Injectable Zn²⁺-Driven carrier-free hydrogel via self-assembly of natural small molecules for liver fibrosis therapy through concurrent autophagy activation and ferroptosis inhibition

Abstract

Activation of autophagy represents a critical pathway for eliminating activated hepatic stellate cells (HSCs). However, excessive autophagy induction may trigger ferroptosis, thereby aggravating liver damage. This study proposes synergistic regulation of HSC autophagy and ferroptosis as a novel strategy against liver fibrosis. We engineered an injectable Zn²⁺-coordination supramolecular hydrogel (termed ASA-Gel) through self-assembly of natural bioactive compounds artesunate (ASA) and glycyrrhizic acid (GA) without exogenous carriers. What's more, Zn²⁺ in ASA Gel could significantly reduce the concentration of hydrogel. Molecular dynamics simulations reveal that ASA Gel formation is primarily governed by Zn²⁺-carboxylate coordination bonds and intermolecular hydrogen bonding. In CCl₄-induced fibrotic mice, ASA Gel showed superior efficacy to ASA monotherapy and GA-Zn²⁺ complexes. Molecular profiling revealed that ASA Gel upregulated Atg9A and Atg101 in the liver, thereby promoting autophagosome formation, inhibiting P62 expression, and enhancing autophagic flux. Moreover, ASA Gel significantly promoted the expression of GPX4 and SLC7A11 genes, indicating that it inhibited the ferroptosis process in liver tissue. This work establishes a carrier-free hydrogel platform for spatiotemporally coordinated regulation of autophagy and ferroptosis in HSCs.

Graphical abstract

1. Introduction

Liver fibrosis is a maladaptive repair process triggered by chronic injury, characterized by excessive deposition of scar tissue and architectural distortion of the hepatic parenchyma [1,2]. In this pathological process, activated hepatic stellate cells (HSCs) serve as the primary drivers of fibrosis by transdifferentiating into collagen-producing myofibroblasts [3]. This activated state is perpetuated through crosstalk between HSCs and immune cells: HSC-derived pro-inflammatory cytokines such as IL-6 and TGF-β recruit macrophages and neutrophils, forming a self-amplifying loop of inflammation and extracellular matrix (ECM) accumulation [4,5]. Notably, autophagy, a key anti-inflammatory pathway in macrophages including Kupffer cells, can disrupt this loop by limiting the production of IL-1β (a critical factor maintaining HSC survival and profibrotic phenotype) [6]. Thus, activating macrophage autophagy has been confirmed to possess significant anti-inflammatory and anti-fibrotic potential [7,8], emerging as a core target for anti-fibrotic therapy (see Scheme 1).

Scheme 1.

Schematic illustration of ASA Gel and anti-liver fibrosis in mice. ASA Gel, a injectable Zn²⁺-driven carrier-free hydrogel via self-assembly of natural small molecules, has excellent anti-liver fibrosis activity through concurrent autophagy activation and ferroptosis inhibition.

The anti-fibrotic role of autophagy is not limited to immune cells; it also plays a crucial role in hepatocytes. Studies have shown that autophagy provides essential survival signals for hepatocytes, thereby maintaining normal liver function [9]. Loss of autophagic function induces spontaneous hepatomegaly and liver injury, while various autophagy activators have been proven to reverse this pathological process [10,11]. In recent years, the crosstalk between autophagy and iron metabolism has become a new research hotspot, particularly its interaction with ferritin deposition—autophagy has been identified as a key target regulating cellular susceptibility to ferritin deposition [12]. This discovery positions the "autophagy-ferroptosis axis" as an innovative and promising dual-target strategy against fibrosis [13], providing a new direction for advancing the understanding of anti-fibrotic mechanisms.

Given the central regulatory role of autophagy in liver fibrosis, identifying highly efficient and target-specific autophagy activators has become a critical task in translational medicine. Artesunate (ASA), an antimalarial sesquiterpene lactone derivative, exhibits excellent autophagy-activating activity; however, its clinical application is limited by poor water solubility and off-target distribution in vivo [14]. To address this challenge, carrier-free natural small molecule self-assembly technology offers a breakthrough solution. This technology can directly convert natural active small molecules into hydrogels without adding exogenous pharmaceutical excipients [15,16], and has shown promising applications in anti-inflammation, wound healing, and anti-tumor fields [15,17]. Among these, glycyrrhizic acid (GA), a licorice-derived triterpenoid glycoside, can self-assemble into low-molecular-weight hydrogels due to its unique amphiphilic structure (a hydrophobic 18β-glycyrrhetinic acid core and two hydrophilic glucuronic acid units) [18]. Such hydrogels possess ideal properties including injectability, biodegradability, and tissue shape adaptability [19]. More importantly, GA itself has dual therapeutic values: it can directly inhibit HSC activation to alleviate liver fibrosis [20], and improve the hepatic microenvironment through anti-inflammatory and hepatoprotective effects [21,22]. Additionally, zinc ions (Zn²⁺), essential cofactors for maintaining enzymatic functions and oxidative balance [23,24], can exert a synergistic effect with GA by reducing its hydrogel formation threshold, further optimizing hydrogel performance.

