Lipid nanoparticle-mediated targeted mRNA delivery and its application in cancer therapy

Abstract

mRNA technology has gained significant attention due to its successful application in COVID-19 vaccines, becoming a vital research area in disease treatment. Achieving therapeutic efficacy requires mRNA to efficiently enter target cells and express functional proteins, highlighting the urgent need for effective delivery systems. Lipid nanoparticles (LNPs) have emerged as a crucial enabler for the clinical translation of mRNA therapies, thanks to their remarkable delivery capabilities. Several mRNA-based therapies have now been approved or are in clinical trials, underscoring the vast potential of mRNA technology. This review provides a comprehensive overview of the application of mRNA–LNP in cancer therapy. It systematically summarizes strategies for optimizing LNP composition, introduces innovative synthesis methods and AI-driven formula optimization, and explores targeted delivery strategies. Additionally, it delves into the various applications of mRNA in cancer treatment, including mRNA tumor vaccines, adoptive cell transfer therapies, restoration of tumor suppressors, immunomodulatory factors, combination therapies, and other emerging treatments. By addressing current challenges and future directions, this review aims to offer valuable insights for further research in this field.

---

1. Introduction

Messenger RNA (mRNA) therapeutics have revolutionized modern medicine, enabling precise, transient protein expression without the risk of genomic integration. Advances in nucleoside modifications, sequence engineering, and delivery technologies have significantly enhanced mRNA stability, translational efficiency, and immunogenicity, accelerating its development for diverse applications.¹ The clinical success of mRNA vaccines during the COVID-19 pandemic not only validated the feasibility of this technology but also spurred rapid innovation in mRNA design and delivery.² Beyond vaccinology, these advancements have expanded the scope of mRNA-based therapies, including cancer immunotherapy,³ protein replacement therapy,⁴ and genome editing.⁵

Among the various delivery systems developed for mRNA, lipid nanoparticles (LNPs) have emerged as the most effective and widely used platform, enabling efficient encapsulation, protection, and cytosolic release of mRNA.⁶ A typical LNP formulation consists of ionizable lipids, helper lipids, cholesterol, and polyethylene glycol (PEG)-lipids, each playing a distinct role in ensuring mRNA stability, endosomal escape, and controlled biodistribution.^7–9 Recent breakthroughs, including combinatorial chemistry-driven screening¹⁰ and machine learning-guided optimization,¹¹ have significantly improved LNP potency, delivery efficiency, and safety. These advancements have cemented LNPs as the leading non-viral carriers for mRNA therapeutics.

However, despite these refinements, conventional LNPs exhibit a strong hepatic tropism, primarily due to interactions with apolipoproteins in circulation.¹² This inherent liver accumulation limits their applicability for extrahepatic delivery, posing a major challenge for expanding mRNA therapeutics beyond hepatic indications.^13,14 To address this, various strategies have been developed to engineer LNPs with enhanced tissue specificity, including physicochemical modifications^15,16 to optimize nanoparticle properties, selective organ targeting (SORT) technology for passive distribution control,^17,18 and ligand-based surface functionalization for active targeting.^19,20 These approaches have enabled efficient mRNA delivery to extrahepatic organs such as the spleen,²¹ lungs,²² and bone marrow,²³ opening new opportunities for applications in cancer therapy and other diseases.^24–26

This review is structured into three main sections. The first section provides a comprehensive overview of LNP composition, the role of each component in mRNA delivery, and recent advancements in chemical design. The second section explores extrahepatic targeting strategies and discusses how different modifications influence biodistribution. The final section focuses on the application of LNP-mediated mRNA delivery in cancer therapy, assessing current progress, challenges, and future directions in tumor-targeted mRNA therapeutics. By integrating insights from chemical design, organotropic modifications, and oncology applications, this review aims to provide a holistic perspective on the development of next-generation LNPs for precision mRNA delivery.

2. Lipid nanoparticle composition and design strategy

LNP has played a pivotal role in the field of mRNA drug delivery, emerging as the most advanced and efficient non-viral delivery platform to date. Early lipid-based carriers primarily relied on cationic lipids, which formed complexes with nucleic acids through electrostatic interactions. While effective in nucleic acid loading, these carriers exhibited high toxicity and poor stability in vivo, limiting their further application. The advent of ionizable lipids has overcome these limitations. Ionizable lipids remain electrically neutral at physiological pH, thereby reducing systemic toxicity. However, in the acidic endosomal environment, they acquire a positive charge, facilitating endosomal escape of mRNA.

In addition to ionizable lipids, LNPs are composed of helper phospholipids, cholesterol, and PEG-lipids, each playing a critical role in mRNA delivery (Fig. 1). Helper lipids enhance the bilayer stability of LNPs, prevent nucleic acid leakage, and promote membrane fusion during cellular uptake. Cholesterol modulates membrane fluidity and permeability while contributing to structural stability through tight lipid packing. PEG-lipids form a hydrophilic barrier on the LNP surface, prolonging circulation time and reducing non-specific protein adsorption. The synergistic interactions among these components determine the physicochemical properties, biological stability, and in vivo delivery efficiency of LNPs.

Fig. 1 Representative structures of LNP components. LNP usually contains four components, namely ionizable cationic lipid, phospholipid, cholesterol, and PEGylated lipid. Here listed ionizable lipids include ALC-0315, SM-102, D-Lin-MC3-DMA, and L319; cholesterol and derivatives include β-sitosterol, stigmastanol, funcostrol, campasterol and stigmasterol; phospholipids include 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), 1-palmitoyl-2-oleoylphosphatidylethanolamine (POPE) and 1,2-dihexanoyl-sn-glycero-3-phosphocholine (SOPE); PEG-lipids include 1,2-dimyristoyl-rac-glycero-3-methoxypolyethylene glycol-2000 (DMG-PEG₂₀₀₀), ALC-0159, 1,2-dioleoyl-sn-glycero-3-phosphoethanolaMine-N-[methoxy(polyethylene glycol)-350] (ammonium salt)(C₁₄-PEG₂₀₀₀) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-amino(polyethylene glycol)-2000 (DSPE-PEG₂₀₀₀). LNP was created in BioRender. Sui, y. (2025) https://BioRender.com/plzve61.

In recent years, advances in combinatorial chemistry, natural product-derived lipid design, and artificial intelligence-assisted optimization have expanded the scope of LNP chemical structure development. High-throughput screening, machine learning, and deep learning approaches have enabled the rapid identification of novel lipid materials with superior delivery performance, shifting LNP design from an empirical-driven to a data-driven paradigm. These advancements have accelerated the development of next-generation mRNA delivery systems. This chapter will provide a comprehensive overview of the composition of LNPs and the functional roles of each component, with a particular focus on cutting-edge strategies in LNP design.

2.1 Ionizable lipids

Ionizable lipids are the key component of LNP for mRNA delivery, with their pH-dependent properties playing a crucial role in the delivery process. Under acidic conditions, such as during LNP formulation, ionizable lipids become protonated and interact electrostatically with negatively charged mRNA to form stable complexes, effectively encapsulating and protecting mRNA from enzymatic degradation. Following dialysis under neutral conditions, these lipids undergo deprotonation and adopt a neutral state, thereby reducing non-specific protein adsorption and minimizing systemic toxicity after administration. Upon cellular uptake, the acidic endosomal environment induces re-protonation of the lipid molecules, enhancing their interaction with the endosomal membrane, disrupting membrane integrity, and facilitating mRNA release into the cytoplasm.

The pK_a of ionizable lipids is considered a critical parameter in balancing in vivo stability and endosomal escape efficiency, with an optimal range typically between 6.0 and 7.0. This range ensures stability in circulation while enabling effective endosomal escape.²⁷ Recent studies have shown that the predicted pK_a values of ionizable lipids correlate well with the actual pK_a values of LNPs, providing valuable insights for the rational design of novel ionizable lipids. Additionally, research suggests that negatively charged phospholipids in the endosomal membrane interact with ionizable lipids, inducing the formation of a hexagonal phase, which further disrupts the endosomal membrane and promotes mRNA release.²⁸ The molecular design of ionizable lipids primarily focuses on optimizing three key structural components: the ionizable head group, the hydrophobic tail, and the linker region.

2.1.1 Ionizable lipid head design. The head group plays a critical role in the functionality of ionizable lipids, as its chemical structure determines the lipid's charge transition under different pH conditions. Based on their chemical composition, ionizable lipid head groups primarily include amines (primary, secondary, and tertiary amines), guanidinium groups, heterocyclic compounds, and their derivatives. Among these, tertiary amines are the most commonly used head groups. Representative molecules include DLin-MC3-DMA, which has been clinically approved for siRNA delivery, as well as ALC-0315 and SM-102, which are widely utilized in mRNA vaccine formulations (Fig. 1). Other ionizable lipids reported in the literature with excellent delivery performance, such as 306O_i10, and C12-200, also feature tertiary amine-based head groups.

In recent years, piperazine-derived ionizable lipids have garnered significant attention. For instance, Ni et al. synthesized a series of piperazine-based ionizable lipids (Pi-lipids) and systematically evaluated 65 chemically distinct LNP formulations for in vivo mRNA translation across 14 different cell types. By analyzing the relationship between lipid structure and cellular targeting, they identified key lipid characteristics enabling effective in vivo delivery without the need for ligand modifications, facilitating nucleic acid delivery to extrahepatic tissues.²⁷ Another study developed the PIP-BP ionizable lipid based on a rigid piperazine-amide scaffold, demonstrating superior bone-targeting capability, high biocompatibility, and low toxicity. This design provides new insights into the development of next-generation bone-targeted delivery systems.²⁹ Furthermore, the stereochemistry of piperazine-based ionizable lipids has been found to significantly influence mRNA delivery efficiency. One study has shown that C12-200-S LNPs achieve 2.8-fold and 6.1-fold higher in vivo mRNA delivery efficiency compared to their racemic mixture and enantiomer C12-200-R, respectively, highlighting the crucial role of stereochemical configuration in lipid-mediated mRNA delivery.³⁰

Certain amino head groups not only facilitate mRNA delivery but also exhibit unique physiological effects. One study identified A2-Iso5-2DC18 and A12-Iso5-2DC18 as optimal lipid structures, with A2 mOVA LNP inducing a stronger antigen-specific CTL response. The only structural difference between the two lies in their head groups, with A2 featuring a heterocyclic amine and A12 a linear amine. Further analysis revealed that heterocyclic lipids significantly enhance IFN-γ secretion compared to linear tertiary amines, with their adjuvant effect mediated by the STING signaling pathway.³¹ Another study constructed an imidazole-based lipid library and employed a multi-step screening strategy, identifying imidazole-headed ionizable lipids with outstanding efficiency in T-cell transfection. Among them, 92-O17S was selected as a Cre mRNA delivery vector, achieving an 8.2% gene recombination rate in T cells following intravenous administration in mice, demonstrating its potential for precise gene editing applications.³² These findings underscore that the role of ionizable lipid head groups in mRNA delivery extends beyond regulating transfection efficiency. They also influence LNP targeting and elicit distinct physiological effects, providing valuable insights for the rational design and therapeutic applications of mRNA delivery systems.

2.1.2 Ionizable lipid tail design. Ionizable lipid hydrophobic tails typically consist of one to four alkyl chains ranging from 8 to 20 carbon atoms, playing a crucial role in nanoparticle formation, transfection efficiency, biodegradability, and cytotoxicity. The lipid tail can adopt various structural forms, including linear or branched, saturated or unsaturated, and biodegradable or non-biodegradable. Modifications to the tail structure can directly impact LNP delivery performance and in vivo behavior.

Tail length is a critical factor influencing mRNA delivery efficiency. Mrksich synthesized a series of ionizable lipids with varying tail lengths based on C12-200 and evaluated their mRNA delivery efficiency in LNP formulations. The results showed that C10-200 LNPs, which have alkyl tails two carbons shorter than those of C12-200 LNPs, achieved a tenfold increase in Fluc mRNA delivery efficiency to the liver. Additionally, C12-200 LNPs and C13-200 LNPs exhibited comparable efficiency in delivering the smaller EPO mRNA. However, when delivering the larger Cas9 mRNA, C9-200 LNPs exhibited over threefold higher gene editing efficiency compared to C12-200 LNPs. These findings suggest that shorter-tailed lipids are more suitable for the delivery of large mRNA molecules, whereas longer-tailed lipids may be more effective for small mRNA fragments. Therefore, the selection of lipid tail length should be tailored to the molecular size of the mRNA cargo to optimize delivery performance.³³ Tail length also affects the organ targeting of LNP. A study by Yusuke Sato's team at Hokkaido University in Japan found that LNPs containing shorter tail lipids preferentially delivered mRNA to the spleen. Further cellular and pharmacokinetic analyses showed that LNPs composed of short-tailed ionizable lipids adsorbed less apolipoprotein E (ApoE) on their surface, which may reduce the uptake of LNPs by the liver, thus altering the in vivo distribution and targeting properties of LNPs.³⁴ Additionally, tail length affects mRNA encapsulation efficiency and the apparent pK_a of LNPs. In the iterative design of ionizable lipids, researchers predicted that alkyl tail length may influence steric hindrance and the lipid–water partition coefficient (logD), thereby modulating lipid–mRNA interactions. Experimental results showed that E10 tails (10-carbon chains) increased mRNA encapsulation efficiency to over 90% while maintaining an optimal particle size distribution. Furthermore, tail length was found to be inversely correlated with the apparent pK_a of LNPs, highlighting the importance of tail structure in optimizing nanoparticle performance.³⁵

The branching of lipid tails also plays a critical role in LNP stability and intracellular delivery efficiency. A study that constructed a library of 32 α-branched lipid tails found that highly symmetrical branched tails enhanced LNP stability and improved intracellular delivery.³⁶ Another study systematically optimized the head group, tail length, and branching pattern, identifying several biodegradable branched lipids with in vivo mRNA delivery efficiency comparable to or even exceeding that of MC3-LNP, the siRNA delivery system based on DLin-MC3-DMA, demonstrating their potential as next-generation delivery carriers (Fig. 2a).³⁷ Furthermore, research has shown that lipids with dual-tail structures can adopt a crown-like conformation, promoting membrane destabilization and enhancing intracellular nucleic acid release.⁸ In addition, the degree of tail unsaturation directly affects membrane stability, thereby modulating nucleic acid delivery efficiency.³⁸ In summary, tail length, branching, and saturation play crucial roles in determining LNP delivery efficiency, organ specificity, and stability, providing essential insights for the rational design of mRNA delivery systems.

Fig. 2 Design and innovative synthesis strategies of ionizable lipids. (a) A schematic showing how a three-dimensional combinatorial library, enabled by a multi-component reaction, allows simultaneous variation of three components (cationic head, linker, and alkyl chain) for higher efficiency in lipid synthesis and screening.³¹ Copyright 2019, Springer Nature. (b) The A³-coupling involves the formation of an alkynyl–Cu intermediate, which reacts with an iminium ion generated in situ from an aldehyde and an amine to yield a propargylamine product, while regenerating the Cu⁺ catalyst. This multicomponent reaction proceeds under mild conditions and allows for efficient construction of structurally diverse lipid amines.⁴⁶ Copyright 2024, Springer Nature. (c) Tandem multicomponent reaction (T-MCR) for the synthesis of biodegradable ionizable lipids (AID lipids). Primary amines, Traut's reagent, and alkyl acrylates undergo amine–thiol–acrylate coupling under mild conditions, enabling rapid lipid synthesis with improved biodegradability and mRNA delivery performance.⁴⁸ Copyright 2024, Springer Nature.

2.1.3 Ionizable lipid linker design. The linker group plays a crucial role in ionizable lipids, typically connecting the head and tail groups while significantly impacting lipid stability, biodegradability, cytotoxicity, and transfection efficiency. Based on their chemical properties, linkers can be categorized into non-biodegradable types (such as ether and carbamate bonds) and biodegradable types (such as ester, amide, and thiol bonds). Biodegradable linkers offer greater potential for application, as they facilitate lipid clearance in vivo, reduce toxicity, and support repeated dosing. The FDA-approved ionizable lipids DLin-MC3-DMA, ALC-0315, and SM-102 all incorporate degradable ester bonds, demonstrating the effectiveness of this design strategy.

In recent years, to further enhance mRNA delivery efficiency and gene editing outcomes, researchers have designed various biodegradable linkers. One study designed a disulfide-bridged ester linker and employed a modular strategy to synthesize a library of 96 biodegradable ionizable lipids (LDILs). Screening for mRNA delivery efficiency both in vitro and in vivo identified LDILs containing a single 4A3 head group, four disulfide-bridged ester linkers, and three 10-carbon chain tails as highly effective for mRNA delivery. Mechanistic studies revealed that this unique glutathione (GSH)-responsive cone-shaped structure endowed LNPs with exceptional endosomal escape and rapid mRNA release capabilities. This rationally designed LNP not only enhanced in vivo mRNA delivery efficiency and gene editing capacity but also enabled precise imaging of cancer metastases.³⁹ In another study, based on the upregulated ROS levels in tumor cells, a series of degradable ionizable lipids were designed and synthesized for selective mRNA delivery to tumor cells, using thioketone-containing acrylate (TK-12) as the linker group. The resulting LNPs efficiently encapsulated mRNA and, under the influence of ROS in tumor cells, triggered lipid oxidation and degradation, promoting mRNA release and selective expression.⁴⁰ These studies highlight the promising potential of degradable linker groups.

Studies have also demonstrated that the choice of linker groups plays a critical role in the organ-specific targeting of LNPs. In one study, a library of LNPs with various degradable linkers, including esters, carbonates, amides, and carbamates, was constructed. The results showed that lipids with amide (lipid 34), carbamate (lipid 35), and reverse amide (lipid 36) linkers exhibited higher stability, with lipid 35 and 36 displaying significant lung-targeting properties and in particular, lipid 35 showed higher transfection efficiency in lung cells.⁴¹ Additionally, even slight modifications in the chemical structure of the linker could significantly affect the delivery targeting. For example, replacing the ester bond (306-O12B LNP) with an amide bond (306-N16B LNP) shifted the mRNA delivery specificity from the liver to the lungs. Proteomic analysis revealed that specific plasma proteins on the surface of liver and lung-targeted LNPs might play a key role in organ targeting. Further alterations in the head structure of the N-series LNPs enabled targeting of different subcellular lung populations.⁴² In another study, researchers synthesized a series of ketal-ester lipids (KELs), which included a biodegradable ketone segment in the linker and an ester tail. The lipid (4S)-KEL12 demonstrated excellent immune organ targeting, with delivery efficiency in the spleen comparable to SM-102, while expression in the liver was significantly reduced (Fig. 2b).⁴³ A separate study revealed that when lipids with branched ester tails were combined with different linkers, mRNA delivery efficiency varied significantly. In particular, the delivery efficiency was higher when paired with ethanolamine linkers, showing preference for specific organs and cells.⁴⁴ These findings collectively underscore the crucial impact of linker chemistry on the organ-targeting and delivery efficacy of LNPs.

In summary, the rational design of ionizable lipids plays a pivotal role in mRNA delivery via LNPs. Modifications to the head group can regulate ionization behavior, enhance mRNA binding affinity, and influence cellular uptake, while alterations to the hydrophobic tail impact LNP stability, transfection efficiency, and organ-specific targeting. Additionally, the choice of linker affects lipid biodegradability, metabolic clearance, and intracellular release. Recent advancements in the chemical design of ionizable lipids, including the incorporation of novel head structures, tailored hydrophobic tails, and biodegradable linkers, have not only improved delivery efficiency but also expanded the potential for precise tissue targeting. These innovations provide valuable insights for the continued development of next-generation ionizable lipids, laying the foundation for safer and more effective mRNA therapeutics.

