Molybdenum-bridged endo-exogenous antioxidant synergy reverses acute kidney injury via mitochondrial homeostasis reconstruction

Abstract

Acute kidney injury (AKI) progression is driven by mitochondrial redox collapse in proximal tubular epithelial cells (PTECs), where reactive oxygen species (ROS) surge and molybdenum (Mo) metabolic dysregulation create an “oxidative storm-defense collapse” cycle. Conventional antioxidant therapies fail to halt AKI chronicity due to their inability to restore Mo-dependent detoxification enzymes (e.g., Mo-containing Amidoxime Reducing Component, mARC). To address this dual pathology, we developed N-acetylcysteine (NAC)-modified molybdenum disulfide quantum dots (NMDs) that implement an endo-exogenous antioxidant collaborative strategy, synergizing exogenous ROS elimination with endogenous Mo enzyme restoration. NMDs achieve triple-tiered targeting: 1) Organ-selective accumulation leveraging NMDs' hydrophilicity and ultrasmall size; 2) Cell-specific internalization through Organic Anion Transporter 1 (OAT1)-mediated active uptake into PTECs; 3) Mitochondrial precision delivery guided by NAC's intrinsic mitochondrial affinity. Within pathological microenvironments, NMDs exhibit multidimensional therapeutic superiority: exposed Mo(Ⅳ) directly quenches mitochondrial ROS via electron transfer (external clearance), while released Mo ions reactivate mARC and NAC supplies glutathione precursors, synergistically rebuilding endogenous antioxidant defenses (internal reinforcement). In vivo validation demonstrated NMDs’ superior therapeutic efficacy, outperforming clinical antioxidant NAC. This work pioneers a “scavenging-fortification” strategy through Mo-centric metabolic regulation and nanotechnology integration, validating Mo-based materials' therapeutic potential and establishing a paradigm for mitochondrial-targeted AKI treatment.

Graphical abstract

Highlights

• •Through constructing N-acetylcysteine-functionalized molybdenum disulfide quantum dots (NMDs), this nanoplatform achieves hierarchical mitochondria-targeted delivery, synergistically restoring redox homeostasis and offering a novel therapeutic paradigm for mitochondrial injury-related diseases like acute kidney injury.
• •NMDs achieve triple-tiered targeting (kidney .→ proximal tubular cells → mitochondria) via OAT1-mediated uptake and mitochondrial protein affinity, overcoming renal delivery barriers.
• •NMDs synergistically eliminate mitochondrial reactive oxygen species (ROS) via Mo(IV)/Mo(VI) valence transition and restore endogenous defenses by supplying Mo cofactors for mARC reactivation and NAC-derived GSH precursors.
• •NMDs (2 mg/kg) reverse AKI in rhabdomyolysis models, reducing serum creatinine and tubular injury scores, outperforming clinical antioxidants NAC.
• •By preserving mitochondrial integrity, NMDs inhibit mtDNA leakage, suppressing cGAS-STING-driven inflammation and intrinsic apoptosis.

1. Introduction

Acute kidney injury (AKI), a clinical syndrome characterized by abrupt renal dysfunction, poses a critical threat to over 17 million lives globally [1]. With an annual incidence of approximately 13 million cases (85 % in developing countries) and a mortality rate of 23 %, AKI represents a pressing unmet medical need [2,3]. Current therapies, including renal dialysis and fluid resuscitation, remain palliative, failing to halt disease progression [4,5]. This unmet clinical need underscores the critical demand for novel therapeutic agents in AKI management.

Recent advances have unmasked mitochondria in proximal tubular epithelial cells (PTECs) as the epicenter of AKI pathogenesis [6]. PTECs, the workhorses of renal reabsorption, exhibit high mitochondrial density among various cell types—a metabolic advantage turned liability under AKI [7,8]. Pathological electron leakage from mitochondrial respiratory chains surges significantly, triggering reactive oxygen species (ROS) avalanches [9]. These highly reactive species diffuse with millisecond-scale kinetics, selectively targeting electrophilic biomolecules such as genomic DNA and unsaturated lipid moieties [10]. Subsequent bioenergetic collapse unleashes cascading insults: apoptotic signals from damaged mitochondria, cGAS-STING-driven sterile inflammation, and irreversible tubular destruction [[11], [12], [13]]. This subcellular-to-organ domino effect underpins AKI progression. Cellular defense systems, however, are not passive bystanders. The intracellular antioxidant defense comprises two major systems: First, the enzyme antioxidant system, among which metal cofactor-dependent enzymes—particularly the Molybdenum (Mo)-containing Amidoxime Reducing Component (mARC) predominantly enriched in kidneys—occupy a unique position in renal antioxidant defense. As a mitochondria-specifically distributed detoxification enzyme, mARC maintains genomic stability by catalyzing the reduction of toxic hydroxylamine substances, thereby eliminating oxidative stress-induced DNA-damaging products [[14], [15], [16]]. Second, the glutathione (GSH)-centric antioxidant network avoids lipid peroxidation through direct neutralization of free radicals and glutathione peroxidase 4 (GPx4)-dependent repair [17,18]. However, during AKI progression, the antioxidant systems in PTECs undergo systemic collapse: the excessively generated ROS far exceed the cellular clearance capacity, while timely replenishment and repair mechanisms are absent [19,20]. Our Inductively Coupled Plasma Mass Spectrometry (ICP-MS) analysis of kidneys from AKI-model mice demonstrated significantly reduced Mo content in renal tissues of AKI. This depletion may originate from renal tubular injury, oxidative stress-induced degradation of Mo cofactors, and compromised mitochondrial membrane integrity. The deficiency of Mo leads to mARC inactivation, while GSH exhaustion paralyzes the lipid repair system, ultimately forming a pathological closed-loop where oxidative damage and defense failure mutually exacerbate each other.

When confronting this intricate pathological mechanism, the limitations of conventional antioxidant strategies—characterized by “external remediation without internal reinforcement”—have become increasingly evident. This “one-legged walking” treatment mode suffers from dual mechanistic deficiencies: on one hand, exogenous antioxidants have limited targeting capabilities for pathological focus, resulting in suboptimal ROS neutralization efficiency [21,22]. Particularly in complex organs like the kidney, achieving hierarchical deep penetration represents a critical challenge in drug design. On the other hand, irreversible Mo depletion-induced mARC inactivation and GSH system collapse persistently drive oxidative damage toward chronic progression. To overcome this therapeutic impasse, an “endo-exogenous antioxidant synergy” must be established—one that simultaneously achieves precision ROS eradication (“exogenous eradication”) and reconstructs impaired endogenous defense mechanisms (“endogenous consolidation”).

Fortunately, as the first renal epithelial cell population to contact therapeutic agents, PTECs express abundant membrane transporters that orchestrate drug uptake, notably the Organic Anion Transporters (OATs)—key carriers mediating renal transport of drugs and metabolites [[23], [24], [25]]. Emerging evidence identifies N-acetylated compounds as potential OAT1 substrates, including established endogenous examples such as mercapturates (N-acetyl-L-cysteine S-conjugates) [26,27]. Herein, we innovatively developed N-acetylcysteine (NAC)-functionalized molybdenum disulfide quantum dots (NMDs), whose design intricately integrates AKI pathobiology with the unique physicochemical properties of molybdenum sulfides. As the only essential 4d transition metal in the human body, molybdenum not only serves as the catalytic core of key redox enzymes such as mARC, but also enables ROS scavenging through valence transitions. NMDs were synthesized through a bottom-up controlled strategy using ammonium molybdate and thiourea as precursors. The ultra-small size and hydrophilic surface of NMDs enable preferential renal accumulation while avoiding hepatic and splenic sequestration. Furthermore, NAC modification not only stabilizes the quantum dot architecture but also enables cell-specific recognition—the N-acetyl group within its structure serves as a ligand for OAT1, driving receptor-mediated cellular uptake in PTECs. Following internalization, NAC's high mitochondrial affinity directs NMDs to achieve precise localization at injury cores. Mechanistically, NMDs reverse pathology through multidimensional synergy: (a) Mo(Ⅳ) efficiently eliminates mitochondrial ROS via electron transfer and valence transitions, directly blocking oxidative stress cascades; (b) The supplied Mo replenishes mARC active centers, while released NAC provides precursors for GSH synthesis, collectively rebuilding endogenous antioxidant defenses; (c) Additionally, mitochondrial homeostasis restoration suppresses cytochrome c (Cyt c)-mediated apoptotic signaling and cGAS-STING-driven inflammatory storms. This integrated “targeted delivery-efficient scavenging-holistic repair” design bridges the physicochemical advantages of nanomaterials with pathological demands, overcoming the dual limitations of nonspecific biodistribution and single-mechanism action of traditional antioxidants while comprehensively repairing AKI's pathological epicenter—the mitochondrion (see Scheme 1). In summary, our work not only validates the translational potential of Mo-based nanomaterials in AKI treatment but also pioneers a paradigm shift in mitochondrial-targeted nanomedicine—transitioning from mere antioxidant delivery to functional reconstruction of endogenous defense systems.

Scheme 1.

Molybdenum-Bridged Endo-Exogenous Antioxidant Synergy Reverses Acute Kidney Injury via Mitochondrial Homeostasis Reconstruction. (A) Synthesis of NMDs. (B) The hierarchical targeting capability of NMDs: NMDs quickly entering the renal circulation; accumulating in PTECs through Organic Anion Transporter 1 (OAT1); and co-locating with mitochondria. (C) NMDs effectively protect kidney through dual-action redox regulation, reducing Deoxyribonucleic Acid (DNA) and lipid oxidative damage, inhibiting intrinsic apoptosis, and preventing inflammation.

