ABSTRACT
Uncontrolled inflammation has been recognized as a vital driver of the pathology of diverse forms of tissue injury, but current anti‐inflammatory drugs have unsatisfactory efficacy and safety issues. Thus, it is important to explore its underlying mechanism and develop more specific and safe therapies. Here, inspired by the intrinsic immune modulatory role of efferocytosis, we report that bioengineered efferocytic cell‐derive nanovesicles can spatiotemporally regulate neutrophil‐macrophage (Neu‐Mac) crosstalk to promote inflammation resolution and tissue repair. We revealed that dysregulated Neu‐Mac interactions drive inflammation in both acute and chronic tissue injury states. As a conserved innate immune response, the efferocytosis process can induce a robust immunoregulatory phenotype in Mac, and efferocytic Mac‐derived artificial nanovesicles (ACDVEffero) can suppress hyperactivated Neus and promote their apoptosis for the subsequent MacEffero. Engineering ACDVEffero with probiotic‐derived membrane contents (BNVEffero) can further increase its active Neu‐targeting capacity, and these BNVEffero can hijack apoptotic Neus to transfer therapeutic payloads into Macs to increase efferocytosis, thereby driving a positive loop for promoting inflammation resolution and tissue repair. This study highlights that bioinspired nanotherapeutics can tightly and spatiotemporally control the crosstalk between different immune cell types, which provides a promising therapeutic avenue for treating diverse inflammatory disorders.
Keywords: drug delivery, efferocytosis, inflammation, macrophage, nanovesicle, neutrophil
Efferocytosis can induce Mac toward a robust immunoregulatory phenotype (A), and ACDVEffero from these MacEffero can suppress hyperactivated Neu and promote their apoptosis by delivering multiple immunoregulatory molecules (e.g., Acod1, Zfp26 and Ier3) (A), thereby broadly inhibiting proinflammatory signalling pathways (e.g., NF‒κB, TNF and cytokine‐receptor signalling) of Neus (B). Bioengineering ACDVEffero with probiotic‐derived membrane components (BNVEffero) increases their active targeting of Neu (B), enabling hijacking of apoptotic Neu to transfer therapeutic payloads into Mac and increase efferocytosis (C). This creates a positive feedback loop that accelerates inflammation resolution and supports tissue repair (D).

1. Introduction
Notably, uncontrolled acute or chronic inflammatory responses are recognized as vital drivers of the pathology of diverse forms of tissue injury (Pahwa et al. 2023; Leuti et al. 2020), ultimately leading to tissue dysfunction and adverse clinical outcomes. For example, uncontrolled acute inflammation is estimated to contribute to more than 11 million deaths per year worldwide, and chronic inflammatory disorders can account for up to ∼15% of global disability‐adjusted life years (DALYs), causing a substantial burden of morbidity and mortality worldwide (Singer et al. 2016; Kelly et al. 2021). However, the clinical management of uncontrolled inflammation still faces challenges because of the complicated signalling networks that involve diverse immune effector cells and their dynamic interactions (Karagiannis and Arnold 2022). Currently, the main anti‐inflammatory drugs (e.g., glucocorticoids) used in the clinic are still unsatisfactory because of multiple concerns. For example, glucocorticoids can significantly increase the risk of infections and serious systemic side effects (e.g., osteoporosis and peptic ulcer disease) (Harirforoosh et al. 2013). The therapeutic efficacy of current single factor‐targeting strategies (e.g., single cytokine blockade) is also limited because of the complex spatial and temporal crosstalk between different immune cell subsets during inflammation. Thus, it is important to explore the detailed mechanism and develop specific therapies for treating uncontrolled inflammation.
