Therapeutic potential for renal fibrosis by targeting Smad3-dependent noncoding RNAs

Abstract

Renal fibrosis is a characteristic hallmark of chronic kidney disease (CKD) that ultimately results in renal failure, leaving patients with few therapeutic options. TGF-β is a master regulator of renal fibrosis and mediates progressive renal fibrosis via both canonical and noncanonical signaling pathways. In the canonical Smad signaling, Smad3 is a key mediator in tissue fibrosis and mediates renal fibrosis via a number of noncoding RNAs (ncRNAs). In this regard, targeting Smad3-dependent ncRNAs may offer a specific therapy for renal fibrosis. This review highlights the significance and innovation of TGF-β/Smad3-associated ncRNAs as biomarkers and therapeutic targets in renal fibrogenesis. In addition, the underlying mechanisms of these ncRNAs and their future perspectives in the treatment of renal fibrosis are discussed.

Graphical abstract

Lan and colleagues find that TGF-β/Smad signaling, specifically Smad3 and Smad3-dependent ncRNAs, including miRNAs and lncRNAs, is a key pathway leading to renal fibrosis. Thus, specifically targeting Smad3 or Smad3-regulated ncRNAs related to fibrosis may represent a promising therapy for CKD.

Introduction

Chronic kidney disease (CKD) is a life-threatening disease with increasing morbidity and mortality, which eventually leads to the development of end-stage kidney disease.¹ Fibrosis is a common pathological feature of CKD and is mediated by many cellular and molecular mechanisms.² However, effective therapy for renal fibrosis remains to be developed. This is largely due to the complexity and heterogeneity of fibrogenesis in CKD patients.³ Recent studies indicate that the use of functional RNA molecules, which specifically target the fibrotic signaling pathways, may offer a promising therapeutic strategy for halting the progression of renal fibrosis.⁴

Transforming growth factor β (TGF-β) is the key cytokine that regulates the fibrogenic process directly or indirectly through TGF-β type I, II, and III receptors.⁵ Of these, both type I and II receptors are essential in the signal transduction of TGF-β, whereas type III receptor (betaglycan and endoglin) is the co-receptor that facilitates the affinity of ligand binding to the type II receptor.^6,7 As a ligand, TGF-β includes three mammalian isoforms, TGF-β1, TGF-β2, and TGF-β3. The three TGF-β isoforms share similarities in sequence and structure but perform distinct functions in regulating fibrogenesis.^8,9,10 Of these, TGF-β1 has been well studied and is considered the most important growth factor in renal fibrosis.¹¹ TGF-β1 is highly expressed in various fibrotic kidney diseases, including diabetic nephropathy (DN), hypertensive nephropathy, obstructive kidney disease, autosomal dominant polycystic kidney disease, immunoglobulin A nephropathy, crescentic glomerulonephritis, and focal segmental glomerulosclerosis (FSGS), and correlates with progressive renal fibrosis.¹² However, the exact mode of TGF-β1-mediated renal fibrosis remains largely unclear.

It is now well recognized that the secreted TGF-β1 is inactive as a latent form of TGF-β1 that consists of two polypeptides (the latency-associated peptide and the latent TGF-β binding protein). Once released under various pathological conditions, it becomes active and binds to the TGF-β receptors to trigger the downstream signaling cascade to exert its general biological effects.¹³

To some extent, the inhibition of activated TGF-β1 type I and II TGF-β receptors seems to be a useful strategy to inhibit renal fibrosis. However, due to the multiple roles and complexity of TGF-β signaling, global knockout (KO) of TGF-β1 or receptors may result in unintended consequences. Evidence from recent clinical trials shows a disappointing outcome. The use of neutralizing antibodies against TGF-β1 fails to ameliorate the proteinuria or estimated glomerular filtration rate due to their low efficacy.^14,15 Given that TGF-β also plays an essential role in regulating immune and inflammatory responses, homozygous TGF-β1 null mice could develop early-onset multiorgan-related inflammation.¹⁶ Moreover, the homozygous Tgfbr1 null mice also die at mid-gestation with abnormal angiogenesis.¹⁷ These findings suggest that targeting the upstream TGF-β signaling may not be a good antifibrosis strategy due to the pleiotropic actions of the TGF-β signaling across the multiple biological process.