2. Results

2.1. The synthesis and characterization of NMDs

NMDs were synthesized via a bottom-up strategy using ammonium molybdate and thiourea as precursors, with NAC as a capping agent. Control MoS₂ nanoparticles were prepared under identical conditions without NAC. Transmission Electron Microscopy (TEM) characterization revealed distinct morphological differences: NMDs exhibited monodisperse spherical morphology with an average diameter of 4∼5 nm (Fig. 1A–B), whereas MoS₂ formed irregular sheet-like aggregates averaging 145 nm (Fig. S1). Thermogravimetric Analysis (TGA) further demonstrated that N-acetylcysteine accounted for approximately 22.52 % of the total mass of NMDs (Fig. S2), confirming successful surface modification. Zeta potential measurements confirmed superior colloidal stability for NMDs (−34.2 mV) compared to MoS₂ (−22.6 mV), with the latter demonstrating significantly higher aggregation propensity (Fig. 1C, Fig. S3). X-ray Photoelectron Spectroscopy (XPS) analysis confirmed identical elemental compositions (C1s, S2p, Mo3d, Mo3p, N1s, O1s) in both nanomaterials (Fig. 1D–F, Fig. S4). Energy Dispersive X-ray Spectroscopy (EDS) spectra further verified the elemental composition of NMDs (Fig. S5). High-resolution Mo3d spectra revealed that molybdenum in ammonium molybdate underwent reduction mediated by S²⁻ generated during thiourea decomposition, resulting in a high proportion of Mo(IV) in NMDs (46.8 %) (Fig. 1E, Fig. S6). Ultraviolet–visible (UV–Vis) spectra of NMDs displayed characteristic absorption bands for ammonium molybdate (∼215 nm), thiourea (∼235 nm), and NAC (∼300 nm), validating precursor integration (Fig. 1G). Moreover, NMDs exhibited significant photoluminescence with excitation/emission maxima at 400/485 nm (Fig. S7), manifesting the quantum confinement effects of NMDs. Fourier Transform Infrared (FT-IR) analysis further distinguished the materials: NMDs exhibited vibrational modes corresponding to NAC (C=O at 1636 cm⁻¹, C-H at 1131 cm⁻¹) and S-Mo-S layers (894 cm⁻¹), while MoS₂ primarily showed S-Mo-S chain vibrations (833 cm⁻¹) with residual solvent signals (Fig. 1H–I). The above results demonstrate that both NMDs and MoS₂ possess the S-Mo-S layered structure, with NMDs successfully modified by NAC. Simultaneously, NMDs exhibit smaller size and lower surface potential compared to MoS₂, granting them greater advantages in circulatory stability and renal accumulation.

Fig. 1.

The Synthesis and Characterization of NMDs. (A) TEM image of NMDs. (B) Size distribution of NMDs. (C) Zeta potential determination of NMDs. (D) XPS spectrum of NMDs. (E) XPS Mo3d spectrum of NMDs. (F) XPS S2p spectrum of NMDs. (G) UV–Vis spectrum of NMDs, ammonium molybdate, thiourea and NAC. (H) FT-IR spectrum of NMDs. (I) FT-IR spectrum of MoS₂. (J) UV–Vis spectrum of O₂^·- reacting with different concentrations of NMDs. (K) XPS Mo3d spectrum of NMDs after reacting with H₂O₂. (L) UV–Vis spectrum of ·OH reacting with different concentrations of NMDs. (M) UV–Vis spectrum of NMDs reacting with different concentrations of H₂O₂. (N) UV–Vis spectrum of NMDs reacting with different concentrations of ONOO⁻. (O) UV–Vis spectrum of dialysate after NMDs reacting with H₂O₂.

Molybdenum can exist in various oxidation states under physiological conditions and couple oxide/proton transfer with electron transfer, making it one of the preferred metals participating in chemical reactions in the human body. Therefore, molybdenum enzymes with molybdenum as their catalytic center are key redox enzymes in humans, mediating multiple oxygen transfer and redox reactions. Building upon this biochemical foundation, molybdenum sulfide materials, benefiting from their unique two-dimensional layered structure (S-Mo-S) and abundant active sites, possess exceptional antioxidant potential. In environments with excessive ROS, Mo(IV) in NMDs is readily oxidized to Mo(VI), neutralizing toxic ROS via electron transfer and stabilization. Using nitroblue tetrazolium (NBT) assays, we demonstrated NMDs' dose-dependent superoxide anion (O₂^·-) scavenging efficacy (Fig. 1J). XPS analysis revealed a significant decrease in surface Mo(IV) content and corresponding increase in Mo(VI) following ROS exposure, confirming that the antioxidant activity originates from the redox cycling of low-valent molybdenum species (Fig. 1K, Fig. S8). Parallel evaluations using tetramethylbenzidine (TMB) oxidation assays and UV–Vis absorption spectroscopy further validated NMDs' broad-spectrum ROS scavenging capacity against ·OH, H₂O₂, and ONOO⁻ (Fig. 1L–N). While large size MoS₂ nanoparticles also exhibited ROS neutralization, their scavenging efficiencies for ROS at equivalent concentrations were inferior to NMDs (Fig. S9), likely due to MoS₂'s lower specific surface area. To further elucidate the structural evolution of NMDs under oxidative conditions, Raman spectroscopy analyses were performed after treatment with H₂O₂, horseradish peroxidase (HRP), and human myeloperoxidase (MPO). As shown in Fig. S10, the characteristic Raman modes of MoS₂ (E¹_2g at ∼385 cm⁻¹ and A_1g at ∼405 cm⁻¹) were significantly attenuated or even disappeared, along with a marked reduction in the S-related vibrational peaks J1 (∼132 cm⁻¹), J2 (∼219 cm⁻¹), and J3 (∼308 cm⁻¹). A broad new band appeared around 250–300 cm⁻¹, indicative of Mo–O–Mo deformation vibrations, suggesting the formation of MoO₃. XPS spectra further confirmed the oxidation of Mo(IV) to Mo(VI) (Fig. S10). In parallel, the S 2p spectra revealed the oxidation of S²⁻ to SO₄²⁻. These findings collectively indicate that the redox transformation of MoS₂ underlies both the ROS-scavenging functionality and the biodegradability of NMDs under oxidative stress. In addition, NAC serves dual roles: as a stabilizing capping agent and a precursor for GSH—the important component of endogenous antioxidant defense. Dialysis experiments under hydrogen peroxide (H₂O₂) stimulation demonstrated time- and concentration-dependent NAC release from NMDs, evidenced by characteristic UV absorption peaks in external solutions (Fig. 1O). This stimuli-responsive release mechanism enables NMDs to dynamically replenish intracellular GSH reserves, synergistically enhancing cellular antioxidant defenses through both direct ROS scavenging and indirect antioxidant system restoration.

In summary, this section details the strategic synthesis and comprehensive characterization of NMDs. Structural analyses confirm that NAC functionalization confers NMDs with ultrasmall size and superior colloidal stability, showing better potential in both renal targeting than uncontrolled MoS₂ aggregates. The synergistic interplay between abundant surface-exposed Mo(IV) ions and stimuli-responsive NAC release equips NMDs with multifunctional redox homeostasis modulation capabilities, addressing both acute ROS neutralization and chronic antioxidant defense restoration.

2.2. NMDs achieved sequential targeting of kidney, PTECs, and mitochondria

Precision-targeted delivery to pathological sites stands as a cardinal objective in therapeutic drug design, particularly for achieving deep tissue penetration in complex organs like kidneys. First, the kidney's inherent hydrophilicity necessitates therapeutic drugs owns high aqueous solubility [28,29]. Second, PTECs represent the primary sites of injury and key therapeutic targets in AKI [30,31]: (1) as frontline interfaces for nephrotoxic agents from plasma/glomerular filtrate, and (2) as mitochondrial-rich cells generating excessive ROS. Notably, the high expression of OAT1 on PTECs provides great convenience for targeted drug design. Building on the above insights, we thus hypothesized that NAC-modified NMDs, leveraging their high hydrophilicity and N-acetyl motif, could undergo OAT1-mediated renal targeting. To validate this, we employed a glycerol-induced rhabdomyolysis AKI (RM-AKI) mouse model and explored the distribution of NMDs in the main organs and cell groups.

First, TEM was employed to observe tissue damage around the glomerular filtration barrier (GFB) and drug distribution in different treatment groups. As shown in Fig. 2A, the Control group exhibited normal GFB structure with orderly arranged podocytes. The AKI group displayed significant GFB damage characterized by irregular podocyte morphology and partial edema adhesion. In contrast, NMDs-treated mice showed restored GFB architecture. NMDs could partially travers the GFB into Bowman's capsule lumen due to their ultrasmall size, whereas MoS₂ was trapped inside the GFB due to size constraints. Pharmacokinetic modeling quantified systemic exposure characteristics: NMDs exhibited a plasma half-life (t_1/2) of 5.412 h in healthy mice and 5.704 h in AKI mice, and is mainly excreted in urine through kidneys. The marginally prolonged retention in AKI mice may align with enhanced PTECs utilization for antioxidant system repair (Fig. S11). Subsequently, Fluorescein isothiocyanate (FITC) labeling was applied to track their biodistribution in main organs (Fig. S12). Stereofluorescence imaging revealed significant renal accumulation of NMDs in both healthy and AKI mice, with less hepatic and pulmonary deposition, and least distribution in cardiac and splenic tissues (Fig. 2B). Time-dependent fluorescence intensity analysis demonstrated detectable renal fluorescence within 1 h post-intravenous administration, peaking at 9 h, followed by gradual renal clearance (Fig. 2C–D, Figure S13-14). This pharmacokinetic profile satisfies two critical requirements for AKI therapeutics: rapid action and absence of long-term metabolic/excretion burden. ICP-MS analysis revealed renal Mo concentrations reaching ∼800 μg/kg in AKI mice at 9 h, decreasing to <200 μg/kg by 24 h, consistent with fluorescence tracking trends (Fig. 2E). To directly visualize in vivo biodistribution, Cy7-labeled NMDs were further used for whole-body imaging. Strong renal signals were observed post-injection, with maximal kidney accumulation at 9 h, closely matching ex vivo imaging results (Fig. S15).

