Biomimetic nanocages with cascade homing for theranostics of ischemia-reperfusion acute kidney injury

Abstract

Acute kidney injury (AKI) affects an estimated 13 million patients each year with a mortality of up to 1.7 million, yet current clinical management remains largely supportive and lacks etiology-directed therapy. Single-cell transcriptomic analysis revealed that mitochondrial dysfunction in proximal tubular cells is a hallmark of AKI pathology. Inspired by the regenerative capacity of axolotls, which relies on robust mitochondrial bioenergetics, we developed a biomimetic nanoplatform for treating AKI. We engineered GIPL nanoparticles comprising gold–silver nanocages loaded with the mitochondriotropic antioxidant idebenone (IDB), PEGylated for stability, and functionalized with LTH peptide for kidney injury molecule-1 (KIM-1) recognition. This design enables cascade targeting: cell-level targeting via KIM-1-mediated uptake into injured tubular epithelial cells, followed by subcellular accumulation of IDB in mitochondria. In vitro studies demonstrated that GIPL effectively scavenges reactive oxygen species, restores mitochondrial membrane potential, and inhibits apoptosis. Mechanistically, GIPL suppresses mTOR phosphorylation, thereby enhancing mitophagy to clear damaged mitochondria while attenuating pro-apoptotic signaling, ultimately restoring mitochondrial homeostasis. In vivo evaluation in Ischemia-reperfusion (IR) AKI mice confirmed preferential renal accumulation of GIPL, with significant amelioration of tubular injury, oxidative stress, and inflammation. The intrinsic X-ray attenuation of gold–silver nanocages additionally allows renal accumulation of GIPL to be visualized by computed tomography (CT). With a favorable biosafety profile, GIPL is a promising cascade-targeted mitochondrial therapy for IR-AKI. Drawing on the mitochondria-centric regenerative principle of the axolotl, this work integrates single-cell transcriptomic target identification, cascade cell-to-mitochondria delivery, and intrinsic CT contrast into a single nanoplatform, providing a new therapeutic strategy for AKI.

Graphical abstract

Axolotl-inspired GIPL nanocages integrate CT imaging, KIM-1-mediated injured tubule targeting, and idebenone-driven mitochondrial protection to alleviate ischemia-reperfusion acute kidney injury by suppressing mTOR activation, enhancing mitophagy, and reducing apoptosis.

Highlights

• •Axolotl-inspired GIPL enables AKI theranostics.
• •LTH peptide directs GIPL to KIM-1-positive injured tubules.
• •Idebenone-loaded nanocages restore mitochondrial homeostasis.
• •GIPL inhibits mTOR signaling to enhance mitophagy and reduce apoptosis.
• •GIPL alleviates renal oxidative stress, inflammation, and dysfunction.

1. Introduction

Nature has long served as a source of inspiration for biomedical innovation. Among the most remarkable phenomena in the animal kingdom is the extraordinary regenerative capacity of the axolotl (Ambystoma mexicanum), a salamander capable of fully regenerating entire limbs and organs throughout its lifespan [1], [2]. Recent single-cell transcriptomic studies have unveiled a previously unrecognized “mitochondria-related cell population” that plays a vital energy-supporting role during axolotl limb regeneration [3]. This specialized cell cluster provides critical metabolic support through enhanced mitochondrial activity, enabling the sustained energy supply required for tissue reconstruction [3]. Similarly, in zebrafish heart regeneration, metabolic reprogramming and mitochondrial function restoration have been identified as essential drivers of cardiomyocyte proliferation and successful tissue repair [4], [5]. These findings from regenerative model organisms underscore a fundamental principle: robust mitochondrial function is indispensable for tissue regeneration and cellular recovery following injury [6], [7].

In stark contrast to these regenerative species, mammalian organs exhibit severely limited repair capacity following acute insults [8], [9]. Acute kidney injury (AKI) has emerged as a major global public health concern, with an estimated annual incidence of 13 million cases and associated mortality reaching 1.7 million deaths [10], [11]. Its clinical diagnosis remains heavily dependent on delayed indicators such as elevations in serum creatinine and reductions in urine output, both of which typically manifest only after considerable tubular injury has already occurred, thereby narrowing the therapeutic window. Current clinical management primarily focuses on supportive care, including fluid resuscitation, hemodynamic stabilization, and renal replacement therapy, none of which specifically targets the underlying pathophysiology of AKI [10], [12]. Consequently, a subset of patients fails to achieve renal recovery, with in-hospital mortality exceeding 50% in severe cases and survivors facing a markedly elevated risk of progression to chronic kidney disease and end-stage renal disease [10], [12]. At the cellular level, renal proximal tubular epithelial cells, owing to their high metabolic activity and dependence on mitochondrial oxidative phosphorylation for energy-intensive reabsorption processes, are the primary targets of AKI-related damage [13], [14]. Ischemia-reperfusion (IR) injury or nephrotoxic drug exposure leads to mitochondrial dysfunction, characterized by excessive reactive oxygen species (ROS) production, membrane potential dissipation, and ATP depletion, ultimately triggering apoptotic cascades [15], [16]. Unlike the axolotl's ability to rapidly activate mitochondrial repair programs upon injury, damaged tubular cells in mammalian kidneys often fail to restore mitochondrial homeostasis, leading to persistent cellular dysfunction and progression to chronic kidney disease [14]. Current clinical approaches, however, engage none of the subcellular pathology that drives tubular injury, leaving the mitochondrial roots of AKI therapeutically unsolved [10], [12]. Therefore, developing therapeutic strategies that target mitochondrial dysfunction in renal tubular cells represents a critical unmet need in AKI treatment.

The concept of cascade targeting, which achieves sequential cell-to-subcellular localization, has emerged as an ideal strategy for precision nanomedicine. At the cellular level, kidney injury molecule-1 (KIM-1), a type I transmembrane glycoprotein minimally expressed in healthy kidneys but dramatically upregulated on injured proximal tubular epithelial cells, serves as an ideal molecular target for selective delivery to damaged cells [17]. This damage-inducible expression pattern transforms the pathological marker into a therapeutic homing beacon, enabling disease signal-driven precision targeting [17], [18], [19]. At the subcellular level, mitochondria, as the epicenter of AKI pathology, represents the ultimate therapeutic destination where intervention can most effectively interrupt the vicious cycle of oxidative stress and cellular damage [16], [20]. Furthermore, integrating imaging and therapeutic functions within a single platform may allow nanoparticle distribution to be visualized alongside targeted drug delivery.

