An inhalable nanozyme for STING blockade to treat radiation-induced lung injury

Abstract

Background

Radiation-induced lung injury (RILI), including radiation pneumonitis (RP) and radiation-induced pulmonary fibrosis (RIPF), represents a major dose-limiting complication in thoracic radiotherapy. Acute RP constrains radiotherapy implementation, while RIPF causes irreversible pulmonary dysfunction. During ionizing radiation, the STING pathway is upregulated, and oxidative stress is induced in macrophages, promoting acute RP and resulting in late RIPF.

Results

In this study, an inhalable ferrous nanozyme encapsulating rosmarinic acid, termed FeRAzyme, was developed to address RILI induced by radiotherapy. Rosmarinic acid was rationally selected through molecular docking and dynamics simulations, revealing its non-covalent binding to STING at residues Ser243-Tyr245-Leu259-Glu260-Asn211, which blocks STING phosphorylation. After screening of a metal-based nanozyme library, FeRAzyme showed remarkable catalase- and superoxide dismutase-mimetic activities, efficiently scavenging reactive oxygen species and thus attenuating radiation-induced oxidative damage. Administered via inhalation, FeRAzyme suppressed STING phosphorylation and alleviated oxidative stress in macrophages, thus inhibiting proinflammatory cytokine secretion.

Conclusions

Both in vitro and in vivo studies demonstrated that FeRAzyme effectively restored macrophage homeostasis and ameliorated both acute RP and myofibroblast activation without systemic toxicity. This work establishes a dual-functional STING-inhibitory nanozyme strategy for mitigating RILI, offering a promising approach for radioprotection in thoracic cancer radiotherapy.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04636-3.

Introduction

Patients receiving radiotherapy for thoracic malignancies, including lung cancer, breast cancer, and esophageal cancer, are at risk of developing radiation-induced lung injury (RILI) [1–4]. Despite advances in precision radiotherapy modalities such as intensity-modulated radiotherapy (IMRT), approximately 5% – 25% of patients continue to develop RILI [5, 6]. The principal manifestations of RILI are acute radiation pneumonitis (RP) and late radiation-induced pulmonary fibrosis (RIPF). RP serves as a precursor to pulmonary fibrosis, with the severity of the inflammatory response directly correlating with fibrotic progression [7]. During radiotherapy, ionizing radiation induces the activation of lung tissue-resident fibroblasts, endothelial cells, and vascular smooth muscle cells, promoting their transdifferentiation into myofibroblasts and initiating fibrogenesis [8]. RIPF typically manifests as irreversible pathological remodeling, characterized by dyspnea, destruction of pulmonary parenchymal architecture, interstitial edema, and potential progression to respiratory failure [9, 10]. These pathological alterations not only constrain the safe escalation of therapeutic radiation doses but also compromise overall treatment efficacy [11]. Currently, no specific pharmacological interventions are available to effectively reverse RIPF [12, 13].

Radiation-induced DNA damage represents a central event in mediating radiotoxicity. Irradiated cells generate cytoplasmic double-stranded DNA, which is recognized by cyclic GMP-AMP synthase (cGAS), the principal sensor for cytoplasmic DNA, leading to the activation of the stimulator of interferon genes (STING) signaling pathway and promoting the secretion of proinflammatory mediators, including tumor necrosis factor-α (TNF-α), interleukin (IL)-6, and IL-1β [14–17]. Emerging evidence demonstrates that the cGAS-STING pathway plays an essential role in the pathogenesis and progression of RILI, with its activation significantly influencing macrophage functional phenotypes [18, 19]. In radiation-induced fibrosis, macrophage infiltration from adjacent tissues or the circulation is critical. Persistent activation of the cGAS-STING pathway induces macrophage dysfunction, resulting in excessive release of inflammatory mediators, which then activate myofibroblasts, initiating dysregulated repair responses that culminate in scar tissue deposition and fibrotic progression [20–22]. Therefore, therapeutic targeting of STING signaling represents a promising strategy for modulating radiation-induced innate immune responses and potentially ameliorating RILI through the restoration of macrophage homeostasis.