Fig. 2.

NMDs Achieved Sequential Targeting of Kidney, PTECs, and Mitochondria. (A) TEM images of GFB in different groups. (B) Representative fluorescence images of different organs in AKI mice 9 h after NMDs administration. (C–D) Representative fluorescence images (C) and corresponding fluorescence intensity (D) of kidneys in AKI mice at various time points (1–72h) after NMDs administration. (E) Content of molybdenum of kidneys 9 h after NMDs administration. (F) Representative immunofluorescence (IF) staining images of FITC-NMDs and FITC-MoS₂ with glomerular (Synaptopodin, purple) and tubular (AQP1, red) markers in mice kidney tissues. (G) Representative IF staining image of FITC-NMDs with OAT1 (red) and CD31 (yellow) in mice kidney tissues. (H) Representative fluorescence images of HK-2 cells incubated with FITC-NMDs and probenecid/NAC co-incubation. (I) Colocalization fluorescence images of FITC-NMDs with different organelles in HK-2 cells. (J–K) Molecular docking results of NAC with TOM20 and VDAC1. (L–M) Molecular docking results of MoS₂ with TOM20 and VDAC1. Data are presented as mean ± SD. One-way ANOVA followed by Student-Newman-Keuls (SNK) test was used for analysis. n = 3.

As previously established, PTECs constitute pivotal targets in AKI pathogenesis and therapeutic intervention. Leveraging the high expression of membrane transporters on PTEC plasmalemma, FITC-NMDs exhibited predominant distribution in renal tubules [Aquaporin-1 (AQP1)-positive] rather than glomeruli (Synaptopodin-positive) (Fig. 2F). Notably, NAC-modified NMDs demonstrated significantly higher fluorescence intensity in tubular regions compared to MoS₂, indicative of enhanced PTEC uptake. Given OAT1’s established affinity for N-acetyl moieties, we further investigated the NMDs internalization mechanism. Immunofluorescence analysis revealed strong colocalization between FITC-NMDs and OAT1 at the basolateral membrane of PTECs, accompanied by observable vascular-to-tubular diffusion patterns (CD31-positive vasculature) (Fig. 2G). The reduced fluorescence intensity detected at brush-border membranes is potentially attributable to the selectivity of GFB against anionic compound. Cellular-level investigations further substantiated OAT1’s critical role in NMDs uptake. Probenecid-mediated OAT inhibition significantly reduced FITC-NMDs internalization in human renal tubular epithelial cells (HK-2) (Fig. 2H, Fig. S16). Competitive inhibition via NAC co-incubation similarly attenuated uptake efficiency. These collective findings demonstrate that NAC functionalization confers active PTEC-targeting capability to NMDs, with OAT1-mediated transmembrane transport constituting a primary cellular internalization mechanism.

Following successful internalization into PTECs, the optimal scenario for NMDs is precise targeting of mitochondria—the primary source of ROS. Co-localization experiments employing FITC-labeled NMDs and organelle-specific probes in HK-2 cells revealed differential subcellular distribution patterns. Fluorescent microscopy demonstrated spatial overlap between FITC-NMDs and mitochondria, nucleus, endoplasmic reticulum (ER), and Golgi apparatus in different degrees (Fig. 2I). Quantitative analysis via Pearson's correlation coefficients identified mitochondria as the predominant localization site (R = 0.91), suggesting preferential mitochondrial accumulation that facilitates targeted ROS neutralization. Notably, NAC co-incubation significantly reduced mitochondrial co-localization (R = 0.36), indicating NAC's essential role in mitochondrial targeting specificity, whereas other organelle distributions remained unaffected (Fig. S17). Molecular docking simulations were conducted to elucidate the mitochondrial targeting mechanism, focusing on interactions between NAC (NMDs' surface ligand) and key mitochondrial outer membrane proteins: Translocase of Outer Mitochondrial Membrane 20 (TOM20), Translocase of Outer Mitochondrial Membrane 34 (TOM34), Voltage-Dependent Anion Channel 1 (VDAC1), and VDAC2. Stable binding was observed between NAC and TOM20 (−3.5 kcal/mol), mediated through three distinct interactions: two hydrophobic contacts (3.8 Å) with LEU54, a salt bridge (5.1 Å) with HIS926, and a carbon-hydrogen bond (2.1 Å) with GLU50 (Fig. 2J). Comparative analysis revealed strong binding affinities to other mitochondrial proteins: TOM34 (−3.9 kcal/mol), VDAC1 (−3.8 kcal/mol), and VDAC2 (−3.7 kcal/mol) (Fig. 2J–K, Figure S18). These computational findings demonstrate spontaneous NAC-protein interactions that mechanistically explain NMDs’ mitochondrial precision. Control experiments with MoS₂ exhibited significantly weaker binding affinities (TOM20: −1.3 kcal/mol; TOM34: −0.9 kcal/mol; VDAC1: −2.6 kcal/mol; VDAC2: −3.6 kcal/mol) (Fig. 2L–M, Fig. S18), conclusively establishing NAC-driven targeting specificity rather than inherent MoS₂ properties.

Collectively, NMDs achieve hierarchical targeting by sequentially reaching renal tissue, PTECs, and mitochondria, precisely aligning with kidney anatomy and AKI pathology to ensure therapeutic precision and efficiency.

2.3. NMDs effectively alleviated renal injury during RM-AKI

The sequential and precise targeting capacity of NMDs establishes the foundation for their therapeutic efficacy at AKI lesion sites (Fig. 3A). To validate this therapeutic potential, we employed a rhabdomyolysis-induced AKI mouse model. During rhabdomyolysis, extensive muscle tissue breakdown leads to massive myoglobin release, which freely traverses the GFB to induce oxidative stress in renal tubules, alter renal hemodynamics, and cause tubular obstruction through cast formation, collectively resulting in acute renal dysfunction [32,33]. Serum creatinine (Cre) and blood urea nitrogen (Bun) levels, as biomarkers of glomerular filtration rate, exhibited significant elevation in AKI mice compared to Control mice, confirming successful model establishment. NMDs administration induced dose-dependent restoration of Cre and Bun levels, with high-dose treatment (2 mg/kg) achieving near-normal values. In contrast, equivalent doses of MoS₂ or NAC failed to improve renal function (Figure S19, Fig. 3B). Histopathological evaluation via H&E staining revealed severe tubular injury in AKI mice, including tubular dilatation, epithelial necrosis, basement membrane detachment, and cast formation (pathological score: 3.8), compared to intact renal architecture in Control mice. NMDs treatment significantly restored renal morphology, outperforming ineffective MoS₂/NAC interventions (Fig. 3C–D). Body weight monitoring further corroborated these findings: Control mice maintained steady weight gain (2–4 g), while AKI mice exhibited progressive weight loss reversed by NMDs but not MoS₂/NAC (Fig. 3E). Renal biomarker analysis further substantiated NMDs’ efficacy. Neutrophil gelatinase-associated lipocalin (NGAL), predominantly secreted by injured tubular epithelia, and kidney injury molecule-1 (KIM-1), undetectable in healthy kidneys but markedly upregulated post-injury, both showed significant elevation in AKI mice. NMDs treatment normalized these biomarkers, whereas MoS₂ and NAC remained ineffective (Fig. 3F). To further verify the therapeutic generalizability, NMDs also significantly alleviated renal injury in a cisplatin-induced AKI model. Both functional biomarkers and histological assessment confirmed notable renoprotection (Fig. S20).

Fig. 3.

NMDs effectively alleviated renal injury during RM-AKI. (A) Schematic illustration of NMDS entering nephron and actively targeting PTECs. (B) Levels of serum Cre and Bun in different treatment groups. (C–D) Kidney injury scores and representative H&E-stained images of kidney tissues from different treatment groups. Arrows indicate damaged PTECs, and asterisks indicate casts. (E) Body weight variation of different treatment groups. (F) Levels of NGAL and KIM-1 in kidney tissues of different treatment groups. Data are presented as mean ± SD. One-way ANOVA followed by SNK test was used for analysis. n = 6, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; ns, not significant (P > 0.05).

Collectively, these findings validate NMDs’ superior renoprotective effects over MoS₂ and NAC, attributable to their precise renal targeting, sustained retention, and enhanced antioxidant capacity.

2.4. NMDs restored redox balance through eliminating mtROS and strengthening endogenous antioxidant system

Integrating pharmacological design, in vitro characterization and biodistribution analysis, we propose that NMDs’ therapeutic efficacy in AKI mice principally stems from their capacity to restore renal redox homeostasis through dual antioxidant mechanisms. The abundant reduced Mo(Ⅳ) in NMDs, combined with their mitochondrial-targeting precision, enable efficient neutralization of mitochondrial ROS (mtROS), thereby preventing its diffusion and damage biomacromolecules. Concurrently, NAC surface modification and intrinsic molybdenum content provide essential precursors for reconstructing intracellular antioxidant systems. This multilevel antioxidant mechanism not only effectively reduces renal ROS burden but also establishes a robust foundation for tissue repair and functional recovery.

