There is a growing interest in small RNA-based therapeutics including Antisense Oligonucleotides (ASOs), microRNAs (miRNAs or miRs), short interfering RNAs (siRNAs) and aptamers. ASOs, miRNAs and siRNAs act by targeting and inhibiting specific target mRNAs, whereas aptamers are 3D-structured oligonucleotides serving as high-affinity ligands and potential antagonists of disease-associated proteins (Figure 1).
2. Therapeutic RNA Stability and Immunogenicity
A major hurdle for the therapeutic use of RNAs is represented by their instability due to the susceptibility to enzymatic degradation. To overcome this limitation, different chemical modifications to enhance RNA resistance to nucleases have been proposed. Substitutions at the 2′-position of the ribose have been frequently employed to address this issue. These include the introduction of 2′-fluoro, 2′-amino or 2′-O-metyl groups or the use of locked nucleic acids (LNAs) that contain a methylene bridge between the 2′-O to the 4′-C of the sugar. Modifications of the phosphate group, as the inclusion of phosphorothioates, or the capping at the 3′-terminus have been applied to improve RNA resistance to degradation, as well
[13]. In addition to the improvement of the resistance, the use of alternative bases (e.g., pseu-douridines) has been found to ameliorate the stability of the duplex structures, enhanc-ing the functional activity of dsRNA molecules
[14].
Importantly, chemical modifications of RNAs also improve their immunogenic profile. Different pathways, including Toll-like receptor (TLR) 3, 7 and 8, dsR-NA-dependent protein kinase (PKR), retinoid-acid-inducible gene I (RIG-I) and/or melanoma differentiation-associated gene 5 (MDA-5) pathways, physiologically recognize RNA molecules inducing innate immune response with the production of interferon (IFN) and pro-inflammatory cytokines. Small synthetic RNAs can be recognized by these systems and activate innate immunity
[15], but this can be easily prevented through the design of non-immunogenic molecules avoiding the presence of immune-stimulatory motifs
[16][17] and introducing modified nucleotides
[18]. Concerning NSCLC, toxicology studies in rodents and primates have been performed for different small RNAs that are in clinical trials, revealing no controllable immune response activation
[19][20][21]. However, the majority of therapeutically relevant RNA molecules proposed for NSCLC are still at an early preclinical phase of investigation and there are currently data on their effects on the host immune system in immunocompetent models available. Regarding the immunostimulatory properties of some small RNAs (mainly siRNAs and miRNAs), it should be considered that in some specific circumstances, the activation of the immune system represents a remedy for therapy, and the design of new RNA-based agents stimulating the host immune system has been explored for cancer and viral infection treatment
[22].
3. ASOs
Antisense oligonucleotide-based therapy is a captivating approach to cancer treatment. ASOs are small (18-20 nucleotides) single-stranded modified oligonucleotides capable of hybridizing specific target mRNAs through base pairing recognition. The main mechanism of action of ASOs is the formation of hetero-duplexes with the mRNA in the cytoplasm, leading to the activation of the ubiquitous endonuclease RNase H and the hydrolysis of the complexes or to the alteration of the correct ribosomal assembly. Alternatively, ASOs can act by entering the nucleus and regulating mRNA maturation through the inhibition of the 5′ cap formation, of the normal splicing, or through the activation of the RNase H
[23][24][25][26] (
Figure 2).
Figure 2. Main mechanisms of action of ASOs. Schematic representation of the different mechanisms of action of ASOs. ASOs can block protein expression by binding to the mRNA and leading to: (i) the RNase H-mediated mRNA degradation; (ii) the steric block of the correct ribosomal assembly; (iii) the alteration of the normal splicing causing exon skipping or inclusion. Alternatively, they can sequestrate miRNAs avoiding their binding to target mRNAs.
ASOs activating RNase H have generally a specific composition with a central region of minimum of five DNA nucleotides called ‘gapmer’, flanked by RNA sequences, which increase the binding to the target. Steric block ASOs, instead, do not contain the DNA strand core, but include mixed modified nucleotides, such as LNAs, morpholinos, peptide nucleic acids and thiophosphoroamidates
[27].
