EDAC-mediated O-acylisourea rearrangement for tertiary amine cationization of hyaluronic acid (HA) and its application as structural backbones in virus-inspired polyplexes

Abstract

Cationic modification of hyaluronic acid (HA) is challenging due to its polyanionic nature, poor reactivity in water, and the instability of conventional coupling intermediates. This limits the development of HA-based components in non-viral gene delivery systems, which already suffer from amorphous morphology and mechanical fragility that reduce their transfection efficiency. Here, we reprogram a classically unfavorable EDAC-mediated rearrangement into a productive synthetic route, enabling direct cationization of hyaluronic acid (HA) through spontaneous O-acylisourea rearrangement. This water-based, catalyst-free process achieves up to 70 % substitution of HA's carboxyl groups—introducing cationic tertiary amine functionalities in water. The resulting aminated-hyaluronic acid (HAA) scaffolds act as rigid structural backbones in virus-inspired polymer–DNA nanoparticles termed as “Skeletoplexes”, with enhanced stability and performance. When incorporated into polyplexes formed from diverse cationic systems—including poly(β-amino esters) and commercial vectors such as BrPERfect, Xfect, jetPEI, and Lipofectamine3000—HAA scaffolds improved in vitro transfection efficiency by up to 4-fold and in vivo gene expression by approximately 2-fold. These results establish a generalizable and green scaffold-based strategy that bridges the structural and functional gap between viral and non-viral gene delivery vectors.

Graphical abstract

Highlights

• •Green, catalyst-free method enables direct cationization of HA in water.
• •EDAC-mediated O-acylisourea rearrangement reprogrammed into productive pathway.
• •Tertiary amine–rich HA acts as rigid backbone in virus-mimetic polyplexes.
• •Skeletoplexes enhance stability, robustness, and gene delivery efficiency.
• •Strategy broadly compatible with commercial and custom gene delivery systems.

1. Introduction

Efficient and safe nucleic acid delivery remains a central challenge in the realization of gene therapy [[1], [2], [3], [4], [5], [6], [7]]. While viral vectors exhibit high transfection efficiency, their use is limited by immunogenicity, insertional mutagenesis risks, and complex manufacturing requirements [8,9]. In contrast, non-viral delivery systems—comprising cationic polymers, lipid-based nanoparticles, and synthetic vectors—offer enhanced safety and scalability. However, these synthetic carriers typically suffer from structural disorder and mechanical instability, leading to premature disassembly, reduced cellular uptake, and suboptimal gene expression [[10], [11], [12], [13], [14], [15], [16], [17]].

Among non-viral approaches, polymer–DNA complexes (polyplexes), formed through electrostatic interactions between cationic polymers and nucleic acids, have garnered considerable interest due to their tuneable chemistry and design versatility [[17], [18], [19], [20], [21], [22], [23], [24], [25]]. Yet, their amorphous morphology and lack of internal rigidity severely hinder their in vivo performance [[26], [27], [28], [29], [30]]. In stark contrast, natural viruses encapsulate their genetic cargo within highly ordered, mechanically robust protein shells that ensure structural integrity throughout the delivery process [[31], [32], [33]]. Inspired by this structural precision, researchers have sought to create virus-mimetic polyplexes (“artificial viruses”) to bridge the efficiency gap between viral and non-viral systems [[34], [35], [36], [37], [38], [39], [40]].

A promising route to such architectures is the incorporation of rigid scaffolding elements into polyplexes to reinforce nanoparticle morphology and improve functional performance. Hyaluronic acid (HA), a naturally occurring, biocompatible, and semi-rigid polysaccharide, represents an appealing structural scaffold candidate. However, its polyanionic nature necessitates cationic modification to enable effective DNA binding and condensation [[41], [42], [43]]. Traditionally, HA cationization is carried out in aqueous/organic mixed solvents to improve reactivity and substitution efficiency, but this raises concerns about environmental sustainability and biocompatibility. Conventional aqueous-phase methods typically employ 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDAC)/N-Hydroxysuccinimide (NHS)-mediated coupling with exogenous amines such as ethylenediamine, PEG-NH₂, or PEI [41,44,45]. These multistep reactions are limited by the instability of O-acylisourea intermediates, demand strict pH control, and often yield low substitution degrees [46,47]. Additionally, extensive dialysis is required to remove residual amines and NHS, complicating purification, and residual molecules like PEI pose cytotoxicity risks.

