Mechanistic insights into mRNA–LNP interactions: role of ionizable lipid content in regulating mRNA intracellular release and translation

Abstract

Messenger RNA (mRNA) therapeutics rely on lipid nanoparticles (LNPs) for effective intracellular delivery, yet the intracellular release of mRNA from these carriers remains an underexplored barrier to translation. In this study, we investigate how the ionizable lipids within clinically relevant LNP formulations influence mRNA delivery and translation. By systematically varying the molar ratio of ionizable lipids (SM-102 and ALC-0315) while maintaining other lipid components constant, we found that LNPs containing lower ionizable lipid levels (~ 30 mol%) achieved significantly higher in vitro transfection efficiency and in vivo mRNA expression compared to standard clinical formulations (50 mol% for Moderna; 46.3 mol% for Pfizer/BioNTech). Notably, this enhancement occurred despite similar cellular uptake and endosomal escape, implicating intracellular mRNA–LNP dissociation as a previously underappreciated determinant of translational efficiency. Fluorescence colocalization analyses confirmed greater cytosolic mRNA release from LNPs with reduced ionizable lipid content. Moreover, these optimized formulations elicited stronger antigen-specific humoral and cellular immune responses in mice, highlighting their potential for improved mRNA performance. Collectively, these findings identify intracellular mRNA dissociation—rather than endosomal escape alone—as a key bottleneck in LNP-mediated delivery and demonstrate that fine-tuning ionizable lipid composition can substantially enhance mRNA performance.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04383-5.

Introduction

Messenger RNA (mRNA) therapeutics have emerged as a transformative platform for the treatment of a wide range of diseases, including infectious diseases, cancer, and autoimmune disorders [1–5]. By encoding disease-targeting proteins, mRNA offers a versatile and rapidly adaptable approach to therapeutic development. Despite its promise, the clinical translation of mRNA faces significant challenges [6]. Notably, mRNA is susceptible to degradation via oxidation, hydrolysis, and nucleases, and its large, negatively charged, hydrophilic structure hinders its ability to cross cell membranes, limiting intracellular delivery.

Lipid nanoparticles (LNPs) have become the gold standard for nucleic acid delivery and are integral to the clinical success of mRNA-based therapeutics [1, 7, 8]. LNPs protect fragile mRNA molecules, facilitate cellular uptake, and promote cytosolic delivery. A typical LNP formulation consists of ionizable lipids, cholesterol, phospholipids, and PEGylated lipids [9–12]. Among these, ionizable lipids play a pivotal role by enabling mRNA encapsulation during formulation, protecting mRNA from degradation in blood circulation, and promoting endosomal escape [13, 14]. Cholesterol contributes to membrane fluidity and fusion, enhancing structural stability and endosomal trafficking [15]. Helper phospholipids, such as 1,2-Distearoyl-sn-glycero-3-phosphocholine (DSPC), provide bilayer structure and mimic natural membrane components [16, 17], while PEGylated lipids stabilize particles, prevent aggregation, and prolong systemic circulation [18, 19].

Ionizable lipids typically comprise the majority of the lipid content in LNPs, approximately 50 mol% in most clinically approved formulations. For example, Moderna’s and Pfizer/BioNTech’s mRNA vaccines contain 50 mol% and 46.3 mol% ionizable lipids, respectively [20]. These lipids are positively charged at low pH (e.g., ~pH 4) during formulation, facilitating electrostatic complexation with negatively charged mRNA. At physiological pH (~ 7.4), they become neutral to reduce systemic toxicity but regain positive charge within the acidic environment of endosomes (pH ~ 5–6), promoting endosomal escape via membrane destabilization and fusion [21].

Efficient mRNA expression relies on three critical steps: cellular uptake, endosomal escape, and intracellular release of mRNA. It is well established that mRNA-LNPs primarily enter cells via endocytosis [22], a process influenced by the physicochemical properties of LNPs, such as surface charge and hydrophobicity [23]. Following endocytosis, mRNA-LNPs are trafficked into endosomes, which maintain an acidic environment (pH 5–6). Within this acidic milieu, ionizable lipids undergo protonation and subsequently interact with anionic endosomal lipids, including phosphatidylserine and bis(monoacylglycero)phosphate [24], promoting structural transitions (e.g., from lamellar to inverted hexagonal phases) that destabilize the membrane and facilitate endosomal escape [6, 22]. The protonation of ionizable lipids also reduces the concentration of free H⁺ ions within the endosome, triggering an influx of counterions, such as chloride, to maintain charge balance. This ionic imbalance drives osmotic water influx, leading to endosomal swelling and rupture, thereby releasing mRNA-LNPs into the cytosol [6]. This mechanism is often referred to as the Proton Sponge Effect. While endosomal escape is widely recognized as a key step for efficient mRNA delivery and expression, the subsequent release of mRNA from the LNP complex remains underexplored. Importantly, successful cytosolic localization does not guarantee translation; mRNA must first dissociate from the LNP to be accessible to the translational machinery. Because mRNA-LNP interactions are primarily governed by electrostatic forces between the mRNA and ionizable lipids, the amount of ionizable lipids may influence mRNA dissociation: higher ionizable lipid content may strengthen mRNA-lipid interactions, potentially impeding release, while lower content may facilitate dissociation.

