1. Introduction
The advancement of antisense oligonucleotides (ASOs) has brought about a profound change in the field of genetic therapeutics, offering a promising avenue for addressing a diverse array of diseases on a molecular level. ASOs are short synthetic nucleic acid analogs that offer a revolutionary means to modulate gene expression by precisely interacting with RNA transcripts. The history of ASO can be traced back to the pioneering work of Zamecnik and Stephenson in early 1970, who first proposed the concept of using synthetic oligonucleotides to regulate eukaryotic gene expression in cultured cells through sequence-specific hybridization with RNA [
1,
2]. Later, the pharmacokinetic properties of ASOs, such as stability, reduced susceptibility to nuclease degradation, specificity, and cellular absorption have been greatly improved by developments in oligonucleotide chemistry, including the introduction of chemical modifications and different backbone structures which transformed them from theoretical concepts into potentially effective therapeutic agents [
3].
ASOs have been successfully employed in treating a wide range of diseases, including Duchenne muscular dystrophy (DMD), spinal muscular atrophy (SMA), amyotrophic lateral sclerosis (ALS), and many more which led to the regulatory approval of 10 ASO-based drugs till now [
4] and many antisense drug candidates to clinical trials to treat cardiovascular, metabolic, endocrine, neurological, neuromuscular, inflammatory, and infectious diseases [
5]. This demonstrates the dynamic nature of ASO-mediated therapy. Despite being a promising approach, it is widely accepted that delivery of ASO treatments to specific tissues is limited by factors such as intracellular trafficking, degradation in biological fluids, and transportation across cellular barriers [
6]. Although chemical modifications have improved their metabolic stabilities significantly as well as their affinities for RNA targets and have to some extent reduced off-target effects, no chemical modification has significantly improved cellular uptake or tissue targeting.
Cell-penetrating peptides (CPPs) or peptide transduction domains (PTDs) are one of the many approaches that have been developed to improve the delivery of oligonucleotides. CPPs are small peptides with the ability to transport cargos, including ASOs across cellular barriers and hereby offer the potential to improve ASOs’ cellular uptake and intracellular distribution, enhancing the therapeutic outcomes and reducing the required dosage [
7]. The first CPP was introduced in 1946 and since then there has been a continuous effort of developing a more efficient cell-penetrating peptide that can ensure increased delivery of oligonucleotides and better pharmacological properties [
8].
Particularly in the context of phosphorodiamidate morpholino oligomers (PMOs), R6G, PiP (PNA/PMO Internalizing Peptides) and DG9 have captured interest among the CPPs for their potential to improve ASO-mediated therapy. PMOs have shown effectiveness in treating genetic diseases, but their poor cellular absorption continues to be a major drawback. Due to the high efficacy and low toxicity, DG9 has become a promising CPP for improving the intracellular transport of PMOs since it holds the prospect of improved therapeutic advantages [
9,
10]. This review offers a thorough analysis of ASO therapies and the difficulties they encounter, highlighting the potential contribution of CPPs, particularly DG9, to overcoming these difficulties and improving ASO efficacy. Through an exploration of CPP-mediated ASO delivery intricacies and focusing on the remarkable properties of DG9, this review seeks to highlight the potential of this approach to transform ASO-mediated therapy more effectively.
5. Challenges associated with ASOs delivery
Although antisense oligonucleotides (ASOs) have great potential as therapeutic agents, their efficient delivery faces several difficulties. These difficulties are associated with the physiochemical characteristics of ASO molecules, such as their large size, molecular weight (single-stranded ASOs are ~4–10 kDa, double-stranded siRNAs are ~14 kDa), and negative charge which hinders passive diffusion across the cell membrane. ASOs predominantly rely on endocytosis for cellular uptake which might be ineffective and lead to entrapment in endosomes or lysosomes, leading to lysosomal degradation. So, once inside the cell, ASO must escape endosomal entrapment to gain access to the target region in the cytoplasm or nucleus [
27]. Apart from that, for the systemically administered ASOs to be effective, it needs to avoid renal clearance [
28,
29], resist nuclease degradation both in extracellular fluid and intracellular compartment [
30] and avoid removal by the reticuloendothelial system, which includes mononuclear phagocytes, liver sinusoidal endothelial cells, and Kupffer cells [
31]. A study reported that intravenous administration of an AON resulted in 40% and 18% accumulation in the liver and kidneys respectively [
32]. Recently ASOs are also being developed for the treatment of Central Nervous System (CNS) related diseases. The additional barrier, in this case, ASOs have to cross the blood-brain barrier (BBB) or brain–cerebrospinal barrier, before they can distribute within the CNS. The vascular barriers of the nervous system are comprised of a monolayer of endothelial cells forming tight junctions through interactions of cell adhesion molecules which prevents most of the ASOs to reach the CNS after systemic injection [
33] (
Figure 2).
