Nucleic acid drugs: recent progress and future perspectives

Abstract

High efficacy, selectivity and cellular targeting of therapeutic agents has been an active area of investigation for decades. Currently, most clinically approved therapeutics are small molecules or protein/antibody biologics. Targeted action of small molecule drugs remains a challenge in medicine. In addition, many diseases are considered ‘undruggable’ using standard biomacromolecules. Many of these challenges however, can be addressed using nucleic therapeutics. Nucleic acid drugs (NADs) are a new generation of gene-editing modalities characterized by their high efficiency and rapid development, which have become an active research topic in new drug development field. However, many factors, including their low stability, short half-life, high immunogenicity, tissue targeting, cellular uptake, and endosomal escape, hamper the delivery and clinical application of NADs. Scientists have used chemical modification techniques to improve the physicochemical properties of NADs. In contrast, modified NADs typically require carriers to enter target cells and reach specific intracellular locations. Multiple delivery approaches have been developed to effectively improve intracellular delivery and the in vivo bioavailability of NADs. Several NADs have entered the clinical trial recently, and some have been approved for therapeutic use in different fields. This review summarizes NADs development and evolution and introduces NADs classifications and general delivery strategies, highlighting their success in clinical applications. Additionally, this review discusses the limitations and potential future applications of NADs as gene therapy candidates.

Introduction

The central dogma of genetics posits that nucleic acids carry human genetic information and play a crucial role in life processes, such as growth, development, and reproduction. Moreover, nucleic acids can be used to modify genetic information to treat various diseases.^1,2 With the advancement of the life sciences, proteomics, and genomics methods, nucleic acid drugs (NADs) have been developed to translate and regulate nucleic acid functions.^3–5 These drugs can achieve long-lasting efficacy through gene repression, replacement, and editing.^4,6 Many studies have shown the feasibility of NADs in disease prevention and treatment.^7–9 Thus, research and development on new classes of functional NADs are gradually emerging.

NADs are a class of gene therapy agents based on DNA, RNA, or synthetic oligonucleotide analogs. They have considerable potential for clinical applications, such as treating bacterial infections, tumors, and neuromuscular diseases.^10–12 However, as negatively charged biological macromolecules, NADs have difficulty in crossing cellular membranes to enter cells. Additionally, they can be easily degraded by endogenous nucleases in plasma and tissues. Furthermore, few amounts of NADs that enter cells often become trapped by endosomes and subsequently degraded by lysosomes, considerably limiting their development and application.^13–15 Currently, two main strategies exist to address the application challenges of NADs. One approach is to modify the nucleic acid structure to stabilize the properties of NADs and avoid recognition by the immune system. The other approach is to use delivery systems that facilitate their passage through cell membranes and ensure their localization to specific subcellular compartments. Consequently, the modification and transformation of NADs and the development of efficient, safe, and targeted delivery systems have become the primary focus of research and development on NADs.^16,17

The accelerated development of the NADs field has relied on innovations and breakthroughs in foundational technologies, such as chemical synthesis, site-specific modifications, and delivery techniques. These advancements are crucial to ensure the safety, effectiveness, targeting, and applicability of NADs.^18,19 Advancements in carrier technology and delivery systems have enhanced the biological activity of NADs, which improves their cellular targeting and uptake. Thus, the concentration and bioavailability of these drugs in target tissues are increased.²⁰ Various delivery systems for NADs have been developed, including lipid nanoparticles (LNPs), cationic polymer complexes, and ligand-mediated nucleic acid molecular targeted delivery systems based on specific receptors, peptides, and other engineered carriers.^21–27 However, these systems have several drawbacks, such as nonspecific distribution, inefficient cytoplasmic delivery, and suboptimal organelle targeting. Several studies have reported that more than one strategy is needed to address the delivery challenges. Thus, combining chemical structure modifications of nucleic acids with advanced drug delivery systems could achieve enhanced therapeutic effects.

This paper outlines the history of several significant molecular biology discoveries related to NADs, tracing key milestones from initial conceptualization to clinical applications. Then, we introduce the various NAD types and their modes of action, with an overview of both approved NADs and those currently in clinical trials. Then, we discuss the challenges associated with NADs development and explore strategies for overcoming the obstacles to in vivo delivery, including chemical modifications and delivery systems. Finally, we highlight the remaining challenges for NADs development, offering references for the design and clinical application of novel NADs.

Concept and historical development of NADs

NADs development is inseparable from the major discoveries in fundamental molecular biology and the continuous observations of life activities (Fig. 1). In 1869, Friedrich Miescher discovered a new molecule called “nuclein” from white blood cells, which marked the beginning of DNA discovery.²⁸ However, owing to the lack of advanced technologies at the time, the critical role of nucleic acids was not fully understood. The subsequent revelation of DNA’s double helix structure and the formulation of the central dogma of genetics clarified that nucleic acids are crucial participants in transmitting genetic information.^1,2 Since then, researchers have understood that genetic information is encoded within nucleic acids and translated into proteins via complex mechanisms, and it plays a vital role in all life processes, such as growth, development, and reproduction.

Fig. 1.

Historical timeline of essential discoveries in fundamental molecular biology theory and critical developments in NADs therapy. The orange boxes represent major biological discoveries in nucleic acids development, including the discovery of DNA and RNA, as well as researchers’ exploration of special biological phenomena such as RNA interference, nucleic acid hybridization, and gene editing. The yellow boxes show the breakthrough progress in the clinical application of NADs based on the aforementioned biological phenomena. These include successful clinical application cases of NADs, such as the first ASO drug Fomivirsen, the first siRNA drug Patisiran, the first aptamer drug Pegaptanib, the COVID-19 mRNA vaccines, as well as clinical trials for NADs in development, such as the saRNA drug MTL-CEBPA

Additionally, there have been significant breakthroughs in NADs development including the discovery of RNA’s double-stranded structure and the phenomenon of nucleic acid hybridization.²⁹ Early studies posited that single-stranded RNA could not form double-stranded structures. However, in 1956, Rich and Davies discovered that RNA could form double-stranded structures similar to those of DNA based on the principle of complementary base pairing. This result laid the foundation for developing RNA double-stranded drugs, including microRNAs (miRNAs) and small interfering RNAs (siRNAs).³⁰ In 1960, Rich³¹ reported the phenomenon of DNA/RNA hybridization. In 1978, building on this discovery, Zamecnik and Stephenson³² used specific oligodeoxynucleotide chains that target the 35S RNA of the Rous sarcoma virus to inhibit virus replication, marking a prototype for the application of antisense oligonucleotide (ASO) drugs in disease treatment. As research on transcription and translation advanced, it was discovered that initial RNA transcripts typically require intron removal and the linkage of exons to form mature messenger RNA (mRNA), a process known as RNA splicing.^33,34 RNA splicing is a crucial step in gene expression, and abnormalities in this process are the main cause of genetic variations and diseases.^35,36 Dominski et al.³⁷ found that ASOs targeting splicing sites can restore correct splicing of defective genes rather than only downregulating gene expression. This discovery provided a novel treatment strategy for diseases related to mis-splicing.

In 1998, Andrew Fire and Craig Mello reported that double-stranded RNA (dsRNA) had potent gene silencing effects in Caenorhabditis elegans,³⁸ a discovery they termed RNA interference (RNAi). RNAi was quickly applied to inhibit the replication of the hepatitis C virus in mice,³⁹ marking the earliest evidence of siRNA-mediated in vivo gene silencing. In 2006, they were awarded the Nobel Prize in Physiology and Medicine for their RNAi technology.⁴⁰ Subsequently, researchers demonstrated that siRNA could be tailored to disrupt the expression of any pathogenic gene, propelling RNAi into the spotlight and fostering its active application in the treatment of various diseases. In 2010, the first human trial using RNAi technology was conducted to evaluate the therapeutic effect of siRNA for targeting the M2 subunit of ribonucleotide reductase in patients with melanoma.⁴¹ Since then, siRNA drugs have had issues with stability, immunogenicity, off-target effects, safety, and the delivery system. However, after advanced chemical modifications and the development of targeted delivery systems, the first siRNA drug approved by the Food and Drug Administration (FDA) in 2018 reignited interest in NADs.⁴²

Notably, the discovery of RNA-dependent RNA polymerase and reverse transcriptase has been crucial for developing subsequent mRNA drugs.^43,44 In 1984, Krieg and Melton used RNA polymerase extracted from viruses for in vitro transcription from engineered DNA templates, successfully achieving mRNA expression in cell-free systems.⁴⁵ Thus, since the 1990s, in vitro-synthesized mRNA has been increasingly applied for protein replacement in preventive and therapeutic vaccines.^46–48 In 2005, the discovery of pseudouridine (Ψ) modification addressed the immunogenicity issue of in vitro-synthesized mRNA,⁴⁹ leading to the initiation of the first human trial of an mRNA vaccine against melanoma in 2008. In 2020, during the COVID-19 pandemic, the FDA authorized the emergency use of mRNA vaccines, providing effective measures for preventing and controlling the virus.^50,51 This resulted in widespread public attention to NADs.⁵² In 2023, the Nobel Prize in Physiology and Medicine was awarded to Katalin Kariko and Drew Weissman for pioneering nucleoside base modification technology to decrease mRNA immunogenicity, further highlighting the critical role of chemical modification technologies in developing mRNA vaccines. This opened new clinical applications for NADs in treating human diseases.^53,54

Simultaneously, discovering other types of nucleic acids and biological phenomena has further expanded the scope of NADs. Specifically, the emergence of gene-editing technology has provided a foundation for developing new therapies for genetic mutation diseases.^55–57 Recently, the first gene therapy based on clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 technology has been approved for marketing,⁵⁸ resulting in revolutionary changes in NADs development.

Classification and therapeutic mechanisms of NADs

NADs can be broadly divided into three categories based on their mechanisms of action. The first category includes NADs that target nucleic acids to regulate protein expression by promoting or inhibiting translation. This category primarily consists of ASOs, siRNAs, miRNAs, small activating RNAs (saRNAs), and the CRISPR/Cas system, which enables precise gene editing of genomic DNA.

The second category includes NADs that target proteins, with aptamers as the main examples. Unlike the first category, aptamers can directly and specifically bind to target proteins, functioning similarly to antibodies by providing a targeting mechanism.

The third category includes NADs that express proteins, such as in vitro-transcribed mRNA, which can produce specific proteins in vivo to exert biological activity. This section briefly introduces the mechanisms of action of these different NAD types and highlights drugs successfully applied in clinical settings (Fig. 2).

Fig. 2.

Classification and therapeutic mechanisms of NADs. a Gapmer ASO (consisting of a DNA-based internal gap with RNA-like flanking regions) binds to target mRNA with high affinity to form an RNA-DNA duplex and participates in RNase H-mediated mRNA degradation. b Steric block ASO regulates functional target gene expression through exon skipping or exon inclusion or interrupts translation initiation by targeting and masking the AUG start codon of the target mRNA. c siRNAs form RISC with AGO2. While the passenger strand is discarded, the antisense strand binds to the target mRNA, downregulating the translation level of the target mRNA. d pri-miRNAs produced by miRNA gene transcription in the non-coding region are processed to form mature miRNAs with the help of a series of complexes (Drosha/DGCR8, Exportin-5/RAN-GTP, and Dicer/TRBP). miRNAs combine with the AGO2 to form miRISC. The activity of miRNAs can be inhibited by miRNA inhibitors that either form a complex with the mature miRNA loaded in the miRISC complex or by masking a target site via interactions with the specific transcript being targeted. e saRNAs recruit the RITA complex (including AGO2, CTR9, RHA, and RNAP II) to stimulate the initiation and extension of transcription. f CRISPR-mediated gene editing mainly uses Cas9 and sgRNA to introduce DSBs at specific positions in the genome effectively. DSBs are generally repaired by HDR or NHEJ, achieving insertion, knockout, and site-specific mutagenesis. g Aptamers screened by SELEX technology can recognize specific proteins by forming 3D structures. h Exogenous mRNAs introduced into cells undergo translation to proteins and facilitate protein function through protein replacement therapy and mRNA vaccines

NADs that target nucleic acids

ASOs

ASOs are artificially synthesized single-stranded oligonucleotide chains that regulate and target RNA’s function via specific binding according to the Watson-Crick base-pairing principle.⁵⁹ The mechanisms of action of clinically used ASOs mainly include ribonuclease H (RNase H)-mediated degradation (Fig. 2a) and steric blockage mechanisms (Fig. 2b).^60–62 RNase H-dependent ASOs, such as gapmers, bind to complementary mRNA, recruit RNase H to cleave mRNA, and thus block translation of the target gene. This results in the inhibition or reduction of the synthesis of the target protein. Representative drugs based on this mechanism include fomivirsen (Vitravene®), mipomersen (Kynamro®, Delisted), inotersen (Tegsedi®), volanesorsen (Waylivra®), and tofersen (Qalsody®).^63–67 Because RNase H is active in the nucleus and cytoplasm, ASOs can also target other transcripts, such as long non-coding RNAs.^68,69 Indeed, the pathogenesis of many diseases without clear protein targets is related to non-coding RNA, which can be used to predict the advantages of ASO in treating diseases.^70,71

Additionally, ASOs can form a double-stranded structure complementary to miRNA, leading to its degradation and causing gene upregulation.^72–74 Additionally, ASOs regulate transcription via steric hindrance, influencing specific splicing of pre-mRNA and selectively altering the expression of specific proteins.^75,76 ASOs that employ this mechanism are splicing-switching oligonucleotides, which modulate functional target gene expression by skipping or including exons.^77–79 For example, Golodirsen, an ASO drug targeting the human Duchenne muscular dystrophy (DMD) gene, was introduced in the USA in 2019 by Sarepta Therapeutics to treat DMD patients with confirmed mutations amenable to exon 53 skipping.⁸⁰ Additionally, several studies have demonstrated that ASOs can disrupt translation initiation by targeting and masking the AUG start codon of the target mRNA.⁸¹ These discoveries have enabled ASOs to be applied in disease treatment via more diverse mechanisms.