Oxidative stress results from an imbalance between highly reactive molecular species, including reactive oxygen species (ROS) and reactive nitrogen species (RNS), and endogenous antioxidant defense systems [23]. Under physiological conditions, basal oxidative stress is maintained within a tightly regulated state of "redox homeostasis". However, radiotherapy disrupts this homeostatic equilibrium in both malignant and normal tissues, inducing redox dysregulation and promoting fibrotic progression [24]. Within this framework, nanozymes with enzyme-like catalytic functions have attracted considerable attention in recent years [25–27]. They combine the advantages of easy preparation and high stability of nanomaterials, enabling them to catalyze enzyme-like reactions under physiological conditions [28–30]. Nanozymes with superoxide dismutase (SOD) and catalase (CAT) activities are of particular interest, as they can effectively scavenge excess ROS such as superoxide anions (·O₂⁻) and hydrogen peroxide (H₂O₂), precisely regulating the redox state [31–35]. Owing to these robust antioxidant properties, ROS-scavenging nanozymes have been widely and successfully deployed against a broad spectrum of ROS-mediated inflammatory diseases. Recent literature highlights their therapeutic efficacy in managing conditions such as inflammatory bowel disease, osteoarthritis, acute kidney injury, and bacterial infections, where they significantly attenuate local and systemic inflammation by restoring redox homeostasis [36–38]. Therefore, based on the excellent catalytic activity of nanozymes, their application in the treatment of RILI shows great promise [39, 40]. By efficiently clearing ROS, these systems are expected to reduce the severity of acute RP and inhibit the development of RIPF, providing a new strategy for the prevention and treatment of this type of injury.

Natural herbal products represent valuable sources of bioactive compounds, showing potential antioxidant, immunomodulatory, and anti-inflammatory properties [41–43]. Considering the pivotal role of STING signaling in RILI pathogenesis, the binding affinity of several natural herbal compounds to STING was evaluated through molecular docking and molecular dynamics (MD) simulations. Rosmarinic acid (RA), a water-soluble natural phenolic compound derived from rosemary, displayed a binding energy of -7.5 kcal/mol to STING, indicating robust and stable non-covalent interaction between the two molecules. Inspired by these findings, a nano-assembly termed FeRAzyme (ferrous nanozyme encapsulating RA) was developed for the treatment of RILI via noninvasive intratracheal instillation therapy through STING targeting and oxidative stress mitigation in macrophages. Macrophage infiltration plays a critical role in RILI pathogenesis through the production of inflammatory cytokines that promote both acute tissue injury and further pulmonary fibrotic responses. Following pulmonary administration, FeRAzyme is efficiently internalized, exploiting the natural phagocytic propensity of lung-resident macrophages. As illustrated in Scheme 1, subsequent to this localized accumulation, FeRAzyme shows significant non-covalent binding to macrophage STING at residues Ser243-Tyr245-Leu259-Glu260-Asn211, inhibiting STING phosphorylation and attenuating pathway activation. Concurrently, FeRAzyme represents CAT- and SOD-mimetic activities that mitigate oxidative stress-induced injury through the elimination of H₂O₂ and ·O₂⁻, suppressing the secretion of proinflammatory cytokines including TNF-α, IL-6, and IL-1β. This inhalable FeRAzyme formulation ameliorates RILI by reducing acute inflammatory responses and restoring macrophage homeostasis, decreasing excessive inflammatory mediator production, and deactivating myofibroblasts. This approach demonstrates significant potential as a noninvasive therapeutic strategy for both RP and RIPF.

Scheme 1.

Schematic illustration of (A) synthesis of metal-based nanozymes and (B) proposed mechanism by which FeRAzyme ameliorates RILI: 1) noninvasive intratracheal instillation increases local accumulation of FeRAzyme within pulmonary tissue; 2) FeRAzyme shows CAT/SOD-mimetic activity with ROS scavenging capability to attenuate oxidative stress; 3) FeRAzyme suppresses cGAS-STING pathway activation; 4) through dual inhibition of oxidative stress injury and STING pathway signaling, FeRAzyme mitigates both RP and RIPF

Results and discussion

Macrophage cGAS-STING pathway activation in the murine RILI model

To investigate whether the cGAS-STING pathway is involved in RILI, a murine RILI model was established through 20-Gy thoracic irradiation (IR). Immunofluorescence co-localization analysis revealed a significant increase in phosphorylated STING (p-STING) expression specifically within macrophages at 30 days and 3 months post-irradiation (Fig. 1A). In vitro experiments further confirmed that 10-Gy irradiation significantly upregulated the protein expression levels of cGAS, p-STING, p-TBK1, and p-NF-κB in murine bone marrow-derived macrophages (BMDMs) (Fig. 1B–C). This upregulation was accompanied by a marked increase in the secretion of inflammatory cytokines (TNF-α, IL-6, and IL-1β) in culture supernatants (Fig. 1D–F). These results indicate that activation of the macrophage cGAS-STING pathway plays a critical role in RILI progression.