It is well documented that TGF-β1 can bind to its receptor II to trigger the canonical signaling by phosphorylating the mother against decapentaplegic (Smad) proteins, including Smad2 and Smad3. After being phosphorylated, Smad2 and Smad3 can form the complex with a common Smad 4 and then subsequently translocate into the nucleus and stimulate transcription of fibrotic target genes.^11,13,18 In addition, the noncanonical TGF-β/ALK1/Smad1/5 signaling plays a critical role in fibrogenesis by activating ERK, JNK, p38, PI3K/Akt, and JAK2/STAT3 to regulate gene expression.^19,20 Under diseased conditions, many mediators such as angiotensin II (Ang II), advanced glycation end products (AGEs), oxidative stress, and inflammatory cytokines could activate Smad2/3 via TGF-β-independent mechanisms.^21,22 Multiple intracellular signaling pathways can regulate fibrosis, which could be a potential explanation for the lack of success in clinical trials that focused on targeting the upstream of TGF-β signaling using the antibody strategy.

Increasing evidence shows that noncoding RNAs (ncRNAs) may participate in inflammation and fibrosis due to their physicochemical and physiological properties.^23,24,25,26 ncRNAs are functional RNA molecules that do not encode proteins but produce noncoding transcripts that regulate complex biological processes through epigenetic mechanisms.²⁴ Recent studies also demonstrated that TGF-β may act via ncRNAs to mediate renal fibrosis, and targeting these ncRNAs may offer a more specific and effective therapy for renal fibrosis. Thus, it is likely that TGF-β/Smad signaling may mediate renal fibrosis via ncRNAs-dependent mechanisms, and targeting these Smad-dependent ncRNAs may represent a promising therapy for CKD.

Pathological characteristics and animal models of renal fibrosis

Renal fibrosis is characterized by the loss of nephrons accompanied by the accumulation of the extracellular matrix (ECM) and ECM-producing cells such as fibroblasts and myofibroblasts with progressive loss of renal function. However, mechanisms that regulate renal fibrosis remain unclear. Thus, understanding the mechanisms of renal fibrogenesis is the first step toward the development of effective therapeutic strategies clinically. The kidneys are vulnerable to a wide range of insults such as toxins, ischemia/reperfusion, and inflammation.^27,28 The process of wound healing triggers repair from the injuries. If the insults are prolonged, the balance between ECM deposition and degradation is severely hampered and fibrosis develops. Moreover, unresolving injuries can induce renal intrinsic cells to release proinflammatory/fibrotic cytokines and chemokines to promote fibrosis. Finally, excessive ECM accumulation in glomerulus and tubulointerstitial area can also cause progressive renal functional injuries, eventually resulting in renal failure.²⁹

To understand the process and mechanisms of renal fibrosis, the use of appropriate animal models is important. In the current experimental system, in vivo models of renal fibrosis can be classified into surgical, nephrotoxic, physical, and genetically modified models.³⁰ For example, the unilateral ureteral obstructive (UUO) rodent model is the most widely used surgical model for studying progressive renal fibrosis.³¹ In this model, one of two ureters is ligated and the urine accumulates within the kidney, leading to remarkable renal hemodynamic and metabolic disorders. Pathologically, cell apoptosis or necrosis of tubular cells and the infiltration of macrophages can be observed. Tubular epithelial cells, endothelial cells, and pericytes can also differentiate into ECM-producing fibroblasts and α -smooth muscle actin (α-SMA) myofibroblasts.³² However, no significant renal dysfunction could be observed systemically because the opposite sham-operated kidney can compensate for the loss of UUO kidney function.

The 5/6 nephrectomy and ischemia-reperfusion (IR) model are well-established surgical models to study the progression of kidney disease both pathologically and functionally.^33,34,35 The IR model is commonly used for studying the disease transition from acute renal injury to CKD with progressive renal fibrosis.

As for the nephrotoxic CKD models, renal inflammation and fibrosis can be induced by several chemicals such as adriamycin, folic acid, uranyl nitrate, and cyclosporine A.^36,37,38,39 In the physical-induced injury model, radiation treatment of a local dose of 10 Gy may induce renal injury and thus promotes interstitial fibrosis and inflammation.^40,41

The genetically modified rodent models have been well established for studying the functional role and mechanisms of the molecules or genes that are involved in the pathogenesis of renal fibrosis. By overexpressing or deleting the specific gene either globally or locally from the kidney, the role and mechanisms of the target gene in kidney disease can be determined with an appropriate animal model. For instance, mice overexpressing TGF-β1 or connective tissue growth factor (CTGF) are vulnerable to developing renal fibrosis following injury under a variety of animal models.^42,43,44 In addition, angiotensin receptor (type I and II)–deficient mice are protective in renal fibrosis when induced by Ang II.^45,46

It should be pointed out that the process of renal fibrosis is complicated and varied between models. However, they offer significant insights into studying fibrogenesis in kidney disease.