To systematically investigate NMDs’ regulatory effects on redox imbalance, we conducted comprehensive cellular-level experiments. An oxidative damage model was established in HK-2 cells using H₂O₂ stimulation. Cell viability assays demonstrated significant H₂O₂-induced cytotoxicity, which was dose-dependently reversed by NMDs treatment. At 40 μg/mL, NMDs restored viability to near-control levels. In contrast, NAC showed no efficacy at tested concentrations, consistent with literature-reported requirements for higher NAC doses (≥800 μg/mL). MoS₂ exhibited partial recovery, though significantly inferior to NMDs, likely due to reduced cellular uptake and limited active site accessibility (Fig. S21). Intracellular ROS quantification using DCFH-DA probe also revealed effective ROS scavenging capability of NMDs. MoS₂ showed modest ROS suppression, while NAC failed to reduce fluorescence intensity (Fig. 4A–B). Analysis of mitochondrial superoxide via MitoSOX staining demonstrated elevated mtROS levels in H₂O₂-injured cells. NMDs treatment effectively normalized mtROS levels, whereas MoS₂ and NAC exhibited minimal mtROS scavenging capacity (Fig. 4C–D). These results highlight NMDs’ unique dual advantage: mitochondrial-targeted delivery coupled with potent ROS neutralizing activity, synergistically addressing oxidative injury at its source.

Fig. 4.

NMDs restored redox balance. (A–B) Representative DCFH-DA fluorescence images (A) and corresponding fluorescence intensity (B) of HK-2 cells in different treatment groups. (C–D) Mito-SOX fluorescence staining images (C) and corresponding fluorescence intensity (D) of HK-2 cells from different treatment groups. (E) TEM images of mitochondria in PTECs from different treatment groups of mice. (F–G) Representative DHE fluorescence images (F) and corresponding fluorescence intensity (G) of kidney tissues in different treatment groups. (H–I) Levels of oxidative products MDA (H) and TBARS (I) in kidney tissues from different treatment groups. (J) IHC results of 8-OHdG in mouse kidneys from different treatment groups. (K–L) GSH/GSSG (K) levels and GPx levels (L) in mouse kidneys from different treatment groups. (M) Levels of molybdenum in mouse kidneys from different treatment groups. (N) Representative IF staining image of mARC1 (green), glomerular (synaptopodin, yellow) and tubular (AQP1, red) markers in mice kidney tissues from different treatment groups. (O) Fluorescence intensity of mARC in different treatment groups. Data are presented as mean ± SD. One-way ANOVA followed by SNK test was used for analysis. n = 3, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; ns, not significant (P > 0.05).

Subsequent in vivo investigations comprehensively validated NMDs’ redox regulatory capacity, encompassing mitochondrial protection, ROS scavenging, and antioxidant system reinforcement. Firstly, TEM images revealed intact mitochondrial ultrastructure in Control mice, characterized by rod-shaped organelles with smooth outer membranes and densely packed, regularly arranged cristae (as indicated by white arrows in Fig. 4E). In contrast, AKI mice exhibited prominent mitochondrial damage, including outer membrane rupture, swelling, cristae disruption, and fragmentation (red arrows, Fig. 4E). Notably, these pathological changes were substantially mitigated in the NMDs-treated group, where most mitochondria retained normal morphology or showed only mild alterations (yellow arrows, Fig. 4E). In contrast, MoS₂ and NAC failed to restore mitochondrial integrity. Moreover, dihydroethidium (DHE) staining demonstrated intense renal ROS fluorescence in AKI mice, normalized by NMDs treatment but unaffected by MoS₂/NAC (Fig. 4F–G), confirming NMDs’ superior ROS neutralization. Given the susceptibility of lipids and DNA to oxidative modification [10,34], lipid peroxidation markers (malondialdehyde [MDA], thiobarbituric acid-reactive substances [TBARS]) and 8-hydroxy-2’-deoxyguanosine (8-OHdG) were detected. AKI mice exhibited elevated MDA/TBARS levels and pronounced 8-OHdG staining, indicating severe oxidative damage. NMDs treatment restored these parameters to near-normal levels, while MoS₂/NAC showed negligible efficacy (Fig. 4H–J, Figure S22). Notably, the body possesses intrinsic antioxidant defenses comprising molecular scavengers and enzymatic systems that detoxify ROS under cellular stress [35]. GSH, a pivotal intracellular antioxidant, mitigates oxidative damage either through direct ROS neutralization or as an essential cofactor for GPx4 [36,37]. And mARC, a molybdenum-dependent enzyme, catalyzes the reduction of N-hydroxylated compounds (e.g., nucleobases and nucleotides), thereby maintaining genomic stability [38]. However, during AKI progression, these systems become overwhelmed: GSH is rapidly oxidized to GSSG, while GPx4 activities and mARC levels decline due to mitochondrial dysfunction and ROS-mediated enzyme inactivation (Fig. 4K–L). Notably, NMDs counteracted these deficits through dual biochemical reconstitution: NAC provides GSH precursor support, while molybdenum replenishes the metal cofactor essential for mARC activity. As a result, after NMDs treatment, the GSH/GSSG ratio and GPx4 activity in renal tissue were significantly restored, indicating that the GSH-centered antioxidant system was effectively repaired. Furthermore, ICP-MS confirmed AKI-induced renal molybdenum depletion, which NMDs effectively reversed (Fig. 4M), paralleled by restored mARC expression (Fig. 4N–O). In contrast, neither MoS₂ nor NAC monotherapy restored antioxidant capacity, likely due to inadequate targeting, insufficient ROS neutralization, and mechanistic singularity.

Collectively, these findings establish that NMDs synergistically restore redox homeostasis through: (1) direct ROS scavenging, (2) GSH system rejuvenation, and (3) molybdenum-dependent enzyme reactivation—a multifaceted therapeutic strategy surpassing conventional antioxidant approaches in AKI management.

2.5. NMDs effectively inhibited mitochondrial apoptosis

Emerging research highlights the pivotal role of mitochondrial-mediated crosstalk between oxidative stress and apoptosis in AKI pathogenesis [12]. During AKI progression, excessive mtROS generation not only directly damages mitochondrial structure and function but also activates downstream apoptotic pathways in PTECs, exacerbating tubular disintegration and renal dysfunction [39]. We hypothesized that NMDs’ mitochondrial enrichment and protective capacity could mitigate apoptosis initiation. Mitochondrial membrane potential (ΔΨm), a vital bioenergetic parameter for cellular viability, was evaluated using JC-1 staining (Fig. 5A–C). H₂O₂ stimulation caused ΔΨm collapse in HK-2 cells, evidenced by increased green fluorescence (monomeric JC-1) and diminished red fluorescence (J-aggregates). NMDs treatment effectively reversed this depolarization through mtROS scavenging, restoring physiological ΔΨm. MoS₂ partially mitigated depolarization, while NAC showed no protective effect. ATP quantification confirmed mitochondrial functional preservation, with NMDs restoring energy production to near-physiological levels, outperforming MoS₂ and NAC (Fig. 5D). Apoptosis progression was analyzed via Annexin V-FITC/PI staining, revealing H₂O₂-induced shifts from viable to apoptotic cell populations (90 % surviving cells in Control group vs 55 % surviving cells in H₂O₂ group). NMDs treatment significantly reduced both early and late apoptosis, restoring viable cell proportions (80 % surviving cells), whereas MoS₂ showed limited efficacy against late apoptosis and NAC provided no protection at the same doses (Fig. 5E–F). Western blot analysis was further conducted to examine apoptosis-related protein expression across treatment groups. The Bcl-2 protein family, comprising pro-apoptotic Bax and anti-apoptotic Bcl-2, serves as the primary regulator of mitochondrial apoptotic factor release. Cyt c, an essential component of the mitochondrial electron transport chain, translocates to the cytoplasm (Cyto-Cyt c) upon apoptotic stimulation, where it binds apoptotic protease-activating factor-1 to form the apoptosome, subsequently activating caspase cascades. Caspase-3, the key executioner protease, cleaves critical intracellular substrates to execute apoptosis. WB results revealed elevated Cyto-Cyt c levels in H₂O₂-treated HK-2 cells versus Control cells, accompanied by increased Bax expression, decreased Bcl-2 levels, and significant accumulation of cleaved caspase-3—collectively indicating mitochondrial apoptosis activation. NMDs and MoS₂ treatments modulated these apoptotic markers: reduced Bax, Cyto-Cyt c, and cleaved caspase-3 alongside restored Bcl-2 expression. Notably, NMDs demonstrated superior anti-apoptotic efficacy compared to MoS₂, while NAC failed to suppress mitochondrial apoptosis in HK-2 cells (Fig. 5G–L, Figure S23).

Fig. 5.