2.1.4 Innovative synthetic strategies for ionizable lipids. Traditional ionizable lipid synthesis strategies primarily rely on classic organic reactions such as Michael addition and ring-opening reactions. Michael addition facilitates the conjugation of enolate compounds with nucleophilic reagents to generate lipid molecules, while ring-opening reactions enable rapid linkage of lipid structural units. However, these methods exhibit limitations in structural tunability and high-throughput optimization, making it challenging to meet the growing demand for lipid diversity and functional optimization in novel delivery systems. In recent years, multicomponent reactions (MCRs) have emerged as a powerful and modular approach for ionizable lipid synthesis, offering significant advantages in efficiency and structural diversity. By enabling the rapid conjugation of multiple functional groups within a single reaction system, MCRs streamline lipid construction and provide new strategies for enhancing the chemical diversity and delivery performance of LNPs.

One pioneering study employed an isocyanide-mediated three-component reaction (3-CR) for the high-throughput synthesis of ionizable lipids. This approach enabled the one-step coupling of primary or secondary amines, ketones, and isocyanides or their derivatives, generating a structurally diverse lipid library of over 1000 variants. High-throughput screening identified lipids with cyclic amine head groups that activated the STING signaling pathway, highlighting their potential for LNP-mediated cancer immunotherapy (Fig. 2c).³¹ Another efficient synthetic strategy, the Ugi four-component reaction (Ugi-4CR), allows for the one-pot coupling of aldehydes, isocyanides, amines, and carboxylic acids, thereby eliminating the need for lengthy purification steps associated with traditional methods.⁴⁵ Leveraging this approach, researchers constructed an ionizable lipid library with diverse structural variants and identified highly effective mRNA delivery systems targeting the liver and spleen. In the context of directed chemical evolution, the amine–aldehyde–alkyne (A³) coupling reaction has been utilized to facilitate the rapid iterative optimization of lipid structures. This method not only operates under mild reaction conditions but also provides precise control over molecular architecture, enabling fine-tuned lipid designs for enhanced mRNA delivery.⁴⁶ Similarly, the van Leusen three-component reaction (van Leusen-3CR) has been applied to efficiently synthesize imidazole-based ionizable lipids (IMILs) under mild conditions, circumventing the stringent anhydrous and oxygen-free requirements of conventional synthesis while expanding the structural versatility of ionizable lipids.⁴⁷ For the development of biodegradable lipids, a study introduced a tandem multicomponent reaction (T-MCR) that enables the rapid synthesis of biodegradable ionizable lipids (AID lipids). This strategy leverages the amine, Traut's reagent (2-iminiothiolane hydrochloride), and alkyl acrylate through amine–thiol–acrylate coupling, achieving efficient synthesis at room temperature within one hour. Notably, this approach not only simplifies lipid preparation but also enhances LNP degradation properties, improving both mRNA delivery efficiency and biocompatibility.⁴⁸

2.2 Cholesterol

Cholesterol is a key component of LNPs and plays an important role in maintaining nanoparticle stability, improving mRNA encapsulation efficiency and regulating membrane properties. As a naturally occurring sterol lipid, cholesterol is capable of embedding in lipid bilayers, filling gaps between lipid molecules, increasing membrane packing density, reducing nucleic acid leakage, and enhancing the structural integrity of nanoparticles.⁴⁴ Its incorporation into LNPs was originally inspired by liposomal drug delivery systems, where cholesterol was found to improve membrane stability, reduce plasma protein adsorption, and extend nanoparticle circulation time.⁴⁹ In LNP formulations, cholesterol typically constitutes 20–50% of the total lipid content, and deviations from this ratio can significantly affect nanoparticle morphology, mRNA encapsulation efficiency, and biodistribution.⁴⁴ Additionally, cholesterol modulates membrane fluidity and bilayer thickness, contributing to the formation of an ordered lipid phase while adapting to different lipid environments. When combined with low-transition temperature (T_m) phospholipids, cholesterol reduces membrane fluidity, whereas with high-T_m lipids, it increases fluidity, balancing membrane stability and dynamic adaptability.⁴⁹ Beyond structural support, cholesterol plays a critical role in LNP uptake and endosomal escape by interacting with cellular membranes, promoting membrane fusion, and enhancing mRNA release into the cytoplasm.²⁵

Beyond its role in structural stability, modifications to the chemical structure of cholesterol can significantly impact the in vivo behavior of LNPs. The James E. Dahlman group observed that LNPs formulated with 20-hydroxycholesterol (20-OH cholesterol) could effectively deliver mRNA to hepatocytes, with this compound existing in two conformations: 20α and 20β. To investigate the effects of these stereoisomers, they prepared LNPs using either 20α-hydroxycholesterol (20α) or a mixture of 20α- and 20β-hydroxycholesterol (20mix) and injected them into mice. The results showed that 20α-LNPs exhibited superior delivery efficiency (Fig. 3a). Furthermore, they discovered that this stereochemistry-dependent delivery effect was not limited to a specific LNP type. The incorporation of 20α enhanced the mRNA delivery efficiency of LNPs formulated with both DLin-MC3-DMA and Moderna lipid H, demonstrating the broad applicability of this phenomenon.⁵⁰

Fig. 3 Cholesterol design. (a) The chemical structures of 20-hydroxycholesterol (20α and 20mix) three days after injection, flow cytometry quantified tdTom⁺ Kupffer cells and liver endothelial cells for MC3-based LNPs and Moderna lipid H-based LNPs.⁵⁰ Copyright 2023, Springer Nature. (b) The chemical structures of β-sitosterol, stigmastanol, campesterol and fucosterol. Representative fluorescent images showing FLuc mRNA (red), cytosolic mRNA (green), and LNP accumulation (blue) in HeLa cells after delivery with LNP or eLNP, visualized using smFISH.⁵¹ Copyright 2020, Springer Nature. (c) The chemical structures of 7α-HC. Screen of LNP library for luciferase mRNA delive in Jurkats to identify top formulations.⁵⁵ Copyright 2022, Elsevier.

In recent years, the application of cholesterol derivatives in LNP formulations has garnered increasing attention, leading to significant advancements. One study demonstrated that incorporating C-24 alkyl phytosterols (eLNP) into LNPs significantly enhances gene transfection efficiency. The key factors maintaining high transfection rates were identified as the alkyl tail length, sterol ring flexibility, and hydroxyl group polarity. Compared to conventional LNPs, eLNPs exhibited greater intracellular diffusion, suggesting a more efficient endosomal escape pathway (Fig. 3b).⁵¹ Additionally, cholesterol derivatives influence LNP morphology, promoting multilayer structures, lipid phase separation, and micelle formation. Specifically, methylation or ethylation at the C-24 position increased micelle formation by over 50%, while the introduction of double bonds enhanced lipid phase separation by 90%, both contributing to improved gene transfection efficiency.⁵²

In other studies on cholesterol alternatives, β-sitosterol incorporation significantly boosted LNP transfection activity without altering particle size or ζ potential.⁵³ Similarly, replacing cholesterol with stigmastanol in LNPs doubled gene editing efficiency in HepG2-GFP cells. However, LNPs formulated with phytosterols exhibited larger particle sizes and slightly reduced encapsulation efficiency compared to cholesterol-based LNPs, highlighting the necessity of balancing delivery efficiency and nanoparticle stability when optimizing sterol composition.⁵⁴ Additionally, replacing 25% or 50% of cholesterol with 7α-hydroxycholesterol (7α-HC) in LNP formulations effectively increased mRNA delivery efficiency to T cells, while maintaining cytotoxicity levels comparable to unmodified formulations (Fig. 3c).⁵⁵ Further advancements have explored ionizable cholesterol derivatives in LNP formulations. A novel cholesterol-based ionizable lipid, 3β[L-histidyl-carbamoyl] cholesterol (Hchol), was developed as a cholesterol replacement in LNPs. Hchol-LNPs, incorporated into MC3- and SM-102-based formulations, demonstrated high biocompatibility under physiological conditions and significantly enhanced mRNA endosomal escape, leading to improving in vivo gene expression.⁵⁶

Overall, cholesterol's role in LNPs extends far beyond structural stability, as its stereochemistry, alkyl tail modifications, and hydroxyl functionalization profoundly impact mRNA delivery efficiency, transfection performance, and organ tropism. Rationally designing cholesterol derivatives provides an effective strategy to enhance LNP-mediated mRNA delivery, offering an optimized platform for mRNA therapeutics and gene editing applications.

2.3 Phospholipids

Phospholipids are key auxiliary lipids in LNPs, playing a crucial role in maintaining nanoparticle stability, improving mRNA encapsulation efficiency, and facilitating intracellular delivery. Among them, DSPC and DOPE are the most used phospholipids in LNP formulations. DSPC, a saturated phospholipid with a high phase transition temperature,⁵⁷ significantly enhances the structural stability of LNPs and has been widely applied in siRNA delivery and COVID-19 vaccine formulations, gaining recognition for its excellent circulation stability. In contrast, DOPE, with its unique cone-shaped structure and smaller headgroup, promotes the formation of the hexagonal (H_II) phase,^9,58 thereby significantly enhancing endosomal escape and accelerating cytoplasmic mRNA release. Additionally, DOPE enhances LNP fusion with the plasma membrane, further improving transfection efficiency.⁵⁹ It was demonstrated that the delivery efficiency of mRNA could be significantly improved after replacing DSPC with DOPE.^60,61

A recent study systematically analyzed the impact of different phospholipid head groups on mRNA delivery, comparing PE-based phospholipids (DOPE, SOPE, POPE, DPPE) and PC-based phospholipids (POPC, DSPC), while also evaluating variations in chain length, unsaturation, and methylation. Additionally, organelle-specific lipids such as BMP and CL were investigated for their potential roles in LNP formulations (Fig. 4a). Compared with PC-based phospholipids, PE-based phospholipids exhibited higher mRNA delivery efficiency in HEK-293T and HeLa cells and enhanced mRNA expression in vivo. In contrast, BMP-LNPs preferentially accumulated in the spleen despite their low fluorescence signal, highlighting their potential for organ-specific targeting.⁶²

Fig. 4 Phospholipids design. (a) Phospholipids containing PE head groups enhance endosomal escape due to their fusogenic properties and enhance LNP-mediated mRNA delivery.⁶² Copyright 2022, Royal Society of chemistry. (b) As ESM helper lipid content increases, LNP morphology transitions from a solid core to a bilayer structure enclosing an aqueous compartment with a small solid core. LNPs with 40 mol% ESM show extended circulation and enhanced gene expression in spleen, bone marrow, and liver compared to 10 mol% DSPC LNPs.⁶³ Copyright 2023, Elsevier.

Furthermore, high phospholipid content significantly influences LNP structure and function. As the phospholipid ratio increases, LNP morphology transitions from a solid-core structure to a bilayer-encased aqueous compartment surrounding a smaller solid core. Cryo-TEM imaging of the MC3-ESM LNP system revealed that 40 mol% ESM-LNPs exhibited the most uniform structural distribution, forming elongated internal solid cores, emphasizing the critical role of phospholipid composition in LNP architecture (Fig. 4b). In terms of delivery performance, 40 mol% DSPC- or ESM-LNPs demonstrated significantly higher mRNA transfection efficiency than 10 mol% DSPC- or ESM-LNPs, with 40 mol% ESM-LNPs achieving the best in vitro performance. Additionally, pharmacokinetic analysis indicated that 40 mol% ESM-LNPs had an extended half-life of 3.7 hours, compared to just 15 minutes for 10 mol% DSPC-LNPs, suggesting that higher ESM content prolongs systemic circulation (Fig. 4b).⁶³ Notably, earlier studies found that 40 mol% DSPC-LNPs, even with a reduced ionizable lipid ratio, exhibited similar structural changes and an extended circulation half-life, regardless of whether siRNA was encapsulated.⁶⁴ These findings underscore the crucial role of phospholipid type and composition in LNP stability, mRNA encapsulation, intracellular delivery, and in vivo circulation, providing valuable insights for LNP formulation optimization.

2.4 PEG-lipids

Polyethylene glycol-modified lipids (PEG-lipids) are a critical component of LNPs, playing a vital role in enhancing nanoparticle stability, prolonging circulation time, and modulating biodistribution. However, PEGylation introduces the so-called “PEG dilemma”, wherein PEG prevents nanoparticle aggregation and reduces serum protein adsorption but simultaneously impairs LNP interactions with target cells, thereby diminishing cellular uptake and endosomal escape efficiency.^57,65 To optimize PEG performance, one study introduced fluorinated modifications to the PEG head group and found that at a molar ratio of 1.5%, mRNA delivery efficiency was significantly improved. Specifically, mRNA expression levels in B16F10 tumor cells and primary dendritic cells increased by fivefold and twofold, respectively, while in vivo studies further confirmed that this strategy enhanced mRNA expression by at least threefold. Mechanistic investigations revealed that fluorinated PEG not only improved LNP cellular uptake but also enhanced endosomal escape, leading to a substantial increase in overall delivery efficiency.⁶⁶ Another study explored the impact of PEG-lipid hydrophobicity (including single- and double-alkyl chains of varying lengths) and terminal functional groups (such as methoxy, carboxyl, and amino groups) on LNP performance. Five types of PEG-lipids were evaluated, including SA-PEG, DMG-PEG, DSPE-PEG, DSPE-PEG2k-NH₂, and DSPE-PEG2k-COOH (Fig. 5a). The results showed that DSPE-based LNPs exhibited poor mRNA delivery across all liver cell types, whereas SA-based LNPs achieved the highest transfection efficiency in liver-resident Kupffer cells. DMG-based LNPs demonstrated superior mRNA expression in hepatocytes and endothelial cells. In terms of circulation half-life, DSPE-LNPs outperformed both DMG- and SA-based LNPs. Furthermore, replacing the methoxy (–OCH₃) terminal group of PEG with carboxyl (–COOH) or amino (–NH₂) groups accelerated PEG shedding, thereby enhancing LNP interactions with hepatic parenchymal cells.¹²

Fig. 5 PEG-lipids design. (a) Schematic of a PEG-lipid toolbox showing the chemical structures of five PEGylated lipids used to optimize mRNA LNPs.¹² Copyright 2025, American Chemical Society. (b) Chemical structure of DMG-PEG and DMG-PEOZ. LNPs with PEOZ form stable nanoparticles at different molar ratios. Mice receiving single or four weekly injections of PEG–LNPs or PEOZ–LNPs showed reduced liver and spleen luminescence in pre-treated compared to first-injection mice.⁶⁷ Copyright 2024, Wiley VCH GmbH. (c) Structure of DMPE-poly (CBMA) and DLPE-poly (CMBA) ZIP–lipids with varying degrees of polymerization. ZIP–lipids are designed for LNPs used in nebulized mRNA delivery.⁶⁹ Copyright 2024, American Chemical Society.

In recent years, researchers have actively explored alternatives to PEGylated lipids to overcome the limitations associated with PEG modification while maintaining the high delivery efficiency of LNPs. One study demonstrated that poly(2-ethyl-2-oxazoline) (PEOZ) lipids could be used to construct LNPs, enabling effective mRNA delivery to the liver and multiple cell types in mice, with notably enhanced delivery efficiency to the spleen. Moreover, PEOZ–LNPs exhibited stable mRNA delivery performance even after repeated dosing, suggesting their potential for sustained administration (Fig. 5b).⁶⁷ Another study developed a PEG-free LNP formulation by replacing PEG-lipids with polyserine (pSar) lipids, creating a novel pSar–LNP system. The results demonstrated that pSar–LNPs exhibited comparable mRNA delivery efficiency and immunogenicity to conventional PEG–LNPs. This suggests that the pSar-based approach can maintain LNP functionality while mitigating PEG-associated adverse effects, providing a promising design strategy for next-generation mRNA delivery systems.⁶⁸ Additionally, a novel LNP formulation utilizing a zwitterionic polymer (ZIP)–lipid conjugate as a PEG lipid substitute was developed to enhance aerosol stability (Fig. 5c). ZIP–LNPs maintained structural integrity during nebulization and achieved efficient mRNA transfection in a mouse model. Notably, in mucus-rich environments, ZIP–LNPs exhibited delivery efficiency comparable to PEG–LNPs, highlighting their potential for mRNA delivery in mucus-obstructive diseases.⁶⁹ These research advances not only broaden the strategies of LNP surface modification, but also provide new directions for optimizing the stability, biodistribution, and repeated drug delivery effects of mRNA delivery, and promote the development of LNP delivery systems toward safer and more efficient.

2.5 AI-driven optimization of LNP

The development of traditional LNPs has relied on empirical screening and iterative optimization, however, the complexity of lipid–mRNA interactions makes this process time-consuming and inefficient. Recent advances in artificial intelligence (AI) have revolutionized LNP design by enabling machine learning (ML) and deep learning (DL) to analyze large datasets, identify structure–activity relationships, predict lipid compositions, and accelerate high-throughput synthesis (HTS). These approaches significantly enhance LNP optimization and improve delivery performance. While AI has been used for mRNA sequence optimization, this review focuses on its applications in LNP development, highlighting its role in lipid design, delivery system engineering, and performance prediction.

One of the key applications of AI in LNP optimization is integrating high-throughput chemical synthesis with ML to screen candidate lipids and accurately predict delivery performance. A study developed a HTS platform based on a 4CR, constructing an initial library of 584 ionizable lipids. The researchers employed an XGBoost model to predict lipid delivery efficiency, followed by computational screening of 40 000 virtual lipid structures. Sixteen top candidates were selected for experimental validation, ultimately identifying lipid 119-23 (Fig. 6a–d), which exhibited significantly enhanced mRNA delivery to muscle and immune cells compared to commercial LNPs, demonstrating the power of AI-driven lipid screening.¹⁰ Recently, another study trained a graph neural network (GNN) and implemented the LiON (lipid optimization using neural networks) strategy to predict nucleic acid delivery efficiency based on LNP chemical composition. Initially, LiON was used to select candidate ionizable lipids targeting the liver, successfully validating its predictive accuracy. The study further utilized LiON to screen over 1.6 million computationally generated lipids, identifying two top-performing lipids, FO-32 and FO-35. Subsequent experiments confirmed that FO-32 and FO-35 efficiently delivered mRNA to the alveoli and conducting airways of ferrets following intratracheal administration, demonstrating the potential of AI-driven LNP design for inhaled mRNA therapeutics.⁷⁰ Another study introduced the AI-guided ionizable lipid engineering (AGILE) platform, which utilizes deep neural networks (DNNs) to predict the delivery efficiency of candidate lipids and rapidly screen for highly effective mRNA-delivering LNPs (Fig. 6e).⁷¹

Fig. 6 AI-driven rational design of ionizable lipids for mRNA delivery. (a) High-throughput screening and evaluation of ionizable lipid libraries were performed to generate data for training machine learning algorithms.¹⁰ Copyright 2024, Springer Nature. (b) A 384-member ionizable lipid library with four-component chemical structures was synthesized using high-throughput methods.¹⁰ Copyright 2024, Springer Nature. (c) Machine learning algorithms were trained and used for predictions based on the screening data.¹⁰ Copyright 2024, Springer Nature. (d) The optimal ionizable lipid structure for mRNA delivery was identified through high-throughput screening and confirmed by machine learning algorithms.¹⁰ Copyright 2024, Springer Nature. (e) AGILE model enables accurate prediction of ionizable lipid efficacy in mRNA delivery LNP formulations, streamlining viable candidate identification from extensive lipid libraries. ML analysis for predicting the mRNA expression efficiency and assessing feature importance of the ionizable lipid substructure and LNP composition.⁷¹ Copyright 2024, Springer Nature.

