Recent Advances and Prospects of Nucleic Acid Therapeutics for Anti-Cancer Therapy

1.1. Overview of Nucleic Acid Therapeutics (NATs)

Nitrogen bases, which are one of the main components of nucleic acids, form hydrogen bonds between complementary bases that are called base pairs. Through interstrand base pairing, they can hybridize into double-stranded structures, and through intrastrand base pairing, they can fold to form secondary structures. This is the most basic working principle of all NATs, which will be described below. NATs are nucleic acid strands that range in length from as short as 8–50 nucleotides to as long as several thousand or more nucleotides and operate through a variety of mechanisms, including regulating gene expression or protein activity. NATs include various forms of DNA and RNA, including plasmid DNA (pDNA), small interfering RNA (siRNA), anti-microRNA (anti-miR), microRNA (miR) mimics, antisense oligonucleotide (ASO), messenger RNA (mRNA), aptamer, catalytic nucleic acid (CNA), and CRISPR cas9 guide RNA (gRNA). NATs have the potential to achieve precision medicine and long-term treatment because they _target disease-related genes or proteins after being delivered intracellularly [9]. However, the application of NATs requires additional delivery strategies due to their low physiological stability owing to the presence of nucleases and low permeability to cell membranes due to a negatively charged phosphate backbone of nucleic acid [10,11]. These are part of the body’s defense mechanism against invasion of exogenous nucleic acid materials, and sometimes they may induce innate immune responses and lead to serious off-_target effects. To meet these requirements, various viral vector-based and nonviral-based delivery systems, such as DNA nanostructures (DNs), inorganic nanoparticles (INPs), and lipid nanoparticles (LNPs), have been developed that can protect NATs from degradation, isolate them from the external environment, increase cell membrane permeability, and maximize delivery to _target cells.

In recent decades, NATs have developed remarkably along with delivery technologies, and in this review, we would like to introduce the current status and recent advances of not only NATs but also the delivery technologies for anti-cancer treatment and discuss the tasks and prospects that lie ahead. The schematic image in Figure 1 summarizes the contents to be introduced in this review.

1.2. Nucleic Acid Therapeutics (NATs)

Since nucleic acids play a role in storing and transmitting genetic information within living organisms, the concept of regulating the expression of genes through the involvement of exogenous nucleic acid molecules has been around since the 1970s [12]. In the 1980s, the presence of ASO in prokaryotic cells revealed that nucleic acid-mediated gene regulation is a naturally occurring mechanism within cells, which gave credence to the development of NATs [13]. Several years later, the discovery that ribozymes, which recognize and cleave RNA at specific base sequences, act as effective gene regulators provided a major advance to the study of NATs as therapeutics [14]. Since then, research on NATs has made a lot of progress, and various forms of DNA and RNA therapeutics to be described below have been developed. NATs are largely divided into four modes of action: gene silencing, protein inhibition, gene editing, and gene expression (Figure 1). Although in vivo application of NATs requires additional intracellular delivery systems, the following sections describe NATs and delivery systems separately. First, we describe the mechanism of action and _target genes of each NAT, and then we describe four NAT delivery systems: viral vectors, DNs, INPs, and LNPs.

