Involvement of dysregulated RNA binding protein and alternative splicing regulatory networks in diabetic nephropathy from type 2 albuminuric cohorts

Abstract

Background: Diabetic nephropathy (DN) is a primary contributor to end-stage renal disease, yet the underlying molecular mechanisms remain incompletely understood. This study aims to elucidate the role of RNA-binding proteins (RBPs) and RBP-alternative splicing (AS) regulatory networks in the pathogenesis of DN.

Methods: Two RNA-seq datasets (GSE117085 and GSE142025) were retrieved from the Sequence Read Archive (SRA) database. Regulated alternative splicing events (RASEs) and genes (RASGs) of RASEs, along with differentiated RBPs, were identified. Validated differentiated RBPs were correlated with clinical features using the Nephroseq v5 online platform. Using the DN mouse model and RT-qPCR, validated the alternative splicing of RNA.

Results: Our analysis revealed 15 differentiated RBP genes and 423 RASEs in the kidney cortex of DN rats compared to controls. Enrichment analysis highlighted lipid metabolism pathways for RASGs. Seven of the identified RBPs were validated in kidney biopsy samples from DN patients versus controls. A co-deregulatory network was constructed based on dysregulated RBPs and RASEs, with select RASGs identified. In vivo experiments, compared to normal mice, the mRNA levels of RPS19 were significantly elevated in the renal tissues of DN mice, while the levels of CPEB4 and CRYZ were markedly decreased.

Conclusion: In conclusion, this study provides evidence implicating dysregulated RBPs and RBP-AS regulatory networks in the development of diabetic nephropathy. The validated RBPs exhibited close associations with clinical biomarkers, reinforcing their potential as therapeutic targets for DN. These findings enhance our understanding of the molecular basis of DN and offer new insights for future research and intervention strategies.

Clinical trial: Not applicable.

Keywords: Alternative splicing; Co-expression; Diabetic nephropathy; RNA binding protein; Transcriptome.

Fig. 1

Transcriptome analysis of differentially expressed genes (DEGs) in non-diabetic rat kidney cortex (Ctrl) and diabetic rat kidney cortex (Disease). (A) The hierarchical clustering heat map shows sample correlation between Disease and Ctrl samples. (B) The volcano plot shows all DEGs between Disease and Ctrl samples. P-value ≤ 0.05 and FC (fold change) ≥ 1.5 or 0.75. (C) Principal component analysis (PCA) based on reads per kilobase of exon per million fragments mapped (FPKM) value of all DEGs. The ellipse for each group is the confidence ellipse. (D) The hierarchical clustering heat map shows expression levels of all DEGs. (E, F) The bar plot shows the most enriched GO biological process results of the up-regulated DEGs (E) and down-regulated DEGs (F)

Fig. 2

Transcriptome analysis of alternative splicing regulation in disease and Ctrl samples. (A) The bar plot shows the number of all significant regulated alternative splicing events (RASEs). X-axis: RASE number. Y-axis: the different types of AS events. (B) PCA based on PSI (percent-spliced-in) value of all non-intron retention (NIR) RASEs. The ellipse for each group is the confidence ellipse. (C) PSI heat map of all significant NIR RASEs. AS filtered should have detectable splice. Junctions in all samples and at least 80% of samples should have > = 10 splice junction supporting reads. (D) Bar plot showing the most enriched GO biological process results of the NIR-regulated alternative splicing genes (RASGs). (E) Bar plot showing the most enriched KEGG pathways results of the NIR RASGs

Fig. 3

Correlation analysis between differentially expressed RNA binding protein genes and NIR RASEs. (A) Venn diagram showing differentially expressed RNA binding protein genes (RBPs). (B) The hierarchical clustering heat map shows expression levels of differential expressed RBPs in A. (C) The bar plot exhibits the most enriched GO biological process results illustrated for RASGs co-disturbed by top hub RBPs. (D) The co-deregulation of alternative splicing network between top hub RBPs (the red color represents the number of connections) and RASEs (the green circles include NIR RASEs) in DN. The top enriched GO biological process of co-disturbed RASGs is shown in blue and brown color. (E) The box plot showed PSI levels of NRG1, which detected mutually exclusive exons (MXE), and the IGV-sashimi plot showed the RASEs and binding sites across mRNA of NRG1. Reads distribution of RASE was plotted in the up panel, and the transcripts of each gene are shown below

Fig. 4

RBPs-alternative splicing co-expression networks on diabetic nephropathy in GSE142025 datasets. (A, B) Box plot shows expression levels of upregulated (A) or downregulated RBPs (B) in Fig. 3A. *** P < 0.001, * P < 0.05. (C) The co-deregulation of alternative splicing network between top hub RBPs (red color) and RASEs (green color, which includes NIR RASEs in GSE142025 datasets) in DN. The top enriched GO biological process of co-disturbed RASGs is shown in blue and brown color. (D) The box plot shows PSI levels of SLCO2B1, which detected alternative 5’splice site (A5SS), and IGV-sashimi plot shows the RASEs and binding sites across mRNA of SLCO2B1. Reads distribution of RASE is plotted in the up panel, and the transcripts of each gene are shown below. *** P < 0.001

Fig. 5

The expression pattern of RBP genes in Nephroseq database. (A) The expression of ENDOU increased in db/db C57BLKS (left panel), eNOS-deficient C57BLKS db/db (middle panel), and DBA/2 mice (right panel). (B) The expression of KHDRBS3 decreased in eNOS‐deficient C57BLKS db/db mice. (C) The expression of AFF3 decreased in eNOS‐deficient C57BLKS db/db mice. (D) The expression of TDRD5 decreased in eNOS‐deficient C57BLKS db/db mice. (E) The expression of ERIC2 decreased in eNOS‐deficient C57BLKS db/db mice. * P < 0.05, ND vs. DN

Fig. 6

The expression pattern of RBP genes in kidney in DN patients. (A) The expression of RPS19 increased in DN patients of Woroniecka Diabetes TubInt (left panel) and Schmid Diabetes TubInt (right panel). (B) The expression of AFF3 increased in DN patients of ERCB Nephrotic Syndrome TubInt. (C) The expression of ZC3H12D increased in DN patients of ERCB Nephrotic Syndrome TubInt. (D) The expression of CPEB1 decreased in DN patients of Woroniecka Diabetes TubInt. * P < 0.05, ND vs. DN

Fig. 7

Correlation of RBP gene with ACR and eGFR in DN mouse model and DN patients. (A) CRYZ is negatively correlated with ACR in DN mouse models. (B) KHDRBS3 is positively correlated with ACR in DN mouse models. (C) ZC3H12D is positively correlated with ACR in DN mouse models. (D) TDRD5 is positively correlated with ACR in DN mouse models. ACR, albumin to creatinine ratio. (E) RPS19 is negatively correlated with eGFR in DN patients. (F) AGO3 is negatively correlated with eGFR in DN patients. (G) NSUN3 is negatively correlated with eGFR in DN patients. (H) RBM11 is negatively correlated with eGFR in DN patients. (I) TDRD5 is negatively correlated with eGFR in DN patients. eGFR, estimated glomerular filtration rate

Fig. 8

Quantitative real-time PCR to detect the alternative splicing in DN model (A), splice variants of CPEB4 in the GSE117085 dataset (B).^aP < 0.05 vs. the Con group, ^bP < 0.05 vs. the normal mouse group, ^nsP > 0.05 vs. the normal mouse group