Based on the above research foundation, this study constructed a novel carrier-free ternary hydrogel (ASA Gel) through supramolecular self-assembly of GA, Zn²⁺, and ASA. In this system, GA forms a nanofibrous network via coordination between glucuronic acid and Zn²⁺, significantly reducing the critical gelation concentration. In a carbon tetrachloride-induced liver fibrosis mouse model, ASA Gel achieved sustained hepatic drug release, reducing collagen deposition by 79.99 % (significantly superior to the 48.19 % in the free ASA group) without obvious hepatotoxicity. Mechanistic studies revealed that ASA Gel exerts efficient anti-fibrotic effects by synergistically activating autophagy and inhibiting ferroptosis in HSCs. This study pioneered a mechanism-driven nanotherapeutic paradigm, integrating the advantages of carrier-free delivery, autophagy activation, and ferroptosis inhibition, thus providing new insights for the clinical treatment of liver fibrosis.

2. Material and methods

2.1. Preparation of ASA gel

Mechanistic studies revealed that specific metal ions differentially regulate the self-assembly of glycyrrhizic acid (GA). Due to the amphiphilic structure, GA hydrogel formation was observed at a minimum concentration of 1.5 % in water (Fig. S1). As evidenced in Supporting Information (Fig. S2, Table S1), divalent cations (Zn²⁺) significantly enhance GA gelation at subthreshold concentrations. In contrast, monovalent Ag⁺ exhibited negligible effects, while bivalent Fe²⁺ induced irreversible precipitation due to excessive charge screening. Notably, trace Zn²⁺ (0.1 mg mL⁻¹) demonstrated exceptional efficacy, triggering GA hydrogelation at the lowest recorded concentration (0.75 %) as validated by tube inversion rheology.

ASA Gel was prepared via a one-pot procedure exploiting the amphiphilic self-assembly of glycyrrhizic acid (GA) as the structural scaffold. ZnSO₄ was dissolved in deionized water to obtain 0.1 mg ml⁻¹ ZnSO₄ aqueous solution. GA was dissolved in the prepared ZnSO₄ in 7.5 mg mL⁻¹ and ASA was dissolved in absolute ethanol in 7 mg mL⁻¹ to obtain solutions of GA and ASA, respectively.

The GA-Zn²⁺ mixture was heated to 60 °C to reduce viscosity, after which the ASA-ethanol solution was introduced at a 1:1 M ratio (GA:ASA) under continuous stirring to facilitate ethanol evaporation. Upon cooling to 25 °C, the ternary system (GA/Zn²⁺/ASA) underwent rapid co-assembly driven by Zn²⁺-carboxyl coordination and ASA-GA hydrophobic interactions, forming a viscoelastic hydrogel. Gelation integrity was confirmed by tube inversion rheology, where the absence of flow after 60 s inversion indicated successful hydrogel formation.

2.2. Characterization of ASA gel

The structural and mechanistic characterization of ASA Gel was systematically executed through a multimodal analytical platform: Zeta potential measurements (Malvern Zetasizer Nano ZS90, UK) at 25 °C, while scanning electron microscopy (ZEISS SUPRA55 SEM, Germany) resolved its hierarchical porous morphology. Rheology tests were performed on the rheometer (MCR 302, Aaton paar, Austria) to investigate mechanical properties and stability of ASA Gel. FT-IR spectra were recorded on a Fourier transform infrared spectrometer (NicoletiS10, Thermo, USA) with the range of 4000-400 cm⁻¹. ¹H NMR (Avance IIIHD 400 MHz spectrometer, Bruker, America) spectra were used to ensure the formation mechanism between GA and ASA. MD simulation was performed by GROMACS 2019.6 software to reveal the self-assembly mechanism of ASA Gel.

2.3. Animal experiments

C57BL/6J mice (8 weeks old, male, 18–22 g) were purchased from SIBEIFU Biotechnology Co, Ltd. (Beijing, China). Mice had free access to normal chow diet and sterile water. Mice were randomly divided into seven groups (n = 5). They were given olive oil (Control), CCl₄ (Model), GA-Zn (15 mg kg⁻¹), ASA free (14 mg kg⁻¹), ASA Gel-L (7.25 mg kg⁻¹), ASA Gel-M (14.5 mg kg⁻¹) and ASA Gel-H (29 mg kg⁻¹), respectively. CCl₄ was dissolved in olive oil (13 % v·v⁻¹, 1 mL kg⁻¹). For CCl₄ model, mice were injected intraperitoneally with CCl₄ twice a week for 4 weeks. GA-Zn, ASA and ASA gel were injected intraperitoneally to mice once a day for 4 weeks. At the end of the experiment, mice were anesthetized with isoflurane and executed for blood and liver and main organs extraction. All animal studies and procedures were approved by the Life Science Ethics Committee of the First Affiliated Hospital of Zhengzhou University (No. 2024-KY-0912).