In this study, we developed a green and efficient “EDAC rearrangement” strategy for direct cationization of hyaluronic acid (HA), eliminating the need for NHS or organic solvents. By reprogramming a conventionally unfavorable side reaction, the transient O-acylisourea intermediate undergoes spontaneous rearrangement in water, covalently introducing N,N-dimethylamine fragments from EDAC itself onto the HA backbone. This one-step, catalyst-free process achieves up to 70 % substitution. Leveraging the resulting aminated HA (HAA), we constructed structurally reinforced, virus-inspired “Skeletoplexes” by embedding HAA scaffolds into polymer–DNA complexes formed from hyperbranched poly(β-amino esters) (HPAE), linear poly(β-amino esters) (LPAE), and commercial transfection reagents including BrPERfect, Xfect, jetPEI, and Lipofectamine 3000. We systematically investigated the structural integration of HAA within polyplexes by analyzing nanoparticle assembly, morphology, colloidal stability, DNA condensation efficiency, and zeta potential across a range of formulations. Transfection performance was evaluated in multiple cell lines and delivery conditions, followed by in vivo assessment of gene expression and biodistribution via systemic administration in mice.

2. Results and discussion

2.1. Direct cationization of hyaluronic acid via EDAC rearrangement

Cationic hyaluronic acid (HAA) was synthesized via a one-pot aqueous reaction using EDAC as both the activating agent and cationic source (Scheme 1). In this process, sodium hyaluronate was first reacted with EDAC in water at room temperature to form O-acylisourea intermediates. After 3 h, a primary amine was introduced to promote rearrangement. Unlike conventional EDAC/NHS amidation that couples exogenous amines to carboxyl groups, this reaction proceeds predominantly through spontaneous intramolecular rearrangement of the O-acylisourea intermediate, covalently transferring dimethylamino and ethyl fragments derived from EDAC itself onto the HA backbone.

Scheme 1.

Comparison of the conventional cationic modification method of HA with the EDAC rearrangement method developed in this work.

The chemical structure and degree of substitution of the resulting HAA were characterized by ¹H NMR spectroscopy (Fig. 1). Peaks corresponding to tertiary amine protons (δ = 2.76) and ethyl protons (δ = 1.07) were clearly observed, confirming successful introduction of cationic functionalities. ¹³C NMR and ¹H–¹H COSY further verified the structure (Fig. S1 and S2). Quantitative integration indicated that up to 70 % of the carboxyl groups were substituted, with approximately 50 % corresponding to tertiary amines and 20 % to ethyl groups. The resulting HAA is rich in tertiary amine groups and retains the semi-rigid backbone of native HA, making it an ideal scaffold for nanoparticle structural reinforcement. To further verify the reaction mechanism, HAA-1 to HAA-11 were synthesized in a single-step process (Table S1 and Table S2). In all cases, no degradation was observed (Fig. S3). The NMR results showed that the reaction was significantly weakened in organic solvents, and no final rearranged products were detected (HAA-1 to HAA-5, Fig. S4–S6). Moreover, this rearrangement occurred only in the presence of amines (HAA-6 to HAA-11, Fig. S4–S6).

Fig. 1.

¹H NMR spectrum of HAA. Tertiary amine functionality was calculated from the integration of (I_d/6)/(I_g/3) × 100 %. The ethyl substitutions were calculated from the integration of (I_f/3)/(I_g/3) × 100 %.

2.2. Construction and structural-transfection evaluation of skeletoplexes

To investigate the structural effects of HAA on polymer–DNA complex formation, we constructed a new class of structurally reinforced nanoparticles termed Skeletoplexes, wherein rigid cationic HAA scaffolds were embedded into conventional polyplex systems. Polyplexes without HAA served as controls (Fig. 2A). Initially, a cationic polymer, HPAE1, synthesized according to previously reported methods [48] (described in detail in Materials and Methods, Fig. S7A and S8), was selected as a model vector for constructing both Skeletoplexes and control polyplexes.

Fig. 2.

Construction and transfection evaluation of Skeletoplexes. (A) Schematic illustration depicting structural differences between Skeletoplexes and polyplexes. (B) Particle size and (C) zeta potential of polyplexes formed by DNA and HAA. (D) Particle size and (E) zeta potential of HPAE1-based Skeletoplexes formulated with varying HAA contents. (F) Quantitative analysis of GFP expression in HEK cells transfected by HPAE1-based Skeletoplexes, and (G) corresponding statistical results from flow cytometry analysis. Transfection performance evaluation in HEK cells of Skeletoplexes based on (H) HPAE2 and (I) LPAE1 polymers. For all Skeletoplexes and polyplexes, the polymer-to-DNA mass ratio (w/w) was fixed at 20:1; Skeletoplexes with an HAA weight ratio of 0 correspond to polyplexes.