We hypothesize that adjusting the ionizable lipid content within LNPs can modulate mRNA dissociation, thereby optimizing transfection efficiency. To test this, we systematically varied the ionizable lipid content, specifically SM-102 and ALC-0315, in clinically approved LNP formulations from Moderna and Pfizer/BioNTech (Fig. S1). We evaluated how changes in ionizable lipid content affect the physicochemical properties of LNPs (size, zeta potential, encapsulation efficiency), cellular uptake, endosomal escape, intracellular mRNA release, protein expression, and immunogenicity. Unexpectedly, LNPs formulated with a lower ionizable lipid content demonstrated significantly enhanced mRNA transfection efficiency and immunogenicity in murine models compared to the clinically approved mRNA/LNP formulations with higher ionizable lipid content (46.3–50 mol%). Notably, while endosomal escape was comparable across groups, formulations with reduced ionizable lipid content showed significantly greater intracellular mRNA release, suggesting that dissociation—rather than escape alone—may be a limiting factor in translation efficiency. These findings underscore the importance of striking a balance between mRNA–LNP interactions to facilitate effective mRNA release and functional translation.

Our results highlight a previously underappreciated aspect of mRNA delivery: the intracellular dissociation of mRNA from LNPs. This insight suggests that future optimization of mRNA-LNP formulations should consider not only cellular uptake and endosomal escape but also the intracellular release of mRNA as a key determinant of transfection efficiency.

Results

Tuning ionizable lipid content alters LNP physicochemical properties and mRNA binding

To investigate how ionizable lipid content influences mRNA-LNP characteristics, we started with benchmark formulations based on the clinically approved Moderna and Pfizer/BioNTech mRNA-LNP formulation. This formulation consists of four lipid components: ionizable lipid (either SM-102 or ALC-0315), helper lipid DSPC, cholesterol, and a PEGylated lipid (DMG-PEG2000 or ALC-0159). The molar ratios for the Moderna version are 50:10:38.5:1.5, while for the Pfizer/BioNTech version, they are 46.3:9.4:42.7:1.6, reflecting their clinically approved formulations. We created a series of mRNA-lipid nanoparticle (mRNA-LNP) formulations by varying the proportions of ionizable lipids while keeping the relative ratios of the other lipid components constant. Consequently, as the proportion of ionizable lipids was reduced, the relative molar ratios of the other lipid components increased in the final mRNA-LNP formulation. Importantly, since the amount of mRNA remained unchanged, a reduction in ionizable lipid also decreased the effective ionizable lipid-to-mRNA ratio. The detailed compositions are presented in Supporting Tables 1 and 2. Formulations were designated based on their ionizable lipid molar percentage (e.g., 50, 30, 20). Dynamic light scattering (DLS) measurements revealed that all formulations had similar hydrodynamic diameters and low polydispersity indices (PDI) (Figs. 1a-b), indicating that reductions in ionizable lipid (SM-102) content did not compromise the LNP formulation. However, the 20-group exhibited a significantly more negative zeta potential compared to other groups, which exhibited slightly positive or near-neutral zeta potentials (Fig. 1c). This shift is likely due to a lower density of cationic charges from SM-102, leading to incomplete neutralization of the negatively charged mRNA. Despite this, all formulations showed comparable mRNA encapsulation (Fig. 1d), suggesting that even reduced ionizable lipid levels are sufficient for complexation. To further evaluate the interaction between mRNA and LNPs, gel retardation assays were performed. As the ionizable lipid content decreased, free mRNA bands became increasingly visible (Figs. 1e-f), indicating a weakening of mRNA-LNP complexation. These findings suggest that reducing ionizable lipid content modulates mRNA-LNP binding strength without compromising particle formation or mRNA encapsulation.

Fig. 1.