Role of TGF-β/Smad signaling in renal fibrosis

Although targeting the upstream of TGF-β signaling by using neutralizing antibodies against TGF-β1 fails to treat patients with FSGS and DN,^14,15 it is well established that the canonical TGF-β/Smad signaling plays a key role in renal fibrosis.¹¹ Thus, identification of the downstream molecules or genes of TGF-β/Smad signaling that specifically mediate renal fibrosis is urgently needed. Under disease conditions, intracellular phosphorylation of Smad2 and Smad3 is crucial in the fibrotic process. Although both Smad proteins share a similar structure, they play a distinct role in tissue fibrosis. Smad3 binds directly to DNA via its MH1 domain, whereas Smad2 binds indirectly to transcriptional factors.⁴⁷ It is well accepted that among the TGF-β/Smad signaling, Smad3 is pathogenic, whereas Smad2 and Smad7 are protective.¹¹ Thus, Smad molecules tightly control the canonical TGF-β signaling through various signaling cascades to either positively or negatively regulate fibrosis. It is worth noting that the regulatory roles of these Smads in renal fibrosis are distinct.

Among the Smad molecules, Smad3 has been well studied and is a key mediator of renal fibrosis. TGF-β1 induces collagen-related genes via a Smad3-dependent mechanism, thereby leading to renal fibrosis.^47,48 Functionally, Smad3-KO mice are protected from progressive renal fibrosis, including the epithelial-mesenchymal transition (EMT), ECM accumulation, and matrix metalloproteinase-1 activities in UUO model.^49,50 Furthermore, Smad3 activation is also a hallmark in patients with hypertensive nephropathy and DN. Under hypertensive and diabetic conditions, Ang II and AGEs can activate Smad3 directly and indirectly via both TGF-β and ERK/p38 MAPK-Smad cross-talk pathways.^51,52 Thus, mice lacking Smad3 are protected against progressive renal fibrosis in hypertensive and DN.^53,54 A recent study also detected that Smad3 is a key regulator in islet β cell development and insulin synthesis and secretion because diabetic db/db mice lacking Smad3 are protected islet β cells from diabetic injury without evidence of insulin resistance and type 2 diabetes and DN.^55,56 Despite Smad3 deficiency appearing to inhibit renal fibrosis, Smad3-KO mice have been found to be vulnerable, with defects in mucosal immunity, cartilage, and bones.^57,58 Therefore, it raises a concern that directly targeting Smad3 may inhibit renal fibrosis while promoting renal inflammation and immune injury in immunologically mediated kidney diseases.

Due to the fatal deficiency of animal models in Smad2-KO mice,^59,60 the function of Smad2 remains unclear. However, the finding that conditional KO of Smad2 from tubular epithelial cells promotes Smad3-mediated renal fibrosis indicates that Smad2 may protect against renal fibrosis.⁶¹ Recent studies show that Sirt6 interacts directly with Smad2 to inhibit its phosphorylation and nuclear localization and hepatic fibrosis,⁶² which highlights the need for further research into the role of Smad2 in tissue fibrosis.

Smad4 has been shown to play a role in TGF-β and bone morphogenetic protein signaling because it binds to Smad2/3 or Smad1/5 complexes to facilitate their nuclear translocation and signaling transcription. One study reported that the Smad3/Smad4/CDK9 complex promotes renal fibrosis in a UUO model,⁶³ but the deficiency of Smad4 can reduce fibrosis by disengaging TGF-β signaling.^64,65 Nonetheless, recent studies revealed that Smad4 is also essential for the development of the neonatal medulla and ureter differentiation during mouse embryogenesis, indicating its diverse role in renal fibrosis. Conditional KO of Smad4 in mesangial cells inhibits COL1A1 expression rather than fibronectin.⁶⁶ This finding is consistent with the observation that conditional deletion of Smad4 can inhibit renal fibrosis by blocking Smad3 binding to collagen promoters.⁶⁷ Similarly, deletion of Smad4 using small interfering RNAs (siRNAs) also inhibits α-SMA⁺ myofibroblasts in a mouse model of folic acid–induced renal fibrosis.⁶⁸ However, the loss of Smad4 may impair the Smad7-mediated suppression of nuclear factor κB (NF-κB)-driven inflammation,⁶⁷ suggesting that Smad4 has dual functions in renal inflammation and fibrosis and may not be a promising therapeutic target for kidney disease.

