During the coronavirus disease 2019 (COVID-19) pandemic, over-prescription and misuse of antibiotics became widespread, largely due to patients’ susceptibility to secondary infections and diagnostic uncertainty (Hsu, 2020). This excessive use of antibiotics during the pandemic is belived to have accelerated the emergence and dissemination of antimicrobial-resistant (AMR) bacteria (Rawson et al., 2020; Pelfrene et al., 2021). Indeed, antibiotic resistance has been described as a “silent pandemic” in itself (Mahoney et al., 2021). Although new antibiotics have been developed over recent decades to combat AMR pathogens, the rate at which resistance emerges has consistently outpaced the development of new antimicrobial agents (Årdal et al., 2020). The growing prevalence of antibiotic resistance thus represents one of the most pressing challenges in modern medicine (Christaki et al., 2020).

To address this global threat, alternative therapeutic strategies have been explored, including the use of nucleic acid–based agents. Delivery platforms such as cell-penetrating peptides (CPPs), extracellular vesicles (EVs), DNA nanostructure, and polymer-based nanoparticles (NPs) have been actively investigated as potential solutions (Figure 1a) (Xue et al., 2018; Moreira et al., 2024). Such platforms aim to overcome biological barriers and improve the intracellular delivery efficiency of therapeutic nucleic acids. However, achieving efficient and safe delivery of these molecules to target sites remains a formidable challenge (Wojciechowska et al., 2020; Bier and Nizet, 2021).

Nucleotide-based antibiotics represent a new class of programmable therapeutics that employ nucleic acids—such as antisense oligonucleotides, small RNAs, or CRISPR–Cas systems—to silence essential genes, inhibit virulence pathways, or eliminate antimicrobial resistance determinants in bacteria (Wan et al., 2021; Moreira et al., 2024). Unlike conventional small-molecule antibiotics, these agents act through sequence-specific base pairing or targeted gene editing, enabling highly precise bacterial gene modulation. However, their clinical potential has been limited by the absence of an efficient delivery system capable of transporting nucleic acids across the complex bacterial cell envelope. In this context, lipid nanoparticles (LNPs) offer a uniquely suitable platform: their modular lipid architecture enables encapsulation, protection, and controlled release of nucleic acid cargo (Figure 1b) (Cullis and Felgner, 2024), providing the essential technological bridge that links nucleotide-based antibiotics to practical antimicrobial applications. Thus, the development of LNPs directly underpins the feasibility and therapeutic utility of nucleotide antibiotics.

The delivery platform is classified into two types: encapsulation and transporter. In the case of encapsulation, nucleic acids are encapsulated in materials that can enter the cytosol, where nucleic acids are released [e.g., EVs, polymer-based NPs, LNPs]. Nucleic acids can be linked to transporter materials that can penetrate the bacterial membrane, carrying the cargo into the target cell [e.g., CPPs and DNA nanostructures] (Table 1).

EVs are lipid-bound vesicles secreted by cells into the extracellular surroundings. Intracellular nucleic acids are wrapped in the cell membrane before release. EVs are produced by encapsulation in the membranes of bacterial cells. Thus, the target cell membrane’s surface selectivity and lipid compatibility represent advantages. The decisive drawback is the difficulty in isolating EVs from cells. Ultracentrifugation and isolation were required to separate microbial EVs. EVs were lost during the isolation process, which reduced production efficiency (Onódi et al., 2018). For this reason, EVs were still being used at the laboratory scale (Bost et al., 2021).

Polymer-based NPs (micelles and dendrimers) are produced by the electrostatic self-assembly of polymers and nucleic acids. NPs can alter the properties of nanoparticles by modifying the molecular structure of the polymer shell (size, surface properties) (Makino et al., 2014). Size control was an advantage compared with EVs, but the control range was narrow (diameter 10−100 nm) (Tapeinos et al., 2017). The use of organic solvents for the preparation of nanoparticles was another disadvantage that can increase toxicity (Tapeinos et al., 2017).