Physicochemical characterization of SM-102-based mRNA-LNPs formulated with varying ionizable lipid ratios. (a) Particle size; (b) Polydispersity index (PDI); (c) Zeta potential; (d) mRNA encapsulation; (e) Gel retardation assay of freshly prepared mRNA-LNPs with a loading of 0.4 µg mRNA per well; and (f) Quantification of free mRNA from gel images in (e) using ImageJ. The percentage of released mRNA from different groups was normalized to the intensity of the mRNA band in the free mRNA group. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Optimal transfection achieved at intermediate ionizable lipid levels

To assess the functional impact of varying ionizable lipid content, we evaluated the in vitro transfection efficiency of each mRNA-LNP formulation. Interestingly, the transfection efficiency followed a biphasic trend: it increased as the SM-102 content was reduced from 50 to 30, but declined with further reduction to 20 (Fig. 2a). This pattern was consistently observed with both ionizable lipids, SM-102- and ALC-3015-based LNPs from Moderna and Pfizer/BioNTech, respectively (Figs. 2a, S2-9). To further confirm the generalizability of these observations, we tested mRNA encoding multiple proteins, enhanced green fluorescent protein (EGFP), firefly luciferase (Fluc), and ovalbumin (OVA). In all cases, the 30-group outperformed the 50-group, the latter representing the clinically used standard (Figs. 2b-c). Additionally, the same transfection pattern was observed in B16F10 tumor cells (Fig. S9). Collectively, these findings confirm that fine-tuning ionizable lipid content can significantly improve mRNA delivery and expression across different LNP platforms, mRNA cargos, and cell types.

Fig. 2.

Transfection efficiency of mRNA-LNPs. (a) EGFP expression in HEK293 cells transfected with EGFP-mRNA-LNPs (Moderna and Pfizer/BioNTech formulations). Expression was quantified by flow cytometry and presented as mean fluorescence intensity (MFI). (b) Luciferase expression following transfection with Fluc-mRNA-LNPs. (c) OVA expression following transfection with OVA-mRNA-LNPs (Moderna formulation). OVA expression was assessed by ELISA. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ****p < 0.0001

Cellular uptake and endosomal escape are not primary drivers of enhanced transfection

To investigate the mechanistic basis for the improved transfection observed with reduced ionizable lipid content, we evaluated cellular uptake and endosomal escape using the 30- and 50-group formulations from Moderna, as well as the 30- and 46.3-group formulations from Pfizer/BioNTech. The 20-group, which exhibited markedly reduced transfection efficiency (Fig. 2a), was excluded from these mechanistic studies due to its suboptimal performance. But we suspect that the reduced efficiency in the 20-group may be attributed to excessively weak mRNA–LNP interactions, leading to insufficient protection of mRNA from nuclease degradation or exchange with anionic biomolecules in the extracellular or intracellular environment. mRNA-LNPs were specifically labeled with NBD-cholesterol by replacing 25% (w/w) of the unlabeled cholesterol in mRNA-LNPs with NBD-cholesterol. Flow cytometry (Figs. 3a, S10) and fluorescence microscopy (Fig. 3b) analysis demonstrated comparable levels of cellular uptake between all groups, suggesting that differential internalization is not responsible for the enhanced transfection observed in the 30-group.

Fig. 3.

Cellular uptake of mRNA-LNPs (Moderna version). (a) Quantification of uptake in HEK293 cells assessed by flow cytometry. mRNA-LNPs were labeled with NBD-cholesterol (Green). (b) Representative fluorescence microscopy images of cells treated with NBD-cholesterol-labeled mRNA-LNPs (green). Nuclei were stained with Hoechst 33,342 (blue). Scale bars, 100 μm. Data are presented as mean ± SD

Next, we assessed endosomal escape, a critical step for cytosolic delivery of mRNA. Confocal microscopy revealed similar colocalization patterns between NBD-cholesterol-labeled mRNA-LNPs (green) and endosomal compartments stained with LysoTracker (red) for both groups (Fig. 4a). Quantitative analysis using Pearson’s correlation coefficients indicated no statistically significant difference in endosomal escape efficiency between the 30- and 50-groups (Fig. 4b). These findings were further supported by a calcein release assay, which confirmed comparable levels of endosomal disruption across both formulations, as shown by fluorescence images (Figs. S11a, S12a) and quantitative measurements by flow cytometry (Figs. S11b, S12b). Notably, despite the reduced ionizable lipid content, endosomal escape efficiency was maintained, indicating that large quantities of ionizable lipids may not be essential for effective endosomal disruption. These results collectively suggest that neither cellular uptake nor endosomal escape accounts for the superior transfection efficiency of the 30-group, pointing instead to post-endosomal processes as the critical determinant.

Fig. 4.