Smad7 is a negative feedback molecule of TGF-β signaling. On the one hand, it can bind to TGF-β receptor type I (TβRI), thereby inhibiting the recruitment and phosphorylation of R-Smads complexes.⁶⁹ On the other hand, Smad7 acts as an adaptor that mediates the recruitment of E3 ubiquitin ligases to promote TβRI degradation,^70,71 or it may impede the formation of functional Smad-DNA complexes within the nucleus.⁷² It is interesting that Smad7 also functions as an anti-inflammatory mediator by regulating NF-κB-driven renal inflammation. By interacting with NF-κB, Smad7 induces the expression of IκBα (nuclear factor of kappa light polypeptide gene enhancer in B cells inhibitor α), thus inhibiting NF-κB-driven inflammatory responses.⁷³ Smad7 is downregulated in a variety of kidney diseases.⁷⁴ The functional role of Smad7 in CKD has been well established in a number of mouse models in which the deletion of Smad7 promotes, whereas overexpression of renal Smad7 inhibits, TGF-β/Smad3-mediated renal fibrosis and NF-κB-dependent renal inflammation.^{49,75,76,77,78,79,80,81} Thus, Smad7 is protective and may be a promising therapeutic molecule for both renal fibrosis and inflammation.⁸²

Although the diverse functions of Smads at the cellular level have been broadly studied, the specific clefts and pockets for druggable targets in Smad proteins have yet to be discovered.⁸³ Over the past decades, with advancements in transcriptomic and bioinformatic techniques, an increasing number of TGF-β/Smad-related ncRNAs have been discovered and some of their functions and mechanisms have been validated. The RNA-based therapeutics related to the Smads have shown great potential in the treatment of renal fibrosis.

Role of Smad3-dependent micro/RNAs (miRNAs) in renal fibrosis

ncRNAs can be categorized into short ncRNAs (<200 nt) and long noncoding RNAs (lncRNAs) (>200 nt). They can be further categorized as either structural or regulatory ncRNAs.⁸⁴ miRNAs are single-stranded RNAs (ssRNAs) that bind to the 3′ UTR of mRNA, thereby regulating a wide range of biological processes.⁸³ Mechanistic studies have demonstrated that a number of miRNAs may interact with TGF-β/Smad3 signaling during renal fibrosis (Figure 1).

Figure 1.

Regulation of miRNAs in TGF-β/Smad3-mediated renal fibrosis

After binding to its receptors, TGF-β1 activates Smad3 to induce fibrogenic miRNAs while suppressing antifibrotic miRNAs to regulate fibrosis, which is reversed by Smad7.

A number of TGF-β/Smad3-dependent miRNAs have been well described.^11,85 Among them, the antifibrotic miR-29 family members (miR-29a, b, and c) are well recognized. In the fibrotic kidney, miR-29 is lost and has been shown to play a role in the pathogenesis of obstructive nephropathy and DN and peritoneal fibrosis.^86,87,88,89 Indeed, Smad3 can mediate TGF-β1-induced downregulation of miR-29 by binding to the promoter of miR-29. Therefore, restored renal miR-29 by the ultrasound microbubble-mediated kidney-specific overexpressing miR-29b technique is capable of inhibiting renal fibrosis in a variety of mouse models with renal fibrosis.^{11,85,86,87,88,89} All of these studies reveal a protective role of miR-29 in CKD.

In addition, a number of miRNAs such as miR-200, miR-15b, miR-26a, miR-let-7, miR-101, and miR-136 can also interact with TGF-β/Smad3 and inhibit renal fibrosis by targeting TGF-β signaling.^{90,91,92,93,94,95} It has been reported that the miR-200 family is downregulated by TGF-β1/2 and in diabetic kidneys, whereas overexpression of miR-200a suppresses fibrosis and EMT. Mechanistically, miR-200a can directly bind to the 3′ UTR of Tgfb2 to decrease its expression.⁹¹ Thus, the miR-200 family may be protective in DN.