In addition, AI models can also be used to predict key parameters affecting mRNA delivery. A study reported AI-based prediction of two key attributes of LNP-apparent pK_a and mRNA delivery efficiency. The team conducted two rounds of iterative screening of nearly 20 million ionizable lipids through an AI-driven generation and screening process, resulting in three and six novel lipid molecules, respectively. Delivery validation in a mouse model showed that one of the lipids from the first round of screening, which contained a benzene ring structure, had a delivery efficiency comparable to that of the control lipid, DLin-MC3-DMA. The six lipids from the second round of screening all met or exceeded the delivery performance of DLin-MC3-DMA, and the delivery efficiency of one of the lipids was close to that of the control lipid SM-102, which demonstrated the great potential of AI in accelerating the design and optimization of LNPs.⁷² Another study applied a random forest regression model to analyze the features of 213 LNPs, successfully predicting mRNA expression efficiency and refining LNP-based mRNA vaccine delivery strategies.⁷³ Additionally, AI has been utilized in mRNA–LNP vaccine quality control.⁷⁴ As ML and DL models continue to evolve and integrate with large-scale experimental datasets, AI-driven LNP development is expected to expand further. In the future, AI may enable personalized LNP design, tailoring delivery strategies to specific tissues or cell types. The continuous advancement of AI technology will further accelerate mRNA vaccine development, gene editing applications, and other mRNA-based therapeutics, facilitating their clinical translation and widespread application.

3. Targeted delivery of LNP

Although LNP vector-based mRNA therapeutics have made significant progress in the fields of gene editing, cancer treatment, and vaccine development. However, achieving organ-specific targeted delivery of LNP remains a significant challenge. After intravenous injection, most LNPs are preferentially enriched in the liver, and about 30–90% of the nanoparticles accumulate in the liver, significantly limiting their application in the treatment of extrahepatic diseases. Therefore, how to break through the limitation of liver enrichment and achieve precise delivery to specific organs or cell types has become a key direction for LNP design optimization. In Chapter 2, the functions of LNP components and the effects of structural modifications on LNP biodistribution at the chemical level were discussed in detail. In this chapter, we will focus on the active strategies that have been used in recent years to promote the targeted delivery of LNP at the formulation level, including the SORT strategy, surface functionalization modifications, altered routes of administration, and high-throughput screening, which will provide new ideas for the application of LNP in extrahepatic gene therapy application to provide new ideas.

3.1 SORT strategies

The SORT strategy represents a significant advancement in overcoming the inherent hepatic accumulation of LNPs and enabling precise organ-specific delivery. Unlike conventional four-component LNPs, SORT LNPs incorporate a fifth lipid component (SORT molecule) to precisely modulate nanoparticle physicochemical properties, thereby reshaping their biodistribution profile. Cheng et al. first introduced this strategy and demonstrated that tuning the charge and composition of SORT lipids effectively directs LNPs to specific non-hepatic tissues.¹⁸ For example, the addition of 50% of a permanent cationic lipid (e.g., DOTAP) to the original 5A2-SC8 LNP significantly enhances LNP accumulation in the lungs, while the addition of 30% of an anionic lipid (e.g., 1,2-dioleoyl-sn-glycero-3-phosphate, 18PA) enables spleen-specific targeting. Importantly, this strategy is remarkably tunable and can be widely applied to different LNP systems (Fig. 7a). Replacement with DLin-MC3-DMA or C12-200 maintains similar organ-specific delivery, resulting in the successful formulation of DLin-MC3-DMA SORT LNP and C12-200 SORT LNP. The research team further compared different LNP formulation techniques, including pipette mixing, vortex mixing, and microfluidic mixing.⁷⁵ All three methods achieved high mRNA encapsulation efficiency, with microfluidic mixing yielding slightly superior encapsulation compared to the other two methods. Additionally, LNPs prepared via pipette and vortex mixing exhibited larger particle sizes, likely due to slower mixing rates and mass transfer effects. However, formulation methods did not impact the applicability of the SORT strategy, and particle size showed no significant influence on organ specificity. The targeting mechanism of SORT LNPs is governed by three key factors. First, PEG lipid desorption is a prerequisite for efficient mRNA delivery to target tissues. Second, the incorporation of SORT lipids substantially alters the serum protein adsorption profile on LNP surfaces, influencing their biodistribution. Finally, the adsorbed proteins on the LNP surface interact with specific receptors expressed in target organs, facilitating the efficient transport and functional expression of mRNA.⁷⁶ These findings elucidate the key regulatory factors governing SORT LNP-mediated targeted delivery and provide a theoretical and technical foundation for further optimizing LNP design to enhance tissue specificity in mRNA therapeutics.

Fig. 7 SORT methodology that added a fifth component to achieve tissue-specific mRNA delivery. (a) Adding a SORT molecule to traditional four-component LNPs. This modification enables targeted delivery of mRNA to specific tissues, including the liver, lungs, and spleen in mice, following intravenous administration.¹⁸ Copyright 2022, Springer Nature. (b) A schematic showing the design of covalent lipids and the construction of bone marrow-targeted LNPs by mixing components for gene editing via IV injection.⁸⁰ Copyright 2024, Springer Nature. (c) A schematic comparing previous and our lung-targeting strategies and showcasing our liver-targeting strategy. By modifying LNP components and lipid structures, authentic targeting of the lung and liver was achieved. Representative images display mRNA accumulation and translation for different LNP formulations.⁸² Copyright 2024, Springer Nature.

The application of SORT strategies has significantly expanded the delivery capabilities of LNPs, overcoming their inherent hepatic accumulation and enabling precise targeting of specific organs and cell types. In recent years, this strategy has demonstrated great potential in fields such as gene editing and cell therapy. For pulmonary delivery in the treatment of cystic fibrosis (CF), Wei et al. developed an optimized second-generation lung-targeted SORT LNP (DOTAP40 LNP), which significantly enhanced the efficacy of gene-editing therapy and improved tolerability for repeated dosing. This system efficiently delivered mRNA to lung basal cells, a key target for CF treatment. Subsequent experiments demonstrated that lung-targeted SORT LNP successfully corrected CFTR (CF transmembrane conductance regulator) mutations in both homozygous G542X mice and patient-derived bronchial epithelial cells carrying the F508del mutation.⁷⁷ However, the study has certain limitations. While CFTR gene correction temporarily restores CFTR protein function, long-term therapeutic efficacy may require the correction of airway stem/progenitor cells. Therefore, direct in vivo gene editing of stem cell populations is considered the optimal strategy for achieving a durable cure for CF.

Encouragingly, Daniel J. Siegwart group further optimized lung-targeted SORT LNPs to achieve efficient genome editing in stem cells, resulting in sustained therapeutic responses. Their study first validated the applicability of lung SORT LNP in gene editing, demonstrating its ability to effectively deliver Cre mRNA, Cas9 mRNA–sgRNA, and ABE mRNA–sgRNA to major lung cell types. In a tdTomato mouse model, the system achieved >70% modification of lung stem cells and maintained tdTomato expression in >80% of lung epithelial cells for up to 660 days. To explore the therapeutic potential of lung SORT LNP for CF, the researchers utilized the LNP-ABE system to correct the CFTR^R553X nonsense mutation, achieving a 60% editing efficiency in patient-derived bronchial epithelial cells, which led to a 5.5-fold increase in mature CFTR protein expression. When combined with standard treatment for patients carrying the F508del mutation, CFTR expression was further enhanced to 7.8-fold. In a humanized CFTR^R553X mutant mouse model, intravenous administration of 1.5 mg kg⁻¹ LNP-ABE over 10 days resulted in a 50% gene editing rate in lung stem cells and a 12.2% overall editing rate in lung tissue. The corrected mutation restored CFTR function without inducing liver or kidney toxicity or tissue damage, demonstrating excellent safety and tolerability.⁷⁸ These findings highlight the significant potential of lung-targeted SORT LNPs in gene editing therapy, providing strong evidence for their application in the long-term treatment of CF and other pulmonary genetic disorders.

Beyond pulmonary applications, SORT strategies have also been leveraged for immune cell delivery, enabling in situ CAR-T cell generation and overcoming the limitations of traditional cell therapies. Researchers developed spleen-targeted SORT LNPs that could directly transfect T cells in vivo, inducing CAR expression without the need for peripheral blood cell isolation and reinfusion. This approach significantly increased tumor-infiltrating lymphocytes in a B-cell lymphoma model, enhancing antitumor efficacy and reducing hepatic tumor metastasis, presenting a novel strategy for simplifying CAR-T therapy.⁷⁹

SORT strategies have also advanced hematopoietic stem cell targeting. Screening a molecular library of 41 compounds, researchers identified the covalent lipid SA-NHS as a key determinant of bone marrow tropism and further optimized its structure to develop LNPs capable of efficiently delivering mRNA to HSCs, leukemia stem cells, and various bone marrow cell populations (Fig. 7b). Proteomic analysis revealed that the bone marrow-targeting capability of these LNPs depended on the enrichment of ApoE on their surface, providing a promising strategy for in vivo gene editing in hematological disorders.⁸⁰ Additionally, SORT LNPs have been demonstrated to efficiently deliver siRNA to the lung (58%) and spleen (45%) and, for the first time, to the kidney (15%).⁸¹

The conventional SORT strategy relies on the incorporation of a fifth lipid component to regulate the organ-specific delivery of LNPs, but this approach increases formulation complexity. Recently, researchers have developed a series of simplified SORT LNPs that achieve efficient organ-targeted delivery while reducing formulation complexity. One study demonstrated that cholesterol and phospholipids are not essential for LNP-mediated delivery. Based on this finding, researchers designed a three-component LNP (3-Comp LNP) composed solely of an ionizable lipid, a permanent cationic lipid, and a PEG-lipid. This formulation exhibited preferential accumulation in pulmonary endothelial and epithelial cells, enabling efficient mRNA delivery and translation (Fig. 7c). Notably, 3-Comp LNPs outperformed traditional four- or five-component SORT LNPs containing cholesterol in pulmonary delivery while maintaining excellent stability furthermore, this lung-targeting formulation strategy is generalizable to other existing cationic lipids and LNPs.⁸² Another study further introduced an organ- and cell-specific mRNA–LNP delivery platform using a simplified three-component LNP composed of an ionizable lipid, a PEG-lipid, and a targeting lipid. This system achieved efficient mRNA delivery to the lungs, liver, and spleen. Additionally, by incorporating specific miRNA target sites into the mRNA sequence, researchers enhanced cell-type-specific protein translation within target tissues.⁸³ This strategy not only optimizes the targeted delivery efficiency of LNPs but also offers a novel approach for precise gene expression regulation.

3.2 Surface modification

The organ-targeting capability of LNPs is not solely determined by their intrinsic lipid composition but is also significantly influenced by surface modifications. By conjugating specific antibodies, peptides, or small-molecule ligands, LNPs can selectively interact with target cell receptors, thereby modulating their biodistribution and enhancing mRNA delivery specificity. For HSC targeting, researchers have developed CD117 antibody-modified LNPs (CD117/LNP-mRNA), which significantly improve their accumulation in HSCs. In vitro studies demonstrated that CD117-modified LNPs exhibited superior HSC-targeting efficiency compared to CD45 and other antibodies, without compromising stem cell potency. In vivo, intravenous administration of CD117/LNP-Cre enabled efficient gene editing in long-term HSCs in a dose-dependent manner while preserving their stemness and function. Additionally, this strategy has been successfully applied to a sickle cell disease model, enabling in vivo gene editing using human CD117/LNPs.⁸⁴ Similarly, another study developed CD117-modified LNPs that, with a single intravenous injection, achieved siRNA or mRNA delivery to HSCs in mice, resulting in >90% gene editing in HSCs and hematopoietic progenitor cells (HSPCs). The edited cells retained their self-renewal capacity and function, ultimately generating a large population of gene-modified immune cells.⁸⁵

Antibody-modified LNPs have also been used for T cell targeting and in vivo CAR-T cell generation. A study designed CD5 antibody-modified LNPs (CD5/LNP) for mRNA-based CAR-T cell induction (Fig. 8a). In vitro, this system enabled 81% of T cells to express the reporter gene, whereas unmodified LNPs achieved only 7% transfection efficiency. After intravenous injection of CD5/LNPs containing luciferase mRNA (CD5/LNP-Luc) into mice, a significant increase in luciferase activity was observed in splenic T cells compared to the control group. Further experiments demonstrated that this approach successfully induced T cells in vivo to produce chimeric antigen receptors (CARs) targeting fibroblast activation protein (FAP) (FAPCAR). In a cardiac injury model, the generated CAR-T cells were able to eliminate cardiac fibroblasts and improve cardiac function (Fig. 8a).⁸⁶ Another work, also through antibody-coupled modification of LNP, achieved targeted delivery of CAR mRNA to T cells, showing antibody-dependent and dose-dependent CAR expression and cytokine release, while achieving up to 90% B-cell clearance.⁸⁷ Additionally, CD4 antibody-modified LNPs (CD4/LNPs) enabled efficient mRNA delivery to CD4⁺ T cells, with intravenous administration leading to significant accumulation in the spleen and a 30-fold increase in mRNA expression in T cells compared to non-targeted LNPs.⁸⁸ Another study developed anti-CD3-modified LNPs (aCD3-LNPs) for T-cell-specific mRNA delivery, further expanding the application of LNPs in immune cell engineering.⁸⁹

Fig. 8 Representative surface modification strategies for LNP targeting. (a) CD5-targeted LNP produce functional, mRNA-based FAPCAR T cells in vitro. Histologic analysis using picrosirius red staining shows reduced extracellular matrix burden in injured mice treated with CD5/LNP-FAPCAR compared to saline or IgG/LNP-FAPCAR controls.⁸⁶ Copyright 2022, The American Association for the Advancement of Science. (b) A schematic showing antibody functionalization with SATA for thiol group introduction, followed by conjugation to maleimide-modified LNPs to create lung-targeted LNPs. Bioluminescence imaging visualizes tissue distribution of firefly luciferase mRNA expression 4.5 hours after IV administration of unconjugated, anti-PECAM-1, and control IgG-conjugated LNPs.⁹⁰ Copyright 2018, Elsevier. (c) A schematic showing the engineering of aEGFR–LNPs, screening for optimal antibody density, evaluating their in vitro and in vivo mRNA delivery efficiency, and characterizing their uptake in the placental microenvironment of non-pregnant and pregnant mice.⁹² Copyright 2024, Elsevier.

In pulmonary disease treatment, antibody-modified LNPs have also demonstrated exceptional delivery efficiency. One study employed PECAM-1 antibody-modified LNPs, which exhibited a 200-fold increase in lung mRNA delivery efficiency and a 25-fold enhancement in protein expression compared to unmodified LNPs, while significantly reducing hepatic uptake (Fig. 8b).⁹⁰ Another study utilized LNPs conjugated with an antibody against the plasma membrane vesicle-associated protein (PV1) for lung-targeted delivery. This strategy markedly improved lung-specific mRNA delivery, with protein expression levels increasing 40-fold relative to unmodified LNPs.⁹¹

Notably, antibody-conjugated LNPs have also shown great potential for placenta-targeted delivery. Researchers developed an epidermal growth factor receptor (EGFR) antibody-conjugated LNP (aEGFR–LNP) for delivery of mRNA to the placenta (Fig. 8c). Optimized aEGFR–LNP mediated an approximately two-fold increase in mRNA delivery in the mouse placenta and an approximately two-fold increase in LNP uptake in EGFR-expressing trophoblasts, as compared to non-targeted LNP.⁹² These findings highlight that antibody modification not only enhances pulmonary accumulation but also minimizes off-target distribution, providing a more precise approach for mRNA therapy in lung diseases. Moving forward, optimizing antibody-lipid conjugation strategies to improve LNP stability and delivery efficiency will be critical for advancing the clinical translation of this technology.

Peptide functionalization has emerged as a promising strategy for targeted mRNA delivery via LNPs. Researchers have developed a peptide-modified LNP platform to enable systemic brain delivery of mRNA by targeting receptors overexpressed in brain endothelial cells and neurons. Four specific peptide sequences were selected: RVG29 (targeting nicotinic acetylcholine receptors), T7 (targeting transferrin receptors), angiopep-2 (AP2, targeting low-density lipoprotein receptor-related protein 1, LRP-1), and mApoE (targeting low-density lipoprotein receptor, LDLR). Compared to non-targeted LNPs, all peptide-modified LNPs significantly enhanced mRNA delivery to the brain. Notably, mApoE-modified LNPs increased the brain-to-liver distribution ratio by approximately eightfold, demonstrating their ability to effectively cross the blood–brain barrier and enhance the therapeutic potential of mRNA for neurological disorders.⁹³ In tumor-targeted delivery, Yang's team employed click chemistry to conjugate PD-L1-binding peptides with PEGylated lipids, which were directly incorporated into LNPs during formulation, generating tumor-targeting peptide-modified LNPs (Pep LNPs). The optimized Pep LNPs markedly improved mRNA delivery to PD-L1-overexpressing tumor cells, further validating the role of peptide functionalization in enhancing LNP tumor-targeting capabilities and offering a novel strategy for mRNA-based cancer therapy.⁹⁴

Modifying the LNP surface with small-molecule ligands, such as folic acid and hyaluronic acid, has also been shown to enhance tumor-targeted delivery.^95–97 However, research on the application of small-molecule ligands in LNP-mediated mRNA delivery remains relatively limited. Further optimization of ligand selection and conjugation strategies is required to improve stability, specificity, and in vivo delivery efficiency. In summary, LNP surface modification strategies have greatly expanded its targeting capabilities. Functionalization with antibodies, peptides, or small-molecule ligands not only reduces liver accumulation, but also precisely delivers mRNAs to hematopoietic stem cells, immune cells, lung tissues, and the placenta, thereby increasing the therapeutic potential of mRNA-based therapies. Looking forward, the development of efficient and biocompatible surface modification strategies and the optimization of conjugation methods hold great promise for advancing the use of LNPs in personalized gene therapy, cancer immunotherapy, and fetal medicine, and ultimately accelerating the clinical translation of mRNA therapeutics.

3.3 Altering administration route

The route of administration plays a critical role in determining the biodistribution, mRNA expression efficiency, and therapeutic efficacy of LNPs. Studies have shown that different administration routes can significantly alter the organ-specific targeting of LNPs.

Despite recent studies demonstrating that modulating the surface charge of nanoparticles can enhance lung delivery of systemically administered LNPs, nebulization remains a more promising strategy for achieving lung-specific mRNA delivery while minimizing off-target effects in other organs. However, the strong shear forces generated during nebulization can lead to LNP aggregation, structural disruption, and even mRNA leakage, ultimately reducing transfection efficiency. In addition, efficient transfection of bronchial epithelial cells requires overcoming the physicochemical barriers posed by the mucus layer, making nebulized LNP-mediated mRNA delivery particularly challenging.

To address the instability of LNPs during nebulization, one study optimized LNP composition and rationally designed buffer systems and excipients to improve nebulization stability. Using an in vitro air–liquid interface (ALI) culture model, two candidate lipids, IR-117-17 and IR-19-Py, were identified, both of which exhibited superior in vivo performance. Notably, nebulized IR-117-17 LNPs achieved a 300-fold increase in pulmonary mRNA delivery efficiency compared to previously reported LNPs.⁹⁸ In another study, researchers developed a charge-assisted stabilization (CAS) strategy, yielding CAS-LNPs, in which negatively charged peptide–lipid conjugates were incorporated to enhance LNP stability under nebulization conditions (Fig. 9a). This approach significantly improved mRNA delivery efficiency upon inhalation. Importantly, the CAS strategy exhibited high versatility, allowing for broad applicability across various LNP formulations. Further investigations revealed that CAS-LNPs primarily targeted pulmonary dendritic cells, underscoring their potential as mRNA vaccine delivery vectors.⁹⁹

Fig. 9 Representative administration routes and their applications. (a) A schematic showing the design of CAS-LNPs, where charged lipids are incorporated into clinical LNP formulations to enhance stability during nebulization by increasing electrostatic repulsions, preventing mRNA leakage.⁹⁹ Copyright 2024, Springer Nature. (b) A schematic showing an inhalable mucus-penetrating mRNA–LNP designed to deliver genetic material to AT2 cells, restoring epithelial stem cell populations and treating idiopathic pulmonary fibrosis.¹⁰¹ Copyright 2024, The American Association for the Advancement of Science. (c) Fluorescence imaging shows the distribution of DOPE, PS, and DOTAP lipids in various tissues, with hematoxylin staining highlighting mLUC expression in the pancreas compared to control, indicating that LNPs transfect primarily pancreatic islets.¹⁰⁴ Copyright 2023, The American Association for the Advancement of Science.