Idebenone (IDB), a synthetic short-chain benzoquinone analog of coenzyme Q10, represents an ideal therapeutic cargo for mitochondria-targeted delivery [21], [22]. Its amphiphilic structure facilitates penetration across mitochondrial membranes, where it functions as an electron carrier in the respiratory chain and scavenges ROS [22], [23], [24]. These properties have led to its clinical approval for mitochondrial diseases such as Leber's hereditary optic neuropathy [25]. However, the therapeutic potential of IDB in AKI remains unexplored due to its poor water solubility and lack of kidney specificity, necessitating advanced delivery platforms to achieve targeted mitochondrial protection in injured tubular cells [26]. Moreover, gold-silver nanocages (GSNCs) represent an attractive nanoplatform for theranostic applications in AKI. Their hollow porous architecture enables efficient encapsulation of hydrophobic drugs, while the high X-ray attenuation coefficient of gold provides excellent computed tomography (CT) contrast for imaging [27], [28], [29], [30]. The combination of CT contrast and therapeutic delivery within a single platform allows nanoparticle distribution to be visualized alongside targeted treatment. However, existing kidney-targeted nanomedicines still face key limitations: most rely on passive accumulation via disrupted glomerular filtration and lack cell-type specificity, few achieve subcellular delivery to injury-relevant organelles such as mitochondria, and virtually none integrate diagnostic imaging with therapy in a single platform. A rationally designed system that simultaneously addresses these three gaps is therefore needed.

We hypothesized that a cascade-targeted nanoplatform coupling KIM-1–mediated tubular recognition, mitochondriotropic delivery of IDB, and intrinsic CT contrast would restore mitochondrial homeostasis in injured tubular cells while allowing non-invasive visualization of its renal biodistribution. To translate this rationale into a testable system, and inspired by the mitochondria-centric regenerative mechanisms observed in salamanders and guided by the principle of cascade-programmed targeting, we herein designed and validated GSNCs@IDB@PEG@LTH nanoparticles (GIPL) to achieve hierarchical cell-to-organelle delivery for therapy of AKI (Fig. 1). GSNCs were selected as the nanocarrier due to their hollow porous architecture for efficient drug encapsulation and high X-ray attenuation coefficient for CT imaging, enabling visualization of nanoparticle biodistribution. The design integrates two sequential targeting steps: (i) surface-conjugated LTH peptide specifically recognizes KIM-1 receptors on damaged tubular epithelial cells, triggering receptor-mediated endocytosis (cell-level targeting); and (ii) released IDB, driven by its inherent mitochondriotropic properties, enriches within mitochondria (subcellular-level targeting). This programmed cascade delivery is designed to direct the therapeutic payload to the precise subcellular site of mitochondrial dysfunction, thereby enhancing therapeutic efficacy while limiting off-target effects. Meanwhile, the intrinsic CT contrast capability of GSNCs enables visualization of nanoparticle accumulation in injured kidneys.

Fig. 1.

Schematic illustration of the axolotl-inspired GIPL nanoplatform for targeted therapy and CT visualization in AKI.

2. Results and discussion

2.1. Biomimetic design rationale and therapeutic target identification

To establish the therapeutic strategy for AKI, we first explored the molecular mechanisms underlying tissue regeneration and injury. Axolotls possess remarkable regenerative capacity, making them ideal models for identifying conserved regenerative pathways. Transcriptomic analysis of axolotl limb regeneration revealed significant upregulation of genes involved in oxidative stress response, apoptotic signaling, and metabolic pathways (Fig. 2A). Using the GSE116615 dataset, differential expression analysis comparing the 24-h and 0-h time points identified 819 downregulated genes and 816 upregulated genes. Gene Ontology (GO) enrichment analysis identified ROS metabolic process (GO:0072,593), intrinsic and extrinsic apoptotic signaling pathways (GO:0097,193, GO:0097,191), and response to oxidative stress (GO:0006979) as key biological processes. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis further highlighted the involvement of FoxO signaling, cellular senescence, apoptosis, and AGE-RAGE signaling pathways in mitochondrial function, suggesting that modulation of mitochondrial function may facilitate tissue repair. We further analyzed the single-cell transcriptomic datasets GSE131685 and GSE174220, which include normal kidney samples and AKI samples, respectively. t-SNE visualization of integrated data identified seven distinct cell populations: proximal tubule cells, collecting duct cells, distal tubule cells, loop of Henle cells, myeloid cells, natural killer T cells, and smooth muscle cells (Fig. 2B). Cell type annotation was validated using canonical marker genes (Fig. 2C). Cell proportion analysis showed that proximal tubule cells constitute the dominant population in kidney tissue, while their proportion markedly decreased in the AKI group (Fig. 2D). Critically, mitochondrial function scoring across cell types demonstrated significantly lower mitochondrial activity in proximal tubule cells of AKI kidneys compared to controls (Fig. 2E and F, ***P < 0.001). Pathway expression analysis further confirmed elevated AGE-RAGE signaling, apoptosis, cellular senescence, and FoxO signaling in AKI tissues (Fig. S1). These findings indicated that enhanced mitochondrial bioenergetics enables regenerative capacity in axolotl, while mitochondrial dysfunction contributes to tubular injury and maladaptive repair in AKI (Fig. 2G). This is in line with recent reports positioning mitochondrial injury not as a downstream casualty but as a central driver of AKI progression [31], [32].

Fig. 2.

Biomimetic therapeutic strategies inspired by axolotl regeneration and AKI mechanisms. (A) Volcano plot and pathway enrichment analysis (GO and KEGG) of differentially expressed genes during axolotl limb regeneration, revealing upregulation of oxidative stress, apoptotic signaling, and metabolic pathways. (B) t-SNE visualization of integrated scRNA-seq data from control and AKI human kidney tissues, identifying seven distinct cell populations. (C) Dot plot displaying canonical marker gene expression profiles across identified cell types. (D) Cell type proportions in control vs AKI groups. (E) Mitochondrial function scores across cell types. (F) Comparison of mitochondrial function scores between AKI and control groups. (G) Schematic model depicting the divergent roles of mitochondrial metabolism in tissue fate determination. Enhanced mitochondrial bioenergetics underlies the regenerative capacity in axolotl, whereas mitochondrial dysfunction, manifested as bioenergetic failure and excessive oxidative stress, drives tubular injury and maladaptive repair in AKI. ***P < 0.001.

