Multi-targeted biomimetic CeO₂-based nanoplatform for sepsis therapy via mitochondrial homeostasis Authors: Xiaolin Tan, Zaili Zhang, Mingzhe Wu, Yuan Tan, Xue Zhang, Chang Li, Zhifeng Wen, Yanxia Lin, Jun Chen, Yao Li Published: 2026 ABSTRACT Sepsis is a life - threatening systemic inflammatory response syndrome triggered by pathogenic infections, which is characterized by severe oxidative stress, immune dysregulation, and multiple organ dysfunction, accompanied by a high mortality rate. Its primary pathogenesis encompasses pathogen - induced immune overactivation, resulting in cytokine storms, oxidative damage, tissue injury, and organ failure.This research developed a bionic nanotherapeutic system for the treatment of sepsis. Cerium dioxide (CeO₂) nanoparticles with inherent antioxidant activity were employed as carriers, loaded with the broad – spectrum antibiotic imipenem - cilastatin sodium hydrate (IC), and further coated with macrophage membranes (MM). The MM conferred immune evasion, homologous targeting, and lipopolysaccharide (LPS) - neutralizing capabilities to the nanosystem.This integrated platform combines CeO₂ - mediated reactive oxygen species (ROS) scavenging, antibiotic - induced bactericidal effects, and MM - mediated inflammatory regulation. In vitro investigations verified its potent ROS - scavenging, LPS - neutralizing, anti - inflammatory, and antibacterial activities. In a cecal ligation and puncture (CLP) - induced mouse sepsis model, IC@CeO₂@MM effectively reduced both systemic and local inflammatory cytokines, alleviated multi - organ injury in the liver, kidneys, and intestines, and enhanced survival rates. Mechanistically, it maintained mitochondrial membrane potential and decreased mitochondrial ROS (mtROS) accumulation in macrophages.This synergistic strategy simultaneously achieves antibacterial, antioxidant, and immunomodulatory effects, presenting a promising and novel approach for the comprehensive management of sepsis. Introduction Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, and is one of the leading causes of mortality in intensive care units worldwide (Fathi et al., 2019; Vincent et al., 2006). Its pathophysiological process begins with the invasion of pathogens (such as bacteria, fungi, or viruses) (Harder et al., 2018). Toxins released by these pathogens (e.g., lipopolysaccharides) are recognized by host pattern recognition receptors, leading to the excessive activation of both innate and adaptive immune systems (Kashif et al., 2025; Shubhra et al., 2025; Zhou et al., 2025). This uncontrolled activation triggers a “cytokine storm,” releasing large amounts of pro-inflammatory mediators such as tumor necrosis factor-α and interleukin-1β, accompanied by severe oxidative stress that causes a sharp increase in reactive oxygen species (ROS) levels (Li et al., 2024). Inflammation and oxidative stress mutually reinforce each other, forming a vicious cycle that results in widespread vascular endothelial damage, increased capillary permeability, microcirculatory dysfunction, and inadequate tissue perfusion, ultimately progressing irreversibly to multiple organ failure (Jarczak and Nierhaus, 2022). Although clinical guidelines recommend early antibiotic use and supportive therapy, simply eliminating pathogens often fails to reverse the initiated intense inflammation and oxidative damage cascades (Zaragoza et al., 2020), leading to persistently high patient mortality rates. This underscores the urgency of developing novel therapeutic strategies capable of simultaneously targeting multiple pathological processes such as infection, inflammation, and oxidative damage (Srdić et al., 2024). To overcome the limitations of conventional treatments, the field of nano-biomaterial offers new opportunities for the comprehensive management of sepsis. Current research primarily focuses on two major directions: first, the development of nano-drugs with intrinsic pharmacological activity, such as cerium dioxide nanoparticles possessing superoxide dismutase and catalase-mimetic activities (Deval et al., 2021; Sarnatskaya et al., 2020). These nanoparticles efficiently scavenge excess ROS through reversible Ce3+/Ce4+ valence cycling and have been demonstrated in multiple studies to effectively mitigate oxidative stress damage in sepsis models. Second, the design of smart nano-drug delivery systems that utilize their targeted delivery properties to enhance therapeutic efficiency and reduce side effects. Among these, biomimetic nanotechnology, particularly cell membrane camouflage technology, has become a cutting-edge hotspot (Fang et al., 2023). As recently reviewed by Li et al., nanomaterial-based encapsulation of bioactive substances—including antioxidants, anti-inflammatory agents, and immunomodulators—enhances the stability, bioavailability, and targeted delivery of therapeutic agents, representing a transformative approach to sepsis therapy (Li et al., 2025). By coating nanoparticle surfaces with natural cell membranes (e.g., red blood cell membranes, neutrophil membranes, or macrophage membranes (MM)), these nanoparticles can inherit the complex bio-interface functions of the source cells, thereby achieving long circulation, immune evasion, and active targeting to inflammatory sites (Yu et al., 2021). However, most existing systems have relatively single functions, focusing either on antibacterial, anti-inflammatory, or antioxidant effects, making it difficult to achieve synergistic therapeutic outcomes in the complex pathological environment of sepsis. Therefore, integrating multiple therapeutic modules to construct an integrated nano-platform capable of precisely identifying lesions and simultaneously performing multiple tasks—“bactericidal-ROS clearance-immune regulation”—is a significant challenge and development trend in the current field. Based on the above background, this study aimed to develop a multifunctional biomimetic nano-therapeutic system for the synergistic treatment of sepsis. Specifically, CeO₂ nanoparticles with strong antioxidant capacity were used as the core carrier to scavenge excessive ROS. The broad-spectrum antibiotic imipenem-cilastatin sodium hydrate (IC) was loaded onto CeO₂ nanoparticles to achieve efficient pathogen eradication. Importantly, the nano-complex was coated with a macrophage membrane to enhance biocompatibility, evade immune clearance, and prolong circulation time. Meanwhile, the chemokine receptors and adhesion molecules on the macrophage membrane enable active targeting and accumulation at inflammatory sites. In addition, membrane-retained receptors can neutralize circulating pro-inflammatory cytokines and endotoxins, thereby mitigating the cytokine storm. This biomimetic system functions as a “Trojan horse” that precisely targets infection sites and