Diabetic nephropathy (DN), a vascular complication of diabetes, is the leading cause of death in patients with diabetes. The contribution of aberrantly expressed circular RNAs (circRNAs) to DN in vivo is poorly understood. Integrated comparative circRNA microarray profiling was used to examine the expression of circRNAs in diabetic kidney of db/db mice. We found that circRNA_010383 expression was markedly downregulated in diabetic kidneys, mesangial cells, and tubular epithelial cells cultured in high-glucose conditions. circRNA_010383 colocalized with miRNA-135a (miR-135a) and inhibited miR-135a function by directly binding to miR-135a. In vitro, the knockdown of circRNA_010383 promoted the accumulation of extracellular matrix (ECM) proteins and downregulated the expression of transient receptor potential cation channel, subfamily C, member 1 (TRPC1), which is a target protein of miR-135a. Furthermore, circRNA_010383 overexpression effectively inhibited the high-glucose–induced accumulation of ECM and increased TRPC1 levels in vitro. More importantly, the kidney target of circRNA_010383 overexpression inhibited proteinuria and renal fibrosis in db/db mice. Mechanistically, we identified that a loss of circRNA_010383 promoted proteinuria and renal fibrosis in DN by acting as a sponge for miR-135a. This study reveals that circRNA_010383 may be a novel therapeutic target for DN in the future.
Introduction
Diabetic nephropathy (DN), which occurs in 20–40% of patients with diabetes, is the most severe microvascular complication of diabetes and a leading cause of death among patients with diabetes (1–3). Pathologically, DN is characterized by the progressive accumulation of extracellular matrix (ECM) in the glomerular mesangium and thickening of the basement membrane, which leads to glomerulosclerosis (4). Previously, miRNAs and long noncoding RNAs have been demonstrated to contribute to pathological processes and the development of renal fibrosis in DN (5–7). Circular RNAs (circRNAs), another class of noncoding RNAs that are widely expressed in mammals, are circularized by the joining of the 3′- and 5′-ends (8,9). Altered circRNA expression profiles have been reported in various diseases, such as ischemic heart disease (10), cancer (11), and diabetes (12). At the molecular level, certain circRNAs contain miRNA response elements (MREs), which suppress their function as miRNA sponges (13,14). Emerging evidence shows that circRNAs play an important role in the development of microvascular complications in diabetes. The downregulation of circRNA DMNT3B expression contributes to diabetic retinal vascular dysfunction by targeting miR-20b-5p and BAMBI (15). cPWWP2A regulates pericyte biology, and cPWWP2A overexpression alleviates diabetes-induced retinal vascular dysfunction (16). Nevertheless, the potential functions of circRNAs and the related mechanisms in DN are still largely unknown.
In this study, we used circRNA expression profiles to identify novel circRNAs in DN. We identified circRNA_010383 as a signature circRNA in DN, as indicated by the downregulation of its expression in diabetic conditions in vitro and in vivo. Next, we showed that circRNA_010383 functions as a sponge for miR-135a, the expression of which is upregulated in DN, which promotes renal fibrosis by regulating the expression of the transient receptor potential cation channel subfamily C, member 1 (TRPC1), as shown in our previous study (6). Thus, we demonstrate that circRNA_010383 is a signature circRNA in DN that negatively regulates renal fibrosis by acting as a sponge for miR-135a.
Research Design and Methods
Animal Studies
Male db/db mice on a C57BKS background and littermate control mice (db/m) were used in this study (GemPharmatech, Nanjing, China). Mice were housed at an ambient temperature of 22°C with a 12-h light-dark cycle and maintained on a normal chow diet. All studies and experimental procedures were approved by the Southern Medical University of Animal Experimentation Ethics Committee, and the experimental methods were performed in accordance with the approved guidelines. Blood glucose was measured with a Touch UltraSmart Blood Glucose Meter. Urinary albumin was measured with a mouse albumin ELISA quantitation kit (Bethyl Laboratories, Montgomery, TX). Urine creatinine levels were determined as described previously (17). Urinary albumin was standardized to urine creatinine and expressed as mg/mmol creatinine.
RNA Extraction and Quantitative Real-time-PCR
Total RNA and miRNA were extracted as previously described (6,18). circRNA was amplified using divergent primers to target the splice junction. β-Actin and U6 (RiboBio, Guangzhou, China) were used as internal controls for RNA and miRNA, respectively. Quantitative (q)real-time-PCR was conducted as previously described (6,19). Primers for miRNAs were designed and synthesized by RiboBio, and the primers used are listed in Supplementary Table 1.
circRNA Microarray and Data Analysis
The RNA isolation and microarray analysis of circRNAs from mouse whole kidney was performed by KangChen BioTech (Shanghai, China). Total RNA was digested with RNase R (Epicenter) to remove linear RNAs. The enriched circRNAs were amplified and transcribed into fluorescent circRNAs by a random priming method and hybridized onto an Arraystar mouse circRNA Array v2. Bioinformatics analysis and visualization of the microarray data were performed with MeV 4.6 (mev.tm4.org).
Human Tissue Specimens
Human DN biopsy sections were obtained from diagnostic renal biopsies performed at Zhujiang Hospital, Southern Medical University. Paracarcinoma normal kidney tissues were obtained from patients who underwent nephrectomy at Zhujiang Hospital. All studies involving human kidney tissues were approved by the Zhujiang Hospital Institutional Ethics Committee.
