular stimuli to maintain tissue homeostasis. In the kidney, primary cilia are located on epithelial cells of tubules projecting into the lumen where they function as fluid sensors (Pluznick and Caplan, 2015 ). Primary cilia are 1–10 µm long and 0.2–0.3 µm wide and comprise four main parts: the axoneme, basal body, ciliary membrane and the transition zone (Zhao et al, 2023 ). The axoneme contains nine microtubules arranged as a 9 + 0 ring structure attached to the cell membrane (Zhao et al, 2023 ). Defects in primary cilia function or structure cause a group of human diseases called ciliopathies including Polycystic Kidney Disease (PKD) (McConnachie et al, 2021 ). A clinical case report of a preterm infant with advanced PKD has documented improved PKD outcomes following treatment with the synthetic glucocorticoid steroid dexamethasone (Katz et al, 1979 ). In addition, a study in human adipose stem cells, showed that dexamethasone treatment in vitro promoted cilia growth and elongation, and dexamethasone has also been shown to restore primary cilia gene expression in cancer cells (Forcioli-Conti et al, 2015 ). Additionally, a recent study demonstrated that aldosterone acting via the mineralocorticoid receptor in renal collecting ducts could regulate cilia length (Komarynets et al, 2020 ), and in a separate study, glucocorticoids regulate cilia gene expression in hippocampal neurones (Mifsud et al, 2021 ). Given these observations we have investigated the impact of glucocorticoid signalling on ciliogenesis in glucocorticoid receptor null mice, glucocorticoid treated human organoid cultures and renal cell lines. Glucocorticoid steroid hormones play essential roles for maturation and growth of many fetal organs including the lung and heart, yet kidney-specific roles are not well characterised (Fowden and Forhead, 2015 ; Whirledge and DeFranco, 2018 ). Glucocorticoids activate the intracellular glucocorticoid receptor (GR) that acts primarily as a nuclear transcriptional regulator (Ackermann et al, 2010 ; Cole and Young, 2017 ; Quax et al, 2013 ). Global GR-null mice die shortly after birth due to respiratory failure (Bird et al, 2014 ; Bird et al, 2007 ; Cole et al, 1995 ) and several studies have characterised mouse models of conditional GR deletion in kidney epithelium. Deletion in the distal nephron using a Ksp- cre driver resulted in a mild increase in blood pressure in adult mice with normal renal histology (Goodwin et al, 2010 ). A more extensive renal tubular deletion of GR using an inducible Pax8 -cre driver altered the expression of several epithelial sodium transporters including the thiazide-sensitive Na + /Cl - cotransporter ( Ncc ), the Na-K-Cl cotransporter-2 ( Nkcc2 ) (Canonica et al, 2019 ). Furthermore, adult mice heterozygous for a global GR-null allele developed salt-sensitive hypertension and reduced mRNA levels of Ncc when on a high salt diet (Ivy et al, 2018 ). Additionally, rodents exposed to synthetic glucocorticoids have reduced birth weight and increased risk of high blood pressure in adulthood (Seckl, 2004 ). It has been shown that both glucocorticoid excess and deficiency can lead to salt-sensitive hypertension (Bailey et al, 2009 ; Ivy et al, 2018 ). We show here that GR-medicated glucocorticoid signalling regulates primary ciliogenesis in the renal tubule. Our results demonstrate a capacity for GR agonists and antagonists to modulate cilia formation and function.

Table EV1 Appendix Peer Review File Source data Fig. 1 Source data Fig. 2 Source data Fig. 3 Source data Fig. 4 Source data Fig. 5 Source data Fig. 6 Source data Fig. 7 Figure EV2 Source Data Figure EV4 Source Data Figure EV5 Source Data Expanded View Figures Table EV1 Appendix Peer Review File Source data Fig. 1 Source data Fig. 2 Source data Fig. 3 Source data Fig. 4 Source data Fig. 5 Source data Fig. 6 Source data Fig. 7 Figure EV2 Source Data Figure EV4 Source Data Figure EV5 Source Data Expanded View Figures