simultaneously delivers antibiotics, scavenges ROS, and modulates inflammation, thereby disrupting the vicious cycle of infection, inflammation, oxidative damage, and organ failure in sepsis. Our results demonstrate that IC@CeO₂@MM exhibits potent antibacterial and anti-inflammatory activities and significantly improves survival in septic models. Mechanistically, IC@CeO₂@MM preserves mitochondrial homeostasis, which contributes to the suppression of the inflammatory cascade in sepsis and providing a novel strategy for synergistic antibacterial, antioxidant, and immunomodulatory therapy. Materials and methods Preparation of CeO₂ nanozymes 0.5 g of cerium (III) nitrate hexahydrate and 0.2 g of PVP K30 were dissolved in 15 mL of ethylene glycol under vigorous stirring. Subsequently, 1 mL of an aqueous solution containing 1 M hydrochloric acid was added dropwise. After stirring for 30 min, the clear solution was transferred into a Teflon-lined autoclave and maintained at 160 °C for 3 h. After cooling to room temperature, the gray nanoparticles were collected and washed three times with deionized water, each time centrifuged at 10,000 rpm for 20 min. Preparation of IC@CeO₂ nanozymes 100 mg of cerium dioxide nanoparticles were dispersed in 10 mL of deionized water. Imipenem-cilastatin sodium hydrate (IC) was dissolved in dimethyl sulfoxide to prepare a 50 mM solution. Then, 4 mL of the IC solution was added dropwise to the cerium dioxide suspension. After stirring for 12 h, IC@CeO₂ was obtained. Hyaluronic acid (HA) was then added at a mass ratio of 1:4 (CeO₂:HA), and stirring continued for another 12 h. The resulting nanoparticles were alternately washed three times with water and ethanol, each time centrifuged at 10,000 rpm for 20 min. The final product was dispersed in deionized water. Extraction of macrophage membranes First, mouse bone marrow-derived macrophages (BMDMs) were isolated as follows: Tibias and femurs were harvested from mice, and the epiphyses were cut under sterile conditions. The bone marrow cavity was flushed with PBS containing 2% FBS. The collected flushing solution was centrifuged at 400 ×g for 5 min, and the supernatant was discarded. The pellet was resuspended in red blood cell lysis buffer and incubated on ice for 3–5 min. An equal volume of PBS was added to terminate the lysis, followed by another centrifugation. The supernatant was removed, and the cells were resuspended in DMEM high-glucose medium supplemented with 10% FBS, 1% penicillin-streptomycin, and 30 ng/mL M-CSF. Cells were cultured in a 37 °C incubator with 5% CO₂. On day 3, half of the medium was replaced with fresh medium. After 6–7 days of differentiation, the adherent cells were identified as mature bone marrow-derived macrophages and used for subsequent experiments. The extracted macrophages were resuspended in PBS and immediately stored at −80 °C. The frozen cells underwent 2–3 freeze-thaw cycles until a viscous, mucus-like consistency was achieved. The resulting lysate was transferred to penicillin vials, mixed with a small amount of PBS, and sonicated in pulse mode (1 s on/1 s off for 2 min) at the lowest power setting (10%). The sonicated suspension was then centrifuged at 3000–4000 g for 10 min to remove cellular debris and organelles, retaining the supernatant. For membrane purification, the supernatant was subjected to high-speed centrifugation (14,800 g, 4 °C, 30 min), followed by a 10-min rest and another 30-min centrifugation. Alternatively, ultracentrifugation at 100,000 g for 30 min could be used. The final membrane pellet was resuspended in pre-cooled PBS and homogenized by gentle pipetting. Preparation of IC@CeO₂@MM nanozymes Macrophage membrane fragments were mixed with IC@CeO₂ nanoparticles at a mass ratio of 1 mg (nanoparticles) to 200 μg (cell membrane, MM) in PBS. The mixture was then subjected to intermittent sonication (25% power, 3 s on/6 s off) for 10 min to promote membrane fusion. Subsequently, the sample was centrifuged at 12,000 rpm (4 °C, 10 min) to remove unbound membrane components, yielding the final IC@CeO₂@MM nanoparticles. ABTS+ scavenging assay First, a 7.4 mM ABTS+ solution was prepared by dissolving 3 mg of ABTS diammonium salt in 0.735 mL of distilled water. Additionally, a 2.6 mM potassium persulfate (K₂S₂O₈) solution was prepared by dissolving 1 mg of K₂S₂O₈ in 1.43 mL of distilled water. To generate ABTS+ cations, 0.2 mL of the ABTS solution was mixed with 0.2 mL of the K₂S₂O₈ solution and incubated in the dark at room temperature for 12 h. The resulting solution was diluted 10 to 20 times with water to achieve an absorbance of approximately 0.7 at 734 nm, forming the working solution. In the experiment, different concentrations of IC@CeO₂@MM nanozymes were incubated with 1 mL of the ABTS+ working solution at room temperature for 15 min, and the absorbance was measured at 734 nm using a UV–Vis spectrophotometer. DPPH· scavenging assay First, a DPPH solution was prepared by dissolving 1 mg of DPPH in 24 mL of ethanol, followed by 5 min of sonication. The solution was then diluted with ethanol to adjust the absorbance at 519 nm to between 0.6 and 1.0, forming the working solution. Different concentrations of IC@CeO₂@MM nanozymes were co-incubated with 1 mL of the DPPH working solution at room temperature for 30 min. The absorbance was then measured at 519 nm using a UV–Vis spectrophotometer. In vitro antibacterial activity evaluation The dilution plate method was used to evaluate the in vitro antibacterial efficacy of the materials, following serial dilution steps. Briefly, E. coli (ATCC 25922) or S. aureus (ATCC 6538) was purchased from China Center of Industrial Culture Collection (CICC) and inoculated into LB broth at a 1:1000 (volume/volume) ratio and cultured overnight at 37 °C with shaking. Bacterial cells were then collected by centrifugation, washed, and resuspended in PBS solution. The bacterial suspension was incubated with different test materials for 12 or 24 h. After treatment, the samples were serially diluted, spread on LB agar plates, and incubated at 37 °C for 12 h. Colony-forming units (CFU) on each plate were counted, and the bacterial survival rate was calculated accordingly. LPS neutralization assay An LPS ELISA kit (Andy Gene, Beijing) was used to evaluate the LPS adsorption capacity of the nanozymes. In the standard experimental procedure, 5 μL of LPS solution (concentration 240 ng/mL) was mixed with 5 μL of IC@CeO₂@MM nanozyme suspension (concentration 5 mg/mL in PBS, pH 7.4; 3 replicates). At specified time points, 100 μL of the supernatant was collected by centrifugation and processed according to the manufacturer's instructions. The absorbance at 450 nm was recorded, and the LPS concentration was calculated using a standard curve. Cell culture The mouse macrophage