2.4. Immunohistochemistry (IHC) assay and immunofluorescence (IF) staining

Model and all treatment groups' hepatic tissues embedded in paraffin and sliced into thick sections. And then proceeded IHC assay and IF staining, methods were supplied in supporting information.

2.5. In vivo biodistribution of ASA gel in liver fibrosis model mice assessed by fluorescence imaging

C57BL/6J mice (8 weeks old, female, 18–22 g) were purchased from SIBEIFU Biotechnology Co, Ltd. (Beijing, China). The liver fibrosis model mice was induced through intraperitoneal administration of CCl₄ (dissolved in olive oil, 13 % v·v⁻¹, 1 mL kg⁻¹) twice weekly for a duration of 4 weeks. Mice were injected with DiR-ASA Gel and free DiR via the tail vein. At the predetermined time points (0, 1, 2, 4, 8, 12, 24 and 48 h), the mice were anesthetized and underwent fluorescence imaging using an in vivo spectroscopic imaging system (IVIS spectrum, PE). To further analyze the in vivo distribution of ASA Gel, the mice were euthanized 48 h post-injection, and major organs were collected for ex vivo imaging.

2.6. Western Blot

Cellular and tissue proteins were extracted by RIPA lysis, and protein quantification was performed using the BCA method. Then 10 μg of denatured proteins were loaded into Tris-Gly gels for electrophoresis. The gel was then transferred to a PVDF membrane, closed with 5 % BSA, and incubated with primary and secondary antibodies. Finally, the bands were visualized using ECL chemiluminescent reagents. Image Lab software was used for quantitative analysis. Antibodies aganst α-SMA (80008-1-RR, 14395-1-AP), Collagen I (14695-1-AP, 67288-1-Ig), GAPDH (60004-1-l g), GPX4 (67763-1-Ig), SLC7A11 (26864-1-AP) and P62 (18420-1-AP) were obtain from Proteintech Group, Inc. (China).

2.7. RNA sequencing analysis of liver

Liver tissues of control, model, and ASA gel groups were collected for RNA-seq and qPCR analysis. The detailed steps were available in the Supporting Information (Tables S2 and S3).

2.8. Cell culture

LX-2 cell line were obtained from the Chinese Academy of Medical Sciences & Peking Union Medical College and cultured in DMEM (Thermo Fisher Scientific) containing 10 % (v·v⁻¹) dialyzed FBS, 100 U mL⁻¹ penicillin and 100 mg mL⁻¹ streptomycin.

The cytotoxicity of these compounds was tested on LX-2 cells in vitro by the Cell Counting Kit-8 (CCK-8) assay. In short, the cells were plated onto 96-well plates, 100 μL of cells with a density of 8 × 10³ cells·mL⁻¹ were added to per well, and incubated for 24 h at 37 °C with 5 % CO₂. Then the cells were exposed to various concentrations of the tested compounds and incubated for 24 h. The CCK-8 solution (20 μL, 5 mg mL⁻¹) was added to each well. After 4 h of incubation, the solution was throwed away and 150 μL of DMSO was added to dissolve the formazan crystals. The absorbance was measured at a wavelength of 450 nm. Wells without drugs were used as the blanks. The IC₅₀ values were defined as the concentration of compounds that produced a 50 % proliferation inhibition of surviving cells and calculated with the GraphPad Prism 8. The inhibitory rate was calculated in the following equation:

$$\%\text{Inhibition}=100\%\times [1-\left(\text{OD}\_\left(\text{Sample}\text{group}\right)-\text{OD}\_\left(\text{Blank}\text{group}\right)\right)/\left(\text{OD}\_\left(\text{Control}\text{group}\right)-\text{OD}\_\left(\text{Blank}\text{group}\right)\right)$$

2.9. Establishment of cell models for liver fibrosis and drug administration

The cells were plated onto 6-well plates, 2 mL of cells with a density of 1 × 10⁵ cells·mL⁻¹ were added to per well, and incubated for 24 h at 37 °C with 5 % CO₂. After the cells adhere to the plate, replace with serum-free, double-antibiotic-free DMEM medium for 12 h. ASA, GA-Zn and ASA Gel group added TGF-β1 5 ng mL⁻¹. During the drug administration period, continue to use serum-free, double-antibiotic-free DMEM medium for culture. After 48 h of culture, discard the culture medium, wash the cells twice with PBS, then collect the cells for the nest experiments.

