Chemically or physically introducing lipids into lysine-histidine-based peptide systems for safe, efficient and targeted mRNA delivery

Abstract

mRNA therapy offers a promising platform for developing highly effective and personalized treatments for diverse diseases. In this study, we integrate structurally simple yet functionally complementary peptide and lipid components to enable the construction of a highly efficient mRNA delivery vector with minimal formulation complexity, making it cost-effective. Using both chemical lipidation (i.e., lipopeptide) and physical lipid incorporation to introduce lipid components with complementary functions into simple lysine–histidine modular peptides, we established a versatile library to develop economical, safe, efficient, and targeted mRNA carriers and systematically investigated how peptide and lipid modules jointly affect mRNA delivery in vitro and in vivo. In vitro results demonstrated that lipid components significantly improved the mRNA encapsulation and transfection efficiency of peptides while maintaining good biocompatibility. The chemical optimization strategy yielded a remarkable 5.3-fold enhancement in transfection efficiency over the pure peptide system. Utilizing the physical strategy, the top-performing formulation exhibited higher transfection efficiency than the pure lipid and 43.9-fold greater efficiency than its pure peptide counterpart, further surpassing commercial reagents in mRNA expression intensity. In vivo, the optimal physical composite system achieved improved efficiency and lung selectivity in mRNA delivery compared to the pure lipid system, demonstrating strong potential for advancing mRNA therapy for lung-related diseases. These findings validate the effectiveness of our complementary design strategy and indicate that this approach can be broadly applied to develop additional simple yet highly functional gene-delivery vectors.

---

New concepts

Developing safe, efficient, and targeted delivery vectors is essential for advancing gene therapy. Peptides offer excellent biocompatibility, biodegradability, and multifunctionality, yet their application is limited by low delivery efficiency and poor stability as single components. Lipids, in contrast, provide strong delivery efficiency and structural stability but raise concerns regarding biosafety. Here, we introduce a complementary peptide–lipid design strategy that enables the construction of extrahepatic delivery vectors with minimal components yet fully integrated functionalities. By designing lysine–histidine peptides and incorporating lipid elements through chemical lipidation or physical incorporation, we created a simple yet versatile library that combines the biosafety, endosomal escape capability, and cost-effectiveness of low-molecular-weight peptides with the hydrophobic assembly and interaction properties of lipids. Through modular regulation and comprehensive characterization, we elucidated how peptide and lipid modules cooperate to define carrier performance. Using dual screening strategies, we identified high-performing mRNA delivery systems exhibiting excellent mRNA encapsulation, structural assembly, and cellular internalization. Moreover, the optimal formulation outperformed both pure lipid systems in endosomal escape, in vitro and in vivo transfection efficiency, and lung-targeting specificity, demonstrating the feasibility and effectiveness of our strategy. This minimalist yet functionally complete design concept provides a promising pathway for developing economical and efficient gene-delivery platforms.

Introduction

mRNA therapy stands as one of the most promising therapeutic strategies for various diseases, and developing safe and efficient mRNA carriers is essential for its clinical application.^1–3 Peptide-based carriers offer advantages such as modular design, bioactivity, and biodegradability, making them promising candidates for gene delivery.^4–6 In particular, low-molecular-weight peptides have good biocompatibility and economical production but poor stability in vivo, which hinders efficient in vivo delivery of genetic drugs—especially large nucleic acids such as mRNA. To address this, researchers have developed polymeric carriers based on basic amino acids (e.g., lysine, arginine, and histidine), which demonstrate improved nucleic acid condensation, transfection efficiency, and in vivo stability.^7–9 However, their complex synthesis and potential cytotoxicity due to their highly positive charge remain significant challenges. Thus, improving the low-molecular-weight peptide system to achieve efficient, safe and economical mRNA delivery remains an attractive direction in biomedical research.

In recent years, lipid-based gene carriers have achieved great success in practical applications due to their high efficiency and in vivo stability.^1,10,11 Meanwhile, their complex formula and potential immune reactions need to be taken into consideration.² It is a potential strategy to combine low-molecular-weight peptides and lipids, which is expected to integrate the advantages of both.^12–20 Jiang's group developed a class of dendritic lipopeptide-based systems for siRNA or small molecular drugs, which can achieve high cell and tissue penetration and mitochondrial targeting ability.^16–23 Our group previously developed lipopeptide-based lipid nanoparticles that enable efficient siRNA and mRNA delivery to the lungs, liver, and spleen, while demonstrating excellent biosafety.²⁴ In addition, the complex system of peptides and cationic lipids has been reported to have a significantly enhanced transfection efficiency.^25,26 To enable simpler carrier design and achieve practical economic benefits, it is valuable to systematically dissect peptide–lipid property complementarity so that the simplest peptide and lipid components can together endow gene vectors with the full set of properties required for high-performance delivery.