cell line RAW264.7 (ATCC) was cultured in DMEM/High glucose medium supplemented with 10% FBS and 1% penicillin/streptomycin at 37 °C under 5% CO₂ in a humidified atmosphere. RNA isolation and real-time quantitative PCR (RT-qPCR) Total RNA was extracted with TRIzol reagent (93,289, Sigma), and 1 μg of RNA was reverse-transcribed into cDNA using 1st Strand cDNA Synthesis SuperMix (11137ES70, Yeasen). RT-qPCR was performed with SYBR Green Master Mix (11201ES, Yeasen) on a QuantStudio 5 system. Gene expression was normalized to ACTIN and calculated using the 2 − ΔΔCT method. Western blot Liver tissue (10 mg) and 106 cells was lysed in 300 μL RIPA buffer (P0013K, Beyotime) containing 1% PMSF (ST506, Beyotime) on ice. After centrifugation (12,000 ×g, 15 min, 4 °C), the supernatant was collected, mixed with 5× SDS loading buffer (20315ES20, Epizyme), and denatured at 100 °C for 10–12 min. Primary antibodies used were: anti-CD86 (CST, #91882, 1:1000), anti-iNOS (CST, # 13120, 1:1000), anti-VDAC (CST, #4866, 1:1000), anti-TFAM (CST, #8076, 1:1000), anti-CD11b (Proteintech, 66519–1-Ig, 1:2000), anti-TOM20 (Proteintech, 11802–1-AP, 1:2000), anti-TLR4 (Proteintech, 19811–1-AP, 1:500), anti-ACTIN (ABclonal, #A11283, 1:20000). Measurement of mitochondrial ROS RAW264.7 cells seeded in 12-well plates were treated with 500 μM H₂O₂ and various formulations for 24 h. Cells were then washed, incubated with 5 μM MitoSOX™ Red (MT14, DOJINDO) at 37 °C for 10 min, and rewashed. Mitochondrial superoxide levels were immediately visualized by confocal microscopy and quantified using ImageJ. Assessment of mitochondrial membrane potential (ΔΨm) ΔΨm was assessed using the JC-1 fluorescent dye. RAW264.7 cells were incubated with JC-1 (5 μg/ml, HY-K0601, MCE) for 15 min at 37 °C. Cells were then imaged, where green fluorescence (∼535 nm, JC-1 monomers) indicates mitochondrial depolarization, and red fluorescence (∼600 nm, JC-1 aggregates) reflects intact ΔΨm. CLP murine model of sepsis Six- to eight-week-old male C57BL/6 J mice were obtained from Jackson Laboratories and housed under specific pathogen-free conditions (22 ± 2 °C, 12-h light-dark cycle, maximum 5 mice per cage). All experimental procedures were approved by the Institutional Animal Care and Use Committee. Inclusion criteria required that mice recover from anesthesia and remain alive at the first assessment time point (1 h post-CLP); exclusion criteria included failed cecal puncture/ligation and procedure-unrelated injuries. Mice were randomly assigned to groups using a computer-generated random number sequence. Under isoflurane anesthesia and local bupivacaine (0.25%, 1–2 mg/kg, HY-B0405, MCE), the cecum was exposed via midline incision, ligated below the ileocecal valve with 4–0 silk. After surgery, mice were monitored every 4 h for the first 24 h and twice daily thereafter. All efforts were made to minimize animal suffering. In addition to buprenorphine analgesia, mice were provided with soft food pellets on the cage floor. Euthanasia was performed by cervical dislocation under isoflurane anesthesia to ensure minimal distress. H&E staining Tissue samples harvested from different experimental groups were fixed in 4% paraformaldehyde and processed using standard histological techniques, which included dehydration, clearing, and paraffin embedding. Sections were then stained with Hematoxylin and Eosin (H&E) for histological evaluation under a light microscope. Immunohistochemical staining Paraffin-embedded tissue sections were deparaffinized, rehydrated, and subjected to antigen retrieval. Endogenous peroxidase activity was blocked with 3% H₂O₂, followed by blocking with 5% normal goat serum. Sections were incubated overnight at 4 °C with a primary antibody against 4-HNE (ab46545, Abcam). After washing, an HRP-conjugated secondary antibody was applied, and signal was developed using DAB. Measurement of oxidative stress and inflammatory secretion Tissue homogenates were prepared in ice-cold PBS. Reduced GSH and MDA levels were measured using commercial colorimetric kits to assess redox state and lipid peroxidation, respectively. Systemic and local inflammation were evaluated by quantifying cytokine levels in serum and tissue homogenates using specific ELISA kits. Determination of bacterial colony-forming units (CFU) Bacterial counts were measured following a previously described protocol. Tissues harvested from control and CLP mice were weighed and homogenized in 1 mL of sterile 0.9% saline using a tissue homogenizer. The homogenates were left to sediment for 10 min at room temperature, after which the supernatants were collected and subjected to serial ten-fold dilutions. Subsequently, 20 μL aliquots of each dilution were spread onto tryptic soy agar plates. Following overnight incubation at 37 °C, the number of CFU was counted. Time-kill curve assay Time-kill curves were performed to evaluate the bactericidal kinetics of different formulations against E. coli (ATCC 25922) and S. aureus (ATCC 6538). Overnight bacterial cultures were diluted to an initial inoculum of approximately 5 × 105 CFU/mL. The tested groups included Con (blank control), free IC, IC@CeO₂, and IC@CeO₂@MM, with IC concentrations strictly normalized across all treatment groups by HPLC quantification. At predetermined time points (0, 1, 2, 4, 6, 8, and 24 h), aliquots were collected, serially diluted 10-fold with sterile PBS, and 100 μL of each dilution was spread onto TSB agar plates. Colonies were counted after overnight incubation at 37 °C, and the results were expressed as log₁₀ CFU/mL. All experiments were performed in triplicate with three independent biological replicates. Statistical analysis Statistical analyses were performed using GraphPad Prism 10. Unpaired t-test (two groups), or one-way ANOVA (≥3 groups) as indicated. A P-value <0.05 was considered significant. Results Preparation of IC@CeO2@MM nanoparticles To address the various pathological features of sepsis, we constructed IC@CeO₂@MM bionic nanoparticles. Fig. 1A illustrates the synthesis process of the nanoparticles. The IC@CeO₂@MM bionic nanoparticles were constructed by extracting MM and then coating them onto CeO₂ nanoparticles loaded with the antibiotic IC. Transmission electron microscopy (TEM) revealed minimal morphological changes in the nanoparticles before and after antibiotic loading, with both appearing as uniform spheres. After membrane coating, the IC@CeO₂@MM bionic nanoparticles exhibited a distinct membrane structure on their surface. Furthermore, TEM-assisted elemental mapping confirmed the presence of specific sulfur elements from CeO₂ and the cell membrane in the bionic nanoparticles (Fig. 1B-C). Dynamic light scattering (DLS) results showed that compared to CeO₂ and IC@CeO₂, the IC@CeO₂@MM bionic nanoparticles exhibited a gradual increase in particle size, while the zeta potentials of all three types of nanoparticles were negative (Fig. S1–2). We also evaluated the changes in IC@CeO₂@MM bionic nanoparticles during one week of storage. The particle size, zeta potential, and PDI of the IC@CeO₂@MM bionic nanoparticles fluctuated only slightly over the week, demonstrating their high stability (Fig. 1D). The drug encapsulation efficiency and drug loading capacity of the IC@CeO₂@MM bionic nanoparticles showed minimal changes, confirming that the nanoparticles did not leak the drug during storage (Fig. 1E and Fig. S3).Fig. 1Preparation of IC@CeO2@MM nanoparticles.