Antisense Oligonucleotide: A Potential Therapeutic Intervention for Chronic Kidney Disease

Abstract

Chronic kidney disease (CKD) is a global public health issue that places an increasing burden on the healthcare systems of both the developed and developing countries. CKD is a progressive and irreversible condition, affecting approximately 10% of the population worldwide. Patients that have progressed to end-stage renal disease (ESRD) require expensive renal replacement therapy, i.e., dialysis or kidney transplantation. Current CKD therapy largely relies on the use of angiotensin-converting enzyme inhibitors (ACEis) and angiotensin receptor blockers (ARBs). However, these treatments by no means halt the progression of CKD to ESRD. Therefore, the development of new therapies is urgently needed. Antisense oligonucleotide (ASO) has recently attracted considerable interest as a drug development platform. Thus far, eight ASO-based drugs have been granted approval by the US Food and Drug Administration for the treatment of various diseases. Herein, we review the ASOs developed for the identification of CKD-relevant genes and/or the simultaneous development of the ASOs as potential therapeutics towards treating CKD.

1. Introduction

1.1. Antisense Oligonucleotide

Antisense oligonucleotides (ASOs) are one of the most classical synthetic therapeutic oligonucleotides that are able to modify gene expression. ASOs are typically designed to regulate gene expression by specifically binding to the pre-mRNA or mRNA of the gene via Watson–Crick base pairing [1,2,3,4]. Since the first report on the ASO-mediated inhibition of gene expression in the late 1970s by pioneering researchers Zamecnik and Stephenson [5,6,7], ASO technology has become a well-established platform for ASO-mediated RNA-targeting therapy [8]. So far, the US Food and Drug Administration (FDA) has granted approval for eight ASO drugs for clinical usage (

Table 1. Antisense oligonucleotide (ASO)-based drugs approved by the FDA for clinical applications.

No.	ASO Drug	Approval Year	Indication	Mechanism	Ref.
1	Fomivirsen (Vitravene®)	1998	Cytomegalovirus (CMV) retinitis	Downregulate the gene encoding CMV immediate-early 2 protein	[9,10]
2	Mipomersen (Kynamro®)	2013	Familial hypercholesterolemia (FH)	Downregulate the gene APOB encoding apolipoprotein B	[11,12]
3	Eteplirsen (Exondys 51®)	2016	Duchenne muscular dystrophy (DMD)	Rescue the expression of dystrophin through exon-51 skipping of the mRNA of DMD gene	[13,14,15,16]
4	Nusinersen (Spinraza®)	2016	Spinal muscular atrophy (SMA)	Increase the production of the survival motor neuron (SMN) protein by exon-7 inclusion of the mRNA of SMN2 gene	[17,18,19,20]
5	Inotersen (Tegsedi®)	2018	Hereditary transthyretin (TTR) amyloidosis	Downregulate the gene TTR encoding transthyretin	[21,22]
6	Golodirsen (Vyondys 53®)	2019	DMD	Rescue the expression of dystrophin through exon-53 skipping of the mRNA of DMD gene	[23]
7	Viltolarsen (Viltepso®)	2020	DMD	Rescue the expression of dystrophin through exon-53 skipping of the mRNA of DMD gene	[24]
8	Casimersen (Amondys 45®)	2021	DMD	Rescue the expression of dystrophin through exon-45 skipping of the mRNA of DMD gene	[25]

and Table S1) [9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25]. These successful clinical translations inspire both the academia and pharmaceutical industry to develop ASO-based drugs for the treatment of various diseases, either by the downregulation of disease-causing gene expression or rescuing the expression of essential but defective genes.

1.2. Chronic Kidney Disease

Chronic kidney disease (CKD) is a leading public health issue worldwide [26]. Approximately 850 million people (~10% adult population) are affected by CKD [27]. CKD accounted for 1.3% of years of life lost (YLL) in 2012 [28] and will become the fifth most common cause of YLL worldwide by 2040 [29]. In the US, CKD affects ~13% of the population [30], with more than 100,000 new patients starting on dialysis every year [31]. CKD is defined as the persistently decreased function of the kidney for more than 90 days [32]. Kidney dysfunction is shown by a glomerular filtration rate (GFR) of less than 60 mL/min/1.73 m², or markers of kidney damage such as hematuria, albuminuria, and a variety of abnormalities detected by histology or imaging [32,33]. In CKD, the kidney undergoes a progressive and irreversible functional decline that can be classified into six stages (G1: normal, GFR ≥ 90 mL/min/1.73 m²; G2: 60–89 mL/min/1.73 m²; G3a: 45–59 mL/min/1.73 m²; G3b: 30–44 mL/min/1.73 m²; G4: 15–29 mL/min/1.73 m²; and G5: end-stage renal disease/kidney failure, <15 mL/min/1.73 m²). Patients who reach end-stage renal disease (ESRD) require kidney replacement therapy, i.e., dialysis or kidney transplantation, as their kidney is no longer able to maintain life in the long term [28]. However, less than half of patients needing kidney replacement therapy have access to treatment [34], as many governments and individuals are faced with the unaffordability of these costly therapies [35]. Moreover, kidney transplantation leads to a high risk of mortality as a result of unavoidable continuous immunosuppression (anti-rejection), which may cause infections and cancer development [36,37]. Therefore, it is imperative to delay or even prevent the progression of CKD from early stage to ESRD.

Although the causes of CKD vary geographically, diabetes accounts for 30–50% of CKD worldwide [28]. In developed countries, diabetes and hypertension are the leading causes of CKD [28]. Diabetic nephropathy is the main cause of ESRD in these countries and the burden of ESRD resulting from type 2 diabetes is projected to increase fourfold in decades to come [38,39], which is partly due to the increased prevalence of type 2 diabetes in young people [40]. In developing countries, CKD from glomerulonephritis and interstitial nephritis are more common as a result of the high prevalence of infections [41,42], such as streptococcal infections, acquired immunodeficiency syndrome (AIDS), schistosomiasis, leishmaniasis, hepatitis B, and hepatitis C. Notably, people affected by CKD are 5–10 times more likely to die prematurely due to the complications of CKD than ESRD [43]. This increased risk of death can be largely attributed to cardiovascular disease and cancer [44].