Endosomal escape of mRNA-LNPs (Moderna version). (a) Representative confocal images of HEK293 cells treated with NBD-cholesterol-labeled mRNA-LNPs (green). Endosomes/lysosomes were stained with LysoTracker Red, and nuclei with Hoechst 33,342 (blue). Scale bars, 25 μm. (b) Quantification of colocalization between mRNA-LNPs and endosomes was performed using Pearson’s correlation coefficient, based on three independent confocal images per group, analyzed with ImageJ software. Data are presented as mean ± SD

Increased intracellular mRNA dissociation enhances translational efficiency

Given that the 30- and 50- groups demonstrated comparable endosomal escape efficiencies, we hypothesized that the improved transfection observed in the 30-group may result from enhanced dissociation of mRNA from the LNP complex within cells. We posited that reduced electrostatic interaction in the 30-group facilitates greater mRNA release in cells, thereby improving accessibility to the translational machinery.

To evaluate intracellular mRNA-LNP dissociation, we employed dual-labeling of mRNA-LNPs with NBD-cholesterol or NBD-SM102 (green)to label the LNP and Cy5 (red) to label the mRNA. HEK-293 cells were incubated with the labeled mRNA-LNPs for six hours, followed by confocal microscopy analysis (Figs. 5, S13). In the 30-group, a spatial separation of the Cy5 and NBD signals was observed, indicating the dissociation of mRNA from LNPs (Figs. 5a, S13). In contrast, the 50-group exhibited strong colocalization of Cy5 and NBD signals (Figs. 5a, S13), indicating that mRNA remained tightly associated with the LNP carriers. Quantitative colocalization analysis confirmed a significantly lower degree of mRNA-LNP association in the 30-group compared to the 50-group (Figs. 5b, S13b), supporting the notion of increased intracellular mRNA release. The same trend was observed in Pfizer/BioNTech version mRNA-LNPs (Fig. S14). Since free mRNA must be available in the cytosol for ribosomal engagement and translation, this enhanced dissociation likely accounts for the higher protein expression observed with the 30-group. Conversely, in the 50-group, strong mRNA-LNP complexation may limit translational access, thereby reducing transfection efficiency. These findings highlight intracellular mRNA release as a critical step in effective mRNA delivery, distinct from endosomal escape, and underscore the importance of tuning mRNA–lipid interactions to optimize translational outcomes.

Fig. 5.

Intracellular dissociation of mRNA from LNPs. (a) Representative confocal images of HEK293 cells after treatment with mRNA-LNP for 6 h: dual-labeled mRNA-LNPs: NBD-cholesterol for LNPs (green) and Cy5 for mRNA (red). Nuclei were stained with Hoechst 33,342 (blue). Scale bars, 25 μm. (b) Quantification of intracellular LNP and mRNA colocalization based on 7–10 images for each group, expressed as Pearson’s correlation coefficient and analyzed using ImageJ. Data are presented as mean ± SD. ***p < 0.001

Reduced ionizable lipid content enhances in vivo mRNA expression

Building on the observed in vitro enhancement in transfection efficiency, we next evaluated the in vivo performance of the mRNA-LNP formulations. Following intramuscular injection in mice, bioluminescence imaging revealed that the 30-group resulted in significantly higher luciferase expression compared to the 50-group (Figs. 6a-b, S15), mirroring the in vitro findings. To further assess systemic delivery, we administered mRNA-LNPs intravenously (Figs. 6c-d, S16). Consistent with intramuscular results, the 30-group exhibited markedly elevated luciferase expression in the liver—the primary site of mRNA-LNP accumulation following intravenous injection [25]. These findings demonstrate that modulating ionizable lipid content to achieve optimal mRNA–lipid binding enhances both in vitro and in vivo mRNA expression.

Fig. 6.

In vivo luciferase expression following administration of mRNA-LNPs. (a) Representative bioluminescence images of mice at 6 h following IM injection of Fluc-mRNA-LNPs (0.25 µg Fluc mRNA per mouse) in C57BL/6 mice. (b) Quantification of luminescence intensity based on images from (A). (c) Representative images at 6 h after IV injection of Fluc-mRNA-LNPs (2 µg Fluc mRNA per mouse) in Balb/c mice. (d) Ex vivo bioluminescence of major organs, and (e) Quantification of organ-specific luminescence. Data are shown as mean ± SD. **p < 0.01, ***p < 0.0001