Firstly, the interaction between the HAA scaffold and DNA was assessed in the absence of HPAE1. According to Fig. 2B, mixtures of HAA and DNA exhibited large particle sizes (>2000 nm) at HAA/DNA mass ratios below 40:1, indicating poor DNA compaction by HAA alone. However, zeta potential measurements (Fig. 2C) showed that the complexes became positively charged at mass ratios above 10:1, confirming electrostatic binding. These results suggest that HAA, while not an effective condensing agent on its own, binds DNA in an extended conformation and can serve as a structural scaffold.

Then, HAA was incorporated into polyplexes by sequentially mixing DNA with HAA and HPAE1 at varying HAA-to-DNA ratios (Fig. 2D and E). Size analysis (Fig. 2D) revealed that with increasing HAA content, nanoparticle size initially increased and then subsequently decreased. At low HAA ratios (5:1–10:1), the scaffold is likely incorporated directly into the original polyplex structure, resulting in the formation of larger skeletoplexes. However, at higher HAA ratios (above 20:1), the addition of HAA dispersed the complexes into smaller structures. In these conditions, each skeletoplex carried a smaller amount of DNA, but the overall system contained more nanoparticle units. This trend indicates a balance: with increasing HAA content, the system transited from fewer large particles to many smaller nanoparticles. It is worth noting that this transition did not affect the zeta potentials of the nanoparticles (Fig. 2E), which remained positive and stable across all formulations. SEM was employed to further examine the structure of the skeletoplexes (Fig. S9). Samples were prepared at HPAE1/HAA/DNA ratios of 20:0:1 and 20:20:1. To prevent charge accumulation, each sample was coated with approximately 2 nm of platinum prior to imaging. However, due to residual water within the skeletoplexes, volume changes occurred during vacuum drying and metal coating, leading to varying degrees of surface swelling and cracking. In comparison, the skeletoplexes containing HA exhibited more pronounced swelling and obvious surface cracking, whereas those lacking HA displayed a relatively smoother morphology with only minor local bulging. This difference indicates that the incorporation of HA substantially increases the water retention capacity of the skeletoplexes, suggesting the formation of a thicker hydration layer. Such a hydration layer is beneficial for enhancing nanoparticle stability, preventing interparticle aggregation, reducing plasma protein adsorption during transfection, decreasing recognition by the immune system, and ultimately improving biocompatibility while lowering cytotoxicity and inflammatory responses.

Subsequently, the cellular uptake (Fig. S10), endocytic pathways (Fig. S11), and lysosomal escape capability (Fig. S12) of the skeletoplexes constructed based on HPAE1 were further investigated. The results showed that the addition of an appropriate amount of HAA (5:1–20:1) enhanced the cellular uptake of the skeletoplexes, while a further increase in HAA content inhibited uptake (Fig. S10). This uptake trend corresponds to the architecture shift induced by HAA content, where moderate scaffold density improves particle compactness and membrane interaction, while excessive scaffold leads to overhydration and reduced cell association. However, regardless of whether HAA was included, both the polyplexes and the skeletoplexes were internalized mainly through macropinocytosis (Fig. S11). It is worth noting that the formation of skeletoplexes greatly enhanced their lysosomal escape capability, even when only a 5:1 ratio of HAA was used (Fig. S12). This improvement is consistent with the increased internal structural stability imparted by the HAA scaffold, which allows the nanoparticles to better withstand endosomal stress and facilitates endosomal membrane disruption.

To further investigate whether nanoparticles with an integrated scaffold exhibited enhanced transfection efficiency, these HAA-HPAE1-based skeletoplexes and control polyplexes were applied to transfect HEK cells with the GFP-coding commercial plasmid gWiz (Fig. 2F). The results showed significantly improved transfection upon introduction of the HAA scaffold. Importantly, HAA alone showed no inherent transfection capability (Fig. S13, tested over HAA/DNA ratios of 5:1–50:1), suggesting that improvements in transfection were solely due to the structural transformation of nanoparticles. However, further increases in HAA content resulted in diminished transfection efficiency. These observations were confirmed via flow cytometry analysis (Fig. 2G).