1.3. Cardiorenal Syndromes

Since the kidneys and heart have a bidirectional interorgan communication, dysfunction in one organ may cause dysfunction in the other organ, resulting in cardiorenal syndromes, a complex disorder of both the heart and kidneys [45]. Cardiorenal syndromes can be classified into five sub-types: type I [heart failure leading to acute kidney injury (AKI)], type II (chronic heart failure leading to CKD), type III (AKI leading to acute heart failure), type IV (CKD leading to heart failure), and type V (systemic condition such as diabetes mellitus leading to both renal and cardiac dysfunction) [45]. The common pathophysiological mechanisms in heart failure and CKD include dysfunction of the neurohormonal system (leading to the activation of the renin–angiotensin–aldosterone system), abnormal endothelial activation, reduced intestinal perfusion, and release of proinflammatory cytokines such as TNF-α, IL-1, and IL-6 [46]. These mechanisms simultaneously and sequentially contribute to cardiorenal syndromes, ultimately resulting in fibrosis and dysfunction in both organs [46].

2. Conventional Therapies and Their Limitations

Current standard of care for the treatment of CKD contains angiotensin-converting enzyme (ACE) inhibitors (ACEis) and angiotensin receptor (AR) blockers (ARBs) [47,48]. Drugs of these classes to some extent reduce the risk of kidney failure and major cardiovascular events [49,50]. The renoprotective effect of ACEis and ARBs is attributable to their ability to normalize glomerular hyperfiltration in kidneys [51,52]. However, both therapies slow but do not halt the progression of CKD from early stage to ESRD [53]. In order to improve their renal protection effect, combinational treatments including ACEi plus ARB, renin inhibitor plus ACEi, or renin inhibitor plus ARB were evaluated; however, the results were unsatisfactory in that unacceptable side effects such as hypotension, increased hyperkalemia, and AKI were observed, leading to the termination of the trials [54,55]. With the lack of clinically validated therapeutic targets (apart from ACE, AR, and renin) of CKD within or beyond the renin–angiotensin system, it is critically important to identify new targets for CKD in an attempt to develop novel therapeutics [53]. The targeted inhibition of gene expression by nucleic acid-based interventions can be a useful strategy for the identification and validation of potential CKD-relevant targets and the simultaneous or subsequent development of ASO-based therapy for CKD.

3. Antisense Oligonucleotide as Therapeutics

3.1. Mechanisms of Action

ASOs are short single-stranded chemically modified DNAs or RNAs that are about 15 to 30 nucleotides in length, complementary to their target pre-mRNA or mRNA. Upon binding specifically to their RNA targets and forming a target/ASO duplex, ASOs are capable of modulating gene expression through different mechanisms of action. These include: (1) induction of Ribonuclease H (RNase H)-mediated mRNA decay; (2) steric blockade that either modifies splicing by hindering the splicing factors from binding to pre-mRNA or represses gene expression by avoiding the 5′-capping of pre-mRNA or impeding the translational machinery from associating with mRNA (Figure 1) [2,56,57,58,59,60,61].

RNase H-mediated mRNA degradation is the most used ASO mechanism for the purpose of gene knockdown, which allows for the exploration of gene function, identification and validation of potential disease-relevant genes, and therapeutic application by downregulating disease-causing gene expression [62]. The FDA has so far granted three RNase H-competent ASO drugs, fomivirsen (Vitravene^®), mipomersen (Kynamro^®), and inotersen (Tegsedi^®) as therapies for cytomegalovirus (CMV) retinitis, familial hypercholesterolemia (FH), and hereditary transthyretin (TTR) amyloidosis, respectively [63]. Splice modulation by steric-blocking ASOs has been widely used to rescue the expression of essential but defective genes due to frameshifting mutations [64]. Splice modulation by exon skipping can restore the open reading frame by removing premature termination codons caused by mutations and restore the production of functional essential proteins. Since 2016, the FDA has approved five splice-modulating ASO drugs, eteplirsen (Exondys 51^®), nusinersen (Spinraza^®), golodirsen (Vyondys 53^®), viltolarsen (Viltepso^®), and casimersen (Amondys 45^®) for the treatment of Duchenne muscular dystrophy (DMD) (eteplirsen, golodirsen, viltolarsen, casimersen) and spinal muscular atrophy (SMA) (nusinersen), respectively.

3.2. Chemical Modification and Rational Design of ASO

ASOs composed of natural deoxyribonucleotides or ribonucleotides are not suitable for research and/or therapeutic purposes as they are easily degraded by extracellular and intracellular nucleases, lacking target-binding affinity and specificity [63]. In order to resolve these problems, chemically modified nucleotides, i.e., nucleotide analogues, are used for constructing ASO sequences [65,66,67,68,69,70] (Figure 2). Each nucleotide comprises a sugar (either ribose or deoxyribose), a phosphate (internucleotide linkage), and a nitrogenous base, and these components can be chemically modified alone or combinedly. The earliest attempt to modify a nucleotide was focused on replacing the non-bridging oxygen atom of phosphate by other atoms or groups. Examples include phosphorothioate (PS), phosphorodithioate (PS2), methyl phosphonate (MP), boranophosphate (PB), and phosphonoacetate (PACE) [71,72,73,74,75] (Figure 2). PS modification (the oxygen is replaced by sulfur) enhances the nuclease stability of ASOs and prolongs their half-life in plasma. However, this advantage is compromised by increased toxicity. Mesyl phosphoramidate (MsPA) has attracted considerable attention recently as it shows significant advantages over PS in nuclease stability, binding affinity, and physiological safety [76,77,78]. Therefore, MsPA is considered a potential alternative to PS [76]. The sugar moiety of the nucleotide is the hotspot of modification. A number of nucleotide analogs with modified sugar have been discovered or invented, such as 2′-O-methyl (2′-OMe) [79,80], 2′-O-methoxyethyl (2′-MOE) [81], 2′-O, 4′-C-ethylenebridged nucleic acid (ENA) [82], locked nucleic acid (LNA) [83,84,85], threose nucleic acid (TNA) [86], 2′-amino (2′-NH₂) [87], 2′-fluoro (2′-F) [88], 2′-fluroarabino nucleic acid (2′-FANA) [89], 1,5-anhydro hexitol nucleic acid (HNA) [90], cyclohexenyl nucleic acid (CeNA) [90], altritol nucleic acid (ANA) [90], and 2′-4′ constrained ethyl (cEt) [91] (Figure 2). Sugar modification may confer ASOs with improved nuclease resistance and enhanced hybridization affinity. Efforts have also been made on developing nucleotide analogues with a completely replaced backbone (i.e., sugar moiety plus phosphate moiety), such as peptide nucleic acid (PNA) [92] and phosphorodiamidate morpholino oligomer (PMO) [93] (Figure 2). A PNA monomer contains an uncharged N-(2-aminoethyl)-glycine, while a PMO monomer consists of a morpholine ring with a neutral phosphorodiamidate linkage. PNA and PMO display excellent target binding affinity and stability against cellular nuclease hydrolysis. Nucleobase modifications are not as commonly used as the above-described modifications on internucleotide linkage, sugar, or backbone in constructing ASO sequences [56].