Fig. 1.

Radiation-induced activation of the cGAS-STING pathway in macrophages. (A) Immunofluorescence images of F4/80 (green) and p-STING (red) in murine lung tissues. (B, C) Western blot analysis and quantification of cGAS, p-STING, p-TBK1, and p-NFκB protein expression in BMDMs (n = 3; Student’s t-test). (D-F) Concentrations of TNF-α, IL-6, and IL-1β in culture supernatants (n = 3; Student’s t-test). (G) Schematic illustration of molecular docking. (H) Heatmap of binding energies. (I) Molecular docking simulations of herb-derived drugs with the target protein. (J) Per-atom binding energy of RA with STING. (K) Cell viability of BMDMs determined by CCK-8 assay, normalized to the IR group. Data are presented as mean ± SD from three independent experiments with statistical significance indicated as *p < 0.05, **p < 0.01, ***p < 0.001

Considering the critical role of the cGAS-STING pathway in RILI, the interactions between selected compounds (1600 FDA-approved drugs and herb-derived drugs) and the STING protein were evaluated using molecular docking simulations, as shown in Fig. 1G, with binding affinities presented in Fig. 1H. The binding energies of six herb-derived drugs to the target protein are shown in Fig. 1I, wherein RA displayed a binding energy of -7.5 kcal/mol (Fig. 1J). Macrophage viability following co-incubation with these herb-derived drugs at various concentrations was assessed (Fig. 1K). Among the tested compounds, RA demonstrated the most significant concentration-dependent protection against cell death. Therefore, RA was selected as the lead compound for cGAS-STING pathway inhibition.

Synthesis and characterization of FeRAzymes

Using RA, FeCl₃, and DSPE-TK-PEG as precursors, FeRAzyme was prepared through a self-assembly approach (Scheme 1). Substitution of ferric ions with copper or manganese ions enabled the synthesis of CuRAzyme and MnRAzyme, establishing a metal-based nanozyme library (Fig. 2A). Transmission electron microscopy (TEM) revealed uniform morphology of FeRAzymes, CuRAzymes, and MnRAzymes (Fig. 2B). As shown in Fig. 2C, elemental mapping of FeRAzyme confirmed the presence of Fe, C, N, and O within the composite structure. The zeta potentials of FeRAzymes, CuRAzymes, and MnRAzymes were -20.5 mV, -15.6 mV, and -16.7 mV, respectively (Fig. 2D). Dynamic light scattering (DLS) analysis indicated particle sizes of 88.1 nm, 82.0 nm, and 91.5 nm for FeRAzymes, CuRAzymes, and MnRAzymes, respectively (Fig. 2E). The ultraviolet–visible absorption profile of FeRAzyme is presented in Fig. 2F. To further characterize the elemental states of FeRAzyme, X-ray photoelectron spectroscopy (XPS) was performed. High-resolution XPS analysis of Fe 2p (Fig. 2G) revealed two distinct peaks at 723.5 eV and 710.22 eV, corresponding to the binding energies of Fe²⁺ 2p_1/2 and 2p_3/2 orbitals, while peaks at 725.5 eV and 712.2 eV, along with four satellite peaks, were assigned to the 2p_1/2 and 2p_3/2 orbitals of Fe³⁺. Moreover, Cu 2p in CuRAzymes and Mn 2p in MnRAzyme were shown in Figure S1 and S2 respectively. The high-resolution XPS spectrum of Cu 2p revealed the presence of both Cu²⁺ and Cu^0/+1, while that of Mn 2p indicated a mixed valence state of Mn⁴⁺ and Mn³⁺. The zeta potential of FeRAzyme, CuRAzyme, and MnRAzyme remained relatively stable over a 7‑day observation period (Figure S3). These results confirm the successful preparation of a series of metal ion-based nanozymes.

Fig. 2.

Physicochemical characterization of FeRAzymes. (A) Schematic illustration of the establishment of the nanozyme library. (B) TEM images illustrating the morphology of FeRAzymes, CuRAzymes, and MnRAzymes. (C) Elemental mapping of FeRAzymes. (D) Zeta potentials of FeRAzymes, CuRAzymes, and MnRA. (E) Size distribution of FeRAzymes, CuRAzymes, and MnRAzymes by DLS. (F) Absorption spectrum of FeRAzyme. (G) High-resolution XPS spectra of Fe 2p