Cell Culture
Mouse glomerular mesangial cells (mMCs; the murine SV40-transfected mouse mesangial cell line SV40 MES 13; ATCC, Manassas, VA) were cultured in DMEM mixed 1:1 (v/v) with F12 medium (Life Technologies, Grand Island, NY) containing 10% FBS (Gibco) and 1% antibiotics (100 units/mL penicillin and 100 μg/mL streptomycin [Life Technologies]) at 37°C in 5% CO2. Mouse tubular epithelial cells (mTECs) were cultured in DMEM/F12 with 10% FBS and 1% antibiotics at 37°C and 5% CO2. The cells were grown to ∼70–80% confluence and deprived of serum for 24 h before the experiments were performed. Cells were stimulated with d-glucose (Life Technologies) at normal (5.5 mmol/L) or high (30 mmol/L) concentrations for different lengths of time in serum-free medium. d-Mannitol (30 mmol/L) (Life Technologies) was used as a control for osmolality.
Biotin-Coupled Probe RNA Pull-Down Assay
The biotin-coupled probe RNA pull-down assay was performed as previously described with modification (20). The circRNA_010383 probe and a negative control (NC) probe were designed and synthesized by RiboBio. Cells overexpressing circRNA_010383 were lysed. The cell lysates were incubated with streptavidin-coated magnetic beads (Thermo Scientific, Shanghai, China) conjugated with the biotin-labeled circRNA_010383 probe or NC probe for 1 h at room temperature and then washed with wash buffer. The mixture was incubated with proteinase K and lysis buffer. Finally, the bound RNAs were extracted to detect the expression of circRNA_010383 and miR-135a. The probe sequences are listed in Supplementary Table 2.
Dual-Luciferase Reporter Assay
The full sequence of circRNA_010383 was amplified by PCR from mouse genomic DNA and cloned into the psiCHECK2 plasmid (GeneSeed, Guangzhou, China) (psiCHECK2-circRNA_010383-wt), whereas the plasmid containing the MRE sequence harboring mutations in the miR-135a site was used as a control (psiCHECK2-circRNA_010383-Mut) (Supplementary Fig. 1A). Cells were seeded in 24-well plates and incubated for 24 h to 70–80% confluence. The psiCHECK2-circRNA_010383-wt plasmid or psiCHECK2-circRNA_010383-Mut plasmid (200 ng/µL) and 50 nmol/L NC, miR-135a mimic, or 200 nmol/L miR-135a inhibitor were cotransfected into the cells. Luciferase activity was measured 48 h after transfection with a Dual-Luciferase Reporter Assay kit (Promega, Madison, WI). The reporter activity was represented by the ratio of the firefly luciferase activity to the Renilla luciferase activity.
RNA Fluorescence In Situ Hybridization
Fluorescence in situ hybridization was performed as previously reported (15). The kidney sections were deparaffinized and permeabilized with 0.2% Triton X-100. Cells underwent fixation with 4% paraformaldehyde, followed by prehybridization with 0.5% Triton X-100. Then, the sections/cells were incubated with RNA probes in hybridization buffer for 12 h. The circRNA_010383 probe was labeled with FITC, and the miR-135a probe was labeled with Cy3. Nuclei were stained with DAPI (Vector Laboratories, Burlingame, CA). Images were captured with a confocal microscope (Leica, Wetzlar, Germany). The probe sequences are listed in Supplementary Table 2.
Transfection of siRNA, Plasmid, and miRNA Mimics
Cells were transfected with circRNA_010383-specific siRNAs (designed and synthesized by GeneSeed) or NC using Lipofectamine RNAiMAX reagent (Invitrogen). In addition, the cells cultured in high-glucose medium were also transfected with the pLCDH-ciR plasmid, a plasmid containing circRNA_010383, a plasmid containing the linear_010383 RNA sequence (all constructed by GeneSeed, Supplementary Fig. 1B and C), or an miR-135a mimic (designed and synthesized by RiboBio) or a TRPC1 siRNA (designed and synthesized by Genepharma, Shanghai, China) using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s manual. To construct the linear-010383 overexpression plasmid, the full-length circRNA_010383 was amplified and then inserted into PCDH-CMV-MCS-EF1a-GFP-Puro (CD513B-1) vector at the EcoRI and BamHI, which do not form a circular sequence.
Western Blot Analysis
Western blotting was performed as previously described (21). Antibodies against collagen I (Col I) were obtained from Abcam, and antibodies against E-cadherin (1:2,500) (BD Biosciences, San Jose, CA), TRPC1 (1:5,000) (Santa Cruz Biotechnology, Dallas, TX), proliferating cell nuclear antigen (PCNA) (Merck Millipore, Billerica, MA), β-actin (1:1,000) (Sigma-Aldrich, St. Louis, MO), fibronectin (1:5,000) (BD Biosciences), and α-smooth muscle actin (α-SMA) (1:5,000) (Sigma-Aldrich) were obtained from the indicated sources. Horseradish peroxidase-conjugated anti-mouse IgG (1:5,000) and horseradish peroxidase-conjugated anti-rabbit IgG (1:5,000) were obtained from Cell Signaling Technology (Beverly, MA). Images were captured with AxioVision Rel. 4.6 (Carl Zeiss, Goettingen, Germany).
Immunofluorescence
Indirect immunofluorescence was performed with an established procedure. Cells were fixed in methanol for 10 min at 20°C. Fixed cells and the renal tissues were washed with PBS, permeabilized in 0.1% TritonX-100 (Biosharp, Anhui, China) for 10 min at room temperature, and incubated with blocking buffer (5% BSA in PBS) for 1 h at room temperature. Cells or renal tissues were incubated with monoclonal mouse anti-TRPC1 (1:100) or monoclonal mouse anti–E-cadherin (1:100) at 4°C overnight, followed by Alexa Fluor 488-conjugated anti-mouse IgG (1:1,000) antibody and then staining with DAPI. Positive staining was captured with a laser-scanning confocal microscope (Zeiss LSM 510 META; Carl Zeiss).