Additionally, to explore the impact of LNP chemical composition on lung delivery, one study introduced an iterative in vivo clustering-based screening method, leading to the development of a novel nebulization-adapted LNP, NLD1. Unlike conventional LNPs, NLD1 is enriched in polyethylene glycol (PEG)-lipid (55%) and cationic helper lipids, enabling sustained pulmonary mRNA expression for 1–7 days post-nebulization, with a 4- to 5-fold higher delivery efficiency than other LNPs. Furthermore, NLD1 was successfully employed to deliver mRNA encoding a broad-spectrum neutralizing antibody, providing effective protection against lethal H1N1 influenza infection in mice.¹⁰⁰ In a more recent study, researchers introduced pulmonary surfactant components into LNP design for the first time. By incorporating dipalmitoyl phosphatidylcholine (DPPC), a phospholipid-rich component of lung surfactant, into conventional lipid formulations, they developed mucus-penetrating surfactant-mimetic LNP (Fig. 9b).¹⁰¹ This system enabled efficient mucus penetration upon inhalation and facilitated co-delivery of dual mRNA cargos, offering a novel strategy to enhance pulmonary mRNA delivery.

In addition to nebulized inhalation, intranasal inhalation is also an effective method for pulmonary mRNA delivery. A recent study proposed an innovative lipid nanoparticle (iLLN) delivery system, featuring a liquid lipid core. This system self-assembles through hydrophobic interactions among lipid components including ALC-0315, DOTMA, β-sitosterol, DOPE, triolein, and DSPE-PEG (Fig. 9c). Notably, triolein, a naturally occurring triglyceride with a low melting point (5.53 °C), forms a liquid lipid core at physiological temperatures. The liquid lipid core design has been shown to enhance nanoparticle deformability, thereby improving their mucosal permeability and facilitating more efficient diffusion to target tissues.¹⁰² Four hours after intranasal administration, the iLLN-2/mRNA complex exhibited strong bioluminescent signals in the nasal cavity, with Fluc mRNA expression predominantly localized to the nasal cavity and lungs. The mRNA was successfully delivered across alveolar epithelial cells, whereas conventional LNPs were mainly confined to the respiratory surface, demonstrating lower permeability. Moreover, compared to conventional ALC-LNPs, the iLLN-2/mRNA complex induced a significantly stronger mucosal IgA and IgG response specific to the SARS-CoV-2 spike protein during the prime-boost intranasal immunization, without causing any noticeable inflammatory reactions.¹⁰³

Among the various administration routes for LNP-mediated mRNA delivery, intraperitoneal injection has received comparatively little attention. However, recent studies have demonstrated that this approach can effectively enhance mRNA accumulation in the pancreas when using LNPs incorporating cationic helper lipids. Compared to intravenous injection, intraperitoneal administration significantly increases pancreatic mRNA delivery but is accompanied by off-target expression in the liver and spleen. Further optimization of the LNP formulation revealed that incorporating 40% of the cationic helper lipid DOTAP markedly reduced off-target expression while ensuring that the majority of the translated protein was localized within insulin-producing β-cells (Fig. 9c). Mechanistic investigations further identified that pancreatic mRNA delivery via intraperitoneal injection is facilitated by horizontal gene transfer through exosomes secreted by peritoneal macrophages.¹⁰⁴ These findings underscore the potential of intraperitoneal injection as an effective strategy for organ-specific mRNA delivery and provide valuable insights for advancing mRNA therapeutics targeting pancreatic diseases.

Beyond these approaches, additional administration routes have been explored to achieve distinct mRNA delivery objectives. Intramuscular injection remains a widely utilized strategy for vaccine administration, as it effectively stimulates antigen-presenting cells (APCs) and elicits a robust immune response.¹⁰⁵ In contrast, subcutaneous injection enhances mRNA accumulation in lymph nodes, thereby promoting immune activation and adaptive immunity.¹⁰⁶ Localized administration, including intraocular,^107–109 intratumoral,¹¹⁰ and intracardiac injection¹¹¹ provides a targeted delivery strategy that maximizes mRNA expression within specific tissues while minimizing systemic off-target effects. Furthermore, novel strategies for central nervous system (CNS) delivery are under investigation, with cerebrospinal fluid injection and intranasal administration emerging as effective techniques for enhancing mRNA distribution in the brain.⁵⁰ Collectively, these alternative administration routes expand the potential applications of LNP-based mRNA therapeutics, offering precise organ targeting and broadening the scope of mRNA-based treatments across diverse physiological and pathological contexts.

3.4 High-throughput screening strategy of targeting LNPs

High-throughput screening enables the parallel testing of a large number of ionizable lipids or LNP formulations, accelerating the identification of optimal ionizable lipids, refining LNP formulations, and facilitating the selection of LNPs with organ-specific targeting capabilities. Consequently, the development of fast and reliable in vivo high-throughput screening platforms is critical for optimizing LNP-mediated mRNA delivery. Among these approaches, DNA barcoding technology has emerged as an innovative high-throughput screening tool, allowing each LNP formulation to be uniquely labeled with a DNA barcode.¹¹² This molecular tagging strategy enables researchers to precisely track LNP biodistribution across different organs and tissues via high-throughput sequencing following in vivo administration. However, previous studies have indicated that LNP accumulation in a given tissue does not necessarily correlate with efficient mRNA expression.¹¹³ In some cases, certain LNPs exhibit higher functional mRNA delivery efficiency in tissues where their accumulation is relatively low.^114,115 To address this limitation, researchers developed the fast identification of nanoparticle delivery (FIND) system, which integrates rationally designed DNA barcodes with the Cre-Lox system to provide simultaneous readouts of LNP biodistribution and functional mRNA delivery efficiency.¹¹⁶ In this system, Cre mRNA and a unique DNA barcode are co-encapsulated within LNPs, with each LNP formulation carrying a distinct barcode. The LNP library is then injected into Lox-Stop-Lox-tdTomato (Ai14) transgenic mice, in which Cre mRNA translation results in Cre protein production. Upon nuclear translocation, Cre protein excises the stop cassette in the tdTomato locus, leading to the activation of tdTomato fluorescence. Following in vivo administration, fluorescence-activated cell sorting (FACS) is used to isolate tdTomato+ cells, and deep sequencing of these cells enables the identification of LNPs capable of achieving efficient mRNA delivery. This approach overcomes the limitation of relying solely on LNP accumulation to assess delivery efficiency, offering a more precise and effective method for LNP screening.

This technology has been applied to multiple LNP optimization studies. For example, the library of Pi-lipids was systematically evaluated for mRNA delivery efficiency using DNA barcoding screening techniques.²⁷ In another study, this method was employed to identify LNPs capable of specifically delivering mRNA to human head and neck squamous cell carcinoma (HNSCC) tumors, while simultaneously reducing off-target accumulation in the liver.¹¹⁷

To further enhance the resolution of LNP delivery analysis at the single-cell level, researchers developed the single-cell nanoparticle targeting sequencing (SENT-seq) platform, which enables the quantitative assessment of LNP-mediated mRNA delivery at single-cell resolution. SENT-seq not only allows for the detection of DNA barcodes within individual cells but also provides insights into target protein expression and transcriptomic changes following LNP administration.¹¹⁸ Using this approach, a recent study screened 105 different LNP formulations for their ability to target bone marrow cells in mice. Among them, LNP emerged as the top-performing candidate. Notably, LNP was capable of efficiently delivering mRNA to mouse bone marrow and human primary HSCs without the need for additional targeting ligands. Further validation in non-human primates (NHPs) demonstrated that LNP could effectively deliver mRNA to CD34⁺ hematopoietic stem cells at low doses (0.25 mg kg⁻¹ and 0.4 mg kg⁻¹), highlighting its potential for gene therapy applications.¹¹⁹

Building upon DNA barcoding technology, Daniel Anderson's research team developed a high-throughput screening system based on peptide barcodes to simultaneously evaluate multiple LNP formulations within a single experimental animal. This approach involves encoding unique peptide barcodes into mRNA, which are then individually encapsulated within distinct LNP formulations. By administering a mixed pool of these LNPs to a single animal, researchers can assess functional mRNA delivery efficiency across different formulations. Following administration, liquid chromatography-tandem mass spectrometry (LC-MS/MS) is employed to quantify the translation levels of each peptide barcode, enabling precise evaluation of mRNA delivery efficiency for each LNP in vivo. Compared to DNA barcoding, this method provides a more accurate assessment of functional delivery while expanding its applicability. Using this screening platform, the research team efficiently evaluated 384 ionizable lipids and over 400 distinct LNP formulations. Remarkably, the entire screening process required only nine mice, significantly reducing animal usage while enhancing research efficiency. Based on these findings, the team further optimized and developed a biodegradable LNP (RM133-3-21), which demonstrated efficient mRNA delivery and functional protein expression in vivo, offering a promising strategy for advancing RNA delivery system optimization.¹¹³

These advancements in high-throughput screening, particularly through DNA and peptide barcoding technologies, have significantly improved the precision and efficiency of identifying LNPs with enhanced organ-specific targeting capabilities. By enabling systematic optimization of LNP formulations and functional mRNA delivery assessment in vivo, these approaches pave the way for the development of next-generation RNA therapeutics with improved targeting efficiency and reduced off-target effects.

4. mRNA-based cancer therapy

Cancer has long been a major global health challenge. It is estimated that by 2070, the number of new cancer cases worldwide will reach 34 million.¹²⁰ The characteristics of malignant tumor cells, such as unlimited proliferation, abnormal differentiation, high invasiveness, and metastasis, make clinical treatment complicated by issues such as interpatient heterogeneity, multidrug resistance, pre-surgical metastasis, post-surgical recurrence, and significant treatment-related side effects.¹²¹ The successful application of mRNA vaccine technology during the COVID-19 pandemic in 2019 has provided valuable insights for cancer treatment.¹²² Building on this foundation, mRNA-based cancer vaccines, adoptive cell therapies, and strategies aimed at restoring tumor-suppressor gene function have become major research focuses. Moreover, the combined use of multiple therapies offers new avenues for cancer treatment.

4.1 mRNA-based tumor vaccine

Immunotherapy is an emerging approach in cancer treatment that enhances the immune system's ability to recognize and eliminate tumors.¹²³ A prominent form of immunotherapy is tumor vaccines, which utilize mRNA encoding tumor antigens to stimulate immune cells to produce tumor-specific proteins. This triggers immune responses aimed at targeting and destroying cancer cells (Fig. 10).¹²⁴ Unlike traditional vaccines, mRNA-based tumor vaccines offer significant advantages, including high flexibility in design and the ability to encode multiple tumor-specific antigens, thereby inducing stronger and broader immune reactions. These vaccines can deliver both antigen fragments and complete antigens, expand antigen diversity and enhance immune recognition. Furthermore, mRNA vaccines have a favorable safety profile with no risk of genomic integration, fast metabolic clearance, and minimal side effects.¹²⁵ Additionally, mRNA tumor vaccines can be tailored to an individual's unique tumor mutation profile, advancing precision medicine. With the addition of mRNA modifications to improve stability and reduce immunogenicity, these vaccines present promising new avenues for cancer immunotherapy and many cases have been tested in clinical trials (Table 1).

Fig. 10 mRNA-based cancer vaccines. Schematic illustration of the mRNA–LNP cancer vaccines. Created in BioRender. Sui, y. (2025) https://BioRender.com/dp46nj1.

Table 1 A summary of clinical trials based on mRNA cancer vaccines

Name	Combination	Condition	Status	NCT number (phase)	Outcome measures
NA and unknown indicate that no information was found on https://ClinicalTrials.gov. DOR, duration of response; OS, overall survival; ORR, objective response rate; PFS, progression-free survival; RFS, relapse-free survival; AE, adverse event; DLT, dose-limiting toxicity; DMFs, distant metastasis-free survival; TEAEs, treatment-emergent adverse events; DCR, disease control rate; MTD, maximum tolerated dose; SAEs, serious adverse events.
Personalized tumor preventive vaccine	Anti-PD1	Pancreatic cancer recurrent	Recruiting	NCT06496373 (I)	RFS and OS
EBV mRNA vaccine	None	EBV-positive advanced malignant tumors	Unknown	NCT05714748 (I)	ORR, PFS and OS
Personalized tumor vaccine	Atezolizumab and mFOLFIRINOX	Pancreatic cancer	Active, not recruiting	NCT04161755 (I)	RFS
Patient-specific neoantigen vaccine	Atezolizumab (anti-PDL1)	Colorectal neoplasms	Active, not recruiting	NCT05141721 (II and III)	DOR and AEs
Personalized cancer vaccine mRNA-4157	Pembrolizumab (anti-PD1)	Melanoma	Recruiting	NCT03897881 (II)	RFS, DMFs, DLTs and AEs
Anti-cancer neoantigen mRNA vaccine	None	Solid tumor	Recruiting	NCT06195384 (I)	DLT
Neoantigen mRNA vaccine	Camrelizumab	Pancreatic cancer	Recruiting	NCT06326736 (I)	TEAEs and DCR
Individualized mRNA neoantigen vaccine mRNA-0523-L001	None	Adrenal cortical carcinoma, medullary thyroid cancer, thymic neuroendocrine carcinoma and pancreatic neuroendocrine tumor	Recruiting	NCT06141369 (NA)	MTD, DLT, ORR, DCR and PFS
COVID-19 vaccine mRNA-1273	Anti-PD1 and anti-PDL1	Solid tumor malignancy, hematologic malignancy, leukemia and lymphoma	Completed	NCT04847050 (II)	AEs
Personalized neoantigen vaccine	Anti-PD1 and anti-PDL1	Gastric cancer, esophageal cancer and liver cancer	Recruiting	NCT05192460 (NA)	AEs, ORR and PFS
BNT113	Pembrolizumab (anti-PDL1)	Unresectable head and neck squamous cell carcinoma, metastatic head and neck cancer and recurrent head and neck cancer	Recruiting	NCT04534205 (II)	OS, ORR, PFS, DCR, DOR and TEAEs
BNT111	Cemiplimab (anti-PD1 and anti-PDL1)	Melanoma stage III, melanoma stage IV and unresectable melanoma	Active, not recruiting	NCT04526899 (II)	ORR, DOR, DCR, TTR, PFS, OS and TEAEs
RO7198457	Pembrolizumab (anti-PD1)	Advanced melanoma	Completed	NCT03815058 (II)	PFS, ORR, OS, DOR and AEs
Personalized mRNA vaccine iNeo-Vac-R01	None	Digestive system neoplasms	Recruiting	NCT06019702 (I)	AEs, DLT, OS, ORR, DCR, DOR and PFS

---

Building on these advances, researchers developed an mRNA vaccine nanoparticle composed of mRNA encoding OVA and the palm acid-modified TLR7/8 agonist R848 (C16-R848). This vaccine NP not only retained the adjuvant activity of C16-R848 but also significantly enhanced mRNA transfection efficiency and antigen presentation. Compared to non-adjuvanted vaccines, the C16-R848-adjuvanted mRNA vaccine notably increased the expansion and tumor infiltration of OVA-specific CD8⁺ T cells, inducing a stronger immune response. The study demonstrated that the C16-R848-adjuvanted mRNA vaccine nanoparticle effectively prevented and suppressed tumor growth in mouse models of OVA-expressing lymphoma and prostate cancer, significantly improving the efficacy of mRNA vaccines in cancer immunotherapy.¹²⁶ Moreover, researchers have developed a lymph node (LN)-targeted mRNA vaccine based on 113-O12B LNP for cancer immunotherapy. Compared to LNPs formulated with ALC-0315, 113-O12B LNP demonstrates higher targeting specificity. Studies have shown that targeting mRNA delivery to LNs enhances CD8⁺ T cell responses to the encoded full-length OVA model antigen and achieves superior protective and therapeutic efficacy in the B16F10 melanoma model.¹⁰⁶ In parallel, recent studies have shown that optimizing PEG linker length in dendritic cells-targeted mRNA nanovaccines can significantly improve mRNA uptake and antigen presentation in vivo. Notably, PEG400-modified nanocarriers achieved superior DC uptake, enhanced mRNA expression, and better tumor suppression in the E.G7-OVA lymphoma and TC-1 cervical tumor mouse model, highlighting the importance of nanocarrier structure design in the efficacy of mRNA tumor vaccines.¹²⁷ Similarly, Jin developed a universal mRNA-based anti-tumor vaccine that leverages the body's immune memory of known pathogens to treat cancer. The vaccine reprograms tumor cells in situ to activate antigen-specific effector T cells, which then target and kill tumor cells. In mouse models of melanoma, breast cancer, and colon cancer, the vaccine showed significant anti-tumor effects and promoted T cell infiltration and activation at the tumor site. RNA-seq and single-cell RNA-seq analyses revealed that the vaccine can reshape the tumor microenvironment (TME), turning it from immunosuppressive to pro-inflammatory, while boosting the cytotoxic functions of memory T cells.¹²⁸

In the context of pancreatic ductal adenocarcinoma (PDAC), a leading cause of cancer-related deaths, Rojas and colleagues developed a personalized neoantigen-targeted vaccine using an mRNA–LNP platform, based on surgically resected PDAC samples. Clinical trials showed that the vaccine induces specific T cell responses, potentially eradicating micro metastases. Additionally, when combined with anti-PD-L1 and mFOLFIRINOX, the vaccine significantly boosted tumor-specific immunity and delayed recurrence.¹²⁹

Unlike traditional LNP delivery systems, Nie and their colleagues have successfully developed a novel mRNA delivery platform (OMV-LL) by using genetic engineering to modify the surface of bacterial outer membrane vesicles (OMVs) with RNA-binding protein L7Ae and the lysosomal escape protein, listeriolysin O. In preclinical models of lung metastatic melanoma and subcutaneous colorectal cancer, OMV-LL-mRNA significantly inhibited the growth and metastasis of tumors. This platform offers a unique “plug-and-display” advantage, enabling efficient and flexible application in the development of personalized mRNA cancer vaccines and providing a new technological pathway for precision medicine.¹³⁰

Further progress was made in 2021 when the FDA granted orphan drug designation to BNT111 for treating stage IIB to IV melanoma. BNT111 is an mRNA tumor vaccine that encodes four melanoma-associated antigens (tyrosinase, NY-ESO-1, MAGE-A3, and TPTE) and is delivered via intravenous injection to activate immune cells to target and kill tumor cells.¹³¹ In a phase 1 trial with 56 patients, the combination of BNT111 with PD-1 monoclonal antibodies showed higher response and disease control rates and good safety. The complementary mechanisms of the vaccine and PD-1 antibodies resulted in a synergistic effect greater than their individual contributions (NCT04526899).

Most recently, LK101 injection from Likang Life Sciences received FDA approval for clinical trials. This is the first Chinese mRNA vaccine for tumor neoantigens. LK101 is an mRNA-dendritic cell vaccine, where mRNA encoding tumor neoantigens is used to transfect dendritic cells in vitro, combining the benefits of both mRNA and dendritic cell vaccines. Early studies show that when combined with ablation therapy, LK101 is safe and significantly reduces recurrence in liver cancer patients (CXSL2200612).

However, despite recent encouraging progress, significant challenges still remain in the development and clinical implementation of mRNA-based cancer vaccines.¹³² A primary concern is the delivery efficiency of mRNA molecules, as these molecules must enter cells effectively to achieve sufficient antigen expression and immune activation.¹³³ Another significant challenge is carefully balancing immune activation to avoid harmful side effects like cytokine release syndrome (CRS) or autoimmunity, while maintaining effective tumor-targeting immunity.¹³⁴ Striking the right balance between immunogenicity and safety is difficult, since overly strong immune responses can cause widespread inflammation and severe side effects, necessitating careful optimization and comprehensive clinical assessment to guarantee safety and efficacy.

