Current clinical insights into circRNAs and signal transduction in diabetic nephropathy

1 Introduction

Diabetic mellitus, a metabolic disorder, is generally associated with other chronic disorders like peripheral neuropathy, retinopathy, and nephropathy. Diabetic nephropathy (DN) is a critical clinical complication in the kidney that results from diabetes. It has been reported that chronic kidney disease (CKD) or end-stage renal failure occurs due to DN around the world. In the US, one in three people with diabetes suffers from DN [1]. Globally, DN occurs in 30%–40% of patients diagnosed with diabetes [2]. Some of the major causes of DN include disorders of glucose metabolism and oxidative stress. Renal fibrosis, extracellular matrix accumulation, and proteinuria are some of the clinical characteristics associated with DN [3].

Circular ribonucleic acids (circRNAs) are closed loop non-coding RNAs formed by backsplicing. Gene regulation can be modulated by circRNAs by sponging microRNAs (miRNAs) and regulating nuclear transcription [4]. Hence, there has been a resurgence of interest in understanding circRNAs that are involved in the pathogenesis of numerous diseases. The characteristics of circRNAs including wide distribution, stability, and diferential expression suggest their potential use as therapeutic and diagnostic biomarkers in the treatment of various diseases including DN [5].

The pathogenesis of DN is complex, an improved understanding of the molecular basis of this disease may help in faster diagnosis and therapeutic intervention. Reports indicate that vitamin K2-7 may be useful in DN and patients on hemodialysis [6, 7, 8]. Indeed, biological networks, whether directed or undirected, are key to identifying novel biomarkers/targets from big data [9]. In this review, we elucidate a circRNA biological network from available clinical evidence in DN and delineate circRNA-regulated signal transduction pathways in DN. Indeed, the identification of circRNA biomarkers will facilitate early diagnosis and therapeutic intervention in DN.

2 Clinical Evidence for CircRNAs in DN

Several upregulated circRNAs have been reported in clinical studies of DN. However, only one downregulated circRNA has been reported in clinical studies of DN till date. A list of circRNAs involved in the modulation of clinical DN are highlighted in Table 1. Further, a biological network of these circRNAs in clinical DN was generated using bioinformatic tools and is illustrated in Figure 1.

Figure 1

Human circRNA-circRNA network in diabetic nephropathy. A human circRNA-circRNA network comprising upregulated and downregulated circRNAs in clinical studies of diabetic nephropathy with 15 nodes and 210 edges was constructed using Cytoscape 3.10.0.[23]. CircRNA: circular ribonucleic acid.

3 Signal Transduction in DN

In this section, we discuss the potential role of circRNAs involved in regulating key proteins that play significant roles in the pathogenesis of DN. CircRNAs and their targets in regulation of DN as well as associated signaling pathways are depicted in Figure 2 and summarized in Table 2 for the benefit of the reader.