To evaluate the generalizability of this skeletoplex strategy, additional formulations were prepared using structurally distinct polymers: a hyperbranched cationic polymer, HPAE2, composed of different monomers to HPAE1 (Fig. S7B and S8)., and a linear cationic polymer, LPAE1 (Fig. S7C and S8) (both synthesized according to procedures detailed in Materials and Methods). Similar improvements were consistently observed (Fig. 2H and I), whereby incorporating the HAA scaffold into polyplexes to form skeletoplexes significantly enhanced transfection performance across different polymers. Specifically, the HPAE2-based skeletoplexes showed an initial increase followed by a decrease in transfection efficiency, peaking at an HAA/DNA mass ratio of 20:1. Conversely, the softer, linear polymer LPAE1 showed progressively improved transfection efficiency with increasing HAA content. This intriguing finding clearly indicates that incorporating a rigid scaffold into relatively soft polyplex structures to form skeletoplexes resulted in improved transfection performance, particularly pronounced in nanoparticles formed from softer polymers such as LPAE1.

In addition, the effect of the skeletoplex structure on mRNA delivery was also evaluated. By incorporating different amounts of HAA into LPAE2 (Fig. S8), the transfection efficiency in HEK and A549 cells was examined. The results showed that once the skeletoplexes were formed, the mRNA delivery efficiency of LPAE2 was significantly improved (Fig. S14).

2.3. Universality of skeletoplex strategy in commercial gene delivery vectors

To further investigate whether the enhanced transfection performance conferred by the Skeletoplex structure is universally applicable, commercially available cationic polymer vectors (BrPerfect, Xfect, and jetPEI) were selected as base materials to construct nanoparticles. HAA scaffolds were introduced into these systems to structurally convert them into Skeletoplexes. The transfection efficiencies of these newly prepared Skeletoplexes were evaluated across three commonly used cell lines (HEK, A549, and HeLa cells), as shown in Fig. 3.

Fig. 3.

Transfection performance evaluation of Skeletoplexes constructed from commercial vectors. Transfection efficiency of Skeletoplexes formulated using commercial reagents (A) BrPERfect, (B) Xfect, (C) jetPEI, and (D) Lipofectamine3000 in HEK cells. Transfection efficiency of Skeletoplexes formulated using commercial reagents (E) BrPERfect, (F) Xfect, (G) jetPEI, and (H) Lipofectamine3000 in A549 cells. Transfection efficiency of Skeletoplexes formulated using commercial reagents (I) BrPERfect, (J) Xfect, (K) jetPEI, and (L) Lipofectamine3000 in HeLa cells.

The results demonstrated that incorporating the HAA scaffold into the formulation of these three commercial polymer vectors significantly enhanced their transfection efficiencies to varying degrees, provided that the HAA ratios were optimized appropriately. Interestingly, the optimal HAA-to-DNA ratios varied across different cell types, possibly reflecting differences in nanoparticle internalization mechanisms and the physical or mechanical properties inherent to each cell line. Generally, for BrPerfect, a branched cationic polymer, the best transfection performance was observed at lower HAA ratios, specifically within the range of 5:1–10:1 (Fig. 3A, E, 3I). In contrast, linear polymers such as Xfect (Fig. 3B, F, 3J) and jetPEI (Fig. 3C, G, 3K) achieved optimal transfection efficiency at relatively higher HAA ratios, typically ranging from 20:1 to 40:1.

The structural enhancement strategy was further extended to liposomal vectors. Lipofectamine3000, a commercially available lipid-based transfection reagent, was modified by incorporating the HAA scaffold into its formulation, converting it into a Skeletoplex structure (the detailed preparation protocol is described in Materials and Methods). The results (Fig. 3D, H, 3L) indicated a notable improvement in transfection efficiency upon introduction of the scaffold, consistent with observations made using cationic polymer-based vectors. However, optimal ratios varied substantially between different cell lines, falling within the broader range of 5:1–40:1. Collectively, these results highlight the versatility of the Skeletoplex strategy and demonstrate its ability to enhance gene delivery across diverse carrier architectures and biological environments.

2.4. High-concentration preparation and In vitro transfection performance evaluation of skeletoplexes

For practical applications, gene delivery nanoparticles must remain stable at high concentrations required for in vivo and clinical use. To evaluate the performance of Skeletoplexes under such conditions, we prepared HPAE1-based formulations at a 7-fold higher DNA concentration compared to standard in vitro settings, and investigated their DNA encapsulation, colloidal stability, surface charge, and transfection efficiency (Fig. 4).

Fig. 4.