2.10. Statistical analysis

All data were expressed as mean ± SD and analyzed using the Graphpad Prism (version 10.4.0). The significance was tested with one-way analysis of variance (ANOVA) among three or more groups. Statistical significance was set at ∗P < 0.05, ∗∗P < 0.01 and ∗∗∗P < 0.001.

3. Results and discussion

3.1. Morphological and mechanical characteristics of ASA gel

The ternary carrier-free ASA hydrogel was prepared through coordinated self-assembly of GA, Zn²⁺, and ASA (Fig. 1A) in aqueous media. As illustrated in Fig. 1B, heating the homogeneous mixture to 60 °C followed by cooling to 25 °C induced rapid supramolecular organization, yielding a transparent hydrogel. The zeta potential of ASA Gel dissolved in aqueous solution was −41.9 ± 0.1156 mV (Fig. 1C), indicating its stability due to electrostatic repulsion. Although the self-assembly behavior of GA has been previously reported [15], we found that Zn²⁺ could greatly promote and enhance the formation of GA Gel (Fig. S2). Furthermore, scanning electron microscopy (SEM) observations revealed that the ternary ASA gel possessed an irregular fibrous structure (Fig. 1D). The Gel was stable, transparent and had the best plasticity under a GA: Zn²⁺: ASA molar ratio of 7 : 1: 7. Furthermore, the Gel made it possible to shape a variety of patterns and letters, such as "LOVE" through syringe at room temperature (Fig. 1E).

Fig. 1.

Characteristics of ASA Gel. A Chemical structure of ASA and GA. B Reversible gel-sol transitions of ASA Gel triggered by shear stress and temperature. C Zeta potential of ASA Gel. D Tyndall effect and SEM image of ASA Gel. E Photographs of the injectability and moldability of ASA Gel. F Rheological analysis of temperature of ASA Gel. G Viscosity measurement of ASA Gel, measured at 0.1 % strain. H Dynamic frequency sweep of the ASA Gel, measured at 0.1 % strain.

The comprehensive rheological profiling of the ternary carrier-free ASA Gel revealed its multifunctional mechanical adaptability and stability (Fig. 1F–H). The viscoelastic properties of the ASA Gel were investigated through rotational rheology measurements. The viscosity of the ASA Gel decreased with increasing frequency, which made it injectable by a syringe The viscosity of the ASA Gel decreased with increasing rotation frequency, which made it injectable by a syringe (Fig. 1G). Additionally, dynamic frequency sweeps showed that the value of G′ was approximately tenfold greater than G'' (Fig. 1H), confirming its structural stability. All results revealed that the ASA Gel had perfect gel properties, including viscoelastic property, injectability, stability. The current results were consistent with the recent reports that GA-component hydrogel had good physical and chemical properties [15].

3.2. Self-assembly mechanism of ASA gel

The Fourier transform infrared (FT-IR) spectra provided critical insights into the molecular interactions (Fig. 2A). In free GA, as well as in the physical mixture of glycyrrhizic acid, zinc ions and artemisinin ester, characteristic peaks of GA were observed at 3233.60 cm⁻¹ (-OH stretching vibration), 1722.94 cm⁻¹ (C=O stretching), 1599.11 cm⁻¹ and 1422.05 cm⁻¹ (asymmetric and symmetric COO⁻ stretching, respectively), and 1036.83 cm⁻¹ (C-O stretching of glucuronic acid). After glycyrrhizic acid forms a hydrogel with zinc ions (Zn²⁺), significant spectral shifts were detected: the -OH peak exhibited a blue shift to 3202.69 cm⁻¹, while the COO⁻ peaks underwent notable changes in both shape and intensity. These observations strongly suggest that the carboxyl groups of GA's glucuronic acid serve as primary binding sites for Zn²⁺ through coordination bonds.

Fig. 2.

Self-assembly mechanism of ASA Gel. A FT-IR, GA-ASA and Free ASA. B¹H NMR spectrum of GA.

For free ASA, two distinct C=O stretching peaks (attributed to vibrational coupling) were observed. However, after gel formation, only a single residual peak at 1725.00 cm⁻¹ remained. Concurrently, the C-O stretching peak of GA-Zn at 1032.37 cm⁻¹ shifted to 1007.80 cm⁻¹ in ASA Gel, accompanied by a marked enhancement of the C=O absorption intensity. These changes indicate successful assembly between ASA and GA-Zn, likely mediated by hydrogen bonding between ASA's carboxyl group and the hydroxyl groups of GA's glucuronic acid. To further validate these interactions, ¹H NMR spectroscopy was employed (Fig. 2B). A pronounced alteration in the proton peaks of GA's glucuronic acid was observed upon ASA binding, coupled with the complete disappearance of the carboxyl proton signal in ASA. This result aligns with the FT-IR data, confirming that ASA's carboxyl group engages in hydrogen bonding with the hydroxyl groups of GA's glucuronic acid during gel assembly.