In contrast, TGF-β/Smad3 also induces renal fibrosis via a number of profibrotic miRNAs, including miR-21, miR-192, miR-433, miR-491-5p, and miR-130b.^{96,97,98,99,100,101,102,103,104,105,106} During renal fibrosis, all of these profibrotic miRNAs are upregulated and are induced by TGF-β/Smad3 signaling. Of them, miR-21 has been well studied. Mechanistically, miR-21 is induced by TGF-β1 via Smad3 and mediates renal fibrosis by targeting Smad7 and phosphatase and tensin homolog (PTEN) in a variety of kidney diseases, including obstructive nephropathy and DN.^{95,96,97,98,99} Thus, targeting miR-21 may offer therapeutic potential for renal inflammation and fibrosis.

Role of Smad3-dependent lncRNAs in renal fibrosis

lncRNAs regulate various biological and pathophysiological processes by interacting with chromatin, transcription factors, or other regulators. In particular, kidney-specific lncRNAs are closely linked to cell types and disease stages. So far, a number of Smad3-dependent lncRNAs have been identified and their functional role in renal inflammation and fibrosis has been demonstrated (Figure 2).

Figure 2.

Regulation of lncRNAs in the TGF-β/Smad3-mediated renal fibrosis

Smad3 is a key pathway in the development of CKD by inducing fibrogenic lncRNAs while inhibiting antifibrotic miRNAs.

By performing RNA sequencing on the UUO kidneys of Smad3-WT and -KO mice, we first identified a novel Smad3-dependent lncRNA, Arid2-IR, which is increased in obstructive and diabetic kidneys.¹⁰⁷ Mechanistically, Smad3 can bind to the promoter region of Arid2-IR to promote fibrosis by enhancing the interaction with early growth response protein 1.^107,108 Functionally, the overexpression of Arid2-IR can promote NF-κB-driven inflammation.¹⁰⁹

By using whole-transcriptome RNA sequencing, we further identified the additional Smad3-dependent transcripts in the diabetic kidneys of Smad3-WT db/db, Smad3-KO db/db, and Smad3(+/−) db/db.¹¹⁰ Among them, lncRNA Erbb4-IR is significantly upregulated in obstructive and diabetic kidney and is induced via a Smad3-dependent pathway following TGF-β1 stimulation. Kidney-specific silencing Erbb4-IR can inhibit albuminuria, serum creatinine, and renal fibrosis in db/db mice and is able to block progressive renal fibrosis in a mouse model of UUO.^111,112 Mechanistic studies reveal that Erbb4-IR mediates progressive renal fibrosis by binding to the 3′ region of miR-29b genomic locus to suppress its transcription and by inhibiting Smad7 transcription through binding to its corresponding genomic sequence of 3′ UTR.^111,112 Therefore, Erbb4-IR functions as a trans-regulator to negatively regulate miR-29b and Smad7 in renal fibrosis. Overexpression of Erbb4-IR results in renal fibrosis by suppressing renal miR-29b and Smad7.

We also find that LRNA9884 is another Smad3-dependent lncRNA. LRNA9884 is upregulated in db/db mice and functions to promote renal inflammation by binding to the promoter of MCP-1 gene to enhance its transcription.¹¹³ Thus, silencing LRNA9884 attenuates albuminuria and serum creatinine in diabetic db/db mice.¹¹³

lncRNA lncTSI (TGF-β/Smad3-interacting lncRNA) is another Smad3-dependent lncRNA.¹¹⁴ lncTSI directly binds to the MH2 domain of Smad3 protein to block its interaction with TβRI. Kidney-specific overexpressing lncTSI attenuates renal fibrosis by repressing TGF-β/Smad signaling via a negative feedback mechanism in a mouse model of UUO.¹¹⁴

RNA-based therapeutics and clinical applications

As described above, TGF-β/Smad signaling plays a diverse role in renal inflammation and fibrosis. Lessons learned from Smad3-KO mice also strongly suggest that therapeutic development by targeting TGF-β/Smad3 signaling should aim at the downstream transcriptomes that specifically relate to renal fibrosis without affecting the systemic immune system. The identification of Smad3-dependent ncRNAs that specifically regulate tissue fibrosis may allow us to do so. There are a variety of RNA-based therapeutics, such as antisense oligonucleotides (ASOs), siRNAs, RNA aptamers, short hairpin RNAs, miRNAs, lncRNAs, circular RNAs, and CRISPR genome editing. Thus, treatment for CKD by targeting Smad-dependent ncRNAs that specifically regulate renal fibrosis may be a promising therapeutic approach (Figure 3).