siRNA

siRNA is a dsRNA molecule that is typically 19–23 base pairs long and is found naturally in various organisms or artificially synthesized.^82,83 As a classical effector of RNAi, siRNA induces gene silencing by blocking mRNA translation.^84,85 Unlike ASOs, siRNA-mediated gene silencing occurs via the RNA-induced silencing complex (RISC), not RNase H. Once mature siRNA enters the cell, it forms a complex with the Argonaute-2 (AGO2) protein (Fig. 2c).^86,87 As the passenger strand of siRNA is degraded, the antisense strand binds to the target mRNA. It guides RISC to cleave the target sequence, thus achieving therapeutic effects by downregulating the translation of specific proteins.^88–90 Based on this mechanism, researchers have designed siRNAs that target specific pathological genes to achieve specific gene silencing using RNAi.^91,92 However, challenges, such as stability, specificity, and delivery obstacles, hindered progress in the early stages of development.^93–96 Advancements in carrier technology and nucleic acid modification techniques have led to significant progress in overcoming these challenges, as exemplified by the first siRNA drug, patisiran (Onpattro®), which has been approved for treating hereditary transthyretin-mediated amyloidosis by degrading mRNA encoding transthyretin (TTR).⁹⁷ Thus far, six siRNA drugs have received international approval. With the ongoing development of novel chemical modifications and targeted delivery systems, more siRNA drugs are anticipated to enter the market soon.

miRNA

Endogenous non-coding RNAs that have been discovered in eukaryotic organisms can act as gene regulators. Since the 1993 discovery of the first miRNA, lin-4, in the nematode C. elegans,⁹⁸ miRNAs have been shown to participate in various biological functions and pathological mechanisms, including cell proliferation, differentiation, migration, disease occurrence, and progression.⁹⁹ There has been extensive research on the regulatory mechanisms of miRNAs. With the assistance of complexes, such as Drosha/DGCR, Exportin-5/RAN-GTP, and Dicer/TRBP, primary miRNAs (pri-miRNAs) transcribed from the non-coding region of miRNA genes are processed to form mature miRNAs (Fig. 2d).^100–105 miRNAs combine with Argonaute proteins to form the miRNA-induced silencing complex (miRISC), which can silence the target transcript via base pair complementation.^106,107 Unlike siRNA-mediated gene silencing, miRNAs can simultaneously recognize and regulate the expression of multiple target mRNAs due to their low complementarity.^108,109 The antisense strand of miRNA can bind to the target mRNA via complete and incomplete complementarity, leading to the cleavage of target mRNA and the inhibition of target gene expression.^110,111 Contrary to the classical gene silencing mechanism, research has shown that miRNAs can interact with 3’-UTR to upregulate gene expression, indicating the complexity and diversity of miRNA regulatory mechanisms.^112–114

As studies on the roles of miRNAs in diseases continue to be reported, there is anticipation for the potential use of miRNAs in different pathological processes. However, thus far, no miRNA drugs have been approved for the market.^115,116 miRNA drugs’ ongoing development primarily includes two categories: miRNA mimics and miRNA inhibitors.^117–121 miRNA mimics are synthetic dsRNA molecules that mimic the function of endogenous miRNAs. Similarly, miRNA inhibitors are single-stranded RNAs that are complementary to endogenous miRNAs, which can reduce the effect of gene silencing by specifically inhibiting miRNA.^122,123 The miRNA drug MRG-201 (Remlarsen), developed by Viridian Therapeutics (formerly known as miRagen Therapeutics), simulates microRNA-29b to reverse regulate fibrosis, thus inhibiting fibrous proliferation in skin wounds. Clinical trials have been conducted to evaluate the efficacy and safety of MRG-201 in treating fibrotic diseases (NCT03601052).¹²⁴ In addition to MRG-201, Viridian Therapeutics has several other miRNA-based drugs, including MRG-106 (Cobomarsen), which targets miR-155 to inhibit tumor development, and MRG-110, which targets microRNA-92 to promote angiogenesis.^125,126 TTX-MC138 (NCT06260774 and NCT05908773), designed by TransCode Therapeutics, targets microRNA-10b in the treatment of pancreatic cancer.¹²⁷ Despite the relatively slow progress in miRNA-related drug development compared with other NADs, their superior performance in treating tumors, heart failure, and diabetes indicates their promising prospects in clinical treatment.

saRNA

In addition to the classical gene silencing mechanism, the discovery of RNA activation has provided a new perspective on gene regulation. RNA activation is a gene regulation phenomenon mediated by a small dsRNA called saRNA, which targets gene promoter sequences to enhance the transcription of the target gene.^128,129 After entering the cell via endocytosis, saRNA binds to AGO2. The sense strand is released and degraded, as AGO2 and the antisense strand are transported into the cell nucleus, where they bind to the promoter region (Fig. 2e). Subsequently, AGO2 recruits RNA polymerase-associated protein CTR9 homolog (part of the polymerase-associated factor 1 complex) and RNA helicase II (RHA) to form the RNA-induced transcriptional activation (RITA) complex. This complex interacts with RNA polymerase II (RNAP II) and stimulates the initiation and elongation of transcription.^130,131 saRNAs are conserved in mammals and participate in the activation of various genes, such as vascular endothelial growth factor A (VEGFA), E-cadherin, progesterone receptor, and Kruppel-like factor 4.^{128,132–135} This conservation allows for the establishment of animal models and preclinical studies for saRNA-based therapies. Targeted activation therapy using saRNA has been validated in various disease animal models.^{129,136–140} However, clinical translation of saRNAs remains challenging due to the absence of an immune system in xenograft models and the complexity of drug action in the human body.¹⁴¹

The leading saRNA drug, MTL-CEBPA, which was developed by MiNA Therapeutics, targets the CCAAT/enhancer-binding protein alpha (CEBPA). Phase I trial results (NCT05097911) have demonstrated that MTL-CEBPA has a good safety profile for the treatment of hepatocellular carcinoma and may enhance the therapeutic effects of tyrosine kinase inhibitors by modulating immune suppression.^142–144 In addition to MTL-CEBPA, MiNA Therapeutics has several saRNA candidates in preclinical development for undisclosed tumor types, metabolic diseases, and genetic disorders. RAG-01 is another saRNA candidate drug designed to treat bladder cancer by activating the expression of the human cyclin-dependent kinase inhibitor 1 A.¹⁴⁵ Ractigen Therapeutics has submitted an Investigational New Drug application for RAG-01, and it is expected to become the second saRNA drug to enter clinical trials internationally.

CRISPR/Cas9 system

Gene editing is a genetic engineering technology that precisely modifies target genes via insertion, deletion, and site-specific mutation. The CRISPR/Cas9 system has gained widespread attention among the various genome-editing methods. As a prokaryotic adaptive immune system, CRISPR/Cas modules establish bacterial defense against exogenous nucleic acids. They have been widely used in gene editing owing to their high efficiency and accuracy.^146–149 The CRISPR-mediated gene-editing system primarily involves Cas9 and single guide RNA (sgRNA) to introduce DNA double-strand breaks (DSBs) at specific positions in the genome. These DSBs are usually repaired by homologous directed repair (HDR) or nonhomologous end joining (NHEJ), achieving mutation or foreign gene insertion (Fig. 2f).^150,151 CRISPR/Cas9 technology has been effective in various fields, including hematologic disorders, malignant tumors, and genetic diseases.^151–153 However, this technology has several challenges, such as low delivery and HDR efficiency, off-target effects, and toxic side effects.^154–157 Currently, the CRISPR/Cas9 system can enter cells for gene editing in three forms: the Cas9: sgRNA ribonucleoprotein complex, mRNA for Cas9 translation alongside a separate sgRNA, and a DNA plasmid that encodes both the Cas9 protein and sgRNA.^158–160 Each method has its pros and cons. The plasmid system is relatively stable but has low cutting and editing efficiency, and continuous expression of Cas9 may increase off-target effects.¹⁶¹ mRNA and sgRNA are susceptible to degradation by nucleases. The Cas9/sgRNA ribonucleoprotein complex, while the most responsive mode of action, has a low off-target rate and low toxicity, but the large size of the complex complicates delivery.¹⁶⁰ The delivery of the CRISPR/Cas9 system is a crucial issue limiting its clinical application. Like other NADs, researchers have focused on nanocarriers based on liposomes, chitosan, and other materials to achieve efficient delivery, laying a solid foundation for the clinical application of the CRISPR/Cas9 system.^162–164 Exa-cel (Casgevy®) was approved in November 2023 as the first gene-editing therapeutic drug based on the CRISPR/Cas9 system for treating sickle cell disease (SCD) and transfusion-dependent β-thalassemia (TDT). Exa-cel stimulates artificial blood stem cells and progenitor cells in vitro to differentiate into red blood cells that produce high fetal hemoglobin levels.^165,166 The launch of Exa-cel has filled the gap in gene-editing drugs and provided considerable motivation for developing other gene-editing therapies.

NADs that target proteins

Aptamers are single-stranded oligonucleotide molecules that have specific recognition functions obtained via iterative screening from large libraries of random oligonucleotides using the systematic evolution of ligands by exponential enrichment (SELEX) technology.^167–169 Unlike other nucleic acid-based drugs, aptamers specifically recognize and bind target molecules, such as peptides, proteins, viruses, bacteria, and cells, relying on their unique three-dimensional conformation. This is similar to the conformational recognition that mediates antibody–antigen interactions and complex formation (Fig. 2g).^170–173 However, compared with antibodies, aptamers have several advantages, including high thermal and physiological stability, low immunogenicity, and a wider range of target specificity. Since their first report in 1990, SELEX technology has been continuously improved,^174,175 diversifying the selection and development of high-affinity aptamers.^176,177 Aptamers have shown promising applications in treating cancer and ophthalmic and cardiovascular diseases (CVDs).^178–180

In 2004, the FDA approved the first aptamer drug, pegaptanib (Macugen®), to treat choroidal neovascularization caused by neovascular age-related macular degeneration (AMD).¹⁸¹ However, Pegaptanib was withdrawn from the market owing to its poor efficacy and competition from anti-VEGF antibody drugs, such as Lucentis.⁶⁴ Despite this, research on aptamers continues. In August 2023, avacincaptad pegol was approved for treating geographic atrophy (GA) secondary to dry AMD. Complex cascade overactivity is likely instrumental in AMD pathology. A crucial complement component, C5, has become a primary therapeutic target for many inflammatory diseases, including AMD.¹⁸² As a C5 inhibitor, avacincaptad pegol has been shown to slow GA progression by targeting the source of retinal cell death. Additionally, AS1411 is a candidate drug that targets nucleolin, which has been confirmed to be effective for treating renal cell carcinoma,¹⁸³ glioma,¹⁸⁴ and acute myeloid leukemia.¹⁸⁵ In addition, aptamers are widely used in drug delivery, clinical diagnostics, and biosensing.^186–188 This paper primarily focuses on their application as therapeutic drugs. Thus, we will not describe their other uses in detail further.

NADs that express proteins

mRNA, a single-stranded polynucleotide that carries genetic information, is essential for expressing encoded proteins within cells, thus exerting corresponding biological functions via these proteins. Consequently, the concept of using mRNA as NADs has been proposed. In 1990, Wolff et al.⁴⁶ injected in vitro-synthesized mRNA into mouse skeletal muscle and successfully induced the expression of specific proteins. They demonstrated the feasibility of using in vitro-synthesized mRNA as an information carrier to guide somatic protein synthesis.

However, the development of mRNA-based drugs has resulted in several challenges. Unmodified mRNA can induce Toll-like receptor-mediated immune responses, leading to blocked protein synthesis. Kariko and Weissman discovered that nucleobase modifications could protect mRNA from triggering inflammatory responses.^49,54 In addition, the large size of mRNA and its susceptibility to degradation by nucleases result in low cellular uptake, further limiting its application. To overcome these obstacles, researchers have developed various delivery systems, including lipids, peptides, and polymers, for mRNA delivery both in vitro and in vivo.^189–191

In contrast to the previously mentioned NADs that exert therapeutic effects by directly binding mRNA or proteins, two strategies for the use of mRNA drugs have been attempted (Fig. 2h). One strategy is protein replacement therapy, which involves introducing exogenous mRNAs into cells to express functional proteins or supplement deficient ones.^192,193 For example, one research group used LNPs to deliver mRNA encoding erythropoietin into mouse fetuses, thus increasing erythropoietin protein levels in the mouse bloodstream.¹⁹² Additionally, mRNA therapy has been applied to treat patients with deficiencies in essential enzyme genes, such as argininosuccinate lyase, ornithine transcarbamylase, and methylmalonyl-CoA mutase (MUT), therefore restoring enzyme levels and mitigating deficiencies.^194–196 Moderna, a company dedicated to mRNA therapy, has several mRNA drugs based on enzyme replacement therapy in clinical trials. For instance, mRNA-3927, which encodes the alpha and beta subunits of the propionyl-CoA carboxylase enzyme, is designed to treat propionic acidemia. A Phase I/II trial (NCT05130437) for this indication has been initiated in pediatric patients to evaluate the long-term safety of mRNA-3927.¹⁹⁷ Additionally, mRNA-3704 and mRNA-3705 encode MUT, which is designed to treat methylmalonic acidemia and is currently under investigation.