To further elucidate the self-assembly dynamics of ASA Gel, molecular dynamics (MD) simulations were performed using a system containing GA, Zn²⁺, and ASA at a molar ratio of 7 : 1: 7. The root-mean-square deviation (RMSD) analysis revealed that all molecules tended to be stable after 20 ns (Fig. 3A), indicating stable aggregation formation. Concurrently, the solvent-accessible surface area (SASA) values demonstrated a progressive reduction over the simulation timeframe (Fig. S3), reflecting increased molecular packing density and system compactness. These results collectively confirm the formation of a thermodynamically stable supramolecular architecture. Binding energy decomposition analysis identified electrostatic interactions and van der Waals forces as the primary drivers of self-assembly between GA and ASA (Fig. 3B). The dynamic self-assembling process was showed in Fig. 3C. Thereafter, we analyzed intermolecular interactions of molecular aggregations at 20 ns. Zn²⁺ formed a coordination bond with carboxyl group from the second glucuronic acid residue of GA; The carboxyl group of ASA established hydrogen bonds with hydroxyl groups of GA's first glucuronic acid (Fig. 3D). MD simulation clearly displayed the ternary self-assembly process and mechanism, which was accordance with the above experimental results.

Fig. 3.

MD simulation of ASA Gel. A Time-dependent changes of RMSD in self-assembly process. B The electrostatic energy and van der Waals energy in selfassembly process. C The structure changes of ASA Gel at 0, 1, 3, 5, 10 and 20 ns. D The self-assembled structure and intermolecular interaction of ASA Gel under a stable state (20 ns).

3.3. ASA gel alleviates CCl₄-induced liver damage in vivo

In order to investigate the therapeutic effects of ASA Gel on hepatic fibrosis and its mechanism, a mouse model of hepatic fibrosis induced by CCl₄ was used to evaluate the anti-liver damage and anti-hepatic fibrosis effects. First, mice were given 0.1 mg mL⁻¹ CCl₄ for 4 weeks to induce liver fibrosis [25]. The mice were then injected with GA-Zn²⁺ (15 mg kg⁻¹, equivalent to 29 mg·kg⁻¹¹ ASA Gel), free ASA (14 mg kg⁻¹, equivalent to 29 mg kg⁻¹ ASA Gel), ASA Gel (7.25 mg kg⁻¹, 14.5 mg kg⁻¹, 29 mg kg⁻¹) at different doses for 4 weeks, while CCl₄ was given continuously (Fig. 4A). As shown in Fig. 4B, compared with the control group, GA-Zn²⁺ group, ASA group and ASA Gel group, the liver index of mice in the model group was significantly increased. ASA Gel not only decreased the ratio of liver/body weight, but also notably reduced the serum levels of AST, ALT and other core indicators of liver function such as albumin and bilirubin (Fig. 4C, D, Fig. S5), the key diagnostic markers used in liver function test, indicating that ASA Gel had certain therapeutic effects. The effect of ASA Gel-M (14.5 mg kg⁻¹) and ASA Gel-H (29 mg kg⁻¹) group was stronger than that in ASA group (29 mg kg⁻¹).

Fig. 4.

ASA Gel alleviates liver damage in CCl₄-induced liver fibrosis mice. A ASA Gel regimen for CCl₄-induced fibrosis in mice. B Ratio of liver to body weight. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗∗P < 0.001. C Serum ALT levels. Each group is n = 5. Compared with the control group, Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗∗P < 0.001. D Serum AST levels. Each group is n = 5. Compared with the control group, Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. E Representative images of liver sections stained by H&E. F Pathological score of liver tissues on H&E-stained sections. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Histological changes were observed by hematoxylin and eosin (H&E) staining. As shown in Fig. 4E, the liver tissue structure in the control group was intact and clear: the hepatic lobule structure was regular, hepatocytes were arranged closely and orderly, without obvious inflammatory cell infiltration, hepatocyte degeneration/necrosis, or fibrosis signs, and the tissue morphology around the central vein was normal, reflecting the typical characteristics of healthy liver tissue. CCl₄ reatment induced hepatic fibrosis injury, which was characterized by disordered arrangement of hepatocytes, destruction of the normal hepatic lobule structure; enlarged inflammatory infiltration area, and abnormal increase of interstitium. In contrast, ASA Gel at doses of 14.5 and 29 mg kg⁻¹ significantly alleviated these histopathological alterations. Quantitative analysis showed that the inflammatory infiltration in the liver fibrosis model group was as high as 24-fold that of the normal group. The drug treatments all showed varying degrees of improvement. Moreover, the ASA Gel-H (29 mg kg⁻¹) group could significantly reduce the inflammatory infiltration caused by liver fibrosis, with a reduction rate of 3.64-fold (Fig. 4F). In conclusion, ASA Gel could mitigate CCl₄-induced liver injury in mice.