Here, we utilized a simple modular peptide library comprising five low-molecular-weight peptides with different lengths composed solely of lysine and histidine (denoted as KHn, where n is the number of residues ranging from 2 to 12) and introduced lipid elements into the KHn systems via chemical lipidation (Fig. S1) and physical lipid incorporation to systematically evaluate the impact of lipid elements on the performance of KHn systems. And, here, the KHn and lipid components are designed to have complementary properties. The results demonstrate that both approaches enhanced the mRNA encapsulation and transfection efficiency of KHn-based carriers while maintaining good biocompatibility. With the chemical strategy, the optimal carrier achieved a transfection efficiency of 71.2% in human cervical cancer cells (HeLa), which is 5.3-fold higher than that of the pure peptide system. Further analysis indicated that both the lipid and KHn module significantly influenced transfection ability. With the physical strategy, the top-performing formulation exhibited 75.9% transfection efficiency in HeLa cells, surpassing both the pure lipid system and exceeding the pure peptide system by 43.9-fold. Moreover, it elicited higher mRNA expression levels than a commercial transfection reagent. The peptide-to-lipid molar ratio was identified as a critical factor, with the KHn structure also playing a notable role. Importantly, compared to the pure lipid system, the optimal physically formulated system mediated more efficient lung-targeted mRNA delivery in vivo, with a better selectivity of 74.8%. Finally, we identified a simple peptide/lipid system with an enhanced assembly structure, mRNA encapsulation effect, endosome escape ability, transfection efficiency, and lung selectivity. At the same time, it also exhibits good biocompatibility, economicity, and storage stability.

Results and discussion

In vitro mRNA delivery of KHn

In the modular KHn peptide library (Fig. 1a and Table S1), the peptides were intentionally restricted to two functional modules: (i) nucleic-acid binding (lysine (K) and histidine (H))²⁷ and (ii) endosomal escape (H).^28,29 This minimalist design ensures high biocompatibility and low synthesis cost. The nucleic acid binding capacity of KHn peptides was initially evaluated using a gel electrophoretic mobility shift assay (Fig. 1b). The results indicated that although nucleic acid encapsulation improved with increasing KHn-to-nucleic acid mass ratios, all peptides exhibited limited binding efficiency. Even at a mass ratio of 50, clear migration bands of free nucleic acids were still detected, suggesting insufficient condensation, likely due to the lack of hydrophobic domains in KHn which impedes effective self-assembly with nucleic acid in an aqueous environment. Longer peptides, particularly KH8 and KH12, demonstrated comparatively better encapsulation performance. Consistent with the gel electrophoretic results, quantitative assessment of mRNA condensation further confirmed that longer KHn peptides achieved higher encapsulation efficiency (ee), with KH8 reaching 75.3% efficiency at a KH8-to-DNA mass ratio of 30 (Fig. 1c). Cytocompatibility evaluation revealed that all KHn peptides exhibited good biocompatibility, maintaining cell viability above 90% at a concentration of 60 µg mL⁻¹ (Fig. 1d). And, safety is an important indicator for drug delivery carriers.³⁰ However, a gradual decrease in cell viability was observed as the peptide length increased. These findings underscore the importance of developing low-molecular-weight cationic peptide-based gene delivery systems for biological safety.

Fig. 1 In vitro mRNA delivery of KHn. (a) Molecular structures of 5 kinds of modular KHn peptides. (b) Gel electrophoresis effects of KHn@DNA complexes at different KHn-to-DNA mass ratios (10, 20, 30, 40, and 50). (c) The mRNA encapsulation efficiency of KHn at a KHn-to-DNA mass ratio of 30 (n = 3). (d) Cell viability of HeLa treated with a KHn@mRNA complex concentration of 60 µg mL⁻¹ for 48 h (n = 3). (e) The uptake efficiency of HeLa to KHn@FAM-siRNA complexes (n = 2). (f) The EGFP expression positive rate in HeLa treated with KHn@mEGFP complexes (n = 2).

To evaluate the potential of KHn as a gene delivery vector, we first examined its ability to transport carboxyfluorescein-labeled small interfering RNA (FAM-siRNA) into cells. As shown in Fig. 1e, longer KHn demonstrated higher cellular uptake, which may be attributed to higher ee and increased cationic modules promoting membrane penetration, but their internalization efficiency did not exceed 30% (Fig. S2). Therefore, the transfection ability of the KHn@mEGFP complexes (mEGFP represents enhanced green fluorescent protein (EGFP) mRNA) in cells was very limited. Only KH12@mEGFP reached approximately 13.5% transfection, while all other constructs yielded below 5% (Fig. 1f and Fig. S3). Given KHn's excellent biocompatibility, economical synthesis, and promising endosomal escape capability, the incorporation of lipid components is promising for enhancing its transfection efficiency to develop safe and efficient gene delivery vectors.