(A) Schematic illustration of preparation of IC@CeO2@MM nanoparticles.(B) TEM images of different nanoparticles.(C) TEM assisted element mapping of IC@CeO2@MM nanoparticles.(D) Particle size, Zeta potential, and PDI stability of IC@CeO2@MM nanoparticles.(E) Encapsulating efficiency stability of IC@CeO2@MM nanoparticles.(F) XRD images of different nanoparticles.(G) SDS-PAGE of macrophage membrane (MM) and IC@CeO2@MM nanoparticles.(H) IC releases of IC@CeO2 (G2) and IC@CeO2@MM nanoparticles (G3).Fig. 1 Preparation of IC@CeO2@MM nanoparticles. (A) Schematic illustration of preparation of IC@CeO2@MM nanoparticles. (B) TEM images of different nanoparticles. (C) TEM assisted element mapping of IC@CeO2@MM nanoparticles. (D) Particle size, Zeta potential, and PDI stability of IC@CeO2@MM nanoparticles. (E) Encapsulating efficiency stability of IC@CeO2@MM nanoparticles. (F) XRD images of different nanoparticles. (G) SDS-PAGE of macrophage membrane (MM) and IC@CeO2@MM nanoparticles. (H) IC releases of IC@CeO2 (G2) and IC@CeO2@MM nanoparticles (G3). XRD results indicated that antibiotic loading and cell membrane coating had little effect on the crystal structure of CeO₂ (Fig. 1F). SDS-PAGE results showed that the protein composition of the IC@CeO₂@MM bionic nanoparticles was highly similar to that of the macrophage membrane, confirming that the bionic nanoparticles successfully inherited the membrane components (Fig. 1G). Finally, the drug release efficiency of the IC@CeO₂@MM bionic nanoparticles was evaluated. The IC@CeO₂ nanoparticles exhibited rapid drug release, whereas the membrane-coated IC@CeO₂@MM bionic nanoparticles showed slowed drug release (Fig. 1H), demonstrating that the membrane coating reduced antibiotic release and achieved sustained release, which is beneficial for enhancing antibacterial efficacy. In Vitro characterization of IC@CeO2@MM nanoparticles The biomimetic nanoparticles possess capabilities for ROS scavenging, polarization of macrophages toward an anti-inflammatory phenotype, antibacterial activity, and the ability to neutralize LPS via the surface macrophage membrane, thereby achieving sepsis treatment (Fig. 2A). First, through DPPH and ABTS assays, we demonstrated that the IC@CeO₂@MM biomimetic nanoparticles can effectively scavenge ROS (Fig. 2B-E). Next, we evaluated the antibacterial capacity of the IC@CeO₂@MM nanoparticles. We selected E. coli and S. aureus as model strains for Gram-negative and Gram-positive bacteria, respectively. After 12 and 24 h of treatment with different formulations, we found that both IC@CeO₂ and IC@CeO₂@MM nanoparticles significantly inhibited the growth of E. coli and S. aureus (Fig. 2F and S4–5). Furthermore, when different materials were added to an LPS solution, we observed that CeO₂ and IC@CeO₂ nanoparticles could not neutralize LPS, whereas IC@CeO₂@MM significantly reduced the LPS concentration in the system. This is attributed to the natural LPS-neutralizing receptors present on the macrophage membrane of IC@CeO₂@MM (Fig. 2G). Finally, we assessed the neutralizing effect on inflammatory cytokines. Similarly, CeO₂ and IC@CeO₂ nanoparticles failed to neutralize the inflammatory factors TNF-α and IL-6, while IC@CeO₂@MM significantly reduced the levels of these inflammatory factors (Fig. S6A). Western blot analysis confirmed that the IC@CeO₂@MM biomimetic nanoplatform successfully inherits key macrophage membrane markers (CD11b and TLR4) while maintaining membrane purity, as evidenced by the absence of nuclear (LAMINB1) and mitochondrial (TOM20) contamination. The comparable expression of these functional receptors between the purified MM vesicles and the final IC@CeO₂@MM construct demonstrates that the membrane coating process preserves the integrity and functionality of membrane proteins (Fig. S6B). Above results indicate that IC@CeO₂@MM can provide a tri-therapeutic approach for the treatment of sepsis.Fig. 2In vitro characterization of IC@CeO2@MM nanoparticles.(A) Schematic illustration of LPS neutralizaiton, anti-inflammatory, and anti-bacteiral effect of IC@CeO2@MM nanoparticles.(B, C) Scavenged DPPH. efficiency of different concentrations of IC@CeO2@MM nanoparticles.(D, E) Scavenged ABTS. + efficiency of different concentrations of IC@CeO2@MM nanoparticles.(F) E. coli growth onto LB agar paltes after different treatments after 12 h treatment.(G) Relative LPS level in different groups.Fig. 2 In vitro characterization of IC@CeO2@MM nanoparticles. (A) Schematic illustration of LPS neutralizaiton, anti-inflammatory, and anti-bacteiral effect of IC@CeO2@MM nanoparticles. (B, C) Scavenged DPPH. efficiency of different concentrations of IC@CeO2@MM nanoparticles. (D, E) Scavenged ABTS. + efficiency of different concentrations of IC@CeO2@MM nanoparticles. (F) E. coli growth onto LB agar paltes after different treatments after 12 h treatment. (G) Relative LPS level in different groups. Cellular uptake, biocompatibility, and in vitro anti-inflammatory effects of IC@CeO₂@MM CCK-8 assays indicated no significant cytotoxicity of IC@CeO₂@MM toward macrophages at tested concentrations (Fig. 3A). Fluorescence microscopy demonstrated efficient uptake of IC@CeO₂@MM by macrophages (Fig. 3B). LPS stimulation significantly increased intracellular ROS levels (Fig. 3C), protein secretion of inflammatory cytokines (Fig. 3D). In addition, DCFH-DA staining further confirmed that IC@CeO₂@MM significantly inhibited LPS-induced total ROS and mtROS generation (Fig. 3E). Regarding the M1 polarization marker iNOS, we applied both RT-qPCR and WB analysis to assess the expression alteration under different IC-based treatments, among which treatment with IC@CeO₂@MM most effectively reversed up-regulation of iNOS, outperforming free antibiotic or IC@CeO₂ groups (Fig. 3F-G). Thus, functional analysis confirmed that IC@CeO₂@MM was non-cytotoxic and efficiently internalized by macrophages, where it significantly inhibited LPS-induced ROS generation and most effectively suppressed the upregulation of the M1 marker iNOS.Fig. 3Cellular uptake, biocompatibility, and in vitro anti-inflammatory effects of IC@CeO₂@MM.(A) Cell viability of macrophages treated with various concentrations of IC@CeO₂@MM, as determined by CCK-8 assay, n = 6.