Multienzyme-like activities of FeRAzymes

The mitigation of oxidative stress plays a crucial role in RILI treatment. A nanozyme library of metal ion-based nanozymes, including FeRAzyme, CuRAzyme, and MnRAzyme, was evaluated for their SOD/CAT cascade activities mimicking natural enzymes. Compared with MnRAzyme, both FeRAzyme and CuRAzyme displayed significantly elevated CAT-like activity as evidenced by H₂O₂ decomposition and O₂ generation assays (Fig. 3A and S4). Meanwhile, increased ·O₂⁻ scavenging capacity was observed in the FeRAzyme treatment group via the xanthine-xanthine oxidase system (Fig. 3B). Further evaluation of 2,2’-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) and 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging ratios demonstrated that FeRAzyme displayed remarkable suppression of nitrogen radicals (Fig. 3C and D). Electron spin resonance (ESR) spectra further confirmed the ROS scavenging capability of FeRAzyme in eliminating ·O₂⁻ and ·OH (Fig. 3E and F). These analyses demonstrate the excellent catalytic stability of FeRAzyme for ROS scavenging (Fig. 3G).

Fig. 3.

Illustration of ROS-scavenging capabilities of FeRAzymes. (A) CAT-like activities of FeRAzymes, CuRAzymes, and MnRAzymes were assessed by monitoring the dissolved oxygen generated from H₂O₂ solutions at pH 7.4, with metal equivalents of 0.3 μg/mL. (B) SOD-mimetic activities measured via xanthine oxidase/cytochrome c system at pH 7.4. (C and D) Radical scavenging efficiency was further evaluated using ABTS (C) and DPPH (D) assays for FeRAzymes, CuRAzymes, and MnRAzymes. (E) The reduction of superoxide radical generation by FeRAzymes was confirmed via ESR spectroscopy. (F) The suppression of hydroxyl radical formation was validated by ESR spectroscopy. (G) Schematic diagram of the enzyme-mimetic activity of FeRAzyme

Next, density functional theory (DFT) calculations were employed to investigate the catalytic processes of CAT and SOD. The proposed mechanisms for the SOD-like and CAT-like activities of FeRAzyme are shown in Fig. 4A and B, respectively. Under physiological conditions, ·O₂⁻ is rapidly protonated to form a hydroperoxyl radical (·OOH). Initially, ·OOH adsorbs onto the surface of FeRAzyme, capturing a hydrogen bond from -OH, leading to the desorption of H₂O₂. Another ·OOH then binds via hydrogen bonding, facilitating hydrogen transfer and resulting in the desorption of an O₂ molecule, completing the catalytic cycle. In the CAT reaction, H₂O₂ undergoes adsorption, followed by hydrogen transfer that leads to H₂O desorption. After the adsorption of a second H₂O₂ molecule, another H₂O is released, and eventually O₂ desorbs. The Gibbs free energy profiles (Figs. 4C and D) indicate that these cycles are highly exergonic, demonstrating a strong thermodynamic driving force.

Fig. 4.

DFT calculations for the catalytic process. (A-B) SOD (A) and CAT (B) mimicking properties and mechanisms. (C-D) Gibbs free energy profile of SOD (C) and CAT (D) like process

FeRAzymes attenuate macrophage cGAS-STING pathway activation in vitro

Given the strong antioxidant properties of FeRAzyme, its ROS-scavenging activity was evaluated in vitro. As shown in Figure S5, radiation induced a significant increase in ROS levels in BMDMs, whereas FeRAzyme treatment effectively reduced intracellular ROS. The effect of FeRAzyme on the cGAS-STING pathway was further investigated. Immunofluorescence and Western blot analyses demonstrated that FeRAzyme significantly suppressed radiation-induced activation of the cGAS-STING pathway in macrophages (Fig. 5A–C). Since activation of this pathway stimulates the production of proinflammatory cytokines, which exacerbate inflammatory responses, cytokine levels in BMDM culture supernatants were analyzed. FeRAzyme treatment significantly decreased TNF-α, IL-6, and IL-1β levels, indicating its potent anti-inflammatory activity (Fig. 5D–F).

Fig. 5.