Based on these results, we employed network pharmacology to evaluate the therapeutic potential of IDB, a mitochondrial-targeted antioxidant, for AKI treatment. Database screening identified 94 overlapping targets between IDB and AKI, representing 1.2% of total AKI-related genes (Fig. S2A and B). Protein-protein interaction (PPI) network analysis identified hub genes with high degree values, including AKT1, EGFR, mTOR, MDM2, PIK3CA, MAPK1, KDR, MAPK14, MAPK8, and CDK2 (Fig. S2C and D). Among these, mTOR emerged as a central hub connecting autophagy regulation, apoptosis, and mitochondrial homeostasis. This is consistent with previous studies demonstrating that aberrant mTOR signaling plays a critical role in AKI pathogenesis: hyperactivation of mTOR suppresses mitophagy and promotes apoptosis, leading to accumulation of damaged mitochondria and subsequent mitochondrial dysfunction [33], [34], [35]. KEGG enrichment analysis revealed that the common targets were significantly enriched in the mTOR signaling pathway, autophagy, MAPK signaling, apoptosis, HIF-1 signaling, and PI3K-AKT signaling pathways (Fig. S2E), which aligned with the dysregulated pathways identified in our AKI scRNA-seq analysis. Molecular docking simulations demonstrated favorable binding affinity between IDB and mTOR (binding energy: −7.5 kcal/mol), with key interactions at ARG-48, ASP-51, and GLN-54 residues (Fig. S3). To confirm that this reduction reflects a dose-dependent action on mTOR, the following experiments were conducted. Cell viability assays first confirmed that free IDB (0–20 μM) had no cytotoxicity to cells, defining a safe working range for the subsequent mechanistic experiments (Fig. S4A). Within this range, free IDB (5, 10, 20 μM) reduced p-mTOR expression in a dose-dependent manner (Fig. S4B and C). These analyses collectively support the hypothesis that IDB may exert therapeutic effects by inhibiting mTOR activation, thereby restoring mitophagy, suppressing apoptosis, and ultimately ameliorating mitochondrial dysfunction in AKI. Taken together, the axolotl transcriptomic profiling and the human AKI scRNA-seq converge on a shared principle: after injury, the balance between mitochondrial energy production and redox homeostasis determines whether injured tubular cells recover or undergo cell death. In this sense, “axolotl-inspired” refers to the biological principle guiding our strategy rather than direct biomimicry. Guided by this principle, we engineered GIPL, a KIM-1-targeted GSNCs loaded with IDB for cascade delivery to mitochondria of injured tubular cells, as detailed in the following sections.

2.2. Synthesis and Characterization of GIPL nanoparticles

To overcome the poor aqueous solubility of IDB and achieve targeted delivery to injured renal tubular cells, we developed a stepwise synthesis strategy for GIPL nanoparticles. KIM-1 is minimally expressed in normal kidneys but dramatically upregulated in injured proximal tubular cells, making it an ideal target for AKI-specific drug delivery [36], [37]. We designed GIPL by conjugating the KIM-1-binding peptide LTH (LTHVVWLC) to PEGylated, IDB-loaded GSNCs. Transmission electron microscopy (TEM) revealed the morphological evolution at each synthesis stage (Fig. 3A and B). Unmodified GSNCs exhibited well-defined, porous cage-like structures with smooth surfaces and an average diameter of 61.36 ± 9.67 nm. Following IDB loading, the porous architecture became less distinct due to uniform physical adsorption of IDB within the nanocages (G@IDB, 60.52 ± 8.65 nm). PEGylation (GI@PEG) introduced a low electron-density corona around the particles, increasing the diameter to 71.01 ± 7.90 nm. Final conjugation with the LTH targeting peptide yielded GIPL nanoparticles with slightly roughened surfaces and a mean diameter of 79.35 ± 12.95 nm (Fig. 3D).

Fig. 3.

Synthesis and Characterization of GIPL Nanoparticles. (A) TEM images of GSNCs, G@IDB, GI@PEG, and GIPL (insets: schematic illustrations). (B) High-magnification TEM images revealing the core-shell structure at each step of GIPL synthesis. (C) TEM image of GIPL with corresponding elemental mapping. (D) Size distribution histograms of GSNCs, G@IDB, GI@PEG, and GIPL nanoparticles. (E) UV-Vis absorption spectra of GSNCs, G@IDB, GI@PEG, and GIPL nanoparticles (20 μg/mL) dispersed in ddH₂O, recorded over the wavelength range of 200–900 nm at room temperature. (F) Zeta potential measurements of GSNCs, G@IDB, GI@PEG, and GIPL. (G) Cumulative release profile of IDB from GIPL nanoparticles in ddH₂O at 37 °C over 96 h. Abbreviations: GSNCs, gold-silver nanocages; G@IDB, idebenone-loaded gold-silver nanocages; GI@PEG, PEGylated idebenone-loaded gold-silver nanocages; GIPL, LTH peptide-functionalized PEGylated idebenone-loaded gold-silver nanocages; IDB, idebenone.

Energy dispersive spectroscopy (EDS) elemental mapping confirmed the hierarchical architecture of GIPL (Fig. 3C). Gold and silver elements were enriched on the nanoparticle surface, forming a well-defined bimetallic alloy. Carbon, nitrogen, oxygen and sulfur signals were also detected in the nanoparticles. These signals may indicate the presence of IDB, PEG and LTH in the final GIPL nanocomposite. Quantitative ICP-OES analysis further revealed that the mass ratio of Au to Ag in the GIPL nanoparticles was 1.10 ± 0.06 (Table S1), which was consistent with the previously reported structure of GSNCs [38].

UV-Vis spectroscopy demonstrated that GSNCs exhibited a characteristic surface plasmon resonance peak at approximately 800 nm (Fig. 3E), which remained stable after IDB loading, indicating preserved electronic structure. PEG modification induced new absorption in the 200–250 nm range, which was consistent with the reference spectrum of NH₂-PEG-SH. The LTH conjugation further introduced a new peak around 280 nm, corresponding to the tryptophan residue in the LTH peptide (Fig. S5). FTIR spectroscopy further confirmed the stepwise functionalization of GIPL nanoparticles (Fig. S6). The broad peak at approximately 3350–3450 cm⁻¹ was attributed to the O–H stretching vibration of adsorbed water. The peaks at 2919 and 2851 cm⁻¹ were assigned to the C–H stretching vibrations of residual PVP rather than the gold–silver framework, which is IR-inactive. The peak at 1657 cm⁻¹ was also attributed to the C=O stretching vibration of residual PVP. The FTIR spectrum of G@IDB was similar to that of GSNCs, with no obvious IDB-specific peaks observed. This was likely because IDB was physically adsorbed within the nanoparticle cages, and its characteristic signals were masked by the strong background signals from PVP and the nanoparticle framework. After PEGylation, new peaks appeared at 1093 and 1014 cm⁻¹, corresponding to the characteristic C–O–C stretching vibrations of PEG. After LTH coupling, new peaks appeared at 1542 cm⁻¹ (amide II) and 1352 cm⁻¹ (amide III), indicating the successful introduction of peptide functionalities. Additionally, Zeta potential measurements demonstrated successful surface modifications: GSNCs exhibited negative potential (−20 ± 0.41 mV), which shifted to positive values after PEG modification (5.35 ± 0.51 mV) and further increased following LTH conjugation (7.30 ± 0.53 mV) (Fig. 3F). These results confirm the successful hierarchical assembly of the GIPL platform.