4.2 Adoptive cell transfer therapy

In the past decade, adoptive cell transfer therapies have transformed the landscape of cancer treatment by harnessing the power of engineered immune cells to target and eliminate tumor cells. The incorporation of mRNA technology has further enhanced these therapies, making the production and application of engineered immune cells, such as CAR-T, CAR-NK, and CAR-modified macrophages (CAR-M), more efficient and versatile.¹³⁵ This progress offers substantial potential for improving antitumor efficacy and patient outcomes.

4.2.1 CAR-T. The emergence of CAR-T cell therapy marks a new breakthrough in cancer treatment. By genetically engineering T cells to express CARs, CAR-T cells can precisely identify and attack cancer cells.¹³⁶ The CAR structure includes an antibody binding domain, a transmembrane domain, and an intracellular signaling domain, which together induce tumor cell death and activate the immune system to eliminate tumors.

Although CAR-T cell therapy has shown significant clinical efficacy, completed trials have revealed substantial adverse events. The most frequently observed adverse events in CAR-T cell therapies are CRS and immune effector cell-associated neurotoxicity syndrome (ICANS).¹³⁷ CRS involves increased levels of various cytokines, such as IL-6, IL-8, IL-10, IFN-γ, GM-CSF, macrophage inflammatory protein-1β, and monocyte chemoattractant protein-1.¹³⁸ The failure of CAR-T cell therapy is primarily attributed to factors such as the functional characteristics of co-stimulatory domains within the CAR, the initial T cell phenotype, and the intrinsic quality of T cells. Another significant contributor to therapeutic failure and adverse outcomes is the off-target effect, which arises when CAR-T cells mistakenly attack healthy cells expressing the target antigen. Therefore, enhancing the specificity of immune cells is essential to minimize damage to normal tissues.¹³⁹ With the advancement of mRNA technology, researchers have discovered that delivering CAR mRNA to T cells via in vitro transcription (IVT) enables transient expression of CARs.¹⁴⁰ This approach avoids the risk of genomic integration while maintaining efficient transfection and translation. mRNA-based CAR-T therapy has demonstrated comparable short-term efficacy to traditional CAR-T therapy in cancers such as acute lymphoblastic leukemia (ALL), melanoma, and Hodgkin's lymphoma.¹⁴¹

Electroporation (EP) is currently used in clinical trials for mRNA delivery to generate CAR-T cells, but it has shown limited antitumor efficacy. In contrast, LNPs have demonstrated significantly prolonged efficacy and lower cytotoxicity in vitro compared to EP-CAR-T cells. The mRNA delivered by LNPs efficiently transfects T cells without integrating into the genome, reducing the risk of long-term toxicity. These results underscore the significant promise of mRNA–LNP delivery for the ex vivo engineering of CAR-T cells.¹⁴²

Recently, preclinical studies have shown that SYS6020 effectively kills BCMA-positive myeloma cells and has demonstrated good safety and efficacy. SYS6020, developed by CSPC group, is the world's first CAR-T cell therapy based on mRNA–LNP to receive clinical trial approval. The drug expresses CARs that specifically recognize BCMA antigens, targeting and killing BCMA-positive tumor cells. Compared to traditional CAR-T therapies, SYS6020 exhibits high cell viability, high CAR positivity, no risk of genomic integration, and lower incidence of CRS and other side effects.¹⁴³

Although mRNA–LNP-based approaches, particularly in the context of CAR-T therapies, have shown great promise in treating various malignancies, tumor cells often develop resistance that limits long-term efficacy. A key resistance mechanism is antigen loss or downregulation, where tumor cells reduce or completely eliminate the expression of targeted surface antigens.¹⁴⁴ This phenomenon is commonly observed in both hematological malignancies and solid tumors. In the case of mRNA-engineered CAR-T cells, the absence of the specific antigen results in immune escape and tumor relapse, especially when targeting a single antigen.¹⁴⁵

In addition to antigen modulation, the TME plays a crucial role in suppressing immune activity.¹⁴⁶ Tumors may recruit regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs), and secrete immunosuppressive cytokines such as IL-10 and TGF-β, which collectively impair the function and persistence of mRNA-generated immune cells.¹⁴⁷ The hypoxic and acidic nature of the TME may further hinder T cell infiltration and survival.

Moreover, epigenetic changes in tumor cells can alter antigen presentation pathways,¹⁴⁸ such as downregulation of MHC class I molecules, reducing recognition by TCR-engineered cells.¹⁴⁹ Another challenge lies in the transient nature of mRNA expression, which may lead to insufficient duration of therapeutic protein or receptor expression, allowing tumors to escape immune pressure once expression wanes.

To overcome these obstacles, several strategies are under investigation, including dual or multi-antigen targeting, mRNA encoding for cytokine adjuvants, armored CAR constructs, and combination with checkpoint inhibitors to reactivate exhausted T cells.¹¹⁰ Rational design of LNPs that can co-deliver mRNA and immune modulators may further enhance therapeutic durability and counteract resistance.

4.2.2 CAR-NK. NK cells are innate immune cells that respond quickly and directly kill pathogens, playing a crucial role in defending against tumors and viral infections. Compared to CAR-T cell therapy, CAR-NK has several key advantages.¹⁵⁰ CAR-NK cells can be derived from diverse sources such as the NK-92 cell line, peripheral blood, and umbilical cord blood, enabling the production of “off-the-shelf” products for immediate patient use. This approach is clinically safer, with lower risks of neurotoxicity and CRS.¹⁵¹ CAR-NK cells also show broad antitumor activity and practical utility, making them a versatile and effective option. Xu's team utilized the NEPA21 mRNA electroporation method, successfully demonstrating that NKG2D CAR-NK cells exhibit significant specific killing activity against tumor cells expressing high levels of NKG2D ligands in both in vitro and in vivo models. The study showed that local administration of NKG2D CAR-NK cells effectively reduced tumor burden in colorectal cancer, with tumor necrosis and complete metabolic remission observed at the injection site, offering a new option for enhanced tumor immunotherapy. Furthermore, the research team led by Vita Golubovskaya expanded NK cells from primary PBMCs and efficiently induced CAR expression using mRNA–LNP technology.¹⁵¹ Experiments showed that NK cells transfected with CD19-CAR mRNA–LNP effectively killed two types of leukemia cell lines in a dose-dependent manner. These CAR-NK cells also exhibited significantly higher secretion of IFN-γ and granzyme B compared to regular NK cells. Both in vitro and in vivo studies demonstrated their strong anti-tumor activity.¹⁵²

Shin and colleagues have designed a novel type of bifunctional lipid nanoparticle containing DOTAP. These nanoparticles are engineered to enhance the antitumor efficacy of CAR-NK cells by facilitating their activation and improving the delivery efficiency of CAR mRNA. In an orthotopic mouse model, CAR-NK cells targeting glypican-3, which is commonly overexpressed in hepatocellular carcinoma, demonstrated significant therapeutic efficacy.¹⁵³

In various pediatric sarcomas, EphA2 receptors are overexpressed and critically involved in tumor progression and angiogenesis. Recently, Pui Yeng Lam et al. demonstrated that EphA2-specific CAR-NK cells show significant cytotoxicity against sarcoma cell lines in vitro and significant antitumor activity in rhabdomyosarcoma and osteosarcoma mouse models.¹⁵²

4.2.3 CAR-M. CAR-M are emerging as a promising cell type in cancer immunotherapy. Similar to CAR-T and CAR-NK cells, CAR-M consists of an extracellular signaling domain that recognizes tumor-specific antigens, a transmembrane region, and an intracellular activation signal. By extracting macrophages from patients and genetically modifying them, CAR-M cells can be engineered to kill tumors.¹⁵⁴ Compared to T cells and NK cells, macrophages are more easily infiltrate the immunosuppressive TME, offering new opportunities for tumor immunotherapy. Utilizing mRNA technology to deliver engineered macrophages enables personalized treatment.

Jiang's team developed liver macrophage-targeted LNPs encapsulating mRNA encoding CAR and CD24-Siglec-G (Siglec-GΔITIMs) lacking ITIMs. In HCC mouse models, these LNPs significantly boosted liver macrophage phagocytosis, effectively reduced tumor burden, and prolonged mouse survival.¹⁵⁵ Recently, a research team developed a CAR-M therapy, CT-0508, targeting HER2-overexpressing advanced solid tumors. This therapy reprograms macrophages to specifically recognize and phagocytose HER2-overexpressing tumor cells, while also activating the TME and inducing anti-tumor immune responses. Clinical results have shown preliminary safety and feasibility of this therapy in solid tumor patients.¹⁵⁶

4.3 Restoration of tumor suppressor

Tumor suppressor genes encode proteins that play a critical role in regulating cellular pathways that prevent tumor growth and progression. In various cancers, key tumor suppressor proteins like p53, phosphatase and tensin homolog deleted on chromosome 10 (PTEN) are often rendered inactive due to mutations or loss of expression, resulting in uncontrolled cell proliferation.¹⁵⁷ By utilizing mRNA-based therapies to restore the function of these genes, it is possible to reactivate important anti-tumor mechanisms, such as inducing apoptosis and suppressing tumor cell proliferation, offering a promising avenue for cancer treatment.

4.3.1 p53. As a critical regulator of cell cycle arrest, senescence, and apoptosis, the p53 tumor suppressor gene plays a pivotal role in preventing tumorigenesis.¹⁵⁸ Consequently, restoring p53 expression has become a promising approach in cancer immunotherapy. Mutations or inactivation of the p53 gene are frequently observed in various cancers, including approximately 36% of hepatocellular carcinoma (HCC) cases and nearly 68% of non-small cell lung cancer (NSCLC) cases.¹⁵⁹ In a recent study led by Na Kong, an innovative redox-responsive nanoparticle system was developed to efficiently deliver synthetic mRNA encoding the p53 protein. The findings revealed that nanoparticles loaded with p53-mRNA significantly inhibited the proliferation of p53-null HCC and NSCLC cells by inducing cell cycle arrest and apoptosis. Moreover, the restoration of p53 function notably enhanced the sensitivity of these cancer cells to everolimus, an mTOR inhibitor that had previously shown limited clinical efficacy in advanced HCC and NSCLC. Importantly, the dual targeting of the p53 tumor-suppressive pathway and the oncogenic mTOR signaling pathway demonstrated remarkable antitumor effects, both in vitro and in diverse animal models of HCC and NSCLC.¹⁶⁰ Additionally, Xiao designed an mRNA nanoparticle platform targeting CXCR4 to deliver p53 mRNA, which was combined with anti-PD-1 monoclonal antibody therapy. This strategy effectively induced p53 expression in an HCC model and facilitated global reprogramming of the TME. Compared to monotherapy with anti-PD-1 or therapeutic p53 expression alone, the combination therapy significantly enhanced antitumor effects (Fig. 11a).¹⁶¹

Fig. 11 mRNA encoding tumor suppressors for cancer therapy. (a) A schematic showing CXCR4-targeted p53 mRNA nanoparticles and their synergistic effect with anti-PD-1 therapy in reprogramming the TME. The nanoparticles activate CD8⁺ T cells and NK cells, repolarize TAMs, and enhance anti-tumor cytokine expression. Right panel shows fluorescence imaging of treatment effects over time, highlighting the superior efficacy of CTCE-p53 NPs combined with anti-PD-1 therapy.¹⁶¹ Copyright 2022, Springer Nature. (b) A schematic showing PTEN/Pep LNPs inducing PTEN-mediated autophagy and cell death, activating immune responses, and significantly inhibiting tumor growth, reducing weight loss, and lowering fluorescence signals compared to controls.⁹⁴ Copyright 2024, Advanced Science published by Wiley-VCH GmbH.

4.3.2 PTEN. PTEN is a key tumor suppressor in the development and progression of different tumor types.¹⁶² Islam demonstrated that PTEN mRNA can be effectively introduced into PTEN-deficient prostate cancer cells both in vitro and in vivo using polymer–lipid hybrid nanoparticles coated with a PEG shell. This approach enables the re-expression of PTEN, leading to significant inhibition of tumor growth in prostate cancer mouse models. Mechanistically, the restoration of PTEN function in these cells is associated with the suppression of the PI3K-AKT pathway and an increase in apoptosis.¹⁶³

Moreover, in a triple-negative breast cancers (TNBC) model, Pep LNP delivers PTEN mRNA to induce autophagy and immunogenic cell death (ICD), activate dendritic cells, and enhance CD8⁺ T-cell infiltration, thereby effectively inhibiting tumor growth and metastasis (Fig. 11b).⁹⁴ Additionally, Yue Hu and colleagues developed an ionizable lipid library and identified PPz-2R1, which efficiently delivers PTEN mRNA to the lungs. In a PTEN-deficient lung cancer model, this delivery restored PTEN function, mitigated the immunosuppressive TME, and enhanced antitumor immunity. Combined with anti-PD-1 therapy, it significantly inhibited tumor growth, surpassing the efficacy of either treatment alone.¹⁶⁴

4.4 Immunomodulatory factors

Immunomodulatory factors play a pivotal role in immune responses by regulating cell proliferation, differentiation, and signaling pathways, thereby maintaining the balance of the immune system. These factors are particularly crucial in cancer immunotherapy, where they can modulate the TME and enhance antitumor immune responses. It is critical for the survival and proliferation of regulatory T cells (Tregs). Seymour de Picciotto et al. designed a long-acting mutant IL-2 fusion protein (HSA-IL2m) by fusing three mutated IL-2 molecules with human serum albumin (HSA). The HSA-IL2m mRNA was encapsulated in LNPs for delivery. This innovative approach selectively activates and expands Tregs in mice and non-human primates, while significantly mitigating disease severity in mouse models of acute graft-versus-host disease (GvHD) and experimental autoimmune encephalomyelitis (EAE). These findings underscore the therapeutic potential of mRNA-encoded HSA-IL2m for the treatment of autoimmune diseases.¹⁶⁵ Also, in a related study, it was found that Bin Wang et al. developed LNP encapsulating IL-12 mRNA (DMT7-IL12 LNP) and used MSA-2 as a STING agonist. In mouse tumor models of B16F10 melanoma and 4T1 breast cancer, the co-delivery of DMT7-IL12 LNP and MSA-2 significantly increased the number of CD8⁺ T cells in the TME and enhanced the secretion of key cytokines, including granzyme B, TNF-α, and IFN-γ. Compared with monotherapy using MSA-2 or DMT7-IL12 LNP alone, the combination immunotherapy (DMT7-IL12 LNP + MSA-2) demonstrated robust antitumor efficacy and effectively reversed the state of T cell exhaustion in the TMEICD (Fig. 12a).¹⁶⁶ Additionally, another study by Yizhou Dong and colleagues proposed a novel antitumor immunotherapy strategy that combines the delivery of mRNA encoding costimulatory molecules with their corresponding agonist antibodies. They synthesized and constructed a library of biomimetic ionizable lipids derived from phospholipids and glycolipids, identifying that the phospholipid derivative PL1 can efficiently deliver mRNA encoding costimulatory molecules CD137 and OX40 to T cells. In A20 and B16F10 tumor models, the combination of PL1-OX40 mRNA and anti-OX40 antibody demonstrated significantly enhanced antitumor efficacy compared to anti-OX40 antibody monotherapy (Fig. 12b).¹⁶⁷ Moreover, Amisaki designed a triad LNP therapy, utilizing LNPs to intratumorally deliver a mixture of nucleoside-modified mRNA encoding the cytokines IL-21 and IL-7, as well as the immunostimulatory molecule 4-1BBL. This triplet LNP formulation leverages the synergistic effects of IL-21 with IL-7 and 4-1BBL to markedly enhance the frequency of tumor-infiltrating CD8⁺ T cells and their ability to produce granzyme B and IFN-γ. Consequently, this approach achieves tumor eradication and establishes long-term immunological memory, effectively preventing tumor recurrence (Fig. 12c).¹¹⁰

Fig. 12 mRNA-based therapeutics via expressing antibodies, cytokines, and costimulatory molecules. (a) A schematic showing DMT7-IL12 LNPs combined with MSA-2 activating the immune system through STING, enhancing CD8⁺ T cells and NK cells, and effectively inhibiting tumor growth.¹⁶⁶ Copyright 2024, American Chemical society. (b) A schematic showing enhanced cancer immunotherapy using PL1-OX40 mRNA combined with anti-OX40 antibody to activate T cells, with a graph illustrating reduced tumor size over time compared to controls.¹⁶⁷ Copyright 2021, Springer Nature. (c) A schematic showing tumor volume changes in treated and untreated mice after E0771 cell inoculation, highlighting the effectiveness of Triplet LNP in inhibiting tumor growth compared to vehicle and control LNPs.¹¹⁰ Copyright 2024, Springer Nature.

In recent clinical settings, several mRNA-based immunomodulatory strategies have entered early-phase trials. For example, mRNA-2752, encoding IL-23, IL-36γ, and OX40L, has been evaluated in patients with solid tumors showing preliminary immune activation and manageable toxicity.¹⁶⁸

Another candidate, MEDI1191, an mRNA-based therapeutic encoding IL-12 and formulated in LNP, has completed phase I clinical trials in combination with the anti-PD-L1 antibody durvalumab for the treatment of advanced solid tumors.¹⁶⁹ Clinical results suggest manageable safety and early signs of antitumor activity, highlighting the translational feasibility of mRNA–LNP platforms for localized immune modulation.

Despite these encouraging developments, several critical obstacles remain that could limit the durability, safety, and broad applicability of mRNA-based immunomodulatory therapies.¹⁷⁰ The transient expression of mRNA and the need for repeated dosing remain challenges for achieving sustained immune stimulation.¹⁷¹ Moreover, balancing immune activation with systemic toxicity is particularly critical when encoding potent cytokines such as IL-12 or IL-15. There is also a risk of overstimulation or CRS, which necessitates careful control of dose, route, and delivery specificity.¹⁷² In addition, tumor heterogeneity and variable immune landscapes may limit the generalizability of responses across patients.