An ASO can be synthesized as a uniformly modified sequence, a gapmer, or a mixmer (Figure 3 and Figure S1). Uniform modification and mixmer-design confer ASO maximized improvement in terms of nuclease stability and binding affinity. However, most of the modifications (except for the above-mentioned modified linkages PS, PS2, PB, PACE, MsPA, excluding MP) are unable to induce target RNA degradation as these analogues do not support RNase H-mediated cleavage. Therefore, uniformly modified ASOs are solely used as a steric blocker to induce splice switching or translational repression. In order to retain the ability of ASOs in recruiting RNase H and improve their nuclease stability at the same time, gapmer design emerges as the perfect solution [94,95,96,97]. A gapmer-like ASO usually consists of an unmodified or PS-modified central sequence (~10 deoxyribonucleotides) and sugar moiety-modified flanking sequences (~5 nucleotide analogues) on both its sides (Figure 3 and Figure S1). The central DNA region ensures that the ASO is capable of inducing RNase H-dependent target hydrolysis, while the flanking modified sequences protect the ASO from nuclease attack and enhance binding affinity. Most recently, Anderson et al. reported that combination of PS and MsPA linkages could further improve the therapeutic index and duration of effect of gapmer-like ASOs [78]. Most of the nucleotide analogues can be used to synthesize chimeric ASOs (i.e., gapmer and mixmer) that contain more than one chemistry in an attempt to optimize the stability and efficacy of ASOs. However, it has been challenging to develop a flexible and robust synthetic route that allows for the generation of chimeric PMO sequencing with other analogues. To overcome this, Veedu et al. developed an analogue of PMO called morpholino nucleic acid (MNA), allowing the synthesis of chimeric ASO with both morpholine ring-containing nucleotide and 2′-OMe nucleotides, providing a promising solution for the incompatibility of PMO [98]. More recently, Caruthers et al. reported the synthesis of a new analogue called thiophosphoramidate morpholino (TMO), which allows the easy incorporation of morpholino–PS moieties and other sugar modifications [99]. To date, PS, 2′-MOE, and PMO are the only three chemistries used in FDA-approved ASO drugs. Specifically, fomivirsen is a 21 mer DNA sequence uniformly modified by PS [9,10]; both mipomersen and inotersen are 20 mer 5-10-5 2′-MOE-PS gapmers (5-10-5 MOE^PS-DNA^PS-MOE^PS design) [11,12,21,22], i.e., they contain a central 10 mer DNA-PS sequence and two 5 mer flanking 2′-MOE-PS sequences; nusinersen is an 18 mer ASO uniformly modified by 2′-MOE-PS [17,18,19,20]; and the four DMD-targeting fully PMO-modified drugs, eteplirsen, golodirsen, viltolarsen, and casimersen have 30, 25, 21, and 22 monomers, respectively [13,14,15,16,23,24,25].

3.3. ASO, siRNA, and miRNA

ASO and small interfering RNA (siRNA) are the two most widely used nucleic acid-based strategies for the transient silencing of gene expression [100]. Different from ASO, siRNA has two strands and degrades target mRNA by RNA-induced silencing complex (RISC) instead of RNase H [101,102,103]. Since unmodified RNAs possess a higher potency than oligodeoxynucleotides, it is relatively easier to obtain a potent siRNA than an ASO. Therefore, siRNA is considered a more preferable option for in vitro studies [101]. Although lower doses of therapeutic siRNAs are sufficient to cause target gene knockdown, and they exhibit a longer duration of activity as compared to common ASOs, off-target effects of siRNAs (such as microRNA-like off-target effects and activation of Toll-like receptors) can lead to toxicities which compromise their therapeutic benefits [104]. Some of the off-target effects could be alleviated by pinpointing chemical modification (e.g., the seed region of siRNA) [105]. However, ASO is the better choice to be developed as nucleic acid-based therapeutics for four reasons: (1) lower cost of production as ASOs are single stranded, (2) in vivo delivery of ASOs is easier than siRNA (ASO delivery does not need a vector while siRNA delivery needs a carrier) [101], (3) ASOs can not only silence gene expression, but also restore gene expression which has been recognized as the most effective strategy for the treatment of DMD and SMA, (4) novel chemical modification (e.g., MsPA) and its rational positioning in gapmer-like ASOs lead to the advent of very long-acting next generation antisense oligomers [78].

MicroRNAs (miRNAs) are short noncoding endogenous RNAs of 18–22 mer in length that play important roles in the up- or down-regulation of genes [106]. miRNA-based therapeutic oligonucleotides can be developed as miRNA mimics or anti-miRNAs (ASOs). miRNA mimics imitate miRNA functions, while anti-miRNAs bind to miRNAs complementarily and deactivate their functions through RNase H-mediated degradation or the steric blockage mechanism [107,108]. Like the chemically modified ASOs targeting mRNAs/pre-mRNAs, anti-miRNAs are usually composed of nucleotide analogues instead of unmodified nucleotides to obtain improved nuclease stability and target miRNA-binding affinity. For example, miRNA-92a is a potential therapeutic target of cardiovascular diseases with CKD as it mediates endothelial dysfunction in CKD [109]. Hinkel et al. developed an LNA-modified ASO as an anti-miRNA-92a (named LNA-92a) [110], which efficiently inhibited miRNA-92a leading to protection against ischemia-reperfusion injury in a pig model. In addition, long non-coding RNAs (lncRNAs) regulate the effects of miRNAs on mRNA expression [111]. Some of the lncRNAs have emerged as potential prognostic biomarkers for CKD progression, such as lncRNAs HCP5, and NOP14-AS1 [112]. Development of ASOs targeting these lncRNAs could facilitate research aiming at elucidating their roles in CKD and developing ASO-based therapeutic strategies.