Enhanced immune response with optimized LNP formulation

To determine whether improved mRNA delivery translated into enhanced immunogenicity, we evaluated immune responses following administration of OVA mRNA-LNPs. Mice were immunized on day 0 and boosted on day 5. Serum samples collected on days 7 and 10 were analyzed by ELISA to quantify OVA-specific IgG levels, while splenocytes harvested on day 10 were restimulated ex vivo with OVA peptide to assess T cell responses via IFN-γ secretion (Fig. 7a). The 30-group elicited significantly higher OVA-specific IgG titers on both days 7 and 10 compared to other groups (Fig. 7b), indicating a more potent humoral immune response. In parallel, splenocytes from the 30-group secreted substantially more IFN-γ in response to OVA-specific peptide stimulation, confirming the induction of a strong, antigen-specific cellular immune response (Fig. 7c). In contrast, although the 50-group showed a trend toward increased IFN-γ secretion, the response was not significantly different from unstimulated controls (Fig. 7c), suggesting a weaker T cell response. These results collectively demonstrate that the 30-group enhances both humoral and cellular immunity, supporting the hypothesis that optimized mRNA–LNP dissociation improves antigen expression and downstream immune activation.

Fig. 7.

In vivo immunogenicity of mRNA-LNPs. (a) Schematic representation of the experimental procedures used to evaluate the in vivo immunogenicity of mRNA-LNPs. (b) OVA-specific IgG titers in serum collected on days 7 and 10 following SC injection of OVA-mRNA-LNPs (2 µg mRNA per mouse). (b) IFN-γ secretion from splenocytes was measured by ELISA after ex vivo restimulation with OVA peptide. Data is presented as mean ± SD. n = 4 mice for PBS group and n = 5 mice for 30- and 50- groups. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Discussion and conclusions

LNPs have become the cornerstone of mRNA therapeutics, primarily due to their ability to encapsulate, protect, and deliver fragile mRNA molecules into target cells [1]. Among LNP components, ionizable lipids are crucial for mRNA encapsulation and endosomal escape, two essential processes that enable cytosolic delivery [6]. Traditionally, a higher ionizable lipid content has been associated with improved mRNA complexation, due to the availability of more positively charged groups that interact with the negatively charged mRNA, as well as enhanced intracellular trafficking resulting from improved endosomal disruption capabilities [26, 27]. However, our findings challenge this conventional view by revealing that an excess of ionizable lipids can actually hinder downstream mRNA translation by restricting intracellular release of mRNA from the LNP complex.

Our study demonstrates that while clinically approved formulations from Moderna and Pfizer/BioNTech contain approximately 50 mol% ionizable lipids, reducing this proportion to around 30 mol% significantly enhances protein expression and immunogenicity, both in vitro and in vivo. Importantly, this improvement is not due to differences in particle size, encapsulation efficiency, cellular uptake, or endosomal escape. Rather, we found that intracellular dissociation of mRNA from LNPs, a step distinct from endosomal escape, is the key limiting factor influenced by ionizable lipid content (Figs. 5, S13, S14).

Gel retardation assay demonstrated that decreasing the ionizable lipid content weakens the interaction between mRNA and LNPs (Figs. 1e-f). This weakened binding facilitates intracellular mRNA release, as evidenced by fluorescent colocalization studies, which show a greater dispersion of intracellular free mRNA (Fig. 5), thereby enabling enhanced access to the cellular translation machinery. In contrast, standard formulations with higher ionizable lipids exhibited stronger electrostatic interactions, likely contributing to mRNA retention within the lipid matrix. This sequestration may impede mRNA translation by reducing its translational accessibility. Notably, this physicochemical barrier has been largely overlooked in the design of mRNA delivery, representing a critical bottleneck in optimizing translational efficiency.

Additionally, our data reveals a biphasic relationship between ionizable lipid content and transfection efficiency. While moderate reductions (30-group) enhance performance, further reduction (20-group) compromises transfection efficiency (Fig. 2), likely due to insufficient mRNA complexation or instability in the extracellular environment. These findings underscore the need for a finely tuned balance: sufficient ionizable lipid to ensure efficient mRNA encapsulation and endosomal escape, but not so much as to hinder cytosolic release. This conclusion is consistent with a recent report highlighting the importance of precisely tuning ionizable lipid levels [27]. Encouragingly, the 30-group formulations enhanced transfection than the 50-group. At the same time, the 30-group maintained physical stability during storage at − 80 °C (Fig. S17), indicating that reducing the amount of ionizable lipids does not compromise the stability of mRNA-LNP, thereby promising practical deployment. Moreover, this optimized formulation elicited stronger humoral and cellular immune responses (Fig. 7), underscoring the functional benefits of tuning ionizable lipid content in mRNA vaccine design.

This study identifies intracellular mRNA dissociation as a previously underappreciated but critical step in the mRNA delivery cascade. We demonstrate that reducing ionizable lipid content below clinically approved levels enhances intracellular mRNA release, leading to improved translation and immunogenicity. These results suggest that endosomal escape, while necessary, is not sufficient for effective mRNA delivery—highlighting mRNA–lipid dissociation as a key design parameter. Future efforts to improve mRNA-LNP platforms should therefore incorporate strategies to balance complexation strength with timely intracellular release. Such refinements could enhance the potency and versatility of mRNA therapeutics across a broad range of applications, from vaccines to protein replacement therapies.