The second approach involves mRNA vaccines, which activate the body’s immune response to combat infectious diseases and tumors by directly translating mRNA-containing antigen proteins. Compared with traditional inactivated vaccines, mRNA vaccines have advantages such as cell-free production, high production efficiency, and low cost. Thus, they are quite promising for addressing sudden epidemic infectious diseases. During the COVID-19 pandemic, mRNA prophylactic vaccines were crucial among all candidate vaccine types. The FDA authorized the emergency use of BNT162b2, which was developed by Pfizer-BioNTech, and mRNA-1273, developed by Moderna.^198,199 Moreover, the development of mRNA vaccine technology has been propelled by large-scale clinical trials of mRNA vaccines. For example, ARCov, which was jointly developed by Abogen Biosciences, the Academy of Military Medical Sciences, and Walvax Biotechnology, has addressed the issue of poor thermal stability of mRNA vaccines.^200,201 Additionally, preventive vaccines against influenza viruses,^202,203 respiratory syncytial virus,²⁰⁴ rabies virus,²⁰⁵ and other viruses, as well as cancer-targeted therapeutic vaccines,^206–208 are being continuously researched and developed, which further reflect the application prospects of mRNA vaccines.

In summary, in recent decades, NADs development has undergone significant progress and achieved considerable results. More than 20 products based on ASOs, aptamers, siRNA, and mRNA have been approved for marketing to treat various rare genetic disorders (Table 1). Several companies worldwide have been active in this field, attracting substantial investment, and the market development space is expected to expand further.

Table 1.

NADs approved for clinical application

Classification	Drug (brand name)	Company	Indication	Dose (route)	Target (organ)	Modification & delivery	Approval year	Ref(s)
ASO	Fomivirsen (Vitravene)	Ionis	CMV retinitis	330 μg per eye once every 4 weeks (ITV)	CMV UL123 (eye)	PS	1998 (delisted)	234,578
	Mipomersen (Kynamro)	Ionis	HoFH	200 mg once weekly (SC)	ApoB-100 (liver)	2’-MOE	2013 (delisted)	579
	Eteplirsen (Exondys 51)	Sarepta Therapeutics	DMD	30 mg kg−1 once weekly (IV)	Exon 51 of DMD (muscle)	PMO	2016	497,498
	Nusinersen (Spinraza)	Ionis & Biogen	SMA	12 mg once every 4 months (IT)	Exon 7 of SMN2 (CNS)	2’-MOE	2016	580,581
	Inotersen (Tegsedi)	Ionis	hATTR	300 mg once weekly (SC)	TTR (liver)	2’-MOE	2018	490,491
	Volanesorsen (Waylivra)	Ionis	FCS	300 mg once weekly (SC)	ApoC-III (liver)	2’-MOE	2019	66,582
	Golodirsen (Vyondy 53)	Sarepta Therapeutics	DMD	30 mg kg−1 once weekly (IV)	Exon 53 of DMD (muscle)	PMO	2019	80
	Viltolarsen (Viltepso)	Nippon Shinyaku	DMD	80 mg kg−1 once weekly (IV)	Exon 53 of DMD (muscle)	PMO	2020	263,583
	Casimersen (Amondys 45)	Sarepta Therapeutics	DMD	30 mg kg−1 once weekly (IV)	Exon 45 of DMD (muscle)	PMO	2021	264,496
	Tofersen (Qalsody)	Biogen	ALS	100 mg once every 28 days (IT)	SOD1(CNS)	PS/ 2’-MOE	2023	67,505
	Eplontersen (Wainua)	Ionis & AstraZeneca	ATTRv-PN	45 mg once monthly (SC)	TTR (liver)	LICA	2023	492
siRNA	Patisiran (Onpattro)	Alnylam	hATTR	0.3 mg kg−1 once every 3 weeks, max:30 mg (IV)	TTR (liver)	2’-OME & LNP	2018	97
	Givosiran (Givlaari)	Alnylam	AHP	2.5 mg kg−1 once monthly (SC)	ALAS1(liver)	PS/ 2’-OME & GalNAc	2019	584
	Lumasiran (Oxlumo)	Alnylam	PH1	3 or 6 mg kg−1 once every 3 months (SC)	HAO1(liver)	PS/ 2’-OME & GalNAc	2020	510
	Inclisiran (Leqvio)	Alnylam & Novaetis	primary hypercholesterolemia	284 mg at 0 and 3 months, and then once every 6 months (SC)	PCSK9(liver)	PS/ 2’-OME & GalNAc	2020	539
	Vutrisiran (Amvuttra)	Alnylam	hATTR-PN	25 mg once every 3 months (SC)	TTR (liver)	PS/ 2’-OME & GalNAc	2022	493
	Nedosiran (Rivfloza)	Novo Nordisk & Dicerna	PH1	128 or 160 mg once monthly (SC)	LDHA (liver)	PS/ 2’-OME & GalXCTM	2023	330
CRISPR/Cas	Exagamglogene autotemcel (Casgevy)	Vertex & CRISPR Therapeutics	TDT & SCD	Once (IV)	BCL11A	/	2023	165,166
Aptamers	Pegaptanib (Macugen)	EyeTech Pharmaceuticals & Pfizer	AMD	0.3 mg once every 6 weeks (ITV)	VEGF-165 (eye)	pegylated	2004 (delisted)	181
	avacincaptad pegol (Izervay)	Iveric Bio	GA	2 mg once monthly (ITV)	complement C5 (eye)	pegylated	2023	585
mRNA	BNT162b2 (Comirnaty)	Pfizer & BioNTech	COVID-19	IM	SARS-CoV-2(liver)	LNP	2020	199
	Elasomeran (Spikevax)	Modern	COVID-19	IM	SARS-CoV-2(liver)	LNP	2020	586
	mRNA-1345 (mRESVIA)	Modern	Respiratory syncytial virus infection	IM	Respiratory syncytial virus (lung)	LNP	2024	587

Current challenges in NADs development

Unlike traditional small molecules and antibody drugs that exert their pharmacological effects on proteins, most NADs directly regulate gene expression, offering a broader range of targets. This is particularly valuable for addressing genes with defective proteins that are difficult to be targeted with conventional drugs, showing considerable potential in treating rare, chronic, infectious diseases and other metabolic disorders. Despite these advantages, researchers must accurately identify the genetic information related to the disease and choose the appropriate type of NADs based on the mechanism of action.^209,210 For targeted NADs, including ASO and siRNA, when the relevant genetic information of the disease is determined, lead compounds can be designed for the gene sequence to avoid off-target effects during development.²¹¹ The efficacy of aptamer drugs is related to their sequence and conformation. SELEX technology has been applied to better screen specific sequences with high affinity for a target from a randomly generated single-stranded nucleic acid sequence library.²¹² Understanding the relationship between the mRNA sequence, structure, function, and stability for mRNA development is important to ensure the maximum functional protein output of delivered mRNA molecules.²¹³ In recent years, the NADs sequence design process has been accelerated by the development of advanced bioinformatics tools, considerably reducing the time and costs.^{209,210,214,215} Nonetheless, the major obstacle in NADs development involves how to reach target cells to fully achieve therapeutic benefits (Fig. 3).^13,216,217 The most relevant challenges can be summarized as follows.

Fig. 3.

Current challenges in NADs delivery. NADs are administered in many ways, such as intravitreal, intramuscular, intrathecal, and intravenous injection. For systemic delivery, NADs must first overcome renal clearance, nuclease degradation, immune system recognition, and drug off-target until reaching target tissues and organs. Subsequently, NADs successfully reach the target cells, enter the cell via endocytosis, enter the endosomes, and escape successfully to achieve the desired therapeutic effect. It is difficult for negatively charged NADs to cross the phospholipid bilayer on the surface of the cell membrane, which usually requires the help of carriers to recognize receptors or chemical modification of NADs to change properties

Pharmacokinetics, stability, and degradation

Naked nucleic acids have poor in vivo stability and can degrade in the bloodstream. They undergo either enzymatic (nucleases and RNAse) or chemical (oxidation and hydrolysis) degradation in the blood and tissue fluids or are filtered and cleared by the kidneys.²¹⁸ Researchers have implemented several chemical modifications to enhance NADs stability, but almost all these methods can affect the efficacy and safety.²¹⁹

Immunogenicity

Exogenous nucleic acids can be recognized as exogenous signals by pattern recognition receptors in the immune system, triggering immune responses that compromise the structural integrity and stability of the nucleic acids. Careful design and modification of NADs can mitigate immunogenicity, but this requires extensive testing.^220,221

Targeting problems

Therapeutic nucleic acids often lack sufficient targeting ability in vivo. This insufficient targeting leads to low NADs concentrations at the disease site and unintended gene silencing or activation, potentially causing safety issues by requiring higher doses. Designing particular nucleic acids and conducting thorough off-target screening are critical for the efficient clinical application of NADs.

Uptake efficiency and endosomal escape

Efficient uptake of NADs into cells is difficult owing to their size, charge, and hydrophilicity. Negatively charged nucleic acids do not easily cross the negatively charged lipid bilayer of the cell membrane. Additionally, endosomal escape is a significant barrier as nucleic acids often become trapped in endosomes and degraded. After entering cells, NADs are often captured by endosomes. Only those NADs that successfully escape the endosomes can exert their therapeutic effects. Inadequate endosomal escape can result in reduced efficacy and increased off-target toxicity.²²²

In addition to the previous issues, NADs face manufacturing and scalability challenges. Producing NADs at scale with consistent quality requires advanced manufacturing technologies and stringent quality control measures. This results in a relatively high cost, affecting accessibility and affordability for patients.

Strategies to improve NADs performance

Because of these challenges, accurately delivering drugs to target tissues and improving patient quality of life have become the core objectives for NADs development and research. Recently, significant advancements have been made in chemical modification technologies and delivery vehicles for NADs, considerably enhancing delivery efficiency.¹⁶ Thus far, dozens of NADs have benefited from successful chemical modification and carrier delivery and have earned FDA approval. For example, the success of COVID-19 vaccines is mainly due to advancements in base modification and LNP delivery systems. Therefore, the focus of NADs research and development has shifted toward improving nucleic acid modifications and developing efficient, safe, and targeted delivery systems.

Chemical modification

Various chemical modification methods have been introduced with the continuous progress of chemical synthesis and modification technologies to provide more precise treatments.^223,224 NADs can be precisely modified to improve their efficacy and stability while decreasing toxicity and immunogenicity.^223,224 The most widespread modification methods include backbone, ribose, and nucleobase modifications.^225–228

Backbone modification

The modification strategies used in first-generation NADs have focused on modification of the phosphate backbone, mainly using other types of groups to replace the non-bridging oxygen atoms in the phosphate backbone, such as phosphorothioate (PS), methyl phosphate, and boranophosphate.²²⁹ The most often used backbone modification method is PS modification, where the oxygen atom is replaced by sulfur.²³⁰ This modification can improve the resistance of nucleic acids to nucleases, enhance blood stability, reduce renal clearance, and prolong the circulation time of drugs in vivo by improving their binding to plasma proteins.^231–233 However, it has been determined that while improving stability, PS modification can induce inflammatory responses and produce hepatotoxic effects.²³¹ Fomivirsen, a representative drug modified by PS, was withdrawn due to its limited therapeutic effect and inflammatory reaction.²³⁴ This has inspired the exploration of new modification technologies to reduce the adverse effects associated with PS modification.

Ribose modification

Ribose modification is another common strategy that has had notable success. Changes to the 2’ position group can affect nucleic acid stability and affinity,²²⁷ with common modifications including 2’-fluoro (2’-F), 2’-O-methoxyethyl (2’-MOE), and 2’-O-methyl (2’-OMe).^17,228,235 These modifications considerably increase nuclease tolerance and prolong the half-life of nucleic acids while effectively avoiding the inflammatory reactions triggered by PS modification, demonstrating a higher safety and activity. Furthermore, synergistic modifications combining PS and 2’-MOE have considerably improved the physicochemical properties and reduced NADs side effects. Dual modifications, such as 2’,4’- and 2’,5’-sugar modifications, have been employed in the development of siRNA therapeutics to further expand the potential of ribose modifications.²³⁶ Beyond specific site modifications, altering the sugar ring provides another NADs design strategy.²³⁷ A locked nucleic acid (LNA) is a nucleic acid analog with a unique bicyclic structure, where the C4’ and O2’ atoms are connected by different methylene bridges, forming a stable C3’-endo conformation.²³⁸ An LNA adheres to Watson-Crick base pairing and has a strong affinity for DNA and RNA. However, owing to its bicyclic backbone structure limitations, an LNA can only spontaneously form A-type hybrid duplexes with target nucleic acid strands.^239,240 These hybrids have strong thermal stability and can activate RNase H degradation activity under specific conditions, indicating LNA’s potential as an antisense drug. Overall, LNA’s good stability, high binding specificity, and strong nuclease resistance make it advantageous for in vitro and in vivo applications.¹⁷ In contrast, unlocked nucleic acid (UNA) is a flexible RNA mimic that lacks chemical bonds between the ribose ring’s C2’ and C3’ atoms. A UNA can attach to oligonucleotide monomers to form hybrids, regulating their flexibility and thermal stability.²⁴¹ A UNA supports RNase H activity, which is beneficial for antisense-based nucleic acid therapies.²⁴² Additionally, UNA modification at the siRNA terminus improves siRNA stability and silencing efficacy while considerably reducing off-target effects, highlighting its potential in the development of new therapeutic siRNAs.^243,244