Antisense strategies have been used for decades with the purpose of downregulating over-expressed genes, but more recently have been employed even in the opposite situation, where the upregulation of therapeutic proteins could lead to a beneficial effect. This new approach is allowed by the development of new ASOs targeting miRNAs (called anti-miRNAs) that avoid miR binding to target mRNAs with a consequent upregulation of the translation. In addition, ASOs could be designed to target the 5′ untranslated region (UTR) of the mRNAs, which is often responsible for translational repression mainly through the formation of stem-loop structures
[28].
Currently, ASOs are among the most advanced small RNA therapeutics in clinical trials, with several molecules approved by the Food and Drug Administration (FDA) for the treatment of different disorders. Among approved ASOs, there are: Fomivirsen for cytomegalovirus retinitis; Mipomersen for homozygous familial hypercholesterolemia; Eteplirsen and Golodirsen for Duchenne muscular dystrophy; Nusinersen for spinal muscular atrophy; Inotersen and Patisiran for polyneuropathy caused by hereditary transthyretin-mediated amyloidosis; Givosiran for acute hepatic porphyria; and Milasen, for neuronal ceroid lipofuscinosis 7
[29].
ASOs have been widely applied to NSCLC therapy by targeting key genes regulators of the malignant phenotype, such as B-cell lymphoma 2 (Bcl2), Protein Kinase B (PKB/Akt-1), Kirsten Rat Sarcoma virus (KRAS), Vascular Endothelial Growth Factor (VEGF), Signal Transducer and Activator of Transcription 3 (STAT3), clusterin and Protein Kinase C alpha (PKC alpha).
KRAS is one of the most mutated genes in different human cancers, including NSCLC
[30]. It is considered a good target for anti-cancer therapies and many KRAS inhibitors are currently under evaluation in clinical trials
[31]. AZD4785 is an ASO containing 2′-4′ constrained ethyl residues (cEt). It targets mutated
KRAS mRNA with high affinity and is able to induce its degradation. This ASO blocks pathways activated by KRAS hampering the proliferation of KRAS-mutated cancer cells
[32]. Ross SJ et al. tested the efficacy of AZD4785 in multiple
xenograft mouse models of NSCLC, demonstrating a good anti-tumor effect upon its systemic injection. In addition, they demonstrated ASO safety in primates, paving the way for the experimentation of its efficacy in a phase 1 clinical trial
[32]. More recently, Wang et al.
[33] reported a novel anti-KRAS ASO with improved antisense activity and pharmacokinetics. They designed a bottlebrush-like complex consisting of a DNA backbone, poly(ethylene glycol) (PEG) side chains and ASO overhanging. In this construct, PEG protects the ASO from enzymatic degradation, while ASO overhangs increase the chance of hybridization with the target. By using NSCLC cell models, authors showed that this molecule allows a higher antisense efficacy than unassembled hairpins and displays an improved retention time in vivo
[33].
The over-expression of Bcl2 and its activator Akt-1 is responsible for the increased cancer cell viability, growth and apoptotic evasion in NSCLC
[34][35]. Specific gapmer-based ASOs, called G3139 and RX-020, have been devised against
Bcl2 and
Akt-1 mRNAs, respectively, and tested in pre-clinical and clinical studies for NSCLC therapy. However, they did not reach success in clinical trials due to improper delivery. In order to improve ASO effectiveness, G3139 and RX-020 have been modified at the 5′ and the 3′ ends with 2′-O-methyl groups, and loaded in lipid NPs
[36]. ASO-NP complexes demonstrated excellent cellular uptake and colloidal stability that let them gain a better anti-tumor effect in
xenograft mouse models. Moreover, G3139 has been also incorporated in DOTAP/egg PC/cholesterol/Tween 80 lipid NPs and tested both in vitro and in vivo in NSCLC models. This complex gives an effective reduction in Bcl2 protein level and cell growth inhibition
[37].
In order to target STAT3, a key transcription factor over-expressed in NSCLC, an ethyl-modified ASO, called AZD9150, has been designed by Hong D et al. and tested in vitro in multiple cell lines
[19]. The ASO was 10–20 times more potent than the earlier generation (Gen 2.0) STAT3 ASOs, showing half-maximal inhibitory concentration (IC50) values for the most sensitive cells in the nanomolar range. Next, AZD9150 gave a strong inhibition of STAT3 in human primary patient-derived
xenograft (PDX) models of NSCLC, colorectal cancer and lymphoma. Most importantly, the authors described a phase I dose-escalation study where AZD9150 antitumor activity as a single agent was observed in patients with highly treatment-refractory lymphomas and NSCLC. Currently, AZD9150 is being studied in a phase I/II trial alone or in combination with other anti-cancer drugs in patients with NSCLC, lymphoma and different advanced solid tumors
[38][39]. In addition, again with the aim of targeting STAT3 in lung cancer, Njatcha and colleagues employed a cyclic 15 nucleotide length decoy called CS3D
[40]. The circular oligonucleotide was obtained through the inclusion of two hexaethyleneglycol spacers that provide flexibility and thermal stability. Effects obtained upon decoy ASO transfection in 201T and H1975 lung cancer cells include the reduction in cell growth and proliferation and the increase in apoptosis.
Another key protein regulating important pro-oncogenic pathways in cancer, which can be therefore considered an interesting therapeutic target, is PKC. A 20-mer phosphorothioate ASO (ISIS 3521) has been synthesized to bind the 3′ UTR region of
PKC alpha mRNA. ISIS 3521 has been tested alone or in combination with chemotherapeutics for the treatment of NSCLC and different refractory cancers
[41]. The ASO entered phase III clinical trials in combination with carboplatin and paclitaxel, or with gemcitabine and cisplatin
[42].
Combined therapeutic regimens in clinical trials for NSCLC include the use of an anti-clusterin ASO. Clusterin is a chaperone heterodimeric glycoprotein upregulated in many cancers, including NSCLC, and involved in the clearance of cellular debris and apoptosis. A 2′-methoxyethyl-modified phosphorothioate ASO (OGX-011) inhibiting the
clusterin gene has been used in different kinds of tumors and is currently in phase I/II trial in combination with cisplatin and gemcitabine, and in phase III in combination with docetaxel for the treatment of advanced NSCLC
[43][44].
An additional interesting application of ASOs in NSCLC was described to counteract angiogenesis. Recent studies have shown that VEGFA is up-regulated in squamous cell carcinoma, due to the overexpression of the long non-coding (lnc) RNA LINC00173.v1, a RNA sponge sequestering miR-511-5p
[45]. Importantly, ASOs inhibiting LINC00173.v1 were developed and tested in mice, alone or in combination with cisplatin. Lung cancer cells were inoculated in mice and one week later, the ASOs were injected intraperitoneally once weekly for 4 weeks. Results showed the ASO ability to effectively inhibit tumor growth, increase mice survival and improve the therapeutic sensitivity to cisplatin
[45].
Self-renewing tumor-initiating cells (TICs) represent a small population within the tumor mass that strongly contribute to tumor progression, recurrence and drug resistance. In order to target this population, Lin J et al. designed innovative splicing-modulating steric hindrance antisense oligonucleotides (shAONs) targeting the glycine decarboxylase (
GLDC) gene, whose protein is a metabolic enzyme overexpressed in TICs of NSCLC, and necessary for their maintenance
[46]. The developed ASOs efficiently induced exon skipping with an IC
50 of 3.5–7 nM, disrupting the
GLDC open reading frame and resulting in the inhibition of cell proliferation and colony formation in A549 cells and NSCLC tumor spheres (TS32). Most importantly, the best shAON strongly inhibited tumor growth in mice
xenografts of TS32.