Although our data demonstrate that intracellular mRNA dissociation from LNPs is a critical determinant of transfection efficiency, a significant limitation is that we did not resolve the specific subcellular location where dissociation occurs. Prior Studies using FRET (Fluorescence Resonance Energy Transfer)-based probes have shown that siRNA-LNP predominantly dissociate within the endosomal compartment [28, 29]. Whether these findings translate to mRNA–LNPs remains uncertain due to key differences in nucleic acid size and the distinct structural organization of mRNA–LNPs compared with siRNA–LNPs [30]. Future studies employing compartment-specific imaging or localized biochemical probes will be necessary to establish where mRNA–LNP dissociation takes place. This information will be essential for designing strategies that more precisely enhance intracellular release and overall mRNA delivery efficiency. Another limitation is that the present study focuses exclusively on mRNA. Given the broad utility of LNPs for delivering diverse nucleic acids [31], additional work is needed to determine whether the dissociation–transfection relationship we observed applies similarly to other modalities such as microRNAs, siRNAs, or antisense oligonucleotides. Addressing these gaps will help generalize the mechanistic insights described here and guide the optimization of LNP systems across a wider range of therapeutic applications.

Materials and methods

Materials

LipidLaunch™ LNP-102 Exploration Kit and LNP-0315 Exploration Kits were purchased from Cayman Chemical (Ann Arbor, MI, USA). The Quant-iT™ RNA Assay Kit (Broad Range), NBD Cholesterol (22-(N-(7-Nitrobenz-2-Oxa-1,3-Diazol-4-yl) Amino)-23,24-Bisnor-5-Cholen-3β-Ol), DAPI (4’,6-diamidino-2-phenylindole), and Calcein (High Purity) were purchased from Thermo Fisher Scientific (Waltham, MA, USA). EGFP (Cap1, 5-MOU, Cat. No.:RP-A00008), and Firefly luciferase (Fluc, Cap1, 5-MOU, Cat. No.: RP-A00022) mRNAs were purchased from GenScript (Piscataway, NJ, USA). OVA mRNA (clean Cap1, 5-MOU, Cat. No.: L-7210) was purchased from TriLink (San Diego, CA). EZ Cap™ Cy5 FLuc mRNA (5-moUTP, Cat. No.: R1010) was purchased from APExBio (Houston, TX, USA). ELISA kits for mouse IFN-γ and IgG, LysoTracker™ Red DND-99, OVA protein, and the SIINFEKL peptide (OVA257–264) were purchased from Invivogen (San Diego, CA, USA). D-Luciferin (potassium salt) was purchased from GoldBio (St Louis, MO). All chemical reagents were used directly without further purification.

Cells and animals

HEK293 human embryonic kidney cells (ATCC, Manassas, VA, USA) were cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin at 37 °C with 5% CO₂. All animal procedures were approved by the Institutional Animal Care and Use Committee in conformity with the NIH guidelines for the care and use of laboratory animals. Male C57BL/6 and female BALB/c mice (5–6 weeks old) were purchased from Jackson Laboratory (Bar Harbor, ME, USA) and acclimated for one week prior to experimentation.

mRNA-LNP preparation and characterization

mRNA-LNPs were prepared using a modified protocol based on the manufacturer’s instructions (Cayman Chemical). Briefly, an ethanol phase containing ionizable lipid (SM-102 or ALC-0315), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), cholesterol, and PEGylated lipid (DMG-PEG2000 or ALC-0159) was rapidly mixed with an aqueous mRNA solution (50 mM sodium acetate, pH 4) by pipetting for at least 15 s and then left undisturbed for 20 min. The starting molar ratio of SM-102: DSPC: cholesterol: PEG-lipid in the Moderna version mRNA-LNP was 50:10:38.5:1.5, while the starting molar ratio of ALC-0315: DSPC: cholesterol: ALC-0159 in the Pfizer/BioNTech was 46.3:9.4:42.7:1.6, based on their clinically approved formulations, with a lipid-to-mRNA weight ratio of 10:1. A series of formulations was generated by systematically reducing the proportion of ionizable lipid, while the amounts of the other lipid components and the mRNA were kept constant. Dynamic light scattering (DLS) was performed using a Zetasizer Nano (Malvern Instruments; Malvern, UK) to measure particle size, zeta potential, and polydispersity index (PDI). mRNA encapsulation efficiency was assessed using the Quant-iT™ RNA Assay Kit (Broad Range) according to the manufacturer’s instructions. A 1% Triton X-100 was used to extract the loaded mRNA from LNPs. The mRNA–LNP complexation strength was evaluated using gel retardation assays with a loading amount of 0.4 µg mRNA per well, running for 30 min at 65 mV. The acquired gel was imaged with a gel documentation system (Bio-Rad Laboratories, Hercules, CA, USA). To calculate the percentage of released mRNA from different groups, the mRNA band was quantified using ImageJ software and normalized to the intensity of the mRNA band in the free mRNA group.

mRNA cell transfection

For EGFP expression studies, HEK293 cells were seeded at a density of 1 × 10⁵ cells per well in 24-well plates and cultured for 24 h. Cells were then treated with mRNA-LNPs containing 50 ng EGFP-encoding mRNA per well and incubated for an additional 24 h. After treatment, cells were collected, and EGFP expression was quantified using flow cytometry (Beckman Coulter, Brea, CA, USA). For luciferase expression assays, HEK293 cells or B16F10 cells were seeded in white, clear-bottom 96-well plates at a density of 1 × 10⁴ cells per well and incubated for 24 h. Cells were subsequently treated with mRNA-LNPs containing 20 ng of Fluc mRNA per well and incubated for an additional 24 h. Luciferase activity was measured using the Luc-Screen™ Extended-Glow Luciferase Assay System (Thermo Fisher Scientific) according to the manufacturer’s protocol. For OVA mRNA expression analysis, HEK293 cells were seeded in 6-well plates at a density of 5 × 10⁵ cells per well and cultured for 24 h. Cells were then treated with mRNA-LNPs at a dose of 500 ng OVA mRNA per well and incubated for 24 h. Following incubation, cells were lysed using 1X RIPA buffer supplemented with 1X protease inhibitors, and total protein was extracted. OVA protein levels in the lysates were quantified using OVA detection assay kit (Chondrex, Inc.).

Stability evaluation

mRNA-LNPs were formulated and suspended in PBS containing 8% sucrose (w/v) solution, then stored at − 80 °C. After three weeks of storage, samples were thawed on ice and analyzed for particle size, zeta potential, polydispersity index (PDI), and transfection efficiency. These parameters were compared with those of freshly prepared mRNA-LNPs to assess stability.

Cellular uptake

For the cellular uptake study, mRNA-LNPs were labeled by substituting 25 mol% of fluorescently labeled cholesterol with unlabeled cholesterol. HEK293 cells were seeded in 24-well plates at a density of 1 × 10⁵ cells/well and cultured for 24 h. Cells were then treated with mRNA-LNPs containing 50 ng of mRNA and incubated for 6 h. Following treatment, cells were harvested and analyzed by flow cytometry to measure green fluorescence from NBD-cholesterol, indicating the extent of cellular uptake.

Endosomal escape

Endosomal escape was assessed using the Calcein assay. HEK293 cells were seeded in 24-well plates (1 × 10⁵ cells per well) and incubated for 24 h. Cells were then treated with 0.5 mL of mRNA-LNPs (50 ng mRNA per well) supplemented with 40 µL of Calcein solution (1 mg/mL) and incubated for 6 h. Post-incubation, cells were washed with PBS, stained with Hoechst 33,342 (3 µg/mL, 15 min), and fixed with 4% paraformaldehyde for 15 min. Calcein fluorescence was visualized using the iRiS™ Digital Cell Imaging System (Logos Biosystems; Anyang, South Korea), and fluorescence intensity was quantified with ImageJ software. In a parallel experiment, flow cytometry was used to quantify Calcein fluorescence. After 6 h of treatment with RNA-LNPs and Calcein, cells were washed, harvested, and analyzed by flow cytometry.

Endosomal escape was further visualized using confocal microscopy. HEK293 cells were seeded in confocal dishes at a density of 1 × 10⁵ cells per well and cultured for 24 h before treatment with NBD-cholesterol-labeled mRNA-LNPs (50 ng mRNA/well) for 6 h. Endosomes/lysosomes were stained with LysoTracker (4 µM), and nuclei were stained with Hoechst 33,342 (3 µg/mL). Imaging was performed using a Leica SP8 laser scanning confocal microscope (Leica Microsystems; Wetzlar, Germany). Pearson’s correlation coefficient was employed to quantify the degree of overlap between NBD-cholesterol–labeled LNPs and LysoTracker-labeled endosomes in confocal images, an indicator of the degree of endosomal escape. The coefficient ranges from 0 to 1, where 0 indicates complete separation of the two channels and 1 represents complete overlap. In this context, a higher Pearson’s correlation coefficient reflects greater colocalization, indicating increased retention of LNPs within endosomes and, consequently, reduced endosomal escape. Pearson’s correlation coefficients were calculated from the acquired images using Leica LAS X software.

mRNA dissociation

To evaluate mRNA dissociation, HEK293 cells were seeded in confocal dishes at a density of 1 × 10⁵ cells and cultured for 24 h, then treated with Cy5-labeled mRNA-LNPs (100 ng mRNA per well) for 6 h. After treatment, cells were stained with Hoechst 33,342 (3 µg/mL) for 15 min, washed with PBS, fixed with 4% paraformaldehyde for 15 min, and imaged using a Leica SP8 laser scanning confocal microscope or Nikon N-STORM Super-Resolution Microscope. Images were analyzed with Leica LAS X software or ImageJ to assess intracellular mRNA release. Pearson’s correlation coefficient was used to indicate the colocalization of the Cy5-labeled mRNA and NBD cholesterol-labeled LNP. A higher Pearson’s correlation coefficient reflects greater colocalization, indicating increased retention of mRNA within LNPs and, consequently, reduced mRNA dissociation.

In vivo mRNA expression

Mice received intramuscular (IM) injections of Fluc mRNA-LNPs (0.25 µg of mRNA/mouse) into the gastrocnemius muscle of C57BL/6 mice. Six hours post-injection, mice were administered luciferin intraperitoneally (100 µL, 30 mg/mL), and bioluminescence imaging was performed 10–20 min later using a Spectral AMI-HT IVIS system (Spectral Instruments Imaging, Tucson, AZ, USA). In a separate experiment, BALB/c mice were intravenously injected with Fluc mRNA-LNPs (2 µg mRNA/mouse) and imaged 6 h after receiving the treatment. For ex vivo imaging of organs, mice received an additional luciferin dose (30 mg/mL; 100 µL per mouse) before euthanasia. Bioluminescence signals were analyzed using Aura software.

In vivo vaccination study

For vaccination studies, C57BL/6 mice were immunized with OVA mRNA-LNP formulations via subcutaneous injection (2 µg mRNA per mouse) on days 0 and 5. Serum samples were collected on days 7 and 10, then diluted 1:10 in ELISA/ELISPOT diluent, and antigen-specific IgG levels were quantified using an in-house ELISA. Briefly, flat-bottom 96-well plates were precoated at 4 °C overnight with 100 µL per well of OVA protein (20 µg/mL) in 100 mM carbonate buffer (pH 9.6). Plates were then washed three times with washing buffer and blocked with ELISA/ELISPOT diluent for 1 h at room temperature, followed by two additional washes. Diluted serum samples (100 µL per well) were added and incubated for 2 h at room temperature. After three washes, 100 µL of horseradish peroxidase–conjugated goat anti-mouse IgG (1:20,000 in ELISA/ELISPOT diluent) was added and incubated for 1 h at room temperature. Plates were washed three times and incubated with 100 µL of tetramethylbenzidine substrate for 15–30 min, depending on color development. The reaction was stopped by adding 100 µL of stop solution, and absorbance was measured at 450 nm with correction at 570 nm. On day 10, mice were euthanized, and splenocytes were harvested and cultured in 24-well plates (1 × 10⁶ cells per well) in MEM supplemented with 10 mM β-mercaptoethanol. Cells were either untreated or stimulated with 2.5 µg/mL OVA_257–264 (SIINFEKL) peptide for 48 h. The culture supernatants were collected and analyzed for IFN-γ secretion using ELISA.

Statistical analysis

Statistical analyses were performed with GraphPad Prism 10 (La Jolla, CA). All data were presented as mean ± standard deviation (SD). Group comparisons were performed using one-way or two-way ANOVA, followed by appropriate post hoc multiple comparisons tests as recommended. In cases involving two groups, either unpaired or paired t-tests were applied, as appropriate. A p-value of less than 0.05 was considered statistically significant. Statistical significance was denoted as follows: ns (not significant), p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****).

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H. T. designed and performed the experiments, analyzed the data, and wrote the original draft of the manuscript. U.K. assisted with LNP preparation, as well as in vitro and in vivo animal studies. C.N. and Y.L. contributed to the in vivo animal studies. F.M. designed the experiment, analyzed data, wrote, and revised the manuscript. All authors have read and approved the final version of the manuscript.

Funding

This work was supported by the start-up package from the University of Massachusetts Lowell and the University of Houston (UH).

Data availability

No datasets were generated or analyzed during the current study.

Declarations

Ethics approval and consent to participate

All animal procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals and the Guidelines of the Institutional Animal Care and Ethics Committee of the University of Massachusetts Lowell and University of Houston.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.