CPPs are characterized by their ability to penetrate the bacterial membrane. CPPs were used as nucleic acid carriers, taking advantage of their membrane-penetrating properties (Xue et al., 2018). CPPs effectively delivered nucleic acids into the bacterial inner membrane under in vitro conditions (Wojciechowska et al., 2020). However, under in vivo conditions, practical applications were not possible due to the side effect of hemolysis (Wojciechowska et al., 2020).

DNA nanostructures are combinations of complementary gene sequences assembled into 2D and 3D nano-transporters (tetrahedron, octahedron, cube, etc.) (Xue et al., 2018). Nucleic acids can be delivered to bacteria in the form of nanostructures-linker-nucleic acids (Wojciechowska et al., 2020). DNA nanostructures are ideal transporters with many advantages. However, it was still only possible to produce them at the laboratory scale, and the unit price was unpredictable (Xue et al., 2018; Wojciechowska et al., 2020).

LNPs are fatty particles used to deliver therapeutic molecules like mRNA by protecting them from degradation and helping them enter eukaryotic cells (Figure 1c). LNPs have demonstrated remarkable promise as delivery vehicles, most notably during the COVID-19 pandemic, when mRNA encapsulated in LNPs was successfully utilized in vaccines. The benefits of LNP-based mRNA vaccines—such as stability, scalability, rapid development, and economic feasibility—have been well documented (Meo et al., 2021). Notably, the pivotal role of LNPs in mRNA technology was recognized through consecutive Nobel Prizes in Physiology or Medicine in 2023 and 2024, underscoring their significance as a transformative platform for future gene delivery (Karikó et al., 2005; Ambros and Ruvkun, 2018).

Building on these advancements, we established a foundation for exploring the potential application of LNPs in targeting bacterial systems (Kim et al., 2025). In this review, we propose novel avenues for applying LNPs to the treatment of AMR bacteria. By integrating physicochemical and biological design strategies, our aim is to optimize LNP formulations for efficient and targeted delivery to resistant bacterial strains. Ultimately, this approach seeks to enhance the therapeutic efficacy of nucleic acid–based agents and contribute to innovative solutions addressing the escalating challenge of antibiotic resistance. Continued investigation in this field may pave the way for next-generation antimicrobial therapies that safeguard global health.

LNPs have emerged as one of the most versatile and clinically validated platforms for nucleic acid delivery, as evidenced by their widespread use in mRNA vaccines and gene therapy applications (Swingle et al., 2021; Barbier et al., 2022). Their modular composition, physicochemical tunability, and manufacturing scalability position LNPs as a promising foundation for the development of nucleotide-based antimicrobial agents. In the context of bacterial systems, where membrane architecture and permeability significantly differ from those of eukaryotic cells, understanding the structural features and delivery mechanisms of LNPs is essential for rational design. This section provides an overview of the fundamental components of LNPs, the mechanistic basis of nucleic acid encapsulation and delivery, and the intrinsic advantages that make LNPs strong candidates for next-generation nucleotide antibiotics.

To effectively control bacterial pathogens, nucleic acids must be designed to target specific bacterial genes. These targets can be broadly classified into three categories: essential genes, virulence factors, and AMR genes. The first strategy involves inhibiting essential genes required for vital cellular processes such as cell wall synthesis, DNA replication, RNA transcription, and protein translation. Silencing these genes leads to bacterial death. However, such broad-spectrum antibacterial activity may also eliminate beneficial members of the human microbiota. To overcome this limitation, anti-virulence therapies have been proposed (Clatworthy et al., 2007). A more refined approach is to suppress bacterial virulence factors that enable pathogens to colonize, invade, and persist within host tissues. Nucleic acid–based strategies have been designed to interfere with biofilm formation, toxin–antitoxin systems, and quorum sensing networks (Da et al., 2017; Równicki et al., 2018; Zhang et al., 2020). The third approach focuses on targeting AMR genes. Rather than killing bacteria directly, gene-editing technologies such as the clustered regularly interspaced short palindromic repeats–CRISPR-associated protein (CRISPR–Cas) system can disable resistance determinants, thereby restoring bacterial susceptibility to antibiotics (Kang et al., 2017; Wan et al., 2021).