Chemical lipidation of KHn for improved transfection

Lipidation of KHn was performed via the ring-opening reaction between amines and epoxides (Fig. 2a and Fig. S1), yielding 25 lipopeptides with a simple structure composed of diverse KHn headgroups and varying lipid tail lengths (denoted as KHn-LP or Ax/Ex-KHn).^13,24,31 The KHn headgroup facilitates nucleic acid adsorption and endosomal escape, while the lipid tail promotes hydrophobic assembly and interaction with the lipid bilayer of cell membranes. Gel electrophoresis confirmed markedly enhanced nucleic acid encapsulation in all KHn-LP formulations (Fig. 2b), with most achieving over 80% mRNA encapsulation. Interestingly, KH2-LP and KH12-LP, with the shortest and longest headgroups, respectively, performed best (Fig. 2c and Fig. S4). In contrast to the pure KHn system, lipopeptides bearing the shortest KHn headgroup exhibited the highest cytotoxicity, while the others maintained good biocompatibility (Fig. 2d). These results indicated that lipidation significantly alters the properties of KHn.

Fig. 2 In vitro mRNA delivery of KHn-LPs. (a) Molecular structures of 5 kinds of 1,2-epoxides, and a schematic diagram of the reaction between KHn and 1,2-epoxides to synthesize KHn-LPs. (b) Gel electrophoresis effects of KHn-LP@DNA complexes. (c) A heatmap of the mRNA ee of KHn-LP (n = 3). (d) A heatmap of the cell viability of HeLa treated with a KHn-LP@mRNA complex concentration of 20 µg mL⁻¹ for 48 h (n = 3). (e) A heatmap of the firefly luciferase activity of HeLa treated with KHn-LP@mFluc and KHn@mFluc complexes after 48 h (n = 2). (f) The inverted fluorescence microscopy image (left) and corresponding flow cytometry characterization result (right) of HeLa treated with A10-KH12@mEGFP complexes after 48 h. (g) Hydrodynamic diameter and polydispersity index (PDI) of A10-KH12@mRNA and KH12@mRNA complexes at neutral pH (n = 3). (h) The transmission electron microscopy (TEM) image of A10-KH12@mRNA complexes at neutral pH. (i) Zeta potentials of A10-KH12@mRNA and KH12@mRNA complexes at neutral pH (n = 3). (j) The mRNA ee of A10-KH12 and KH12 (n = 3).

We further evaluated the intracellular uptake efficiency of the 25 KHn-LP formulations. The results showed that longer KHn headgroups and tails (especially the A12 tail) promote the uptake (Fig. S5–S7). Then, we used the 25 KHn-LP formulations to deliver firefly luciferase (Fluc) mRNA (mFluc) into cancer cells (HeLa and human lung cancer cells A549) for preliminary screening. As shown in Fig. 2e and Fig. S8, lipopeptides with KH2 or KH12 headgroups exhibited superior mRNA delivery efficiency, consistent with the trend observed in mRNA ee across the KHn series; in contrast, lipopeptides with A10, A12, or A14 tails also showed enhanced transfection, but this trend did not strongly correlate with the ee of different tails. Notably, the A12 tail showed the best overall performance, exhibiting the highest encapsulation, uptake, and transfection (Fig. 2c and Fig. S4, S5). We speculate that these structures achieve an optimal balance among nucleic acid encapsulation, endosomal escape, and cytoplasmic release, contributing to their high transfection performance.^20,32,33 Importantly, the mRNA transfection efficiency did not perfectly positively correlate with the uptake efficiency. For instance, KHn of medium length performed well in cellular uptake but do not exhibit satisfactory mRNA ee, and A10 showed strong transfection performance despite poor cellular uptake. Therefore, the peptide headgroup structure and tail structure have significant influences on the performance of lipopeptides. The top-performing formulations—A10-KH12, A12-KH2, A12-KH12, and A14-KH2—were selected for further screening to deliver mEGFP in HeLa cells. Among these, A10-KH12 emerged as the optimal candidate, achieving a transfection efficiency of 71.2%, which is 5.3 times higher than that of unmodified KH12 (Fig. 2f and Fig. S9).

The lead formulation A10-KH12 was further analyzed using dynamic light scattering (DLS) and transmission electron microscopy (TEM). Compared with KH12@mRNA complexes, A10-KH12@mRNA complexes exhibited smaller (∼131 nm), more uniform nanoparticles (polydispersity index (PDI) < 0.2) and well-defined spherical morphologies favorable for cellular uptake (Fig. 2g and 2h), whereas KH12@mRNA complexes existed as irregular micron-scale aggregates (Fig. 2g, and Fig. S10). A10-KH12@mRNA also exhibited a moderate positive zeta potential (∼15 mV), supporting membrane interaction while maintaining biocompatibility (Fig. 2i). In contrast, KH12@mRNA complexes exhibited a highly negative surface charge, likely resulting from surface-bound but insufficiently encapsulated nucleic acids (Fig. 2i). mRNA ee had increased by 24.1% for A10-KH12 compared to KH12 (Fig. 2j). These results further suggest that lipidation enhances nucleic acid packaging and thereby improves transfection efficiency. Collectively, lipidation of low-molecular-weight KHn peptides is an effective strategy to enhance transfection efficiency, holding potential for the design of peptide-based functional delivery systems.