3.4. ASA gel alleviates liver fibrosis in vivo and TGF-β1-induced cell model in vitro

ASA Gel significantly reduced the deposition of collagen fibers as shown by masson staining (Fig. 5A). Masson trichrome staining quantified a 79.99 % reduction in collagen deposition following high-dose ASA Gel treatment (Fig. 5B). Western Blot analyses revealed ASA Gel could suppress the α-smooth muscle actin (α-SMA) and collagen type I (COL1) protein expression (Fig. 5C). Immunofluorescence corroborated these findings, showing a 4.7-fold reduction in α-SMA⁺/COL1⁺ cell density in peri-sinusoidal regions, compared with the model group, indicative of hepatic stellate cell (HSC) deactivation (Fig. 5D and E). In summary, ASA Gel attenuated CCl₄-induced liver fibrosis in mice.

Fig. 5.

ASA Gel alleviates CCl₄-induced liver fibrosis in mice and TGF-β1-induced cell model in LX-2 cell line. A Representative images of liver sections stained with masson staining. GA: 15 mg kg⁻¹; ASA free: 14 mg kg⁻¹; ASA Gel-L: 7.25 mg kg⁻¹; ASA Gel-M: 14.5 mg kg⁻¹; ASA Gel-H: 29 mg kg⁻¹. B Collagen deposition of masson pine. C α-SMA and COL 1 protein expression in liver tissue. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗P < 0.01. D The relative fluoresence intensity of COL1 and α-SMA in control, model, GA-Zn, ASA free and ASA Gel-H groups.Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗P < 0.01, ∗∗∗P < 0.001. E Immunofluorescence staining of liver COL 1 (green) and α-SMA (red), and nuclear DAPI reverse staining (blue). F-H The CCK-8 assay curves of GA-Zn, ASA and ASA Gel against LX-2 cells for 48 h of treatment.Compared with the Control group, ∗∗P < 0.01, ∗∗∗P < 0.001. I Protein expression of COL1 and α-SMA in LX-2 cell line. J-K The relative mRNA level of COL1 and α-SMA in LX-2 cell line. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗P < 0.01, ∗∗∗P < 0.001.

To comprehensively evaluate the regulatory effects of ASA Gel on hepatic stellate cells (HSCs), we induced human HSCs (LX-2 cell line) with transforming growth factor-β1 (TGF-β1) to simulate the hepatic fibrosis process in vitro. Firstly, the cytotoxicity of ASA, GA-Zn, and ASA Gel against LX-2 cells was assessed using the Cell Counting Kit-8 (CCK-8) assay. The results showed that the half-maximal inhibitory concentration (IC₅₀) values of GA-Zn, ASA, and ASA Gel were 357.6 μg mL⁻¹, 177.1 μg mL⁻¹, and 449.9 μg mL⁻¹, respectively (Fig. 5F, G, Fig. S6). At a concentration of 25 μg mL⁻¹, all three formulations exhibited minimal cytotoxicity toward LX-2 cells. Therefore, we selected concentrations of 6.25 μg mL⁻¹, 12.5 μg mL⁻¹ and 25 μg mL⁻¹ for further experiments. Subsequently, LX-2 cells were treated with 5 ng mL⁻¹ TGF-β1 for 48 h to induce their activation. Total cellular protein and RNA were then extracted separately for Western Blot and qPCR analyses, respectively.

Consistent with the in vivo experimental findings, ASA Gel effectively downregulated the expression of hepatic fibrosis-related indicators. The protein levels of COL1 and α-SMA were significantly upregulated in the TGF-β1-induced group (3.35-fold and 3.01-fold of the control group, respectively). ASA Gel intervention dose-dependently reduced these protein levels: in the ASA Gel-H (25 μg mL⁻¹) group, COL1 and α-SMA protein expression decreased to 0.37-fold and 0.29-fold of those in the Model group, with extremely significant differences (Fig. 5H). The qPCR results also showed the same trend differences (Fig. 5I and J).

Biological safety is a crucial factor in determining the clinical translation of drugs. We further conducted a safety assessment on multiple internal organs through H&E staining, as shown in Fig. S7. Four weeks after treatment, no significant pathological changes were observed in the heart, liver, spleen, lung, and kidney tissues collected from all treatment groups compared to the control group.