2.1. RNA Interference (RNAi)

RNAi is an intracellular negative feedback process induced by double-stranded RNA, causing the degradation of specific mRNA _targets, which are defense mechanisms against external nucleic acid invasion [16,17]. siRNAs and miRNAs are non-coding RNAs that mediate RNAi and have received great attention for their high potential as therapeutics for various diseases [18]. As shown in Figure 2A, both siRNA and miRNA have RNAi activity in combination with the Argonaute protein (Ago), a component of the RNA-induced silencing complex (RISC). Although siRNA and miRNA have similar functions involved in RNAi in cytoplasm, they are of different origins. The differences between siRNA and miRNA are that siRNA is an externally derived exogenous RNA, while miRNA is an endogenous RNA derived from the genome. Additionally, siRNA inhibits a perfectly complementary mRNA, resulting in less off-_target effects, while miRNA inhibits multiple mRNAs that are partially complementary to the seed sequence [7,18,19]. Naturally occurring siRNAs are derived from exogenous long dsRNAs, and long dsRNAs are cleaved in the cytoplasm by the riboendonuclease Dicer to produce siRNAs with 21 to 23 base pairs [20]. After double-stranded siRNA is loaded onto Ago2 within the RISC, one strand is selected as the guide RNA, and the other strand, the passenger RNA, is cleaved and released. In this process, the RNA strand loaded onto the Ago2 is selected based on the difference in thermodynamic stability at the ends of the two small RNA strands [21]. Synthetic siRNAs can be treated as single-stranded or double-stranded; single-stranded siRNA is unstable and thus has a relatively low RNAi effect but exhibits RNAi activity quickly, whereas double-stranded siRNA exhibits RNAi activity relatively late but exhibits a longer RNAi effect with higher efficiency [22]. Although the double-stranded siRNA has higher physiological stability, there is also a concern that off-_target effects may occur, in which mRNA expression is suppressed by the passenger strand [23]. The approaches to anti-cancer therapy with synthetic siRNA can be carried out by internalizing synthetic siRNA into the _target cancer cells, thereby silencing the _target mRNA involved in cancer. These siRNA-_target genes are mainly those that promote cell differentiation and proliferation and inhibit apoptosis, which are overexpressed in cancer cells, such as Bcl-2 [24], survivin [25], epidermal growth factor receptor (EGFR) [26], Janus kinase 1 (JAK1) [27], and polo-like kinase 1 (PLK1) [28]. In addition, the JIP1 and CLN3 genes involved in resistance to chemo drugs and the PD-L1 and CTLA-4 genes acting on immune avoidance can also be _targeted by siRNA for anti-cancer therapy [29,30,31].

miRNAs are about 22 nt long RNAs that are synthesized through multiple processing steps in various pathways within living cells [32]. In the major synthesis pathway, miRNAs are first transcribed by RNA polymerase II (RNAP II) from genome to primary miRNA (pri-miRNA) with a double-stranded hairpin-loop structure, typically over 1 kb long. Afterwards, the pri-miRNA undergoes a post-transcriptional modification process, is capped and polyadenylated, and is further processed by the microprocessor complex consisting of Drosha (ribonuclease III enzyme) and DGCR8 (DiGeorge syndrome critical region 8) proteins to form precursor miRNA (pre-miRNA) having a 70–100 bp length with a 3′ end overhang [33]. The pre-miRNA is transported from the nucleus to the cytoplasm by the Exporting 5 protein, and in the cytoplasm, the 5′ and 3′ ends of the pre-miRNA hairpin are cleaved by Dicer, forming a mature miRNA duplex with about 22 bp length [33,34,35]. Among the two strands of the miRNA duplex, the one originating from the 5′ end of the pre-miRNA hairpin is called the 5p strand, and the strand originating from the 3′ end is called the 3p strand. In contrast to siRNA, miRNAs can be loaded onto the Ago protein on both strands derived from duplex, and the selection of the guide strand is influenced by its thermodynamic properties [36]. The 5p and 3p strands _target different mRNAs because they have sequences complementary to each other. When one strand of miRNA duplex is loaded onto the Ago protein, the miRNA duplex is unwound to single strands, and the unloaded strand is released and degraded. Depending on the type of Ago onto which the miRNA is loaded, it either destabilizes the _target mRNA or cleaves the _target mRNA [37]. Anti-cancer therapy approaches with synthetic miRNAs are called miRNA (miR) mimics because they make cancer cells have the same sequence as the inhibited or inactivated tumor suppressor miRNAs. miR mimics suppress cancer cells by mimicking the biological functions of the endogenous tumor inhibitor miRNAs, such as Let-7 [38], miR-29 [39], miR-34 [40], miR-101 [41], miR-126 [42], and miR-497 [43]. Chemically synthesized miR mimics can be treated in a single-stranded form, as needed, but RNAi efficiency is reduced compared to the double-stranded form, as in the case of siRNA [44]. On the other hand, single-stranded miR mimics have relatively high cell membrane permeability compared to the double-stranded miR mimics and may be useful in reducing off-_target effects [45].

