podDKD mice develop a renal phenotype by 24 weeks

A. and B. Body weight (A) and blood glucose (B) are not significantly different in podDKD mice at 24 weeks compared with littermate controls (n=7-8 each group). C. Urinary Albumin Creatinine ratio (uACR) is significantly increased in podDKD mice at 24 weeks. Unpaired t-test, **p<0.01 (n=7-8 mice per group). D. Images and quantification of PAS staining show tubular protein casts (indicated by arrows) and glomerulosclerosis in podDKD mice. Scale bar=25 μm. Unpaired t-test *p<0.05. E. Masson’s trichrome staining shows increased fibrosis (blue staining) in podDKD mice at 24 weeks. Scale bar=25 μm. F. Transmission Electron Micrograph (TEM) images of glomerular filtration barrier (GFB) (scale bar left panels=5 μm, right panels =500 nm) show ultrastructural damage to the GFB in podDKD mice with significantly increased foot process width. Unpaired t-test **p<0.01. G. Immunofluorescent staining and quantification of WT1 in podDKD and Cre negative control mice at 24 weeks of age shows significant reduction in % podocytes per glomerulus in podDKD mice. Nuclei counterstained with DAPI. Scale bar=100 μm. Unpaired t-test ****p<0.0001, ≥8 glomeruli analysed per mouse, 3 mice per group. Scale bar=100 μm.

Simultaneous knockout of podocyte IR and IGF1R in vitro is highly detrimental

A. Western blot shows >80% reduction of IR and IGF1R protein in ciDKD cells. t-test, ****p<0.00001, n=18. B. Phosphorylation of AKT and p44/42MAPK in response to acute insulin stimulation at 10 nM and 100 nM for 10 minutes was significantly reduced in ciDKD podocytes. One way ANOVA, **p<0.01, *p<0.05, n=3. C. Phosphorylation of AKT and p44/42MAPK in response to acute IGF1 stimulation at 10 and 100 ng/ml for 10 minutes was reduced in ciDKD podocytes. One way ANOVA, *p<0.05, n=3. D. Fewer than 50% of ciDKD cells survive 7 days after gene excision. t-test, ****p<0.0001, n=3-4 independent experiments.

Proteomic analysis of ciDKD podocytes reveals downregulation of spliceosomal proteins

A. Schematic outlining workflow for proteomic analysis. B. Heat map showing hierarchical clustering of ciDKD vs wild-type podocyte proteomes (decreased protein expression in green, increased expression in red) and the GO and KEGG terms enriched in four major clusters. C. and D. KEGG enriched terms in STRING. Downregulated pathways (green) are associated with higher enrichment scores in comparison to upregulated pathways (red). Enrichment scores are computed by STRING using the Kolmogorov-Smirnoff test. KEGG term “Spliceosome” (in black) is associated with a high enrichment score and the highest false discovery rate across all terms. E. Western blots show significantly reduced levels of EIF4A, SF3B4 and PTBP2 in ciDKD podocytes compared with wild-type cells. Unpaired t-test, ***p<0.001, **p<0.01, *p<0.05, n=3 independent experiments. F. Representative immunohistochemistry and quantification using an antibody to Sf3b4 shows reduced expression in the glomeruli of podDKD mice compared to littermate controls. T-test ****p<0.0001.

Exposure of cultured podocytes, but not glomerular endothelial cells, to the spliceosome inhibitor pladienolide B results in dose dependent cell death

A. Schematic overview of the spliceosome pathway showing a majority of spliceosome proteins are significantly downregulated (green) in ciDKD podocytes. B-D. Pladienolide B exposure for 48 hours in HeLa cells (B), glomerular endothelial cells (GenC) (C) and podocytes (D). One way ANOVA, ****p<0.0001, **p<0.01, *p<0.05, n=3 independent experiments. E. Bright field images of wild-type podocytes exposed to the indicated concentrations of pladienolide B. Scale bar=100 μm.

Loss of podocyte IR and IGF1R is associated with increases in intron retention and unproductive transcript expression

A. Schematic outlining long-read RNA sequencing workflow. B. An overview of the alternative splicing events quantified by FLAIR. Boxes represent exons (blue=constitutive exons; yellow=alternative exons), lines represent introns. C. Boxplot showing the fraction of transcripts with intron retention events in ciDKD and wild-type podocytes. t-test, **p<0.01, n=4 for each experimental condition. D. Box plot of the fraction of productivity events in the ciDKD and wild-type transcriptomes shows a higher proportion of unproductive transcripts in ciDKD podocytes. Includes no start codon; start codon but no stop codon; productive transcripts; premature termination codon i.e. unproductive transcript. t-test, ***p<0.001, **p<0.01, *p< 0.05, n=4 for each experimental condition. E. Proportional stacked bar graph of productivity events in the ciDKD and wild-type transcriptomes.

Compound knockdown of the IR and IGF1R alters the podocyte transcriptome

A. UCSC genome browser tracks of transcripts annotated to Fn1. Red lines indicate the number of reads in each sample containing the EDA and EDB exons (n=4 each group). B. Boxplot showing differential expression of an Fn1 transcript containing EDA/EDB exons in ciDKD podocytes. C. Exon/intron structure of Fn1 transcript differentially expressed in ciDKD vs wild-type podocytes. Expression of an EDA and EDB exon-containing transcript (not normally expressed in mature podocytes and associated with fibrosis) is increased in ciDKD podocytes. D. qPCR shows decreased expression of Hcfc1r1 mRNA in ciDKD podocytes. t-test, ****p<0.00001, n=4 independent experiments. E. qPCR using primers specific to the Fn1 EDB exon shows an increase in the expression of fibrosis associated EDB exon-containing transcripts in ciDKD podocytes. t-test, **p<0.001, n=4 independent experiments.