Characterization and transfection performance of Skeletoplexes formulated at high concentration. (A) DNA encapsulation efficiency, (B) particle size, and (C) zeta potential of Skeletoplexes with varying HAA content. Transfection efficiency of Skeletoplexes with different HAA contents in (D) A549, (E) HEK, and (F) HeLa cells. Unlike Fig. 2, which reports transfection under standard low-concentration preparation, Fig. 4D–F presents transfection results for high-concentration formulations to evaluate performance under conditions relevant for in vivo dosing, where nanoparticle stability is a limiting factor.

Initially, the DNA encapsulation capability of Skeletoplexes at high concentration was assessed. As shown in Fig. 4A, DNA encapsulation efficiency slightly increased with rising HAA content, reaching >70 % at an HAA-to-DNA mass ratio of 50:1. Dynamic light scattering measurements of nanoparticle sizes (Fig. 4B) revealed that particle diameters progressively increased with HAA content, from approximately 220 nm (polyplex without HAA) to about 380 nm. Interestingly, at an HAA-to-DNA mass ratio of 50:1, the particle size slightly decreased again (around 350 nm). However, the nanoparticle distribution results (Fig. S15) showed that when the HAA ratio exceeded 20:1, large aggregates with sizes of approximately 2–6 μm appeared, which were formed due to excessive HAA binding. To further examine nanoparticle stability, polyplexes and Skeletoplexes were incubated at 37 °C in cell culture medium. After 4 h, the polyplexes exhibited significant aggregation, with notable increases in particle size. In contrast, Skeletoplexes showed remarkable improvement in stability upon the addition of HAA, maintaining relatively stable particle sizes without significant enlargement over time; in fact, sizes even decreased slightly at higher HAA contents.

Subsequently, the zeta potential and its stability over time were evaluated (Fig. 4C, Fig. S16). Results indicated a slight increase in nanoparticle zeta potential upon addition of HAA. However, excessively high HAA ratios (>40:1) slightly reduced the zeta potential. Notably, there was a clear difference in zeta potential stability between polyplexes and Skeletoplexes. The zeta potential of polyplexes rapidly decreased over time, dropping to about 17 mV after 4 h, possibly contributing to their aggregation. In contrast, Skeletoplexes maintained higher and more stable zeta potentials. Even at the highest HAA ratio tested (50:1), Skeletoplexes retained a zeta potential greater than 30 mV after 4 h of incubation. These findings indicate that transitioning from polyplex to Skeletoplex effectively prevented nanoparticle aggregation, which is particularly beneficial for high-concentration nanoparticle formulations.

The stability advantage of Skeletoplexes was further examined through in vitro transfection experiments. Nanoparticles prepared under high-concentration conditions were employed to deliver gWiz plasmid into A549 (Fig. 4D), HEK (Fig. 4E), and HeLa (Fig. 4F) cells. GFP expression analysis revealed that, compared to polyplexes, the introduction of the HAA scaffold to form Skeletoplexes significantly improved transfection efficiency at relatively low ratios (5:1–20:1). However, the optimal HAA ratio and the extent of transfection enhancement varied among different cell types. For A549 cells, the optimal transfection efficiency was achieved at an HAA ratio of 20:1, resulting in a nearly 4-fold enhancement compared to conventional polyplexes. These results validate the structural reinforcement function of HAA under clinically relevant formulation conditions.

2.5. In vivo transfection performance evaluation of skeletoplexes

The in vivo transfection performance of HPAE1-based polyplexes and skeletoplexes was evaluated via tail vein injection into mice. As shown in Figs. 5A, 5B, S17A, clear luciferase expression was observed in all groups 6 h post-injection when delivering luciferase-encoding DNA. Quantitative analysis (Fig. 5B) revealed that with the transformation from polyplex to Skeletoplex and the increase in HAA content, transfection efficiency initially increased but subsequently declined. Peak luciferase expression was achieved at an HAA-to-DNA ratio of 10:1, resulting in approximately a two-fold increase in transfection performance compared to conventional polyplexes.

Fig. 5.

In vivo transfection performance evaluation of nanoparticles. (A) Bioluminescence images of hairless, immunocompetent mice at 6h after tail-vein injection of polymers complexed with DNA encoding luciferase. (B) Quantitative analysis of the corresponding bioluminescence signals. (C) Luminescence images of major organs of mice post DNA. (D) Quantitative analysis of the corresponding bioluminescence signals from different organs. (E) Quantitative analysis of bioluminescence intensity over time, normalized to the signal at 6 h. (F) Quantitative analysis of the original (non-normalized) bioluminescence signal. 40 μg DNA per mouse. n = 3.

Subsequently, mice were dissected, and luciferase expression was compared across different organs including the heart, liver, spleen, lung, and kidney (Fig. 5C and D, S17B). Quantitative analysis (Fig. 5D) demonstrated that the incorporation of the HAA scaffold, converting polyplexes into Skeletoplexes, altered their organ-targeting profiles. The strongest luciferase expression was consistently observed in the spleen, with notable expression also detected in the lung and liver. The long-term gene expression over time (6 h–72 h) was further investigated. The results (Fig. 5E and F) showed that after reaching a peak at 6 h, luciferase expression gradually decreased. However, when the HAA ratio was 5:1, the longest-lasting expression was observed, with approximately 30 % of the initial 6 h luciferase expression still remaining at 72 h post-transfection. At 72 h post-transfection, the luciferase expression levels across different organs were nearly comparable (Fig. S18). Next, transfection outcomes at different doses were evaluated while keeping the ratio between DNA and other components constant (HPAE1/HAA/DNA = 20:10:1). The DNA dose per mouse was gradually increased from 40 μg to 100 μg. The results showed that as the dose increased, the overall luciferase expression exhibited a slight increase (Fig. S19). In addition, the organ distribution at 6 h shifted from being primarily in the spleen to being primarily in the liver (Fig. S20). Finally, an in vivo safety assessment was conducted. Histological analysis of the liver and kidney at different doses (40 μg–100 μg DNA per mouse) showed no noticeable toxicity (Fig. S21).

Taken together, both the in vitro and in vivo results show that transfection exhibits a ratio-dependent profile, with an optimal HAA/DNA ratio of 10–20:1 in vitro and 10:1 consistently providing both the highest and most sustained gene expression in vivo. These results suggest that transforming polyplexes into Skeletoplexes via incorporation of the HAA scaffold represents a beneficial strategy that enhances transfection efficiency without significantly altering their intrinsic organ-targeting characteristics, thereby eliminating the need for additional targeting adjustments for intended organ transfection applications.

3. Conclusion

This work reported an innovative synthetic method—the one-step aqueous EDAC rearrangement—for facile, efficient, and environmentally benign synthesis of rigid, cationically functionalized hyaluronic acid scaffolds (HAA). By eliminating conventional organic solvents, catalysts, and auxiliary reagents, our synthesis method substantially simplifies the production and purification processes. The resultant rigid HAA scaffolds, when integrated into standard polymer-DNA polyplexes, effectively transformed them into structurally reinforced, virus-inspired "Skeletoplexes". These novel nanoparticles exhibited significantly enhanced mechanical robustness, colloidal stability, and transfection efficiency both in vitro and in vivo. Remarkably, Skeletoplexes maintained optimal biological activity and original organ-targeting profiles, demonstrating the broad applicability and versatility of our scaffold-based structural enhancement strategy. This synthesis-driven innovation provides a robust and universally adaptable platform for developing stable and potent non-viral vectors, paving the way for advanced gene therapy applications.

CRediT authorship contribution statement

Yinghao Li: Writing – original draft, Visualization, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Liang Yao: Methodology, Investigation. Jiahao Liu: Methodology, Investigation. Yi Situ: Methodology. Chunyu Zhao: Visualization, Methodology. Tianyu Mao: Investigation, Methodology. Xi Wang: Resources, Methodology. Rijian Song: Resources, Methodology. Hongyun Tai: Validation, Resources. Zhonglei He: Validation, Resources. Jing Lyu: Writing – review & editing, Visualization, Validation, Project administration. Wenxin Wang: Supervision, Project administration, Funding acquisition.

Ethics approval and consent to participate

All animal experiments were approved by the Biomedical Research Ethics Committee of Anhui University of Science and Technology (Approval No. SZ2024-004) and conducted in accordance with relevant guidelines and regulations.

Declaration of competing interest

The authors declare the following personal relationships which may be considered as potential competing interests: Yinghao Li, Xi Wang and Hongyun Tai are currently employed by Branca Bunús Ltd.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Multimedia component 1

ASSOCIATED CONTENT: GPC traces and chemical structures of the synthesized PAEs. In vivo transfection performance of complexes with different HAA weight ratios. Additional characterization data, including NMR spectra, particle size and zeta potential distributions, cellular uptake analysis, endosomal escape evaluation, and biocompatibility assessments, are also provided.