Due to these challenges, to date, most of the approved oligonucleotide treatments are delivered either locally (for example- to the eye or spinal cord) or to the liver. The eye is chosen as a target for ASO delivery (for example- Pegaptanib and Fomivirsen) due to its accessibility, anatomical considerations, and immune-privileged status [
12]. Although ocular delivery of ASOs has benefits, there are still obstacles to be overcome, including getting through anatomical obstacles (such as the blood-retinal barrier), maximizing ASO stability, and pharmacokinetics for long-lasting therapeutic effects. For ASOs targeting the CNS, direct delivery into the cerebrospinal fluid via lumber puncture is most commonly used (for example-Nusinersen) [
34]. It should be noted, however, that this method requires expertise, and specialized equipment, and carries a small risk of complications associated with invasive procedures.
7. Overcoming the limitations of PMO by conjugating it with Cell Penetrating Peptides
As mentioned earlier, PMOs shows low efficacy as therapeutic agent due to their poor cellular uptake, less permeability of membrane barriers, rapid clearance from the systemic circulation, inability to cross blood-brain barriers, and the requirement of repetitive administration and/or high dosage of the drug for executing its function. Apart from that, due to the hydrophobicity of the plasma membrane and the neutral charge of PMO, only small portions of internalized PMOs can escape endosomes and reach their intended target [
83]. A promising utilization of CPP is their ability to directly conjugate with neutrally charged PMO and PNA and increase the delivery efficacy [
109,
110,
111].
A promising utilization of CPP is their ability to directly conjugate with neutrally charged PMO and PNA by using several methods, including maleimide linkage, disulfide linkage, click chemistry or amide linkage, and enhances the pharmacokinetic properties of PMO and PNA. Among various therapeutic purposes, this approach has been extensively explored most for Duchenne muscular dystrophy (DMD), which affects approximately 1 in 3500 newborn boys and is caused by out-of-frame deletions in the
Dmd gene resulting in the loss of dystrophin, the structural muscle protein [
112]. Lack of dystrophin results in progressive muscular degeneration, which impairs ambulation and causes mortality from cardiac and respiratory failure [
113]. The mRNA reading frame around the deletion can be restored by the "exon-skipping" approach where pre-mRNA splicing is modulated to produce smaller but functional proteins. This approach has been used successfully with naked PMO that resulted in the conditional approval of four PMO-based drugs for DMD (e.g.- eteplirsen, golodirsen viltolarsen, and casamirsen) [
73,
74]. Eteplirsen has been found to restore an average of 0.9% dystrophin of normal levels after 180 weeks of treatment which indicates low treatment efficacy despite the safe profile of this drug [
102]. For Golodirsen, according to the trial results dystrophin expression increases by ~0.9% after the demonstration of the drug [
73], and the casimersen-treated group saw a 0.81% increase in dystrophin production [
114]. Whether such a tiny increase in dystrophin expression is enough to slow down disease progression and provide clinical benefits, is still a big question. Apart from that Vitolarsen has limited efficacy in cardiac tissue due to poor uptake. As the primary cause of mortality in the DMD patient population is cardiorespiratory complications, the low efficacy of the drug in the heart is a serious concern in exon-skipping therapy [
115].
Therefore, there is still a need for a more potent substance to raise dystrophin levels and thereby maximize the functional advantages of this strategy.