Ultrasound-Mediated Gene Transfer of circRNA_010383 Plasmids Into the Kidneys of db/db Mice
As previously described (22), the circRNA_010383 expression plasmid (pLCDH-circRNA_010383) was combined with SonoVue (Bracco, Milan, Italy) at a ratio of 1:1 (v/v). Eight db/db mice (20–22 g body weight, 10 weeks old) were treated with the mixed solution (200 μL) via tail vein injection. An ultrasound transducer (Therasonic; Electro-Medical Supplies, Wantage, U.K.) was placed directly on the skin of the back over the kidneys with a continuous wave output of 1 MHz at 1 W power output for a total of 5 min per side. Eight db/db mice and age-matched db/m mice were treated with a mixture containing the same amount of empty control plasmid (pLCDH-EV) using the same procedure and were used as two control groups. To maintain the transgene expression levels, gene therapy was given at ages 10, 12, 14, 16, and 18 weeks. The body weight and blood glucose levels were monitored, and urine was collected every 2 weeks. All mice were killed at age 20 weeks, and their kidneys and serum were collected for analysis.
Histology Examination
Kidneys were fixed in 4% paraformaldehyde overnight at 4°C and then embedded in paraffin. Paraffin-embedded tissues (4-μm sections) were stained with periodic acid Schiff (PAS) and Sirius red stains for subsequent immunohistochemical examination. The mesangial matrix index was calculated using Image-Pro Plus software (Media Cybernetics, Bethesda, MD) and the PAS-stained sections in 10 random fields in each animal. For quantitation of the fibrotic area, the collagen-stained area was calculated as a percentage of the total area by using Image-Pro Plus software in Sirius red-stained sections in 10 random fields in each animal.
Statistical Analysis
All data are shown as the mean ± SD. Statistical analyses were performed with the Student t test or one-way ANOVA, followed by multiple comparison from GraphPad Prism 5.0 (GraphPad Software, San Diego, CA). A repeated-analysis ANOVA was used for blood glucose, body weight, and albumin excretion. Statistical significance was defined as P < 0.05.
Data and Resource Availability
The resources generated during and/or analyzed during the current study are available from the corresponding authors on reasonable request.
Results
circRNA Expression Profiles in DN
Kidneys were collected from a well-known established mouse model of DN that included five db/db mice and five littermates aged 10 weeks. circRNA expression profiles were evaluated by microarray hybridization. The variation in circRNA expression between the two groups is shown in Supplementary Fig. 2. In total, 140 differentially expressed circRNAs with fold-changes >2.0 were identified (Supplementary Table 3). A hierarchical clustering analysis was performed based on the 140 differentially expressed circRNAs (Fig. 1A). We selected eight circRNAs to validate the microarray results by real-time PCR. We confirmed that the expression of circRNA_000551, circRNA_009326, circRNA_002722, and circRNA_010009 was upregulated, whereas the expression of circRNA_013935, circRNA_010964, circRNA_010383, and circRNA_000903 was downregulated in diabetic mice compared with control mice (Fig. 1B and C).
circRNA expression profile in a mouse model of DN. A: The cluster heat map shows differentially expressed circRNAs with a greater than 2.0-fold change. The red and green colors indicate high and low expression levels, respectively (n = 5 per group). B and C: Eight differentially expressed—upregulated (B) or downregulated (C)—circRNAs were validated by real-time PCR, with β-actin as a normalization control (n = 6 per group). D: Expression and localization of human homolog of circRNA_010383 was detected by in situ hyperdilation. E: qRT-PCR analysis of human homolog of circRNA_010383 levels in normal kidney tissues (n = 5) and DN (n = 8). F: circRNA_010383 exemplifies the validation strategy. Convergent (divergent) primers were used to detect total (circular) RNAs. Sanger sequencing confirmed the head-to-tail splicing. G: Divergent primers amplified circRNAs (circR) in the cDNA but not the gDNA in mMCs and mTECs. H: Real-time PCR analysis of circRNA_010383 and Akap7 levels in mMCs and mTECs upon RNase R treatment (n = 3). I: qRT-PCR analysis of circRNA_010383 levels in mMCs and mTECs, respectively, cultured in normal glucose medium (5.5 mmol/L), d-mannitol (30 mmol/L), or high-glucose medium (30 mmol/L) for the indicated time. Data represent the mean ± SD (n = 3). *P < 0.05, **P < 0.01, and ***P < 0.001 vs. normal glucose at the same time point.
circRNA expression profile in a mouse model of DN. A: The cluster heat map shows differentially expressed circRNAs with a greater than 2.0-fold change. The red and green colors indicate high and low expression levels, respectively (n = 5 per group). B and C: Eight differentially expressed—upregulated (B) or downregulated (C)—circRNAs were validated by real-time PCR, with β-actin as a normalization control (n = 6 per group). D: Expression and localization of human homolog of circRNA_010383 was detected by in situ hyperdilation. E: qRT-PCR analysis of human homolog of circRNA_010383 levels in normal kidney tissues (n = 5) and DN (n = 8). F: circRNA_010383 exemplifies the validation strategy. Convergent (divergent) primers were used to detect total (circular) RNAs. Sanger sequencing confirmed the head-to-tail splicing. G: Divergent primers amplified circRNAs (circR) in the cDNA but not the gDNA in mMCs and mTECs. H: Real-time PCR analysis of circRNA_010383 and Akap7 levels in mMCs and mTECs upon RNase R treatment (n = 3). I: qRT-PCR analysis of circRNA_010383 levels in mMCs and mTECs, respectively, cultured in normal glucose medium (5.5 mmol/L), d-mannitol (30 mmol/L), or high-glucose medium (30 mmol/L) for the indicated time. Data represent the mean ± SD (n = 3). *P < 0.05, **P < 0.01, and ***P < 0.001 vs. normal glucose at the same time point.