Some miRNAs act as oncogenes that _target mRNAs in tumor suppressor genes and are overexpressed in cancer. In this case, anti-miRNA (anti-miR) is used to irreversibly inhibit the gene-silencing activity of these onco-miRs [46]. Therefore, the use of anti-miRs leads to the upregulation of one or more tumor suppressor genes that were being inhibited by the _target onco-miRs, and thus they have a significant potential for anti-cancer therapy. The anti-miR sequence is designed to have a high affinity with the miRNA; thus, it competitively hybridizes to the miRNA–RISC (miRISC) along with the _target mRNA (Figure 3) [47]. The miRISCs hybridized with anti-miR can no longer cleave the _target mRNA, resulting in an upregulation of the _target gene expression. Among various oncogenic miRs, miR-21 in particular is upregulated in various types of cancer and has therefore been widely _targeted in anti-miR-based cancer suppression studies [48,49,50,51]. miR-21 mainly causes tumor development, growth, and metastasis by downregulating the genes related to apoptosis, such as phosphatase and tensin homolog (PTEN) and programmed cell death protein 4 (PDCD4), as well as tumor growth and metastasis inhibitory genes, such as reversion-inducing cysteine-rich protein with Kazal motifs (RECK) and tropomyosin 1 (TPM1) [52,53]. In addition, miR-155 [54], miR-191 [55], miR-203 [56], miR-214 [57], and miR-221 [58] are also known to downregulate major tumor suppressor genes, which makes them a promising _target for anti-miR therapeutics. Since one miRNA can effectively regulate the expression of multiple _target proteins, the delivery of miR mimics related to tumor suppression or anti-miRs that block tumor-induced miRNAs has great potential for anti-cancer therapy.

2.2. Antisense Oligonucleotide (ASO)

ASO is a short, single-stranded deoxyribonucleotide with 12~24 nt in length that hybridizes by Watson–Crick base pairing with specific _target mRNAs with complementary sequences. Upon being introduced into the cytoplasm and nucleus, ASO inhibits the translation of _target mRNA, mainly through RNase H-mediated cleavage [59,60]. ASO is introduced as a single strand and hybridized to the _target RNA, resulting in the formation of an RNA–DNA heterodimer, which is cleaved by RNase H (Figure 2B). RNase H is an endonuclease that is distributed in both the nucleus and cytoplasm and specifically recognizes RNA–DNA heteroduplexes and cleaves the RNA strand selectively. ASO, with a chemically modified sugar backbone, such as 2′-O-methyl (2′-OMe) or 2′-O-methoxyethyl (2′-MOE), inhibits the translation of _target mRNA through steric hindrance because it interferes with the binding and cleavage of RNase H to the RNA–DNA heteroduplex [61,62,63]. These chemical modifications provide additional advantages in increasing the physiological stability of ASOs. Similar to siRNA, ASO _target genes are involved in cancer differentiation, growth, proliferation, and survival, such as Bcl-2 [64], survivin [65], EGFR [66], signal transducer and activator of transcription 3 (STAT3) [67], and PD-L1 [68]. Gene silencing via ASOs has an advantage over siRNAs in that it can _target precursor mRNA (pre-mRNA) or long non-coding RNA (lncRNA) within the nucleus [69,70]. lncRNAs are long non-coding RNAs with a length of greater than 200 nt involved in the regulation of gene expression through various biological mechanisms and are clearly distinguished from the highly conserved short non-coding RNAs, such as miRNAs and siRNAs [71,72]. Therefore, abnormally expressed lncRNAs are accompanied by DNA damage, immune evasion, and cellular metabolic disorders, which contribute to various diseases, including cancer [73,74,75]. Selective inhibition of abnormally expressed lncRNA via ASO is an effective approach to cancer treatment [76]. Meanwhile, pre-mRNA is one of the primary transcripts transcribed from a gene and is an mRNA that contains introns before the splicing process [77]. Sometimes, to increase the expression rate of proteins required for the disease treatment, ASOs that cause steric hindrance are hybridized to the splice site of pre-mRNA to regulate the splicing process or to the upstream open reading frames (uORFs) of mRNA [78,79,80]. In summary, this therapeutic approach uses ASO to treat cancer by suppressing oncogenic genes or increasing the expression of tumor suppressor genes by _targeting various _target RNAs present in the cytoplasm and nucleus.