Figure 3.

RNA-based therapeutics in the treatment of renal fibrosis

There are several RNA-based therapies for renal fibrosis, including (A) antisense oligonucleotides; (B) RNAi; (C) RNA aptamers; (D) ncRNAs; (E) CRISPR-Cas gene editing; (F) RNA nanoparticles.

The ASOs are single-stranded DNA that can be synthesized to target any complementary nucleotide sequence of interest RNA. Approved by the US Food and Drug Administration in 1988, the first-generation ASO, fomivirsen, was administered clinically for the treatment of retinitis.¹¹⁵ In the field of kidney disease, the development of TGF-β1/Smad-targeted oligodeoxynucleotide has also been shown to inhibit TGF-β1 and Smad transcription, thereby reducing EMT and renal fibrosis.¹¹⁶

siRNAs are double-stranded RNAs (dsRNAs) that interfere with the expression of specific genes with fully complementary nucleotide sequences. By degrading the mRNA, siRNAs can stop the translation of specific genes. By using synthetic siRNA against Smad3, a study has shown that overexpressing siRNA is able to inhibit endogenous Smad3 and Smad3 signaling.¹¹⁷ SRN-001 is a siRNA drug targeting amphiregulin. Amphiregulin is a downstream gene induced by TGF-β during fibrosis and functions to promote the fibroblast-to-myofibroblast transition. SRN-001 is undergoing a Phase IA clinical trial that is promising for treating patients with lung and kidney fibrosis. This trial was registered at ClinicalTrials.gov (NCT05984992). LEM-S401 is a potential siRNA that has completed a Phase I clinical trial. This trial was registered at ClinicalTrials.gov (NCT04707131). LEM-S401 effectively silences the CTGF gene in TGF-β-induced lung and skin fibrosis in vitro, as well as inhibits the accumulation of ECM in wound healing model in vivo.¹¹⁸ Another potential candidate of RNA-based therapeutics is ND-L02-s0201, a lipid nanoparticle encapsulating siRNA against heat shock protein 47 for the treatment of hepatic fibrosis and idiopathic pulmonary fibrosis (the trials were registered at ClinicalTrials.gov under NCT03241264, NCT01858935, NCT02227459, and NCT03538301).¹¹⁹ These ongoing research studies represent one of the most promising RNA-based therapeutics in treating disease-associated fibrosis.

RNA aptamers are short, single-stranded RNA molecules that can bind specifically to the target molecules. RNA aptamers can be designed to inhibit the activity or function of TGF-β signaling. A novel anti-TGF-β1 RNA aptamer, APT-β1, is able to bind human and mouse active TGF-β1 with high affinity and shows specificity and significantly inhibits TGF-β1-induced cell morphology in a xenograft mouse model of non-small cell lung cancer.¹²⁰ This study has highlighted the inhibitory effect of TGF-β-targeted RNA aptamer on TGF-β signaling.

The CRISPR-associated protein (Cas) systems are widely applied to precisely edit genome sequence and specifically KO or knockin of the target gene.¹²¹ This kind of new system manipulates nucleic acids with different methods. The Cas9 system is able to target dsDNA and ssRNA,¹²² whereas the Cas13 system only targets RNA. The Cas13d is of high efficiency and specificity to knock down the target RNA. By using the Cas13d method, Wang’s group demonstrated the pathogenic role of a Smad3-dependent miRNA, let-7i-5p, in TGF-β1-stimulated renal tubular cells and in the UUO model, indicating that let-7i-5p could be the therapeutic target for CKD-related fibrosis.¹²³