(B) Fluorescence microscopy images demonstrating the efficient cellular uptake of fluorescently labeled IC@CeO₂@MM by macrophages.(C—D) Intracellular ROS levels (C) and secretion of inflammatory cytokines (D) in macrophages following LPS stimulation and treatment with different formulations, n = 6.(E) Representative images and quantification of DCFH-DA and MitoSOX Red staining showing that IC@CeO₂@MM significantly inhibits LPS-induced total ROS and mtROS generation, n = 6.(F-G) The mRNA (F) and protein (G) expression levels of the M1 macrophage marker iiNOS, analyzed by RT-qPCR and western blot, respectively. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)Fig. 3 Cellular uptake, biocompatibility, and in vitro anti-inflammatory effects of IC@CeO₂@MM. (A) Cell viability of macrophages treated with various concentrations of IC@CeO₂@MM, as determined by CCK-8 assay, n = 6. (B) Fluorescence microscopy images demonstrating the efficient cellular uptake of fluorescently labeled IC@CeO₂@MM by macrophages. (C—D) Intracellular ROS levels (C) and secretion of inflammatory cytokines (D) in macrophages following LPS stimulation and treatment with different formulations, n = 6. (E) Representative images and quantification of DCFH-DA and MitoSOX Red staining showing that IC@CeO₂@MM significantly inhibits LPS-induced total ROS and mtROS generation, n = 6. (F-G) The mRNA (F) and protein (G) expression levels of the M1 macrophage marker iiNOS, analyzed by RT-qPCR and western blot, respectively. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Therapeutic efficacy of IC@CeO₂@MM in a murine sepsis model In the CLP-induced septic mouse model, treatment with IC@CeO₂@MM significantly improved the 24-h survival rate compared to the CLP, free antibiotic, and IC@CeO₂ groups (Fig. 4A). Given CLP inflicted major organ dysfunction to various extent, IC@CeO₂@MM, we accessed serum creatinine and blood urea nitrogen to gauge renal injury and AST and ALT levels for hepatic injury. IC@CeO₂@MM significantly lowered serum creatinine and blood urea nitrogen (Fig. 4B) as well as AST and ALT levels, suggesting protection of parenchymal organs (Fig. 4C). Furthermore, serum analysis showed that IC@CeO₂@MM most effectively reduced the levels of IL-1β, IL-6, IL-18, and TNF-α in septic mice (Fig. 4D), which was further verified by CFU analysis in collected ascitic fluid and peripheral blood (Fig. S8A&B). In the CLP-induced septic mouse model, IC@CeO₂@MM treatment significantly improved the 24-h survival rate, protected against renal and hepatic dysfunction.Fig. 4Therapeutic efficacy of IC@CeO₂@MM in a murine CLP-induced sepsis model.(A) Kaplan-Meier survival curves of mice over 24 h following CLP surgery and treatment with different formulations.(B) Serum levels of hepatic injury markers, AST and ALT, n = 6.(C) Serum levels of renal injury markers, creatinine and BUN, in the experimental groups, n = 6.(D) Serum concentrations of pro-inflammatory cytokines (IL-1β, IL-6, IL-18, TNF-α) measured by Elisa assay, n = 6. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001 vs. CLP group.Fig. 4 Therapeutic efficacy of IC@CeO₂@MM in a murine CLP-induced sepsis model. (A) Kaplan-Meier survival curves of mice over 24 h following CLP surgery and treatment with different formulations. (B) Serum levels of hepatic injury markers, AST and ALT, n = 6. (C) Serum levels of renal injury markers, creatinine and BUN, in the experimental groups, n = 6. (D) Serum concentrations of pro-inflammatory cytokines (IL-1β, IL-6, IL-18, TNF-α) measured by Elisa assay, n = 6. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001 vs. CLP group. Suppression of M1 macrophages and oxidative damage mediates the hepatoprotective effect of IC@CeO₂@MM in murine sepsis CLP mice exhibited severe liver inflammation and vacuolar degeneration, renal tubular damage, and intestinal villi destruction with inflammatory infiltration. Histopathological analysis via H&E staining demonstrated that IC@CeO₂@MM most effectively mitigated CLP-induced parenchymal liver injury, observed as reduced hepatocellular damage and inflammatory cell infiltration (Fig. 5A). Consistently, western blot analysis of whole liver lysates revealed a significant attenuation in the protein expression of the M1 polarization markers iNOS and CD86 following IC@CeO₂@MM treatment (Fig. 5B). Immunofluorescence co-staining further showed that IC@CeO₂@MM markedly decreased the infiltration of inflammatory macrophages (F4/80+CD86+) as well as neutrophils and natural killer cells (Ly6G+ and CD56+, respectively) in liver tissues (Fig. 5C-D). This anti-inflammatory effect was corroborated by a reduction in oxidative stress, evidenced by diminished dihydroethidium (DHE) fluorescence (Fig. S7A) and improved redox balance through lowered malondialdehyde (MDA) and elevated glutathione (GSH) levels (Fig. S7B—C), collectively indicating that IC@CeO₂@MM conferred comprehensive hepatic protection by suppressing inflammation, immune cell recruitment, and oxidative damage.Fig. 5IC@CeO₂@MM attenuates hepatic injury by suppressing M1 macrophage polarization and oxidative stress.(A) Representative H&E-stained liver tissue sections showing parenchymal injury and inflammatory cell infiltration.(B) Western blot analysis of iNOS and CD86 protein expression in whole liver lysates.(C—D) Representative immunofluorescence images of liver sections co-stained for macrophages (F4/80) and the M1 marker CD86 (C), or for neutrophils (Ly6G) and natural killer cells (CD56) (D). Corresponding quantification of IF intensity was normalized to Sham group. Nuclei are counterstained with DAPI (blue). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)Fig. 5 IC@CeO₂@MM attenuates hepatic injury by suppressing M1 macrophage polarization and oxidative stress. (A) Representative H&E-stained liver tissue sections showing parenchymal injury and inflammatory cell infiltration. (B) Western blot analysis of iNOS and CD86 protein expression in whole liver lysates. (C—D) Representative immunofluorescence images of liver sections co-stained for macrophages (F4/80) and the M1 marker CD86 (C), or for neutrophils (Ly6G) and natural killer cells (CD56) (D). Corresponding quantification of IF intensity was normalized to Sham group. Nuclei are counterstained with DAPI (blue). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) IC@CeO₂@MM attenuates sepsis-induced intestinal injury by preserving epithelial integrity and tight junction expression Histopathological and molecular analyses revealed that IC@CeO₂@MM conferred significant protection against CLP-induced intestinal injury. H&E staining demonstrated that IC@CeO₂@MM most effectively attenuated intestinal damage, characterized by reduced epithelial cell disruption, villus shortening, and lamina propria edema compared to other treatment groups (Fig. 6A). Assessment of key tight junction proteins, critical for maintaining the intestinal epithelial barrier, showed that IC@CeO₂@MM treatment markedly reduced the loss of Occludin and ZO-1, as evidenced by immunofluorescence and immunohistochemistry, respectively (Fig. 6B and C). Collectively, these findings indicate that IC@CeO₂@MM mitigates sepsis-induced intestinal injury by preserving the structural integrity of the intestinal epithelium and maintaining the expression of crucial tight junction proteins.Fig. 6IC@CeO₂@MM preserves intestinal epithelial integrity in sepsis.(A) Representative H&E-stained sections of intestinal tissue, demonstrating structural damage such as villus shortening and epithelial disruption.(B—C) Assessment of tight junction integrity via immunofluorescence staining for Occludin (B) and immunohistochemistry for ZO-1 (C). Corresponding quantification of IF intensity was normalized to Sham group.Fig. 6 IC@CeO₂@MM preserves intestinal epithelial integrity in sepsis. (A) Representative H&E-stained sections of intestinal tissue, demonstrating structural damage such as villus shortening and epithelial disruption. (B—C) Assessment of tight junction integrity via immunofluorescence staining for Occludin (B) and immunohistochemistry for ZO-1 (C). Corresponding quantification of IF intensity was normalized to Sham group. IC@CeO₂@MM confers renoprotection in sepsis by attenuating tubular injury Firstly, we collected the kidney lysate to assess the intra-kidney bacterial load. The spiked CFU levels was most efficiently diminished by IC@CeO₂@MM application (Fig. S8C). Moreover, time-kill curve assays were performed to compare the bactericidal kinetics of the different formulations against E. coli and S. aureus. Free IC rapidly reduced bacterial counts within the first 2–4 h, but notable regrowth was observed at 24 h, particularly for E. coli. IC@CeO₂ exhibited a sustained release profile with slower initial killing but prolonged antibacterial activity, showing reduced regrowth compared to free IC. Notably, IC@CeO₂@MM demonstrated the most potent bactericidal effect, with no evidence of bacterial regrowth (Fig. S8D). Histopathological and biochemical analyses consistently demonstrated the potent renoprotective effect of IC@CeO₂@MM in septic mice. IC@CeO₂@MM demonstrated the most potent tubular cell protection potency (Fig. S9A&B). To further investigate whether the reduced pro-inflammatory activation was accompanied by a shift toward a protective macrophage phenotype, we examined the expression of the M2-associated marker Arg1 by Western blot and the anti-inflammatory cytokine IL-10 by ELISA. As shown in Fig. S9C, IC@CeO₂@MM treatment significantly increased Arg1 protein expression in liver tissue compared to the CLP group. Similarly, serum IL-10 levels were markedly elevated in the IC@CeO₂@MM group relative to CLP controls (Fig. S9D). Co-staining with Lotus tetragonolobus lectin (LTL) and neutrophil gelatinase-associated lipocalin (NGAL) revealed that IC@CeO₂@MM most effectively attenuated CLP-induced parenchymal kidney injury, characterized by reduced damage to renal tubular epithelial cells, a finding corroborated by quantitative analysis of fluorescence intensity (Fig. 7A). Furthermore, immunohistochemistry for 4-hydroxynonenal (4-HNE) showed a marked reduction in lipid peroxidation accumulation within tubular epithelial cells (Fig. 7B&C). This alleviation of oxidative stress was further quantified by a significant decrease in malondialdehyde (MDA) and a concomitant increase in glutathione (GSH) levels in the IC@CeO₂@MM group compared to other treatments (Fig. 7D &E). Collectively, these results indicate that IC@CeO₂@MM protects against septic acute kidney injury by mitigating tubular damage, suppressing pro-inflammatory macrophage polarization, and counteracting oxidative lipid damage.Fig. 7IC@CeO₂@MM confers renoprotection by mitigating tubular injury and oxidative stress.(A) Representative confocal microscopy images of kidney sections co-stained with Lotus tetragonolobus lectin (LTL, green) for tubules and neutrophil gelatinase-associated lipocalin (NGAL, red) for injury. The fluorescence intensity of NGAL was quantified.(B) Immunohistochemical staining for 4-hydroxynonenal (4-HNE), a marker of lipid peroxidation, in renal tubular epithelial cells.(C—F) Biochemical quantification of NGAL signal intensity, n = 6 (C) and oxidative stress markers in kidney tissue: MDA levels, n = 6 (D) and GSH levels, n = 6 (E). Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001 vs. CLP group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)Fig. 7 IC@CeO₂@MM confers renoprotection by mitigating tubular injury and oxidative stress. (A) Representative confocal microscopy images of kidney sections co-stained with Lotus tetragonolobus lectin (LTL, green) for tubules and neutrophil gelatinase-associated lipocalin (NGAL, red) for injury. The fluorescence intensity of NGAL was quantified. (B) Immunohistochemical staining for 4-hydroxynonenal (4-HNE), a marker of lipid peroxidation, in renal tubular epithelial cells. (C—F) Biochemical quantification of NGAL signal intensity, n = 6 (C) and oxidative stress markers in kidney tissue: MDA levels, n = 6 (D) and GSH levels, n = 6 (E). Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001 vs. CLP group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) IC@CeO₂@MM exerts anti-inflammatory effects by maintaining mitochondrial homeostasis Mechanistic exploration revealed that LPS stimulation caused mitochondrial dysfunction in macrophages, evidenced by decreased mitochondrial membrane potential, characterized by increased JC-1 monomers (Fig. 8A). Moreover, we observed reduced mitochondrial mass, reflected by downregulated expression of mitochondrial proteins (TFAM, VDAC) and accumulated mtROS intensity (Fig. 8B and C). Treatment with IC@CeO₂@MM effectively mitigated these dysfunctions. To ascertain the spiked inflammation was mediated by mitochondrial damage, we applied mito-TEMPO to the cultured RAW264.7 cells. A differential effect in mitochondrial protection was observed between IC@CeO₂@MM and IC@CeO₂ treatments, which was evened out by mito-TEMPO application. In WB analysis gauging the mitochondrial mass, we observed the mitochondrial-targeted antioxidant mito-TEMPO recapitulated the protective effects of IC@CeO₂@MM and showed additive benefits when co-administered (Fig. 8D), which was similar in the mito-SOX intensity analysis (Fig. 8E). Mito-TEMPO treatment reduced the differential effect between IC@CeO₂@MM and IC@CeO₂ in inflammatory polarization, evidenced by reduced iNOS and CD86 expression (Fig. 8F) and inflammatory secretion (Fig. 8G). The additive effect observed when MitoTEMPO was combined with IC@CeO₂@MM suggests that while mitochondrial ROS contribute to the anti-inflammatory effect, other pathways are also involved.Fig. 8IC@CeO₂@MM maintains mitochondrial homeostasis to exert anti-inflammatory effects in macrophages.