Figure 2

Signal transduction associated with circRNAs involved in diabetic nephropathy. PORCN aids the translocation of Wnt from endoplasmic reticulum to cell membrane where Wnt binds to frizzled and LRP activating Wnt pathway; circ_0008529 and circ_0000064 upregulate Wnt pathway by downregulating miR-485-5p and miR-424-5p respectively. Further, Wnt/Frizzled/LRP complex activates Dishevelled which inhibits GSK-3β and upregulates β-catenin; circ_012448 upregulates GSK-3β by downregulating miR-29b-2-5p; SOX6 inhibits β-catenin, circ_0080425 and circ_0006916 upregulates SOX6 by sponging miR-185-5p and miR-137 respectively. Moreover, Wnt binds to Frizzled and activates NFAT pathway via Calmodulin and Calcineurin; circ_0037128 inhibits miR-497-5p resulting in upregulation of NFAT. In addition, AGE binds to RAGE leading to activation of PI3K/AKT/NF-кB pathway initiating an inflammatory response; circ_0004442 upregulates AKT/NF-кB by sponging miR-126-5p and miR-204-5p; SIRT1 inhibits PGC1alpha which leads to mitochondrial dysfunction and progression of DN; circ_0000284 upregulates SIRT1 by sponging miR-326. Also, SGLT2 transports glucose and sodium to cytoplasm via the SGLT2; circ_000166 upregulates SGLT2 by inhibiting miR-296. Further, TGF-β1 binds to TGF-β receptor that activates Smad pathway; circ_0008529 upregulates Smad2 by inhibiting miR-185-5p. Moreover, ROCK1 and ROCK2 are activated by GPCR and TGFR through Rho-GTP and ANXA2 respectively resulting in activation of Smad, JNK and ERK pathways; circ_0068087 upregulates ROCK2 by inhibiting miR-106a-5p; circ_0003928 upregulates ANXA2 by inhibiting miR-151-3p. ERK activates AP-1 that results in transcription of inflammation and fibrosis-related genes; ERK activates MMP9 and circ_0001162 upregulates MMP9 by inhibiting miR-149-5p. CircRNAs, circular ribonucleic acids; PORCN, Porcupine; LRP, low-density lipoprotein receptor related protein; GSK-3β, glycogen synthase kinase-3β; CK1, casein kinase 1 enzyme; APC, adenomatous polyposis coli; SOX6, SRY-Box transcription factor 6; IP3, inositol trisphosphate; NFAT, nuclear factor of activated T cells; PGC1α, peroxisome proliferator-activated receptor-gamma coactivator-1α; SIRT1, silent information regulator sirtuin 1; AGE, advanced glycation end product; RAGE, receptor for AGE; PI3K, phosphoinositide 3-kinase; NF-кB, nuclear factor κB; SGLT2, sodium glucose cotransporter; TGF-β1, transforming growth factor-β1; TGFBR, TGF-β receptor; Smad2, suppressor of mothers against decapentaplegic 2; LPA, lysophosphatidic acid; GPCR, G-protein-coupled receptor; GDP, guanosine diphosphate; GEF, guanine-nucleotide exchange factor; GTP, guanosine triphosphate; ROCK1, Rho associated coiled-coil containing protein kinase 1; MEK, mitogen-activated protein kinase; JNK, c-Jun N-terminal kinase; ERK, extracellular signal-regulated kinase; MMP9, matrix metallopeptidase 9; TCF, T-cell factor; ANXA2, AnnexinA2; ICAM-1, intercellular adhesion molecule-1; AP-1, activator protein 1.

Liu et al.[24] reported the overexpression of circular homeodomain-interacting protein kinase 3 (circHIPK3, hsa_circ_0000284) in high glucose (HG)-treated human podocyte cell line (HPC). It was also found that circHIPK3 promotes the growth of fused in sarcoma (FUS) on the ectodysplasin A2 receptor (EDA2R) promoter, upregulated the expression of the EDA2R and activated apoptotic signaling. It was concluded that circHIPK3/FUS/EDA2R can be used as potential targets for diabetic kidney disease. Further, a study performed by Wang et al.[25] in HG-treated Human kidney cells 2 (HK-2 cells) demonstrated the upregulation of circ_0008529. It was concluded that suppression of circ_0008529 inhibited inflammatory injury by targeting miR-485-5p/Wnt2B pathway in HK-2 cells. Hence, circ_0008529 plays a significant role in DN progression. In addition, Chen et al.[26] studied the role of circ_000166 in DN renal fibrosis in HG-induced HK-2 cells which showed upregulation of circ_000166. Also, it was demonstrated that miR-296 can be regulated by circ_000166 as competing endogenous RNA (ceRNA) resulted in overexpression of sodium-glucose cotransporter 2 (SGLT2). Therefore, suppression of circ_000166 can lead to a decrease in DN by regulating circ_000166/miR-296/SGLT2 signaling pathway.

Ye et al.[27] demonstrated upregulation of hsa_circ_0001162 in HG-treated HPC. In addition, mechanistic studies revealed that hsa_circ_0001162 upregulated matrix metallopeptidase 9 (MMP9) by inhibiting miR-149-5p thus promoting HG-induced podocyte injury. Therefore, it was concluded that silencing of hsa_circ_0001162 can be a potential therapeutic target in DN. Interestingly, Song et al.[28] reported overexpression of hsa_circRNA_012448 in HG-treated HK-2 cells. Also, suppression of hsa_miR-29b-2-5p and over-expression of glycogen synthase kinase 3β (GSK3β) was observed in HK-2 cells. GSK3β plays a critical role in oxidative stress, leading to the progression of DKD. Further, HG induced HK-2 cells treated with dapagliflozin demonstrated downregulation of GSK3β. It was concluded that DN can be treated by targeting hsa_circRNA_012448/hsa_miR-29b-2-5p/GSK3β pathway.