The innate immune system plays a critical role in regulating host defence, tissue injury and repair processes, with neutrophils (Neu) and macrophages (Mac) serving as major effector cells in these processes (Rosales 2020; Wang et al. 2020). As the first line of response to infection or tissue injury, Neu cells can rapidly reach injury sites and interact with many other immune cells (e.g., monocytes and macrophages) to drive inflammatory responses with context‐specific outcomes. Under controlled conditions (e.g., mild or transient injury), the activated Neus that fulfil their role promptly undergo apoptosis and are then recognized, engulfed and digested by local Macs (Mac efferocytosis, MacEffero), which is a critical immune action that drives inflammation resolution and tissue repair after injury (Doran et al. 2020; Gerlach et al. 2021). The engulfment of dying cells by Mac can lead to the acquisition of excess cellular materials (e.g., lipids, carbohydrates, proteins and nucleic acids), which require efficient intracellular degradation and efflux‐related pathways to process metabolic overload (Gerlach et al. 2021; Lin et al. 2026). Therefore, proper metabolic regulation of Mac cells is required for sustained efferocytosis and its related inflammation resolution and tissue repair processes. However, under uncontrolled conditions (e.g., severe or persistent injury), dysregulated Neu cells exhibit a hyperactivated phenotype and defects in apoptosis, which can amplify inflammation and tissue injury via the release of excessive amounts of proinflammatory substrates (e.g., cytokines and ROS) and the induction of proinflammatory (M1‐like) Mac rather than MacEffero (Doran et al. 2020; Gerlach et al. 2021). Therefore, precise regulation of Neu‐Mac crosstalk may be an advanced strategy for resolving inflammation.
Here, inspired by the intrinsic immune‐modulating role of MacEffero, we report that bioengineered efferocytic nanovesicles can spatiotemporally regulate Neu‐Mac communication to promote inflammation resolution and tissue repair. We revealed that dysregulated Neu‒Mac interactions drive both acute and chronic inflammatory injury. Efferocytosis can induce an unconventional immunoregulatory Mac phenotype, and efferocytic Mac‐derived artificial nanovesicles (ACDVEffero) can suppress Neu overactivation and promote their apoptosis for the subsequent MacEffero. Engineering ACDVEffero with probiotic‐derived membrane contents (BNVEffero) can further increase its active Neu‐targeting capacity, and these BNVEffero can hijack apoptotic Neus to transfer therapeutic payloads into macrophages to increase efferocytosis, thereby driving a positive loop for promoting inflammation resolution and tissue repair. This study highlights that bioinspired nanotherapeutics can tightly and spatiotemporally control crosstalk between different immune cell types, providing a promising therapeutic avenue for treating diverse inflammatory disorders.
2. Materials and Methods
2.1. Single‐Cell RNA‐seq Data Analysis
Single‐cell RNA sequencing data were retrieved from the kidneys of I/R injury model mice (accession number: GSE174324). Data preprocessing, quality control, normalization and dimensionality reduction were performed via the Seurat package (version 4.3.0). Cell clustering was conducted to identify distinct cell populations, which were subsequently annotated on the basis of canonical marker gene expression, via UMAP. Cell‒cell interactions were analysed via CellChat tools.
2.2. Animal Model Establishment
All animal experiments were performed following the Guidelines of the National Institutes of Health on the Use of Laboratory Animals and were approved by the Animal Care and Use Committee of West China Hospital, Sichuan University (Permit NO. 20241230003). Male C57BL/6 and BALB/c mice (∼20–25 g) were purchased from Ensiweier (Chengdu, China) and housed in a special pathogen‐free facility. The ALI mouse model was established via intraperitoneal (i.p.) injection of LPS (15 mg/kg, Sigma‒Aldrich, St. Louis, USA). A CKD mouse model was established via sodium pentobarbital anaesthesia, and UUO was induced via ligation of the left ureter with 7–0 silk through a left flank incision.
2.3. Histological Examination
Mouse tissues fixed in 4% paraformaldehyde were embedded in paraffin and sliced into 4 µm thick sections for immunohistochemistry (IHC) and haematoxylin and eosin (H&E) staining. For IHC, tissue sections were deparaffinized in xylene and rehydrated in graded ethanol, and the antigens were retrieved with citrate buffer. After inactivation of endogenous peroxidase with 3% H2O2, the sections were blocked with 1% bovine serum albumin (BSA), incubated with diluted rabbit anti‐Ly6G (Abcam, UK), anti‐F4/80 (CST, USA), or anti‐MPO (Abcam) primary antibodies overnight at 4°C, and then stained with horseradish peroxidase (HRP)‐conjugated secondary antibodies (ABClonal, Wuhan, China) and 3,3'‐diaminobenzidine (DAB) substrate. Images of the stained sections were captured with Pannoramic SCANII (3D HISTECH, Budapest, Hungary) and quantified via NIH ImageJ software (NIH, Bethesda, USA).