The choice of nucleic acid type is also crucial. Regulatory small RNAs (sRNAs) can modulate gene expression post-transcriptionally by base-pairing with target mRNAs through antisense interactions (Lalaouna et al., 2013; Hegarty and Stewart, 2018). This pairing either blocks ribosome access to the ribosome-binding site (RBS) or triggers degradation of the sRNA–mRNA duplex via RNase H, ultimately inhibiting translation of the target gene (Figure 3a). Alternatively, antisense oligonucleotides (ASOs), chemically modified DNA or RNA analogs, can achieve similar regulatory effects. Structural modifications such as phosphorothioate (PS), peptide nucleic acid (PNA), or locked nucleic acid (LNA) backbones enhance nuclease resistance and cellular stability during delivery into prokaryotic cells (Hegarty and Stewart, 2018; Xue et al., 2018). To suppress multidrug-resistant bacteria by gene disruption, CRISPR–Cas systems have been widely explored (Figure 3b). The bacterial RNA-guided endonuclease Cas9 cleaves genomic DNA at sequences complementary to the guide RNA (gRNA), resulting in targeted killing or removal of antibiotic resistance genes (Ran et al., 2013; Bier and Nizet, 2021). A catalytically inactive variant, dCas9, retains DNA-binding capability but lacks nuclease activity; it represses transcription by sterically blocking RNA polymerase at promoter regions; a mechanism known as CRISPR interference (CRISPRi) (Zuberi et al., 2017; Zhang et al., 2021a). Since many antimicrobial resistance (AMR) genes are plasmid-encoded, cleavage by Cas9 alone may be insufficient to eradicate bacteria carrying high-copy plasmids (Citorik et al., 2014). To address this, CRISPR–Cas13a has been introduced owing to its collateral single-stranded RNA (ssRNA) cleavage activity (Figure 3b) (Kiga et al., 2020). Once activated by recognition of AMR gene transcripts, Cas13a indiscriminately degrades surrounding RNAs, effectively killing bacteria that harbor resistance genes (Figure 3b).

Potential challenges in developing bacteria-targeting LNPs arise from the fundamental structural differences between bacterial and eukaryotic cell membranes (Figures 1d, e). Unlike eukaryotic membranes, bacterial membranes exhibit unique compositions and architectures (Sun et al., 2022). Consequently, it remains uncertain whether LNPs originally designed for eukaryotic cells can effectively deliver cargo into the bacterial cytosol. In Gram-negative bacteria, the presence of lipopolysaccharides (LPS) on the outer membrane (Figure 1d) can physically hinder LNP permeation. Another major obstacle is the inherently low delivery efficiency associated with the double-layered membrane structure of these bacteria. The distinct organization of bacterial membranes is believed to influence their mechanical strength and elasticity, further complicating nanoparticle penetration (Jauffred et al., 2007; Sun et al., 2022). However, how LNPs interact with and modulate the mechanical properties of the bacterial double membrane remains poorly understood.

Additional barriers arise from the thick peptidoglycan layer, particularly in Gram-positive bacteria, where the cell wall is substantially thicker than that of Gram-negative species (Figure 1d). Beyond these physical impediments, challenges also stem from the chemical composition of LNP lipids themselves (Chaudhary et al., 2021). The endosomal escape mechanism (Figure 1c)—in which LNP lipids fuse with cellular membranes to release encapsulated nucleic acids—is governed by the molecular structures of ionizable and helper lipids (Csiszár et al., 2015; Bost et al., 2021; Chaudhary et al., 2021). Under acidic conditions, the fusogenic properties of these lipids become critical, enabling the release of nucleic acids through membrane fusion with the target bacterial envelope (Csiszár et al., 2015).