4. Antisense Oligonucleotides Targeting Chronic Kidney Disease

ASO-mediated gene knockdown by specifically reducing the mRNA production of genes enables the identification and/or verification of disease-relevant genes in CKD, leading to the expansion of knowledge regarding the molecular basis underlying the disease, thereby bringing new hope to identifying promising target genes for CKD therapy aimed at delaying or halting the progressive decline of kidney function. Furthermore, upon recognition of genes as potential therapeutic targets for CKD, the identified ASO-based gene inhibitors can directly serve as lead compounds, which speeds up the drug development process and bypasses the potential difficulties that development of small molecule-based inhibitors may face, such as the undruggability of target proteins and the toxicity of small molecules.

In fact, uniformly PS-modified ASOs and 2′-MOE-PS gapmer-like ASOs have been developed by different groups to target genes involved in the progression of CKD by downregulating the expression of those genes through the RNase H mechanism. In this section, we review the ASO-based research on genes that are relevant to CKD, including THBS1 (also known as TSP1) that encodes thrombospondin-1 (TSP1), CCN2 (also known as CTGF) encoding connective tissue growth factor (CTGF), KRAS encoding Kirsten rat sarcoma viral oncogene homolog (KRAS), MTOR encoding mammalian target of rapamycin (mTOR), AGT encoding angiotensinogen (AGT), and APOL1 that encodes apolipoprotein L1 (APOL1). These studies provide new insights into the molecular pathogenesis of CKD and explore the therapeutic potential of ASO-based gene inhibitors for the targeted therapy of CKD. In vitro screening of ASO candidates and subsequent in vivo investigations in these studies are shown in

Table 2. In vitro screening of ASOs and the identified best-performing candidates for subsequent in vitro and/or in vivo studies. Purple asterisks “*” in ASO sequences represent PS modification, red color represents 2′-MOE modification, blue color represents 2′-4′ constrained ethyl modification, black color represents deoxyribonucleotide. 2′-MOE: 2′-O-methoxyethyl, PS: Phosphorothioate internucleotide linkage, PO: Phosphodiester internucleotide linkage.

In Vitro Study (Initial Screen of ASO Candidates)			Best-Performing ASO Candidates
Target Gene	Chemistry	Cellular Target	ASO Sequences	Ref.
TSP-1	Eleven 14–25 mer DNAPS ASOs	Mesangial cell	5′-T*T*C*T*C*C*G*T*T*G*T*G*A*T*T*G*A*A-3′ 5′-C*A*C*C*T*C*C*A*A*T*G*A*G*T*T-3′	[113]
CTGF	20 mer 4-12-4 MOEPO-DNAPS-MOEPO and MOEPS-DNAPS-MOEPS ASOs	Rat mesangial cell line	5′-CCACA*A*G*C*T*G*T*C*C*A*G*T*CTAA-3′ 5′-C*C*A*C*A*A*G*C*T*G*T*C*C*A*G*T*C*T*A*A-3′	[114]
KRAS	20 mer 5-10-5 MOEPS-DNAPS-MOEPS ASOs	Rat renal fibroblast (NRK-49F)	5′-A*T*T*C*A*C*A*T*G*A*C*T*A*T*A*C*A*C*C*T-3′ 5′-C*A*C*A*C*T*T*A*T*T*C*C*C*T*A*C*T*A*G*G-3′	[115]
MTOR	~150 20 mer 5-10-5 MOEPS-DNAPS-MOEPS ASOs	Primary murine hepatocytes (for screening), type 1 Madin-Darby Canine Kidney cells (for other in vitro experiments)	5′-T*C*C*A*C*T*T*T*T*C*A*C*A*G*C*A*C*T*G*C-3′	[116]
AGT	~150 20 mer 5-10-5 MOEPS-DNAPS-MOEPS ASOs	Primary murine hepatocytes	5′-T*C*T*T*C*C*A*C*C*C*T*G*T*C*A*C*A*G*C*C-3′	[117]
APOL1	Over 4000 16 mer MOEPS-DNAPS-MOEPS or 2′-4′ constrained ethyl (cEt)PS-DNAPS-cEtPS ASOs	A-431 cell line	N/A	[118]

and

Table 3. Representative results of in vivo studies of ASO-mediated silencing of CKD-related genes.