2.4. Immunofluorescence Staining
After fixation (4% paraformaldehyde for 10 min), permeabilization (0.3% Triton X‐100 for 10 min), and blocking (1% BSA for 1 h), the cells or sections were incubated with anti‐Ly6G (Abcam), anti‐CD68 (Abcam), anti‐IL‐1β (ABClonal) and anti‐ICAM‐1 (Proteintech, Wuhan, China) antibodies overnight at 4°C. After being washed with PBS, the slides were incubated with fluorescein isothiocyanate (FITC)‐ or rhodamine‐conjugated secondary antibodies (Life Technologies, CA, USA) at 37°C for 1 h. The nuclei were visualized by staining with DAPI (Beyotime Biotechnology, Shanghai, China). Digital images were captured via a confocal laser scanning microscope (Nikon).
2.5. Cell Culture
Mouse macrophage lines (RAW264.7) were cultured in RPMI 1640 (Gibco, CA, USA) supplemented with 10% (v/v) FBS (ExCell Bio, Shanghai, China) and 1% penicillin‒streptomycin (Beyotime Biotechnology). Jurkat T‐lymphocyte cells were cultured in RPMI 1640 supplemented with 10% (v/v) FBS and 1% penicillin‒streptomycin. Bone marrow‐derived primary Neu cells were collected from male C57BL/6 mice (25 g) via a mouse neutrophil isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's specifications. The isolated cells were cultured in RPMI 1640 medium supplemented with 10% FBS and 1% penicillin‒streptomycin.
2.6. Establishment of MacEffero In Vitro
ACs were prepared from Jurkat cells suspended at a density of 3 × 106 cells/mL and treated with STS (500 nM, MedChem Express, Shanghai, China) for 4 h at 37°C. The degree of apoptosis was determined via an Annexin V/PI kit (Oriscience, Guangzhou, China) according to the product manual. Following induction, the ACs were collected by gentle centrifugation, washed twice with prewarmed phosphate‐buffered saline (PBS), and labelled with DiD dye (5 µM, Sigma–Aldrich). DiD‐labelled ACs were immediately cocultured with RAW264.7 Mac cells at a 1:1 cell ratio in complete RPMI 1640 medium supplemented with 10% FBS and 1% penicillin‐streptomycin for 4 h at 37°C. After 4 h of incubation, the supernatants were discarded, and the cells were washed twice with PBS. The attached cells were collected and subjected to FCA analysis via a flow cytometer (LSRFortessaTM, BD Biosciences, San Jose, USA).
2.7. RNA Sequencing (RNA‐seq) of Cells
MacCon and MacEffero cells were collected for RNA sequencing analysis. In brief, total RNA was extracted from samples via TRIzol, and genomic DNA was removed via DNase I (Takara, Shiga, Japan). The RNA‐seq transcriptome library was constructed via a TruSeq RNA Sample Preparation Kit (Illumina, San Diego, CA, USA), and the paired‐end RNA‐seq library was sequenced on an Illumina HiSeq Xten/NovaSeq 6000 sequencer by Shanghai Majorbio Biopharm Technology Co., Ltd. (Shanghai, China). The expression levels of the transcripts were calculated via the transcripts per million reads (TPM) method. DEGs (|FC| > 1.5 and p‐adjusted < 0.05) were considered significant. For mechanistic investigations, LPS‐activated Neus were treated with ACDVEffero (20 µg/mL, 4 h) and then subjected to RNA‐seq. DEGs (|FC| > 2 and p‐adjusted <0.05) were considered significant. Venn diagrams, principal component analysis (PCA) plots, volcano plots and heatmaps were generated via a free online platform (https://www.omicsolution.org/wkomics/main/). GO and KEGG analyses of the DEGs were performed with DAVID Bioinformatics Resources (https://david.ncifcrf.gov/home.jsp), and the results were visualized on a free online platform (https://www.omicsolution.org/wkomics/main/).