NMDs effectively inhibited mitochondrial apoptosis. (A–B) JC-1 cytometry results (A) and corresponding statistical data (B) of HK-2 cells from different treatment groups. (C) JC-1 fluorescence staining images of HK-2 cells from different treatment groups. (D) ATP production of HK-2 cells from different treatment groups. (E–F) Annexin-V/PI flow cytometry results (E) and corresponding apoptosis ratio (F) in HK-2 cells from different treatment groups. (G–K) Western Blotting analysis (G) of apoptosis-related protein expression levels in HK-2 cells from different treatment groups and grayscale analysis of cyto-Cyt c(H), Bax (I), Bcl-2 (J), and cleaved-Caspase 3 (K). (L) Schematic illustration of NMDs inhibiting intrinsic apoptosis. (M–N) TUNEL fluorescence staining results (M) and quantification of TUNEL-positive cells (N) in mouse kidneys from different treatment groups. Data are presented as mean ± SD. One-way ANOVA followed by SNK test was used for analysis. n = 3, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; ns, not significant (P > 0.05).

Moreover, in vivo investigations in RM-AKI mice confirmed NMDs’ inhibition of mitochondrial apoptosis as well. Western blot analysis of renal tissue homogenates revealed elevated levels of pro-apoptotic Bax, Cyto-Cyt c, and cleaved caspase-3 alongside reduced Bcl-2 and procaspase-3 in AKI mice, indicating the activation of apoptotic pathway. NMDs treatment restored these protein levels toward physiological ranges, whereas MoS₂ and NAC interventions showed no significant modulation compared to AKI group, underscoring their inability to mitigate mitochondrial damage due to deficient renal and mitochondrial targeting (Fig. S24). Terminal-deoxynucleoitidyl Transferase Mediated Nick End Labeling (TUNEL) staining demonstrated abundant apoptotic cells in AKI renal tubules, which NMDs treatment markedly reduced. MoS₂ and NAC failed to diminish TUNEL-positive signals, further validating NMDs’ mitochondrial protection and apoptosis suppression (Fig. 5M–N).

Collectively, these findings establish mitochondria as central executors of apoptosis through regulated factor release. By combining potent antioxidant activity with mitochondrial specificity, NMDs intercept AKI progression through scavenging mtROS, preserving mitochondrial structure, and blocking apoptotic cascade, thereby rescuing critical renal cells.

2.6. NMDs inhibited mtDNA leakage-induced cGAS-STING pathway activation

Mitochondrial DNA (mtDNA), being proximal to ROS generation sites near the inner mitochondrial membrane and lacking histone-mediated protection, exhibits heightened vulnerability to oxidative damage [40]. ROS-induced mitochondrial damage facilitates mtDNA leakage into the cytosol via permeability transition pores or mitochondrial fragmentation, activating innate immune sensors that trigger inflammatory cascades. The cGAS-STING pathway, a validated mediator of AKI-related inflammation, becomes activated when cGAS binds cytosolic double-stranded DNA (dsDNA), catalyzing cGAMP synthesis. The second messenger cGAMP combines to STING on ER membranes, inducing conformational changes that activate downstream NF-κB and IRF3 signaling, ultimately driving renal inflammation [41]. Fluorescent microscopy images revealed strict nuclear/mitochondrial dsDNA localization in Control HK-2 cells, whereas H₂O₂ treatment caused cytosolic dsDNA dispersion. NMDs treatment effectively prevented the cytoplasmic leakage of dsDNA, whereas MoS₂ and NAC failed to show considerable effect (Fig. 6A). Correspondingly, H₂O₂-injured cells exhibited elevated cGAMP levels due to cytosolic DNA-cGAS interaction, which NMDs treatment normalized to Control levels (Fig. 6B). Western blot analysis demonstrated basal cGAS/STING/IRF3 expression under physiological conditions. H₂O₂ injury upregulated cGAS, STING expression, and increased phosphorylated IRF3/NF-κB (Fig. 6C–D, Figure S25). This correlated with elevated proinflammatory mediators (TNF-α, IL-6, IL-1β, CXCL1, CXCL2, CCL2) in injured cells (Fig. 6E–G, Figure S26). NMDs treatment suppressed both cGAS-STING activation and inflammatory cytokine overproduction, confirming its capacity to intercept pathological innate immune signaling in AKI.

Fig. 6.

NMDs inhibited mtDNA leakage-induced cGAS-STING pathway activation. (A) Fluorescence staining results of dsDNA (green), TOM20 (red), and DAPI (blue) in HK-2 cells from different treatment groups. (B) Levels of 2′-3′cGAMP in kidneys from different treatment groups. (C) WB analysis of cGAS-STING related protein expression in HK-2 cells. (D) Schematic illustration of NMDs inhibiting mtDNA leakage-induced cGAS-STING pathway activation. (E–G) Levels of inflammatory cytokines TNF-α (E), IL-6 (F) and chemokines CXCL1 (G) in HK-2 cells from different treatment groups. (H) Fluorescence staining results of dsDNA (green), TOM20 (red), and DAPI (blue) in kidneys from different treatment groups. (I) IHC results of cGAS and STING in kidneys from different treatment groups. (J–L) Levels of inflammatory cytokines TNF-α (J), IL-6 (K) and chemokines CXCL1 (L) in kidneys from different treatment groups. (M) IHC results of F4/80 and Ly-6G in kidneys from different treatment groups. Data are presented as mean ± SD. One-way ANOVA followed by SNK test was used for analysis. n = 3, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; ns, not significant (P > 0.05).

At the animal level, AKI mice exhibited mitochondrial structural disruption and substantial cytosolic dsDNA leakage, which was effectively prevented by NMDs treatment. In contrast, neither MoS₂ nor NAC demonstrated discernible inhibition of mitochondrial dsDNA release (Fig. 6H). Correspondingly, WB analysis of renal tissue homogenates revealed elevated cGAS-STING pathway protein levels in AKI mice, normalized by NMDs intervention but unaffected by MoS₂/NAC treatments (Fig. S27). Immunohistochemical (IHC) validation further confirmed NMDs’ regulatory effects: AKI kidneys displayed aberrantly high cGAS/STING expression with extensive tubular cytoplasmic positivity, while NMDs treatment substantially reduced positive staining areas, indicating pathway suppression (Fig. 6I). This suppression directly attenuated downstream inflammatory cascades, evidenced by significantly reduced renal inflammatory cytokines/chemokines (Fig. 6J–L, Figure S28) and diminished interstitial infiltration of macrophages (F4/80-positive) and neutrophils (Ly6G-positive) (Fig. 6M). MoS₂/NAC groups maintained pronounced inflammatory activation.

Collectively, these results demonstrate that NMDs preserve mitochondrial integrity through maintaining redox balance, thereby preventing mtDNA leakage, suppressing pathological cGAS-STING activation, and mitigating inflammatory responses—critical processes for halting fibrotic progression in AKI.

2.7. RNA-seq analysis

Building upon NMDs’ demonstrated therapeutic efficacy in animal models, we performed whole-transcriptome sequencing to systematically decode their regulatory effects on AKI pathological networks. Sample correlation heatmaps confirmed high consistency among biological replicates within groups (Fig. S29), while principal component analysis distinctly separated transcriptional profiles of Control, AKI, and NMDs-treated groups (Fig. 7A), validating data robustness for mechanistic exploration. Differential gene expression analysis (criteria: |log₂FC|>1, Q < 0.05) identified 4,709 significant differentially expressed genes (DEGs) between AKI and Control groups (2,331 upregulated; 2,378 downregulated), with 1415 DEGs (814 upregulated; 601 downregulated) modulated by NMDs treatment (Fig. 7B–C, Figure S30). Intersection analysis revealed 1,111 overlapping DEGs, potentially representing core pathways underlying NMDs’ therapeutic actions. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses functionally characterized these DEGs. GO chord diagrams highlighted predominant associations with ROS metabolism, apoptotic regulation, inflammatory responses, mitochondrial function, and proliferative processes (Fig. 7D). KEGG pathway analysis further implicated cell cycle control, oxidative stress, cytoplasmic DNA sensing, neutrophil chemotaxis, and inflammatory cascades (Fig. 7E), collectively verifying NMDs’ coordinated blockade of oxidative damage, cell death, mitochondrial function and sterile inflammation at molecular network levels. Gene expression clustering of core functional modules revealed distinct expression patterns across groups (Fig. 7F–I, Figure S31). AKI mice exhibited marked transcriptional divergence, while Control and NMDs-treated groups showed closer expression profiles, particularly in four functional categories critical to therapeutic efficacy. This multifaceted transcriptional reprogramming—including redox homeostasis restoration, mitochondrial functional repair, and inflammatory cascade suppression—was validated at the genome scale, providing solid evidence for NMDs’ integrated therapeutic mechanism.

Fig. 7.

RNA-seq revealed the therapeutic mechanisms of NMDs. (A) PCA results of different samples in RNA-seq of mouse kidney tissues. (B) Venn diagram of DEGs in RNA-seq of kidney tissues. (C) Histogram of DEGs between different groups. (D–E) GO enrichment chord diagram (D) and KEGG pathway enrichment bubble chart (E) for DEGs between the AKI and NMDs groups. (F–I) Cluster heatmap of gene expression across different groups of four functional categories: oxidative stress (F), apoptosis (G), mitochondrial function (H), and DNA sensing (I). n = 3, |log₂FC|>1, Q < 0.05.

2.8. NMDs exhibited excellent biocompatibility

The distinct physicochemical properties of nanomedicines compared to traditional small-molecule drugs raise critical biosafety considerations, particularly given the high reactivity of metal-based nanomaterials [42]. Previous studies have demonstrated that titanium dioxide nanoparticles can trigger immune dysregulation via TLR pathways, while lanthanide nanoparticles accumulate in the liver, leading to functional impairment [43,44]. Notably, molybdenum—an essential trace element in humans—plays vital roles in metabolic processes. Despite its low dietary requirement [Recommended dietary allowance (RDA): 45 μg/day], molybdenum exhibits a substantial safety margin with an Upper Tolerable Intake Level (UL) of 2,000 μg/day. This wide RDA-UL interval reflects high body tolerance, enabling efficient elimination of excess molybdenum at elevated intake levels while maintaining homeostasis at lower exposures—a dual mechanism ensuring both deficiency prevention and biosafety of molybdenum-based therapeutics [45].