In vitro Inhibition of Macrophage STING by FeRAzyme Attenuates Fibroblast Activation. (A and B) Immunofluorescence analysis of p-STING in BMDMs. (C) Western blot analysis of cGAS, p-STING, p-TBK1, and p-NFκB in BMDMs. (D–F) Levels of TNF-α, IL-6, and IL-1β in culture supernatants (n = 3; one-way ANOVA). (G) The scheme for the generation of CM. (H–L) Immunofluorescence (H–K) and Western blot (L) analyses of α-SMA and COL1A1 were conducted in MLFs. (M–O) qPCR analysis of COL1A1,α-SMA, and TGF-β1 (n = 3; one-way ANOVA). (P and Q) Transwell migration assay with corresponding statistical analyses evaluated fibroblast migratory capacity (n = 3; one-way ANOVA). G1: Control; G2: IR; G3: Fezyme; G4: RA; G5: FeRAzyme. Data are presented as mean ± SD from three independent experiments, with significance indicated as *p < 0.05, **p < 0.01, and ***p < 0.001

During the progression of RILI, inflammatory cytokines induce the activation of fibroblasts into myofibroblasts via signaling pathways, which in turn secrete large amounts of extracellular matrix (ECM) and drive its remodeling. The aberrantly deposited ECM sustains the activated state of fibroblasts through mechanical signaling and biochemical feedback, and these two processes synergistically drive the initiation and irreversible progression of pulmonary fibrosis. To examine the effect of macrophage STING pathway activation on fibroblast activation, conditioned medium (CM) was collected from macrophages subjected to IR with different treatments and applied to cultured mouse lung fibroblasts (MLFs) for 48 h (Fig. 5G). The expression of key fibroblast activation markers, which are associated with ECM, α-smooth muscle actin (α-SMA), and collagen I (COL1A1), was then evaluated. Both immunofluorescence and Western blot analyses demonstrated that CM from irradiated macrophages significantly increased the protein levels of α-SMA and COL1A1 in MLFs. In comparison, these markers were reduced to varying degrees across the treatment groups, with FeRAzyme producing the most significant decrease (Fig. 5H–L). Consistent results were obtained by qPCR, which showed that FeRAzyme significantly downregulated transforming growth factor-β1 (TGF-β1), a key profibrotic mediator (Fig. 5M–O). Furthermore, transwell (Fig. 5P and Q) and wound-healing (Figure S6) assays were used to assess fibroblast migratory capacity. CM from irradiated macrophages increased MLF migration, whereas FeRAzyme treatment effectively suppressed this effect. These findings indicate that inflammatory cytokines induced by radiation-activated macrophage STING signaling promote fibroblast activation, while FeRAzyme mitigates pulmonary inflammation and fibrosis by inhibiting this pathway.

Therapeutic effects of FeRAzyme on acute RP

To comprehensively evaluate the therapeutic efficacy of FeRAzyme across the entire pathological continuum of RILI, our in vivo study was designed to span three critical stages: 30 days post-irradiation to assess acute RP, 3 months to evaluate the transitional phase of early RIPF, and 6 months to investigate the established chronic RIPF [44].

The acute phase of RILI is characterized by a robust inflammatory response. An acute RP model was established to assess the treatment effect of FeRAzyme. First, to evaluate the pulmonary distribution and retention of FeRAzyme, the nanozyme was administered intratracheally in mice, and its fluorescence intensity and localization were monitored using an in vivo imaging system (IVIS). Ex vivo imaging of the lungs and quantitative analysis demonstrated that following inhalation administration, the FeRAzyme achieved prolonged retention in the lungs, maintaining high levels for up to 24 h post-dose (Fig. 6A and B). This finding was further corroborated by fluorescence imaging of frozen lung tissue sections (Fig. 6C). Following this, the therapeutic effect of FeRAzyme on acute RP was assessed in the mouse model. After thoracic irradiation, mice were randomly assigned to IR, Fezyme, RA, and FeRA groups, with untreated mice utilized as the healthy control group. Treatments were initiated on the day of irradiation via intratracheal instillation and administered every other day until day 14 post-irradiation. On day 30, lung tissues were collected to evaluate the effect of FeRAzyme on acute RP (Fig. 6D). Micro-CT imaging revealed significant inflammatory features in the IR group, including diffuse or patchy ground-glass opacities and areas of consolidation, whereas these pathological changes were significantly attenuated in the FeRAzyme-treated group (Fig. 6E and F). H&E staining revealed that radiation induced alveolar wall thickening and extensive inflammatory cell infiltration. In comparison, FeRAzyme treatment significantly reduced inflammatory infiltration and preserved alveolar architecture near normal levels (Fig. 6G). Surfactant protein C (SFTPC), a key marker of type II alveolar epithelial cells critical for lung function, was significantly decreased in the radiation group. Immunofluorescence analysis showed that FeRAzyme treatment restored SFTPC expression, indicating that FeRAzyme supports the functional recovery of lung epithelial cells (Figure S7). The effect of FeRAzyme on the STING pathway in lung tissues was further examined. Immunofluorescence and Western blot analyses confirmed that radiation activated the cGAS-STING pathway, whereas FeRAzyme treatment significantly suppressed this activation (Fig. 6H–J). Consistently, ELISA measurements showed that serum levels of the proinflammatory cytokines TNF-α, IL-6, and IL-1β were significantly decreased following FeRAzyme administration (Fig. 6K–M). Given the pivotal role of macrophages in RP progression, lung macrophage populations were quantified by flow cytometry. Radiation induced a significant increase in total macrophage numbers compared with controls, while FeRAzyme treatment significantly reduced macrophage infiltration (Fig. 6O). Pulmonary interstitial macrophages (IMs) are reported to expand during acute lung injury, contributing to inflammatory responses. IMs were significantly reduced following FeRAzyme treatment (Fig. 6P). Moreover, M1 macrophages, which mediate early inflammatory responses in RILI, were significantly increased in lung tissues 30 days post-radiation, whereas FeRAzyme administration effectively reduced their levels (Fig. 6Q). These results demonstrate that intratracheal delivery of FeRAzyme effectively suppresses early inflammation in RILI.