Conjugation of CPPs to PMO is one such approach to improve PMO delivery. This strategy was first demonstrated with an arginine-rich peptide, (RXR)4 which was administered to the
mdx mouse model of DMD in a variety of doses, time intervals, and delivery methods and it was observed that a single intravenous administration can high dystrophin exon skipping in skeletal muscle, the diaphragm, and for the first time in the heart [
116]. Another arginine-rich peptide, (RXRRBR)2 peptide (B-peptide) identified from a screen using the EGFP-654 splicing reporter mouse model to ensure PPMO entry to cells and notable exon-skipping in the heart after retro-orbital injection resulting in improved cardiac function specifically end-systolic volume and end-diastolic volume and resistance to dobutamine [
117]. In another study, intravenous injection of a single 25 mg/kg dose of B-peptide conjugated PMO to
mdx mouse confirmed approximately 50% wild-type dystrophin levels along with restoration in cardiac function [
116,
117] and improved muscle function whereas weekly administration of naked PMO at 200 mg/kg for 12 weeks could only achieve 10% wild-type dystrophin levels [
118]. Fusion of muscle-specific peptide (MSP) with B-peptide through a phage display has been found to improve activity 2- to 4-fold after multiple 6 mg/kg dosing [
119]. Interestingly, another study revealed that a specific orientation (B-MSP-PMO) can lead to a 2–5-fold improvement in skeletal muscle restoration compared to B- PMO [
120]. Additionally, B-PMO has also been used for research in canine models of DMD that better mimic the pathophysiology of human illness and serve as a more rigorous evaluation of the efficacy of CPP-PMOs in restoring dystrophin expression. Repeat low dose (4 mg/kg per ASO) B-PMO intravenous injection has been found to restore 5% dystrophin of wild-type levels throughout the body, including in the heart where improvement of cardiac conduction defects was seen after therapy [
121]. Another arginine-rich peptide- R6G is also currently being explored for the treatment of DMD [
122]. R6G peptide is a modification of the conventional R6 peptide with the glycine residue. that has been extensively studied for various neuromuscular disorders [
123,
124]. When conjugated with PMO, it has shown promise in exon skipping efficacy specifically in cardiac muscle [
122].
Recent research has led to the development of several peptide series known as "Pip’s" (PMO/PNA internalization peptides), which are generated from the parent peptide penetratin [
125,
126] and consist of the amino acids arginine (R), 6-aminohexanoic acid (X), and ß-alanine spacer (B), with an internal core containing hydrophobic residues [
12]. The most recent Pip-PMO conjugates are significantly more effective than naked PMO and, more critically, reach cardiac muscle following systemic administration in dystrophic animal models. A single intravenous injection of the Pip5e peptide conjugated PMO induced the highest amounts of exon skipping and dystrophin restoration throughout the body, including in the heart of
mdx mice [
127]. To increase homogenous dystrophin repair and to target the heart muscle more effectively, the Pip6 series of peptides were generated by further iterations of core design [
128]. In a study, it was observed that inversion of the Pip5e-PMO hydrophobic core (Pip6a) resulted in a cardiac dystrophin recovery score of up to 37% in the
mdx animal model [
92]. In another study by the same group, it has been demonstrated that administering Pip6f-PMO (scrambled peptide core) can increase the levels of the protein dystrophin by up to 28% in the heart of
mdx mice who had previously undergone a forced exercise regimen to cause changes resembling the DMD cardiac phenotype [
129]. Additionally, injection of Pip2a or Pip2b conjugated PPMOs in the tibialis anterior of the
mdx mouse has also been found to induce an effective exon 23 skipping and a noticeable increase in dystrophin rescue [
130]. Another CPP created to target muscle is M12 which was discovered through phage display conducted on C2C12 myoblasts upon conjugation to PMO, M12 achieved approximately 10–25% of wild-type dystrophin levels following a single systemic administration although at dosing levels 5- to 6-fold higher than those required for comparable efficacy [
102].
CPP-PMO has also been used as a therapeutic approach for myotonic dystrophy type I, where a CTG expansion in the DMPK gene’s 3′ untranslated region causes a pathogenic transcript that interacts with RNA-binding proteins like muscleblind-like 1 (MBNL1) to cause widespread aberrant splicing abnormalities. Systemic administration of B-PMO targeting this repeat element causes blocking of Mbnl1 sequestration, resulting in normal nuclear distribution and subsequent correction of abnormal RNA splicing, including for chloride channel 1 gene, which is a primary contributor to myotonia [
131].
One of the biggest challenges of nucleic acid therapeutic is to cross the blood-brain barrier to reach the central nervous system (CNS) after systemic delivery. CPPs have been identified as promising medicines in the treatment of central nervous system (CNS) diseases due to their demonstrated transmembrane transporting ability. It is assumed that small-size cationic or amphipathic CPPs may exhibit greater affinity for negatively charged endothelial cells on the blood-brain barrier [
132,
133]. CPP-PMOs have recently been investigated in preclinical models of spinal muscular atrophy (SMA), an autosomal recessive neuromuscular disorder that results in premature death [
134]. This disease is caused by mutations in the survival of the motor neuron 1 (SMN1) gene. A paralogous gene, SMN2, encodes a vital SMN protein but generates only minimal levels due to a sequence variant leading to the exclusion of exon 7 from approximately 90% of mature transcripts. Consequently, a truncated, non-functional protein is produced [
135,
136]. To address the functional deficiency caused by the loss of SMN1 protein in patients, ASOs have been employed to facilitate the inclusion of exon 7 in SMN2 transcripts, thereby enhancing the production of SMN2 protein [
137]. However, the limited delivery of the currently used ASO in the rostral spinal and brain has reduced the therapeutic efficacy [
104]. Nusinersen, a modified 2’-MOE PS ASO has been recently approved by FDA for the treatment of SMA. Intrathecal injection of Nusinersen can significantly improve motor function and increase the lifetime of SMA patients [
138,
139]. However, this procedure is invasive and is linked to unpleasant post-lumbar puncture adverse effects for the patients [
137]. Therefore, to address this, PPMO trials have been conducted. Intravenous administration of Pip6A-PMO in the Taiwanese severe SMA mouse model increased mean survival and SMN2 expression in the brain and spinal cord, and improved neuromuscular junction morphology [
140]. Due to the mouse model’s severity, the drug has to be administered before postnatal day 2 to demonstrate functional benefit. It is likely that the BBB may not be fully formed at that time, as a result, does not accurately represent the clinical condition for therapeutic intervention [
102]. To prove the blood-brain-barrier crossing capacity of PPMO, a study has been conducted where symptomatic SMA mice were administered RXR-MO and r6-MO (morpholino oligomer) conjugates intraperitoneally at PD-5 with a completely closed BBB. The treated mice showed improved median survivals of 41.4 and 23 days, respectively which is significantly higher compared to the naked MO (~17 days). Additionally, RXR-MO and r6-MO conjugates were found in the central nervous system in a symptomatic phase. Pathological studies demonstrated that CPP-MOs mitigate the degradation of neuromuscular connections more efficiently than scrambled or naked MOs [
124]. Another study demonstrated that a derivative of an ApoE could induce a 0.25-fold increase in exon 7 inclusion in the pre-mRNA of the spinal cord and to a lesser extent in the brain of a spinal muscular atrophy mouse model, improving the diseased mice’s phenotype [
141].
CPP-PMO strategies have also been developed for the treatment of other neurological diseases like Huntington’s disease (HD) and Amyotrophic Lateral Sclerosis (ALS) [
83] as well as for use as antibacterial agents because ASOs by themselves are not very effective at penetrating bacterial cell walls [
102]. It is evident that CPPs have a great deal of therapeutic potential in delivering and increasing the efficacy of ASOs specifically the PMO-based strategy.
8. DG9: A CPP for enhancing the delivery and cellular uptake of ASO and proteins
Although CPPs hold promise in facilitating the transport of biologically active cargo across cell membranes, including the notorious blood-brain barrier and other challenging barriers within the body, CPPs also pose a number of difficulties and issues that require careful study. The primary obstacle to completing clinical trials for PPMO-based medications right now is their toxicity and immunogenicity. Toxicity can be variable depending on several factors, including species, treatment duration, frequency of systemic administration, dosage, exons skipped, and the cationic nature of the peptide [
83]. Additionally, first-generation arginine-rich peptides were found to be more immunogenic than PMOs [
142], suggesting that the toxicity may result from immunogenic processes such as complement activation [
121,
143]. Due to severe side effects, a pre-clinical experiment using an arginine-rich PPMO by Sarepta had to be stopped. It is assumed that the side effects were partially attributable to the high dosage employed [
144]. In a separate study, rats given high doses of B-peptide-PMO experienced a loss of body weight and an increase in serum blood urea nitrogen and creatinine in a dose-dependent manner indicating decreased renal output [
145]. Therefore, the quest for cell-penetrating peptides is still ongoing in order to overcome the challenges. A peptide found recently in this search is- DG9.
DG9 is a cell-penetrating peptide derived from the protein transduction domain (PTD) of the human Hph-1 transcription factor, which facilitates the cell membrane penetration of its protein cargos in the lungs (
Figure 4). Two of these Hph-1 domains constitute DG9 [
9]. In a study by Choi et al., it has been shown that intraperitoneal injection of fusion proteins conjugated with the Hph-1 domain has enhanced the delivery in a wide range of organs, including the heart and brain which are apparently challenging to be delivered. Additionally, according to the study, cell viability was not affected, and behavioral abnormalities, cytotoxic effects, and immunogenicity were not observed after 1.6 mg/kg of intravenous administration of Hph-1 fused protein into mice for 14 days or 100 ug of intraperitoneal injection two times a week for 2 weeks [
146]. The same author later reported another study where they used the same protein transduction domain (PTD) of the human Hph-1 transcription factor, but this time with two tandem sequences (HHph-1-PTD) and fused it with Foxp3 (target protein of the study) protein to increase the cell permeability of Foxp3. In
in vitro study, HHph-1-Foxp3 was detected in the nucleus as well as in the cytoplasm within 30 minutes of transduction, suggesting that Foxp3 protein is efficiently delivered to cells and is localized in the nucleus. The delivery efficacy of HHph-1 was also proved
in vivo as HHph-1-Foxp3 treated mice lived longer and the phenotype got improved compared to the control groups. They also found that two repeats of Hph-1-PTD (HHph-1) resulted in optimal intracellular transduction and rapid delivery compared to one Hph1 domain [
147].
A separate study reported that a conjugate of PMO and a unique peptide, derived from a human T cell and a near dimer of the PTD is at least 10- to 100-fold more efficient than the prior peptides at delivering the PMO into bacteria and ultimately causing the bacterial death. The only difference between the DG9 and the peptide used in the study is only L forms of amino acid residues were used in the peptide [
148]. Kim et al. previously demonstrated that DG9 can deliver a PMO to the zebrafish heart and can cause strong exon skipping in the heart while also inducing exon skipping at significantly greater levels in the skeletal muscle [
149].
The FDA has approved exon skipping as a promising therapy for DMD which utilizes phosphorodiamidate morpholino oligomer (PMO) to target and modulate gene expression. Yokota’s research group has identified DG9 peptide conjugation as a powerful way of enhancing the exon-skipping efficacy of PMO
in vivo [
9]. The positive aspect of the DG9 peptide used in this study is that it has a potentially better toxicity profile compared to other peptides. As mentioned earlier, Peptide-conjugated PMOs have been found to induce dose-dependent toxic effects in pre-clinical studies which is thought to be linked with their amino acid compositions [
83,
144,
150]. It has also been reported that substituting D-amino acid for L-amino acid in polymer-peptide conjugates attenuated anti-polymer antibody generation and toxicity and exhibit good tolerance
in vivo even after repeated administration [
151]. Therefore, certain L-arginine residues in DG9 were converted to D- arginine (DG9 (sequence N-YArVRRrGPRGYArVRRrGPRr-C; uppercase: L-amino acids, lowercase: D-amino acids)) [
9]. This conversion has been shown to improve the viability of peptide-conjugated PMO- treated cells
in vitro along with increasing serum stability [
152]. Additionally, this DG9 does not contain any 6-aminohexanoic acid residues (often represented by “X” in peptide sequences), which have also been linked to higher toxicity [
152]. In the study of DG9-PMO mediated efficient exon skipping by Lim et al., it has been demonstrated that retro-orbital injection of DG9-conjugated PMO into hDMDdel52;
mdx mice can increase skipping efficiency 2.2 to 12.3-fold and 14.4-fold compared to the unconjugated PMO with a dystrophin restoration amount of 3% and 2.5% of wild-type levels in skeletal muscles and heart, respectively. Skeletal muscles produced 2.8 to 3.9% more dystrophin and had an exon 51 skipping level of 55 to 71% after receiving repeated injections of DG9-PMO once each week for three weeks. Most notably, hDMDdel52;
mdx mice treated repeatedly with DG9-PMO showed a considerable improvement in forelimb and total limb grip strength indicating the improvement of the muscle function of the treated mice. Additionally, the tibialis anterior DG9-PMO intramuscular injection was successful and demonstrated dystrophin restoration, suggesting the possibility of DG9-PMO for DMD therapy. There was no significant toxicity observed after the injection of DG9-PMO [
9]
Another study by Yokota’s research group tested the effectiveness of DG9 peptide in an SMA mouse model (
Figure 4). In the study, Tejal et al. showed that after a single subcutaneous administration of DG9-PMO into SMA mice (Taiwanese model) FL-SMN2 (Full-length SMN2) expression was increased ~5-fold compared to unconjugated PMO in the majority of the tissues including brain and spinal cord. The results indicated improved motor and breathing function and muscle strength with an increased mean survival of 58 days for DG9-PMO treated mice which were significantly higher compared to untreated (8 days) and unconjugated PMO treated mice (12 days). The fact that DG9 greatly improved the uptake of PMO in the CNS and peripheral tissues at PD7, despite subcutaneous treatment at PD5, indicates that DG9-PMO can assure extensive distribution of the PMO to both the peripheral and CNS tissues. The toxicological studies show that DG9-PMO does not appear to be adverse or to impair mice’s immune systems [
10].