The identification of ncRNAs in the pathogenesis of renal fibrosis has highlighted their potential as therapeutic targets for CKD.¹²⁴ We and other investigators have shown that treatment by targeting Smad3-dependent miRNAs that specifically regulate renal fibrosis is able to attenuate kidney injury (Figure 1). For example, the inhibition of miR-21 can protect against the development of renal fibrosis and improve renal dysfunction in mouse models of diabetes and UUO.^{96,97,98,99,100} miR-192, miR-433, and miR-130b are also Smad3-dependent miRNAs related to renal fibrosis, and the inhibition of these profibrotic miRNAs can significantly inhibit renal fibrosis in a number of animal models, including obstructive nephropathy and DN.^{101,102,103,104,105,106} In contrast, the overexpression of Smad3-dependent miRNAs that are protective for renal fibrosis can also combat renal fibrosis and improve renal dysfunction in DN and obstructive nephropathy.^86,87,88,89 Thus, the inhibition of profibrotic miR-21 and the overexpression of miR-29 could be the most promising antifibrotic approaches. Clinically, an inhibitor of miR-21 (RG-012) has been shown to inhibit renal fibrosis in patients with Alport syndrome.¹²⁵ This study was registered at ClinicalTrials.gov as NCT03373786.Therapeutic delivery of miR-29 mimics (MRG-201, Remlarsen) can inhibit bleomycin-induced pulmonary fibrosis in mice.¹²⁶ A Phase I clinical trial (NCT02603224) has demonstrated the efficacy of the miR-29b mimic (MRG-201) on patients with fibrotic scarring (hypertrophic scar or keloid).¹²⁷ Recently, the miR-29 mimic was further developed into the MRG-229, a next-generation miR-29 mimic, and has been demonstrated to have in inhibitory effect on patients with idiopathic pulmonary fibrosis.¹²⁸ These studies offer promising options for miRNA-based therapeutics in halting the progression of renal fibrosis.

However, there have been clinical concerns regarding the specificity and potential off-target effects of anti-miRNA therapy. This is largely due to the findings that one miRNA can regulate many individual mRNAs and one targeted gene can also be regulated by many miRNAs. In addition, there is concern about the differences between mice and humans in genes regulated by individual miRNAs. Thus, targeting Smad3-dependent lncRNAs may offer more specific antifibrosis therapy (Figure 2). Experimentally, we find that kidney-specific targeting profibrotic and proinflammatory lncRNAs such as Erbb4-IR and LRNA 9884 are capable of inhibiting renal inflammation and fibrosis in mouse models of obstructive nephropathy and DN.^111,112,113 In contrast, the overexpression of lncRNA lncTSI can inhibit progressive renal fibrosis. It should be pointed out that although lncRNAs have been widely explored as significant biomarkers in fibrotic diseases, most of the lncRNA-related therapeutics are in the preclinical stage.¹²⁴ The development of lncRNA-based therapies for fibrosis remains a significant challenge. This is largely attributed to the species specificity, tissue and cell specificity, and disease specificity.^129,130 It has been reported that 30% of lncRNA transcripts are primate specific. Of the lncRNAs, 0.7% are specific to the human lineage, and only ∼1% of lncRNAs are expressed in all of the species analyzed.¹³¹ In contrast to miRNAs, only 11% of lncRNAs can be detected in all human tissues, whereas 21% of lncRNAs are not detected in any human tissue. lncRNA expression is also disease dependent. For example, Arid2-IR is highly expressed in UUO-induced acute renal inflammation, but this is not found in ischemic acute kidney injury.¹⁰⁹ Thus, these lncRNA specificities may also limit the development of lncRNA-based therapies clinically.

In addition, the RNAs delivery efficiency into the specific organ or cells remains challenging clinically. Although improved delivery systems such as lipid nanoparticles, polymers, viral or bacterial vector systems, and exosomes can significantly improve the target delivery, it remains of low efficacy. In this regard, we have established a noninvasive ultrasound system to deliver lipid microbubble–containing specific genes into the kidneys. Lipid microbubbles containing the target ncRNA vector can be injected into the mice via the tail vein, followed by placing the ultrasound probe on the back over the kidneys with a pulse-wave output of 1 MHz at 2 W/cm² output for several minutes (the duration time depends on different disease models).¹³² By using this technique, we can specifically transfect Smad7-expressing plasmid,^75,76,77,81 miRNAs (miR-21, miR-29b, miR-192, or miR-433),^{86,87,88,96,97,104} and lncRNAs (Arid2-IR, Erbb4-IR, or LRNA9884)^{109,110,111,112,113,114} into the kidneys to effectively treat the kidney diseases, including obstructive nephropathy and DN.

Small molecules and posttranslational modification by targeting Smad3

The druggable proteins that specifically target TGF-β/Smad signaling by small molecules (i.e., Smad3 inhibitor) can be an alternative approach for the treatment of renal fibrosis. Small molecules can effectively penetrate cell membranes and interact with target proteins within the cell. In this respect, small molecules targeting Smad3 are designed to inhibit the activation of Smad3 and its downstream signaling and represent an effective treatment for CKD by targeting TGF-β/Smad3 signaling (Figure 4).

Figure 4.

Treatments for renal fibrosis by targeting Smad3

There are a number of therapeutic approaches for renal fibrosis, including a variety of Smad3 inhibitors, posttranslational modifications, and RNA-based therapeutics.

SIS3 is the specific inhibitor of Smad3 that reduces the phosphorylation of Smad3, ameliorates renal inflammation, podocyte injury, and inhibits EMT and macrophage-to-myofibroblast transition in kidney disease.^{133,134,135,136,137} However, the poor water solubility of SIS3 limits its clinical application. Researchers discovered a novel water-soluble compound, which effectively blocks the phosphorylation of Smad3 in natural killer cells and could serve as an alternative treatment for renal fibrosis.¹³⁷ Moreover, halofuginone, a competitive prolyl-tRNA synthetase inhibitor, decreases Col I synthesis by inhibiting Smad3 phosphorylation, thereby improving renal function by reducing proteinuria in a 5/6 nephrectomy rat model.^138,139 Nevertheless, further mechanistic studies are necessary to ascertain the efficacy and safety of these Smad3 inhibitors for treating renal fibrotic diseases. In addition, naringenin, a natural flavonoid abundant in grapefruit and other fruits and herbs, has been shown to exert a protective effect against organ fibrosis and metastasis of cancer cells as a specific Smad3 inhibitor.^140,141 Furthermore, the combination of naringenin (Smad3 inhibitor) and asiatic acid (Smad7 agonist) demonstrated a dual effect on renal fibrosis and DN-related islet injury by restoring the Smad3/Smad7 imbalance.^142,143 Hence, future research should evaluate the full potential of these compounds as therapeutic agents.

Protein phosphatase magnesium-dependent 1A (PPM1A) plays an important role in regulating TGF-β signaling. PPM1A is a Smad3 dephosphatase that causes the acceleration of Smad2/3 dephosphorylation in TGF-β-induced fibrosis.^144,145 Downregulation of PPM1A occurs in the tubulointerstitium of obstructive and aristolochic acid–induced nephropathy, which leads to the phosphorylation of Smad3 and renal fibrosis.¹⁴⁶ In addition, PPM1A and PTEN act together to reduce both Smad3 phosphorylation and activation of fibrotic genes.¹⁴⁷ Vitamin D analog maxacalcitol recruits the PPM1A/vitamin D receptor complex to promote Smad3 dephosphorylation, ameliorating tubulointerstitial fibrosis in obstructive nephropathy.¹⁴⁸ Since PPM1A dephosphorylates Smad2/3 at the C-terminal SXS motif, molecules that target this region could be identified as therapeutic drugs for antifibrosis treatment. One such drug candidate is small C-terminal domain phosphatase 1, which has been found to specifically dephosphorylate Smad2/3.¹⁴⁹ In addition, small molecular compounds such as GQ5 can be isolated as specific inhibitors of Smad3 to halt renal fibrosis.¹⁵⁰ Targeting Smad3 for ubiquitination and degradation represents another promising antifibrotic strategy. Proteolysis targeting chimeric molecules can degrade Smad3 through ubiquitination.¹⁵¹ Also, a study demonstrated that resveratrol inhibited obstructive renal fibrosis by promoting Smad3 deacetylation.¹⁵² Smad3 deacetylation also prevents kidney fibrosis.

Conclusion and future perspectives

Renal fibrosis is a serious complication of CKD, and understanding the pathogenic mechanisms of fibrosis is crucial in the development of effective therapies. Smad3 is an important mediator involved in canonical TGF-β/Smad signaling, and targeting this molecule has been considered a promising therapeutic strategy.

ncRNAs have been identified as key regulators of the TGF-β/Smad3 signaling pathway. Therefore, restoring the imbalance of TGF-β/Smad signaling and targeting TGF-β/Smad3-dependent ncRNAs are attractive therapeutics for the treatment of fibrosis. These ncRNAs have been proven to act as both profibrotic and antifibrotic regulators, as well as the biomarkers for the early detection of renal fibrosis. Further studies are urgently needed to fully elucidate the diverse role of TGF-β signaling in kidney diseases, paving the way for a translational leap from laboratory discovery to clinical application.