Fig. 6.

Therapeutic Effects of FeRAzyme on Acute RP. (A) Schematic diagram of intratracheal instillation in mice. (B) The pulmonary distribution of FeRAzyme evaluated by IVIS imaging and quantitative analysis. (C) Fluorescence distribution of FeRA in the lung sections. (D) The experimental schedule for establishing the acute RP mouse model and treatment regimen. (E and F) Micro-CT scan analysis of mouse lungs. (G) H&E staining of lung sections. (H and I) Immunofluorescence analysis of SFTPC in lung tissues. (J) Western blot analysis of cGAS, p-STING, p-TBK1, and p-NFκB in lung tissues (n = 5; one-way ANOVA). (K–M) Serum levels of TNF-α, IL-6, and IL-1β. (n = 5; one-way ANOVA). (O–Q) Flow cytometric analysis of total macrophages (O), IMs (P), and M1 macrophages (Q) in lung tissues. G1: Control; G2: IR; G3: Fezyme; G4: RA; G5: FeRAzyme. Data are presented as mean ± SD from 5 mice of each group, with statistical significance indicated as *p < 0.05, **p < 0.01, and ***p < 0.001

Molecular mechanisms underlying FeRAzyme therapeutic activity

To further elucidate the therapeutic mechanism of the FeRAzyme for RP, RNA sequencing of mouse lung tissues was performed. Volcano plot analysis identified 970 DEGs between the IR and control groups, including 598 upregulated and 372 downregulated genes. Comparatively, comparison between the FeRAzyme and control groups revealed 376 DEGs, with 150 upregulated and 226 downregulated. Between the FeRAzyme and IR groups, a total of 692 DEGs were detected, comprising 174 upregulated and 518 downregulated genes (Fig. 7A–C). Venn diagram analysis illustrated the overlap of DEGs across comparisons. A total of 301 genes upregulated by radiation were downregulated following FeRAzyme treatment, while 50 genes downregulated by radiation were upregulated, indicating that FeRAzyme mitigates radiation-induced changes in gene expression (Fig. 7D–F). Gene Ontology (GO) enrichment analysis of the upregulated DEGs in the IR vs. control groups revealed significant enrichment of terms associated with immune response, chemotaxis, and inflammatory processes (Figure S8A). These inflammation-related processes were also enriched among the genes downregulated by FeRAzyme treatment, indicating that FeRAzyme suppresses radiation-induced inflammation in lung tissues (Fig. 7G). The KEGG pathway analysis also showed activation of multiple inflammation-related signaling pathways in the IR group, including the chemokine, TNF, JAK–STAT, NF-κB, and cytosolic DNA-sensing pathways, whereas FeRAzyme treatment inhibited the activation of these pathways (Figure S5B and H). To further validate these findings, gene set enrichment analysis (GSEA) was conducted, revealing significant enrichment of the chemokine, TNF, NF-κB, and cytosolic DNA-sensing pathways in the IR group. These pathways were upregulated in RILI (Figure S8C–F) but were markedly downregulated following FeRAzyme treatment (Fig. 7I–L). Radiation-induced DNA damage is detected by cytosolic sensors, such as cGAS, which activate the STING signaling pathway. This activation triggers downstream NF-κB mediated inflammatory cascades, resulting in the elevated release of proinflammatory cytokines, including TNF and chemokines, thus aggravating RILI. These findings indicate that FeRAzyme suppresses cGAS–STING pathway activation, reduces inflammatory cytokine production, and mitigates the inflammatory response.