Nucleobase modification

Nucleobases are essential components of nucleic acids, and changes to their structure can affect the stability, biological activity, and immunogenicity of nucleic acids.^245,246 By modifying specific sites on nucleobases, the stability and affinity of nucleic acids can be greatly improved. Canonical nucleoside analogs formed from nucleobase modifications include 5-methylcytidine (m5C), 5-fluorouracil (5-FU), N7-methylguanosine (m7G), pseudouridine (Ψ), N6-methyladenosine (m6A), and 2’-deoxy-2’-fluoro-uridine (2’-FU).^247–250 Modifications, such as m5C and Ψ, reduce the activity of cytokines and biomarkers in dendritic cells, helping mRNA evade the immune system.^49,251,252 Yoshida et al. found that base modification could considerably reduce the hepatotoxicity of gapmer ASO,²⁴⁵ providing insights for developing new gapmer ASOs. Additionally, in RNAi processes, the 5’ nucleobase affects the binding activity of AGOs to siRNA, thus reducing target cleavage activity.^253,254 Therefore, chemical modification at this specific position could improve siRNA binding affinity.²²⁵

Nucleic acid analogs offer new modification technologies, considerably enhancing the stability and reducing the immunogenicity of NADs in vitro and in vivo.^255,256 Replacing the phosphate backbone by other group improves target affinity, nuclease resistance, and pharmacokinetic properties of nucleic acids. For peptide nucleic acids (PNAs), the sugar-phosphate backbone is replaced by a peptide backbone, which retains the specific binding ability with DNA/RNA and offers higher affinity and better stability than natural nucleic acids.²⁵⁷ Due to these backbone modifications, PNAs show improved resistance to nuclease and protease digestion.²⁵⁸ These characteristics make PNAs powerful tools for disease diagnosis and treatment.^259,260 However, their poor distributions in vivo and low cellular uptake are challenges for their clinical application. Phosphorodiamidate morpholino oligonucleotides (PMOs) are another class of nucleic acid analogs with significant potential, which are characterized by a six-membered morpholine ring backbone.²⁶¹ Several PMO-based NADs have been approved for treating DMD, including eteplirsen (Exondys 51®),²⁶² golodirsen (Vyondys 53^®),⁸⁰ viltolarsen (Viltepso^®),²⁶³ and casimersen (Amondys^®).^264,265 Both PMOs and PNAs are neutral nucleic acid analogs with weak binding to plasma proteins, allowing them to be easily cleared by the kidneys.²⁶⁶ In practice, high doses are required to maintain the therapeutic effects, which can lead to corresponding toxicities and side effects.²⁶⁷

Delivery systems

LNPs

LNPs are self-assembled nanostructures with diameters of approximately 100 nm, capable of combining with negatively charged nucleic acids through electrostatic interactions. They have been widely used for NADs delivery due to their excellent compatibility with cell membranes.^268,269 Classical LNPs typically include ionizable lipids (ILs) or cationic lipids (CLs), auxiliary lipids, cholesterol, and Polyethylene glycol (PEGylated) lipids (Fig. 4a).²⁷⁰ These components self-assemble into monodisperse nanoparticles in specific proportions through intermolecular interactions, encapsulating NADs in their core to protect them from nuclease degradation during delivery. Furthermore, modifying the surface properties of LNPs can enhance the uptake by specific cells and alter the distribution of NADs.²⁷¹

Fig. 4.

Chemical structure of NADs delivery systems. a There are four types of LNPs: ILs (or CLs), auxiliary lipids, cholesterol, and PEGylated lipids. b Schematic and molecular structural formula of cationic polymeric nanoparticles. c Triantennary GalNAc moiety conjugated to siRNA or ASO. d Engineered exosome with RVG-LAMP2B displayed on the outer surface. The exosome contains therapeutic nucleic acids, such as siRNA, microRNA, and ASO. e Schematic of inorganic nanoparticles. f Peptide-assisted NADs delivery strategies. The methods of covalent conjugation include disulfide, amide, maleimide, thiazolidine, oxime, and thioether bond. The methods of non-covalent complexation include hydrophobic and electrostatic interactions

In early designs, CLs were permanently charged to bind the cell membrane and NADs effectively. However, their cytotoxicity limited their application in nanoparticle design.^272,273 ILs are crucial for delivery efficacy, as they can encapsulate nucleic acids in LNPs while affecting the uptake and endosomal escape of NADs. ILs remain neutral at physiological pH, reducing the toxicity and immunogenicity of drugs. At low pH values, ILs acquire a positive charge, interacting with the negatively charged endosomal membrane and transforming its planar bilayer structure into a more hexagonal configuration. This transformation promotes the escape and release of NADs encapsulated in LNPs.^274,275 1,2-Dioleyloxy-3-dimethylaminopropane (DODMA) and its analog 1,2-dioleyloxy-3-(dimethylamino) propane (DODAP) were among the first ILs used for RNA delivery.²⁷⁶ Continuous optimization of ILs revealed that the pKa of ILs is a critical factor in delivery efficacy. An optimal pKa value of 6.2–6.5 maximizes NADs efficacy.²⁷⁷ Researchers have synthesized ILs, such as DLin-KC2-DMA and DLin-MC3-DMA, with the latter used as a novel component for encapsulating nucleic acids. DLin-MC3-DMA was integral to the development of the first FDA-approved siRNA drug, Onpattro®, and has been proven effective for mRNA delivery in vivo.^278,279 In addition, this technology is used in the clinically approved vaccines ALC-0315 (BNT162b2-Comirnaty®) and SM-102 (mRNA-1273-Spikevax®), where the pKa value ranges from 6.1 to 6.7.^278,280 Unlike DLin-MC3-DMA, these vaccines incorporate an ester-based structure in the hydrophobic tail of lipids, enabling faster lipid clearance and improving product tolerance.²⁸¹

PEGylated lipids are another crucial component of LNPs that significantly influence their size, stability, and in vivo distribution.²⁸² However, they have encountered several challenges.^283–285 PEG modification can inhibit the binding of apolipoprotein E to LNPs, a key mechanism for LNP uptake by the liver, thereby affecting liver uptake of LNPs. Moreover, repeated administration of PEGylated products may produce PEG antibodies, potentially reducing drug efficacy.^286,287

1,2-Distearoyl-sn-glycero-3-phosphocholine (DSPC) and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) are common helper lipids in commercially available LNPs. DSPC, a phospholipid molecule with a choline head group and two saturated octadecyl chains, is a critical excipient in Onpattro® and COVID-19 mRNA vaccines.²⁸⁸ Alternatively, DOPE disrupts the endosomal membrane structure by forming a hexagonal crystalline phase, facilitating the release of encapsulated RNA from endosomes.²⁷⁵ Addressing the limitations of traditional phospholipid structures, Liu et al.²⁸⁹ developed hundreds of ionizable phospholipids called iPhos, which were capable of promoting endosomal membrane fusion and inducing hexagonal phase transitions. Cholesterol possesses a strong membrane fusion capacity, facilitating the internalization of NADs and their entry into the cytoplasm. Studies indicate that apolipoprotein E can cause cholesterol to migrate from the core of LNPs to the outer lipid layer, potentially altering the surface properties of LNPs.^290,291 Additionally, Siegwart’s team introduced the fifth component, selective organ targeting, to achieve precise and universal extrahepatic-targeted mRNA delivery strategies.²⁹²

Approximately 150 molecules employ LNP technology, with nearly 75% dedicated to delivering NADs in RNA therapy, including ASOs, siRNA, and mRNA.^293,294 Several new RNA therapeutics are in late-stage clinical trials, such as mRNA-1944 (NCT03829384)²⁹⁵ and NTLA-2001 (NCT06128629).²⁹⁶ Vaccines, such as BNT162b2 and mRNA-1273, have been developed based on mRNA LNP technology against COVID-19, which demonstrate the feasibility of LNPs in delivering mRNA encoding the spike protein of syndrome coronavirus 2 (SARS-CoV-2), thus eliciting an immune response against the virus.²⁹⁷

Cationic polymeric nanoparticles

Delivery systems for NADs based on synthetic or natural cationic polymers are complex structures formed through electrostatic interactions between polymers carrying cationic groups and NADs. Upon cell internalization via endocytosis mechanisms, these nanoparticles disrupt endosomal membranes through the proton sponge effect, facilitating the intracellular delivery of exogenous nucleic acids. Various cationic polymers, including polyethylenimine (PEI), polyamidoamine dendrimers (PAMAM), poly-L-lysine (PLL), poly-β-aminoester (PBAE), and poly-lactic-co-glycolic acid (PLGA), have been developed for this purpose (Fig. 4b).^233,298 As a representative cationic polymer, PEI is widely used in non-viral gene vectors in linear PEI (LPEI) or branched PEI (BPEI).^299,300 BPEI exhibits stronger gene compression ability than LPEI due to its structure containing primary, secondary, and tertiary amine groups every two carbon atoms, enhancing electrostatic interactions with nucleic acids.³⁰¹ With a molecular weight of 25 kDa, PEI is considered the “gold standard” for gene transfection due to its high efficiency, but it induces significant cytotoxicity and lacks biodegradability.^302,303 PLL, another widely used cationic polymer, is synthesized through ring-opening polymerization of its monomers.³⁰⁴ Its positively charged amino acids interact electrostatically with the phosphate backbone of NADs, enabling the delivery of RNA and DNA.^305,306 Despite the biodegradability and biocompatibility of PLL, its in vivo activity remains limited due to its high toxicity, low transfection efficiency, and poor endosomal escape ability. Functional modifications, such as chloroquine, histidine, PEG, or PEI, have been incorporated into PLL to address these limitations.^304,307,308 Recent advancements include PLL nanoparticles coated with hyaluronic acid shells, significantly reducing cytotoxicity, enhancing cellular uptake, and improving gene expression.³⁰⁵

N-Acetylgalactosamine (GalNAc)

Most approved products based on bio-conjugated delivery systems use the GalNAc-coupled delivery system developed by Alnylam Pharmaceuticals. GalNAc, a carbohydrate compound, exhibits a high affinity for the asialoglycoprotein receptor (ASGPR).³⁰⁹ ASGPR is an endocytic receptor notably overexpressed on the hepatocyte membrane surface.^310,311 Interaction between GalNAc and ASGPR facilitates the internalization of GalNAc-bound compounds from the cell surface into endosomes via clathrin-dependent receptor-mediated endocytosis.³¹² As the endosome matures and pH decreases, ASGPR dissociates from the GalNAc conjugate and recycles back to the hepatocyte surface.^313–315 At the same time, GalNAc is degraded, releasing NADs into the cytoplasm to initiate gene regulatory activity.

Leveraging this mechanism, Alnylam has extensively explored GalNAc conjugation with NADs, particularly siRNA, to achieve liver-specific delivery (Fig. 4c).^312,316,317 Notably, the nucleic acid molecules in these conjugates are directly exposed to serum; enhancing their stability in the physiological environment is a crucial challenge. Studies have demonstrated that stability can be significantly enhanced by extensive chemical modification at the 2’ position of nucleotide sugars and by replacing the phosphodiester bond with a thiophosphate bond.^318–320 Furthermore, stability can be improved without compromising drug activity by optimizing the number and modification positions of 2’-F and 2’-OME groups on both strands of double-stranded siRNA, leading to substantial efficacy improvements.^319,321,322 These advancements have accelerated the shift from standard template chemistry to enhanced stability chemistry. Rapid ligand-receptor binding and efficient uptake by target cells are critical for receptor-targeted delivery systems to avoid in vivo clearance.^{315,323–325} Multivalent ligands show significantly enhanced binding affinity for ASGPR compared with monovalent GalNAc units, with affinity rankings in the order of tetraantennary > triantennary » biantennary » monoantennary.³²⁶ Studies by Biessen³²⁷ have suggested that a 2-nm interval between GalNAc and dendritic branch points may optimize NADs transportation efficacy. Additionally, hydrophobic linkers have been found to enhance GalNAc–ASGPR interactions.³²⁸

With the success of Alnylam in GalNAc, other companies have focused on RNA therapy and developed their delivery systems. Dicerna Pharmaceuticals has made considerable efforts to advance GalNAc delivery technology by adopting tetraantennary GalNAc conjugates (GalXC) for NADs delivery. In this conjugate, a unique four-ring structure is introduced on the passenger chain, enhancing the stability of the conjugate.^310,329 Additionally, it accurately targets multiple GalXC ligands and successfully delivers siRNA to hepatocytes. Many candidates that involve this technology are currently undergoing clinical evaluation. Nedosiran,³³⁰ a siRNA targeting the silencing of lactate dehydrogenase A in hepatocytes (LDHA) with four covalently linked GalNAc, was approved by the FDA in 2023.