According to the circBank database (https://www.circbank.cn/), circRNA_010383 is conserved between human and mouse species. In situ hybridization of the FITC-labeled homolog of circRNA_010383-probe showed that circRNA_010383 was predominately localized to renal tubular epithelial cells and glomerular mesangial cells and that circRNA_010383 expression in kidney tissues from patients with DN was significantly downregulated compared with that of normal kidney tissues (Fig. 1D), which was confirmed by real-time PCR (Fig. 1E). According to the reference genome in the University of California, Santa Cruz genome database (https://genome.ucsc.edu/), we assumed that the genomic length of circRNA_010383 is 22,423 base pairs and that the spliced length is 579 base pairs. Sanger sequencing showed that head-to-tail splicing occurred in the exons from Akap7 (Fig. 1F). We found that circRNA_010383 was amplified with divergent primers in cDNA but not genomic DNA (gDNA) in mMCs and mTECs (Fig. 1G). In addition, we found that the mRNA of Akap7 was digested by RNase R but that circRNA_010383 was resistant to RNase R treatment (Fig. 1H). Furthermore, high glucose significantly downregulated circRNA_010383 in mMCs and mTECs (Fig. 1I). These data indicate the circRNA_010383 is in a circular form and may function stably in DN.
circRNA_010383 Acts as a Sponge for miR-135a
We hypothesized that circRNA_010383 plays a role in DN through miRNAs. The interactions between circRNA_010383 and its target miRNAs were predicted using the Arraystar homemade miRNA target prediction software. Notably, miR-135a, the expression of which we previously showed to be upregulated in DN, which promotes renal fibrosis in DN by targeting TRPC1 (6), is one of the predicted miRNA targets of circRNA_010383 (Fig. 2A). We tested the interactions between circRNA_010383 and miR-135a. The in situ hybridization of a digoxigenin-labeled circRNA_010383 probe and a biotinylated miR-135a probe showed that circRNA_010383 and miR-135a predominantly colocalize in the cytoplasm of mMCs and mTECs (Fig. 2B). Luciferase reporter assays were used to confirm the direct interaction between miR-135a and circRNA_010383. The transient cotransfection of a miR-135a mimic and luciferase reporter in mMCs and mTECs resulted in a significant reduction of the luciferase activity (Fig. 2C), and this inhibitory effect of miR-135a on the activity of luciferase reporter linked with circRNA_010383 was abolished by a miR-135a inhibitor (Fig. 2C).
circRNA_010383 acts as a sponge for miR-135a. A: Bioinforming miR-135a binding sites in the 3′-untranslated region (UTR) of circRNA_010383. B: Fluorescence in situ hybridization assay shows miR-135a (red), circRNA_010383 (green), and nuclei (DAPI, blue) in mMCs and mTECs. C: Luciferase assay of mMCs and mTECs transfected with the psiCHECK2 vector, psiCHECK2-circRNA_010383, or psiCHECK2-circRNA_010383 mutant reporter with a miR-135a mimic, NC oligonucleotide, or miR-135a inhibitor (n = 3). **P < 0.01, *P < 0.05 vs. NC. circRNA_010383 (D) and miR-135a (E) in mMCs and mTECs was pulled down by a circRNA_010383-specific probe and detected by a real-time PCR assay. **P < 0.01 vs. miR-NC/miRNA inhibitor NC, and *P < 0.05 vs. miRNA inhibitor NC. F: miR-135a expression was evaluated by real-time PCR assay in mMCs and mTECs transfected with circRNA_010383 siRNA. **P < 0.01 vs. control probe. G: circRNA_010383 expression in mMCs and mTECs transfected with miR-135a mimic. Real-time PCR analysis of miR-135a (H) and TRPC1 (I) levels in mMCs and mTECs cultured in normal glucose medium (5.5 mmol/L), d-mannitol (30 mmol/L), or high-glucose medium (30 mmol/L) for the indicated time. Data represent the mean ± SD (n = 3). *P < 0.05 and **P < 0.01 vs. normal glucose at the same time point. ns, not significant.
circRNA_010383 acts as a sponge for miR-135a. A: Bioinforming miR-135a binding sites in the 3′-untranslated region (UTR) of circRNA_010383. B: Fluorescence in situ hybridization assay shows miR-135a (red), circRNA_010383 (green), and nuclei (DAPI, blue) in mMCs and mTECs. C: Luciferase assay of mMCs and mTECs transfected with the psiCHECK2 vector, psiCHECK2-circRNA_010383, or psiCHECK2-circRNA_010383 mutant reporter with a miR-135a mimic, NC oligonucleotide, or miR-135a inhibitor (n = 3). **P < 0.01, *P < 0.05 vs. NC. circRNA_010383 (D) and miR-135a (E) in mMCs and mTECs was pulled down by a circRNA_010383-specific probe and detected by a real-time PCR assay. **P < 0.01 vs. miR-NC/miRNA inhibitor NC, and *P < 0.05 vs. miRNA inhibitor NC. F: miR-135a expression was evaluated by real-time PCR assay in mMCs and mTECs transfected with circRNA_010383 siRNA. **P < 0.01 vs. control probe. G: circRNA_010383 expression in mMCs and mTECs transfected with miR-135a mimic. Real-time PCR analysis of miR-135a (H) and TRPC1 (I) levels in mMCs and mTECs cultured in normal glucose medium (5.5 mmol/L), d-mannitol (30 mmol/L), or high-glucose medium (30 mmol/L) for the indicated time. Data represent the mean ± SD (n = 3). *P < 0.05 and **P < 0.01 vs. normal glucose at the same time point. ns, not significant.