Liu et al.[4] found upregulation of circ_0003928 and HDAC4 and downregulation of miR-506-3p in HG-treated HK-2 cells. HDAC4 promotes apoptosis and oxidative stress that helps in progression of DN. Silencing of circ_0003928 led to suppression of HDAC4. It was concluded that circ_0003928 sponged miR-506-3p that led to upregulation of HDAC4. Feng et al.[12] found overexpression of circ_0037128 in HG-induced HK-2 cells. Further studies demonstrated that circ_0037128 inhibited miR-497-5p and that miR-497-5p targeted the protein expression of NFAT5. It was concluded that DN can be regulated by miR-497-5p/NFAT5 signaling pathway. In order to determine the role of circ-FBXW12 in DN, a study was performed by Sun et al.[3] in human mesangial cells (HMCs). It was observed that circ-FBXW12 was over-expressed and suppression of circ-FBXW12 decreased cell proliferation, oxidative stress, and extracellular matrix (ECM) production. In addition, it was also demonstrated that the knockdown of circ-FBXW12 led to upregulation of miR-31-5p further suppressing LIN28B. Therefore, it was concluded that miR-31-5p/LIN28B signaling pathway can be a potential target in treating DN.

A study performed by Li et al.[29] found upregulation of circ_0000064 and Wnt2B and downregulation of miR-424-5p in HG-treated HMCs. Also, it was observed that silencing of circ_0000064 suppressed ECM accumulation and inflammatory response in HG-induced HMCs. Further, it was concluded that silencing of circ_0000064 upregulated miR-424-5p led to inhibition of Wnt2B. Similarly, Wang et al.[10] reported overexpression of circ_0000064 in HK-2 cells. It was summarized that circ_0000064 upregulated the expression of ROCK1 by sponging miR-532-3p, circ_0000064 can be used as a potential target for treating DN. Circ_0068087 was upregulated in an experiment carried out by Feng et al.[30] in HG-induced HK-2 cells. Circ_0068087 silencing led to the suppression of fibrosis, inflammation, and oxidative stress. Knockdown of circ_0068087 led to downregulation of ROCK2 by miR-106a-5p suggesting a potential therapy for DN through targeting miR-106a-5p/ROCK2 pathway.

Yun et al.[5] reported overexpression of circ-ACTR2 in HG-treated HMCs. Circ-ACTR2 inhibition led to decrease in oxidative stress, cell proliferation, ECM deposition and inflammation. Also, it was evident that circ-ACTR2 competitively inhibited miR-205-5p, and that miR-205-5p further inhibited HMGA2 which led to progression of DN. Shu et al.[31] found overexpression of circHOMER1 (hsa_circ_0006916) in HG-induced HMCs. It was demonstrated that overexpression of circHOMER1 promoted SRY-Box transcription factor 6 (SOX6) expression by inhibiting miR-137 which led to progression of DN. DN can possibly be treated by regulating miR-137/SOX6 axis. Qui et al.[14] demonstrated upregulation of circTLK1 in HG-induced HMCs. Further, it was observed that overexpression of miR-126-5p and miR-204-5p suppressed oxidative stress, inflammation and ECM deposition. Further, it was concluded that silencing of circTLK1 blocked AKT/nuclear factor kB (NF-κB) pathway by inhibition of miR-126-5p/miR-204-5p. Li et al.[32] reported overexpression of circRNA Williams-Beuren syndrome chromosome region 17 (circ_WBSCR17, hsa_circ_0080425) in HG-treated HK-2 cells. Silencing of circ_WBSCR17 or SOX6 suppressed cell apoptosis, fibrosis and inflammatory response. Moreover, circ_WBSCR17 regulated SOX6 expression by the inhibition of miR-185-5p suggesting a potential target for treating DN.