Physical lipid incorporation into KHn for improved transfection

1,2-Dioleoyl-3-trimethylammonium-propane (DOTAP), a widely used cationic lipid, was chosen for co-delivery with KHn owing to its high transfection efficiency, despite limited biocompatibility (denoted as KHn/DOTAP) (Fig. 3a).^10,33 Its cationic and amphiphilic properties enable effective nucleic acid encapsulation even at a low mass ratio (5 : 1, DOTAP to nucleic acid) (Fig. S11). For a KHn/DOTAP composite system with different KHn-to-DOTAP molar ratios (4 : 1, 3 : 2, 1 : 1, 2 : 3, and 1 : 4), gel shift and mRNA ee assay results showed that DOTAP improved mRNA condensation, particularly for the shorter KHn system (Fig. 3b, c and Fig. S12). Notably, the KH2/DOTAP system exhibited an ee comparable to pure DOTAP (∼85%), whereas the KH12/DOTAP system displayed reduced ee relative to pure DOTAP, which was opposite to the trend observed in pure KHn formulations (Fig. 1c, 3c and Fig. S12). At the same time, compared with other KHn/DOTAP@mRNA complexes, which had a similar particle size of less than 200 nm and a similar PDI of less than 0.4, the KH12/DOTAP@mRNA complexes had a significantly larger particle size and a PDI higher than 0.4 (Fig. 3d, e and Fig. S13a, S14a). We speculate that this is because short KHn have a smaller molecular volume, allowing them to more flexibly assemble with DOTAP. However, the KH12 has a larger molecular volume and a more rigid long-chain structure, which hinder its assembly with DOTAP.

Fig. 3 In vitro mRNA delivery of KHn/DOTAP. (a) Molecular structure of DOTAP, and a schematic diagram of the physical mixing process of KHn and DOTAP. (b) Gel electrophoresis effects of KHn/DOTAP@DNA complexes with different KHn/DOTAP molar ratios (1 : 4, 2 : 3, 1 : 1, 3 : 2, and 4 : 1) at neutral pH. (c)–(e) Heatmaps of the encapsulation efficiency (c), size (d), and PDI (e) of the KHn/DOTAP@mRNA formulations with different KHn-to-DOTAP molar ratios (1 : 0, 4 : 1, 3 : 2, 1 : 1, 2 : 3, and 1 : 4) (n = 3). DOTAP@mEGFP complexes were used as a control. (f) A heatmap of the cell viability of HeLa treated with the KHn/DOTAP@mRNA complex concentration of 20 µg mL⁻¹ for 48 h (n = 3). (g) and (h) Heatmaps of the EGFP expression positive rate (g) and the mean fluorescence intensity (h) in HeLa treated with KHn/DOTAP@mEGFP complexes with different KHn-to-DOTAP molar ratios (4 : 1, 3 : 2, 1 : 1, 2 : 3, and 1 : 4) after 48 h (n = 2). DOTAP@mEGFP complexes and Lipo 2000 complexes were used as controls.

Additionally, when the molar fraction of DOTAP in the KHn/DOTAP system reached 40% (KHn : DOTAP = 3 : 2) (Fig. S12b), it maintained a relatively stable high ee level (> 80%), along with the low particle size and low PDI (Fig. S13b and S14b). Notably, when the molar fraction of DOTAP increased from 40% (KHn : DOTAP = 3 : 2) to 80% (KHn : DOTAP = 1 : 4), the PDI of the complexes showed an increasing trend, i.e., the KHn/DOTAP formulation with a 3 : 2 molar ratio exhibited the best particle uniformity (Fig. S14b). These suggest that the addition of DOTAP is beneficial for the assembly of the system and mRNA, but very high content of DOTAP is detrimental to the formation of uniform particles.

Unlike the negative surface charge of the KHn@mRNA complexes, all the KHn/DOTAP complexes exhibited positive surface charge, which was beneficial for entering cells (Fig. S15). This change is likely due to the introduction of DOTAP, which improved mRNA encapsulation, resulting in a decrease in free or surface-bound mRNA; on the other hand, there might be DOTAP with strong positively charged groups distributed on the surface of the complexes. Based on these findings, we proposed a possible assembly process between KHn and DOTAP: both KHn and DOTAP possess positively charged structures, enabling them to jointly bind negatively charged nucleic acids through electrostatic interactions, forming the nucleic acid complex nucleus; different from the pure KHn system that lacks hydrophobic domains, here, DOTAP, with two hydrophobic tails, wraps around and stabilizes the formed nucleic acid complex nucleus, and therefore, there is a portion of DOTAP located on the outermost layer of the KHn/DOTAP complex. This might explain why the incorporation of DOTAP enables complexes to assemble into uniform, positively charged, small-sized nanoparticles, similar to the pure DOTAP system and significantly different from the pure KHn system (Fig. S12–S15).²⁵

The cytotoxicity test results indicated that as the content of DOTAP increased, the toxicity of the KHn/DOTAP system also increased (Fig. 3f and Fig. S16). Additionally, the KH6/DOTAP system exhibited the highest cytotoxicity (Fig. S16a). The intracellular uptake results demonstrated that incorporating DOTAP also significantly enhanced internalization, and efficiencies exceeded 90% once the DOTAP molar fraction reached 40% (KHn : DOTAP = 3 : 2) (Fig. S17 and S19).