To confirm the direct interaction between miR-135a and circRNA_010383, the MRE of miR-135a in the luciferase reporter was mutated (Supplementary Fig. 3). However, the cotransfection of the miR-135a mimic and the mutated luciferase reporter did not have a significant effect on the luciferase activity (Fig. 2C). In addition, circRNA_010383 and miR-135a were abundantly pulled down by the biotinylated circRNA_010383 probe in mMCs and mTECs (Fig. 2D and E) upon circRNA_010383 overexpression. Moreover, knockdown of circRNA_010383 did not modulate the expression of miR-135a in mMCs and mTECs (Fig. 2F). Additionally, overexpression of miR-135a had no effects on the expression of circRNA_010383 in mMCs and mTECs (Fig. 2G). Furthermore, we confirmed that high glucose levels downregulated the expression of TRPC1 and induced miR-135a expression in mMCs and mTECs (Fig. 2H and I). These results indicate that circRNA-0101383 may act as a sponge for miR-135a.
circRNA_010383 Inhibits the Accumulation of ECM in Mesangial Cells
The key feature of DN is progressive ECM accumulation in the glomerular mesangium. To analyze the role of circRNA_010383 plays in mesangial cells in DN, two siRNAs targeting the backsplice junction sequence of circRNA_010383 were designed (Fig. 3A). We found that both siRNAs reduced circRNA_010383 expression by 45–50% (data not shown) and did not affect the linear-010383 expression (data not shown). The Western blot analysis showed that the knockdown of circRNA_010383 promotes the synthesis of fibronectin, Col I, and α-SMA in mMCs (Fig. 3B and G). Moreover, Western blot revealed that the knockdown of circRNA_010383 increased the level of PCNA and decreased the TRPC1 level (Fig. 3B, F, and G), and the effect of circRNA_010383 on TRPC1 was confirmed by immunofluorescence (Fig. 3H). Additionally, the transfection of the circRNA_010383 expression vector into mMCs cultured in high-glucose medium abolished the synthesis of fibronectin, Col I, α-SMA, and PCNA induced by the high-glucose levels and restored the expression of TRPC1 (Fig. 3I–O). Furthermore, the linear-010383 expression vector did not have an effect similar to that of the circRNA_010383 expression vector on mMCs cultured in high-glucose medium (Supplementary Fig. 4A–F). Thus, these data suggest circRNA_010383 inhibits the progression of glomerular sclerosis.
circRNA_010383 inhibits the accumulation of ECM and upregulates the expression of TPRC1 in mMCs. A: Schematic model of the siRNAs. Two si-circRNA_010383 constructs targeting the backsplice junction of circRNA_010383 were designed and synthesized. Representative Western blot (B) and quantitative analysis of fibronectin (C), Col I (D), α-SMA (E), TRPC1 (F), and PCNA (G) in mMCs. Each bar represents the mean ± SD (n = 3). *P < 0.05, **P < 0.01, and ***P < 0.001 vs. siRNA-control (si-con). H: Representative confocal microscopy images show the cellular localization of TRPC1 (red) and nuclear staining by DAPI (blue). Representative Western blot (I) and quantitative analysis of fibronectin (J), Col I (K), α-SMA (L), PCNA (M), and TRPC1 (N) of mMCs with circRNA_010383. Data are presented as the mean ± SD (n = 3). **P < 0.01, *P < 0.05 vs. normal glucose+vector; #P < 0.05 vs. high glucose+vector. O: Representative confocal microscopy images show the cellular localization of TRPC1 (red) and nuclear staining by DAPI (blue).
circRNA_010383 inhibits the accumulation of ECM and upregulates the expression of TPRC1 in mMCs. A: Schematic model of the siRNAs. Two si-circRNA_010383 constructs targeting the backsplice junction of circRNA_010383 were designed and synthesized. Representative Western blot (B) and quantitative analysis of fibronectin (C), Col I (D), α-SMA (E), TRPC1 (F), and PCNA (G) in mMCs. Each bar represents the mean ± SD (n = 3). *P < 0.05, **P < 0.01, and ***P < 0.001 vs. siRNA-control (si-con). H: Representative confocal microscopy images show the cellular localization of TRPC1 (red) and nuclear staining by DAPI (blue). Representative Western blot (I) and quantitative analysis of fibronectin (J), Col I (K), α-SMA (L), PCNA (M), and TRPC1 (N) of mMCs with circRNA_010383. Data are presented as the mean ± SD (n = 3). **P < 0.01, *P < 0.05 vs. normal glucose+vector; #P < 0.05 vs. high glucose+vector. O: Representative confocal microscopy images show the cellular localization of TRPC1 (red) and nuclear staining by DAPI (blue).