In addition to the qualitative confirmation, quantitative analysis further verified the composition of GIPL. The drug loading content (DLC) and encapsulation efficiency (EE) of IDB in GIPL nanoparticles were quantified by HPLC. GIPL exhibited a DLC of 19.80 ± 1.32% (w/w) and an EE of 24.70 ± 2.05% (Table S2, n = 3), confirming that IDB was efficiently and reproducibly introduced into the hollow porous cavity of the GSNCs. The bicinchoninic acid assay analysis revealed that the coupling efficiency of the LTH peptide on the GIPL was 20.12 ± 1.36% (Table S3), providing quantitative evidence for the successful functionalization of LTH. The residual reagent analysis further confirmed the medicinal-grade purity of the GIPL. The headspace gas chromatography-mass spectrometry quantitative results showed that the residual organic solvents (ethylene glycol: 385 ± 52 ppm, methanol: 148 ± 22 ppm, acetone: 65 ± 12 ppm) were all far below the daily allowable exposure limit stipulated by ICH Q3C (R8) (Table S4). The ICP-MS analysis also indicated that the free metal ions (Ag⁺: 1.24 ± 0.35 ppm, Au³⁺: 0.42 ± 0.08 ppm) were all below the element impurity limit stipulated by ICH Q3D (R2) (10 ppm).

Time-dependent drug release results in ddH₂O at 37 °C revealed a sustained release characteristic of IDB from GIPL nanoparticles, with cumulative release reaching 45.94 ± 5.79% at 24 h and 71.80 ± 12.06% at 48 h. The release entered a plateau phase after 72 h (91.91 ± 6.13%), indicating essentially complete drug release (Fig. 3G). This sustained release behavior is attributed to the diffusion of IDB through the porous cavities of the GSNCs driven by concentration gradients. Consistent release curves were also observed in the SBF and DMEM/F12 + 10% FBS culture systems (Fig. S7), with both media reaching a plateau at 72 h (90.24 ± 6.15% and 93.85 ± 9.25%, respectively) and showing no significant differences among the three media (P > 0.05). These results indicated that GIPL could achieve stable and controllable release of IDB under physiological relevant conditions. It is worth noting that IDB is encapsulated within the GIPL nanoparticle structure and primarily released intracellularly through endocytosis. This design minimizes the potential instability of free IDB in the extracellular environment and represents an inherent advantage of the GIPL nanocarrier.

DLS analysis showed that GIPL maintained excellent colloidal stability in ddH₂O, DMEM/F12, and DMEM/F12 + 10% FBS at 37 °C over 7 days, with Dh changes < 5% and PDI < 0.25 (Table S5). Three independent batches showed consistent Dh and PDI values (inter-batch RSD < 5%), confirming batch-to-batch reproducibility (Table S6). Molecular docking analysis confirmed specific binding between the LTH peptide and the KIM-1 receptor, with a predicted total binding free energy (ΔG) of −33.81 kcal/mol and key interactions at Gly94, Glu91, and Gln41 (2.8–3.0 Å) (Table S7 and Fig. S8), supporting targeted delivery to injured proximal tubular cells. These results confirm successful GIPL assembly and validate our dual-targeting design, in which LTH enables injury-responsive tubular localization and IDB delivers mitochondria-specific protection.

2.3. GIPL nanoparticles attenuate oxidative stress and mitochondrial dysfunction In vitro

Given that oxidative stress–mediated mitochondrial damage plays a pivotal role in the occurrence and progression of AKI, we examined the capacity of GIPL to attenuate the ROS–mitochondrial depolarization–apoptosis cascade in renal tubular epithelial cells [39], [40]. The therapeutic efficacy of GIPL was evaluated using H₂O₂-induced oxidative injury in HK-2 human renal tubular epithelial cells. Biosafety assessments confirmed that GIPL and its intermediate formulations exhibited no significant cytotoxicity at the working concentrations (Figs. S9 and S10), with hemolysis rates below 5% (Fig. S11). Cellular uptake studies using Cy5.5-labeled nanoparticles demonstrated significantly enhanced accumulation of GIPL in H₂O₂-treated cells compared to other formulations, attributable to upregulated KIM-1 expression following oxidative injury (Fig. S12) [36], [37]. This KIM-1-mediated targeting strategy ensures preferential drug delivery to injured tubular cells, potentially improving therapeutic specificity while minimizing off-target effects. Western blot analysis revealed a marked upregulation of KIM-1 following H₂O₂ treatment (Fig. S13A and B). Consistently, confocal imaging demonstrated significantly enhanced uptake of Cy5.5-GIPL in H₂O₂-treated HK-2 cells compared with untreated controls (Fig. S13C and D), confirming that KIM-1 mediates the targeting process. While cellular uptake confirms nanoparticle entry, successful delivery to the therapeutic target requires further verification at the subcellular level. As IDB exerts its function within mitochondria [41], we next assessed whether GIPL enables preferential mitochondrial accumulation of the released IDB. Further confocal analysis showed pronounced co-localization between BODIPY-GIPL (red) and MitoTracker Green in H₂O₂-injured HK-2 cells, yielding a Pearson's correlation coefficient of 0.78 (Fig. S14A and B), in agreement with the well-documented mitochondriotropic behavior of IDB. Together, these results establish that GIPL achieves efficient cell-to-mitochondria delivery, thereby providing the structural foundation for the mitochondrial protective effects characterized in the following section.

DCFH-DA staining revealed that H₂O₂ treatment dramatically increased intracellular ROS levels, which were progressively attenuated by GSNCs, G@IDB, GI@PEG, and maximally reduced by GIPL treatment (Fig. 4A–E, *P < 0.05). MitoSOX staining specifically demonstrated that GIPL treatment reduced mitochondrial superoxide generation by approximately 75% compared to the H₂O₂ group (Fig. 4B–F, ****P < 0.0001). DCFH-DA reflects total intracellular ROS levels, whereas MitoSOX selectively detects mitochondrial superoxide. The substantial attenuation of MitoSOX signal suggests that GIPL targets mitochondrial ROS production at its source, consequently reducing the overall cellular oxidative stress detected by DCFH-DA. JC-1 staining revealed that oxidative injury caused mitochondrial membrane potential (ΔΨm) dissipation, evidenced by a shift from red (J-aggregates, high ΔΨm) to green (J-monomers, low ΔΨm) fluorescence. GIPL treatment effectively restored the red/green fluorescence ratio, indicating preserved mitochondrial membrane integrity (Fig. 4C–G, ***P < 0.001). Maintenance of ΔΨm is critical for ATP synthesis and prevention of mitochondrial permeability transition pore opening, which triggers the intrinsic apoptotic cascade [42].

Fig. 4.

GIPL nanoparticles attenuate H₂O₂-induced oxidative stress and mitochondrial dysfunction in vitro. (A) Representative fluorescence images of intracellular ROS levels detected by DCFH-DA staining. Scale bar: 200 μm. (B) Confocal microscopy images of mitochondrial superoxide detected by MitoSOX staining (red) with DAPI nuclear counterstaining (blue). Scale bar: 30 μm. (C) Mitochondrial membrane potential assessed by JC-1 staining. Red fluorescence indicates JC-1 aggregates (high membrane potential), green fluorescence indicates JC-1 monomers (low membrane potential). Scale bar: 30 μm. (D) Flow cytometry analysis of cell apoptosis using Annexin V-FITC/PI double staining. (E–H) Quantitative analysis of (E) relative ROS fluorescence intensity, (F) MitoSOX fluorescence intensity, (G) JC-1 red/green fluorescence ratio, and (H) apoptosis rate. Data are presented as mean ± SD (n = 3). ns, not significant; *P < 0.05, ***P < 0.001, ****P < 0.0001.

Flow cytometry analysis using Annexin V-FITC/PI double staining demonstrated that GIPL significantly reduced apoptosis rates from approximately 23% in the H₂O₂ group to approximately 8% (Fig. 4D–H, ***P < 0.001). Inflammation is a major contributor to AKI and other ROS-driven pathologies, with the interplay between renal tubular epithelial cells and infiltrating immune cells playing a complex role in tissue repair and functional recovery [43]. The qRT-PCR analysis revealed that GIPL treatment suppressed inflammatory cytokine mRNA expression, including IL-1β, IL-6, TNF-α, and MCP-1 (Fig. S15, *P < 0.05 to ****P < 0.0001). Furthermore, Western blot analysis confirmed that GIPL downregulated pro-apoptotic proteins (cleaved Caspase-3, Bax, Cytochrome c(Cyto-C)) while upregulating anti-apoptotic Bcl-2 (Fig. S16). The observed reduction in Cyto-C release is consistent with preserved mitochondrial membrane integrity, as Cyto-C translocation from mitochondria to cytosol is a hallmark of intrinsic apoptosis initiation [40], [44]. Collectively, these findings demonstrate that GIPL confers robust protection against oxidative stress-induced mitochondrial damage and apoptosis. To elucidate the underlying molecular mechanisms, we performed transcriptomic analysis to identify the key signaling pathways mediating these protective effects. It should be noted that the H₂O₂-induced injury model used in this study recapitulates the oxidative-stress component of tubular damage but does not fully reproduce the hypoxia/reoxygenation insult or the inflammatory milieu in vivo. The therapeutic efficacy of GIPL was further evaluated in vivo using the IR-AKI mouse model, as detailed in the following sections.

2.4. Transcriptomic analysis reveals mTOR-Mediated regulation of mitophagy and apoptosis

To elucidate the molecular mechanisms underlying GIPL's therapeutic effects, RNA sequencing was performed on Control, H₂O₂-treated, and GIPL-treated HK-2 cells. Principal component analysis (PCA) demonstrated distinct clustering of the three groups, with PC1 and PC2 explaining 51.17% and 18.25% of total variance, respectively (Fig. S17). Differential expression analysis identified substantial transcriptomic alterations: 2503 upregulated and 1246 downregulated genes in Control vs AKI comparison, and 1323 upregulated and 1561 downregulated genes in AKI vs Treatment comparison (Fig. 5A and B, Fig. S18). Venn diagram analysis identified 833 genes that were upregulated in AKI and subsequently downregulated following GIPL treatment, representing genes rescued by therapy (Fig. 5C). Conversely, 2087 genes were shared between those downregulated in Control vs AKI and upregulated in AKI vs Treatment comparisons (Fig. 5D). GO and KEGG enrichment analysis of overlapping differentially expressed genes (DEGs) revealed significant involvement of mitochondrial-related processes, including mitochondrial genome maintenance, mitochondrial DNA replication, mitochondrial matrix, and ATPase complex (Fig. 5E). Critically, the mTOR signaling pathway, autophagy, mitophagy, and apoptosis pathways were significantly enriched, supporting our hypothesis that mTOR serves as a central upstream regulator controlling both mitophagy and apoptosis, which in turn determines mitochondrial fate and function.

Fig. 5.

Transcriptomic analysis reveals that GIPL nanoparticles regulate autophagy and apoptosis via the mTOR signaling pathway. (A, B) Volcano plots showing differentially expressed genes (DEGs) in Control vs AKI (A) and AKI vs Treatment (B) groups. (C) Venn diagram illustrating the overlap between upregulated genes in Control vs AKI and downregulated genes in AKI vs Treatment groups. (D) Venn diagram showing the overlap between upregulated genes in AKI vs Treat and downregulated genes in Control vs AKI groups. (E) Gene Ontology (GO) and KEGG pathway enrichment analysis of overlapping DEGs. BP, biological process; CC, cellular component; MF, molecular function. (F) Heatmap displaying the expression profiles of apoptosis- and autophagy-related genes. (G) Western blot analysis of mTOR, p-mTOR, P62, and LC3-I/LC3-II protein expression levels. GAPDH served as the loading control. (H–J) Quantitative analysis of p-mTOR/mTOR ratio (H), P62/GAPDH (I), and LC3-II/LC3-I ratio (J). Data are presented as mean ± SD (n = 3). ns, not significant; **P < 0.01.

Heatmap visualization of autophagy- and apoptosis-related genes demonstrated coordinated regulation by GIPL treatment (Fig. 5F). Pro-apoptotic genes (BID, BAX, BAD, CASP9) were downregulated, while autophagy-related genes (LC3B, ATG3, ATG5, ATG2B) showed differential modulation. Western blot validation confirmed that GIPL treatment reduced p-mTOR/mTOR ratio (Fig. 5G and H, **P < 0.01). Consistently, GIPL also exhibited a dose-dependent suppression of p-mTOR expression at 25, 50, and 100 μg/mL (Fig. S19). The consistent, dose-dependent suppression of p-mTOR by both free IDB and GIPL further supports mTOR inhibition as a key mechanism of GIPL-mediated renal protection. In the meanwhile, GIPL treatment decreased P62/GAPDH (Fig. 5I, **P < 0.01), and increased LC3-II/LC3-I ratio (Fig. 5J, **P < 0.01), collectively indicating enhanced autophagic flux. These findings reveal a coherent mechanistic cascade: GIPL inhibits mTOR phosphorylation, which relieves suppression of autophagy/mitophagy while simultaneously attenuating pro-apoptotic signaling. Specifically, enhanced mitophagy promotes the selective clearance of damaged mitochondria, preventing the accumulation of dysfunctional organelles that would otherwise amplify ROS production and trigger apoptosis. Concurrently, the suppression of pro-apoptotic signaling preserves mitochondrial membrane integrity, thereby blocking Cyto-C release and downstream caspase activation. Together, these coordinated effects restore mitochondrial homeostasis and function, thereby protecting renal tubular epithelial cells from oxidative injury.

2.5. In vivo biodistribution and renal targeting of GIPL

The renal targeting capability of GIPL was evaluated in a murine IR-AKI model. C57BL/6 mice underwent bilateral renal pedicle clamping for 30 min followed by reperfusion and nanoparticle administration via tail vein injection (Fig. 6A and B). In vivo fluorescence imaging using Cy5.5-labeled nanoparticles demonstrated progressive renal accumulation over 24 h (Fig. 6C). Quantitative analysis revealed significantly higher renal fluorescence intensity in GIPL-treated mice compared to GI@PEG at all time points, with peak accumulation at 6 h post-injection (Fig. S20, ***P < 0.001 to ****P < 0.0001).

Fig. 6.

In vivo biodistribution and CT imaging of GIPL in AKI mice. (A) Schematic of the renal ischemia-reperfusion injury model and treatment protocol. (B) Representative photographs of the kidney during ischemia and reperfusion. (C) In vivo and ex vivo fluorescence imaging at 3, 6, 12, and 24 h post-injection. (D) Fluorescence images show the accumulation of Cy5.5-labeled nanoparticles (red) in kidney sections. Nuclei were stained with DAPI (blue). Scale bar: 200 μm. (E) CT phantom images of GIPL at various concentrations. (F) Linear relationship between GIPL concentration and CT value. (G) In vivo CT imaging of AKI mice after different treatments. White arrows and red dashed circles indicate the kidneys.

At this peak time point, Ex vivo fluorescence imaging of excised kidneys confirmed enhanced GIPL accumulation (Fig. 6D), with quantification demonstrating approximately 3-fold higher signal intensity in GIPL vs GI@PEG groups (Fig. S21, ***P < 0.001). Confocal imaging of kidney sections harvested 6 h post-injection showed that the Cy5.5-GIPL signal was predominantly confined to KIM-1-positive tubular segments and co-localized with LTL (a proximal tubule marker) in the merged images (Fig. S22), confirming that the cascade targeting mechanism characterized in vitro is preserved in the IR-AKI kidney. To further provide direct quantitative evidence of renal targeting, we performed ICP-MS analysis of Au content in major organs at 6 h post-injection (Table S8). The GIPL group exhibited the highest renal Au accumulation (43.48 ± 1.80 μg/g), approximately 2.2-fold higher than the GI@PEG group (19.42 ± 1.15 μg/g), while non-targeted formulations (GSNCs, G@IDB, GI@PEG) showed comparable kidney accumulation. These ICP-MS results are consistent with the fluorescence imaging data, further confirming the KIM-1-mediated active targeting conferred by LTH peptide functionalization.

To further clarify the in vivo clearance pathway of GIPL nanoparticles, we conducted a dynamic quantitative analysis of Au concentrations in blood, major organs, urine, and feces using ICP-MS within 14 days post-injection (Fig. S23). The Au concentration in blood reached a peak on the first day (6.42 ± 0.85 μg/mL), and then rapidly decreased to 0.45 ± 0.18 μg/mL by the 14th day, indicating that GIPL has an efficient systemic clearance ability. It is notable that although GIPL significantly targets damaged renal tubular cells through the LTH peptide-mediated KIM-1 recognition (Fig. 6, Table S8), its final excretion mainly occurs through the hepatic-biliary-fecal pathway, with urine excretion being an auxiliary pathway. The concentrations of Au in feces reached a peak on day 3 (16.28 ± 1.62 μg/g), substantially exceeding the peak observed in urine (3.24 ± 0.48 μg/mL) on day 1. Urine excretion may reflect the increased permeability of the glomerular basement membrane (GFB) under AKI. Changes such as endothelial fenestrations, relaxation of the basement membrane, and degeneration of podocyte foot processes all increase the effective retention size of the GFB, thereby allowing some of the originally retained nanoparticles to be excreted from the body [45]. Overall, GIPL mainly targets damaged renal tubules, but is mainly cleared through the hepatic-biliary-fecal pathway, with urine being the secondary route. This supports its good biological safety and clinical translational potential.

Importantly, the intrinsic X-ray attenuation properties of GSNCs endow GIPL with CT imaging capability, enabling visualization of nanoparticle biodistribution. This theranostic capability allows nanoparticle biodistribution to be visualized alongside targeted delivery, which may facilitate future development of image-guided therapeutic strategies for AKI. CT imaging of GIPL solutions at different concentrations demonstrated a linear relationship between GIPL concentration and CT values (Fig. 6E, F, R² > 0.99), confirming the quantitative imaging potential of GIPL for assessing drug accumulation in target tissues. In vivo CT imaging clearly visualized GIPL accumulation in the kidneys of AKI mice, with markedly enhanced contrast compared to other treatment groups (Fig. 6G). This non-invasive imaging capability enables visualization of nanoparticle biodistribution, providing a foundation for future development of image-guided therapeutic strategies. Longitudinal correlation of CT signals with therapeutic response will require further validation. These results confirm that LTH peptide functionalization confers effective renal targeting via KIM-1 binding in injured proximal tubular epithelium. More importantly, combining CT visualization with targeted drug delivery in a single nanoplatform allows nanoparticle biodistribution and therapeutic delivery to injured tubules to be assessed together. This dual functionality supports further preclinical development of GIPL as a cascade-targeted theranostic for AKI.

2.6. GIPL ameliorates renal injury via mTOR-mitophagy-apoptosis axis In vivo with favorable biosafety

The therapeutic efficacy of GIPL was comprehensively evaluated in the IR-AKI mouse model. Histopathological analysis by H&E and PAS staining revealed severe tubular injury in AKI mice, characterized by tubular epithelial cell vacuolization, brush border loss, tubular dilation, and cast formation (Fig. 7A and B). GIPL treatment significantly reduced histological injury scores compared to other treatment groups (Fig. 7F and G, **P < 0.01). Renal function biomarkers, including serum creatinine (SCr) and blood urea nitrogen (BUN), were markedly elevated in AKI mice and significantly reduced by GIPL treatment to near-normal levels (Fig. 7K and L, ****P < 0.0001 and ***P < 0.001, respectively). The superior therapeutic efficacy of GIPL over intermediate formulations (GSNCs, G@IDB, GI@PEG) underscores the importance of the cascade targeting strategy: LTH peptide-mediated KIM-1 recognition ensures preferential accumulation in injured tubular cells, while the mitochondriotropic properties of IDB enable subcellular enrichment at the site of dysfunction. The parallel improvements in histology and renal function demonstrate that our cascade targeting approach effectively bridges mitochondrial protection to overall kidney recovery [46], [47], [48].

Fig. 7.

GIPL nanoparticles ameliorate renal injury, oxidative stress, and inflammation in AKI mice. (A, B) Representative images of H&E (A) and PAS (B) staining of kidney tissue sections. Scale bar: 100 μm. (C) DHE staining (red) showing ROS levels in kidney tissues with DAPI nuclear counterstaining (blue). Scale bar: 100 μm. (D) Immunofluorescence staining of kidney injury marker KIM-1 (red) with DAPI (blue). (E) TUNEL staining (green) showing apoptotic cells with DAPI (blue). (F, G) Quantitative scoring of H&E (F) and PAS (G) staining. (H–J) Quantitative analysis of DHE fluorescence intensity (H), KIM-1 fluorescence intensity (I), and TUNEL-positive cells (J). (K, L) Serum levels of blood urea nitrogen (BUN) (K) and serum creatinine (SCr) (L). Renal tissue levels of glutathione (GSH) (M). Data are presented as mean ± SD (n = 6). ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Dihydroethidium (DHE) staining demonstrated that GIPL effectively reduced renal ROS levels (Fig. 7C–H, ***P < 0.001). Immunofluorescence staining for KIM-1 revealed dramatically elevated expression in AKI kidneys that was substantially attenuated by GIPL treatment (Fig. 7D–I, ****P < 0.0001). The reduction in KIM-1 expression following GIPL treatment not only reflects amelioration of tubular injury but also suggests a potential feedback mechanism: as tubular damage resolves, the targeting ligand's receptor becomes downregulated, which may help prevent excessive nanoparticle accumulation in recovering tissue. TUNEL staining confirmed that GIPL significantly reduced renal tubular cell apoptosis (Fig. 7E–J, ****P < 0.0001). Biochemical analysis revealed that GIPL treatment restored glutathione (GSH) levels (Fig. 7M, *P < 0.05), increased superoxide dismutase (SOD) activity (Fig. S24A, **P < 0.01), and reduced malondialdehyde (MDA) content (Fig. S24B, ****P < 0.0001) in renal tissues. Furthermore, serum pro-inflammatory cytokines TNF-α and IL-6 were significantly decreased following GIPL treatment (Fig. S24C and D, ****P < 0.0001 and ***P < 0.001, respectively), with qRT-PCR analysis confirming reduced mRNA expression of IL-1β, IL-6, TNF-α, and MCP-1 in kidney tissues (Fig. S25). This anti-inflammatory effect may be attributed to both direct mitochondrial protection (damaged mitochondria release DAMPs that activate inflammatory pathways) and reduced tubular cell death, which otherwise triggers sterile inflammation and immune cell infiltration [49], [50], [51], [52].

TEM examination of renal tubular cells provided direct evidence of mitochondrial protection. AKI mice displayed severe mitochondrial ultrastructural damage, including matrix swelling, cristae disorganization, and outer membrane rupture (Fig. 8A). In contrast, control kidneys exhibited elongated mitochondria with well-organized parallel cristae. Remarkably, GIPL treatment substantially restored mitochondrial ultrastructure, preserving cristae integrity and preventing matrix swelling, while intermediate formulations (GSNCs, G@IDB, GI@PEG) showed only partial protective effects. The preservation of cristae architecture is particularly significant, as cristae morphology directly influences respiratory chain supercomplex assembly and ATP synthesis efficiency [53], [54]. These ultrastructural findings provide morphological evidence supporting the functional restoration of mitochondrial bioenergetics observed in our in vitro studies.

Fig. 8.

GIPL nanoparticles restore mitochondrial ultrastructure and regulate autophagy and apoptosis via the mTOR signaling pathway in vivo. (A) Transmission electron microscopy (TEM) images of renal tubular cells showing mitochondrial ultrastructure at different magnifications. Red dashed boxes indicate the magnified regions. Scale bars: 10 μm (top), 4 μm (middle), and 1 μm (bottom). (B–D) Immunohistochemical staining of kidney tissue sections for p-mTOR (B), P62 (C), and LC3B (D). Scale bar: 100 μm. (E–H) qRT-PCR analysis of relative mRNA expression levels of Bax (E), Bcl-2 (F), Lc3 (G), and P62 (H) normalized to GAPDH. Data are presented as mean ± SD (n = 6). ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001.

To validate the mTOR-mitophagy-apoptosis-mitochondrial dysfunction axis in vivo, immunohistochemical analysis was performed. AKI induced marked elevation of p-mTOR expression in renal tubular cells, which was significantly attenuated by GIPL treatment (Fig. 8B). Consistent with the proposed mechanism whereby mTOR inhibition relieves suppression of autophagy, P62 expression was reduced (Fig. 8C) and LC3B expression was increased (Fig. 8D) in GIPL-treated kidneys, indicating enhanced autophagic/mitophagic activity. qRT-PCR quantification validated these observations at the transcriptional level: GIPL treatment significantly reduced Bax mRNA levels (Fig. 8E, ***P < 0.001), increased Bcl-2 mRNA levels (Fig. 8F, **P < 0.01), elevated Lc3 mRNA levels (Fig. 8G, ***P < 0.001), and decreased P62 mRNA expression (Fig. 8H, *P < 0.05). Additionally, Caspase-9 mRNA expression was significantly reduced (Fig. S26, **P < 0.01), indicating suppression of the intrinsic mitochondrial apoptotic pathway. The specific downregulation of Caspase-9, an initiator caspase activated by Cyto-C release from damaged mitochondria, further corroborates that GIPL protects mitochondrial membrane integrity and prevents activation of the intrinsic apoptotic cascade [55], [56]. Collectively, these in vivo results demonstrate that GIPL inhibits mTOR phosphorylation, thereby enhancing mitophagy to clear damaged mitochondria while simultaneously suppressing apoptosis to prevent further mitochondrial injury, ultimately restoring mitochondrial function and protecting against AKI.

Finally, comprehensive biosafety assessment was performed to evaluate the translational potential of GIPL. H&E staining of major organs (heart, liver, spleen, lung, kidney) revealed no significant histopathological abnormalities in healthy mice receiving intravenous GIPL administration (Fig. S27). Body weight monitoring showed no significant differences between groups (Fig. S28A), and hematological parameters including white blood cell (WBC) and red blood cell (RBC) counts remained within normal ranges (Fig. S28B–C). Liver function indicators (ALT, AST) and renal function markers (SCr, BUN) showed no significant alterations (Figure S28D–E, H–I). Serum inflammatory cytokines (IL-6, TNF-α) were not elevated (Fig. S28F–G), and oxidative stress markers (SOD, MDA, GSH) in healthy kidney tissues remained stable (Fig. S28J–L). These results collectively demonstrate excellent biocompatibility and safety of GIPL nanoparticles, supporting their potential for clinical translation. Nevertheless, this study has several limitations. The therapeutic efficacy of GIPL was evaluated only in the IR-AKI model, and its effects on other subtypes of AKI, including nephrotoxic AKI and sepsis-associated AKI, have yet to be determined. In addition, all in vivo investigations were performed in mice. Further studies using large-animal models are warranted to confirm the safety and efficacy of GIPL and advance the clinical translation of this strategy.

3. Conclusion

In this study, we developed GIPL, a biomimetic theranostic nanoplatform inspired by the mitochondria-centered regenerative mechanisms of axolotls. Three key innovations highlight the present work: (i) identification of mTOR as a central therapeutic target through integrated single-cell transcriptomic and network pharmacology analyses, with experimental validation demonstrating that mTOR inhibition simultaneously enhances mitophagy and suppresses apoptosis to restore mitochondrial homeostasis; (ii) engineering of a cascade-targeting strategy achieving hierarchical cell-to-organelle delivery via KIM-1-mediated tubular cell recognition and IDB-driven mitochondrial accumulation; and (iii) integration of CT visualization of nanoparticle accumulation with targeted therapy within a single platform. GIPL demonstrated robust therapeutic efficacy both in vitro and in vivo, effectively scavenging ROS, restoring mitochondrial membrane potential, inhibiting mTOR phosphorylation to enhance mitophagy, suppressing apoptosis, and significantly ameliorating tubular injury and renal dysfunction in IR-AKI mice. The favorable biosafety profile further supports its translational potential.

Experimental procedures

A murine ischemia-reperfusion acute kidney injury model was established using bilateral renal pedicle clamping. Briefly, mice were fasted for 6 h and deprived of water for 4 h before surgery. Anesthesia was induced by intraperitoneal injection of 1% sodium pentobarbital at 50 mg/kg. After abdominal disinfection, a midline abdominal incision was made to expose both kidneys. Bilateral renal pedicles were clamped with microvascular clamps for 30 min to induce renal ischemia, as confirmed by the kidneys turning dark purple. The clamps were then removed to allow reperfusion, as indicated by the gradual recovery of kidney color. Sham-operated mice underwent the same surgical procedure without renal pedicle clamping.

Immediately after reperfusion, mice received tail vein injection of PBS or different nanoparticle formulations, including GSNCs, GSNCs@IDB, GSNCs@IDB@PEG, and GIPL. At predetermined time points, mice were subjected to in vivo imaging or sacrificed for sample collection. Blood, kidneys, and major organs were harvested for renal function assessment, histological analysis, immunofluorescence staining, oxidative stress evaluation, inflammatory cytokine detection, transmission electron microscopy, and molecular analyses.

CRediT authorship contribution statement

Zihao Xu: Writing – original draft, Data curation, Conceptualization. Yao Wan: Writing – original draft, Data curation, Conceptualization. Yuhang Shang: Writing – original draft, Data curation. Kangjie He: Formal analysis. Chao Jiang: Formal analysis. Gengyu Du: Formal analysis. Yijun Zhao: Formal analysis. Qibo Hu: Formal analysis. Yang Zhang: Formal analysis. Gensheng Zhang: Conceptualization. Xiang Yan: Supervision.

Consent to participate

This study did not involve newly recruited human participants, identifiable human specimens, or individual-level clinical data. Publicly available transcriptomic datasets were analyzed in this study; therefore, additional ethical approval and informed consent for human participants were not required.

Ethical approval and regulatory compliance

All animal experiments in this study were approved by the Institutional Animal Care and Use Committee of Zhejiang University (Approval Number: ZJU20230344). All procedures were performed in accordance with institutional guidelines for the care and use of laboratory animals. This study was conducted in compliance with the ARRIVE guidelines to ensure transparent, complete, and accurate reporting of all in vivo experiments.

Sample size justification and the 3Rs principle

The sample size was determined based on previous experience with renal ischemia-reperfusion injury models and the need to obtain reliable statistical power while minimizing animal use. For therapeutic efficacy experiments, six animals were included in each group. This study was designed in accordance with the 3Rs principle: in vitro experiments using HK-2 cells were performed before in vivo validation to reduce unnecessary animal use; experimental procedures and group allocation were optimized to minimize the number of animals required; and refined surgical techniques, standardized perioperative care, and appropriate anesthesia were used to minimize pain and distress.

Declaration

The authors confirm that all animal experiments described in this manuscript were carried out in compliance with institutional ethical guidelines for animal research. The welfare of the animals was given the highest priority throughout the study.

Funding

This work was supported by National Natural Science Foundation of China (No. 82273014) and Zhejiang Provincial “Pioneer” and “Leading Goose” R&D Program (Grant No. 2024C03196).

Declaration of competing interest

The authors 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 are the Supplementary data to this article:

Data availability

Data will be made available on request.