(A) Analysis of mitochondrial membrane potential (ΔΨm) using JC-1 staining (monomer form indicated in green), n = 6.(B—C) Western blot analysis of mitochondrial proteins (TFAM, VDAC) (B) and quantitative analysis of mtROS intensity using MitoSOX Red staining, n = 6 (C).(D-E) Western blot analysis of mitochondrial mass markers (D) and MitoSOX Red intensity, n = 6 (E) in macrophages co-treated with IC@CeO₂@MM and the mitochondrial antioxidant mito-TEMPO.(F-G) Western blot analysis of iNOS and CD86 expression (F) and Elisa analysis assessing secretion of inflammatory cytokines, n = 6, (G) under the same co-treatment conditions. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001 vs. CLP group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)Fig. 8 IC@CeO₂@MM maintains mitochondrial homeostasis to exert anti-inflammatory effects in macrophages. (A) Analysis of mitochondrial membrane potential (ΔΨm) using JC-1 staining (monomer form indicated in green), n = 6. (B—C) Western blot analysis of mitochondrial proteins (TFAM, VDAC) (B) and quantitative analysis of mtROS intensity using MitoSOX Red staining, n = 6 (C). (D-E) Western blot analysis of mitochondrial mass markers (D) and MitoSOX Red intensity, n = 6 (E) in macrophages co-treated with IC@CeO₂@MM and the mitochondrial antioxidant mito-TEMPO. (F-G) Western blot analysis of iNOS and CD86 expression (F) and Elisa analysis assessing secretion of inflammatory cytokines, n = 6, (G) under the same co-treatment conditions. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001 vs. CLP group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Discussion The biomimetic IC@CeO₂@MM nanoparticles integrate the functions of ROS scavenging, antibacterial activity, and macrophage membrane-mediated neutralization of LPS and inflammatory cytokines. In vitro experiments confirmed that the nanoparticles are efficiently internalized by macrophages without significant cytotoxicity and most effectively reverse LPS-induced cytokine storms and M1 polarization. In a CLP-induced septic mouse model, nanotherapy significantly improved animal survival, reduced systemic inflammation, and alleviated multi-organ damage in the liver, kidney, and intestine. Mechanistically, the nanoparticles maintain mitochondrial membrane potential and reduce mitochondrial ROS accumulation in macrophages. These results demonstrate that IC@CeO₂@MM blocks the inflammatory cascade in sepsis, an effect associated with the preservation of mitochondrial homeostasis, providing a novel strategy for synergistic antibacterial, antioxidant, and immunomodulatory therapy. In sepsis, macrophages serve as pivotal orchestrators of the immune response, exhibiting remarkable functional plasticity that critically determines the balance between tissue injury and host protection. The functional role of macrophages spans a broad spectrum—from driving damaging inflammation and impairing bacterial clearance to executing essential anti-inflammatory and metabolic adaptations for pathogen removal (Shubhra et al., 2025; Zhou et al., 2025). Understanding the signals and mechanisms that dictate these divergent macrophage functions is key to developing targeted therapies for this complex syndrome. In sepsis-related acute lung injury, neutrophils release exosomes that directly target macrophages, driving them into a pro-inflammatory M1 state and priming them for pyroptosis. Exosomal miR-30d-5p activates the NF-κB pathway in macrophages, upregulating the NLRP3 inflammasome (Jiao et al., 2021). Thus, macrophages act as active drivers of inflammation and tissue damage, transforming a neutrophil signal into amplified injury within the sepsis-damaged lung. The damage-associated protein eCIRP directly impairs the macrophage's core function of bacterial phagocytosis, by activating STAT3, which disrupts actin remodeling and membrane protrusion necessary for engulfing bacteria (Zhou et al., 2023). A specific subset of macrophages marked by CD169 plays a critical protective role by producing the anti-inflammatory cytokine IL-10. IL-10 positive macrophages are essential for controlling excessive inflammation, as their absence leads to a lethal cytokine storm and increased mortality (Yeung et al., 2023). In sepsis, macrophages rely on the TRPV4 channel to regulate their metabolic function, enhancing glucose uptake via GLUT1 to fuel glycolysis and support phagolysosome maturation. This TRPV4-driven metabolic adaptation is critical for effective bacterial clearance and for limiting associated lung injury (Orsini et al., 2024). In sepsis-induced liver injury, the receptor ASGR1 on monocytes promotes their differentiation into inflammatory macrophages through the ATF5/NF-κB pathway, driving the production of damaging cytokines. Inhibiting ASGR1 curbs this monocyte-to-macrophage differentiation, reducing the number of inflammatory macrophages and the levels of IL-1β, IL-6, and TNF-α in the liver (Shi et al., 2023). Collectively, macrophages exhibit a critical duality: macrophages act as harmful drivers of tissue damage by amplifying inflammation via pathways like NF-κB and by losing their phagocytic ability; while IL-10 positive macrophages also serve as essential protectors by producing anti-inflammatory IL-10 and adapting metabolically to clear bacteria. In CLP-induced acute kidney injury (SAKI), dysregulated macrophage function, characterized by an imbalance between pro-inflammatory (M1) and anti-inflammatory (M2) polarization as well as inflammation promotion, serves as a central driver of renal inflammation and tissue damage. Macrophage infiltration and subsequent pyroptosis contribute significantly to renal pathology. The protective effect of piplartine (PLG) is associated with its inhibition of macrophage pyroptosis, which is mediated through targeting the translocator protein (TSPO), as evidenced by the TSPO agonist RO5–4864 reversing PLG's benefits (Gong et al., 2025). In another study, cichoric acid (CA) was shown to protect against septic AKI by rebalancing macrophage polarization, shifting from a pro-inflammatory (M1) to a reparative (M2) phenotype. This effect is achieved through a dual mechanism: CA inhibits HIF-1α-driven glycolysis and disrupts the KAT2A/α-tubulin complex to suppress NLRP3 inflammasome activation in macrophages (Zhu et al., 2024). Single-cell analysis reveals that Saa3hi Ccl2hi monocyte-derived infiltrating macrophages (IMs) are a central pro-inflammatory effector cell population in CLP-induced AKI. The macrophage-derived cytokine Saa3 plays a key role in driving inflammation and mediating crosstalk with resident renal cells, contributing to tissue injury (Peng et al., 2024). Heat shock binding protein 21 (HBP21) attenuates sepsis-induced AKI by promoting a protective shift in macrophage polarization from the pro-inflammatory M1 phenotype to the anti-inflammatory M2 phenotype via activation of the PI3K/AKT signaling pathway. This HBP21-driven M2 polarization in macrophages directly protects renal tubular epithelial cells, significantly reducing their apoptosis and improving their viability in a coculture model (An et al., 2025). Therapeutic strategies that target these mechanisms—by modulating metabolic pathways (e.g., HIF-1α/glycolysis), promoting M2 polarization via signals like PI3K/AKT, or directly inhibiting inflammasome activity and pyroptosis—effectively restore macrophage homeostasis and alleviate renal pathology in CLP-AKI. IC@CeO₂@MM alleviates CLP-induced acute kidney injury through a multi-targeted mechanism focused on reprogramming renal macrophage function. IC@CeO₂@MM effectively shifts macrophages from a pro-inflammatory to a protective phenotype, evidenced by reduced expression of iNOS and CD86. Furthermore, IC@CeO₂@MM mitigates tissue damage by suppressing key injury pathways: reducing oxidative stress, inhibiting mitochondrial ROS, and potentially modulating inflammasome-related pathways. This combined action—modulating polarization, quenching oxidative stress, and dampening inflammation—synergistically protects renal tubular cells and restores tissue homeostasis. Hepatic macrophages serve as central effectors in the pathogenesis of CLP-induced acute liver injury, where their recruitment, polarization state, and inflammatory output are critically regulated by multiple signaling pathways and tissue-derived signals. In CLP-induced acute liver injury, hepatocyte-derived HMGB1 drives macrophage recruitment and activation, exacerbating inflammation. HSPA12A in hepatocytes suppresses glycolysis and lactate production, thereby reducing HMGB1 lactylation and its subsequent secretion (Du et al., 2023). Pregnane X receptor (PXR) activation in macrophages promotes a protective shift from pro-inflammatory M1 to anti-inflammatory M2 polarization, which is essential for alleviating endotoxemic acute liver injury. This beneficial effect is mediated through the STAT6 signaling pathway, as the interaction between PXR and p-STAT6 drives the M2 polarization program (Zhao et al., 2024). In another study, the mitochondrial protein TSPO plays a critical role in exacerbating CLP-ALI by driving the pro-inflammatory M1 polarization of liver macrophages, thereby amplifying the local inflammatory cascade and tissue damage. The protective effect of Koumine (KM) against liver injury is mediated through its action as a TSPO ligand, which modulates macrophage polarization by inhibiting the M1 phenotype and promoting the reparative M2 phenotype (Jin et al., 2022). Maresin 1 (MaR1) protects the liver by directly modulating macrophage inflammatory responses, shifting their polarization from a damaging M1 phenotype toward a reparative M2 state. This protective effect is achieved by suppressing the MAPK/NF-κB signaling pathway and inhibiting the activation of the NLRP3 inflammasome, thereby reducing macrophage pyroptosis and the release of pro-inflammatory cytokines like IL-1β (Yang et al., 2022). Our data show that IC@CeO₂@MM effectively inhibits the expression of the M1 macrophage markers iNOS and CD86 in the liver, suggesting a shift away from the pro-inflammatory state driven by pathways like TSPO, etc. Furthermore, IC@CeO₂@MM significantly reduces oxidative stress markers (e.g., MDA) and elevates antioxidants (e.g., GSH) in liver tissue. This potent antioxidative capacity, attributable to the CeO₂ component, may influence metabolic pathways such as glycolysis, thus reducing a key driver of macrophage recruitment. Collectively, IC@CeO₂@MM confers hepatoprotection by synergistically modulating macrophage polarization toward an M2 phenotype and scavenging ROS to disrupt the oxidative stress -inflammasome axis. This dual action mitigates both the inflammatory drive, ultimately preserving hepatocyte function and reducing histopathological damage, as evidenced by the lowered serum AST/ALT levels and improved liver morphology in our study. In a complementary study, Zhuang et al. developed a self-assembled metabolic regulator (LDO) that concurrently targets glycolysis and the STING pathway to reprogram macrophages, attenuate CCL2-driven cytokine storms, and improve survival in sepsis models (Zhuang et al., 2025). While LDO acts upstream by inhibiting glycolytic flux, our IC@CeO₂@MM directly scavenges mtROS and preserves mitochondrial membrane potential. Together, these studies suggest that mitochondrial protection and metabolic reprogramming represent two convergent but distinct mechanistic nodes for macrophage immunomodulation in sepsis. We acknowledge that extended stability and safety characterization—including serum stability, protein corona formation, hemolysis, complement activation, endotoxin contamination, and batch reproducibility—are essential for the translational development of IC@CeO₂@MM. While these parameters are beyond the scope of the current mechanistic study, they will be addressed in detail in our future translational research, where we will also perform structural optimization of this nanoparticle. Conclusion This study developed an innovative biomimetic nanoparticle system (IC@CeO₂@MM) that synergistically combines antibacterial activity, reactive oxygen species scavenging, and immunomodulation for the treatment of sepsis. The system demonstrated significant efficacy in neutralizing pathogens and inflammation in vitro, improved survival and organ protection in septic mice, and was found to exert its effects by contribution of mitochondrial homeostasis maintenance. These findings present a novel, multi-targeted therapeutic strategy that addresses the complex pathophysiology of sepsis. Clinical trial number Not applicable. CRediT authorship contribution statement Xiaolin Tan: Writing – original draft. Zaili Zhang: Writing – original draft, Investigation, Formal analysis. Mingzhe Wu: Writing – original draft. Yuan Tan: Writing – original draft. Xue Zhang: Investigation, Formal analysis. Chang Li: Investigation, Formal analysis. Zhifeng Wen: Writing – review & editing, Supervision, Project administration. Yanxia Lin: Writing – review & editing, Supervision, Project administration. Jun Chen: Writing – review & editing, Supervision, Project administration. Yao Li: Writing – review & editing, Supervision, Project administration. Patient consent statement Not applicable. Ethics approval and consent to participate Not applicable. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Declaration of competing interest Not applicable.