circRNA_010383 Inhibits the Synthesis of ECM Protein in Renal Tubular Epithelial Cells
Tubulointerstitial fibrosis due to the increased deposition of ECM proteins is another key pathologic feature of DN (23). As shown in Fig. 4A–F, Western blot revealed that knockdown of circRNA_010383 dramatically increased the synthesis of ECM-related proteins, such as fibronectin, Col I, and α-SMA, in renal tubular epithelial cells. In addition, Western blot and immunofluorescence revealed that knockdown of circRNA_010383 significantly decreased levels of E-cadherin and TRPC1 in mTECs (Fig. 4A and E–G). Moreover, overexpression of circRNA_010383 in mTECs abolished the synthesis of fibronectin, Col I, and α-SMA induced by the high glucose and restored the expression of TRPC1 (Fig. 4H–L), while the linear-010383 expression vector did not have an effect similar on mTECs cultured in high-glucose medium (Supplementary Fig. 4G–K). Thus, these data suggest that circRNA_010383 inhibits the progression of tubulointerstitial fibrosis.
circRNA_010383 inhibits the accumulation and synthesis of ECM and upregulates the expression of TPRC1 and E-cadherin in mTECs. Representative Western blot (A) and quantitative analysis of fibronectin (B), Col I (C), α-SMA (D), TRPC1 (E), and E-cadherin (F) in mTECs. Data are presented as the mean ± SD (n = 3). *P < 0.05 and **P < 0.01 vs. siRNA-control (si-con). G: Representative confocal microscopy images show the cellular localization of TRPC1 and E-cadherin and nuclear staining by DAPI. Representative Western blot (H) and quantitative analysis of fibronectin (I), Col I (J), α-SMA (K), and TRPC1 (L) of mTECs with circRNA_010383. Data are presented as the mean ± SD (n = 3). ***P < 0.001, **P < 0.01, *P < 0.05 vs. normal glucose+vector; #P < 0.05 vs. high glucose+vector.
circRNA_010383 inhibits the accumulation and synthesis of ECM and upregulates the expression of TPRC1 and E-cadherin in mTECs. Representative Western blot (A) and quantitative analysis of fibronectin (B), Col I (C), α-SMA (D), TRPC1 (E), and E-cadherin (F) in mTECs. Data are presented as the mean ± SD (n = 3). *P < 0.05 and **P < 0.01 vs. siRNA-control (si-con). G: Representative confocal microscopy images show the cellular localization of TRPC1 and E-cadherin and nuclear staining by DAPI. Representative Western blot (H) and quantitative analysis of fibronectin (I), Col I (J), α-SMA (K), and TRPC1 (L) of mTECs with circRNA_010383. Data are presented as the mean ± SD (n = 3). ***P < 0.001, **P < 0.01, *P < 0.05 vs. normal glucose+vector; #P < 0.05 vs. high glucose+vector.
circRNA_010383 Inhibits the Accumulation of ECM by Targeting miR-135a and TRPC1
To assess whether circRNA_010383 inhibits the accumulation of ECM synthesis by targeting miR-135a, the circRNA_010383 expression vector and miR-135a mimic were cotransfected into mMCs cultured in high-glucose medium. Western blotting revealed that the miR-135a mimic significantly promoted fibronectin, Col I, α-SMA, and PCNA synthesis, inhibited TRPC1 expression, and abolished the protective effect of circRNA_010383 overexpression in mesangial cells (Supplementary Fig. 5). Moreover, the circRNA_010383 expression vector and TRPC1 siRNA were cotransfected into mMCs and mTECs cultured in high-glucose medium. As shown in Fig. 5, knockdown of TRPC1 significantly promoted ECM proteins and abrogated the circRNA_010383- mediated reduction of ECM proteins in mMCs and mTECs. These results indicate that circRNA_010383 inhibits ECM accumulation in mMCs and mTECs by targeting miR-135a and TRPC1.
circRNA_010383-induced inhibition of the synthesis of ECM proteins is reversed by the knockdown of TRPC1. The mMCs and mTECs cultured in high-glucose medium were cotransfected with a circRNA_010383 vector or empty vector and siRNA-control (si-con) or TRPC1 siRNA. Representative Western blots (A) and quantitative analysis of TRPC1 (B), fibronectin (C), Col I (D), α-SMA (E), and PCNA (F) of mMCs. Representative Western blots (G) and quantitative analysis of TRPC1 (H), fibronectin (I), Col I (J), and α-SMA (K) of mTECs. Data are presented as the mean ± SD (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 vs. si-con+vector; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. si-TRPC1+vector; &&P < 0.01, &P < 0.05 vs. si-con+circRNA_010383.
circRNA_010383-induced inhibition of the synthesis of ECM proteins is reversed by the knockdown of TRPC1. The mMCs and mTECs cultured in high-glucose medium were cotransfected with a circRNA_010383 vector or empty vector and siRNA-control (si-con) or TRPC1 siRNA. Representative Western blots (A) and quantitative analysis of TRPC1 (B), fibronectin (C), Col I (D), α-SMA (E), and PCNA (F) of mMCs. Representative Western blots (G) and quantitative analysis of TRPC1 (H), fibronectin (I), Col I (J), and α-SMA (K) of mTECs. Data are presented as the mean ± SD (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 vs. si-con+vector; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. si-TRPC1+vector; &&P < 0.01, &P < 0.05 vs. si-con+circRNA_010383.
circRNA_010383 Gene Therapy In Vivo
As previously described (24), a well-established ultrasound-microbubble-mediated gene transfer technique was used to deliver the circRNA_010383 expression plasmid into the kidney to determine the role circRNA_010383 plays in DN in db/db mice (Fig. 6A). We first examined the efficiency of ultrasound-microbubble-mediated circRNA_010383 transgene expression in the kidney. The results revealed that ultrasound-mediated circRNA_010383 transfer markedly restored the levels of circRNA_010383 in db/db mice (Fig. 6B). Moreover, the upregulation of circRNA_010383 expression restored TRPC1 mRNA expression in db/db mice (Fig. 6B), indicating that circRNA_010383 can act as a miR-135a sponge in vivo. The db/db mice developed similar levels of hyperglycemia and similar body weights regardless of the treatment, indicating that the delivery of circRNA-0100383 did not affect blood glucose levels or body weights (Supplementary Fig. 6). Pathologically, the mesangial matrix and thickness of the glomerular basement membrane were significantly increased in db/db mice and were reduced in db/db mice treated with the circRNA_010383 expression plasmid (Fig. 6C–E). We did not detect the effect of circRNA_010383 on podocytes, abundance, or injury. Additionally, db/db mice treated with the circRNA_010383 expression plasmid developed less severe microalbuminuria than db/db mice over the 20-week disease course (Fig. 6F). These results suggest that the in vivo restoration of circRNA_010383 expression improves renal function in diabetic mice. Western blotting revealed that the restoration of circRNA_010383 levels in vivo reduced PCNA, fibronectin, α-SMA, and Col I expression in the diabetic mouse kidneys (Fig. 7A–E). Moreover, Western blotting and immunofluorescence revealed that the restoration of circRNA_010383 levels in vivo restored the levels of TRPC1 protein (Fig. 7A, F, and G). Therefore, the restoration of circRNA_010383 expression attenuated the progression of renal fibrosis as demonstrated by the decreased synthesis and deposition of ECM.
In vivo expression of circRNA_010383 ameliorates the progression of DN. A: Schematic model of the overexpression of circRNA_010383 in vivo. B: Real-time PCR showing that ultrasound-microbubble-mediated circRNA_010383 expression vector treatment, compared with empty vector control treatment, effectively upregulates circRNA_010383 expression and TRPC1 expression in the kidneys of db/db mice at the age of 20 weeks. C: PAS staining and Sirius red staining showing changes in renal histology after the overexpression of circRNA_010383 in db/db mouse kidneys. D and E: Quantification analysis of the mesangial matrix index by PAS staining and renal fibrosis by Sirius red staining in mouse kidneys at the age of 20 weeks. F: Urinary albumin-to-creatinine ratio (ACR) in circRNA_010383-treated db/db mice, db/db mice, and db/m littermates was measured at the indicated weeks of age. Data are shown as mean ± SD (n = 6–8 per group). *P < 0.05, **P < 0.01, ***P < 0.001 vs. db/m mice at the same time point; &P < 0.05, &&P < 0.01, &&&P < 0.001 vs. db/db mice at the same time point. w, weeks.
In vivo expression of circRNA_010383 ameliorates the progression of DN. A: Schematic model of the overexpression of circRNA_010383 in vivo. B: Real-time PCR showing that ultrasound-microbubble-mediated circRNA_010383 expression vector treatment, compared with empty vector control treatment, effectively upregulates circRNA_010383 expression and TRPC1 expression in the kidneys of db/db mice at the age of 20 weeks. C: PAS staining and Sirius red staining showing changes in renal histology after the overexpression of circRNA_010383 in db/db mouse kidneys. D and E: Quantification analysis of the mesangial matrix index by PAS staining and renal fibrosis by Sirius red staining in mouse kidneys at the age of 20 weeks. F: Urinary albumin-to-creatinine ratio (ACR) in circRNA_010383-treated db/db mice, db/db mice, and db/m littermates was measured at the indicated weeks of age. Data are shown as mean ± SD (n = 6–8 per group). *P < 0.05, **P < 0.01, ***P < 0.001 vs. db/m mice at the same time point; &P < 0.05, &&P < 0.01, &&&P < 0.001 vs. db/db mice at the same time point. w, weeks.
In vivo expression of circRNA_010383 ameliorates the renal fibrosis of DN. Representative Western blots (A) and quantitative analysis of renal fibronectin (B), Col I (C), α-SMA (D), PCNA (E), and TRPC1 (F). Data are presented as the mean ± SD (n = 6–8 per group). **P < 0.01 vs. db/m mice; &&P < 0.01, &&&P < 0.001 vs. db/db mice. G: Representative confocal microscopy images show the expression of TRPC1 (red) and nuclear staining by DAPI (blue) on renal tissue of mice.
In vivo expression of circRNA_010383 ameliorates the renal fibrosis of DN. Representative Western blots (A) and quantitative analysis of renal fibronectin (B), Col I (C), α-SMA (D), PCNA (E), and TRPC1 (F). Data are presented as the mean ± SD (n = 6–8 per group). **P < 0.01 vs. db/m mice; &&P < 0.01, &&&P < 0.001 vs. db/db mice. G: Representative confocal microscopy images show the expression of TRPC1 (red) and nuclear staining by DAPI (blue) on renal tissue of mice.
Discussion
Recently, circRNAs have been recognized as important regulators of physiological and pathological processes in human health and disease. In this study, we identified circRNA_010383 as an important signature circRNA in DN. The expression of circRNA_010383 was found to be markedly downregulated in the kidneys of diabetic mice, mesangial cells, and tubular epithelial cells cultured in high-glucose conditions, and this downregulated expression was associated with the development of proteinuria and renal fibrosis. More importantly, our study results showed that the in vivo overexpression of circRNA_010383 ameliorated the progression of DN in db/db mice. Mechanistically, circRNA_010383 blocks the progression of DN by acting as a miR-135a sponge. The overexpression of circRNA_010383 may represent an effective therapeutic approach for DN.