Fig. 7.

Identification of Molecular Mechanisms of FeRAzyme by RNA-Seq Analysis. (A–C) Volcano plots showing the DEGs between the IR and control groups (A), FeRAzyme and control groups (B), and FeRAzyme and IR groups (C). (D–F) Venn diagrams illustrating the overlap of DEGs between the IR vs. control comparison and the FeRAzyme vs. IR comparison. (G and H) GO enrichment analysis (G) and KEGG pathway analysis (H) of downregulated genes in the FeRAzyme vs. IR comparison are presented. (I–L) GSEA of the chemokine signaling pathway (I), TNF signaling pathway (J), NF-κB signaling pathway (K), and cytosolic DNA-sensing pathway (L) between the FeRAzyme and IR groups

Therapeutic efficacy of FeRAzyme on RIPF

As the inflammatory response persists, the pathological landscape evolves into a transitional phase where active inflammation and early fibrotic remodeling coexist. During this period, sustained inflammatory signaling drives fibroblast activation and initiates the deposition of extracellular matrix components, reflecting the concurrent progression of both pneumonitis and fibrosis. Based on this progression pattern, the therapeutic activity of FeRAzyme was assessed at the 3-month time point (Fig. 8A). Micro-CT imaging displayed that irradiation significantly increased lung tissue density, whereas FeRAzyme treatment reduced this elevation (Fig. 8B and C). At 3 months post-irradiation, significant alveolar structural disruption and thickening of alveolar walls were observed, whereas these pathological alterations were mitigated following FeRAzyme administration (Fig. 8D). Masson’s trichrome staining showed that FeRAzyme treatment significantly reduced collagen deposition in lung tissues compared with the IR group (Fig. 8E). Ashcroft scoring further indicated significantly lower scores in FeRAzyme-treated mice, supporting the conclusion that FeRAzyme ameliorates pulmonary fibrosis (Fig. 8F).

Fig. 8.

Therapeutic Effects of FeRAzyme on RILI at 3 Months. (A) Experimental timeline for treatment. (B and C) Micro-CT analysis of lung morphology. (D) H&E-stained lung sections. (E) Masson’s trichrome–stained lung sections. (F) Ashcroft scoring of lung sections (n = 5; one-way ANOVA). (G–L) IHC analysis of α-SMA (G and H), COL1A1 (I and J), and FN1 (K and L) in lung sections (n = 5; one-way ANOVA). (M–P) Immunofluorescence analysis of SFTPC and p-STING in lung sections. (Q) Serum levels of TNF-α, IL-6, and IL-1β (n = 5; one-way ANOVA). G1: Control; G2: IR; G3: Fezyme; G4: RA; G5: FeRAzyme. Data are presented as mean ± SD from 5 mice of each group, with statistical significance indicated as *p < 0.05, **p < 0.01, and ***p < 0.001

The expression of profibrotic markers, including α-SMA, COL1A1, and Fibronectin 1 (FN1), was assessed by immunohistochemical staining. These analyses showed significant downregulation of all three proteins following FeRAzyme treatment (Fig. 8G–L), indicating suppression of fibroblast activation and excessive extracellular matrix accumulation. Immunofluorescence staining revealed increased SFTPC expression and reduced p-STING levels in the FeRAzyme group (Fig. 8M–P). Serum concentrations of TNF-α, IL-1β, and IL-6 were elevated in irradiated mice but were significantly decreased following FeRAzyme administration (Fig. 8Q).

In the established fibrotic stage, fibrotic remodeling becomes the dominant pathological feature. Extensive deposition of extracellular matrix, significant thickening of the alveolar septa, and substantial disruption of lung architecture collectively define this phase. To assess the antifibrotic activity of FeRAzyme during late-stage RIPF, lung tissues were collected 6 months after irradiation (Fig. 9A). Micro-CT imaging showed that severe pulmonary consolidation was significantly reduced in the FeRAzyme-treated group (Fig. 9B and C). H&E and Masson’s trichrome staining further indicated significant attenuation of alveolar structural damage, architectural disruption, and extensive collagen accumulation in mice receiving FeRAzyme (Fig. 9D and E). Significant reductions in hydroxyproline levels (Fig. 9F) and Ashcroft scores (Fig. 9G) were detected in the FeRAzyme-treated group. Immunohistochemical analysis (Fig. 9H–M), Western blotting (Fig. 9P), and qPCR (Fig. 8Q) further confirmed suppression of α-SMA, COL1A1, and FN1. These proteins are core markers of fibroblast activation and extracellular matrix deposition. The results collectively demonstrate the inhibitory effect of FeRAzyme on fibrosis progression. An increase in SFTPC expression was detected in the FeRAzyme group by immunofluorescence analysis (Fig. 9N and O). The serum concentration of TGF-β1 also remained significantly lower following FeRAzyme treatment (Fig. 9R). These findings indicate potent therapeutic activity of FeRAzyme in the RILI mouse model, preventing the progression from inflammation to fibrosis.

Fig. 9.

Therapeutic Effects of FeRAzyme on Late-Stage RIPF. (A) Experimental timeline for establishing the late-stage RIPF mouse model and administering treatment. (B and C) Micro-CT analysis of lung morphology. (D) H&E-stained lung sections. (E) Masson’s trichrome-stained lung sections. (F) Quantification of hydroxyproline levels in lung tissues (n = 5; one-way ANOVA). (G) Ashcroft scoring of lung sections (n = 5; one-way ANOVA). (H–M) IHC assessment of α-SMA (H and I), COL1A1 (J and K), and FN1 (L and M) in lung sections (n = 5; one-way ANOVA). (N and O) Immunofluorescence staining of SFTPC in lung sections. (P) Western blot analysis of α-SMA and COL1A1 in lung tissues. (Q) qPCR quantification of α-SMA, COL1A1, and FN1 in lung tissues (n = 5; one-way ANOVA). (R) Serum TGF-β1 levels (n = 5; one-way ANOVA). G1: Control; G2: IR; G3: Fezyme; G4: RA; G5: FeRAzyme. Data are presented as mean ± SD from 5 mice of each group, with statistical significance indicated as *p < 0.05, **p < 0.01, and ***p < 0.001

The potential toxicity of the FeRAzyme was comprehensively evaluated. H&E staining performed at day 30 post-treatment revealed no apparent tissue damage or histological abnormalities in the major organs of healthy mice (Figure S9). Routine hematological and biochemical parameters displayed no significant differences between the two groups (Figure S10). Thus, FeRAzyme shows potent therapeutic efficacy against RILI, coupled with excellent biosafety.

While this study demonstrates the robust in vivo therapeutic efficacy and biosafety of FeRAzyme, certain limitations should be acknowledged. First, the non-covalent interactions between RA and STING were primarily elucidated through computational molecular docking and dynamics simulations. Future studies incorporating direct structural biology evidence are required to precisely validate the binding conformation. Furthermore, in our current in vivo experimental design, FeRAzyme administration was initiated concurrently with irradiation, serving primarily as a prophylactic and early-intervention model. Therefore, investigating delayed therapeutic interventions will be necessary to comprehensively evaluate the clinical efficacy of FeRAzyme in future studies.

Conclusion

In this study, a multifunctional inhalable FeRAzyme was developed using the natural product RA for the treatment of RILI. The nanozyme provides dual therapeutic activity by modulating the cGAS-STING pathway and oxidative stress. Molecular docking and dynamic simulation analyses showed that RA binds to STING at the Ser243-Tyr245-Leu259-Glu260-Asn211 residues, inhibiting its phosphorylation. The FeRAzyme also possesses intrinsic SOD- and CAT-like activities, supporting the restoration of redox balance through the removal of excess H₂O₂ and ·O₂⁻. Following non-invasive intratracheal instillation, FeRAzymes accumulated efficiently in lung tissue and restored macrophage function by suppressing STING activation and reducing oxidative stress. This regulatory effect lowered the production of proinflammatory cytokines, alleviated acute RP, and prevented fibroblast activation associated with late-stage pulmonary fibrosis. The treatment also showed a favorable biosafety profile without evidence of systemic toxicity. This work provides mechanistic and translational insights into the development of STING-inhibitory nanozymes for effective RILI therapy.

Author contributions

Y.H., M.Luo, and Z.B. performed the experiments, collected and analyzed the data, and drafted the initial manuscript. B.L., X.Y., K.F., and M.Lyu conceived and designed the study, supervised the research, and acquired funding. All authors reviewed, revised, and approved the final manuscript.

Funding

The authors gratefully acknowledge the financial support provided by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (2024ZD0519900, 2024ZD0519902), National Natural Science Foundation of China (No. 12005158, 82403797, 82303678, 82404196), Hubei Provincial Natural Science Foundation of China (No. 2025AFB163), Tongji Hospital Fund Cultivation Project (No. 2024B08).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All animal procedures were approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology (IACUC number: 4283).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.