2.8. Quantitative Real‐Time Polymerase Chain Reaction (qPCR)
Total RNA was extracted from tissues or cells via TRIzol reagent (Gibco) and reverse transcribed into cDNA via a commercial kit (Vazyme, Nanjing, China). qPCR was carried out on a CFX96 real‐time PCR detection system (Bio‐Rad, Hercules, CA, USA) using SYBR Green (Vazyme, Nanjing, China). The primer sequences are available in Table S1. Rps18 or β‐actin served as internal controls.
2.9. Western Blotting Analysis
Total protein was extracted via radioimmunoprecipitation assay (RIPA) buffer (Beyotime) supplemented with protease inhibitors (Calbiochem, San Diego, USA). The protein concentration was determined via a BCA protein assay kit (Thermo Fisher Scientific, Carlsbad, USA). Equal amounts of cellular protein were electrophoresed via sodium dodecyl sulphate‒polyacrylamide gel electrophoresis (SDS‒PAGE) and then transferred to nitrocellulose membranes (NC, Cytiva, Wilmington, USA). The membranes were blocked in western blocking buffer (P0252, Beyotime) and incubated with anti‐ARG1 (Proteintech) and anti‐Tubulin (ABClonal) primary antibodies at 4°C overnight. Afterward, the membranes were washed and incubated with HRP‐conjugated secondary antibodies at 37°C for 1 h. The protein bands were detected via a chemiluminescence kit (MedChem Express, Shanghai, China) and analysed via NIH ImageJ software.
2.10. Mitochondrial Respiratory Capacity Assay
The bioenergetic profile of Mac cells was assessed using a Seahorse XF Flux Analyzer (Seahorse Biosciences, Agilent, USA) according to the manufacturer's instructions. Mac cells were seeded at a density of 10,000 cells per well (for the XF96 Pro analyser) in cell culture microplates. After 1 h of treatment with apoptotic T cells or Neu cells, the culture medium of Mac cells was replaced with complete Seahorse XF DMEM. Following equilibration, the cells were sequentially treated with oligomycin (5 µM), FCCP (1 µM) and rotenone‐antimycin (1 µM) to measure the oxygen consumption rate (OCR). On the day of the assay, each well on the plate was verified to be subconfluent with comparable cell density, and the results were normalized to the cell number. The data were analysed using Agilent Seahorse XF Wave Pro software (version 10).
2.11. Cellular Metabolic Assessments
To determine the cellular lipid contents, the cells were stained with 1 µg/mL BODIPY 493/503 (4,4‐Difluoro‐1,3,5,7,8‐Pentamethyl‐4‐Bora‐3a,4a‐Diaza‐s‐Indacene; Thermo Fisher Scientific) and 1 µM BODIPY FL C16 (4,4‐Difluoro‐5,7‐Dimethyl‐4‐Bora‐3a,4a‐Diaza‐s‐Indacene‐3‐Hexadecanoic Acid; Thermo Fisher Scientific) for 30 min at 37°C. To assess cellular glucose uptake, the cells were stained with 1 µM 2‐NBDG (2‐deoxy‐2‐((7‐nitro‐2,1,3‐benzoxadiazol‐4‐yl) amino)‐D‐glucose; Cayman Chemical, Ann Arbor, USA) for 30 min at 37°C. The cells were then washed with PBS twice to remove free dye. The positive signals of the stained cells were analysed via a flow cytometer (LSRFortessa).
2.12. Macrophage Characteristic Similarity Analysis
MacIL4 cells were generated by incubating RAW264.7 cells with recombinant mouse IL‐4 protein (20 ng/mL; Novoprotein, Shanghai, China) for 24 h. Both MacIL‐4 and MacEffero cells were then subjected to RNA‐seq analysis. The similarity between the two cell populations was subsequently assessed via MacSpectrum software by calculating the MPI and AMDI indices (https://macspectrum.uconn.edu/).