Numerous studies have shown that circRNAs potentially exert their effects by acting as miRNA sponges. A circRNA named ciRS-7 was first reported to function as a sponge of miR-7, increasing the levels of miR-7 targets (13). Sex-determining region Y, a testis-specific circRNA, has also been reported to act as a sponge for miR-138 (13). circRNA AKT3 was found to act as a sponge for miR-296-3p and inhibit cancer metastasis (25). Recently, evidence suggesting circRNAs play a role in DN has begun to increase. circRNA LRP6 was found to be a sponge for miR-205 and to regulate high glucose-induced ECM accumulation in mesangial cells (26). In addition, circRNA_15698 aggravates the ECM accumulation in mesangial cells in DN via miR-185/transforming growth factor-β1 (27). These findings are indicative of the important roles circRNAs play in DN. In the current study, we found that circRNA_010383 functions as a decoy of miR-135a to regulate TRPC1 expression and inhibit ECM accumulation induced by high-glucose conditions in mesangial cells and tubular epithelial cells. More importantly, the current study also provides evidence that the overexpression of circRNA_010383 inhibits proteinuria and renal fibrosis in vivo.
We identified the biological function of circRNA_010383 in the development of DN and found that the knockdown of circRNA_010383 promotes the accumulation of ECM components, including fibronectin, Col I, and α-SMA in cultured mesangial cells and tubular epithelial cells. Moreover, the overexpression of circRNA_010383 effectively inhibits the high glucose-induced accumulation of ECM. Furthermore, we demonstrated that the overexpression of circRNA_010383 in diabetic kidneys restores TRPC1 levels and confirmed that the overexpression of miR-135a in kidney cells promotes renal fibrosis, as observed by circRNA_010383 knockdown. These findings are significant, because during DN, renal miR-135a expression is upregulated, and gene therapy with chemically modified nucleic acid oligonucleotides complementary to the mature miR-135a sequence ameliorates proteinuria and suppresses renal fibrosis (6). Studies of miR-135a suggest that it plays an important role in diabetes and associated complications. Increased miR-135a levels have also been reported in human diabetic gastrocnemius skeletal muscle, and in vivo miR-135a silencing alleviates hyperglycemia and improves glucose tolerance (28). A recent study showed that miR-135a promotes inflammatory responses in vascular smooth muscle cells from db/db mice (29). We previously showed that the downregulation of miR-135a expression can upregulate TRPC1 expression (6), which is repressed in the kidneys of DN patients (6,30) and that the renal expression of TRPC1 is downregulated in animal models of DN (30,31). Strikingly, a study found that TRPC1 is located on human chromosome 3q22-24 (i.e., a region considered to be a hotspot for DN) (30), and TRPC1 genetic polymorphisms are associated with DN in patients with type 2 diabetes in the Han Chinese population (32). Notably, in the current study, we show that the renal fibrosis induced by diabetic conditions associated with circRNA_010383 overexpression can be reversed by the restoration of miR-135a expression or knockdown of TRPC1. circRNA_010383 functions as a miR-135a sponge to regulate TRPC1 levels and may be a novel factor in the mechanism underlying DN.
However, the precise mechanism by which circRNA_010383 affects fibrosis may be related to other target genes of miR-135a and additional putative miRNAs and their targets. Studies of miR-135a have shown that it regulates multiple target genes, including SIRT1 (33,34), HIF-1α (35), SMAD5 (36), and protein tyrosine phosphatase receptor delta (PTPRD) (37). We also predicted by MRE analysis that miR-135b superfamily members might be potential circRNA_010383 miRNA targets. miR-135b expression is upregulated in models of podocyte injury and in glomeruli isolated from patients with focal segmental glomerulosclerosis (38). The ectopic expression of miR-135b leads to severe podocyte injury and the dysregulation of the podocyte cytoskeleton (38). Consequently, circRNA_010383 might regulate the expression of miR-135 members to affect the development of DN.
More importantly, the current study also provides evidence that the overexpression of circRNA_010383 can inhibit proteinuria and renal fibrosis in vivo. These results suggest that circRNA_010383 may be an effective therapeutic agent for DN. In conclusion, in the current study, we identified that circRNA_010383 expression is downregulated in the diabetic conditions of both a DN mouse model and mMCs and mTECs cultured in high-glucose medium. circRNA_010383 negatively regulates the high glucose-induced accumulation of ECM and renal fibrosis by acting as a sponge of miR-135a in vitro and in vivo. The overexpression of circRNA_010383 may represent an effective therapeutic approach for DN.
This article contains supplementary material online at https://doi.org/10.2337/figshare.13177115.
F.P., W.G., and S.L. contributed equally to this study.
Article Information
Acknowledgments. The authors thank Prof. Xueqing Yu, Prof. Jinjin Fang, and Ning Luo from The First Affiliated Hospital Sun Yat-Sen University for their suggestions during this work.
Funding. This work was supported by the National Natural Science Foundation of China (NSFC, no. 81600624, no. 81673792, no. 81873346, no. U1801288, no. 81900607, and no. 82071563), the Natural Science Foundation of Guangdong Province, China (no. 2017A030313708 and no. 2014A030310065), the Science and Technology Planning Project of Guangdong Province, China (no. 2017A020215158), and the Science and Technology Planning Project of Guangzhou, China (no. 201707010286).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. F.P. designed the experiments, analyzed and interpreted the data and wrote the paper. W.G., S.L., B.Y., and C.Z. conducted the experiments and analyzed the data. W.L., X.C., C.L., Q.H., T.C., L.S., S.F., and W.Z. interpreted the data and offered advice. Z.L. and H.L. provided reagents, supervised all parts of the study, and/or edited the paper. Z.L. and H.L. are the guarantors of this work and, as such had full access to all the data in the study and take responsibility for the data and the accuracy of the data analysis.