3.5. ASA gel alleviates the liver fibrosis by concurrent autophagy activation and ferroptosis inhibition

To further investigate the potential molecular mechanisms underlying the effects of ASA Gel, RNA sequencing analysis was performed on liver tissue. A total of 16868 genes were matched from Ensemble: GRCm39, a Mus musculus genome. Differential gene analysis between groups was performed using DESeq2 with a screening threshold of P < 0.05 and a fold change >1.2. As shown in Fig. 6A, there were 7322 differentially expressed genes (DEGs) changed between the control and model groups, with 3753 genes upregulated and 3569 genes downregulated. Compared to the model group, ASA Gel group showed 2404 DEGs, with 1301 genes unregulated and 1103 genes down-regulated (Fig. 6B). The Principal Component Analysis (PCA) followed by ANOSIM analysis confirmed significant differences in gene expression levels among the three groups (Fig. 6C).

Fig. 6.

RNA expression profile of livers in different groups of mice. A Volcano plots of DEGs between control and model group. B Volcano plots of DEGs between model and ASA Gel group. C Principal component analysis of the three groups. D KEGG analysis of gene enrichment in the signaling pathway. E Heatmap of the expression levels in model and ASA Gel.

To identify the signaling pathways affected by ASA Gel, the KEGG analysis was conducted. The deferentially expressed genes (DEGs) were mainly enriched in the ferroptosis signaling pathway and autophagy (Fig. 6D). Then, heatmap of genes related to autophagy and ferroptosis regulation was generated and shown in Fig. 6E.

ASA Gel exerted dual modulation of ferroptosis and autophagy to mitigate CCl₄-induced hepatic fibrosis, as evidenced by multilevel molecular profiling (Fig. 7A–G). ASA Gel transcriptionally activated autophagy initiation machinery, elevating Atg9A (2.2-fold) and Atg101 (2.7-fold) mRNA levels (Fig. 7A and B). Concurrently, ransmission electron microscopy corroborated this mechanism, and the typical structure of autophagy was observed in ASA Gel-treated hepatocytes. Importantly, untreated hepatocytes showed fewer autophagic vesicles in the cytoplasm (Fig. 7C). Whereas treatment with ASA Gel resulted in a significant increase in autophagic vesicles, representing enhanced autophagic flux. Immunofluorescence quantitative analysis showed the model group exhibited a 1.69-fold increase in P62 fluorescence intensity versus controls, demonstrating significant blockade of autophagosome-lysosome fusion. ASA Gel-H (29 mg kg⁻¹) treatment reduced P62 fluorescence intensity significantly compared to the model group, indicating that ASA Gel has a promoting effect on autophagy (Fig. 7D and E). qPCR analyses revealed a 2.4-fold upregulation of GPX4 and 1.8-fold increase in SLC7A11 in the liver tissues of ASA Gel-treated mice, confirming ferroptosis suppression (Fig. 7F and G).

Fig. 7.

ASA Gel alleviates the liver fibrosis by synergistic autophagy activation and ferroptosis inhibition. A-B The relative mRNA level of Atg9A and Atg101 in control, model, GA-Zn, free ASA and ASA Gel groups.Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗∗P < 0.001. C Morphological changes of liver tissue under autophagy formation under transmission electron microscope (TEM), A: Autophagosome; AL:Autophagolysosome. D Immunofluorescence staining of liver P62 (green) and nuclear DAPI reverse staining (blue).Scale bar, 100 μm. E The relative fluorescence intensity of liver P62 in Control, Model, GA-Zn, ASA free and ASA Gel-H groups. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗∗P < 0.001. F-G The relative mRNA level of GPX4 and SLC7A11 in Control, Model, GA-Zn, free ASA and ASA Gel-H groups.Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗∗P < 0.001.

In addition to in vivo studies, in vitro experiments have also confirmed that the ASA Gel can exert an anti-liver fibrosis effect by regulating autophagy and ferroptosis. The protein expression of ferroptosis markers GPX4, SLC7A11 and autophagy-related marker P62 were markedly changed in the TGF-β1 group. ASA Gel treatment reversed this change in a concentration-dependent manner: the ASA Gel-H (25 μg mL⁻¹) group increased to 3.13-fold and 3.66-fold of those in the TGF-β1 group (Fig. 8A), and decreased to 2.85-fold of those in the TGF-β1 group (Fig. 8D) Consistent with the Western Blot findings, qPCR results further confirmed the regulatory effects of ASA Gel at the transcriptional level (Fig. 8B, C, E).

Fig. 8.

In vitro, ASA Gel alleviated TGF-β1-induced hepatic fibrosis by activating autophagy and ferroptosis pathways. A Protein expression of GPX4 and SLC7A11 in LX-2 cell line. B Protein expression of P62 in LX-2 cell line. C-D The relative mRNA level of GPX4 and SLC7A11 in LX-2 cell line. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. E The relative mRNA level of P62. Compared with the Control group, ^###P < 0.001. Compared with the Model group, ∗∗P < 0.01, ∗∗∗P < 0.001.

Collectively, our in vivo and in vitro datas demonstrate that ASA Gel modulates the activation status of HSCs by regulating autophagic and ferroptotic signaling pathways at both the protein and transcriptional levels, thereby exerting anti-hepatic fibrosis effects.

3.6. ASA gel achieved targeted hepatic delivery in liver fibrosis model mice

In addition, to clarify the in vivo biodistribution characteristics of ASA Gel, we evaluated its in vivo distribution using DiR-ASA Gel. The results showed that there were significant differences in biodistribution between free DiR and DiR-ASA Gel (Fig. 9A). During the process of liver fibrosis, the enhanced permeability and retention (ERP) effect in the liver lesion area led to the passive accumulation of both free DiR and DiR-ASA Gel in the liver. Therefore, in the livers of experimental mice with liver fibrosis, both free DiR and DiR-ASA Gel exhibited fluorescence signals. Specifically, the free DiR group reached its peak at approximately 8 h and then rapidly declined, while the fluorescence signal of the DiR-ASA Gel group peaked at 12 h and was more concentrated in the liver region. Quantitative analysis indicated that the fluorescence expression of DiR-ASA Gel was more than twice that of free DiR at the same time point, and at 12 h, it was 4.2-fold than that of free DiR (Fig. 9C). After 48 h, independent fluorescence imaging of each organ showed that fluorescence was concentrated in the livers, spleens, and lungs (Fig. 9B). At 48 h, the fluorescence intensity of DiR-ASA Gel in the liver was twice that of free DiR (Fig. 9D). The non-specific accumulation in the spleen and the trace distribution in the lungs are normal behaviors of nanocarriers in vivo and do not affect the enrichment advantage in the liver [26,27].Compared with free DiR, DiR-ASA Gel can more effectively target and accumulate in the lesioned liver of mice with liver fibrosis models, and its retention time in the liver is significantly prolonged.

Fig. 9.

ASA Gel achieved targeted hepatic delivery in liver fibrosis model mice. A Real-time fluorescence images of mice injected with free DiR and DiR-ASA Gel. B Ex vivo imaging of major organs after injected with free DiR and DiR-ASA Gel. C Ex vivo Statistics of fluorescence intensity at different time points. Compared with the Model + free DiR group, ∗∗∗P < 0.001. D Ex vivo Statistics of fluorescence intensity in the location of the liver, spleen and lung. Compared with the Model + free DiR group, ∗∗∗P < 0.001.

4. Conclusion

The ternary carrier-free ASA Gel hydrogel was synthesized via ligand-mediated hydrogen bonding among GA, Zn²⁺, and ASA, forming a synergistic therapeutic triad with excellent biocompatibility. In CCl₄-induced hepatic fibrosis mice, ASA Gel demonstrated good therapeutic efficacy, which could significantly reduce the serum ALT, AST and other core indicators of liver function such as albumin and bilirubin, decrease the expression of α-SMA, COL 1 fibrosis marker protein, and alleviate collagen deposition. Liver transcriptomics and in vivo studies identified ASA Gel's dual modulation of autophagy-ferroptosis. ASA Gel could up-regulate the autophagosome assembly genes Atg9A and Atg101, promote the formation of autophagosomes and enhance autophagic flux. Meanwhile, ASA Gel could up-regulate the expression of GPX4 and SLC7A11, rescuing hepatocytes from ferroptotic death. This self-assembled hydrogel platform synergizes phytochemical bioactivity with metal ion coordination, offering a precision antifibrotic therapy via autophagy potentiation and redox homeostasis restoration.

CRediT authorship contribution statement

Chunsheng Zhu: Conceptualization, Funding acquisition, Methodology, Validation, Writing – original draft. Shuhe Jia: Methodology, Validation, Writing – original draft. Jie Wang: Methodology, Validation, Writing – original draft. Feng Gao: Investigation, Visualization. Yixuan Lin: Investigation, Visualization. Anzheng Nie: Investigation, Visualization. Zhuoqian Guo: Investigation, Visualization. Zheng Zhou: Supervision, Writing – original draft. Bing Xu: Funding acquisition, Project administration, Supervision, Writing – original draft. Haimin Lei: Funding acquisition, Resources, Supervision.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Bing Xu reports financial support was provided by National Key Research and Development Program of China (2022YFC3502100). Bing Xu reports financial support was provided by National Natural Science Foundation of China (No. 82104365). Haimin Lei reports financial support was provided by National Natural Science Foundation of China (No. 82274082). Haimin lei reports financial support was provided by Beijing Key Laboratory for Basic and Development Research on Chinese Medicine (Beijing, 100102). Chunsheng zhu reports financial support was provided by National Natural Science Foundation of China (No. 82204701). If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.