Preliminary biosafety assessments of NMDs revealed no acute cytotoxicity in vitro, with HK-2 cell viability remaining unaffected after 24-h exposure to 320 μg/mL NMDs (Fig. S32). In acute in vivo safety evaluations, single high-dose intravenous administration (10 mg/kg) induced no histopathological abnormalities in major organs (Fig. 8A–B) or significant alterations in hematological parameters (leukocytes, erythrocytes, platelets) and hepatic/renal function markers (AST, ALT, Bun, Cre) within 24 h (Fig. 8C–E, S33). Key inflammatory cytokines (TNF-α, IL-6, IL-1β) remained comparable to controls (Fig. 8F), confirming the absence of systemic inflammatory responses. Moreover, long-term safety evaluation revealed no histopathological abnormalities in major organs following four weekly administrations of therapeutic-dose NMDs (2 mg/kg at 7-day intervals) over the 28-day observation period (Fig. 8B). Hematological parameters (Fig. 8G, Fig. S34), hepatic/renal function indices, and inflammatory cytokine levels (Fig. 8H–J) remained comparable to PBS group. These findings systematically confirm NMDs’ absence of hepatorenal toxicity or immune activation under both acute and chronic exposure conditions, with their exceptional biocompatibility establishing critical groundwork for clinical translation.

Fig. 8.

NMDs exhibited excellent biocompatibility. (A) Schematic illustration of short-term and long-term treatment protocol for biosafety evaluation of NMDs. (B) Representative H&E-stained images of major organs from KM mice after short-term (S-) and long-term (L-) administration of NMDs or PBS. (C–F) Hematological parameters (C), liver function indicators (D), kidney function indicators (E), and inflammatory cytokine levels (F) of short-term treated mice in different treatment groups. (G–J) Hematological parameters (G), liver function indicators (H), kidney function indicators (I), and inflammatory cytokine levels (J) of long-term treated mice in different treatment groups. Data are presented as mean ± SD. n = 6.

3. Conclusion and discussion

The therapeutic management of AKI has long been constrained by conventional approaches’ failure to simultaneously address core pathological processes—targeted oxidative stress control at its source and activation of endogenous repair systems. In recent years, a growing number of antioxidant nanomedicines have emerged for AKI treatment, including various molybdenum-based platforms [[46], [47], [48]] and novel carbon dot systems [6]. These nanotherapies leverage the redox reactivity of the materials to effectively scavenge ROS and alleviate renal injury. Building upon this foundation, our study further expands the biomedical potential of Mo-based nanomaterials. We innovatively developed NMDs—a nanoplatform that integrates the dual functionalities of exogenous ROS clearance (via Mo valence transition) and endogenous antioxidant system restoration (via mARC and GSH regeneration)—to interrupt the cascade of redox imbalance and mitochondrial dysfunction in AKI. Notably, NMDs achieve precise therapeutic delivery through a sequential “kidney-PTECs-mitochondria” targeting strategy. This hierarchical targeting begins with rapid renal accumulation enabled by high hydrophilicity and ultrasmall size. NAC modification then promotes OAT1-mediated uptake into injured tubules, followed by mitochondrial anchoring via high affinity to membrane protein. Compared to passive enrichment, this spatial precision allows NMDs to neutralize mitochondrial ROS directly while reactivating endogenous enzymes (e.g., GPx4/mARC) through antioxidant precursor and cofactor supply. The synergy between these mechanisms creates a self-sustaining therapeutic loop: ROS clearance stabilizes the microenvironment for antioxidant system recovery, while enzyme reactivation reinforces mitochondrial integrity. This dual action restores ATP synthesis, suppresses apoptosis/inflammation, and halts AKI progression. Compared to nanomaterials with limited targeting and single mechanism, NMDs demonstrate superior therapeutic breadth by combining spatial precision with functional synergy, enabling both acute damage control and cellular self-defense restoration. Crucially, NMDs’ biocompatible design leverages endogenous metabolic pathways, avoiding nanoparticle accumulation risks critical for clinical translation. In summary, this study introduces a novel nanomedicine paradigm integrating precise delivery with multifunctional therapeutics to target complex disease networks. The “endo-exogenous antioxidant synergy” not only advances AKI treatment but also informs development for other oxidative stress-related conditions. Future work will explore its applications in CKD, advancing toward therapies that combine etiology targeting, functional repair, and systemic homeostasis.

4. Materials and methods

4.1. Materials

The chemicals and reagents used are listed in Table 1.

Table 1.

List of reagents.

Reagent	Manufacturer
Ammonium Molybdate Tetrahydrate	Macklin Biochemical (Shanghai)
Thiourea	Macklin Biochemical (Shanghai)
NAC	Yesen Biotechnology (Shanghai)
FITC	Solarbio Science & Technology (Beijing)
Cy7-NHS	Duofluor (Wuhan)
hMPO	Athens Research and Technology (USA)
HRP	Selleck (USA)
Electron Microscopy Fixative	Servicebio Technology (Wuhan)
DAPI Staining Solution	Beyotime Biotechnology (Shanghai)
DHE Fluorescent Probe	Yesen Biotechnology (Shanghai)
MitoSOX™ Red Mitochondrial Superoxide Indicator	Yesen Biotechnology (Shanghai)
Mouse NGAL ELISA Kit	Elabscience Biotechnology (USA)
Mouse KIM-1 ELISA Kit	Elabscience Biotechnology (USA)
Mouse 8-OHdG ELISA Kit	Elabscience Biotechnology (USA)
Mouse/Human TNF-α ELISA Kit	Elabscience Biotechnology (USA)
Mouse/Human IL-1β ELISA Kit	Elabscience Biotechnology (USA)
Mouse/Human IL-6 ELISA Kit	Elabscience Biotechnology (USA)
Mouse/Human CXCL1 ELISA Kit	Elabscience Biotechnology (USA)
Mouse/Human CXCL2 ELISA Kit	Elabscience Biotechnology (USA)
Mouse/Human CCL2 ELISA Kit	Elabscience Biotechnology (USA)
Pierce™ BCA Protein Assay Kit	Thermo Fisher Scientific (USA)
TUNEL Assay Kit	Beyotime Biotechnology (Shanghai)
CCK8 Assay Kit	Beyotime Biotechnology (Shanghai)
Annexin V-FITC Apoptosis Detection Kit	Beyotime Biotechnology (Shanghai)
Mitochondrial Membrane Potential Assay Kit (JC-1)	Beyotime Biotechnology (Shanghai)
Cre Assay Kit	Jiancheng Bioengineering Institute (Nanjing)
BUN Assay Kit	Jiancheng Bioengineering Institute (Nanjing)
TBARS Assay Kit	Elabscience Biotechnology (USA)
MDA Assay Kit	Jiancheng Bioengineering Institute (Nanjing)
GPX4 Activity Assay Kit	Elabscience Biotechnology (USA)
GSH/GSSG Assay Kit	Jiancheng Bioengineering Institute (Nanjing)
DCFH-DAROS Assay Kit	Yesen Biotechnology (Shanghai)
AST Assay Kit	Jiancheng Bioengineering Institute (Nanjing)
ALT Assay Kit	Jiancheng Bioengineering Institute (Nanjing)
Bax Antibody	Abcam (UK)
Bcl-2 Antibody	Affinity
Cyt c Antibody	Proteintech
Caspase-3 Antibody	Proteintech
Cleaved caspase-3 Antibody	Wanleibio
cGAS Antibody	Proteintech
STING Antibody	Proteintech
IRF3 Antibody	Affinity
p-IRF3 Antibody	Affinity
P65 Antibody	Affinity
p-P65 Antibody	Affinity)
α-Tubulin Antibody	Affinity
dsDNA Antibody	Abcam (UK)
TOM20 Antibody	Abcam (UK)
Ly-6G Antibody	Abcam (UK)
F4/80 Antibody	Proteintech

4.2. NMDs, MoS₂, FITC-NMDs, Cy7-NMDs synthesis

NMDs were synthesized via a hydrothermal growth process using ammonium molybdate and thiourea as molybdenum and sulfur sources, respectively, with NAC serving as a capping agent. Specifically, 0.65 g of ammonium molybdate tetrahydrate, 0.20 g of NAC, and 0.08 g of thiourea were accurately weighed using an analytical balance. Ammonium molybdate and NAC were dissolved in 40 mL of ice-cold ultrapure water, and the mixture was stirred vigorously in an ice bath. Thiourea was then added to the precursor solution under continuous stirring until fully dissolved. The solution was subjected to a hydrothermal reaction at 200 °C for 4 h. The resulting solution was centrifuged at 12,000 rpm for 10 min, and the supernatant, a brown-yellow transparent liquid, was collected. This liquid was dialyzed using a 3.5 kDa dialysis membrane under anaerobic conditions at room temperature for 2 h. After dialysis, the solution was centrifuged again at 12,000 rpm for 10 min, and the supernatant was freeze-dried to obtain solid NMDS powder. For comparison, MoS₂ powder was synthesized under the same conditions without the addition of NAC.