Taken together, while mRNA–LNP mediated delivery of immunomodulatory factors represents a promising frontier in cancer immunotherapy, ongoing clinical trials are essential to optimize delivery parameters, refine dosing regimens, and ensure long-term safety.¹⁷³ The integration of rational combination strategies and biomarker-guided patient selection will be key to translating these approaches into effective and personalized cancer treatments.¹⁷⁴

4.5 Combination therapy

While innovative cancer therapies are constantly emerging, their efficacy is often compromised by immune evasion, limited antigen coverage, or a suppressive TME. To address these challenges, combination therapies based on mRNA–LNP platforms have emerged as a versatile and powerful strategy.¹⁷⁵ These approaches leverage the inherent flexibility of mRNA delivery to integrate complementary therapeutic modalities, thereby enhancing antitumor immunity through multiple synergistic mechanisms. Recent studies have demonstrated the feasibility and potential of such combination therapies in diverse preclinical models and clinical settings.¹⁷⁶

Photodynamic therapy (PDT), which uses laser-activated photosensitizers to generate ROS to kill cancer cells, is highly safe and selective but limited by poor tissue penetration. To address this, Zhou developed a novel nanoparticle system that co-delivers p53 mRNA and the photosensitizer indocyanine green (ICG). This ROS-responsive nanoparticle platform enables precise drug release. PDT-induced ROS not only causes direct oxidative damage to tumor cells but also enhances cell membrane permeability and promotes ICD, increasing the exposure of damage-associated molecular patterns (DAMPs) such as calreticulin. The study showed that this combined strategy significantly reduced the viability of tumor cells in vitro and effectively inhibited tumor growth in a lung cancer mouse model.¹⁷⁷ Moreover, Rao Fu and et al. developed rVSV-LCMVG, a low-immunogenic oncolytic virus that's unlikely to trigger virus–neutralizing antibodies, and combined it with adoptive T-cell transfer. The combination therapy significantly increased intratumoral cytokines and chemokines, recruited CD8⁺ T-cells to the TME, and initiated an anti-tumor immune response. In B16 tumor-bearing mice, it showed better anti-tumor effects than single treatments.¹⁷⁸ This synergy results from the combined actions of the oncolytic virus and mRNA-transfected T cells. The virus selectively replicates in and lyses tumor cells, releasing tumor-associated antigens, while the T cells recognize and eliminate these antigen-expressing cells. Furthermore, virus-induced inflammation enhances antigen presentation and T-cell recruitment via type I interferon signaling, strengthening adaptive immunity. Additionally, mRNA-4157 (V940), a personalized neoantigen therapy targeting up to 34 patient-specific tumor neoantigens, is designed to enhance antitumor activity by inducing T-cell responses. Research by Weber has shown that in patients with high-risk melanoma, the combination of mRNA-4157 and pembrolizumab significantly extends recurrence-free survival compared to monotherapy, while demonstrating manageable safety.¹⁷⁹ This combination works through the synergistic activation of both the innate and adaptive immune systems. mRNA-encoded neoantigens are presented by dendritic cells through MHC class I/II pathways, priming CD8⁺ and CD4⁺ T cells to recognize and attack tumor cells. Pembrolizumab, an anti-PD-1 antibody, blocks the PD-1/PD-L1 interaction, preventing T-cell exhaustion and sustaining effector T-cell activity, which allows the T-cells to maintain their antitumor functions over time. In another study, Wang developed the DPPA/IL-15 NP nanoplatform, which has ultrasound-responsive and CEUS imaging properties. It can target PD-L1 to deliver IL-15 mRNA to tumors. The platform effectively protects IL-15 mRNA and uses UTMD to specifically transfect it into tumor cells. This activates IL-15-related immune effector cells and blocks the PD-1/PD-L1 pathway. Studies show the combination therapy can trigger a strong systemic immune response and enhance antitumor efficacy.¹⁸⁰ The combination therapy also blocks the PD-1/PD-L1 pathway, further enhancing T-cell-mediated immunity and overcoming tumor-induced immune suppression. For instance, in a clinical study, Shen's team treated two end-stage cancer patients with a KRAS G12V neoantigen mRNA vaccine plus pembrolizumab, a PD-1 inhibitor. Results showed this combination shrank the patients' tumors.¹⁸¹ The KRAS G12V neoantigen mRNA vaccine induces the presentation of mutant KRAS peptides on tumor cells, triggering a specific immune response targeting the mutant protein, which is not expressed in normal cells. Pembrolizumab, by blocking the PD-1/PD-L1 axis, prevents immune evasion by the tumor, allowing T cells to remain activated and efficiently kill tumor cells expressing KRAS G12V. In conclusion, mRNA-based combination therapies show great potential in cancer treatment and offer new strategies. By integrating mRNA vaccines with immune checkpoint inhibitors, cytokines, and oncolytic agents, these combinations not only boost the immune system's ability to target cancer cells but also provide a multifaceted approach to overcome tumor resistance and immune evasion.

4.6 Other therapies

Gene-editing technologies hold great promise in cancer therapy by precisely targeting oncogenes or tumor suppressor genes, thereby directly modulating tumor cell behavior and reshaping the immune microenvironment.¹⁸² Rosenblum and colleagues utilized LNPs to deliver Cas9 mRNA and sgRNA. A single intracranial injection of LNP-formulated sgPLK1 and Cas9 mRNA (sgPLK1-cLNPs) into a glioblastoma model achieved a 70% PLK1 gene-editing efficiency, significantly inducing tumor cell apoptosis, reducing tumor growth by 50%, and increasing survival rates by 30%. Additionally, intraperitoneal injection of EGFR-targeted sgPLK1-cLNPs in a metastatic ovarian cancer model resulted in an 82% gene-editing efficiency, markedly inhibiting tumor growth and raising survival rates by 80%.¹⁸³

mRNA-encoded antibodies can be used to target specific antigens on the surface of tumor cells, thereby activating the immune system to attack the tumor.¹⁸⁴ To date, mRNA has been designed to express various types of antibodies, including monoclonal antibodies (mAbs) and bispecific antibodies. Bevacizumab, an anti-VEGF antibody approved by the US FDA for cancer therapy in 2004, can specifically bind to vascular endothelial growth factor (VEGF) and block its interaction with receptors, thereby inhibiting tumor angiogenesis. Le designed a novel lung-selective PBAE (poly(β-amino ester)) nanoparticle to encapsulate mRNA encoding bevacizumab. Results showed that in an orthotopic NSCLC mouse model, bevacizumab mRNA delivered by lung-targeted PBAE nanoparticles more significantly inhibited tumor proliferation and angiogenesis compared to recombinant bevacizumab protein.¹⁸⁵

Similarly, Rybakova developed an IVT-mRNA-based system for delivering anti-HER2 antibodies (trastuzumab). By optimizing the mRNA sequence and encapsulating it in LNPs, the system achieved efficient in vivo antibody expression, effectively inhibiting the growth of HER2-positive tumors and improving survival rates.¹⁸⁶

5. Conclusion and perspective

LNP-mediated mRNA delivery systems have shown significant potential in cancer therapy. Recent advancements, particularly the success of mRNA technology in COVID-19 vaccines, have accelerated its application in disease treatment. In cancer therapy, mRNA-based strategies such as tumor vaccines, adoptive cell transfer therapies, restoration of tumor suppressors, and immunomodulatory factor delivery have achieved notable progress. However, challenges like the hepatic tropism of LNPs, which limits their application in non-hepatic tissues, and the need to optimize mRNA immunogenicity and delivery efficiency, still remain.

Future research should focus on key areas: developing novel LNP components and structures for precise organ and cell targeting, reducing non-specific accumulation and toxicity; incorporating biodegradable materials and new lipid alternatives (e.g., PEG alternatives) to lower LNP immunogenicity and toxicity; optimizing administration routes and dosing strategies to enhance mRNA stability and delivery efficiency; and leveraging AI and high-throughput screening to expedite LNP optimization and personalized therapy development.

In cancer treatment, mRNA–LNP systems can trigger specific immune responses and enhance anti-tumor effects through combination therapies. For instance, integrating mRNA vaccines, CAR-T cell therapies, and gene editing technologies offers innovative cancer treatment approaches. Nevertheless, achieving long-term mRNA expression and precise organ-targeted delivery to avoid off-target effects are ongoing challenges. We expect these issues to be resolved with the development of organ-targeted LNPs and long-term mRNA expression technologies.

Notably, recent clinical trials evaluating mRNA-based immunotherapies have confirmed their translational potential, demonstrating preliminary safety and biological activity in patients with various solid tumors. However, these studies have also underscored several limitations, such as the transient nature of mRNA expression, the need for repeated dosing, and the risk of systemic toxicity when delivering potent immunostimulatory cytokines.¹⁸⁷ In addition, heterogeneity in tumor immune microenvironments may lead to inconsistent clinical responses, highlighting the necessity for biomarker-guided patient selection and individualized treatment design.¹⁸⁸

Looking ahead, the field of mRNA-based cancer therapy is poised for significant breakthroughs. The continuous refinement of LNP formulations and delivery methods will likely expand the range of treatable cancers and improve therapeutic outcomes. Exploring new mRNA formats, such as self-amplifying RNA (saRNA),¹⁸⁹ trans-acting RNA (taRNA),¹⁹⁰ and circular RNA (circRNA),¹⁹¹ could extend the duration of mRNA translation products, further enhancing therapeutic stability and efficacy. Combining mRNA delivery with other treatments like chemotherapy, radiotherapy, and immunotherapy also holds promise for overcoming resistance and achieving synergistic effects. As research into mRNA biology and delivery mechanisms advances, these scientific breakthroughs will translate more rapidly into clinical applications, bringing us closer to fully realizing the potential of mRNA-based therapies in cancer treatment.

In conclusion, LNP-mediated mRNA delivery systems offer new hope for cancer therapy. Despite existing challenges, ongoing technological advancements and multidisciplinary collaboration are expected to drive more breakthroughs, providing more effective treatment options for cancer patients.

Author contributions

S. Y., H. X. W. and Z. J. wrote the manuscript. Z. X. D., W. H. B., H. X. C. and X. Y. L. revised the manuscript. X. Y. L. conceived the review topic.

Conflicts of interest

The authors declare no competing financial interest.

Data availability

No new data were generated or analyzed in this study. All information discussed in this review is derived from previously published sources, which are properly cited in the references.

Acknowledgements

The work was supported by the National Key R&D Program of China (2023YFC3605500), National Natural Science Foundation of China (22477129, 82273796, 82372005, and 82171986), Taishan Scholar Project of Shandong Province (TSQN202306320), Special Supporting Funds for Leading Talents at or above the Provincial Level in Yantai, Natural Science Foundation of Shandong Province (ZR2023MB085, ZR2024QH253), Shandong Laboratory Program (SYS202205).