2.13. Isolation of Mac‐Derived ACDVs
Mac‐derived ACDVs were prepared via a previously established method with slight modifications (Lv et al. 2024). In brief, the cells were collected at 300×g for 3 min and washed with PBS once, after which the dead cells were discarded. The cells were resuspended in PBS, disrupted via ultrasonication and centrifuged at 1000×g for 10 min to remove the nuclei. The supernatants were serially extruded through 5, 1 and 0.4 µm nanopore‐sized membrane filters (Whatman, Maidstone, UK) via a miniextruder (Avanti Polar Lipids, USA). For ultracentrifugation, extruded cell fragments were ultracentrifuged at 110,000×g for 70 min at 4°C using an SW32Ti rotor (Beckman Coulter, Brea, CA, USA). After being washed with PBS, the ACDV pellets were recentrifuged at 110,000 × g for 70 min at 4°C. The pure ACDV pellets were then resuspended in PBS and stored at −80°C until further use. The protein concentrations of the ACDVs were quantified via a BCA protein assay kit.
2.14. Transmission Electron Microscopy (TEM)
The morphology of the ACDVs or muscle tissues was observed via transmission electron microscopy (TEM) (HT7800, Hitachi, Ltd., Tokyo, Japan) as previously described (Wang et al. 2024). To observe the ACDV samples, the suspension was dropped on a carbon‐coated copper grid, followed by washing with pure water to remove unattached particles. The grid was then stained with 2% phosphotungstic acid and air‐dried before being subjected to TEM.
2.15. Nanoparticle Tracking Analysis (NTA)
The size distribution and particle number of the ACDVs were analysed via a Zeta View PMX 120 (Particle Metrix, Meerbusch, Germany) as previously described (Lou et al. 2025). In brief, ACDVs were diluted with pure water and analysed via the following parameters (max area 1000, minimum area 5 and minimum brightness 20) at 25°C.
2.16. Cellular Uptake of ACDVs
The ACDVs were labelled with Memglow dye (200 nM, Cytoskeleton, Denver, USA) following the manufacturer's instructions. The free dye was removed via ultracentrifugation. Neus isolated from bone marrow were treated with the labelled ACDVs (20 µg/mL) or an equivalent amount of control dye for 4 h at 37°C with 5% CO2. After incubation, both adherent and suspended cells were collected and washed with PBS. The positive signals of the stained cells were analysed via a flow cytometer (LSRFortessa).
2.17. Evaluating the Biological Properties of ACDVs
To evaluate the anti‐inflammatory effects, the isolated Neu was activated with LPS (100 ng/mL, Sigma‐Aldrich), treated with ACDVCon or ACDVEffero (20 µg/mL) for 4 h and then subjected to qPCR analysis. To evaluate the promotion of apoptosis, ACDVCon and ACDVEffero (20 µg/mL) were added to Neu cells, and the degree of apoptosis was detected with an Annexin V/PI kit (Oriscience) via a flow cytometer (LSRFortessaTM). To obtain a more beneficial phenotype, MacCon and MacIL4 cells were cocultured with ACs at a ratio of 1:1 for 4 h, after which the NVs were prepared as previously described. Afterward, ACDVEffero or ACDVEffero+IL4 (20 µg/mL) was added to the LPS‐induced Neu cells, and the cells were collected for qPCR analysis after 4 h of incubation.
To evaluate the effects of ACDVEffero in vivo, ACDVCon and ACDVEffero (50 µg in 100 µL per mouse) were i.p. injected into mice treated with LPS (15 mg/kg). After 4 h of treatment, the mice were euthanized under deep anaesthesia, and the lung tissues were collected and divided into three portions: paraffin embedding, frozen OCT embedding, and qPCR analysis. RNA was extracted for qPCR to assess the expression of cytokines. Paraffin sections were subjected to IHC to detect MPO and Ly6G+ Neu and F4/80+ Mac cell infiltration. H&E staining was used to observe pathological structural changes in the lung tissue.
2.18. LC‒MS/MS‐Based Proteomic Analysis of ACDVs
Total ACDV protein was extracted from the ACDV samples and quantified via a BCA kit (Thermo Scientific). One hundred micrograms of protein samples were removed for reductive alkylation, digested overnight at 37°C for reductive alkylation, and digested with trypsin overnight at 37°C. The peptides were desalted and vacuum dried, and the concentrations were determined via a peptide quantification kit (Thermo Scientific). Afterward, 30 µg of trypsin‐digested peptides were analysed via an EASY nLC‐1200 system (Thermo, USA) coupled with a timsTOF Pro2 (Bruker, Germany) mass spectrometer at Majorbio Bio‐Pharm Technology Co. Ltd. (Shanghai, China). Briefly, the tryptic peptides were separated by an ultrahigh‐performance liquid phase system and then analysed with a capillary ion source and a timsTOF Pro2 at a flow rate of 250 nL/min (the electrospray voltage was 1.5 kV). The peptide parent ions and their secondary fragments were detected and analysed via high‐resolution TOF. MS/MS spectra were searched via MaxQuant version 2.0.3.1 software.
Bioinformatic analysis of the proteomic data were performed with a free online platform (https://www.omicsolution.org/wkomics/main/). DEPs with a |FC| > 1.5 and p‐adjusted< 0.05 were identified via a free online platform (https://www.omicsolution.org/wkomics/main/) with limma analysis. PCA plots, volcano plots and heatmaps were generated via a free online platform (https://www.omicsolution.org/wkomics/main/). GO analysis of the DEPs was performed via STRING (https://cn.string‐db.org/) and the Metascape tool (v3.5), and a false discovery rate (FDR) <0.05 was considered to indicate statistical significance.
2.19. Isolation and Characterization of Probiotic‐Derived DMVs
The probiotic strain ECN 1917 cultured in lysogeny broth was collected and washed three times with prechilled PBS. ECN 1917‐derived DMVs were prepared as previously reported (Chen et al. 2023). The cells were resuspended in PBS, disrupted via ultrasonication and centrifuged at 6000 × g for 30 min to remove debris. The supernatant was ultracentrifuged at 100,000 × g for 35 min, and the pellet was subsequently washed and resuspended in PBS. The collected DMV pellets were characterized via TEM and NTA and stored at −80°C until further use. The protein concentrations were quantified via a BCA protein assay kit. To evaluate the potential effects of the DMV on Neu activation, Neu cells were stimulated with LPS and subsequently treated with the DMV. LPS‐induced Neu cells were treated with DMV (20 µg/mL) for 4 h and then subjected to qPCR analysis.
2.20. Construction and Characterization of BNVs
DMVs and ACDVs were mixed at a defined ratio to form a homogeneous suspension and ultrasonicated on ice as previously described (Rayamajhi et al. 2019). The sonicated mixture was sequentially extruded through nanopolycarbonate membranes via a miniextruder (Avanti Polar Lipids) at a constant rate. After extrusion, the suspension was centrifuged at 110,000 × g for 70 min at 4°C. The resulting pellet was subsequently washed once with PBS, resuspended, and stored at −80°C until further use. The collected BNVs were characterized via TEM, NTA, and protein quantification via a BCA protein assay kit. To quantitatively evaluate the fusion efficiency between ACDVs and DMVs, the two vesicle populations were differentially labelled with fluorescent membrane dyes. ACDVs labelled with DiD (APC) and DMVs (PE) were labelled with MemGlow as previously described. After the removal of excess free dye, the labelled ACDVs and DMVs were subjected to the membrane fusion procedure described above to generate dual‐labelled BNVs. The resulting BNVs were analysed by flow cytometry (BD FACSymphony A1), and the proportion of DiD+MemGlow+ events was used to quantify the fusion efficiency.