Exosomes

Exosomes are vesicle-like structures with a diameter of 40–160 nm released after the fusion of intracellular multivesicular bodies (MVBs) with the cell membrane.^331,332 As a bridge of intercellular communication, classical exosomes have a monolayer structure, which can encapsulate bioactive substances, such as proteins, lipids, DNA, and RNA, in the core and deliver them to effector cells to play specific biological functions (Fig. 4d).^333–336 They have been considered good candidates for NADs delivery.^337–340 For example, Kaban et al.³⁴¹ used exosomes derived from natural killer (NK) cells as a carrier of siRNA targeting BCL-2 to treat patients with estrogen receptor-positive (ER + ) breast cancer, resulting in enhanced apoptosis of breast cancer cells. Additionally, another study used plasma exosomes to deliver siRNA to T cells and monocytes and caused post-transcriptional gene silencing in recipient cells.³⁴² Furthermore, exosomes can easily cross biological barriers, such as the blood-brain barrier (BBB), which has good application potential in extra-hepatic-targeted NADs delivery.^343–345

When exosomes are used as NADs delivery vehicles, it is essential to consider the strategy of efficient drug loading. As a natural barrier of exosomes, the membrane structure of the lipid bilayer can protect NADs from external influences. Still, the existence of the membrane structure makes it difficult for exosomes to load drugs efficiently. The strategies of drug loading by exosomes are mainly divided into two categories.^346–348 The first is the exogenous route, in which the NADs are directly introduced into the obtained exosomes via electroporation, co-incubation, sonication, extrusion, and freeze-thaw cycling.^349–352 Although cargo uploading exogenously is simple and convenient to operate, the integrity of exosomes may be damaged during the loading process, which may affect the effect and require additional purification steps to remove the unloaded drugs.^353,354 The other method is the endogenous pathway, which uses the endogenous pathway of exosome generation to indirectly improve the production of exosomes by promoting the expression of target nucleic acids and exosome secretion in productive cell lines.³⁵⁵ However, due to our lack of understanding of exosome biology, structure, and biogenesis, the strategy of endogenous drug loading still requires further research and optimization.

One common disadvantage of exosomes is their random movement in vivo and lack of specific targeting.^355,356 The abundant lipids and membrane-bound proteins on the surface of exosomes provide binding sites for targeting ligands, such as peptides, antibodies, and aptamers. These ligands can be stably attached to the surface of exosomes through covalent bonds to enhance their targeting ability. For example, Kim et al.³⁵⁷ modified exosomes with transferrin receptor-binding peptide (T7 peptide), which efficiently delivered AMO-21 into glioblastoma (GBM) cells in vitro. Furthermore, in vivo delivery results have shown that T7 peptide-modified exosomes effectively reduced the level of miR-21 in tumor cells and inhibited tumor growth compared with unmodified exosomes. Another study used a single-chain variable fragment (scFv) to modify exosomes derived from human umbilical cord blood mesenchymal stem cells (MSCs).³⁵⁸ In addition, exosomes can be genetically engineered to express ligands on their surfaces simultaneously. The most commonly used exosome surface protein is lysosome-associated membrane protein (LAMP). The N-terminus of LAMP-2B is located on the surface of exosomes and can specifically target given sequences.^359–362 After screening for cell-specific binding peptides for particular organs or tissues, such as the rabies virus glycoprotein peptide, LAMP-2B can be genetically modified to achieve targeted delivery (Fig. 4d).

Thus far, approximately 40 companies worldwide, including Codiak Biosciences, Evox Therapeutics, Tavec Pharmaceuticals, Carmine Therapeutics, Anjarium, and Micromedmark Biotech, have developed exosome-based therapies, which are expected to provide cost-effective and more accurate targeted therapies in the clinic (https://bioinformant.com/companies-developing-exosome-technologies/). However, challenges, such as large-scale production, purity, and batch homogeneity of exosomes, still limit their clinical application. Additionally, there are no regulations for the control of therapeutic drugs based on exosomes that consider safety, effectiveness, and quality control, highlighting the urgent need for standardized methods and principles to manage these molecules.³⁶³

Inorganic nanoparticles (INPs)

INPs are nanocarriers based on inorganic substances that have attracted considerable attention due to their unique electrical and optical properties, biocompatibility, and low cytotoxicity. Commonly used INPs include gold nanoparticles (AuNPs), silica nanoparticles (SiNPs), magnetic nanoparticles (MNPs), and carbon nanotubes (CNTs) (Fig. 4e).^364–366 Unlike other delivery carriers, INPs possess a stable and robust structure, a large specific surface area, and tunable surface properties, allowing precise control of the drug delivery process through surface functionalization and controlled release modifications.

AuNPs are widely developed INPs with excellent biocompatibility and low toxicity. Their flexible surfaces enable nucleic acids to bind to the gold nanoparticles directly.^367,368 For example, Shrestha et al.³⁶⁹ developed a gold nanoparticle-mediated drug delivery platform for the co-delivery of doxorubicin and polo-like kinase 1 (Plk1) siRNA, offering an adaptable and straightforward platform for studying drug-siRNA combinations in cancer treatment. Notably, NU-0129, composed of siRNA targeting the GBM oncogene Bcl2Like12 (Bcl2L12) and a gold nanoparticle core, was the first spherical nucleic acid (SNA) drug administered systemically.³⁷⁰ The Phase 0 clinical study (NCT03020017) in eight patients with GBM showed that NU-0129 could pass through the BBB and accumulate in tumors, reducing the abundance of BCL2L12 protein and demonstrating its potential as an innovative therapy for GBM.³⁷¹

MNPs are typically made of magnetic materials, such as iron oxide or iron platinum, and are usually coated with biocompatible materials to enhance their stability and biocompatibility. Under the influence of an external magnetic field, therapeutic drugs or molecules can be loaded onto magnetic nanoparticles and targeted to precise regions in the body.^372,373 The unique superparamagnetic nature, lower toxicity, and site-specific targeting capabilities of MNPs make them excellent nanocarriers for NADs delivery.³⁷⁴

CNTs are cylindrical structures composed of a hexagonal arrangement of sp2 hybridized carbon atoms, also known as graphene.³⁷⁵ CNTs can be categorized into single-walled CNTs (SWCNTs) and multi-walled CNTs (MWCNTs). SWCNTs consist of a single layer of graphene sheets rolled seamlessly into a cylindrical tube. In contrast, MWCNTs are composed of multiple graphene layers wrapped around each other in a cylindrical shape. Although CNTs have poor solubility in both water and organic media, they can be chemically modified to improve their solubility, degradation ability, and drug-loading capacity while reducing toxicity.³⁷⁶ This simple surface functionalization has made CNTs promising carriers for NADs delivery in various diseases.³⁷⁷

Moreover, SiNPs are considered excellent carriers for NADs delivery.³⁷⁸ Research has shown that porous silicon nanoparticles (pSiNPs) can replace commonly used viral vectors or lipid transfection reagents as novel vectors for delivering siRNA to dendritic cells.³⁷⁹ Luo designed pSiNPs with ASO as targeted gene and drug delivery platforms for GBM treatment. These pSiNPs penetrate the BBB monolayer in vitro and target the brain after intravenous injection in an in situ GBM mouse model. This indicates that pSiNPs and their multifunctional strategies have strong potential for cancer treatment and gene delivery research.³⁸⁰

Peptides

Peptides comprise fewer than 100 amino acid residues, and they are between small molecules and proteins in size. Historically, peptides have been recognized for their diverse roles as hormones, signaling molecules, carriers, and supplements.^381–384 They are easily synthesized, possess relatively stable chemical properties, and exhibit a high selectivity. Due to their low immunogenicity and strong specific targeting capacity, peptides have emerged as promising carriers for selectively delivering NADs through various modalities (Fig. 4f).^385,386 Peptides exhibit multiple functions, including tissue targeting,^387–389 membrane penetration,^390–392 endosome escape,^393–395 and nuclear localization,^396–398 depending on their amino acid composition. Several peptides have been developed specifically for NADs delivery based on these unique properties for cell targeting, penetration, endosomal escape, and sub-organelle targeting (Fig. 5).^25,399–403

Fig. 5.

Schematic of peptide-assisted NADs delivery. a Classes of peptides facilitating the delivery of NADs across biological barriers. b Classes of NADs. c The delivery methods of peptide-based carriers include peptide–oligonucleotide conjugates, peptide-based nanoparticles, and peptides in combination with other delivery systems. d Peptides mediate the entry of NADs into cells and transfer them across the cell membrane, complete endosomal escape, and eventually release NADs in the cytoplasm, mitochondrion, nucleus, endoplasmic reticulum, lysosomes, and Golgi apparatus

Targeting peptides

Nonspecific cell binding is the leading cause of the off-target effects observed with several therapeutic molecules. The arginine-glycine-aspartic acid (RGD) sequence has been extensively applied in tumor-targeted therapy, mainly targeting αvβ₃ receptors overexpressed on several tumor cells.^404,405 A recent study demonstrated that siRNA conjugated to cyclic RGD (cRGD) could selectively enter cells that express αvβ₃ integrin.⁴⁰⁶ Intravenously injected cRGD-siRNA molecules produced no innate immune response in mice with transplanted tumors.⁴⁰⁷ Additionally, researchers have found that incorporating siRNA into PEG-modified polylysine with cRGD can enhance gene silencing ability, improve cell uptake, inhibit glioma angiogenesis, and delay tumor progression after systemic administration in mice.⁴⁰⁸

The acidic extracellular microenvironment has become an effective disease diagnosis and treatment target.⁴⁰⁹ pH low insertion peptides (pHLIPs), derived from the C-helix of bacteriorhodopsin, can sense the pH value near the plasma membrane and deliver drugs to pathological tissues.⁴¹⁰ In acidic environments, the elevated proton concentration increases the hydrophobicity of peptides. pHLIPs spontaneously fold to form α helices, which insert and cross the cell membrane to transport conjugated NADs into cells.^191,411 In anti-tumor therapy, it has been demonstrated that pHLIPs can deliver PNAs to cancer cells due to their natural targeting abilities.^412,413 For example, in a mouse model of lung adenocarcinoma, the delivery of PNAs targeting CEACAM6 using pHLIPs resulted in CEACAM6 gene silencing and tumor growth inhibition.⁴¹⁴

Cell-penetrating peptides (CPPs)

CPPs are a short peptide class consisting of 5–30 amino acids that can cross the cell membrane directly.³⁹⁹ They can penetrate the cell membrane by themselves and effectively enhance the intracellular delivery of NADs, aiding their interaction with the target to achieve therapeutic goals.^415–417

Despite their significant therapeutic potential, CPPs have some limitations. One major issue is their lack of selectivity, as they can carry NADs into almost all cell types. To address this, researchers have developed strategies such as local administration, targeting ligands on CPPs, or using activatable CPPs that penetrate only under specific biological stimuli.^418,419 Ensuring the stability and biological activity of CPPs during the delivery process is also a challenge. Modifications can be made to the termini of existing CPPs or the peptide backbone can be adjusted, such as through peptide cyclization, to improve stability and cell permeability.^420,421 Another limitation of CPPs is their limited clinical efficacy, often requiring high concentrations and doses to achieve a therapeutic effect, which can cause significant toxicity and side effects, hindering the development of clinical applications. Current research has found that inducing peptides to form multimers and using multifunctional fusion peptides can improve the delivery efficiency of CPPs to some extent.^422–424

Endosomal escape peptides

Endosomal escape is a significant challenge for non-viral NADs delivery systems. Typically, after arriving at a specific cell, NADs pass through the cell membrane and reach the cytoplasm through various mechanisms, such as direct penetration or endocytic uptake.^309,422 However, peptide carriers and their cargos often become trapped in acidified endosomes, preventing their entry into the cytoplasm, nucleus, and other subcellular compartments.^14,425

Fusogenic peptides are short peptides that promote endosomal release by enhancing interactions with endosomal membranes.^426,427 In a low pH environment, fusogenic peptides undergo pH-dependent conformational changes to form α-helices, which insert into the endosomal membrane, causing instability and decomposition of the membrane structure. This allows internalized NADs to dissociate from the fusogenic peptide delivery system and escape into the cytoplasm to exert their therapeutic effects.⁴²⁷ Many fusiform peptides of viral and non-viral origin, such as the INF7 family,⁴²⁸ L17E family,⁴²⁹ GALA/KALA family,^430–432 and HA family,⁴³³ can disrupt endosomal membranes and increase cytoplasmic delivery. One study evaluated the ability of the H5WYG peptide to deliver NADs and promote endosomal release, confirming the elevated endosomal escape function of this clostridial peptide.⁴³⁴ Many studies have shown that self-assembled peptide carriers modified with fusogenic peptides can overcome intracellular delivery obstacles.^435,436 For instance, the fusogenic peptide L17E was linked to peptide self-assembled disks using click chemistry. Compared with unmodified disks carrying plasmid DNA, nanodisks modified with L17E demonstrated enhanced endosomal escape and improved transfection efficiency in cell culture.⁴²⁹

Peptide-targeting subcellular organelles

The structural integrity and functional stability of organelles are essential for maintaining normal cellular physiological functions. Dysfunction of these organelles can lead to various diseases, including diabetes, neurodegenerative diseases, CVDs, and cancer, making them potential therapeutic targets.^437–439 As mentioned earlier, the efficacy of NADs therapy largely depends on the efficient and safe delivery of NADs to their specific action sites within the body.¹⁴ After escaping from endosomes, the next challenge is transporting NADs quickly and effectively to the nucleus or specific organelles.⁴²⁵ Peptide targeting of subcellular organelles offers a promising strategy to guide NADs to dysfunctional organelles, thus achieving optimal therapeutic effects. This approach provides a potential method to overcome the intracellular delivery obstacles and enhance the efficacy of NAD-based therapies.

Multifunctional peptides

In previous clinical trials, using individual functional peptides for NADs delivery showed limited efficacy due to their inability to overcome various delivery obstacles.⁴⁴⁰ To address this issue, researchers have integrated different peptides or functional domains capable of overcoming transfer barriers into a single peptide-based carrier to enhance delivery efficiency.^441–444 Compared with individual functional peptides, multifunctional peptide carriers considerably improve transfection efficiency and have demonstrated outstanding potential in clinical applications.^445–447 For example, amino acid pairing (AAP) peptides are a novel class of self-assembled biomolecules comprising two main structural domains: an amino acid pairing domain and a cell permeability domain.⁴⁴⁸ AAP peptides possess recognition and membrane-targeting functions, facilitating gene delivery through interactions with DNA and siRNA.⁴⁴⁸

The discovery of peptide-based NADs delivery systems is accelerating by the use of various peptide combinations and high-throughput screening, providing new directions for optimizing and developing peptide carriers.^449–451 For example, Li et al.⁴⁵² designed a multifunctional peptide vector (PLD-R9-G-NLSW) containing the CPP R9, NLS, and 2,3-dimethylmaleic anhydride-modified PLL. This vector condenses with the pIRES-VEGF plasmid to form a complex for gene delivery. In vitro results showed that the complex considerably improved gene internalization and transfection efficiency while reducing cytotoxicity. However, researchers have discovered that the immediate coupling of individual peptides can affect the function and activity of the units. Given the likelihood that distinct peptides may interact with each other during the fusion process, the final functional characteristics of the peptides may not involve a simple integration.^12,453 This phenomenon suggests that the interaction mechanism between combined peptides and the structure-activity relationship needs to be explored in more detail.

Recently, peptides have been applied as ligands to achieve delivery functions in complexes known as peptide-oligonucleotide conjugates (POCs).^454,455 In this approach, oligonucleotides or their analogs are covalently linked to one or more peptide residues to form POCs. Various methods, including disulfide bonds, thioether bonds, thiol-maleimide bonds, phosphodiester bonds, and click chemistry, are used for this covalent connection.^456–458 The primary advantage of covalent conjugation is that the resulting product is a single compound with defined structural and stoichiometric characteristics, aligning with the ideal properties for drug design and in vivo application.^459,460 However, covalent conjugation also has drawbacks. Interactions between cationic peptides and negatively charged oligonucleotides can enhance toxicity, affect pharmacokinetic and pharmacological properties, and limit the efficacy of these therapies.⁴⁶¹ Additionally, the interaction between anions and cations can complicate large-scale preparation and purification of these conjugates.⁴⁶²

The most advanced application of POCs is the delivery of charge-neutral oligonucleotides, such as PNA and PMO, using CPPs.^463–465 CPP-PNA conjugates with Tat and R8 have been successfully applied to PNA delivery^466,467 and showed sound therapeutic prospects in anti-virus, antibacterial, and anti-inflammatory therapy. Beyond CPPs, other functional peptides have also been exploited for PNA delivery. For instance, Soudah et al.⁴⁶⁸ used a conjugate of the cytoplasmically localized internalized peptide (CLIP6) and PNA to treat GBM cells, considerably upregulating the tumor suppressor Mnk2a and promoting cancer cell death. Kaplan et al.⁴⁵⁷ showed that pHLIP-αKu80 (γ), which was created by covalently conjugating pHLIPs with PNA targeting KU80 via disulfide bonds, could selectively reduce Ku80 expression under acidic conditions. In mice, intravenous injection of fluorescently labeled pHLIP-αKu80 (γ) targeted tumors and reduced Ku80 expression.

The development of peptide-phosphoryl diamine morpholino oligonucleotide (PPMO) conjugates has further advanced NADs delivery, particularly for neuromuscular diseases, such as spinal muscular atrophy (SMA) and DMD.^469,470 A recent study demonstrated that subcutaneous injection of a CPP-DG9 and PMO conjugate improved muscle strength and innervation in mice with severe SMA, significantly extending their median survival with no apparent side effects.⁴⁷¹ To enhance the delivery of neutrally charged oligonucleotides, Gait and Wood⁴⁷² developed a series of arginine-rich CPPs named [PNA/PMO]-internalized peptides (Pip), including Pip2a, Pip2b, Pip5e, and Pip6a. Studies have shown that Pip6a-PMO effectively rescues the disease phenotype and improves the survival rate of severe SMA mice, showing strong efficacy in the central nervous system and peripheral tissues.^473,474 Pip6a-PMO predicts a potent therapeutic option that combines the genetic precision of SSO with the systemic delivery efficacy of peptides.⁴⁷⁵

With technological advances, various novel POCs have demonstrated the feasibility of delivering siRNA, ASO, and miRNA in vitro.^406,476 Studies have found that bivalent cRGD successfully transports VEGF receptor 2 (VEGFR2)-siRNA into tumor cells in a mouse non-small cell lung cancer xenograft model, downregulating VEGFR2 expression and significantly inhibiting cancer progression.⁴⁰⁶ Kim et al.⁴⁰² designed a dual-targeting drug delivery system for miR-21 inhibitors, consisting of a PDL1-binding peptide covalently linked with an anti-miR-21 inhibitor via a click reaction. Pep-21 treatment specifically silenced target miR-21 in B16 melanoma cells and M2 macrophages, reducing tumor cell migration and inhibiting tumor progression. PepGen developed a POC drug, PGN-EDO51, using its enhanced delivery oligonucleotide platform, which skips exon 51 in the DMD gene. In Phase I clinical trials, PGN-EDO51 showed good safety, tolerability, and efficacy, and Phase II trials are ongoing.

Electrostatic or hydrophobic interactions between peptides and NADs create self-assembling peptide-based nanoparticles.^477–479 These complexes protect nucleic acids from nuclease degradation and reduce the likelihood of adverse reactions.⁴⁸⁰ Using this strategy, Ryu et al.⁴⁸¹ developed an S-R11 fusion peptide (space peptide bound to polyarginine) that forms stable nanocomposites with siRNA molecules through electrostatic attraction and hydrogen bonding, facilitating siRNA delivery to difficult-to-transfect immune cells. Sirnaomics’ histidine-lysine copolymer peptide nanoparticle (PNP) delivery platform has demonstrated clinical therapeutic potential for RNAi therapy.⁴⁸² In contrast, non-covalent strategies have been used for delivering siRNA, plasmid DNA, and splicing correction oligonucleotides.^483,484 The peptide p5RHH has been efficiently used to transfect siRNA and mRNA into human cartilage to reduce chondrocyte apoptosis and prevent cartilage degeneration^485–487 as well as into cardiac tissue for treating abdominal aortic aneurysm to reduce the risk of aortic rupture and sudden death in mice.⁴⁸⁴

Clinical applications of NADs

Theoretically, NADs can cure any disease by allowing the selection of the correct nucleotide sequence on the target gene. Unlike conventional therapeutics, NADs induce long-lasting or curative effects due to their distinct physicochemical and biological properties. Numerous NADs have transitioned from bench to bedside and have successfully been approved for clinical trials (Fig. 6). Several NADs, including ASOs, aptamers, siRNAs, and mRNA vaccines, have been adopted as vaccines for treating rare genetic diseases, cancer, ophthalmic diseases, CVDs, and infections. Table 2 summarizes NADs successfully applied in clinical experiments for treating human diseases.

Fig. 6.

Clinical application of NADs. Several NADs have been adopted as vaccines for treating rare genetic diseases, cancer, ophthalmic diseases, cardiovascular diseases, and infection diseases, and they have shown remarkable therapeutic effects. The box summarizes the therapeutic NADs for various diseases in the clinic or on the market

Table 2.

Status of NADs-based therapeutics clinical trials

Drug (alternate name)	Company	Classification	Indication (Phase)	Target	Modification & Delivery	Clinical Trail Number	Ref(s)
Rare genetic disorders
NTLA-2001	Intellia Therapeutics	CRISPR-Cas9	ATTR-CM (III)	TTR	LNP	NCT06128629	296,588
Zilganersen (ION-373)	Ionis	ASO	Alexander disease (III)	GFAP mRNA	2’-MOE	NCT04849741	589
Fitusiran (ALN-AT3)	Sanofi	siRNA	Hemophilia A/B (III)	Serpin C1	GalNAc	NCT03417102 NCT03417245	590
Donidalorsen (IONIS-PKK-LRx)	Ionis	ASO	HAE (III)	PKK	GalNAc, 2’-MOE	NCT05392114	591,592
Ulefnersen (ION-363)	Ionis	ASO	ALS (III)	FUS	PS, 2’-MOE	NCT04768972	506
Cemdisiran (ALN-CC5)	Alnylam & Regeneron	siRNA	PNH, Myasthenia gravis (III), IgAN, (II)	Complement C5	2ʹ-OME,2’-OF & GalNAc	NCT05070858 NCT05744921 NCT05133531	593,594
Vesleteplirsen (SRP-5051)	Sarepta Therapeutics	ASO	DMD (II)	Exon 51 of DMD	PMO/ PPMO	NCT04004065	500
PGN-EDO51	PepGen	ASO	DMD (II)	Exon 51 of DMD	EDO	NCT06079736	501
ATL-1102 (ISIS-107248)	Antisense Therapeutics	ASO	DMD (II)	ITGA4	2’-MOE	NCT05938023	595
Zorevunersen (STK-001)	Stoke Therapeutics	ASO	Dravet Syndrome (II)	SCN1A	2’-MOE	NCT04740476 NCT04442295	596
IONIS-MAPTRx (BIIB-080)	Ionis & Biogen	ASO	Mild Alzheimer’s Disease (II)	MAPT	2’-MOE, PS	NCT05399888 NCT03186989	/
SQY51	Sqy Therapeutics	ASO	DMD (I/II)	Exon 51 of DMD	tricyclo-DNA	NCT05753462	/
NTLA-2002	Intellia Therapeutics	CRISPR-Cas9	HAE(I/II)	KLKB1	LNP	NCT05120830	597
mRNA-3705	Moderna	mRNA	MMA (I/II)	MUT	LNP	NCT05295433 NCT04899310	598
Cancer
mRNA-4157 (V-940)	Merck & Moderna	mRNA	NSCLC, Melanoma (III)	/	LNP	NCT03313778 NCT06077760 NCT05933577 NCT03897881	599
Autogene Cevumeran (BNT-122)	BioNTech	mRNA	Colorectal, PDAC (II)	unspecified TAAs	LNP	NCT03289962 NCT03815058 NCT05968326	600
IGV-001	Imvax	ASO	GBM (II)	IGF type 1 receptor	PS, Goldspire™	NCT04485949	520
Cotsiranib (STP-705)	Sirnaomics	siRNA	Keloid, BCC, hypertrophic scars, isSCC (II)	TGF-β1、COX-2	PNP	NCT04669808	482
QN-165 (AS1411)	Qualigen Therapeutics	Aptamer	RCC, glioma and AML (II)	Nucleolin	/	NCT00881244 NCT01034410	183,184
Cobomarsen (MRG-106)	miRagen Therapeutics	miRNA	MF, CTCL, DLBCL, CLL, ATLL (I/II)	miR-155 inhibitor	LNA	NCT02580552 NCT03713320 NCT03837457	125,126
olaptesed pegol	TME Pharma	Aptamer	Tumor, GBM (I/II)	CXCL12	/	NCT03168139 NCT01486797	601
TTX-MC138	TransCode	miRNA	advanced solid tumors (I/II)	miRNA-10b inhibitor	iron oxide nanocarrier	NCT06260774 NCT05908773	127
STP-707	Sirnaomics	siRNA	solid tumors (I)	TGF-β1、COX-2	PNP	NCT05037149	/
MTL-CEBPA	MiNA Therapeutics	saRNA	HCC (I)	CEBPA	/	NCT02716012	142–144
RAG-01	Ractigen Therapeutics	saRNA	NMIBC (IND)	CDKN1A	/	/	122
Ophthalmic diseases
Tivanisiran (SYL1001)	Sylentis, S.A.	siRNA	DED (III)	TRPV1	/	NCT05310422 NCT03108664	528
Sepofarsen (QR-110)	ProQR Therapeutics	ASO	LCA10 (III)	CEP290	2’-OME	NCT03913143	602
Ultevursen (QR-421a)	ProQR Therapeutics	ASO	Usher syndrome and nsRP (II/III)	USH2A	2’-OME	NCT05158296	533
SYL1801	Sylentis	siRNA	AMD (II)	NRARP	/	NCT05637255	526
RBM-007 (APT-F2)	Ribomic	Aptamer	wet AMD (II)	FGF2	/	NCT04640272 NCT04895293	525
IONIS-FB-LRx (RO7434656)	Roche & Ionis	ASO	GA (II)	Complement Factor B	2’-MOE & GalNAc	NCT03815825	603
Cardiovascular diseases
Pelacarsen (AKCEA-APO(a)-LRx)	Novartis AG	ASO	CVD (III)	ApoA	2’-MOE & GalNAc	NCT05900141 NCT04023552	542,604
Lepodisiran	Lilly	siRNA	ASCVD (III)	ApoA	2’-MOE, 2’-F & GalNAc	NCT06292013	546
Olpasiran (AMG 890)	Amgen Biopharmaceuticals	siRNA	ASCVD (III)	ApoA	2’-MOE & GalNAc	NCT04270760 NCT03626662 NCT05581303	547,548
IONIS-FB-LRx (RO7434656)	Roche & Ionis	ASO	Primary IgA Nephropathy (III)	Complement Factor B	2’-MOE & GalNAc	NCT05797610	/
Olezarsen (IONIS-APOCIII-LRx)	Akcea Therapeutics & Ionis	ASO	SHTG & FCS (III)	ApoC-III	GalNAc	NCT05681351 NCT05552326 NCT05079919 NCT05130450	605,606
Zodasiran (ARO-ANG3)	Arrowhead Pharmaceuticals	siRNA	HoFH (II)	ANGPTL3	2’-OME, 2’-F & GalNAc	NCT05217667	607
Zilebesiran (ALN-AGT01)	Alnylam	siRNA	Hypertension (II)	Angiotensinogen	2’-MOE, 2’-F & GalNAc	NCT06272487 NCT04936035 NCT05103332	608
MRG-110 (S-95010)	miRagen Therapeutics	miRNA	Chronic heart failure (I)	miR-92a Inhibitors	LNA	NCT03603431	125
Infection diseases
mRNA-1345	Moderna	mRNA	Respiratory syncytial virus infection (III)	RSV fusion	LNP	NCT06067230	609
mRNA-1647	Moderna	mRNA	Cytomegalovirus infection (III)	HCMVgB & HCMVgH	LNP	NCT05085366	610
Bepirovirsen (GSK-3228836)	GSK & Ionis	ASO	HBV (III)	HBV RNA	2’-MOE	NCT05630807 NCT05630820	559
GSK-5637608 (JNJ-73763989)	GSK	siRNA	HBV (II)	HBV RNA	GalNAc	NCT05275023 NCT0512359	560
ALN-HBV02 (VIR-2218)	Alnylam & Brii Biosciences Ltd & Vir Biotechnology	siRNA	HBV (II)	HBV RNA	2’-OME, 2’-F & GalNAc	NCT05612581	561
Imdusiran (ARB-270729)	Arbutus Biopharma	siRNA	HBV (II)	hepatitis B surface antigen	2’-F & GalNAc	NCT06154278 NCT06245291 NCT06277037	562
mRNA-1944	Moderna	mRNA	Chikungunya Virus (I)	CHKV-24	LNP	NCT03829384	295
Other diseases
Sapablursen (IONIS-TMPRSS6-LRx)	Ionis	ASO	PV (II)	TMPRSS6	2’-MOE & GalNAc	NCT05143957	611
CWT001 (TenoMiR)	Causeway Therapeutics	miRNA	Lateral epicondylitis (II)	microRNA-29a mimic	2’-OME, 2’-F	NCT06192927	/
ION224	Ionis	ASO	MASH (II)	DGAT2	2’-MOE & GalNAc	NCT04932512	612
Rapirosiran (ALN-HSD)	Alnylam	siRNA	NASH (II)	HSD17B13	2ʹ-MOE, 2’-F & GalNAc	NCT05519475	/
ALN-KHK	Alnylam	siRNA	T2DM (I/II)	KHK	GalNAC	NCT05761301	/
OLX10010 (BMT-101)	Olix Pharmaceuticals	siRNA	Hypertrophic scars (II)	CTGF	2’-MOE	NCT04877756 NCT03569267	/
Remlarsen (MRG-201)	miRagen Therapeutics	miRNA	Keloid disorder (II)	microRNA-29 mimic	/	NCT03601052	124
AON-D21	Aptarion Biotech	Aptamer	Community-acquired pneumonia (II)	C5a	/	NCT05962606	613