Notes and references

1. X. Hou, T. Zaks, R. Langer and Y. Dong, Nat. Rev. Mater., 2021, 6, 1078–1094.
2. X. Huang, N. Kong, X. Zhang, Y. Cao, R. Langer and W. Tao, Nat. Med., 2022, 28, 2273–2287.
3. L. J. Kubiatowicz, A. Mohapatra, N. Krishnan, R. H. Fang and L. Zhang, Exploration, 2022, 2, 20210217.
4. A. Magadum, K. Kaur and L. Zangi, Mol. Ther., 2019, 27, 785–793.
5. B. Li, R. S. Manan, S.-Q. Liang, A. Gordon, A. Jiang, A. Varley, G. Gao, R. Langer, W. Xue and D. Anderson, Nat. Biotechnol., 2023, 41, 1410–1415.
6. R. Tenchov, R. Bird, A. E. Curtze and Q. Zhou, ACS Nano, 2021, 15, 16982–17015.
7. Y. Eygeris, M. Gupta, J. Kim and G. Sahay, Acc. Chem. Res., 2022, 55, 2–12.
8. Y. Xu, A. Golubovic, S. Xu, A. Pan and B. Li, J. Mater. Chem. B, 2023, 11, 6527–6539.
9. C. Hald Albertsen, J. A. Kulkarni, D. Witzigmann, M. Lind, K. Petersson and J. B. Simonsen, Adv. Drug Delivery Rev., 2022, 188, 114416.
10. B. Li, I. O. Raji, A. G. R. Gordon, L. Sun, T. M. Raimondo, F. A. Oladimeji, A. Y. Jiang, A. Varley, R. S. Langer and D. G. Anderson, Nat. Mater., 2024, 23, 1002–1008.
11. R. Van Der Meel, F. Grisoni and W. J. M. Mulder, Nat. Mater., 2024, 23, 880–881.
12. M. Gao, J. Zhong, X. Liu, Y. Zhao, D. Zhu, X. Shi, X. Xu, Q. Zhou, W. Xuan, Y. Zhang, Y. Zhou and J. Cheng, ACS Nano, 2025, 19, 5966–5978.
13. K. Su, L. Shi, T. Sheng, X. Yan, L. Lin, C. Meng, S. Wu, Y. Chen, Y. Zhang, C. Wang, Z. Wang, J. Qiu, J. Zhao, T. Xu, Y. Ping, Z. Gu and S. Liu, Nat. Commun., 2024, 15, 5659.
14. A. Akinc, M. A. Maier, M. Manoharan, K. Fitzgerald, M. Jayaraman, S. Barros, S. Ansell, X. Du, M. J. Hope, T. D. Madden, B. L. Mui, S. C. Semple, Y. K. Tam, M. Ciufolini, D. Witzigmann, J. A. Kulkarni, R. Van Der Meel and P. R. Cullis, Nat. Nanotechnol., 2019, 14, 1084–1087.
15. W. Kong, Y. Wei, Z. Dong, W. Liu, J. Zhao, Y. Huang, J. Yang, W. Wu, H. He and J. Qi, J. Nanobiotechnol., 2024, 22, 553.
16. S. Zhao, K. Gao, H. Han, M. Stenzel, B. Yin, H. Song, A. Lawanprasert, J. E. Nielsen, R. Sharma, O. H. Arogundade, S. Pimcharoen, Y.-J. Chen, A. Paul, J. Tuma, M. G. Collins, Y. Wyle, M. G. Cranick, B. W. Burgstone, B. S. Perez, A. E. Barron, A. M. Smith, H. Y. Lee, A. Wang and N. Murthy, Nat. Nanotechnol., 2024, 19, 1702–1711.
17. Y. Fei, X. Yu, P. Liu, H. Ren, T. Wei and Q. Cheng, Adv. Mater., 2024, 36, 2409812.
18. Q. Cheng, T. Wei, L. Farbiak, L. T. Johnson, S. A. Dilliard and D. J. Siegwart, Nat. Nanotechnol., 2020, 15, 313–320.
19. Y. Lin, Q. Cheng and T. Wei, Biophys. Rep., 2023, 9, 255.
20. H. Parhiz, V. V. Shuvaev, N. Pardi, M. Khoshnejad, R. Y. Kiseleva, J. S. Brenner, T. Uhler, S. Tuyishime, B. L. Mui, Y. K. Tam, T. D. Madden, M. J. Hope, D. Weissman and V. R. Muzykantov, J. Controlled Release, 2018, 291, 106–115.
21. F. Wang, J. Lou, X. Lou, F. Wu, X. Gao, X. Yao, J. Wan, X. Duan, W. Deng, L. Ma, L. Zhang, G. He, M. Wang, C. Ni, N. Lei and Z. Qin, Adv. Sci., 2025, 12, 2412543.
22. L. Xue, A. G. Hamilton, G. Zhao, Z. Xiao, R. El-Mayta, X. Han, N. Gong, X. Xiong, J. Xu, C. G. Figueroa-Espada, S. J. Shepherd, A. J. Mukalel, M.-G. Alameh, J. Cui, K. Wang, A. E. Vaughan, D. Weissman and M. J. Mitchell, Nat. Commun., 2024, 15, 1884.
23. H. Kim, R. Zenhausern, K. Gentry, L. Lian, S. G. Huayamares, A. Radmand, D. Loughrey, A. R. Podilapu, M. Z. C. Hatit, H. Ni, A. Li, A. Shajii, H. E. Peck, K. Han, X. Hua, S. Jia, M. Martinez, C. Lee, P. J. Santangelo, A. Tarantal and J. E. Dahlman, Nat. Biotechnol., 2024, 42, 1–8.
24. W. Park, J. Choi, J. Hwang, S. Kim, Y. Kim, M. K. Shim, W. Park, S. Yu, S. Jung, Y. Yang and D.-H. Kweon, ACS Nano, 2025, 19, 6412–6425.
25. Y. Zong, Y. Lin, T. Wei and Q. Cheng, Adv. Mater., 2023, 35, 2303261.
26. J. Yu, Q. Li, C. Zhang, Q. Wang, S. Luo, X. Wang, R. Hu and Q. Cheng, Biomaterials, 2025, 317, 123047.
27. H. Ni, M. Z. C. Hatit, K. Zhao, D. Loughrey, M. P. Lokugamage, H. E. Peck, A. D. Cid, A. Muralidharan, Y. Kim, P. J. Santangelo and J. E. Dahlman, Nat. Commun., 2022, 13, 4766.
28. S. C. Semple, A. Akinc, J. Chen, A. P. Sandhu, B. L. Mui, C. K. Cho, D. W. Y. Sah, D. Stebbing, E. J. Crosley, E. Yaworski, I. M. Hafez, J. R. Dorkin, J. Qin, K. Lam, K. G. Rajeev, K. F. Wong, L. B. Jeffs, L. Nechev, M. L. Eisenhardt, M. Jayaraman, M. Kazem, M. A. Maier, M. Srinivasulu, M. J. Weinstein, Q. Chen, R. Alvarez, S. A. Barros, S. De, S. K. Klimuk, T. Borland, V. Kosovrasti, W. L. Cantley, Y. K. Tam, M. Manoharan, M. A. Ciufolini, M. A. Tracy, A. De Fougerolles, I. MacLachlan, P. R. Cullis, T. D. Madden and M. J. Hope, Nat. Biotechnol., 2010, 28, 172–176.
29. I. Yoon, L. Xue, Q. Chen, J. Liu, J. Xu, Z. Siddiqui, D. Kim, B. Chen, Q. Shi, E. Laura Han, M. Cherry Ruiz, K. H. Vining and M. J. Mitchell, Angew. Chem., Int. Ed., 2025, 64, e202415389.
30. A. J. Da Silva Sanchez, K. Zhao, S. G. Huayamares, M. Z. C. Hatit, M. P. Lokugamage, D. Loughrey, C. Dobrowolski, S. Wang, H. Kim, K. Paunovska, Y. Kuzminich and J. E. Dahlman, J. Controlled Release, 2023, 353, 270–277.
31. L. Miao, L. Li, Y. Huang, D. Delcassian, J. Chahal, J. Han, Y. Shi, K. Sadtler, W. Gao, J. Lin, J. C. Doloff, R. Langer and D. G. Anderson, Nat. Biotechnol., 2019, 37, 1174–1185.
32. X. Zhao, J. Chen, M. Qiu, Y. Li, Z. Glass and Q. Xu, Angew. Chem., 2020, 132, 20258–20264.
33. K. Mrksich, M. S. Padilla, R. A. Joseph, E. L. Han, D. Kim, R. Palanki, J. Xu and M. J. Mitchell, J. Biomed. Mater. Res., Part A, 2024, 112, 1494–1505.
34. K. Hashiba, M. Taguchi, S. Sakamoto, A. Otsu, Y. Maeda, Y. Suzuki, H. Ebe, A. Okazaki, H. Harashima and Y. Sato, Nano Lett., 2024, 24, 12758–12767.
35. G. Tilstra, J. Couture-Senécal, Y. M. A. Lau, A. M. Manning, D. S. M. Wong, W. W. Janaeska, T. A. Wuraola, J. Pang and O. F. Khan, J. Am. Chem. Soc., 2023, 145, 2294–2304.
36. K. Hashiba, Y. Sato, M. Taguchi, S. Sakamoto, A. Otsu, Y. Maeda, T. Shishido, M. Murakawa, A. Okazaki and H. Harashima, Small Sci., 2023, 3, 2200071.
37. X. Han, J. Xu, Y. Xu, M.-G. Alameh, L. Xue, N. Gong, R. El-Mayta, R. Palanki, C. C. Warzecha, G. Zhao, A. E. Vaughan, J. M. Wilson, D. Weissman and M. J. Mitchell, Nat. Commun., 2024, 15, 1762.
38. Y. Zhang, C. Sun, C. Wang, K. E. Jankovic and Y. Dong, Chem. Rev., 2021, 121, 12181–12277.
39. Z. Chen, Y. Tian, J. Yang, F. Wu, S. Liu, W. Cao, W. Xu, T. Hu, D. J. Siegwart and H. Xiong, J. Am. Chem. Soc., 2023, 145, 24302–24314.
40. W. Cai, T. Luo, X. Chen, L. Mao and M. Wang, Adv. Funct. Mater., 2022, 32, 2204947.
41. G. Somu Naidu, R. Rampado, P. Sharma, A. Ezra, G. R. Kundoor, D. Breier and D. Peer, ACS Nano, 2025, 19, 6571–6587.
42. M. Qiu, Y. Tang, J. Chen, R. Muriph, Z. Ye, C. Huang, J. Evans, E. P. Henske and Q. Xu, Proc. Natl. Acad. Sci. U. S. A., 2022, 119, e2116271119.
43. K. Lv, Z. Yu, J. Wang, N. Li, A. Wang, T. Xue, Q. Wang, Y. Shi, L. Han, W. Qin, J. Gong, H. Song, T. Zhang, C. Chang, H. Chen, X. Zhong, J. Ding, R. Chen, M. Liu, W. Zhang, S. Cen and Y. Dong, Adv. Sci., 2024, 11, 2404684.
44. X. Huang, Y. Ma, G. Ma and Y. Xia, Research, 2024, 7, 0370.
45. Z. He, Z. Le, Y. Shi, L. Liu, Z. Liu and Y. Chen, Angew. Chem., 2023, 135, e202310401.
46. X. Han, M.-G. Alameh, Y. Xu, R. Palanki, R. El-Mayta, G. Dwivedi, K. L. Swingle, J. Xu, N. Gong, L. Xue, Q. Shi, I.-C. Yoon, C. C. Warzecha, J. M. Wilson, D. Weissman and M. J. Mitchell, Nat. Biomed. Eng., 2024, 8, 1412–1424.
47. W. Dong, Z. Li, T. Hou, Y. Shen, Z. Guo, Y.-T. Su, Z. Chen, H. Pan, W. Jiang and Y. Wang, J. Am. Chem. Soc., 2024, 146, 15085–15095.
48. X. Han, M.-G. Alameh, N. Gong, L. Xue, M. Ghattas, G. Bojja, J. Xu, G. Zhao, C. C. Warzecha, M. S. Padilla, R. El-Mayta, G. Dwivedi, Y. Xu, A. E. Vaughan, J. M. Wilson, D. Weissman and M. J. Mitchell, Nat. Chem., 2024, 16, 1687–1697.
49. C. Hald Albertsen, J. A. Kulkarni, D. Witzigmann, M. Lind, K. Petersson and J. B. Simonsen, Adv. Drug Delivery Rev., 2022, 188, 114416.
50. M. Z. C. Hatit, C. N. Dobrowolski, M. P. Lokugamage, D. Loughrey, H. Ni, C. Zurla, A. J. Da Silva Sanchez, A. Radmand, S. G. Huayamares, R. Zenhausern, K. Paunovska, H. E. Peck, J. Kim, M. Sato, J. I. Feldman, M.-A. Rivera, A. Cristian, Y. Kim, P. J. Santangelo and J. E. Dahlman, Nat. Chem., 2023, 15, 508–515.
51. S. Patel, N. Ashwanikumar, E. Robinson, Y. Xia, C. Mihai, J. P. Griffith, S. Hou, A. A. Esposito, T. Ketova, K. Welsher, J. L. Joyal, Ö. Almarsson and G. Sahay, Nat. Commun., 2020, 11, 983.
52. Y. Eygeris, S. Patel, A. Jozic and G. Sahay, Nano Lett., 2020, 20, 4543–4549.
53. S. Douka, L. E. Brandenburg, C. Casadidio, J. Walther, B. B. M. Garcia, J. Spanholtz, M. Raimo, W. E. Hennink, E. Mastrobattista and M. Caiazzo, J. Controlled Release, 2023, 361, 455–469.
54. R. Palanki, E. L. Han, A. M. Murray, R. Maganti, S. Tang, K. L. Swingle, D. Kim, H. Yamagata, H. C. Safford, K. Mrksich, W. H. Peranteau and M. J. Mitchell, Lab Chip, 2024, 24, 3790–3801.
55. S. K. Patel, M. M. Billingsley, C. Frazee, X. Han, K. L. Swingle, J. Qin, M.-G. Alameh, K. Wang, D. Weissman and M. J. Mitchell, J. Controlled Release, 2022, 347, 521–532.
56. O. Jung, H. Jung, L. T. Thuy, M. Choi, S. Kim, H.-G. Jeon, J. Yang, S.-M. Kim, T.-D. Kim, E. Lee, Y. Kim and J. S. Choi, Adv. Healthcare Mater., 2024, 13, 2303857.
57. C. Hald Albertsen, J. A. Kulkarni, D. Witzigmann, M. Lind, K. Petersson and J. B. Simonsen, Adv. Drug Delivery Rev., 2022, 188, 114416.
58. J. Li, X. Wang, T. Zhang, C. Wang, Z. Huang, X. Luo and Y. Deng, Asian J. Pharm. Sci., 2015, 10, 81–98.
59. Z. Du, M. M. Munye, A. D. Tagalakis, M. D. I. Manunta and S. L. Hart, Sci. Rep., 2014, 4, 7107.
60. B. Li, X. Luo, B. Deng, J. Wang, D. W. McComb, Y. Shi, K. M. L. Gaensler, X. Tan, A. L. Dunn, B. A. Kerlin and Y. Dong, Nano Lett., 2015, 15, 8099–8107.
61. K. J. Kauffman, J. R. Dorkin, J. H. Yang, M. W. Heartlein, F. DeRosa, F. F. Mir, O. S. Fenton and D. G. Anderson, Nano Lett., 2015, 15, 7300–7306.
62. E. Álvarez-Benedicto, L. Farbiak, M. M. Ramírez, X. Wang, L. T. Johnson, O. Mian, E. D. Guerrero and D. J. Siegwart, Biomater. Sci., 2022, 10, 549–559.
63. N. Chander, G. Basha, M. H. Y. Cheng, D. Witzigmann and P. R. Cullis, Mol. Ther.–Methods Clin. Dev., 2023, 30, 235–245.
64. I. V. Zhigaltsev and P. R. Cullis, Langmuir, 2023, 39, 3185–3193.
65. W. Zai, M. Yang, K. Jiang, J. Guan, H. Wang, K. Hu, C. Huang, J. Chen, W. Fu, C. Zhan and Z. Yuan, Signal Transduction Targeted Ther., 2024, 9, 150.
66. H. Zhang, C. Meng, X. Yi, J. Han, J. Wang, F. Liu, Q. Ling, H. Li and Z. Gu, ACS Nano, 2024, 18, 7825–7836.
67. A. J. D. S. Sanchez, D. Loughrey, E. S. Echeverri, S. G. Huayamares, A. Radmand, K. Paunovska, M. Hatit, K. E. Tiegreen, P. J. Santangelo and J. E. Dahlman, Adv. Healthcare Mater., 2024, 13, 2304033.
68. D. D. Kang, X. Hou, L. Wang, Y. Xue, H. Li, Y. Zhong, S. Wang, B. Deng, D. W. McComb and Y. Dong, Bioact. Mater., 2024, 37, 86–93.
69. A. Y. Jiang, S. Lathwal, S. Meng, J. Witten, E. Beyer, P. McMullen, Y. Hu, R. S. Manan, I. Raji, R. Langer and D. G. Anderson, J. Am. Chem. Soc., 2024, 146, 32567–32574.
70. J. Witten, I. Raji, R. S. Manan, E. Beyer, S. Bartlett, Y. Tang, M. Ebadi, J. Lei, D. Nguyen, F. Oladimeji, A. Y. Jiang, E. MacDonald, Y. Hu, H. Mughal, A. Self, E. Collins, Z. Yan, J. F. Engelhardt, R. Langer and D. G. Anderson, Nat. Biotechnol., 2024, 42, 1–10.
71. Y. Xu, S. Ma, H. Cui, J. Chen, S. Xu, F. Gong, A. Golubovic, M. Zhou, K. C. Wang, A. Varley, R. X. Z. Lu, B. Wang and B. Li, Nat. Commun., 2024, 15, 6305.
72. W. Wang, K. Chen, T. Jiang, Y. Wu, Z. Wu, H. Ying, H. Yu, J. Lu, J. Lin and D. Ouyang, Nat. Commun., 2024, 15, 10804.
73. S.-H. Bae, H. Choi, J. Lee, M.-H. Kang, S.-H. Ahn, Y.-S. Lee, H. Choi, S. Jo, Y. Lee, H.-J. Park, S. Lee, S. Yoon, G. Roh, S. Cho, Y. Cho, D. Ha, S.-Y. Lee, E.-J. Choi, A. Oh, J. Kim, S. Lee, J. Hong, N. Lee, M. Lee, J. Park, D.-H. Jeong, K. Lee and J.-H. Nam, Small, 2025, 21, 2405618.
74. R. Maharjan, K. H. Kim, K. Lee, H.-K. Han and S. H. Jeong, J. Pharm. Anal., 2024, 14, 100996.
75. X. Wang, S. Liu, Y. Sun, X. Yu, S. M. Lee, Q. Cheng, T. Wei, J. Gong, J. Robinson, D. Zhang, X. Lian, P. Basak and D. J. Siegwart, Nat. Protoc., 2023, 18, 265–291.
76. S. A. Dilliard, Q. Cheng and D. J. Siegwart, Proc. Natl. Acad. Sci. U. S. A., 2021, 118, e2109256118.
77. T. Wei, Y. Sun, Q. Cheng, S. Chatterjee, Z. Traylor, L. T. Johnson, M. L. Coquelin, J. Wang, M. J. Torres, X. Lian, X. Wang, Y. Xiao, C. A. Hodges and D. J. Siegwart, Nat. Commun., 2023, 14, 7322.
78. Y. Sun, S. Chatterjee, X. Lian, Z. Traylor, S. R. Sattiraju, Y. Xiao, S. A. Dilliard, Y.-C. Sung, M. Kim, S. M. Lee, S. Moore, X. Wang, D. Zhang, S. Wu, P. Basak, J. Wang, J. Liu, R. J. Mann, D. F. LePage, W. Jiang, S. Abid, M. Hennig, A. Martinez, B. A. Wustman, D. J. Lockhart, R. Jain, R. A. Conlon, M. L. Drumm, C. A. Hodges and D. J. Siegwart, Science, 2024, 384, 1196–1202.
79. E. Álvarez-Benedicto, Z. Tian, S. Chatterjee, D. Orlando, M. Kim, E. D. Guerrero, X. Wang and D. J. Siegwart, Angew. Chem., Int. Ed., 2023, 62, e202310395.
80. X. Lian, S. Chatterjee, Y. Sun, S. A. Dilliard, S. Moore, Y. Xiao, X. Bian, K. Yamada, Y.-C. Sung, R. M. Levine, K. Mayberry, S. John, X. Liu, C. Smith, L. T. Johnson, X. Wang, C. C. Zhang, D. R. Liu, G. A. Newby, M. J. Weiss, J. S. Yen and D. J. Siegwart, Nat. Nanotechnol., 2024, 19, 1409–1417.
81. A. Vaidya, S. Moore, S. Chatterjee, E. Guerrero, M. Kim, L. Farbiak, S. A. Dilliard and D. J. Siegwart, Adv. Mater., 2024, 36, 2313791.
82. K. Su, L. Shi, T. Sheng, X. Yan, L. Lin, C. Meng, S. Wu, Y. Chen, Y. Zhang, C. Wang, Z. Wang, J. Qiu, J. Zhao, T. Xu, Y. Ping, Z. Gu and S. Liu, Nat. Commun., 2024, 15, 5659.
83. Y. Fei, X. Yu, P. Liu, H. Ren, T. Wei and Q. Cheng, Adv. Mater., 2024, 36, 2409812.
84. L. Breda, T. E. Papp, M. P. Triebwasser, A. Yadegari, M. T. Fedorky, N. Tanaka, O. Abdulmalik, G. Pavani, Y. Wang, S. A. Grupp, S. Chou, H. Ni, B. L. Mui, Y. K. Tam, D. Weissman, S. Rivella and H. Parhiz, Science, 2023, 381, 436–443.
85. D. Shi, S. Toyonaga and D. G. Anderson, Nano Lett., 2023, 23, 2938–2944.
86. J. G. Rurik, I. Tombácz, A. Yadegari, P. O. Méndez Fernández, S. V. Shewale, L. Li, T. Kimura, O. Y. Soliman, T. E. Papp, Y. K. Tam, B. L. Mui, S. M. Albelda, E. Puré, C. H. June, H. Aghajanian, D. Weissman, H. Parhiz and J. A. Epstein, Science, 2022, 375, 91–96.
87. M. M. Billingsley, N. Gong, A. J. Mukalel, A. S. Thatte, R. El-Mayta, S. K. Patel, A. E. Metzloff, K. L. Swingle, X. Han, L. Xue, A. G. Hamilton, H. C. Safford, M. Alameh, T. E. Papp, H. Parhiz, D. Weissman and M. J. Mitchell, Small, 2024, 20, 2304378.
88. I. Tombácz, D. Laczkó, H. Shahnawaz, H. Muramatsu, A. Natesan, A. Yadegari, T. E. Papp, M.-G. Alameh, V. Shuvaev, B. L. Mui, Y. K. Tam, V. Muzykantov, N. Pardi, D. Weissman and H. Parhiz, Mol. Ther., 2021, 29, 3293–3304.
89. A. Kheirolomoom, A. J. Kare, E. S. Ingham, R. Paulmurugan, E. R. Robinson, M. Baikoghli, M. Inayathullah, J. W. Seo, J. Wang, B. Z. Fite, B. Wu, S. K. Tumbale, M. N. Raie, R. H. Cheng, L. Nichols, A. D. Borowsky and K. W. Ferrara, Biomaterials, 2022, 281, 121339.
90. H. Parhiz, V. V. Shuvaev, N. Pardi, M. Khoshnejad, R. Y. Kiseleva, J. S. Brenner, T. Uhler, S. Tuyishime, B. L. Mui, Y. K. Tam, T. D. Madden, M. J. Hope, D. Weissman and V. R. Muzykantov, J. Controlled Release, 2018, 291, 106–115.
91. Q. Li, C. Chan, N. Peterson, R. N. Hanna, A. Alfaro, K. L. Allen, H. Wu, W. F. Dall’Acqua, M. J. Borrok and J. L. Santos, ACS Chem. Biol., 2020, 15, 830–836.
92. H. C. Geisler, A. A. Ghalsasi, H. C. Safford, K. L. Swingle, A. S. Thatte, A. J. Mukalel, N. Gong, A. G. Hamilton, E. L. Han, B. E. Nachod, M. S. Padilla and M. J. Mitchell, J. Controlled Release, 2024, 371, 455–469.
93. E. L. Han, S. Tang, D. Kim, A. M. Murray, K. L. Swingle, A. G. Hamilton, K. Mrksich, M. S. Padilla, R. Palanki, J. J. Li and M. J. Mitchell, Nano Lett., 2025, 25, 800–810.
94. Y. Kim, J. Choi, E. H. Kim, W. Park, H. Jang, Y. Jang, S.-G. Chi, D.-H. Kweon, K. Lee, S. H. Kim and Y. Yang, Adv. Sci., 2024, 11, e2309917.
95. Z. R. Cohen, S. Ramishetti, N. Peshes-Yaloz, M. Goldsmith, A. Wohl, Z. Zibly and D. Peer, ACS Nano, 2015, 9, 1581–1591.
96. Y. Zhang, X. Cao, G. Hu, R. Ye, L. Zhang and J. Song, Adv. Healthcare Mater., 2024, 13, 2401376.
97. R. Peng, Q. Huang, L. Wang, G. Qiao, X. Huang, J. Jiang and X. Chu, Angew. Chem., Int. Ed., 2024, 63, e202402715.
98. A. Y. Jiang, J. Witten, I. O. Raji, F. Eweje, C. MacIsaac, S. Meng, F. A. Oladimeji, Y. Hu, R. S. Manan, R. Langer and D. G. Anderson, Nat. Nanotechnol., 2024, 19, 364–375.
99. S. Liu, Y. Wen, X. Shan, X. Ma, C. Yang, X. Cheng, Y. Zhao, J. Li, S. Mi, H. Huo, W. Li, Z. Jiang, Y. Li, J. Lin, L. Miao and X. Lu, Nat. Commun., 2024, 15, 9471.
100. M. P. Lokugamage, D. Vanover, J. Beyersdorf, M. Z. C. Hatit, L. Rotolo, E. S. Echeverri, H. E. Peck, H. Ni, J.-K. Yoon, Y. Kim, P. J. Santangelo and J. E. Dahlman, Nat. Biomed. Eng., 2021, 5, 1059–1068.
101. Y. Wang, J. Zhang, Y. Liu, X. Yue, K. Han, Z. Kong, Y. Dong, Z. Yang, Z. Fu, C. Tang, C. Shi, X. Zhao, M. Han, Z. Wang, Y. Zhang, C. Chen, A. Li, P. Sun, D. Zhu, K. Zhao and X. Jiang, Sci. Adv., 2024, 10, eado4791.
102. C. Wang, J. Xiao, X. Hu, Q. Liu, Y. Zheng, Z. Kang, D. Guo, L. Shi and Y. Liu, Adv. Healthcare Mater., 2023, 12, 2201889.
103. N. Maniyamgama, K. H. Bae, Z. W. Chang, J. Lee, M. J. Y. Ang, Y. J. Tan, L. F. P. Ng, L. Renia, K. P. White and Y. Y. Yang, Adv. Sci., 2025, 12, 2407383.
104. J. R. Melamed, S. S. Yerneni, M. L. Arral, S. T. LoPresti, N. Chaudhary, A. Sehrawat, H. Muramatsu, M.-G. Alameh, N. Pardi, D. Weissman, G. K. Gittes and K. A. Whitehead, Sci. Adv., 2023, 9, eade1444.
105. W. Dowell, J. Dearborn, S. Languon, Z. Miller, T. Kirch, S. Paige, O. Garvin, L. Kjendal, E. Harby, A. B. Zuchowski, E. Clark, C. Lescieur-Garcia, J. Vix, A. Schumer, S. K. Mistri, D. B. Snoke, A. L. Doiron, K. Freeman, M. J. Toth, M. E. Poynter, J. E. Boyson and D. Majumdar, ACS Nano, 2024, 18, 33058–33072.
106. J. Chen, Z. Ye, C. Huang, M. Qiu, D. Song, Y. Li and Q. Xu, Proc. Natl. Acad. Sci. U. S. A., 2022, 119, e2207841119.
107. S. Patel, R. C. Ryals, K. K. Weller, M. E. Pennesi and G. Sahay, J. Controlled Release, 2019, 303, 91–100.
108. J. Devoldere, K. Peynshaert, H. Dewitte, C. Vanhove, L. De Groef, L. Moons, S. Y. Özcan, D. Dalkara, S. C. De Smedt and K. Remaut, J. Controlled Release, 2019, 307, 315–330.
109. D. Loughrey and J. E. Dahlman, Acc. Chem. Res., 2022, 55, 13–23.
110. A. E. I. Hamouda, J. Filtjens, E. Brabants, D. Kancheva, A. Debraekeleer, J. Brughmans, L. Jacobs, P. M. R. Bardet, E. Knetemann, P. Lefesvre, L. Allonsius, M. Gontsarik, I. Varela, M. Crabbé, E. J. Clappaert, F. Cappellesso, A. A. Caro, A. Gordún Peiró, L. Fredericq, E. Hadadi, M. Estapé Senti, R. Schiffelers, L. A. van Grunsven, F. Aboubakar Nana, B. G. De Geest, S. Deschoemaeker, S. De Koker, F. Lambolez and D. Laoui, Nat. Commun., 2024, 15, 10635.
111. I. C. Turnbull, A. A. Eltoukhy, K. M. Fish, M. Nonnenmacher, K. Ishikawa, J. Chen, R. J. Hajjar, D. G. Anderson and K. D. Costa, Mol. Ther., 2016, 24, 66–75.
112. J. E. Dahlman, K. J. Kauffman, Y. Xing, T. E. Shaw, F. F. Mir, C. C. Dlott, R. Langer, D. G. Anderson and E. T. Wang, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 2060–2065.
113. L. H. Rhym, R. S. Manan, A. Koller, G. Stephanie and D. G. Anderson, Nat. Biomed. Eng., 2023, 7, 901–910.
114. O. S. Fenton, K. J. Kauffman, J. C. Kaczmarek, R. L. McClellan, S. Jhunjhunwala, M. W. Tibbitt, M. D. Zeng, E. A. Appel, J. R. Dorkin, F. F. Mir, J. H. Yang, M. A. Oberli, M. W. Heartlein, F. DeRosa, R. Langer and D. G. Anderson, Adv. Mater., 2017, 29, 1606944.
115. O. S. Fenton, K. J. Kauffman, R. L. McClellan, J. C. Kaczmarek, M. D. Zeng, J. L. Andresen, L. H. Rhym, M. W. Heartlein, F. DeRosa and D. G. Anderson, Angew. Chem., Int. Ed., 2018, 57, 13582–13586.
116. C. D. Sago, M. P. Lokugamage, K. Paunovska, D. A. Vanover, C. M. Monaco, N. N. Shah, M. Gamboa Castro, S. E. Anderson, T. G. Rudoltz, G. N. Lando, P. Munnilal Tiwari, J. L. Kirschman, N. Willett, Y. C. Jang, P. J. Santangelo, A. V. Bryksin and J. E. Dahlman, Proc. Natl. Acad. Sci. U. S. A., 2018, 115, 9944–9952.
117. S. G. Huayamares, M. P. Lokugamage, R. Rab, A. J. Da Silva Sanchez, H. Kim, A. Radmand, D. Loughrey, L. Lian, Y. Hou, B. R. Achyut, A. Ehrhardt, J. S. Hong, C. D. Sago, K. Paunovska, E. S. Echeverri, D. Vanover, P. J. Santangelo, E. J. Sorscher and J. E. Dahlman, J. Controlled Release, 2023, 357, 394–403.
118. C. Dobrowolski, K. Paunovska, E. Schrader Echeverri, D. Loughrey, A. J. Da Silva Sanchez, H. Ni, M. Z. C. Hatit, M. P. Lokugamage, Y. Kuzminich, H. E. Peck, P. J. Santangelo and J. E. Dahlman, Nat. Nanotechnol., 2022, 17, 871–879.
119. H. Kim, R. Zenhausern, K. Gentry, L. Lian, S. G. Huayamares, A. Radmand, D. Loughrey, A. R. Podilapu, M. Z. C. Hatit, H. Ni, A. Li, A. Shajii, H. E. Peck, K. Han, X. Hua, S. Jia, M. Martinez, C. Lee, P. J. Santangelo, A. Tarantal and J. E. Dahlman, Nat. Biotechnol., 2024, 42, 1–8.
120. I. Soerjomataram and F. Bray, Nat. Rev. Clin. Oncol., 2021, 18, 663–672.
121. C. Liu, Q. Shi, X. Huang, S. Koo, N. Kong and W. Tao, Nat. Rev. Cancer, 2023, 23, 526–543.
122. C. P. Arevalo, M. J. Bolton, V. Le Sage, N. Ye, C. Furey, H. Muramatsu, M.-G. Alameh, N. Pardi, E. M. Drapeau, K. Parkhouse, T. Garretson, J. S. Morris, L. H. Moncla, Y. K. Tam, S. H. Y. Fan, S. S. Lakdawala, D. Weissman and S. E. Hensley, Science, 2022, 378, 899–904.
123. Y. Zhang and Z. Zhang, Cell. Mol. Immunol., 2020, 17, 807–821.
124. C. Chong, G. Coukos and M. Bassani-Sternberg, Nat. Biotechnol., 2022, 40, 175–188.
125. L.-J. Duan, Q. Wang, C. Zhang, D.-X. Yang and X.-Y. Zhang, Front. Immunol., 2022, 13, 923647.
126. M. A. Islam, J. Rice, E. Reesor, H. Zope, W. Tao, M. Lim, J. Ding, Y. Chen, D. Aduluso, B. R. Zetter, O. C. Farokhzad and J. Shi, Biomaterials, 2021, 266, 120431.
127. W. Xiao, F. Wang, Y. Gu, X. He, N. Fan, Q. Zheng, S. Qin, Z. He, Y. Wei and X. Song, Chin. Chem. Lett., 2024, 35, 108755.
128. J. Fu, S. Wu, N. Bao, L. Wu, H. Qu, Z. Wang, H. Dong, J. Wu and Y. Jin, Adv. Sci., 2025, 12, 2401287.
129. L. A. Rojas, Z. Sethna, K. C. Soares, C. Olcese, N. Pang, E. Patterson, J. Lihm, N. Ceglia, P. Guasp, A. Chu, R. Yu, A. K. Chandra, T. Waters, J. Ruan, M. Amisaki, A. Zebboudj, Z. Odgerel, G. Payne, E. Derhovanessian, F. Müller, I. Rhee, M. Yadav, A. Dobrin, M. Sadelain, M. Łuksza, N. Cohen, L. Tang, O. Basturk, M. Gönen, S. Katz, R. K. Do, A. S. Epstein, P. Momtaz, W. Park, R. Sugarman, A. M. Varghese, E. Won, A. Desai, A. C. Wei, M. I. D’Angelica, T. P. Kingham, I. Mellman, T. Merghoub, J. D. Wolchok, U. Sahin, Ö. Türeci, B. D. Greenbaum, W. R. Jarnagin, J. Drebin, E. M. O’Reilly and V. P. Balachandran, Nature, 2023, 618, 144–150.
130. Y. Li, X. Ma, Y. Yue, K. Zhang, K. Cheng, Q. Feng, N. Ma, J. Liang, T. Zhang, L. Zhang, Z. Chen, X. Wang, L. Ren, X. Zhao and G. Nie, Adv. Mater., 2022, 34, e2109984.
131. S. Chandra, J. C. Wilson, D. Good and M. Q. Wei, Oncol. Res., 2024, 32, 1543–1564.
132. A. V. Yaremenko, M. M. Khan, X. Zhen, Y. Tang and W. Tao, Med, 2025, 6, 100562.
133. C. Galassi, T. A. Chan, I. Vitale and L. Galluzzi, Cancer Cell, 2024, 42, 1825–1863.
134. A. J. Barbier, A. Y. Jiang, P. Zhang, R. Wooster and D. G. Anderson, Nat. Biotechnol., 2022, 40, 840–854.
135. K. Pan, H. Farrukh, V. C. S. R. Chittepu, H. Xu, C. Pan and Z. Zhu, J. Exp. Clin. Cancer Res., 2022, 41, 119.
136. D. J. Baker, Z. Arany, J. A. Baur, J. A. Epstein and C. H. June, Nature, 2023, 619, 707–715.
137. L. Wang, Y. Lv, L. Zhou, S. Wu, Y. Zhu, S. Fu, S. Ding, R. Hong, M. Zhang, H. Yu, A. H. Chang, G. Wei, Y. Hu and H. Huang, Exp. Hematol. Oncol., 2024, 13, 28.
138. H. Murthy, M. Iqbal, J. C. Chavez and M. A. Kharfan-Dabaja, ImmunoTargets Ther., 2019, 8, 43–52.
139. K. Ai, B. Liu, X. Chen, C. Huang, L. Yang, W. Zhang, J. Weng, X. Du, K. Wu and P. Lai, J. Hematol. Oncol., 2024, 17, 105.
140. J. Yang, Y. Chen, Y. Jing, M. R. Green and L. Han, Nat. Rev. Clin. Oncol., 2023, 20, 211–228.
141. K. Xiao, Y. Lai, W. Yuan, S. Li, X. Liu, Z. Xiao and H. Xiao, Interdiscip. Med., 2024, 2, e20230036.
142. R. Kitte, M. Rabel, R. Geczy, S. Park, S. Fricke, U. Koehl and U. S. Tretbar, Mol. Ther.–Methods Clin. Dev., 2023, 31, 101139.
143. W. Wang, S. He, W. Zhang, H. Zhang, V. M. DeStefano, M. Wada, K. Pinz, G. Deener, D. Shah, N. Hagag, M. Wang, M. Hong, R. Zeng, T. Lan, Y. Ma, F. Li, Y. Liang, Z. Guo, C. Zou, M. Wang, L. Ding, Y. Ma and Y. Yuan, Ann. Rheum. Dis., 2024, 83, 1304–1314.
144. H. Lin, X. Yang, S. Ye, L. Huang and W. Mu, Biomed. Pharmacother., 2024, 178, 117252.
145. M. C. Milone, J. Xu, S.-J. Chen, M. A. Collins, J. Zhou, D. J. Powell and J. J. Melenhorst, Nat. Cancer, 2021, 2, 780–793.
146. F. Nasiri, P. Safarzadeh Kozani, F. Salem, M. Mahboubi Kancha, S. Dashti Shokoohi and P. Safarzadeh Kozani, Cancer Cell Int., 2025, 25, 64.
147. A. M. Erasha, H. EL-Gendy, A. S. Aly, M. Fernández-Ortiz and R. K. A. Sayed, Int. J. Mol. Sci., 2025, 26, 2716.
148. M. Li, Y. Xie, J. Zhang, X. Zhou, L. Gao, M. He, X. Liu, X. Miao, Y. Liu, R. Cao, Y. Jia, Z. Zeng and L. Liu, Cancer Lett., 2024, 598, 217111.
149. S. Kong, J. Zhang, L. Wang, W. Li, H. Guo, Q. Weng, Q. He, H. Lou, L. Ding and B. Yang, Cancer Lett., 2025, 611, 217432.
150. X. Lin, Y. Sun, X. Dong, Z. Liu, R. Sugimura and G. Xie, Biomed. Pharmacother., 2023, 165, 115123.
151. F. Liu, X. Miao, L. Han and X. Song, Front. Oncol., 2024, 14, 1390006.
152. P. Y. Lam, N. Omer, J. K. M. Wong, C. Tu, L. Alim, G. R. Rossi, M. Victorova, H. Tompkins, C.-Y. Lin, A. M. Mehdi, A. Choo, M. R. Elliott, E. Coleborn, J. Sun, T. Mercer, O. Vittorio, L. J. Dobson, A. D. McLellan, A. Brooks, Z. K. Tuong, S. W. Cheetham, W. Nicholls and F. Souza-Fonseca-Guimaraes, Clin. Transl. Med., 2025, 15, e70140.
153. H. E. Shin, J. Han, J. D. Park, M. Park, J. Han, M. Kang, J. S. Lee, C. G. Park, J. Park, H. Kim, D. Cho and W. Park, Adv. Funct. Mater., 2024, 34, 2315721.
154. K. M. Maalej, M. Merhi, V. P. Inchakalody, S. Mestiri, M. Alam, C. Maccalli, H. Cherif, S. Uddin, M. Steinhoff, F. M. Marincola and S. Dermime, Mol. Cancer, 2023, 22, 20.
155. Z. Yang, Y. Liu, K. Zhao, W. Jing, L. Gao, X. Dong, Y. Wang, M. Han, C. Shi, C. Tang, P. Sun, R. Zhang, Z. Fu, J. Zhang, D. Zhu, C. Chen and X. Jiang, J. Controlled Release, 2023, 360, 718–733.
156. K. A. Reiss, M. G. Angelos, E. C. Dees, Y. Yuan, N. T. Ueno, P. R. Pohlmann, M. L. Johnson, J. Chao, O. Shestova, J. S. Serody, M. Schmierer, M. Kremp, M. Ball, R. Qureshi, B. H. Schott, P. Sonawane, S. C. DeLong, M. Christiano, R. F. Swaby, S. Abramson, K. Locke, D. Barton, E. Kennedy, S. Gill, D. Cushing, M. Klichinsky, T. Condamine and Y. Abdou, Nat. Med., 2025, 31, 1171–1182.
157. L. Chen, S. Liu and Y. Tao, Signal Transduction Targeted Ther., 2020, 5, 90.
158. Y. Liu, Z. Su, O. Tavana and W. Gu, Cancer Cell, 2024, 42, 946–967.
159. E. Cerami, J. Gao, U. Dogrusoz, B. E. Gross, S. O. Sumer, B. A. Aksoy, A. Jacobsen, C. J. Byrne, M. L. Heuer, E. Larsson, Y. Antipin, B. Reva, A. P. Goldberg, C. Sander and N. Schultz, Cancer Discovery, 2012, 2, 401–404.
160. N. Kong, W. Tao, X. Ling, J. Wang, Y. Xiao, S. Shi, X. Ji, A. Shajii, S. T. Gan, N. Y. Kim, D. G. Duda, T. Xie, O. C. Farokhzad and J. Shi, Sci. Transl. Med., 2019, 11, eaaw1565.
161. Y. Xiao, J. Chen, H. Zhou, X. Zeng, Z. Ruan, Z. Pu, X. Jiang, A. Matsui, L. Zhu, Z. Amoozgar, D. S. Chen, X. Han, D. G. Duda and J. Shi, Nat. Commun., 2022, 13, 758.
162. E. Sellars, M. Gabra and L. Salmena, Cold Spring Harbor Perspect. Med., 2020, 10, a036236.
163. M. A. Islam, Y. Xu, W. Tao, J. M. Ubellacker, M. Lim, D. Aum, G. Y. Lee, K. Zhou, H. Zope, M. Yu, W. Cao, J. T. Oswald, M. Dinarvand, M. Mahmoudi, R. Langer, P. W. Kantoff, O. C. Farokhzad, B. R. Zetter and J. Shi, Nat. Biomed. Eng., 2018, 2, 850–864.
164. Y. Hu, X. He, P. Chen, X.-L. Tian, R. Wang, X. Song, X.-Q. Yu and J. Zhang, Acta Biomater., 2025, 194, 442–454.
165. S. de Picciotto, N. DeVita, C. J. Hsiao, C. Honan, S.-W. Tse, M. Nguyen, J. D. Ferrari, W. Zheng, B. T. Wipke and E. Huang, Nat. Commun., 2022, 13, 3866.
166. B. Wang, M. Tang, Q. Chen, W. Ho, Y. Teng, X. Xiong, Z. Jia, X. Li, X. Xu and X.-Q. Zhang, ACS Nano, 2024, 18, 15499–15516.
167. W. Li, X. Zhang, C. Zhang, J. Yan, X. Hou, S. Du, C. Zeng, W. Zhao, B. Deng, D. W. McComb, Y. Zhang, D. D. Kang, J. Li, W. E. Carson and Y. Dong, Nat. Commun., 2021, 12, 7264.
168. K. Ramalingam, R. Woody, A. Glencer, C. J. Schwartz, H. Mori, J. Wong, G. Hirst, J. Rosenbluth, N. Onishi, J. Gibbs, N. Hylton, A. D. Borowsky, M. Campbell and L. J. Esserman, JAMA Oncol., 2025, 11, 288.
169. T. Marron, V. Subbiah, O. Hamid, S. Goel, M. D. Hellmann, B. A. Carneiro, S. P. Pate, E. De Vries, N. Luheshi, O. Hamid and A. G. Hernández, Ann. Oncol., 2020, 31, S730.
170. J. Han, J. Lim, C.-P. J. Wang, J.-H. Han, H. E. Shin, S.-N. Kim, D. Jeong, S. H. Lee, B.-H. Chun, C. G. Park and W. Park, Nano Convergence, 2023, 10, 36.
171. F. Zhou, L. Huang, S. Li, W. Yang, F. Chen, Z. Cai, X. Liu, W. Xu, V.-P. Lehto, U. Lächelt, R. Huang, Y. Shi, T. Lammers, W. Tao, Z. P. Xu, E. Wagner, Z. Xu and H. Yu, Exploration, 2024, 4, 20210146.
172. S. Fernandes, M. Cassani, F. Cavalieri, G. Forte and F. Caruso, Adv. Sci., 2024, 11, 2305769.
173. W. Chen, Y. Zhu, J. He and X. Sun, Theranostics, 2024, 14, 96–115.
174. X. Pan, Y.-W.-Q. Zhang, C. Dai, J. Zhang, M. Zhang and X. Chen, Int. J. Nanomed., 2025, 20, 3339–3361.
175. C. Jin, Y. Zhang, B. Li, T. Gao, B. Wang and P. Hua, Mater. Today Bio, 2024, 27, 101136.
176. K. Qiu, X. Duan, M. Mao, Y. Song, Y. Rao, D. Cheng, L. Feng, X. Shao, C. Jiang, H. Huang, Y. Wang, H. Li, X. Chen, S. Wu, D. Luo, F. Chen, X. Peng, Y. Zheng, H. Wang, J. Liu, Y. Zhao, X. Song and J. Ren, npj Vaccines, 2023, 8, 144.
177. H. Zhou, Y. Liao, X. Han, D. S. Chen, X. Hong, K. Zhou, X. Jiang, Y. Xiao and J. Shi, Nano Lett., 2023, 23, 3661–3668.
178. R. Fu, R. Qi, H. Xiong, X. Lei, Y. Jiang, J. He, F. Chen, L. Zhang, D. Qiu, Y. Chen, M. Nie, X. Guo, Y. Zhu, J. Zhang, M. Yue, J. Cao, G. Wang, Y. Que, M. Fang, Y. Wang, Y. Chen, T. Cheng, S. Ge, J. Zhang, Q. Yuan, T. Zhang and N. Xia, Signal Transduction Targeted Ther., 2024, 9, 118.
179. J. S. Weber, M. S. Carlino, A. Khattak, T. Meniawy, G. Ansstas, M. H. Taylor, K. B. Kim, M. McKean, G. V. Long, R. J. Sullivan, M. Faries, T. T. Tran, C. L. Cowey, A. Pecora, M. Shaheen, J. Segar, T. Medina, V. Atkinson, G. T. Gibney, J. J. Luke, S. Thomas, E. I. Buchbinder, J. A. Healy, M. Huang, M. Morrissey, I. Feldman, V. Sehgal, C. Robert-Tissot, P. Hou, L. Zhu, M. Brown, P. Aanur, R. S. Meehan and T. Zaks, Lancet, 2024, 403, 632–644.
180. X. Wang, F. Li, J. Zhang, L. Guo, M. Shang, X. Sun, S. Xiao, D. Shi, D. Meng, Y. Zhao, C. Jiang and J. Li, J. Controlled Release, 2024, 367, 45–60.
181. X. Wang, W. Wang, S. Zou, Z. Xu, D. Cao, S. Zhang, M. Wei, Q. Zhan, C. Wen, F. Li, H. Chen, D. Fu, L. Jiang, M. Zhao and B. Shen, Cell Res., 2024, 34, 661–664.
182. D. Zhang, G. Wang, X. Yu, T. Wei, L. Farbiak, L. T. Johnson, A. M. Taylor, J. Xu, Y. Hong, H. Zhu and D. J. Siegwart, Nat. Nanotechnol., 2022, 17, 777–787.
183. D. Rosenblum, A. Gutkin, R. Kedmi, S. Ramishetti, N. Veiga, A. M. Jacobi, M. S. Schubert, D. Friedmann-Morvinski, Z. R. Cohen, M. A. Behlke, J. Lieberman and D. Peer, Sci. Adv., 2020, 6, eabc9450.
184. J. D. Beck, M. Diken, M. Suchan, M. Streuber, E. Diken, L. Kolb, L. Allnoch, F. Vascotto, D. Peters, T. Beißert, Ö. Akilli-Öztürk, Ö. Türeci, S. Kreiter, M. Vormehr and U. Sahin, Cancer Cell, 2024, 42, 568–582.e11.
185. N. D. Le, B. L. Nguyen, B. R. Patil, H. Chun, S. Kim, T. O. O. Nguyen, S. Mishra, S. Tandukar, J.-H. Chang, D. Y. Kim, S. G. Jin, H.-G. Choi, S. K. Ku, J. Kim and J. O. Kim, ACS Nano, 2024, 18, 8392–8410.
186. Y. Rybakova, P. S. Kowalski, Y. Huang, J. T. Gonzalez, M. W. Heartlein, F. DeRosa, D. Delcassian and D. G. Anderson, Mol. Ther., 2019, 27, 1415–1423.
187. Y. Shi, M. Shi, Y. Wang and J. You, Signal Transduction Targeted Ther., 2024, 9, 322.
188. Y. Hong, Y. Liu, H. Shen, B. Li and Q. Li, J. Transl. Med., 2025, 23, 354.
189. N. V. Bathula, J. J. Friesen, I. C. Casmil, C. J. Wayne, S. Liao, S. K. V. Soriano, C. H. Ho, A. Strumpel and A. K. Blakney, J. Controlled Release, 2024, 374, 28–38.
190. J. Li, D. Liu, X. Li, J. Wei, W. Du, A. Zhao and M. Xu, Hum. Vaccines Immunother., 2025, 21, 2469333.
191. M. Peng, S. Zhang, P. Wu, X. Hou, D. Wang, J. Ge, H. Qu, C. Fan, Y. Zhou, B. Xiang, Q. Liao, M. Zhou, M. Tan, G. Li, W. Xiong, P. Chen, Z. Zeng and Z. Gong, Mol. Cancer, 2025, 24, 67.

---

Footnote

† These authors contributed equally.

---