2.3. Catalytic Nucleic Acid (CNA)

CNAs are short single-stranded RNA (ribozyme) or DNA (DNAzyme) molecules with biological catalytic functions, such as RNA cleavage, in combination with a metal cofactor (Figure 2C) [81,82]. A CNA-based gene-silencing strategy has the advantage of not interfering with the endogenous RNAi pathway even when treated at high concentrations because it does not rely on endogenous protein enzymes. In addition, CNAs lacking catalytic activity sometimes exhibit the gene-silencing activity through the antisense effect [83,84]. Before the discovery of CNAs, it was thought that all enzymatic processes within cells were protein-based. This assumption has changed since ribozyme was discovered in the early 1980s, in which the self-splicing RNA or the catalytic site of RNase P was an RNA moiety [85,86]. In contrast to ribozymes found in nature, DNAzymes, which first appeared in the 1990s, are selected using in vitro selection and chemically synthesized [87]. The in vitro selection process prepares a DNA library containing 10^13–10¹⁶ random sequences and incubates them with the _target under the applicable conditions. In the next step, nonspecifically reacting sequences are removed, and sequences that react specifically are recovered and amplified. This process is repeated for 5–15 rounds to enrich the winner DNA strands and perform sequence analysis. Finally, the sensitivity and selectivity between the winner DNA strands are compared to select the desired DNAzyme. Ribozymes can be found not only in nature but also synthesized through in vitro selection [88]. In addition to CNAs that catalyze RNA cleavage reactions, there are CNAs with various other functions, such as alkylation, ligation, and oxidation, but this review will focus on the RNA-cleaving CNAs used in gene silencing [88,89].

Among ribozymes, the hammerhead ribozyme (HHR) is the most extensively studied in gene therapy. The activated structure of HHR consists of a conserved catalytic core and three helixes surrounding it, out of which two helixes, helix I and III, are formed by hybridization with the _target RNA (Figure 4A) [90]. The advantages of HHR as a gene-silencing agent are that the sequences of helix I and III can be easily controlled without inhibiting the activity of the ribozyme, and the total length of the ribozyme is 35 nucleotides, making it easy to chemically synthesize [91]. HHR recognizes and binds to a _target RNA with a sequence complementary to the binding site and cleaves specific sites of _target RNA through phosphate diester isomerization. Divalent metal cations are not essential for the catalytic reaction of hammerhead ribozyme but have been reported to stabilize the activated structure [91]. Similar to other gene-silencing agents, _targets of ribozyme for anti-cancer therapy are genes that are cancer-inducing, such as mutant K-ras [92] and fibroblast growth factor-2 (FGF-2) [93], or important for cancer survival, such as γ-Glutamylcysteine synthetase (γ-GCS) [94] and multiple drug resistance 1 (MDR-1) [95]. Sometimes, ribozymes recognize cancer biomarkers, such as mutant p53 protein or human telomerase reverse transcriptase (hTERT) mRNA, and then activate cancer-killing genes, such as diphtheria toxin (DT) or herpes simplex virus thymidine kinase (HSVtk), that are delivered together to kill cancer cells directly [96,97].

RNA-cleaving DNAzymes (RCDs) are suitable for use as gene-silencing agents because they are physiologically stable compared to ribozymes, can cleave _target mRNA more precisely, and are cost-effective to synthesize compared to other nucleic acid therapeutics [98]. Similar to ribozyme, RCDs have binding arms that can bind to substrate RNA on both sides of the conserved catalytic core (Figure 4B). RCDs cleave the substrate RNA bound to the binding arm through a transesterification reaction with the assistance of a divalent metal cation bound as a cofactor to the catalytic core [99]. The 10–23 RCD is suitable for intracellular gene silencing because it can be designed to cleave almost all _target mRNAs with purine–pyrimidine (R-Y) junctions utilizing Mg²⁺ ions as cofactors with little effect on catalytic function [100,101,102]. Anti-cancer strategies using RCDs work by _targeting genes involved in cancer differentiation, growth, and proliferation, such as Bcl-2 [103], survivin [104], early growth response protein 1 (EGR1) [105], and urokinase plasminogen activator surface receptor (uPAR) [106], or genes involved in cancer survival, such as PD-L1 [107], c-jun [108], and insulin-like growth factor II promoter 3 (IGF-IIP3) [109]. RCDs have a unique advantage over other gene-silencing agents in that their gene-silencing efficacy can be improved by adjusting the length of the binding arm, chemical modification, or using a combination of multiple RCDs [110,111,112,113]. In addition, RCDs are able to activate the catalytic core by oncogenic mRNA to simultaneously achieve down-regulation of the _target mRNA, additional anti-cancer strategies, and diagnosis of cancer [114,115]. However, since CNAs are less efficient in gene silencing compared to other endogenous gene-silencing pathways, they have recently been mainly used as a trigger for drug release or as a signal generation tool for disease diagnosis.