Rare genetic diseases

Transthyretin amyloidosis (ATTR) is a rare systemic disease characterized by the progressive deposition of misfolded TTR protein in the heart and peripheral nerves, leading to transthyretin amyloid polyneuropathy (ATTR-PN) and transthyretin amyloid cardiomyopathy (ATTR-CM).^488,489 Several NADs have been approved for ATTR treatment, including ASOs, such as inotersen^490,491 and eplontersen,⁴⁹² as well as siRNAs, such as patisiran⁹⁷ and vutrisiran.⁴⁹³ These therapies work by disrupting the relevant mRNA, thus inhibiting TTR synthesis. Additionally, researchers are exploring new methods to block TTR production via gene-editing techniques. For instance, a Phase III clinical trial evaluated the efficacy and safety of NTLA-2001, a CRISPR-Cas9-based therapy, in participants with ATTR-CM. In a previous Phase I trial, six patients treated with a single dose of NTLA-2001 showed significant reductions in serum TTR protein levels, with no serious adverse effects or liver damage reported.⁴⁹⁴

DMD is a rare genetic muscle disorder caused by mutations in the dystrophin gene, leading to the absence of the dystrophin protein. Patients typically exhibit symmetric, progressive muscle weakness and atrophy, ultimately leading to premature death due to respiratory and cardiovascular complications.⁴⁹⁵ Significant progress has been made in DMD therapeutic strategies, including NADs. Approved ASO drugs target exon 45 (casimersen),^264,496 exon 51 (eteplirsen),^497,498 and exon 53 (golodirsen and viltolarsen)^80,263 to induce specific exon skipping in the dystrophin gene, thus delaying disease progression. Researchers have developed second-generation skipping drugs, called PPMOs, to enhance targeting and skipping activity.⁴⁹⁹ Sarepta’s SRP-5051, an ASO conjugated with a targeting peptide, has shown higher levels of dystrophin expression and exon skipping rates than eteplirsen. This drug is currently in Phase II clinical trials (NCT04004065).⁵⁰⁰ Additionally, PepGen’s PGN-EDO51, designed to skip exon 51 in the DMD gene, has shown promising results in Phase I clinical trials and is now in Phase II (NCT06079736).⁵⁰¹

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by the dysfunction of upper and lower motor neurons.⁵⁰² Approximately 10% of ALS cases are familial ALS, with various disease-causing genes identified, including superoxide dismutase 1 (SOD1), chromosome 9 open reading frame 72 (C9ORF72), and fused in sarcoma (FUS).^503,504 The first NAD approved for ALS was tofersen (Qalsody®), a PS 2’-MOE ASO targeting SOD1 mutations.^67,505 Ulefnersen (ION-363), an antisense therapy targeting the FUS gene, is currently in Phase III trials for FUS-ALS.⁵⁰⁶ However, clinical trials for C9ORF72-targeted therapies WVE-004 and Tadnersen (IONIS-BIIB5Rx) have been terminated due to poor efficacy.

Primary hyperoxaluria type 1 (PH1) is an autosomal recessive disorder caused by defects in hepatic glycoxidation metabolism, leading to excessive endogenous oxalate production.⁵⁰⁷ It is characterized by recurrent urinary calculi, nephrocalcinosis, and progressive renal damage. PH1 is associated with a deficiency or functional defect in alanine glyoxylate aminotransferase.^508,509 Lumasiran, a siRNA drug that targets hydroxy acid oxidase 1 to reduce hepatic oxalate production, was the first drug approved for PH1.⁵¹⁰ Nedosiran, another RNAi drug developed by Dicerna Pharmaceuticals that targets LDHA, is currently being evaluated for its efficacy in reducing oxalate production.³³⁰

SMA is a genetic neuromuscular disease caused by the deletion or mutation of the survival motor neuron 1 (SMN1) gene, leading to the loss of alpha motor neurons and progressive muscle atrophy.⁵¹¹ The SMN2 gene, highly similar to SMN1, produces mostly non-functional SMN protein due to differences in exon 7 splicing.^512,513 Correcting exon 7 splicing in SMN2 can compensate for the lack of SMN1, producing functional SMN protein and improving symptoms.^514,515 Nusinersen, a modified ASO targeting the intronic splicing silencer N1 of the SMN2 gene, promotes the production of the full-length SMN protein. Several clinical trials have confirmed the efficacy and safety of nusinersen, leading to better motor milestone responses and higher event-free survival rates.^516–518

Cancer

The occurrence and progression of cancer are closely tied to the activation of oncogenes and the loss of tumor suppressor genes. NADs, which target specific genes, offer new methods for cancer treatment. For example, transforming growth factor beta 1 (TGFβ1), a multifunctional cytokine, can alter the tumor microenvironment when dysregulated, promoting angiogenesis and immune suppression, thus influencing tumor progression.^482,519 Therefore, blocking TGF-β signal transduction is a critical strategy in cancer treatment.

Cotsiranib is a siRNA therapy developed by Sirnaomics that uses the PNP delivery platform, containing two active siRNAs targeting TGFβ1 and cytochrome C oxidase subunit 2 mRNA.⁴⁸² A Phase II clinical trial (NCT04669808) for the in situ treatment of basal cell carcinoma showed a positive therapeutic effect and achieved 100% complete clearance. Another anticancer candidate, STP707, is currently being evaluated in a Phase I trial (NCT05037149) for safety, tolerability, and anti-tumor activity in participants with advanced/metastatic or surgically unresectable solid tumors.

In addition, tumor vaccines carrying NADs are considered feasible therapeutic approaches for most solid tumors. IGV-001 is a vaccine targeting GBM, designed based on the Imvax platform Goldspire™. Unlike traditional tumor vaccines, IGV-001 is prepared by mixing tumor tissue removed from the patient’s brain with an insulin-like growth factor 1 receptor antisense oligonucleotide (IMV-001), which inhibits tumor growth. Then, this mixture is encapsulated in biodiffusion chambers with a 0.1-μm pore size, allowing large molecules to permeate and be implanted into the patient’s abdomen.⁵²⁰ This aims to induce a tumor-specific immune response. In earlier clinical trials, IGV-001 demonstrated good tolerance and safety (NCT02507583).⁵²⁰ Imvax is conducting a randomized, multicenter, double-blind, placebo-controlled Phase IIb study (NCT04485949) to evaluate IGV-001 in patients with newly diagnosed GBM.⁵²¹ Additionally, personalized mRNA neoantigen vaccines, such as mRNA-4157, encode multiple antigens to stimulate a patient-specific immune response. Clinical trials are underway for melanoma, non-small cell lung cancer, and other solid tumors. A Phase IIb trial of mRNA-4157 in combination with pembrolizumab for patients with resected high-risk melanoma showed that this combination therapy prolonged recurrence-free survival compared with pembrolizumab monotherapy and had a manageable safety profile.⁵²² Combination therapies for melanoma and non-small cell lung cancer have entered Phase III clinical trials (NCT05933577 and NCT06077760). This is expected to be the first mRNA tumor vaccine on the market.⁵²³

Ophthalmic diseases

AMD is a complex eye disease primarily categorized into dry and wet/neovascular AMD.⁵²⁴ Aptamers have shown promise in AMD treatment. Pegaptanib, a VEGF antagonist with a three-dimensionally structured aptamer, was initially used to inhibit pathological neovascularization and reverse disease progression but was later removed from the market. RBM-007, an RNA aptamer designed to target fibroblast growth factor 2 (FGF2), is currently in Phase II clinical trials for treating neovascular AMD (nAMD).^178,525 FGF2 promotes angiogenesis and retinal fibrosis by stimulating vascular endothelial cell proliferation and inducing VEGF secretion. Unlike traditional anti-angiogenic therapies that target VEGF, Sylentis developed SYL1801, an siRNA targeting NOTCH Regulated Ankyrin Repeat Protein (NRARP), to prevent and control nAMD progression. SYL1801 is administered as eye drops rather than via intravitreal injection, downregulating NRARP expression and inhibiting retinal neovascularization.⁵²⁶ Clinical trials testing various doses of SYL1801 eye drops in healthy volunteers showed good safety and tolerability. Four of the 36 patients tested experienced mild adverse events such as blepharitis, keratitis, hyperemia, and ocular irritation, all of which resolved within 72 h.⁵²⁷ Currently, the drug is in Phase II clinical recruitment to compare the safety and efficacy of three different doses in patients with nAMD (NCT05637255).

Dry eye disease (DED) is a common ocular condition characterized by tear film alterations, ocular inflammation, and neurosensory abnormalities.^528–530 Tivanisiran (SYL1001), an unmodified siRNA eye drop developed by Sylentis, targets the transient receptor potential vanilloid 1 mRNA to alleviate ocular discomfort and pain.⁵²⁸ It has improved the quality of eye redness and tear film in human and animal models. It is undergoing Phase III trials to evaluate its safety in treating DED caused by dry eye syndrome (NCT05310422 and NCT04819269).

QR-110, an ASO drug in Phase III trials, was designed to address Leber congenital amaurosis type 10 caused by the p.Cys998X mutation in the centrosomal protein 290 gene.⁵³¹ QR-110 binds to pre-mRNA, restoring proper splicing and producing fully functional proteins.⁵³² Another candidate, ultevursen (QR-421a), is an ASO therapy for treating vision loss in patients with pigmentary retinopathy due to mutations in exon 13 of the USH2A gene. This drug aims to halt or reverse vision loss by restoring the function of the usherin protein through exon skipping in Usher syndrome type 2a and non-syndromic retinitis pigmentosa.⁵³³

CVDs

CVDs remain a leading cause of mortality worldwide and pose a severe public health challenge. These diseases, including atherosclerosis, myocardial infarction, cardiac hypertrophy, and heart failure, are closely linked to genetic, metabolic, and environmental factors.⁵³⁴ Researchers continually explore new disease targets and develop precision-targeted NADs to address CVDs. Advancements in nucleic acid chemistry have considerably improved drug stability and pharmacokinetics, offering more effective long-term treatment options for reducing the incidence and mortality of CVDs.

Genetic, epidemiological, and clinical research has indicated that lipid abnormalities, such as in low-density lipoprotein cholesterol (LDL-C), are critical factors in the occurrence and progression of CVDs.^535,536 Proprotein convertase subtilisin/kexin type 9 (PCSK9) is an essential regulator of LDL-C metabolism. By reducing the number of LDL receptors (LDLR) on the surface of liver cells, PCSK9 leads to the accumulation of LDL-C, increasing the risk of CVDs.^537,538 NADs targeting PCSK9 work by blocking the synthesis of PCSK9 protein and preventing LDLR degradation, thus promoting LDL-C absorption and reducing its levels in the bloodstream. Inclisiran, the world’s first siRNA drug for CVDs, is administered via subcutaneous injection and conjugated with N-acetyl galactosamine to enhance its targeted uptake by liver cells.⁵³⁹ Results from several Phase III clinical trials (ORION-9/10/11, NCT03397121, NCT03399370, and NCT03400800) have shown that subcutaneous injections of inclisiran every 6 months considerably reduce LDL-C levels by 48–52%, demonstrating sustained efficacy and good tolerability. Injection site reactions were more common in the inclisiran group but were generally mild and not persistent (5.0% vs. 0.7%).^540,541 Inclisiran has been approved for treating various cardiovascular conditions, including atherosclerotic cardiovascular disease (ASCVD), primary hypercholesterolemia, mixed dyslipidemia, and heterozygous familial hypercholesterolemia (HeFH).

Lipoprotein(a) [Lp(a)] is a cholesterol-rich LDL-like particle formed by the covalent bonding of apolipoprotein(a) [Apo(a)] and Apo B100 via disulfide bonds. Elevated Lp(a) levels are considered independent risk factors for ASCVD, heart failure, and aortic valve stenosis.^542–544 Although no targeted drug for reducing Lp(a) levels has been approved thus far, several candidates have shown promise in clinical trials.⁵⁴⁵ For example, pelacarsen, an Lp(a) ASO drug developed by Ionis in collaboration with Novartis AG, effectively lowers Lp(a) levels by selectively cleaving Apo(a) gene (LPA) mRNA. In a Phase II trial, pelacarsen reduced Lp(a) levels by 35–58% depending on the dose, with no significant differences in safety indicators compared with the placebo.⁵⁴² The Phase III trial Lp(a)-HORIZON is underway, involving 8,323 patients with established ASCVD, and is expected to be completed in 2025 (NCT04023552). In addition to ASO drugs, siRNA therapies offer unique solutions for CVDs. Olpasiran and lepodisiran are GalNAc conjugates containing siRNAs that target the LPA gene, promoting RISC-mediated degradation of Apo(a) mRNA and preventing Lp(a) particle assembly in hepatocytes.^546,547 Olpasiran demonstrated a dose-dependent reduction in Lp(a) levels, with a single dose of 9 mg or higher reducing Lp(a) concentrations by over 90% for several months (NCT03626662).⁵⁴⁸ Phase II trials of olpasiran [OCEAN(a)-DOSE] (NCT04270760) showed that a higher dose (75 or 225 mg) could reduce the Lp(a) level of patients by more than 95% compared with the placebo. Regarding safety, the overall incidence of AEs was similar between the dose and placebo groups. Among them, the most common AE of olpasiran was injection site reaction (primarily pain).⁵⁴⁷ The Phase III trial of olpasiran aims to compare its effects with the placebo on the risk of coronary heart disease death, myocardial infarction, and urgent coronary revascularization in ASCVD patients with elevated Lp(a), with results expected by 2026 (NCT05581303). Similarly, lepodisiran has shown well-tolerated and dose-dependent reductions in serum Lp(a) concentrations, with a Phase III trial currently recruiting participants and expected to be completed by 2029 (NCT06292013 and NCT04914546).⁵⁴⁶

Beyond lipid-lowering therapies, there are positive developments in other areas of NADs application. Zilebesiran (ALN-AGT01), which targets angiotensinogen (AGT), is being developed for hypertension treatment. It suppresses the generation of angiotensin I and II, inhibiting the renin-angiotensin-aldosterone system to reduce blood pressure.⁵⁴⁹ Phase I studies showed that a single subcutaneous dose of zilebesiran of 200 mg or more considerably decreased serum AGT levels and 24 h ambulatory blood pressure for up to 24 weeks, with only mild injection site reactions observed.⁵⁵⁰ In Phase II trials (KARDIA-1), zilebesiran reduced systolic blood pressure in all treated patients, with effects lasting 6 months, opening new possibilities for NADs use in treating hypertension and CVDs.⁵⁵¹

Infectious diseases

The COVID-19 pandemic, caused by the severe acute respiratory SARS-CoV-2, has considerably advanced the development of mRNA vaccines. These candidate vaccines are LNP-encapsulated mRNA-based formulations that encode the full-length SARS-CoV-2 spike protein and its receptor-binding domain. They induce an immune response again st the spike protein on the virus’s surface, blocking its entry into cells and providing an antiviral effect.^552,553 mRNA-1273 was one of the first COVID-19 vaccines to enter clinical trials, taking only 42 days from the publication of the virus’s genetic sequence to the production of the first batch of samples. The final analysis from its Phase III clinical trials showed a 94.1% efficacy in preventing COVID-19 illness and a 100% efficacy in preventing severe COVID-19.⁵⁵⁴ Similarly, BioNTech’s mRNA vaccine BNT162b2 demonstrated a 95% efficacy against COVID-19, with protection rates of over 94% for individuals aged 65 and older.⁵⁵⁵ These vaccines were designed, manufactured, evaluated, and brought to market in an extremely short period compared with traditional vaccine development timelines, highlighting the unique advantages of mRNA vaccines, including high safety, simple design, broad target range, and ease of scaling up production.

Chronic hepatitis B, caused by hepatitis B virus (HBV) infection, is a significant global public health problem. Current clinical treatment guidelines aim for a functional cure for CHB, which involves sustained negativity for hepatitis B surface antigen (HBsAg), undetectable HBV DNA, normal liver biochemical indicators, and improvement of liver tissue lesions after treatment cessation.⁵⁵⁶ Despite the efficacy of existing drugs, achieving a functional cure remains challenging. Bepirovirsen, developed by GlaxoSmithKline (GSK), is a 2’-MOE-modified ASO that targets all HBV RNAs, promoting RNase H-mediated RNA degradation, reducing HBV replication, and inhibiting HBsAg production.⁵⁵⁷ Additionally, bepirovirsen has immunostimulatory activity through Toll-like receptor 8, potentially helping the immune system permanently clear the virus from the bloodstream. Data from a Phase IIb clinical trial indicated that after 24 weeks of treatment, bepirovirsen can reduce levels of HBsAg and HBV DNA.⁵⁵⁸ Furthermore, another Phase II trial (B-Sure) evaluated the durability of bepirovirsen’s antiviral effect (NCT04954859). Recruitment for Phase III clinical trials to assess the efficacy and safety of bepirovirsen is also underway, with results expected in early 2026 (NCT05630807 and NCT05630820).⁵⁵⁹ Several candidate siRNA drugs, including GSK-5637608 (JNJ-73763989, GSK),⁵⁶⁰ VIR-2218 (ALN-HBV02, Alnylam & Brii Biosciences Ltd. & Vir Biotechnology),⁵⁶¹ and imdusiran (ARB-270729, Arbutus Biopharma),⁵⁶² have entered Phase II clinical trials, demonstrating a strong ability to reduce HBsAg levels. In practical clinical applications, these siRNA therapies often form combination treatments with nucleoside analogs, TLR agonists, or peginterferon, considerably enhancing the possibility of a functional cure for HBV. For example, VIR-2218, a GalNAc-siRNA conjugate that targets HBV, uses enhanced stability chemistry plus technology to stabilize the siRNA in vivo and reduce off-target effects.⁵⁶¹ Currently, combination therapies involving VIR-2218 are in multiple Phase II trials for HBV patients, and results are eagerly awaited (NCT04856085 and NCT04412863).

Future perspectives of NADs development

Although many NADs have shown promising preclinical results, the number of NADs approved by the FDA for clinical use remains limited. The complexity of nucleic acids and their types, sizes, and mechanisms of action have made the development of NADs systematically very complex. For successful NADs design, critical factors must be considered regarding sequence design, modification, and delivery, as well as the clinical translation, medical indication, and scale production feasibility.

Sequence design of NADs

Typically, NADs are designed and screened based on the pathogenic genes of specific diseases. By leveraging existing sequence design and screening techniques, AI-assisted designs have accelerated the design and screening process, making it more accurate, and have been adapted to the needs of personalized therapy.^215,563 Furthermore, patenting key modification sites in the selected target sequences is crucial to maintaining a competitive advantage.

Structure modification and delivery system

Considering the inherent physicochemical properties of NADs, the critical issue for their application is whether they can reach the target site and exert the expected therapeutic effects. Much evidence supports that it is unrealistic to address all drug delivery issues through a single modification method or a “universal” delivery vector. NADs should exhibit good stability to withstand nucleases and immune system clearance to achieve sound therapeutic effects. In addition to being delivered explicitly to the required tissues, they must reach the targeted cells, efficiently released into the cytoplasm, and exhibit in vivo biocompatibility. Initially, choosing appropriate strategies for the design of NADs is necessary based on the disease type and the required functional nucleic acid. For example, from the approved NADs, ASO drugs often do not require carriers for efficient delivery and only require appropriate modifications, such as PS and 2’ position modifications. Small NADs, such as siRNA, and large ones, such as mRNA, in addition to the necessary modifications, still require the aid of delivery vectors. siRNA typically uses nucleic acid conjugations for delivery, while mRNAs often rely on LNPs for compression and encapsulation. Furthermore, understanding the structure-activity relationship between nucleic acids and the delivery vector and different interactions (hydrophobic, electrostatic, and covalent interactions) and their effect on drug stability is crucial for developing stable and effective NADs. NADs interact with various environments during delivery, and individual chemical modifications or delivery carriers may not suffice to overcome physiological barriers. Combining nucleic acid chemical modifications with drug delivery systems is promising for better therapeutic outcomes. Moreover, researchers are developing various combined carrier-use modes to facilitate effective drug loading, precise targeting, and release. The emergence of intelligent, responsive nanocarriers may help NADs interact with complex environmental changes, responding to specific in vivo changes, such as pH, redox conditions, or external stimuli (ultrasound, light, magnetic fields, and electric fields). These carriers can improve transfection efficiency in target cells and reduce toxic side effects on normal tissues and cells.^564,565 However, hybridizing multiple carriers and modifying functional ligand molecules increase the technical difficulties and manufacturing costs.

Clinical application

Safety

A serious issue in developing these therapeutics is in vivo safety. Despite the detailed understanding of the physicochemical properties of NADs, the potential immunogenicity and toxicity of carriers and related modifications may pose additional challenges to safe and effective NADs delivery.^500,566 For example, early application of relatively mature PPMO technology showed that high doses of CPP-PMO conjugates caused kidney injury in rats.⁵⁶⁷ Cationic polymers, such as PLL/PEI,^305,568 have induced cell apoptosis and inflammation in vivo.

Pharmacokinetic properties and adverse reactions

Advanced technologies have generated NADs complexes with different properties.⁵⁶⁹ Their chemical structure, dosage form, and administration route are the main determinants of their absorption, distribution, metabolism, and excretion, thereby affecting their efficacy and safety.^570,571 For example, based on the unique mechanism of action of ASOs, the main safety issues are severe hepatotoxicity and nephrotoxicity caused by drug accumulation.⁶⁴ Therefore, evaluating NADs delivery systems’ drug metabolism and pharmacokinetics is essential for clinical development. Various bioanalytical methods, such as chromatographic techniques for determining drug plasma concentrations and metabolism^569,572 and imaging technology for estimating drug distribution,⁵⁷³ are used, although each has limitations. Low sensitivity and poor discrimination affect the accuracy of determination results,⁵⁷⁴ highlighting the importance of developing novel analytical methods. Although previous studies have used nano-fluorescent probes to detect the cellular delivery efficiency of oligonucleotide molecules,^403,457,575 more pharmacokinetic data on biodistribution, immune compatibility, and toxicity are needed for clinical translation and practical therapeutic application of these systems.^266,576

Pathological indications

Most diseases exhibit variability in their phenotypes, making it difficult to achieve a curative effect by targeting a single gene. Issues, such as a small patient population and unique targets, often result in a lack of attention from pharmaceutical companies. Therefore, analyzing the pathogenesis of diseases at the genetic level and developing a personalized treatment plan may offer the possibility of a permanent cure for rare and currently incurable genetic diseases. Compared with antibody drugs and small molecule drugs, NADs are specifically favorable for personalized therapies owing to their strong specificity for diverse targets. They have been explored in the development of new therapies for various diseases. Thus far, genetic diseases have been the most approved indication category for NADs. Additionally, breakthroughs in gene therapy and gene-cell combined therapy are promising for the large-scale clinical application of NADs. For instance, Casgevy®, based on the CRISPR/Cas9 gene-editing system, was the first successful attempt at disease treatment, promising a “one-time treatment for a lifelong effect.” However, the complexity of clinical applications and the difficulty of target selection in gene editing have led to long treatment cycles and high costs. Moreover, the application range of NADs is gradually expanding from rare diseases to common diseases such as chronic diseases, infectious diseases, and ophthalmic diseases. The development of curative drugs for indications with large patient populations, such as potential cures for hepatitis B, may address broader clinical needs.

Scale production

Like other drugs, the production of NADs involves multiple technical stages, including raw material collection, synthesis, separation, purification, transportation, and storage, requiring large-scale production capabilities and stringent quality control.⁵⁷⁷ Additionally, the specific requirements for sequence size, purity, and modification and delivery methods of NADs, based on different research needs, increase the demands on the production model.²¹⁴ Therefore, to achieve large-scale clinical application, NADs must ensure easy production, quality control, and transportation. Moreover, as regulatory policies become more refined and the industry chain further develops, companies should seek collaboration opportunities while increasing research and development investments. Collaborations between internationally renowned enterprises in terms of sharing advanced technology and management experience can enhance research and development capabilities and production capacity.

In summary, the development of NADs has undergone a long process, with research on the mechanisms of action of different types of drugs and breakthroughs in chemical modification and delivery technologies. These have transformed NADs from conceptual to practical clinical treatment tools. However, some clinical promotion and application shortcomings still require further development. The organic combination of specific therapeutic nucleic acid molecules, targeted modifications, and functionalized delivery carriers may become vital to achieving personalized nucleic acid therapy and addressing unmet clinical needs.

Author contributions

X.S. developed the concept and wrote the paper, and C.L., S.S., and J.H. discussed, edited, and contributed to the writing. H.X. supervised and revised the paper. All authors have read and approved the article.

Competing interests

The authors declare no competing interests.