Multiple spliceosomal-protein post-translational modifications (PTM) occur in response to insulin and IGF1 stimulation

A. Heat map showing differential phosphorylation at the indicated phosphosites of spliceosome-related proteins in ciDKD podocytes and control cells stimulated with 10 nM insulin or 10 ng/ml IGF1 for 10 minutes (decreased phosphorylation in green, increased phosphorylation in red). Those phosphorylation events that also occurred in Turewicz study [37] highlighted in yellow boxes. B. Differentially phosphorylated proteins identified in A occur throughout the spliceosome cycle (red boxes). C. Heat map showing differential phosphorylation at the indicated phosphosites of serine/arginine repetitive matrix 2 (Srrm2) in ciDKD podocytes and control cells stimulated with 10 nM insulin or 10 ng/ml IGF1 for 10 minutes (decreased phosphorylation in green, increased phosphorylation in red).

Relative contributions of the IR and IGF1R to spliceosomal modulation.

A. Heatmap of hierarchical clustering of ciDKD vs ciIGF1RKD vs ciIRKD vs wild-type podocytes (n=9) proteomes (high relative expression in red, low relative expression in green, equal expression in black). Top of the heat map contains most significantly enriched groups, with enrichment intensity decreasing further down. The spliceosomal targets Sf3b4, Eif4a3, and Ptbp2 were mapped onto the proteomic heatmap comparing wild-type podocytes to cells with knockdown of the IR and/or the IGF1R. Sf3b4 expression is reduced in all three knockdown cell lines. B. Schematic showing signaling pathways uniquely enriched in ciDKD podocytes including the downregulation of spliceosomal tri-snRNP complex assembly. C. Heat map showing differential phosphorylation events in ciIGF1RKD; ciDKD; ciIRKD and wild-type podocytes (decreased phosphorylation in green, increased phosphorylation in red).

Generation of podDKO mice. Related to Figure 1.

A. Breeding scheme used to generate podDKD mice. B. Progression of albuminuria in podDKD mice. Showing that uACR becomes significantly increased by 24 weeks. Unpaired t test, **p<0.005. C. Western blot of primary podocytes derived from podDKD and Cre negative littermate control mice D. No difference in uACR between Cre negative podDKD littermate controls, Cre negative and pod2.Cre expressing mice at 6 months. uACR is significantly increased in podDKD mice. t-test, **p<0.01, n=5-8 mice per group. E. Graph showing serum creatinine levels in subset of control and podDKD mice age (n=5 each group).

Related to Figure 2.

A Development of ciDKD, ciIRKD and ciIGF1RKD podocytes. Primary culture podocytes (homozygous floxed) isolated from transgenic mouse. Conditional immortalisation with temperature sensitive SV40 construct and then excision of IR and/or IGF1R using lentiviral delivered Cre recombinase. B Survival rate of Wild-type and IR/IGF1R floxed podocytes 3 days and 7 days after Cre recombinase induced gene knockdown.t test, ***p<0.001, ****p<0.0001. (n=3 each group)

Proteomic analysis of ciDKD podocytes. Related to Figure 3.

A Heat map of sample clustering B Plot of principal component analysis C Volcano plot showing changes in the proteome of ciDKD podocytes relative to wild-type control cells; log2 fold change (FC) vs −log10 p-value of the scaled abundances. Shaded areas show proteins with FC < and > 2.

Long read RNA sequencing analysis of ciDKD podocytes. Related to Figures 5 and 6.

A Plot indicating the number of reads sequenced. B. Plot of variance transformed data to normalise for library size. C Plot of principal component analysis shows clustering of experimental groups. D-F KEGG enriched terms in ciDKD podocytes genes with intron retention (D), cassette exon splicing (E) and alternative 5’ splicing (F).

Spliceosomal-protein related kinase post-translational modifications (PTM) occuring in response to insulin and IGF1 stimulation. Related to Figure 7.

Graphs showing the relative expression of Cdk11b s578 (A), Srpk1 s311 (B) and Pacsin3 s383 (C) in wild-type and ciDKD podocytes stimulated with 10 nM insulin or 10 ng/ml IGF1. Unpaired t test *p<0.05. **p<0.005, n=3.

Proteomic analysis of ciIRKD, ciIGF1RKD and ciDKD podocytes. Related to Figure 8.

A. Western blot shows >95% reduction of IR expression in ciIRKO podocytes, t test. ****p<0.0001. B. Western blot shows >95% reduction of IGF1R expression in ciIGF1RKD podocytes, t test. ****p<0.0001. C-E. Principal Component Analysis. ciDKD and ciIGF1RKD groups are significantly different to each other and the ciIRKD and wild-type groups (C). IRKD and WT groups show a moderate degree of overlap but do have a clear degree of difference. (D). 2D PCA analysis using third most significant component (Component 3), which demonstrates that along components 2 and 3, all 4 groups distinctly cluster separate to one another (E).