Recently, the GTP-binding protein ADP-ribosylation factor (ARF)-like (ARL) 4C (
ARL4C) has been explored as a potential therapeutic target for NSCLC treatment. An ASO targeting
ARL4C, called ASO-1316, was synthesized and tested in NSCLC evaluating cell proliferation, migration and tumor sphere formation
[47]. Obtained results demonstrated the ASO-1316 ability to suppress lung cancer cell proliferation and migration in vitro, regardless of KRAS or EGFR mutation.
4. MicroRNAs
MiRNAs are a group of endogenous short ncRNAs (18-25 nucleotides in length) that negatively regulate gene expression at a post-transcriptional level. They have attracted great attention as anticancer therapeutics given to their pivotal role in tumorigenesis
[48]. Indeed, miRNAs are involved in the regulation of numerous metabolic and cellular processes, including cell proliferation, differentiation, migration and survival, and can act as oncosuppressors or oncogenes (oncomiRs)
[49]. One main feature of miRNA function is their ability to target multiple pathways simultaneously, thus offering the possibility to improve the therapeutic efficacy of the treatment, decreasing the occurrence of drug resistance, a frequent event when only one pathway is targeted.
MiRNA biogenesis (Figure 3) starts in the nucleus where the RNA polymerase II transcribes a pri-miRNA molecule, a double-stranded precursor with a polyadenylated stem-loop structure at the 3’ and a cap at the 5’. Two RNases III sequentially process this pri-miRNA: Drosha, which forms the pre-miRNA, that is then translocated into the cytoplasm, and Dicer, which processes the pre-miRNA into a mature miRNA. At this point, the non-functional strand (passenger) in the mature miRNA is degraded, while the functional one (guide strand) is loaded into a complex called RISC (RNA-induced silencing complex) and, based on its complementarity with the target mRNA, mediates mRNA degradation or translation inhibition.
Figure 3. microRNA biogenesis. Main steps of microRNA processing. Briefly, miRNA transcripts (pri-miRNAs) are produced and cleaved by Drosha into pre-miRNAs. The pre-miRNAs are then exported into the cytoplasm by exportin-5–Ran-GTP and further processed into mature miRNA duplexes. The functional strand of the mature miRNA is thus guided by the RISC complex to the target mRNA permitting translational repression or mRNA cleavage.
Depending on the extent of miRNA expression in the tumor, different therapeutic approaches can be considered. When miRNAs are downregulated, they can be replaced by oligonucleotides that mimic their action or by viral vectors that allow miRNA expression to be restored. On the contrary, when miRNAs are over-expressed, they can be inhibited by: (i) synthetic anti-miRNAs, also known as antagomirs, that are ASOs complementary to the miRNA of interest able to induce the duplex formation and miRNA degradation; (ii) miRNA sponges that are competitive inhibitors containing multiple miRNA binding sites, allowing the simultaneous sequestering of multiple miRNAs
[50]. The involvement of miRNAs in lung carcinogenesis has been widely documented by several studies focused on the differential expression pattern of miRNAs between lung tumor tissues and normal tissues
[51].
Of note, despite different clinical trials for miRNA therapeutics are currently ongoing for a variety of cancers, including liver cancer, T-cell lymphoma, mesothelioma and lymphoma, only a few molecules are under clinical evaluation for lung cancer therapy, and only one for NSCLC
[52].
The clinical study performed by van Zandwijk and colleagues involves the use of a miR-16 mimic enclosed in enveloped delivery vehicles (EDV), whose surface was functionalized with antibodies specifically targeting the epidermal growth factor receptor (EGFR), over-expressed in different cancer types, including NSCLC
[53]. These complexes, called “TargomiR” showed encouraging results in a Phase I clinical trial in patients with recurrent malignant pleural mesothelioma and NSCLC.
MiR-21 is commonly over-expressed in solid tumors and targets oncosuppressor genes, such as Phosphatase and TENsin homolog (
PTEN), Programmed Cell Death 4 (
PDCD4) and Tropomyosin 1 (
TPM1), promoting cell growth, metastasis and escape from the apoptosis. The involvement of miR-21 in NSCLC tumorigenesis and drug resistance has been widely studied
[54]. Therefore, its modulation by anti-miRNAs encapsulated in various delivery systems has been investigated and showed promising results. In this regard, an interesting study comes from Wang et al.
[55], who employed a PEG-functionalized nanographene oxide (NGO-PEG) dendrimer to deliver the anti-miR-21 to NSCLC. In order to monitor the delivery efficacy, authors included an activatable luciferase reporter gene, containing three miR-21 complementary sequences in the 3′ UTR, so that miR-21 inhibition results in an increase in the luciferase activity. The complexes effectively delivered the anti-miR-21 into the cytoplasm upregulating luciferase intensity, restoring PTEN expression and inhibiting cell migration and invasion. Moreover, the intravenous administration of the complex in NSCLC mouse xenografts allowed efficient delivery of the anti-miRNA.
MiR-150 has been proposed as a therapeutic target in different cancers, including NSCLC
[56]. In particular, it has been reported to directly regulate the expression of Forkhead-box O 4 (
FOXO4) and enhance tumor cell metastasis in NSCLC, both in vitro and in vivo in mouse
xenograft models
[57].
MiR-221/222 has been found to be upregulated in TNF-related apoptosis-inducing ligand (TRAIL)-resistant NSCLC cells. Its inhibition has been demonstrated to improve TRAIL sensitivity by targeting cyclin-dependent kinase inhibitor 1B (
p27Kip1)
[58]. In addition to its therapeutic potential, miR-221 has been also proposed as a marker for early diagnosis and NSCLC screening
[59].
MiR-96 is a potent oncomiR, which increases NSCLC chemoresistance by reducing the expression of Sterile Alpha Motif Domain Containing 9 (
SAMD9)
[60]. It has been demonstrated that anti-miR-96, in combination with cisplatin, is able to upregulate
SAMD9 expression dramatically enhancing cisplatin-induced apoptosis. Although only in vitro data were provided, the study suggests a new approach to improve response to chemotherapy in NSCLC.
MiR-17/92 cluster includes six family members: miR-17, miR-18a, miR-19a, miR-19b-1, miR-20a and miR-92a. Among them, miR-19 is the leading oncogenic player. The miR-17/92 cluster is typically amplified in lung cancer and targets different genes, such as Hypoxia-inducible factor 1-alpha (
HIF-1α),
PTEN, BCL2-like 11 (
BCL2L11), cyclin dependent kinase inhibitor 2A (
CDKN2A) and Thrombospondin-1 (TSP-1), which are involved in cancer cell death and proliferation
[61]. Matsubara et al. found that ASO-inhibiting miR-17-5p and miR-20a induced apoptosis in cancer cells overexpressing miR-17-92
[62].
Furthermore, several studies described the deregulation of miR-19a in various cancer types, and demonstrated that its upregulation strongly correlates with increased metastasis, invasiveness, and a poor prognosis in NSCLC
[63]. The study conducted by Jing Li and colleagues showed that the upregulation of miR-19 (miR-19a and miR-19b-1) in A549 and HCC827 lung cancer cells triggered EMT transition, which led to mesenchymal-like morphological changes and increased cell migration and invasion. The authors also demonstrated that the miR-19 target, PTEN, had an important role in these processes, which were reverted when the miRNA was silenced
[64].
In addition to over-expressed miRNAs, many others are downregulated, acting as tumor-suppressor genes. Below, researchers discuss some of the most interesting tumor-suppressor miRNAs in NSCLC.
MiR-Let-7 family includes 11 members, among which let-7a, let-7c, and let-7g are frequently downregulated in lung cancer patients. MiR-let-7 regulates the expression of several genes involved in cell proliferation and cell cycle, including cMyc, high-mobility group A (
HMGA), STAT3, Janus-activated kinase 2 (
JAK2), KRAS, Cyclin-Dependent Kinase 6 (
CDK6), Homeobox A9 (
HOXA9), transforming growth factor beta receptor 1 (
TGFBR1), B-cell lymphoma-extra-large (
BCL-XL) and Mitogen-activated protein kinase 3 (
MAP4K3)
[65][66][67][68]. It has been widely demonstrated that transfection of miR-Let-7 family members significantly decreased the proliferation rate of NSCLC cells and improved the sensitivity to chemo and radiotherapy
[69][70][71]. Further, it has been reported that Let-7 reduces tumor formation in vivo, in NSCLC orthotopic models following intranasal administration, modulating
KRAS expression. Notably, an aptamer-based strategy for the targeted delivery of miR-Let-7g has been developed by Esposito CL and colleagues and successfully applied to NSCLC both in vitro and in vivo (see the paragraph on Aptamers for more details)
[72].
MiR-29 family, including miR-29a, 29b and 29c, has been widely studied in NSCLC. A first report showed its ability to revert aberrant DNA methylation in lung cancer, by targeting the de novo DNA methyltransferases (DNMTs) 3A and 3B. This leads to the re-expression of methylation-silenced tumor-suppressor genes, such as fragile histidine triad protein (
FHIT),
CDKN2A, cadherin-13 (
CDH13), cadherin-1 (
CDH1) and Ras Association Domain Family Member (
RASSF1A) both in vitro and in vivo
[73]. Next, Wu et al. demonstrated that miR-29 transfection by cationic lipoplexes significantly reduced the expression of cell division protein kinase 6 (
CDK-6),
DNMT3B and myeloid leukemia cell differentiation protein 1 (
MCL1) and inhibited tumor growth in subcutaneous
xenograft models of A549 NSCLC cells
[74]. More recently, Xing et al. confirmed the therapeutic potential of miR-29, by employing an
N-isopropylacrylamide-modified PEI (namely PEN) as a carrier for miR-29a. The authors demonstrated the complex efficacy in allowing miR-29a delivery and processing with a consequent anti-proliferative and anti-migratory effect on A549 NSCLC cells
[75].
MiR-200 family is known as a negative regulator of the EMT due to its ability to target Zinc finger E-box-binding homeobox (
ZEB) family members, acting as translational repressors that promote EMT
[76]. MiR-200 decreases tumor survival by reducing the expression of
VEGF and
VEGF-R1 [77]. In a mouse model of lung adenocarcinoma (KrasLSL-G12D/+; Trp53flox/flox), metastasis was significantly increased by the deletion of miR-200c/141, resulting in a desmoplastic tumor stroma
[78]. Within the stroma, the NOTCH signaling pathway is promoted by the downregulation of miR-200 in cancer-associated fibroblasts, improving the ability of cancer cells to metastasize
[78]. In a group of NSCLC patients with adenocarcinoma, Tejero and colleagues recently showed that high levels of miR-141 and miR-200c are associated with shorter overall survival
[79]. In addition, Cortez and co-workers have shown that miR-200c promotes radiosensitivity of lung cancer cells by controlling the oxidative stress response through the direct regulation of Peroxiredoxin 2 (
PRDX2), GA-binding protein (
GAPB/Nrf2) and Sestrin 1 (
SESN1), and by preventing the repair of radiation-induced double-strand breaks
[80]. The therapeutic potential of miR-200c in combination with other treatments, such as radiation, was also demonstrated in vivo.
MiR-34 family has been found to be deregulated in several human malignancies, and is considered a tumor-suppressive miRNA group due to its synergistic effect with the tumor suppressor gene
p53 [81][82]. The miR-34 family has been shown to inhibit the translation of genes involved in cell growth, proliferation, cell cycle control and anti-apoptotic signaling
[83]. In addition, several preclinical studies have shown the big potential of miR-34a in cancer treatment
[84]. Kasinski et al. showed that miR-34a is a valuable therapeutic option for NSCLC treatment. They concluded that the administration of miR-34a to lung cancer patients with KRAS -positive and p53-negative tumors, slows tumor growth and potentially increases survival
[85]. Notably, miR-34a was the first miRNA restoration therapy investigated in a clinical trial. Although the trial was ultimately unsuccessful
[86], miR-34a is expected to be included in further miRNA mimic studies due to its potent tumor-suppressive effect in a variety of tumor subtypes.