To summarize, anti-cancer therapies through gene silencing introduced above commonly go through the following steps: (1) identifying genes involved in cancer development, (2) designing and synthesizing NATs that specifically inhibit cancer-related genes, and (3) safely delivering the synthesized NATs to the cytoplasm of _target cells. _target genes include genes related to cancer growth, differentiation and metastasis, angiogenesis, immune avoidance, and drug resistance.

3.1. Aptamer-Based Protein Inhibition

The anti-cancer strategy of interfering with the biological function of a _target protein involves inhibiting the growth, metastasis, and invasion of cancer by binding to growth factor receptors or chemokine receptors to block cell signaling [126]. Platelet-derived growth factor (PDGF) is one of the growth factors that regulates cell growth and division. Since it especially regulates angiogenesis, which is essential for cancer growth, the overexpression of PDGF is found in various cancers. Therefore, PDGF was considered a promising _target for cancer treatment, and in fact, blockade of the interaction between PDGF and PDGF receptor through aptamer resulted in the inhibition of colon cancer proliferation [127,128]. CXC motif chemokine 12 (CXCL12) is a chemokine protein, also known as stromal cell-derived factor 1 (SDF-1), that provides cancer drug resistance and immune checkpoint inhibitor resistance. NOX-A12 is an aptamer for CXCL12, which binds to CXCL12 and interferes with its interaction with CXCR4 and CXCR7 receptors [129]. When NOX-A12 was used in combination with radiotherapy, immune checkpoint inhibitors, or cytotoxic drugs, effective inhibition of cancer growth was observed [129,130,131]. The overexpression of the human epidermal growth factor receptor (HER2) is also associated with cancer growth and survival, and tumor suppression can be achieved through protein interference with aptamers _targeting HER2 [132,133]. The protein inhibition efficiency of these aptamers eventually depends on the affinity between the aptamers and _target proteins; hence, multiplexing of aptamers has been shown to be more effective in inhibiting surface proteins, such as HER2, PDGF, and epithelial cell adhesion molecules (EpCAM) (Figure 5A) [133,134].

In addition to strategies in which aptamers bind to _target proteins and interfere with their interactions, bispecific aptamers can be utilized to control cell signaling by regulating the spatial arrangement of cell surface receptors. This strategy allows for more precise _targeting because it _targets multiple types of cancer-related proteins simultaneously, thereby minimizing any off-_target effects [135]. The mesenchymal epithelial transition (Met) receptor is considered one of the _targets for preventing cancer metastasis because it is known to be involved in cell growth and specifically in the metastasis of cancer cells. Studies have reported that pairing the Met receptor aptamer with a cancer cell biomarker, CD71 or Transferrin receptor (TfR), using a bispecific aptamer interferes with the interaction between the Met receptor and hepatocyte growth factor (HGF), thereby inhibiting cancer cell metastasis [136,137]. Han et al. reported a platform technique for inducing lysosomal degradation of _target proteins by utilizing bispecific aptamers that simultaneously _target Met receptors and lysosome-shuttling receptors (IGF-IIRs) (Figure 5B) [138]. Hoshiyama et al. achieved the inhibition of tumor growth signaling through aptamer-mediated cleavage using a bispecific aptamer that simultaneously _targets fibroblast growth factor receptor 1 (FGFR1) and thrombin, a protease that selectively cleaves Arg–Gly bonds (Figure 5C) [139].

3.2. Aptamer-Based Immunotherapy

Immunotherapy is a treatment for cancer that stimulates the immune system to induce immune cells to attack cancer cells, unlike conventional cancer treatments that directly attack cancer [140]. When cancer cells appear, they are recognized by lymphocytes, including natural killer (NK) cells and T cells, and these immune cells kill cancer cells by secreting cytokines, such as Interferon-γ, and cytotoxic proteins, such as perforin, which is called tumor immune surveillance [141]. Cancer cells actively resist immune cells and ultimately escape immune surveillance through an immune editing process in which cancer cells with low immunogenicity survive and grow [142]. Immunotherapy is an anti-cancer therapy that induces cancer cell death by immune cells through neutralizing the immune suppression or immune evasion mechanisms of cancer cells or enhancing the tumor recognition ability of immune cells [143]. The major aptamer-based anti-cancer therapeutic strategies can be divided into two categories: (1) aptamers that inhibit immune checkpoints of cancer cells and (2) aptamers that improve the ability of immune cells to _target cancer cells.

The immune checkpoint is an immunosuppressive pathway that regulates the antigen recognition of immune cells to prevent the indiscriminate destruction of normal cells by excessive immune responses and is an important regulator of the immune system. PD-1/PD-L1 and CTLA-4/CD80(86) are major _targets of aptamers for immune checkpoint inhibition, and immune checkpoints in cancer cells are well-known pathways for cancer to evade immune surveillance [144]. Programmed cell death 1 (PD-1) is expressed on lymphocytes, including T cells and NK cells, and inactivates lymphocytes when it interacts with PD-L1 of cancer cells. Cytotoxic T lymphocyte antigen-4 (CTLA-4) is also a T cell surface receptor that suppresses immune responses when bound to CD80 or CD86. Therefore, blocking of the PD-1/PD-L1 or CTLA-4/CD80(86) interaction via aptamer upregulates inflammatory cytokine secretion and inhibits tumor growth (Figure 6A) [145,146,147].

Immunomodulating bispecific aptamers that simultaneously _target cancer cells and immune cells have the effect of recruiting immune cells around the tumor cells and accumulating them at the tumor site (Figure 6B). Fcγ receptor III (CD16) of NK cells and L-selectin (CD62L) of T cells can be _targets of aptamers for recruiting immune cells, and membrane receptors overexpressed in cancer cells, such as PTK7 [148] and c-Met [149] or PD-L1 [150], an immune checkpoint, can be _targets of aptamers for binding to cancer cells. _targeting PD-L1 can inhibit the immune avoidance pathway of cancer cells, further increasing the immunotherapy effect of bispecific aptamers. These immunomodulating bispecific aptamers can be categorized with [m + n] nomenclature, where [m] represents the valency of the tumor _target aptamer and [n] represents the valency of the immune cell _target aptamer, and higher order valencies indicate not only higher acquired affinity up to a specific upper limit but additionally acquired resistance to nucleases through steric hindrance [151].

3.3. Aptamer-Based _targeted Therapy

The molecular recognition properties of aptamers to _target specific _target proteins can be combined with other therapeutic elements and used in cancer-_targeting drug delivery systems. Aptamers, which can be chemically synthesized, are attractive candidates for tumor-_targeting therapy as an alternative to conventional antibodies because they can be produced inexpensively and with high reproducibility, and they facilitate the functionalization of therapeutic elements at desired locations [152]. Conventional chemotherapy and radiotherapy, which have problems, such as short half-life, high side effects, immune responses, and insufficient therapeutic effects, can have a greatly improved therapeutic effect through combination with aptamer [125,153]. Lo et al. reported an aptamer-based therapeutic that effectively inhibits hepatocellular carcinoma by conjugating the anti-cancer drug 5-fluorouracil (5-FU) to an aptamer that dually _targets CD44E/s, a surface glycoprotein overexpressed in cancer cells (Figure 7A) [154]. Jeong et al. developed an aptamer-based therapeutic that selectively inhibits breast cancer by conjugating maytansinoid (DM1), a selective microtubule inhibitor, to an aptamer _targeting HER2 [155]. Doxorubicin (Dox) is one of the most widely used anti-cancer drugs for various solid and metastatic tumors, especially due to its intercalation properties in the nucleic acid duplex; it damages the genes in the nucleus to induce apoptosis of cancer cells, and when used in combination with NATs, it has the advantage that no additional bonding process is required [156]. Yao et al. reported an aptamer-based therapeutic that can effectively _target colorectal cancer by self-assembling four AS1411 aptamers that _target nucleolin overexpressed in cancer and intercalating Dox [157]. Tan et al. reported an aptamer-based breast cancer treatment that combines a chemosensitizer and an anti-cancer agent by co-delivering a peptide that specifically inhibits heat shock protein 70 (HSP70), a cytoprotective protein overexpressed in cancer, an aptamer that _targets the oncoprotein mucin-1 (MUC-1), and Dox [158]. Omer et al. used trimers of A9g aptamer _targeting PSMA to deliver Dox, thereby increasing the affinity for _target cancer cells and resulting in higher efficacy in anti-cancer effects than when using A9g aptamer monomers [159]. Aptamer is also conjugated with radiosensitizers, such as 1,10-Phenanthroline and gold nanoclusters, increasing the radiation sensitivity of tumors and enabling radiation therapy with fewer side effects [160,161].

In addition to these chemo drugs or metal nanoclusters, aptamers can be used with other NATs, such as siRNAs, ASOs, or DNA genes. In particular, aptamers have the advantage that no additional process is required for conjugation with other NATs because they consist of nucleic acid. Dassie et al. reported significant inhibition of PSMA-positive tumors when treated with a conjugating RNA aptamer _targeting prostate-specific membrane antigens (PSMA), specifically expressed in prostate cancer cells with siRNA _targeting PLK1 mRNA [162]. Wang et al. conjugated the AS1411 aptamer that _targets nucleolin and DNAzyme that recognizes and cleaves survivin mRNA and introduced a pair of fluorophore and quencher at both ends of the DNAzyme to enable simultaneous silencing and detection of tumor genes (Figure 7B) [163]. Strategies to attach two or more NATs to aptamers have also been attempted. Liu et al. reported a chimeric structure that combined an RNA aptamer _targeting PSMA with two siRNAs _targeting survivin and EGFR, demonstrating the therapeutic effect of the combined siRNA [164]. Hong et al. reported a chimeric construct in which two ASOs were introduced into the AS1411 aptamer by conjugating an ASO _targeting the oncogenic gene galectin-1 and a G-rich sequence that can form AS1411 through G-quadruplex (G4) interaction [165]. In addition, anti-miR, which blocks onco-miR, is conjugated to an aptamer _targeting cancer-related membrane proteins, such as EGFR and CD133 and is used as a _target anti-cancer treatment [48,49].

5.1. Plasmid DNA (pDNA)

Plasmid DNA (pDNA) is a small circular DNA molecule found in bacteria that replicates independently of the genome and expresses genes. For therapeutic utilization of pDNA, the optimization of the plasmid backbone, including the insertion site of the desired genetic information and the arrangement of key elements, such as promoters and internal ribosome entry site (IRES), is important, which affects the immunogenicity and therapeutic effects of pDNA [187]. Transfecting the protein-encoding plasmid DNA, which has a direct anti-cancer effect, into _target cancer cells can effectively suppress cancer cells (see (1) in Figure 9). Candidate proteins with such anti-cancer effects include proapoptotic proteins [189]; tumor suppressor proteins, such as p59 [190], PTEN [191], and PDCD4 [192]; cytotoxic proteins, such as saporin (SAP) [193]; proteins that enhance immunogenicity, such as RNA replicase [194]; bacterial toxins, such as diphtheria toxin [195]; and inhibitors of cancer-related receptors, such as HER2 antibody [196] and PSMA antibody [197]. Another anti-cancer strategy utilizing pDNA can induce the destruction of the cancer with the immune system by generating specific proteins called DNA cancer vaccines (see (2) in Figure 9). DNA vaccines are considered safer and more effective treatment strategies than conventional vaccines because they directly transmit the genetic information of antigens [198]. DNA vaccines use pDNA that encodes cytokines, such as interferons (INF-β, INF-γ) [199] and tumor necrosis factor-α (TNF-α) [200]; antibodies to block immune checkpoints, such as PD-L1 [201] and PD-1 [202]; and antigens, such as ovalbumin (OVA) [199], TERT [203], and glycoprotein-100 (gp-100) [204]. Next, pDNA therapeutics can be designed to express RNA-based NATs through their property to transcribe RNA after transfection (see (3) in Figure 9). short hairpin RNA (shRNA), which has an RNAi mechanism similar to siRNA, is one of the RNA-based NATs transcribed in the cell nucleus from pDNA and consists of 19–22 bp double-stranded RNA linked by short hairpins [205]. The shRNA is treated with siRNA by dicer in the cytoplasm and subsequently complexed with RISC to exhibit gene-silencing activity. pDNA encoding shRNA that silences genes involved in cancer survival and growth, such as AKT Serine/Threonine Kinase 1 (Akt1) [192], Vascular Endothelial Growth Factor (VEGF) [206], and EGFR [207], exhibits anti-cancer effects. Additionally, pDNA encoding both Cas9 protein and gRNA is a very simple and convenient strategy that avoids repetitive transfection of each component of the Cas9 system [170]. Zhao et al. achieved permanent immune checkpoint inhibition by encoding gRNA _targeting the PD-L1 gene and Cas9 proteins in pDNA and observed the resultant activation of T-cell-mediated anti-tumor responses [208]. However, these strategies raise concerns that the long-lasting activity of Cas9 may increase off-_target effects compared to strategies that directly deliver Cas9-gRNA complexes [209].

5.2. Messenger RNA (mRNA)

Messenger RNA (mRNA) is a type of RNA that carries genetic information that can direct the synthesis of proteins, and therefore it is the _target of many NATs to silence genes. However, mRNA encoding for therapeutic proteins can be used as therapeutics. Unlike pDNA therapeutics, mRNA therapeutics do not need to enter the nucleus to function, making transfection relatively easy and safe [210]. In addition, mRNA therapeutics have great potential for personalized medicine because they can be produced at lower production costs in a shorter period of time than pDNA therapeutics [211]. mRNA therapeutics are considered an alternative to pDNA therapeutics, which have low transfection efficiency and can cause permanent side effects [212]. mRNA therapeutics production involves multiple unit operations, including pDNA preparation, pDNA linearization, in vitro mRNA transcription (IVT) reaction, pDNA digestion, and mRNA purification [213]. The completed mRNA therapeutic contains a 5′ cap, a 3′ poly adenine (A) tail at both ends to enhance mRNA physiological stability and protein translation efficiency, and a 5′ untranslated region (UTR), a gene of interest (GOI), and a 3′ UTR are located between them. Similar to pDNA, protein-encoded mRNAs with anti-cancer effects induce cancer cell apoptosis (see (1) in Figure 9). mRNA encoding tumor inhibitors, such as P53 [214], PTEN [215], tuberous sclerosis complex 2 (TSC2) [216], mothers against decapentaplegic homolog 4 (SMAD4) [217]; proteins that induce cell membrane disruption, such as mixed lineage kinase domain-like protein (MLKL) [218]; or cancer-related receptor inhibitors, such as HER2 antibody [219], exhibit direct anti-cancer effects. The mRNA cancer vaccine, which induces anti-tumor activity by the immune system, is currently one of the most actively studied treatments [220]. As cancer vaccines, mRNA encoding cytokines, such as Interleukin-12 (IL-12), IL-27 [221], IL-23, IL-36γ [222], and INF-α [223]; antigens, such as OX40L [219], OVA [224], cytokeratin 19 (CK19) [225], and MUC1 [226]; or antibodies that block immune checkpoints, such as anti-PD-1 and anti-PD-L1 [227], can be used (see (2) in Figure 9). In addition, mRNA encoded with the Cas9 protein is co-delivered with gRNA and used for gene editing. Wang et al. reported that the Cas9 protein-encoded mRNA is co-delivered with gRNAs that _target legumain (LGMN) genes, which are known to cause various diseases, including cancer, effectively suppressing the metastatic properties of breast cancer [228].