2.21. Evaluation of the Neu‐Targeting Efficiency of BNVs
For in vitro evaluation of uptake efficiency, BNVs were labelled with Memglow and incubated with isolated Neu at a concentration of 20 µg/mL for 1 h. The cells were then collected, and washed with PBS, after which the fluorescence was analysed via a flow cytometer (LSRFortessa). For in vivo evaluation of Neu‐targeting efficiency, BNVs were labelled with sulfo‐Cy7‐NHS ester (LiTTLE‐PA Life Sciences, Wuhan, China) according to the manufacturer's instructions, and unbound dye was removed via ultracentrifugation. Cy7‐labelled ACDVs or BNVs (100 µg/100 µL per mouse) and an equal volume of free Cy7 solution (negative control) were i.p. administered to the mice after LPS injection for 1 h. The peritoneal dialysis fluid of the mice was collected as previously described (Mao et al. 2020). The collected dialysis fluid was centrifuged at 1500 rpm for 3 min at 4°C to obtain peritoneal cells, which were then washed and resuspended in PBS. The prepared single‐cell suspensions were stained with AF700 anti‐CD45 (BD Pharmigen), APC anti‐Ly6G (BD Pharmigen), BV421 anti‐Ly6C (BD Pharmigen), and PE anti‐F4/80 antibodies (BD Pharmingen) for 30 min, washed with PBS, and analysed by flow cytometry (LSRFortessa, BD Biosciences). To directly compare the in vivo immune‐cell uptake ability of the DMV and BNV, both were labelled with Cy7 and intraperitoneally administered to LPS‐pretreated mice (100 µg/100 µL of vesicles per mouse). At 1 h after administration, peritoneal cells were collected and analysed by flow cytometry (LSRFortessa, BD Biosciences).
To observe the targeting efficiency to injured tissues, Cy7‐labelled ACDVs and BNVs (100 µg/100 µL per mouse) were i.p. injected into the mice. Four hours after injection, the mice were sacrificed by an overdose of pentobarbital sodium anaesthesia, and the injured lung tissues were collected and observed via an in vivo animal imaging system (Biolight, Guangzhou, China). For IF staining, lung tissue sections were stained with anti‐Ly6G (Abcam) to identify Neu. The stained tissue sections were observed via a SpinSR spinning disk confocal superresolution microscope (Olympus) and analysed via OlyVIA 3.3 software (Olympus).
2.22. In Vitro Drug Loading and Release Test
Rosiglitazone maleate (RSG; MedChemExpress, Shanghai, China) was incorporated into BNVs at a mass ratio of 2:1 and incubated at 37°C for 30 min. Unbound RSG was removed by washing and ultracentrifugation (110,000 × g, 70 min). The resulting purified RSG‐loaded BNVs (RSG@BNVs) were resuspended in PBS. The quantification of RSG loading was performed via a SCIEX ExionLC UHPLC system (Waters, USA) equipped with an ACQUITY UPLC HSS C18 column (100 × 2.1 mm, 1.8 µm; Waters) maintained at 40°C. The mobile phases were 100% H2O with 0.1% formic acid (A) and 100% acetonitrile (B). The drug loading efficiency was calculated as follows: drug loading efficiency (%) = 100% × (total RSG − RSG in the supernatant)/total RSG. For the drug release assay, RSG@BNVs were added to a dialysis unit (cutoff MW of 3.5 kDa, Thermo Fisher Scientific) and incubated with release PBS buffer. The release buffer was collected and replaced with an equal volume of fresh buffer, and the RSG content in the release buffer was analysed.
2.23. Drug Transfer Assay and Functional Evaluation
Fluorescent curcumin (Cur; MedChemExpress) was used as a model drug to simulate the transfer process. Cur@BNVs were prepared as RSG@BNVs and applied to Neu cells at a final concentration of 20 µg/mL for 1 h. After incubation, the Neu cells were collected and washed once. The collected Neu cells were then examined under a light microscope and analysed for fluorescence signals via flow cytometry (LSRFortessa). Cur@BNV (20 µg/mL) was subsequently applied to apoptotic Neu cells, which were induced with 250 nM STS for 1 h. The treated cells were collected and designated NeuCon (control) or NeuCur@BNV. Equal numbers of NeuCon and NeuCur@BNV cells were then coincubated with Mac cells for 1 h. Following coincubation, Mac cells were collected and washed once. The collected Mac cells were then examined under a light microscope and analysed for fluorescence signals via flow cytometry (LSRFortessa).