Dong et al.[11] found overexpression of circNUP98 in HG-treated HMCs. It was also observed that silencing of circNUP98 led to suppression of inflammation, fibrosis and oxidative stress by targeting miR-151-3p-HMGA2 pathway. Thus, circNUP98/miR-151-3p/HMGA2 can be targeted to ameliorate DN. An et al.[2] reported increased levels of hsa_circ_0003928 and ANXA2 along with decreased levels of miR-151-3p in HG-treated HK-2 cells. Suppression of circ_0003928 decreased the levels of interleukin-1β (IL-1β), IL-6, tumor necrosis factor-α (TNF-α), and reactive oxygen species (ROS). Further, it was concluded that hsa_circ_0003928 can inhibit miR-151-3p, and miR-151-3p suppressed ANXA2 suggesting circ_0003928 can be used as a potential target for treating DN. A study performed by Zhuang et al.[33] showed downregulation of circHIPK3 in HG-treated HK-2 cells. Further studies demonstrated that circHIPK3 inhibited miRNA-326 and miR-487a-3p. Also, it was observed that overexpression of miRNA-326 or miR-487a-3p led to downregulation of silent information regulator sirtuin 1 (SIRT1) which eventually led to DN progression. Niu et al.[34] used HG-treated HK-2 cells and observed upregulation of circ_0008529 and suppressor of mothers against decapentaplegic 2 (Smad2), and downregulation of miR-185. It was further concluded that circ_0008529 regulated the expression of Smad2 through miR-185-5p, circ_0008529 can be used as a potential target in DN treatment. Zhou et al.[35] reported overexpression of circ_0060077 in HG-treated HK-2 cells. Knockdown of circ_0060077 led to suppression of oxidative stress, fibrosis, and inflammation. Furthermore, it was observed that silencing of circ_0060077 led to upregulation of miR-145-5p that helped in the recovery of HG-treated HK-2 cells via targeting vasorin (VASN).

4 Conclusions and Future Perspectives

Despite recent advances in the therapeutic management of DN, there are still numerous challenges to be addressed. Augmenting our understanding of the molecular mechanisms of DN and the key roles played by circRNAs in regulating DN is essential to improve the prognosis of DN. Non-coding RNAs including circRNAs involved in DN play critical role in gene regulation by sponging miRNA. Also, circRNAs are involved in various signaling pathways that can be used as potential targets for therapeutic treatment in DN. Numerous studies suggest that various circRNAs including circ_0001831/circ_0000867 [13] and circANKRD36 [20] can be used as diagnostic markers. Epistemological evidence from clinical studies point to a critical role for circRNAs in DN dysregulation. Recent studies demonstrated that various circRNAs including circ_0008529 [25], circ_0001162 [27] and circ_012448 [28] can be used as targets for treatment of DN. There is also a growing body of evidence from in vitro cell line studies and in vivo animal experiments that underscore the importance of circRNAs in DN.

CircRNAs are comparatively more stable than linear RNAs that makes them less resistant to degradation. Moreover, they are specific and abundantly available in tissues and biological fluids that renders them useful as biomarkers. However, there are also limitations including lack of clarity in understanding the exact mechanisms of circRNAs involved in pathogenesis of DN. Also, the degradation mechanisms of circRNAs have not yet been characterized. Indeed, there is limited clinical evidence till date of circRNAs used as biomarkers as there are various factors to be considered like patients' disease condition, comorbidities and racial/ethnic diversity. Further, interaction of circRNAs with non-coding RNAs other than miRNAs remain to be explored [38].

However, clinical evidence suggests that modulation of circRNAs plays important roles in the pathogenesis of DN. Our circRNA biological network based on clinical evidence may provide a proof-of-concept circRNA signature for theranostic intervention in DN. Moreover, our discussion on circRNA associated signal transduction provides a summary of potential targets for therapeutic intervention in DN. Indeed, exploiting our current knowledge of circRNAs in DN dysregulation may throw open newer avenues for faster diagnosis and appropriate therapeutic intervention for DN patients and prevent/delay progression to end-stage kidney disease.