The components used to encapsulate nucleic acids in LNPs include ionizable lipids, helper lipids, cholesterol, PEG–lipids, and targeting moieties (Figure 1b). Ionizable lipids electrostatically self-assemble with negatively charged nucleic acids under acidic conditions, forming the core structure of LNPs (Chaudhary et al., 2021). This self-assembly process occurs almost instantaneously within microfluidic mixing systems. In typical formulations, lipids and cholesterol are dissolved in ethanol, whereas RNA or DNA intended for bacterial delivery is dispersed in an acidic aqueous phase (Figure 1b). When these two fractions are combined using microfluidic mixing, spontaneous self-assembly takes place, leading to the rapid formation of LNPs (Shepherd et al., 2021). Finally, targeting moieties are conjugated to the LNP surface to enable specific interactions with bacterial cells (Chaudhary et al., 2021).

This review highlights the emerging concept of nucleotide-based antibiotics delivered by lipid nanoparticles, a strategy that integrates programmable gene-targeting therapeutics with a clinically validated delivery platform. However, their successful commercialization critically depends on the optimization of efficient delivery systems capable of overcoming the unique structural and physicochemical barriers of bacterial cells. While we have demonstrated the feasibility of using LNPs for delivering nucleic acids into bacteria (Kim et al., 2025), several unresolved challenges hinder broader application. These include the limited understanding of LNP–bacterial membrane interactions, the absence of bacterial membrane–specific lipid design principles, variability in delivery efficiency across species, and the need for scalable manufacturing processes that ensure consistent particle properties suitable for in vivo use.

Future progress will require a mechanistic framework explaining how LNPs traverse bacterial outer membranes, peptidoglycan layers, and cytoplasmic membranes. Integrating molecular dynamics simulations, lipid–membrane interaction assays, and high-resolution structural analyses may accelerate the discovery of lipid motifs with enhanced fusogenicity tailored to bacterial systems. In parallel, constructing bacteria-specific lipid libraries informed by the chemical diversity of bacterial envelope components—such as lipid A or teichoic acids—could provide a foundation for rational LNP design. Strategies incorporating targeting moieties or helper molecules may further improve delivery efficiency, particularly for species with robust outer membranes or thick cell walls.

From an engineering perspective, advances in microfluidic manufacturing and rheology-informed process control will be essential for producing LNPs with the particle sizes, homogeneity, and stability required for therapeutic deployment. Simulation-based mixing optimization, automated pressure stabilization, and high-throughput formulation screening may further support scalable and reproducible production. As these technologies evolve, comprehensive in vivo evaluations—including toxicity, biodistribution, pharmacokinetics, and microbiome impact—will be critical for safe clinical application. In addition, evaluating long-term safety including potential immunogenicity, lipid accumulation, and microbiome alterations will be essential, particularly for repeated or chronic administration.

Looking ahead, nucleotide-LNP technologies are expected to extend far beyond antimicrobial applications. CRISPR-based tools was used to modulate gut microbiota and enhance the beneficial functions of commensal bacteria (Hidalgo-Cantabrana et al., 2017). Building upon these advances, LNPs carrying engineered CRISPR–Cas systems could be designed not only to suppress pathogenic traits but also to fine-tune bacterial gene expression or endow microbes with new, host-protective functionalities. Such approaches may enable precise reprogramming of microbial communities in ways that complement traditional antimicrobial therapies.

Furthermore, clinical translation of nucleotide-based antibiotics is anticipated to benefit from programmable sequence design and predictive computational analyses. With the rapid expansion of genomic and metagenomic datasets for host and microbial species, off-target interactions can increasingly be forecasted and minimized through in silico optimization. These developments position LNP-mediated nucleic acid delivery as a transformative platform—not only for treating antibiotic-resistant infections, but also for engineering microbial ecosystems to promote long-term human health.