Then, the KHn/DOTAP formulations were used to deliver mEGFP into HeLa cells to identify their transfection ability (Fig. 3g, h and Fig. S20, S21). We found that the formulations containing KH2 showed optimal transfection performance across the KHn series, along with the highest ee, the smallest particle size, the best biocompatibility, as well as satisfactorily low PDI, high intracellular uptake, and moderate zeta potential; the formulations with a molar ratio of 3 : 2 also performed best in transfection across the tail series, having the lowest PDI, along with the high ee, small size, good biocompatibility, high intracellular uptake, and moderate zeta potential, and these are all favorable characteristics for mRNA transfection (Fig. 3g, h, and Fig. S12–S21). Therefore, we speculate that these formulations achieve the best synergistic assembly effect, contributing to their high transfection performance. In contrast, the 4 : 1 formulation exhibited the worst transfection performance, which can be attributed to its suboptimal assembly structure—evidenced by the lowest ee, the largest size, and the weakest cellular uptake (Fig. 3g and h). Although the 1 : 1 and 2 : 3 formulations showed comparable characteristics to 3 : 2 formulation in ee, size, PDI, zeta potential, and cellular uptake, their delivery efficiency was still inferior to that of the 3 : 2 system, likely due to the less effective endosomal escape and cytosolic nucleic acid release (Fig. 3g, h and Fig. S12–S21). At a molar ratio of 1 : 4, the transfection performance was nearly identical to that of DOTAP alone, indicating that here DOTAP dominated the delivery performance, with KHn contributing only marginally (Fig. 3g and h). These results underscore the importance of the peptide-to-lipid ratio in achieving high transfection efficiency.

Among all 25 formulations, KH2/DOTAP_3 : 2 yielded the highest mEGFP transfection efficacy (75.9%), outperforming DOTAP alone and exhibiting 43.9-fold higher efficiency than KH2 alone (Fig. 3g and Fig. S22). It also surpassed both DOTAP and the commercial reagent Lipofectamine 2000 (Lipo 2000) in EGFP expression intensity (Fig. 3h and Fig. S22). Moreover, KH2/DOTAP_3 : 2 mediated high transfection in multiple cell lines—including CHO (Chinese hamster ovary cells), DC 2.4 (mouse bone marrow-derived dendritic cells), PC 12 (rat adrenal gland pheochromocytoma cells), and Vero (African Green monkey renal cells)—demonstrating its potential for wide application (Fig. 4a and Fig. S23).

Fig. 4 The performance and properties of the KHn/DOTAP_3 : 2 system. (a) The inverted fluorescence microscopy images of CHO, DC 2.4, PC 12, and Vero treated with KH2/DOTAP_3 : 2@mEGFP complexes after 48 h (n = 2 or 3). (b) Confocal microscopy images show the endosomal escape of the KH2/DOTAP_3 : 2@FAM-siRNA complexes (top) and the DOTAP@FAM-siRNA complexes (bottom) in HeLa. Nucleus was stained blue, acidic endosomes were stained red, and FAM-siRNA were stained green. (c) Hydrodynamic diameter and PDI of KH2/DOTAP_3 : 2@mRNA, DOTAP@mRNA, and KH2@mRNA complexes at neutral pH (n = 3). (d) The TEM image of KH2/DOTAP_3 : 2@mRNA complexes at neutral pH. (j) Zeta potentials of KH2/DOTAP_3 : 2@mRNA, DOTAP@mRNA, and KH2@mRNA complexes at neutral pH (n = 3). (f) The mRNA ee of KH2/DOTAP_3 : 2, DOTAP, and KH12 (n = 3).

We examined the endosomal escape of the complexes using confocal microscopy. As shown in Fig. 4b, the nucleus was stained blue, acidic endosomes were stained red, and FAM-siRNA was stained green. Compared with cells treated with KH2/DOTAP_3 : 2@FAM-siRNA complexes, cells treated with the DOTAP@FAM-siRNA complexes exhibited more pronounced colocalization of green and red fluorescence (appearing yellow) (Fig. 4b). This indicates improved endosomal escape of the KH2/DOTAP_3 : 2 due to the presence of KHn compared with DOTAP. Furthermore, this is most likely attributed to the pH-responsive protonation of histidine.^28,29 From the results of DLS and TEM, it can be seen that compared with the KH2@mRNA complexes and the DOTAP@mRNA complexes, the KH2/DOTAP_3 : 2@mRNA complexes had a nanoparticle structure with a favorable nanosize (94 nm), a lower PDI (0.191), and a moderate positive surface charge (28 mV) (Fig. 4c–e and Fig. S24). In terms of mRNA ee, KH2/DOTAP_3 : 2 was close to DOTAP and showed a 40.6% increase compared to KH2 (Fig. 4f). And the KH2/DOTAP_3 : 2@mRNA complexes also exhibited comparable storage stability compared to DOTAP@mRNA complexes (Fig. S25). When stored at 4 °C and pH 7.4 for 4 days, their size, PDI and zeta potential remained at relatively stable levels (Fig. S25), demonstrating their potential for practical application. Importantly, the KH2/DOTAP_3 : 2 system had better biocompatibility than the DOTAP (Fig. S26).