Next, 10 mg of NMDS powder was accurately weighed and dissolved in 8 mL of ultrapure water, and 2 mg of fluorescein isothiocyanate (FITC) was dissolved in 2 mL of dimethyl sulfoxide (DMSO). The two solutions were then combined in a beaker and stirred in a water bath at 25 °C in the dark for 5 h. The resulting bright yellow solution was dialyzed in a 3.5 kDa dialysis membrane at room temperature for 24 h to obtain FITC-NMDs.

NMDs (10 mg) was dispersed in PBS and mixed with Cy7-NHS (2 mg) under gentle stirring. The reaction was allowed to proceed for 5 h at room temperature in the dark to prevent photobleaching of Cy7. Following conjugation, the mixture was dialyzed for 24 h to remove unreacted Cy7 dye and byproducts.

4.3. Nanomaterials characterization

XPS was performed using a VG MKII spectrometer, and the high-resolution spectra were deconvoluted with XPS software (version 5.991). The Zeta potential and hydrodynamic diameter were measured using the Nano ZS90. TEM images were obtained with a TECNAI G2 high-resolution TEM. Aberration-corrected STEM-EELS analysis was carried out on a JEOL ARM300F microscope. XRD patterns were recorded using a Bruker D8 Advance diffractometer. TGA was conducted on a TA Q500 analyzer under airflow. Raman spectra were collected using a Renishaw inVia Raman microscope with a 532 nm laser.

4.4. In vitro ROS scavenging detection

The H₂O₂ and ONOO⁻ scavenging activities of NMDs and MoS₂ were measured via UV–visible spectrophotometry. Briefly, NMDs and MoS₂ were incubated with varying concentrations of H₂O₂ or ONOO⁻ in the dark for 12 h. The UV absorption spectra were analyzed to determine the scavenging rates of H₂O₂ and ONOO⁻.

The ·OH scavenging efficiency of NMDs and MoS₂ was assessed by measuring the inhibition of TMB oxidation. In brief, TMB (400 μL, 10 mM), ferrous sulfate (40 μL, 50 mM), and hydrogen peroxide (20 μL, 100 mM; final concentrations of 1, 2, 4, and 6 μg/mL, respectively) were added to PBS solution (0.01 mol/L, pH = 7.4), adjusting the total volume to 3 mL. Following a 3-min dark reaction, the mixture was transferred to a cuvette, and the absorbance change at 652 nm was recorded.

The O₂^·- scavenging activity of NMDs and MoS₂ was evaluated using the NBT assay. Briefly, methionine (390 μL, 0.1 M), riboflavin (6 μL, 20 μM), NBT (22.5 μL, 0.01 M), and ANT were added to a PBS solution (1.5 mL, 0.1 mol/L, pH = 7.4) to reach final concentrations of 25, 50, 100, and 200 μg/mL, adjusting the volume to 3 mL with deionized water. The mixture was exposed to UV light for 5 min, and the UV absorption spectrum was then recorded. The inhibition rate was determined by the decrease in absorbance at 560 nm.

4.5. Enzymatic degradation of NMDs by HRP and hMPO

To investigate HRP-catalyzed degradation, 200 mL of NMDs suspension (0.1 mg/mL in PBS) was incubated with 7 mg of HRP under constant stirring at room temperature in the dark. Hydrogen peroxide (6 %, w/v) was freshly added every 6 h over a 24-h period to maintain enzymatic activity. For hMPO-mediated degradation, 200 mL of NMDs solution (0.1 mg/mL in 50 mM PBS containing 140 mM NaCl) was mixed with 200 μg of hMPO. Hydrogen peroxide (6 %, w/v) was added every 6 h for a total duration of 24 h. The mixture was maintained at 37 °C under gentle shaking in the dark.

4.6. Establishment of RM-AKI mouse model and treatment strategy

AKI mouse model induced by rhabdomyolysis: KM mice (male, 4–6 weeks old, 20–25 g) were purchased from Hunan STA Laboratory Animal Co., Ltd. (Changsha, China) and housed in groups of six per cage with adaptive feeding for three days prior to study inclusion. After 15 h of water deprivation, 50 % glycerol (8 mg/kg) was injected intramuscularly in equal amounts into the bilateral hind limbs to induce AKI. Healthy control mice received bilateral intramuscular injections of saline in place of glycerol, after which water access was restored. Treatments began 2 h post-model induction. Group 1: Healthy mice, treated with a single intravenous injection of saline; Group 2: AKI mice, treated with a single intravenous injection of saline; Group 3: AKI mice, treated with NMDs (0.5 mg/kg); Group 4: AKI mice, treated with NMDs (1 mg/kg); Group 5: AKI mice, treated with NMDs (2 mg/kg); Group 6: AKI mice, treated with MoS₂ (2 mg/kg); Group 7: AKI mice, treated with NAC (2 mg/kg).

All treatments were administered as a single injection. Mice were euthanized 24 h post-treatment. All animal experiments were approved by the Animal Care and Use Committee of Central South University.

4.7. Establishment of cisplatin-induced AKI mouse model and treatment strategy

KM mice (male, 4–6 weeks old, 20–25 g) were purchased from Hunan STA Laboratory Animal Co., Ltd. (Changsha, China) and housed in groups of six per cage with adaptive feeding for three days prior to study inclusion. To induce cisplatin-induced AKI, mice received a single intraperitoneal (i.p.) injection of cisplatin (20 mg/kg, diluted in 0.9 % saline). Treatments began 2 h post-model induction. Group 1: Healthy mice, treated with a single intravenous injection of saline; Group 2: AKI mice, treated with a single intravenous injection of saline; Group 3: AKI mice, treated with NMDs (1 mg/kg); Group 4: AKI mice, treated with NMDs (2 mg/kg); Group 5: AKI mice, treated with NMDs (4 mg/kg); Group 6: AKI mice, treated with MoS₂ (4 mg/kg); Group 7: AKI mice, treated with NAC (4 mg/kg).

All treatments were administered as a single injection. Mice were euthanized 72 h post-treatment. All animal experiments were approved by the Animal Care and Use Committee of Central South University.

4.8. Evaluation of treatment efficacy in AKI mice

The primary indicators for evaluating treatment efficacy were Cre and Bun. Additional indicators included body weight, kidney injury factor levels, and kidney tissue damage. Kidney tissue injury was assessed using H&E staining.

Tubular injury scoring was based on the structural integrity of renal tubules, tubular dilation, cell shedding, and inflammatory infiltration, on a scale from 0 to 4, as follows: Score 0: Tubular structures intact, with no evident signs of injury. Score 1: Mild injury, with less than 25 % of the tubular area showing slight dilation or cell shedding. Score 2: Moderate injury, with 25–50 % of the tubular area exhibiting moderate dilation, cell shedding, or mild inflammatory infiltration. Score 3: Severe injury, with 50–75 % of the tubular area affected, showing significant dilation, cell shedding, and notable inflammatory cell infiltration. Score 4: Extreme injury, with more than 75 % of the tubular area severely damaged, extensive dilation, cell shedding, and severe inflammatory infiltration.

Levels of inflammatory cytokines (IL-1β, IL-6, TNF-α) and chemokines (CCL2, CXCL1, CXCL2) were measured using ELISA kits according to the manufacturer's instructions. The levels of CRE, BUN, MDA, HO-1, and GSH/GSSG were also measured using corresponding assay kits.

4.9. Pharmacokinetic study of NMDs

To evaluate the pharmacokinetic profile and plasma clearance of NMDs, KM mice (25 ± 2 g, n = 6 per group, with equal numbers of males and females) were randomly divided into two groups: Control and RM-AKI. Mice in both groups received a single intravenous injection of NMDs at a dose of 2 mg/kg. Blood samples were collected from the orbital sinus at predetermined time points (0, 5 min, 15 min, 30 min, 1 h, 3 h, 6 h, 9 h, 12 h, and 24 h) post-injection. Following centrifugation (12,000 rpm, 10 min), the plasma was carefully collected and diluted with 2 % v/v HNO₃ for elemental analysis. The Mo content in plasma was quantified using ICP-MS, which reflects the circulating level of NMDs. The obtained Mo concentration–time data were processed using DAS 2.0 software to fit the pharmacokinetic model and calculate parameters including half-life (t_1/2), clearance (CL), and area under the concentration–time curve (AUC).

4.10. Excretion study of NMDs

To investigate the excretion profile and elimination pathways of NMDs, KM mice (25 ± 2 g, n = 6 per group, with equal numbers of males and females) were randomly assigned to Control and AKI groups. After intravenous administration of NMDs at a dose of 2 mg/kg, mice were individually housed in metabolic cages for urine and feces collection at three time intervals: 0–6 h, 6–12 h, and 12–24 h. At each time point, the total volume of urine and weight of feces were recorded. Samples were then acid-digested with % v/v HNO₃, followed by centrifugation (12,000 rpm, 10 min). The supernatants were collected and analyzed for Mo content using ICP-MS.

4.11. Fluorescence imaging using stereoscopic and in vivo systems

Two hours post-model induction, mice were treated once with FITC-labeled NMDs (2 mg/kg) and euthanized at various time points to collect organs, including the heart, kidneys, lungs, spleen, and liver. Fluorescence intensity in each organ was analyzed using the Xenogen IVIS Lumina in vivo imaging system with an FITC filter channel and a 0.5-s exposure time. Additionally, fluorescence imaging of organs was performed using a stereoscopic fluorescence microscope.