Target Gene	In Vivo Studies of ASO Mediated Gene Silencing		Ref.
TSP-1	Type of CKD model	Animal model	[113]
	Induced experimental mesangial proliferative glomerulonephritis (the anti-Thy1 model)	Sprague-Dawley rats (150–200 g)
	Therapeutic regimen of ASO
	ASOs were transferred into renal glomeruli via left renal artery perfusion. Five days after the administration, kidneys were isolated for analysis.
	Renal function and/or renal damage markers
	Inhibited glomerular extracellular matrix accumulation determined by significantly reduced collagen IV positive glomerular area (%): TSP-1 ASO-treated group (~16%), scrambled ASO-treated group (~31%), p < 0.01. Markedly reduced mesangial cell activation determined by significantly reduced smooth-muscle-actin positive glomerular area (%): TSP-1 ASO-treated group (~15%), scrambled ASO-treated group (~39%), p < 0.01.
CTGF	Type of CKD model	Animal model	[114]
	Mice received streptozotocin (STZ) to develop an experimental model of type 1 diabetes induced diabetic nephropathy, and db/db mice with naturally developed diabetic nephropathy	C57BL/6 mice
	Therapeutic regimen of ASO
	Mice with type 1 diabetes: 20 mg/kg (twice a week) for 16 weeks. db/db mice: 5, 10, 20 mg/kg (twice a week) for 8 weeks
	Renal function and/or renal damage markers
	Mice with type 1 diabetes: Reduced kidney hypertrophy determined by reduced ratio (kidney weight/body weight): CTGF ASO-treated group (1.4%), vehicle-treated group (1.9%), p < 0.02. Attenuated mesangial matrix expansion (a.u.): CTGF ASO-treated group (~1.8), vehicle-treated group (~3.2), p < 0.05. Significantly reduced urinary albumin determined by reduced 24 h urinary albumin excretion (urinary albumin/urinary creatinine, ug/mg): CTGF ASO-treated group (~1.5), vehicle-treated group (~4.0), p < 0.05. db/db mice: Matrix expansion (%): 10 mg/kg CTGF ASO-treated group (~60%), vehicle-treated group (~100%), p < 0.05. Urinary albumin/urinary creatinine (ug/mg): 20 mg/kg CTGF ASO-treated group (~1.2), vehicle-treated group (~2.4), p < 0.05.
KRAS	Type of CKD model	Animal model	[115]
	Unilateral ureteric obstruction (UUO) model	Male Wistar rats
	Therapeutic regimen of ASO
	12.5 mg/kg for six days (administration was performed on alternate days)
	Renal function and/or renal damage markers
	Significantly reduced fibrosis determined by reduced fibrosis score (%): KRAS ASO-1-treated group (17%), scrambled ASO-1 (~40%), p < 0.001; KRAS ASO-2-treated group (20.3%), scrambled ASO-2 (~36%), p < 0.01.
MTOR	Type of CKD model	Animal model	[116]
	An orthologous model of human autosomal dominant polycystic kidney disease (ADPKD) caused by a mutation in the Pkd2 gene	C57BL/6 Pkd2WS25/− mice
	Therapeutic regimen of ASO
	Intraperitoneal injection at 100 mg/kg/week for the first 4 weeks and 50 mg/kg/week for the remaining 8 weeks
	Renal function and/or renal damage markers
	Improved kidney function determined by reduced ratio (kidney weight/body weight): MTOR ASO-treated group (1.5%), scrambled ASO-treated group (2.4%), p < 0.001; and cyst volume density: MTOR ASO-treated group (15.1%), scrambled ASO-treated group (34.1%), p < 0.01.
AGT	Type of CKD model	Animal model	[117]
	An orthologous model of human ADPKD caused by a mutation in the Pkd2 gene	C57BL/6 Pkd2WS25/− mice
	Therapeutic regimen of ASO
	Intraperitoneal injection at 100 mg/kg/week for the first 4 weeks and 50 mg/kg/week for the remaining 8 weeks
	Renal function and/or renal damage markers
	Improved kidney function determined by reduced ratio (kidney weight/body weight): AGT ASO-treated group (1.5%), scrambled ASO-treated group (2.4%), p < 0.01; and cyst volume density: AGT ASO-treated group (22%), scrambled ASO-treated group (34.1%), p < 0.05.
APOL1	Type of CKD model	Animal model	[118]
	Human APOL1-transgenic mice with induced proteinuria by IFN-γ challenge	Human APOL1-transgenic C57BL/6 mice
	Therapeutic regimen of ASO
	Intraperitoneal injection at 50 mg/kg/week for four weeks
	Renal function and/or renal damage markers
	Prevention of IFN-γ induced proteinuria determined by urinary albumin-to-creatinine ratio (ACR) (ug Alb/mg Cre): APOL1 ASO-treated group (0), control ASO-treated group (~1000), p < 0.001.

, respectively.

4.1. Thrombospondin-1 (TSP1)

Renal fibrosis, defined as the pathological accumulation of extracellular matrix, is the common hallmark of CKD [119]. Transforming growth factor-β (TGF-β), a profibrotic cytokine, plays a major role in experimental renal disease when TGF-β is overexpressed in the anti-Thy1 model (induced mesangial proliferative glomerulonephritis in rats) [120,121], and accumulation of mesangial cell matrix and interstitial fibrosis could be identified in transgenic mice expressing active TGF-β1 [122]. Furthermore, in human CKD, the upregulation of TGF-β is correlated with excess extracellular matrix, providing evidence that TGF-β plays a vital role in mediating fibrosis [123]. To this end, Akagi et al. directly blocked TGF-β action in the anti-Thy1 model by the administration of TGF-β-targeting PS-modified DNA ASOs, which achieved markedly reduced extracellular matrix accumulation [124]. However, TGF-β is not a suitable therapeutic target as it is a pleiotropic cytokine that exhibits other essential biofunctions in mammals, and TGF-β knockout mice survive for only a few weeks after birth [125,126,127].

Thrombospondin-1 (TSP1) is a major activator of TGF-β1 [128,129,130,131]. Nevertheless, unlike the TGF-β null mice, TSP1 knockout mice do not die prematurely and are healthy [130,131]. De novo TSP1 expression in mesangial cells colocalizes with the upregulation of TGF-β1 in different experimental kidney disease models [132], including the anti-Thy1 model [121,133]. In order to investigate the role of TSP1 as a TGF-β activator in the development of renal fibrosis, Daniel et al. screened out two PS-modified DNA ASO sequences (an 18 mer ASO and a 15 mer ASO) from 11 candidates that were designed to inhibit TSP1 expression [113]. ASOs were selectively transferred to the glomeruli of Sprague-Dawley (SD) rats with anti-Thy1 antibody-induced mesangial proliferative glomerulonephritis through renal artery perfusion followed by electroporation. Six days after treatment, TSP1-specific ASOs achieved the efficient reduction of TSP1 expression by more than 60%, and the inhibition of active TGF-β secretion by 50%, while not affecting the total expression level of TGF-β [113]. The ASO treatment also led to an evident reduction in the glomerular number of nucleic positive for the phospho-Smad2/3 (a TGF-β-signaling molecule that is known as a marker of TGF-β activation), indicating a markedly decreased glomerular TGF-β activity, which was associated with a marked reduction in mesangial cell activation [113]. Furthermore, TSP1-targeting ASO therapy inhibited the accumulation of glomerular extracellular matrix so that extra-domain A of fibronectin was markedly reduced by nearly 96%, as well as the reduced accumulation of other extracellular matrix proteins such as collagen I and collagen IV [113]. This study reveals that TSP1 is a tightly regulated activator of TGF-β in the anti-Thy1 model, and is responsible for the accumulation of glomerular extracellular matrix by activating TGF-β. Moreover, the data of the study suggests that the ASO-based TSP1 inhibitor may be a feasible therapeutic for fibrotic renal disease.

4.2. Connective Tissue Growth Factor (CTGF)

Over 30% of ESRD is caused by diabetic nephropathy. Glomerulosclerosis, the pathological hallmark of diabetic nephropathy, is characterized by the extracellular matrix accumulation of mesangial cells and tubulointerstitial fibrosis [134]. Both in vitro and in vivo studies have established that TGF-β contributes to glomerulosclerosis and that overexpression of TGF-β is associated with fibrosis and scarring in response to renal injury in diabetes [135,136,137,138,139]. However, as mentioned in the previous section, TGF-β is not a suitable target for drug development due to its multifunctionality. CTGF, a prosclerotic cytokine overexpressed during diabetes and acting downstream of TGF-β [140], directly contributes to the accumulation of extracellular matrix and tubulointerstitial fibrosis in diabetic nephropathy [140,141,142]. Okada et al. downregulated the expression of CTGF in tubular epithelium by the intravenous administration of a CTGF-targeting, PS-modified 18 mer DNA ASO in mice with subtotal nephrectomy (SNx) [143]. They found that decreased expression of CTGF caused by the ASO is associated with attenuated interstitial fibrosis resulting from the downregulated expression of genes involved in the expansion of the glomerular extracellular matrix, demonstrating that CTGF is a direct and significant contributor of TGF-β-dependent renal fibrogenesis [143]. This study suggested that the development of an ASO-based CTGF inhibitor may be a promising strategy for antifibrotic therapy in TGF-β-dependent CKD such as diabetic nephropathy.

Later, Guha et al. investigated the role of CTGF in the progression of diabetic nephropathy by administrating a CTGF-specific 20 mer ASO (either 4-12-4 MOE^PO-DNA^PS-MOE^PO or 4-12-4 MOE^PS-DNA^PS-MOE^PS gapmer) to nephropathic mice with streptozotocin-induced type 1 diabetes and db/db mice (mice with type 2 diabetes) with a dosage of 20 mg/kg (twice weekly) for 16 weeks and 8 weeks, respectively [114]. The CTGF-targeting ASO-inhibited hyperglycemia induced overexpression of CTGF in both diabetic mouse models. In the mice with type 1 diabetes, ASO treatment inhibited a variety of indices of renal disease, for instance, the development of renal hypertrophy was significantly attenuated in that the diabetes-induced increase of kidney weight was reduced by 32%, increases of serum creatinine and urinary albumin which are pathological features of diabetic nephropathy were reduced by 32% and 52%, respectively, and the diabetes-induced expansion of the mesangial matrix was attenuated by 43%, which was associated with the reduced synthesis of collagen 1, fibronectin, and TGF-β1 [114]. Furthermore, ASO treatment achieved the inhibition of profibrotic p38 mitogen-activated protein kinase (MAPK) and its downstream target transcription factor cAMP-response element binding protein (CREB) so that the activation of p38 MAPK and CREB was attenuated by 54% and 74%, respectively, indicating that the progression of diabetic nephropathy may be inhibited [114]. In addition, in the db/db mice, CTGF-targeting ASO reduced serum creatinine, urinary total protein, and urinary albumin by 37%, 41%, and 48%, respectively [114]. This study provides sound scientific evidence that the specific knockdown of CTGF by gapmer-like ASO holds significant promise as a potential therapy for diabetic nephropathy.

4.3. Kirsten Rat Sarcoma Viral Oncogene Homolog (KRAS)

Tubulointerstitial fibrosis, characterized by the excessive deposition of extracellular matrix resulting from an increased number of activated interstitial myofibroblasts, is a key determinant of progressive CKD [144]. RAS proteins, termed small guanosine triphosphatases (GTPases), play essential roles in the regulation of cell survival, proliferation, and differentiation by acting as signal transduction molecules in various extracellular pathways [145]. Bechtel et al. demonstrated that the activation of RAS is directly associated with renal fibrogenesis [146]. Therefore, RAS proteins may be potential therapeutic targets for the renal fibrosis of CKD [147,148,149,150,151,152,153,154]. Sharpe et al. demonstrated that KRAS is the predominant isoform of RAS expressed in human renal fibroblasts and that PS or 2′-MOE-PS-modified ASO-induced KRAS knockdown significantly suppresses the proliferation of fibroblasts [155,156]. Later, in order to investigate the profibrotic role of KRAS in CKD, the same group silenced the KRAS expression in rats with unilateral ureteric obstruction (unilateral ureteric obstruction is considered a model of renal fibrosis and CKD [157]) by the subcutaneous administration of KRAS-specific 20 mer gapmer-like ASOs (ISIS 104440 or ISIS 104419, 5-10-5 MOE^PS-DNA^PS-MOE^PS design) on alternate days (for six days) at a dosage of 12.5 mg/kg [115]. Treatment of KRAS-specific ASOs significantly reduced the level of KRAS mRNA by 61% (ISIS 104440) and 97% (ISIS 104419) compared to their correspondent scrambled ASOs (negative control), which was associated with reduced renal fibrosis (fibrosis score was reduced to 17% by ISIS 104440 and 20.3% by ISIS 104419) and collagen deposition (collagen deposition score was reduced to 18.4% by ISIS 104440 and 17% by ISIS 104419) [115]. Furthermore, the upregulation of α-smooth muscle actin (α-SMA, a marker of myofibroblast activity) induced by obstructive nephropathy was inhibited by the KRAS-targeting ASOs so that the α-SMA expression was reduced from 53% (negative control group) to 3.9% and 20% by ISIS 104440 and ISIS 104419, respectively, indicating that an ASO-mediated KRAS knockdown could prevent the onset of fibrosis in rat models of unilateral ureteric obstruction [115]. This study suggests that ASO-induced KRAS inhibition could be a novel antifibrotic strategy for CKD as ASO treatment markedly inhibited renal fibrosis.

It is worth mentioning that Ross et al. recently evaluated another highly potent KRAS-specific ASO candidate (AZD4785 or ISIS 651987) for its anti-tumour effects in vitro and in vivo [158]. We recommend researchers compare the efficacy between AZD4785 and ISIS 104440/ISIS 104419 to identify the best-performing lead candidate for further anti-CKD therapeutic development.

4.4. Mammalian Target of Rapamycin (mTOR)

Autosomal dominant polycystic kidney disease (PKD), caused by a mutation in the PKD1 or PKD2 gene, with 50% of patients developing CKD, accounts for ~5–10% of ESRD in the US requiring dialysis or renal transplantation [159]. The mTOR signaling pathway plays a role in the regulation of cell growth and proliferation [160]. Specifically, mTOR complex 1 (mTORC1) controls protein synthesis and cell proliferation, and the hyperactivation of the mTORC1 signal is a feature of PKD [161,162,163,164]. mTOR complex 2 (mTORC2) modulates cell survival and arrangement of actin cytoskeleton by phosphorylating AGC kinases such as the pro-survival kinase pAkt^Ser473 [165,166,167]. Activation of mTORC2 is upregulated in PKD [162,163,164,165,166,167,168,169]. Sirolimus, an mTORC1 inhibitor, has been demonstrated to have a therapeutic effect on mice with PKD resulting from Pkd1 inactivation via the reduction of cyst growth and preservation of renal function [162]. However, sirolimus had no effect on renal function in Pkd2WS25/− mice (a model of human autosomal dominant PKD resulting from mutation in the Pkd2 gene) [170], and a clinical study showed that sirolimus treatment did not halt the growth of polycystic kidneys in humans [171]. One possible reason of the inefficacy is that sirolimus directly inhibits mTORC1 but not mTORC2, therefore, it is unable to inhibit mTORC2-dependent Akt-induced proliferation [172].

In order to determine the therapeutic effect of the combined inhibition of mTORC1 and mTORC2 on PKD, Ravichandran et al. developed an mTOR-specific 20 mer gapmer-like ASO (5-10-5 MOE^PS-DNA^PS-MOE^PS design) through the screening of ∼150 ASO candidates [116], as mTOR exists in both mTORC1 and mTORC2 [173]. ASOs were administered into Pkd2WS25/− mice via intraperitoneal injections at a sequential two-stage dosage of 100 mg/kg/week (first 4 weeks) and 50 mg/kg/week (the remaining 8 weeks) [116]. Treatment of mTOR ASO led to a significant reduction in the expression levels of mTOR, pS6 (a marker of mTORC1 signaling), and pAkt^Ser473 [116], which was associated with the reduced ratio of two kidneys/total body weight from 2.4% (control ASO group) to 1.5% (mTOR ASO group), significantly decreased cyst volume density from 34.1% (control ASO group) to 15.1% (mTOR ASO group), and significantly reduced blood urea nitrogen from 43.4 mg/dL (control ASO group) to 29 mg/dL (mTOR ASO group) [116], indicating an ASO-induced melioration of PKD and normalization of renal function. Furthermore, mTOR ASO treatment significantly inhibited both the proliferation and apoptosis of tubular epithelial cells (proliferation and apoptosis of epithelial cells lining the renal tubular plays a key role in cyst growth [174]) so that the number of proliferating cell nuclear antigen positive cells was reduced from 1.9 per cyst (control ASO group) to 0.8 per cyst (mTOR ASO group), and the number of apoptotic cells was reduced from 3.6 per non-cystic tubule (control ASO group) to 1.2 per non-cystic tubule (mTOR ASO group) [116], suggesting that ASO therapy could inhibit cyst growth. This study demonstrated that the combined inhibition of both mTORC1 and mTORC2 holds therapeutic potential for autosomal dominant PKD [116], and an ASO-based mTOR inhibitor can be a promising approach for treating the disease owing to its capability of combined mTORC1/2 knockdown.

4.5. Angiotensinogen (AGT)

Enlargement of renal cysts in autosomal dominant PKD is associated with the activation of the renin–angiotensin system and the resultant production of proinflammatory and profibrotic angiotensin II [175,176,177], which contributes to cystogenesis by the induction of cellular proliferation, inflammation, and fibrosis [175,178]. The renin–angiotensin system also plays an important role in type IV cardiorenal syndrome and CKD [179]. Angiotensin II induces the differentiation of renal fibroblasts into myofibroblasts and stimulates the expression and activation of TGF-β [180,181]. Angiotensin II leads to renal damage by increasing the expression of proinflammatory cytokines and chemokines and renal leukocyte infiltration [182]. Despite the importance of the renin–angiotensin system in the pathogenesis of PKD, the single or combined use of renin–angiotensin system inhibitors, including ACEis, ARBs, and renin inhibitor, have limited efficacy and induce side effects [54,55]. Identification of other targets within the renin–angiotensin system may result in the development of more efficient therapeutics.

Angiotensinogen (AGT) is a substrate of the peptidase renin in the renin–angiotensin system, and AGT is cleaved by renin forming angiotensin I, which is then converted to angiotensin II by ACE [181]. In order to investigate the potential therapeutic effects of direct AGT inhibition on PKD, Ravichandran et al. developed an AGT-specific 20 mer gapmer-like ASO (5-10-5 MOE^PS-DNA^PS-MOE^PS design) through the screening of ∼150 ASO candidates [117]. The administration of ASO was performed by intraperitoneal injections into Pkd2WS25/− mice at a sequential two-stage dosage of 100 mg/kg/week (first 4 weeks) and 50 mg/kg/week (the remaining 8 weeks) [117]. AGT-specific ASO treatment significantly reduced the AGT expression, which was associated with a reduced two kidney/total body weight ratio (AGT ASO group: 1.5%, control ASO group: 2.4%), decreased cyst volume density (AGT ASO group: 22%, control ASO group: 34.1%) and blood urea nitrogen (AGT ASO group: 34 mg/dL, control ASO group: 47 mg/dL), indicating that AGT-specific ASO treatment could lead to decreased PKD and normalization of renal function [117]. Furthermore, significant decreases in proinflammatory cytokines including C-X-C motif chemokine ligand 1 (CXCL1), interleukin 12 (IL-12), and the profibrotic TGF-β were observed in the AGT ASO treatment groups (CXCL1: 0.6 pg/mg, IL-12: 8.8 pg/mg, TGF-β: 32 pg/mg) in contrast to the control ASO treatment groups (CXCL1: 3.4 pg/mg, IL-12: 37.3 pg/mg, TGF-β: 102 pg/mg), indicating that AGT-specific ASO treatment could lead to decreased proinflammatory and profibrotic molecules [117]. Although further investigation is required to elucidate the mechanism underlying the ASO-induced inhibition on cyst growth, chronic AGT inhibition by ASO may be a possible therapeutic strategy for autosomal dominant PKD in the future.

4.6. Apolipoprotein L1 (APOL1)

APOL1 is a newly evolved gene that is only present in a few primates such as humans, baboons, and gorillas [183,184]. The APOL1 protein functions as the trypanolytic factor in serum that lyses trypanosomes against African trypanosomiasis (sleeping sickness), a disease endemic to Africa [185,186,187]. One of the trypanosome species, Trypanosoma brucei rhodesiense, has evolved to resist the wild type APOL1 (G0), while the G1 and G2 mutants of APOL1 (commonly found in populations of African ancestry), discovered in 2010, overcome the resistance of Trypanosoma brucei rhodesiense [188]. However, mutant APOL1 also leads to toxic gain of function when overexpressed so that [189,190], compared to the wild type APOL1 (G0), the G1/G2 mutants are associated with an increased risk of CKD by 7- to 30-fold [188,191,192,193]. Furthermore, these mutants accelerate the GFR decline, and thus the progression of CKD [192,194]. In addition, G1/G2 mutants are strongly associated with various forms of nondiabetic nephropathy such as focal segmental glomerulosclerosis (FSGS) and interferon (IFN) therapy-related collapsing glomerulopathy [188,189,192]. As APOL1 is not essential for kidney development and function [183,195,196,197], reducing the expression level of APOL1 therapeutically will not result in any harmful effects other than increased susceptibility to African sleeping sickness in specific geographical regions [118]. Therefore, in an attempt to study APOL1 systemically and achieve proof of concept for APOL1 inhibition by antisense oligomer, Aghajan et al. established a transgenic mouse model for APOL1-related CKD (C57BL/6 mice with human APOL1 G1 mutant gene were challenged by IFN-γ leading to induced proteinuria in mice), developed a APOL1 specific, 2′-MOE or 2′-4′ constrained ethyl (cEt)-modified 16 mer gapmer ASO on a PS backbone (IONIS-APOL1_Rx) through the screening of over 4000 ASO candidates, and treated APOL1 G1 transgenic mice with IONIS-APOL1_Rx at a weekly dose of 50 mg/kg for four weeks prior to IFN-γ challenge [118]. ASO treatment led to a significant decrease in the APOL1 mRNA levels in both the kidney (by ~50%–60%) and liver (by 95%), which completely prevented the occurrence of IFN-γ-triggered proteinuria, indicating that the CKD-relevant cell types (podocytes, endothelial cells, and mesangial cells that constitute the renal filtration barrier) are sensitive to ASO treatment [118]. It was also found that administration of IONIS-APOL1_Rx at a weekly dosage as low as 6.25 mg/kg led to the significant inhibition of induced proteinuria, demonstrating the potency of ASO therapy and the renoprotective effect that it provided [118]. Although further study is definitely required to reveal the pathogenesis of the toxic gain of function resulting from mutant APOL1, IONIS-APOL1_Rx holds promise to be developed as an efficient anti-CKD therapeutic option for patients with APOL1 nephropathies.

5. Potential Problems of ASO-Based CKD Therapy and Possible Solutions

The efficient delivery of therapeutic ASOs to their target tissues remains a major challenge. Although all FDA-approved ASO drugs are administered directly by injection, intense research efforts have yielded different strategies aimed at improving the in vivo delivery of ASOs, such as nanocarriers, viral delivery, nanoparticles, antibodies, and aptamers. The advancement in the delivery methods of therapeutic oligonucleotides has been discussed in a number of recent reviews [198,199,200,201,202].

The ASO-mediated knockdown of CKD relevant gene expression may lead to severe adverse effects due to the multifunctionality of target genes. For example, although TGF-β inhibition could suppress progression to glomerulosclerosis [123], TGF-β knockout could be deadly due to its essential roles in multiple developmental processes [124,125,126]. Possible solutions include:

(1)

Exploration of target genes that are located upstream or downstream of TGF-β pathway, such as TSP1 (upstream) [113] and CTGF (downstream) [114].

(2)

Exploration of target genes that are newly evolved or less conserved, such as APOL1 [118], so that inhibition of such genes is probably less risky compared with the genes that are functionally conserved.

(3)

Identification of causative genes for diabetes and hypertension as these diseases are the leading causes of CKD; ASO-based targeted therapies may prevent the induction of CKD [8,26,28,203].

(4)

Investigation of AKI-related biomarkers such as neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule-1 (KIM-1) [204,205], since these biomarkers could be potential therapeutic targets that could be involved in the AKI-to-CKD transition.

Off-target effects of ASO that are either sequence dependent or independent may result in in vivo toxicity. Possible solutions include: (1) extending the length of ASO from 15 mer to ~25–30 mer, thus increasing its specificity and reducing potential hybridization between ASO and non-target RNAs; (2) employment of nucleotide analogues of improved safety profile, such as PMO to reduce the sequence independent toxicity of ASO. One example of this solution is the rejection of drisapersen (2′-OMe-PS modified ASO, BioMarin Pharmaceutical) due to life-threatening toxicity and the approval of eteplirsen (PMO modified ASO, Sarepta Therapeutics) by the FDA in 2016 (both drisapersen and eteplirsen are DMD mRNA exon-51-targeting drugs) [56].

6. Conclusions

Patients with CKD progressed to ESRD require expensive dialysis or kidney transplantation; however, current CKD therapy relying largely on the use of renin–angiotensin system inhibitors, such as ACEis and ARBs, does not halt the progression of CKD. Therefore, the development of new therapies with an improved potency and safety profile is urgently needed. Chemically modified ASOs have been used in the functional study of genes involved in the pathogenesis of CKD and the subsequent or coinstantaneous development of ASO-based inhibitors of the target genes. Compared with conventional small molecule-based protein inhibitors, ASOs possess an advantage due to their capability of inducing direct targeted mRNA degradation, bypassing potential obstacles facing “undruggable” proteins. As a result, a few ASO-based inhibitors of genes responsible for the pathogenesis of CKD have been developed. It is recommended that the antifibrotic effects of these ASOs in the heart should also be investigated since cardiac and renal dysfunction share common mechanisms. Furthermore, efforts have been made in the identification of novel therapeutic targets (other than the ACE and AR) within or beyond the renin–angiotensin system. Given the convenience and flexibility of developing an ASO-based inhibitor of a specific gene, there is no doubt that more CKD-relevant genes will be identified as potential therapeutic targets, which may lead to the development and approval of antisense anti-CKD drugs in the future.