In conclusion, the KHn/lipid composite system combines the advantages of both, resulting in an improved assembly structure, enhanced endosomal escape and transfection efficiency, and good biocompatibility, making it a highly efficient and safe gene delivery platform with broad application potential.

In vivo mRNA delivery of screened KHn-based formulations

The two top-performing systems, A10-KH12 and KH2/DOTAP_3 : 2, were selected for in vivo mRNA delivery to evaluate their potential as gene therapy vectors. Cyanine5 (Cy5)-labeled mFluc encapsulated in A10-KH12 or KH2/DOTAP_3 : 2 nanoparticles was intravenously (i.v.) injected into C57BL/6 mice to assess their biodistribution. After 2 hours, ex vivo IVIS imaging of major organs (heart, lungs, spleen, kidneys, and liver) revealed no significant fluorescence signals in mice treated with A10-KH12@Cy5-mFluc complexes, suggesting minimal accumulation—likely due to rapid systemic clearance (Fig. 5a). Consistent with this, A10-KH12 mediated negligible mFluc expression in all examined organs (Fig. 5a and b). Therefore, individual lipopeptides struggle to penetrate multiple biological barriers and achieve effective in vivo mRNA delivery, despite their outstanding performance in in vitro delivery.

Fig. 5 In vivo mRNA delivery of A10-KH12 and KH2/DOTAP_3 : 2. (a) Fluorescence (top) and bioluminescence (bottom) imaging of major organs (heart, lungs, spleen, kidney, liver) following systemic delivery of Cy5-mFluc (2 h) or mFluc (6 h) using A10-KH12 (i.v., 0.75 mg kg⁻¹, n = 3 biologically independent animals). (b) Bioluminescence imaging of live mice after systemic A10-KH12-mediated mFluc delivery (i.v., 0.75 mg kg⁻¹, 6 h, n = 3 biologically independent animals). (c) Fluorescence (top) and bioluminescence (bottom) imaging of major organs after KH2/DOTAP_3 : 2-mediated delivery of Cy5-mFluc (2 h) or mFluc (6 h) (i.v., 0.75 mg kg⁻¹, n = 3 biologically independent animals). (d) Bioluminescence imaging of live mice after KH2/DOTAP_3 : 2-mediated systemic delivery of mFluc (i.v., 0.75 mg kg⁻¹, 6 h, n = 3 biologically independent animals). (e) and (f) Pie charts based on the average radiance statistics showing the proportional fluorescence (e) and bioluminescence (f) distribution across major organs after KH2/DOTAP_3 : 2-mediated systemic Cy5-mFluc (2 h) and mFluc (6 h) delivery (i.v., 0.75 mg kg⁻¹, n = 3 biologically independent animals). (g) Quantified average radiance of major organs after KH2/DOTAP_3 : 2-mediated systemic mFluc delivery (i.v., 0.75 mg kg⁻¹, n = 3 biologically independent animals). One-way ANOVA with Tukey's correction was used. Probability P < 0.05 was considered to be statistically significant.

In contrast, mice treated with KH2/DOTAP_3 : 2@Cy5-mFluc showed strong fluorescence signals in both the lungs and the liver, with notably higher intensity in the lungs, indicating predominant accumulation in these organs (Fig. 5c–e). And, this formulation showed a higher proportion of accumulation in the lungs compared to DOTAP@Cy5-mFluc (Fig. 5e and Fig. S27a, b). Furthermore, KH2/DOTAP_3 : 2 mediated significant mFluc expression in the lungs (Fig. 5c, d, f and g). Although the complexes also accumulated in the liver, no substantial expression was detected, possibly due to limited cellular uptake in hepatic tissue. Therefore, the KH2/DOTAP_3 : 2 system exhibited good lung-targeting specificity, with a selectivity of 74.8%, higher than that of the DOTAP system (44.2%) (Fig. 5f and Fig. S27c–e). According to previous reports, DOTAP has a pulmonary affinity, so that the lung-targeting performance of the KH2/DOTAP_3 : 2 system may be contributed by the DOTAP component.^34,35 However, the KH2/DOTAP system exhibits higher in vivo delivery efficiency and lung selectivity than the DOTAP system, which may be due to their different characteristics, including the particle size, surface charge, surface group composition, and endosome escape ability. Therefore, modulating the composition, structure and proportion of the peptide/lipid system is expected to enable the construction of more selective and efficient organ-targeting carriers. In short, the physically hybrid binary system demonstrates superior in vivo performance and targeting capability compared to the chemically lipid-modified peptide system.

Conclusions

In summary, based on the peptide–lipid complementation strategy, we initially designed five low-molecular-weight peptides (KHn) composed solely of lysine and histidine with varying lengths. These peptides have promising nucleic acid-binding and endosomal escape capabilities, along with advantages in terms of cost-effectiveness and biosafety. However, experimental results revealed that due to the lack of hydrophobic domains, the KHn exhibited unsatisfactory structural assembly and mRNA transfection performance.

To address this limitation, we incorporated lipid elements with complementary functions into the KHn system through two approaches: chemical lipidation and physical lipid incorporation. This led to improved mRNA delivery performance both in vitro and in vivo. Specifically, chemical lipidation of KHn was employed to introduce alkyl tails, which provide hydrophobic interactions and enhance membrane affinity. The resulting lipopeptides demonstrated significantly improved structural assembly, encapsulation efficiency, cellular uptake, and mRNA transfection efficiency compared to the pure peptide system, while maintaining good biocompatibility.

In parallel, the cationic lipid DOTAP—known for its strong mRNA encapsulation and cellular uptake capabilities—was physically incorporated into the KHn system. Through systematic screening in vitro and in vivo, an optimal formulation was identified that exhibited excellent mRNA encapsulation, structural assembly, and cellular internalization. Moreover, this formulation outperformed the pure DOTAP system in terms of endosomal escape, mRNA transfection efficiency in vitro and in vivo, and lung-targeting specificity, while retaining favorable biocompatibility and storage stability.

Meanwhile, extensive characterization further elucidated the significant influence of both KHn and lipid modules on the structure, properties, and performance of the carrier system. These results validate the feasibility of our strategy, demonstrating that a simple system can achieve efficient mRNA delivery. The developed formulations exhibited superior mRNA delivery performance compared to several reported peptide-based systems (Table S2). This strategy holds broad potential for the development of functional gene delivery carriers.

Ethical statement

Animal experiments were performed at Xuhe (Tianjin) Pharmaceutical Technology Co., Ltd, and were approved by the company's Ethics Committee in compliance with the Laboratory Animal Law (LLSC-2024022001).

Author contributions

Chuanmei Tang: conceptualization, data curation, formal analysis, investigation, methodology, and writing – original draft; Yuzhi Ye: data curation, investigation, and writing – original draft; Yaohui Du and Yulin Sun: investigation and methodology; Rongxin Su: supervision, validation, and visualization; Wei Qi: supervision, visualization, and funding acquisition; Yuefei Wang: conceptualization, methodology, funding acquisition, and writing – review and editing. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

All the data supporting the findings of this study are available within the article and its supplementary information (SI) and from the corresponding author upon reasonable request. Supplementary information: experimental details, additional characterization and in vivo analysis. See DOI: https://doi.org/10.1039/d5nh00656b.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 22278306 and 22278314) and the Tianjin Municipal Science and Technology Bureau (No. 23JCQNJC01870).