4.12. In vivo tracking of Cy7-NMD distribution in mice

KM mice were intravenously injected with Cy7-NMDs at a dose of 2 mg/kg. Real-time fluorescence imaging was conducted at 1, 3, 6, 9, 12, 24, 48 and 72 h post-injection using an In Vivo Imaging System (IVIS) with excitation and emission filters set for Cy7. Mice were anesthetized before imaging to minimize movement and ensure accurate visualization. Fluorescence intensity was captured and analyzed in real time to assess the temporal biodistribution of Cy7-NMDs throughout the body.

4.13. Tracking of Mo in vivo by ICP-MS

Two hours post-model induction, mice were treated once with NMDs and euthanized at 6 and 9 h post-treatment. Organs, including the heart, kidneys, lungs, spleen, and liver, were collected, and the Mo content in each organ was analyzed using the ICP-MS detection system.

4.14. Pharmacokinetic study of NMDs

To evaluate the pharmacokinetics of NMDs, Control or RM-AKI mice received a single tail vein injection of NMDs (2 mg/kg). Blood samples were collected at 5 min, 15 min, 0.5 h, 1 h, 3 h, 6 h, 9 h, 12 h, and 24 h post-injection. Plasma was obtained by centrifugation, followed by acid digestion for elemental analysis. The Mo concentration in plasma was quantified using ICP-MS, and pharmacokinetic parameters were calculated using DAS software.

4.15. TEM imaging of renal tissue

Renal tissues from each group were collected, fixed immediately for transmission electron microscopy, dehydrated, resin-embedded, and polymerized. Resin blocks were sectioned onto 150 grids, stained, and observed under a transmission electron microscope (HITACHI HT7800/HT77000) to capture images. Field-emission TEM imaging was performed using the Talos F200X system.

4.16. Molecular docking

AutoDock Vina (version 1.1.2), which utilizes a semi-flexible docking approach, was used for molecular docking in this study. The structure of the small molecule NAC was downloaded from the PubChem database, and docking was performed with four proteins—Tom20, Tom34, Vdac1, and Vdac2—designating them as receptors and acetylcysteine as the ligand. PyMOL (version 4.3.0) was used to prepare the protein structures by removing the original ligands, water, and organic molecules, while AutoDockTools was used for hydrogen addition, charge calculation, and setting atom types to AD4. A docking grid box was constructed for each protein structure, and the small molecule ligand was prepared by defining its root and selecting rotatable bonds in AutoDockTools. Protein and ligand files were converted from “.PDB” to “.PDBQT” format for docking. After docking with Vina, binding scores for each protein-ligand complex were calculated, and interaction analysis was visualized in both 3D and 2D perspectives using PyMOL and Discovery Studio 2019.

4.17. In vivo biocompatibility assessment

Long-term: Healthy KM mice were randomly assigned into two groups, receiving either saline or NMDs (2 mg/kg) once every seven days for a total of four treatments. Mice were euthanized one day after 28 days, and major organs (heart, liver, spleen, lungs, and kidneys) as well as blood samples were collected. Histological changes in organs were evaluated using H&E staining, and hematological analysis was performed on whole blood samples using the Sysmex XE-2100 system (Kobe, Japan). The blood parameters assessed included red blood cells (RBC), white blood cells (WBC), and hemoglobin (HGB). Alanine aminotransferase (ALT), aspartate aminotransferase (AST), and blood urea nitrogen (BUN) levels were determined using colorimetric assays.

Short-term: Healthy KM mice were randomly divided into two groups, receiving a single treatment of either saline or a higher dose of NMDs (10 mg/kg). One day post-treatment, mice were euthanized, and organ and blood samples were collected and analyzed as described above.

4.18. Cell culture

HK-2 cells were cultured in a DMEM/F-12 medium supplemented with 10 % fetal bovine serum. All cells were maintained at 37 °C in a humidified atmosphere with 5 % CO₂. The cell line was obtained from the National Certified Cell Culture Collection Center of China.

4.19. Intracellular ROS clearance

HK-2 cells were seeded into 24-well plates at a density of 2 × 10⁵ cells per well and incubated overnight. Cells were then treated with H₂O₂ (500 μM), or a combination with NMDs (40 μg/mL), MoS₂ (40 μg/mL) and NAC (40 μg/mL) for 4 h. Afterward, cells were labeled with the DCFH-DA ROS probe (CA1410, Beijing, China) and observed under a fluorescence microscope.

4.20. Assessment of mitochondrial function in cells

HK-2 cells were treated as 4.13. After removing the medium, cells were labeled with JC-1 working solution or MitoSOX™ reagent and incubated in the dark at 37 °C for 20 min, followed by staining with Hoechst 33342 for 20 min. Cells were then washed three times with HBSS and observed under a fluorescence microscope, followed by flow cytometry analysis.

For ATP production assays, after the same treatments for 4 h, HK-2 cells were collected and washed three times with cold HBSS. ATP levels were determined and calculated for each group following the manufacturer's protocol of the enhanced ATP detection kit.

4.21. Western Blotting

Cell or tissue samples were lysed in RIPA buffer with PMSF (1 mM), and supernatants were collected by centrifugation at 12,000 rpm for 20 min. Protein concentrations were quantified using the BCA method, mixed with loading buffer (1:1), and denatured at 95 °C for 7 min. Western blotting was then performed. In brief, after electrophoresis, membrane transfer, and blocking, membranes were incubated with primary antibodies (diluted 1:500–2000) overnight. After washing three times with TBST, membranes were incubated with secondary antibodies (diluted 1:5000) at room temperature for 1 h, followed by additional washes. Finally, membranes were visualized immediately with ECL reagents using the Bio-Rad Gel Imaging System (Bio-Rad, Harkles, CA, USA).

4.22. H&E staining

Renal tissue samples were sliced into 5 μm sections, mounted on slides, and dried at 55 °C for 30 min. Sections were deparaffinized in xylene (2 × 5 min) and rehydrated through a graded ethanol series: 100 % ETOH (2 × 20 dips), 95 % ETOH (2 × 20 dips), and 80 % ETOH (2 × 20 dips). The staining process included an 8-min hematoxylin stain, followed by thorough rinsing, 1-min differentiation, and an additional rinse. Sections were then incubated in 95 % ethanol for 5 min, counterstained with eosin for 1 min, dehydrated in absolute ethanol, and cleared in xylene (100 % ETOH (3 × 20 dips), xylene (3 × 20 dips)). Finally, coverslips were applied using neutral resin.

4.23. Immunofluorescence staining

Cells were washed three times in cold PBS with 1 % Tween 20, then fixed in 100 % methanol (pre-cooled at −20 °C) at room temperature for 5 min. After washing, permeabilization, and blocking, cells were incubated overnight at 4 °C with the primary antibody solution (1:200 dilution). Following washes, cells were incubated with cross-absorbed secondary antibodies (1:1000 dilution) for 1 h. For dual staining, the blocking and staining procedures were repeated for the second target. Cells were finally stained with DAPI (1 μg/mL) for 1 min and observed under a fluorescence microscope.

For tissue staining, kidney tissue was fixed in 10 % formalin and embedded in paraffin. Sections (5 μm) were subjected to immunostaining after deparaffinization, following a similar protocol as cells’ immunofluorescence staining.

4.24. Immunohistochemistry staining

Kidney tissue sections were cut into 5 μm slices, mounted on slides, and dried at 55 °C for 30 min. Deparaffinization, antigen retrieval, and pretreatment with proteinase K (20 μg/mL) at 37 °C for 20 min were performed. Sections were quenched in 0.3 % hydrogen peroxide for 10 min, washed three times in PBS, and blocked. Primary and biotin-labeled secondary antibodies (1:100 dilution) were applied sequentially. The sections were processed with an ABC kit and DAB substrate. Finally, hematoxylin was applied for counterstaining (30 s–1 min), and sections were rinsed in PBS for visualization under a microscope.

4.25. RNA-seq analysis

Total RNA was extracted from kidney tissue using TRIzol reagent (Invitrogen, CA, USA) following the manufacturer's instructions. After assessing RNA purity, quantification, and integrity, libraries were prepared with the VAHTS Universal V6 RNA-seq Library Prep Kit as per the manufacturer's protocol. BGI Genomics Co., Ltd. (Shenzhen, China) conducted transcriptomic sequencing and analysis.

4.26. Statistical analysis

All experiments were repeated independently at least three times. Quantitative data are expressed as mean ± standard deviation (SD). Statistical analysis was conducted using SPSS 23.0 software. Comparisons between two groups were performed with an independent-sample t-test, and multiple group comparisons were conducted via one-way ANOVA, followed by the Student-Newman-Keuls (SNK) post hoc test for multiple comparisons. Statistical significance was defined as P < 0.05.

CRediT authorship contribution statement

Qiaohui Chen: Writing – original draft, Investigation. Zuoxiu Xiao: Investigation. Xiaohong Ying: Investigation. Yongqi Yang: Investigation. Jianlin Chen: Investigation. Ziyu Wu: Investigation. Wan Zeng: Investigation. Chenxi Miao: Investigation. Yayun Nan: Investigation. Qiong Huang: Writing – review & editing, Writing – original draft, Supervision, Funding acquisition. Kelong Ai: Writing – review & editing, Writing – original draft, Supervision, Project administration, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

This study and included experimental procedures were approved by the institutional animal care and use committee of Xiangya School of Medicine, Central South University (approval no. XMSB-2024-0263). All animal housing and experiments were conducted in strict accordance with the institutional guidelines for care and use of laboratory animals.

Declaration of competing interests

The authors declare no conflict of interest, financial or otherwise.

Appendix A. Supplementary data

The following is the Supplementary data to this article: