2.1.1. siRNA

Short interference RNAs are short (20–27 nucleotides) double-stranded RNAs that target and degrade mRNA in a sequence-specific manner. The guide (antisense strand) is loaded onto Argonaute 2 protein (AGO2), forming the RNA induced silencing complex (RISC), whereas the sense strand is cleaved. The guide strand targets the specific mRNA by complete complementarity and AGO2 catalyzed mRNA cleavage. The RISC and guide strand can be recycled to target multiple mRNA molecules leading to efficient gene silencing ^[11][13][59]. Alternatively, siRNAs can also be designed to target long non-coding RNAs (lncRNAs), often involved in transcriptional repression, reversing the effects of this negative regulation ^[4]. siRNAs can sometimes be encoded in the form of short-hairpin RNAs (shRNAs), which are usually delivered to the cell by transduction with viral vectors and, in some cases, integrated into the host genome. shRNA is first expressed as a miRNA and processed into a siRNA duplex by the enzymes Drosha and Dicer, which then follows the same interference mechanism previously described ^[13]. Two siRNA-based therapies, Patisiran and Givosiran, have already been approved by the regulatory agencies in May 2020, both targeting the liver ^[4][5]. Interestingly, siRNA technology has also been explored for kidney diseases, showing potential as a therapeutic agent as well as contributing to the understanding of the molecular mechanisms of renal diseases (Table 1) ^[60]. One of the earliest studies successfully demonstrated the feasibility of using a siRNA-based therapy to ameliorate glomerular sclerosis in a mouse model of glomerulonephritis, by modulating the transforming growth factor beta (TGFβ) pathway as result of Mapk1 silencing ^[17]. Likewise, Morishita et al. ^[19] prevented renal fibrosis by using a siRNA against Smad4, suggesting it could be a crucial therapeutic target for renal fibrosis in vivo. Other studies using siRNA-based drugs have focused on reducing the extent of acute kidney injury (AKI), an unavoidable side effect of numerous medical treatments and surgical procedures which deprive the kidney of oxygen. For instance, Glebova et al. ^[61] evaluated the potential of 53 different siRNA targets, mainly related to apoptosis, inflammation and immune rejection pathways after ischemia-reperfusion caused by transplantation. This approach is still under development but has already shown promising results in a mouse model ^[22]. In a similar study, the authors effectively evaluated the prophylactic role of siRNA targeting meprin-1β and p53 expression in a cisplatin-induced murine model of AKI. These two proteins play key roles in depolarization and apoptosis after kidney injury ^[23]. Narváez et al. ^[25] also demonstrated that the administration of a siRNA therapy against Cd40 in a mouse model of AKI induced by unilateral ureteral obstruction (UUO) significantly reduced inflammation and promoted kidney repair. Importantly, upon demonstration of a lack of reduction in megalin protein expression in vivo using a siRNA, another recent study has pointed out that the use of siRNA-based therapies in kidney diseases might, indeed, be more suited to prevention of upregulation than reduction of constitutive baseline mRNA expression ^[24]. Although there is still scope for improvement, siRNA-based therapies in the kidney have shown promising results, especially in the prevention of AKI where some clinical trials have already been carried out ^[60].

2.1.2. saRNA

Small activating RNAs are double-stranded RNAs of 21 nucleotides in length that possess two nucleotide overhangs on both ends. Comparable to siRNA, saRNAs are loaded onto AGO2, where the sense strand in cleaved. Then, the saRNA–AGO2 complex is translocated to the nucleus, binding to complementary promoter regions and recruiting key elements for transcription initiation ^[62]. Thus, saRNA has an identical structure and chemical components as siRNA, but its biological function is the opposite of siRNA, since it acts to enhance gene transcription. The use of saRNA with therapeutic purposes has been recently tested in an in vivo model of ethylene glycol (EG)-induced calcium oxalate (CaOx) kidney crystal formation in rats. Using this approach, the authors significantly enhanced the expression of Trpv5, a key protein mediating calcium transport and reabsorption in the kidney, and achieved a reduction in CaOx crystal formation by promoting calcium reabsorption ^[34].

2.1.3. miRNA

miRNA constitute a class of single-stranded non-coding RNAs (ncRNAs) with a length of approximately 22 nucleotides after maturation. Their natural biogenesis starts in the nucleus, where they are transcribed as pri-miRNAs and cleaved into pre-miRNAs (~70 nucleotides in length) by a multiprotein complex mainly comprised of the Drosha and Dcrg8 proteins. Double-stranded pre-miRNAs are then exported to the cytoplasm and processed by Dicer into mature miRNAs (~22 nucleotides in length) ^[63]. For their mechanism of action, miRNAs are loaded onto AGO2 to form RISC, guiding the complex to its complementary binding site in the target transcript, commonly found in the 3′ untranslated region (UTR). In contrast with siRNAs, miRNAs typically bind with partial complementarity and usually promote translational repression by triggering mRNA decay through deadenylation and decapping. Different miRNAs can bind to the same transcript by overlapping or non-overlapping sites ^[11][13]. In general, there are three approaches to developing miRNA-based therapeutics: (1) anti-miRNA oligonucleotides (AMOs) or miRNA antagomirs, also known as antisense oligonucleotides (ASOs), which antagonize endogenous miRNAs by steric blocking of the miRNA within the RISC complex ^[4][11]; (2) miRNA sponges, which are genetically engineered competitive miRNA inhibitors designed by insertion in tandem of multiple binding sites of targets of analogous miRNAs or mRNAs ^[64]; (3) miRNA mimics, which are engineered double-stranded miRNAs that replace, improve or supplement the function of native miRNAs ^[11][59]; (4) target site blockers (TSB), designed to recognize and mask the regulatory sequences of miRNAs within a specific mRNA, with the potential to maintain the rest of the mRNA network unaffected ^[4].

miRNAs play an important role in the negative regulation of post-transcriptional gene expression. Consequently, aberrant miRNA expression is implicated in the development and progression of numerous diseases, and multiple families of miRNAs are shown to be dysregulated in kidney disease. The role of miRNAs in molecular pathology has already been reported in AKI and kidney transplantation ^[65][66], polycystic kidney disease (PKD) ^[67][68], and renal fibrosis ^[69][70], which represents the final outcome and most relevant pathological event of CKD. This, together with the fact that miRNAs can be detected in exosomes in plasma and urine, indicate them as suitable and attractive new biomarkers for diagnostic purposes and disease monitoring ^[71][72]. Importantly, up/downregulated miRNAs also represent novel therapeutic targets for kidney diseases, whose potential has already begun to be explored in animal models (Table 1). One of the earlier studies was carried out by Chau et al. ^[35], who successfully used an anti-miR21 antagomir, which limited injury and kidney fibrosis in two murine models of AKI. Similarly, renal fibrosis was also ameliorated by an anti-miR192 antagomir in a mouse model of diabetic nephropathy (DN) ^[36]. The use of miR204 and miR211 mimics has also proved effective at reducing the severity of kidney injury in a mouse model of systemic candidiasis ^[37] and inhibition of miR107 using an antagomir significantly prevented tubular cell injury in a mouse model of AKI induced by sepsis ^[39]. Moreover, the use of miRNA-based approaches has also shown potential in the treatment and understanding of the basis of ischemic-kidney injury, as demonstrated by a collection of recent studies (Table 1). One of the most common causes of ischemic AKI is transplantation. In that regard, a recent study has demonstrated the applicability of using antisense technology against the miR182-5p target to improve kidney function and morphology, employing a model of ischemic AKI in rats ^[40]. In addition, Wei et al. ^[41][42] elucidated the molecular mechanisms of miR489 and miR668 in the protection of the kidney during ischemia, indicating the possibility of using these miRNA or miRNA mimetics as therapeutic agents. Similarly, another study demonstrated the protective role of miR199a-3p in an ischemia-reperfusion model in vivo ^[45]. Taken together, these findings are promising for future evaluation of the clinical utility of miRNA mimetics and inhibitors targeting key pathologic renal pathways. However, translation of preclinical findings is sometimes complicated, as a deep understanding of the miRNA regulatory networks underlying the disease is needed. In some cases, targeting key points in the same network may prove more effective. Additionally, most miRNAs are regulated in a cell-type or organ-specific manner; thus, the possibility of off-target and undesired effects in unrelated organs is high. This problem could explain why few investigations using miRNAs-based therapies move forward to the clinical stage. In fact, only four miRNA-based therapies have reached clinical development, two of them dedicated to renal disease ^[66]. One of the drugs was developed for Alport nephropathy, a genetic disorder characterized by chronic glomerulonephritis that progresses to end-stage renal disease in young adult life. The antagomir against miR21 was effectively evaluated in a mouse model, where the animals displayed substantially milder disease and significantly improved survival after treatment ^[38]. The drug RG012 targeting miR21 is currently undergoing a phase 2 clinical trial (clinical trial identifier NCT02855268). The second drug is RGLS4326, an antagomir inhibiting miR17 developed for the treatment of autosomal dominant PKD (ADPKD), a genetic disorder caused by mutations in either PKD1 or PKD2 genes resulting in hyperproliferation of the renal tubular cells and cyst formation ^[43]. Treatment with RGLS4326 attenuated cyst growth in several PKD mouse models and human ADPKD models in vitro and is now in a phase 1 clinical trial (clinical trial identifier NCT04536688).

2.2. RNA/Protein-Based Strategies (CRISPR)

The development of the revolutionary CRISPR/Cas9 gene editing technology is driving the progress of RNA therapeutics forward in a similar way. This system, initially discovered as a form of acquired immunity in bacteria and archaea, consists of a protein (CRISPR-associated nuclease Cas9) and an oligonucleotide guide RNA (sgRNA). While the gRNA directs the nuclease to a specific genomic location adjacent to a protospacer adjacent motif (PAM) sequence, correct base-pairing activates Cas9 nuclease domains, which, cut the DNA, resulting in a double strand-break. In an attempt to repair the damage, eukaryotic cells can use two different mechanisms, nonhomologous end joining (NHEJ) and homology-directed repair (HDR). The NHEJ repair mechanism is prone to introduce small insertion or deletion errors, causing frameshift mutations and leading to gene knockout by disruption of the open reading frame. Conversely, when a donor template is introduced, it can be utilized by the HDR mechanism and harnessed to introduce a new sequence bearing a mutational correction or sequence knock-in in the desired loci ^[73][74]. A great advantage of the system, as consequence of its modularity, is that it allows for the testing of many different potential sgRNA while maintaining the protein component invariant. Importantly, although this system was initially developed to target DNA, RNA-targeting and interference has also been possible due to the development of new engineered nucleases such as Cas13 and RNA-targeting Cas9 (rCas9). Similarly, nuclease deficient Cas9 variants (dCas9) restricted from generating DNA breaks have been fused to transcriptional activation (VPR) ^[75], silencing proteins (KRAB) ^[76], or epigenetic modifiers ^[73][74], which can then be targeted to specific gene promoters, regulating gene expression. Catalytically inactive Cas9 and Cas13 variants have also been fused to other types of functional domains, such as base-pair editors or deaminases able to catalyze A-T to G-C transitions, allowing for single-base edition at the DNA and RNA level without the need to generate double or single-strand breaks ^[74], or reporter proteins, to visualize DNA or RNA ^[73]. Nonetheless, despite the fact that the CRISPR/Cas9 system has emerged as a promising platform with a wide variety of applications in biology research and therapeutics of human disease, efficient and safe delivery of its components to target cells in vivo remains challenging ^[77].

The use of the CRISPR/Cas9 system as a therapeutic approach for renal diseases has great potential, as a significant proportion of these diseases, such as autosomal dominant PKD or Alport syndrome, arise as consequence of genetic mutations. Nevertheless, gene editing in solid organs still faces the challenge of effective delivery to specific cells or tissues. Thus, the use of this tool in kidney research has so far been limited to its application in the development of novel in vitro (using human organoids) and in vivo models of renal disease. Such models are very useful in understanding the molecular mechanisms underlying renal diseases, as well as in the identification of new genes responsible for their progression and that could represent potential therapeutic targets ^[78][79]. Another exciting potential use of CRISPR/Cas9 technology is focused on expanding the available sources of kidneys for transplantation. Some authors have proposed the possibility of transplanting organs from other species such as pigs (xenotransplantation), an approach that, to this day, would lead to an extreme human immune response and rejection of the donor organ. However, CRISPR/Cas9 has appeared as a promising tool that could circumvent this limitation. In this direction, some authors have already employed CRISPR/Cas9 to genetically modify swine eggs in order to generate animals lacking carbohydrate xenoantigen, whose recognition by human and non-human primate antibodies was effectively diminished ^[80], or to generate I MHC null pigs ^[81]. The first demonstration of the feasibility of this approach was published by Higginbotham et al. ^[82], where the authors achieved effective pig-to-primate long-term transplantation (>125 days).