Notes and references

1. K. Paunovska, D. Loughrey and J. E. Dahlman, Nat. Rev. Genet., 2022, 23, 265–280.
2. X. N. Huang, Y. S. Ma, G. H. Ma and Y. F. Xia, Research, 2024, 7, 0370.
3. L. Gál, T. Bellák, A. Marton, Z. Fekécs, D. Weissman, D. Török, R. Biju, C. Vizler, R. Kristóf, M. B. Beattie, P. J. C. Lin, N. Pardi, A. Nógrádi and K. Pajer, Research, 2023, 6, 0056.
4. M. K. Monroe, H. Wang, C. F. Anderson, H. P. Jia, C. Flexner and H. G. Cui, J. Controlled Release, 2022, 348, 1028–1049.
5. P. Boisguérin, S. Deshayes, M. J. Gait, L. O'Donovan, C. Godfrey, C. A. Betts, M. J. A. Wood and B. Lebleu, Adv. Drug Delivery Rev., 2015, 87, 52–67.
6. S. Eskandari, T. Guerin, I. Toth and R. J. Stephenson, Adv. Drug Delivery Rev., 2017, 110, 169–187.
7. Q. R. Chen, L. Zhang, S. A. Stass and A. J. Mixson, Nucleic Acids Res., 2001, 29, 1334–1340.
8. L. Xie, R. Liu, X. Chen, M. He, Y. Zhang and S. Y. Chen, Front. Bioeng. Biotechnol., 2021, 9, 744657.
9. M. Lakshmanan, Y. Kodama, T. Yoshizumi, K. Sudesh and K. Numata, Biomacromolecules, 2013, 14, 10–16.
10. X. C. Hou, T. Zaks, R. Langer and Y. Z. Dong, Nat. Rev. Mater., 2021, 6, 1078–1094.
11. S. A. Dilliard and D. J. Siegwart, Nat. Rev. Mater., 2023, 8, 282–300.
12. J. J. Sonju, A. Dahal, S. S. Singh and S. D. Jois, J. Controlled Release, 2021, 329, 624–644.
13. Y. Z. Dong, K. T. Love, J. R. Dorkin, S. Sirirungruang, Y. L. Zhang, D. L. Chen, R. L. Bogorad, H. Yin, Y. Chen, A. J. Vegas, C. A. Alabi, G. Sahay, K. T. Olejnik, W. H. Wang, A. Schroeder, A. K. R. Lytton-Jean, D. J. Siegwart, A. Akinc, C. Barnes, S. A. Barros, M. Carioto, K. Fitzgerald, J. Hettinger, V. Kumar, T. I. Novobrantseva, J. N. Qin, W. Querbes, V. Koteliansky, R. Langer and D. G. Anderson, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 5753.
14. C. Y. Deng, R. Zhuang, Z. Y. Ying, J. S. Tu, X. H. Xu, C. M. Sun and L. Jiang, Adv. Funct. Mater., 2022, 32, 2206876.
15. M. Zhu, X. X. Wang, R. S. Xie, Y. Y. Wang, X. H. Xu, J. Burger and S. Q. Gong, ACS Appl. Mater. Interfaces, 2023, 15, 10464–10476.
16. L. Lei, X. L. Wang, Y. H. Zhou, R. Cong, S. S. Zhou, R. F. Gao, J. S. Tu, X. Q. Jiang and L. Jiang, Adv. Funct. Mater., 2024, 34, 2313036.
17. L. Jiang, S. S. Zhou, X. K. Zhang, C. Li, S. L. Ji, H. Mao and X. Q. Jiang, Nat. Commun., 2021, 12, 2390.
18. Y. F. Ni, W. Z. Zhao, W. J. Cheng, C. Y. Deng, Z. Y. Ying, L. Li, X. L. Wang, C. M. Sun, J. S. Tu and L. Jiang, J. Controlled Release, 2022, 351, 245–254.
19. R. Zhuang, M. Q. Chen, Y. H. Zhou, W. J. Cheng, T. H. Zhang, Y. F. Ni, C. H. Guo, J. S. Tu and L. Jiang, ACS Macro Lett., 2021, 10, 215–222.
20. Y. W. Dong, Y. Chen, D. D. Zhu, K. J. Shi, C. Ma, W. J. Zhang, P. Rocchi, L. Jiang and X. X. Liu, J. Controlled Release, 2020, 322, 416–425.
21. L. Jiang, W. Z. Chen, S. S. Zhou, C. Li, X. K. Zhang, W. Wu and X. Q. Jiang, Biomater. Sci., 2018, 6, 774–778.
22. S. Zhou, C. Li, Y. Yuan, L. Jiang, W. Z. Chen and X. Q. Jiang, Biomater. Sci., 2022, 10, 7032–7041.
23. Q. H. Xu, C. Shu, Y. C. Li, R. Zhuang, L. Jiang and X. H. Xu, Chem. Commun., 2021, 57, 4859–4862.
24. C. M. Tang, Y. X. Zhang, B. W. Li, X. W. Fan, Z. X. Wang, R. X. Su, W. Qi and Y. F. Wang, Adv. Mater., 2025, 37, 2415643.
25. J. J. Zhang, Z. X. Wang, J. W. Min, X. L. Zhang, R. X. Su, Y. F. Wang and W. Qi, Langmuir, 2023, 39, 7484–7494.
26. Y. H. Du, C. M. Tang, B. W. Li, L. Yue, R. X. Su, W. Qi, X. B. Cui and Y. F. Wang, Langmuir, 2025, 41, 12931–12942.
27. P. S. Ghosh, C. K. Kim, G. Han, N. S. Forbes and V. M. Rotello, ACS Nano, 2008, 2, 2213–2218.
28. Z. X. Wang, J. X. Zhang, Y. F. Wang, J. L. Zhou, X. H. Jiao, M. S. Han, X. L. Zhang, H. L. Hu, R. X. Su, Y. M. Zhang and W. Qi, ACS Nano, 2024, 18, 10324–10340.
29. S. E. Hooshmand, M. J. Sabet, A. Hasanzadeh, S. M. K. Mousavi, N. H. Moghaddam, S. A. Hooshmand, N. Rabiee, Y. Liu, M. R. Hamblin and M. Karimi, J. Gene Med., 2022, 24, e3415.
30. H. Bahaar, S. G. Reddy, B. S. Kumar, K. Prashanthi and H. C. Ananda Murthy, Eng. Sci., 2023, 24, 914.
31. K. T. Love, K. P. Mahon, C. G. Levins, K. A. Whitehead, W. Querbes, J. R. Dorkin, J. Qin, W. Cantley, L. L. Qin, T. Racie, M. Frank-Kamenetsky, K. N. Yip, R. Alvarez, D. W. Y. Sah, A. de Fougerolles, K. Fitzgerald, V. Koteliansky, A. Akinc, R. Langer and D. G. Anderson, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 9915.
32. Y. B. Zhang, C. Z. Sun, C. Wang, K. E. Jankovic and Y. Z. Dong, Chem. Rev., 2021, 121, 12181–12277.
33. M. Jeong, Y. Lee, J. Park, H. Jung and H. Lee, Adv. Drug Delivery Rev., 2023, 200, 114990.
34. Q. Cheng, T. Wei, L. Farbiak, L. T. Johnson, S. A. Dilliard and D. J. Siegwart, Nat. Nanotechnol., 2020, 15, 313–320.
35. T. Wei, Q. Cheng, Y. L. Min, E. N. Olson and D. J. Siegwart, Nat. Commun., 2020, 11, 3232.

---

Footnote

† These authors contributed equally to this work.

---