To directly evaluate the efferocytosis‐mediated transfer of RSG from apoptotic Neu cells to Mac cells, apoptotic Neu cells were incubated with BNVs or RSG@BNVs for 1 h. The Neu cells were then collected and washed with PBS to remove free vesicles and extracellular RSG. The resulting NeuBNV and NeuRSG@BNV were subsequently cocultured with Mac for 1 h to allow efferocytosis. After coculture, Mac cells were collected, washed, and lysed for RSG extraction and LC–MS/MS analysis. To assess the effect of the delivered RSG on Mac cells, inflammatory Mac cells induced with 40 ng/mL LPS were treated with 10 µM RSG for 4 h. The efferocytotic capacity of these cells was determined as previously described, and inflammatory cytokine levels were measured by qPCR.
2.24. Mouse Tissue Injury Models and Treatments
For the ALI mouse model, the treatment groups received an i.p. injection of BNVEffero or RSG@BNVEffero (50 µg in 100 µL per mouse). The control (Con) and ALI groups received 100 µL of PBS, whereas the RSG group was given free RSG in an amount equivalent to the drug loading in RSG@BNV. Additionally, to evaluate the therapeutic efficacy of RSG@DMV in vivo, RSG@DMV was prepared using the same procedure as described for RSG@BNV. ALI mice were then i.p. administered RSG@DMV (50 µg in 100 µL per mouse). After 4 h of treatment, the mice were euthanized under deep anaesthesia, and the lung tissues were collected and divided into three portions: paraffin embedding, frozen OCT embedding, and qPCR analysis. RNA was extracted for qPCR to assess the expression of tnf‐α, mpo and cxcr2. Paraffin sections were subjected to IHC to detect MPO and Ly6G+ Neu and F4/80+ Mac cell infiltration. IF staining was performed to assess IL‐1β and ICAM‐1 expression, as well as the colocalization of Ly6G+ Neu and CD68+ Mac cells. TUNEL staining was performed to evaluate apoptosis, and H&E staining was used to observe pathological structural changes in the lung tissue.
For the mouse CKD model, the treatment groups received i.p. injections of 100 µL of RSG@BNVEffero (50 µg in 100 µL per mouse), whereas the control (sham) and CKD groups were administered 100 µL of PBS, with injections given every other day. After 3 days, the mice were euthanized under deep anaesthesia, and the left kidney was harvested and divided into three portions for paraffin embedding, frozen OCT embedding, and qPCR analysis. RNA was extracted from kidney tissue to determine the expression levels of proinflammatory factors (tnf‐α, il‐1β, mpo and cxcr2) via qPCR. Paraffin sections were subjected to IHC staining to detect MPO expression and infiltration of inflammatory cells (Ly6G+ Neu and F4/80+ Mac). IF staining was performed to assess IL‐1β and ICAM‐1 expression, as well as the colocalization of Ly6G+ Neu and CD68+ Mac cells. TUNEL staining was used to evaluate the degree of apoptosis in the injured kidney. Sirius Red staining was performed to examine collagen alterations in kidney tissue.
2.25. Statistical Analysis
All the data are presented as the means ± standard deviations (SDs) and were analysed via GraphPad software (version 8.0.2; IBM Corporation, USA) with t‐tests (for two‐group comparisons) or one‐way analysis of variance (ANOVA) (for more than two group comparisons), and p < 0.05 was considered to indicate a significant difference. The data were obtained from at least three biological replicates, and “n” represents the number of independent samples for each group.
3. Results
3.1. Dysregulated Neu–Mac Interactions Drive Inflammation in Diverse Forms of Tissue Injury
Upon tissue injury, immune cell subsets and their interactions can dynamically change in different phases and/or conditions, and such intercellular crosstalk can affect the outcomes of whether injured tissue progresses toward more severe damage or repair. To explore the possible changes in immune cell crosstalk, we analysed single‐cell RNA sequencing (scRNA‐seq) data from kidneys with progressed unilateral renal ischemia reperfusion injury (uIRI) (Yao et al. 2022) (GSE174324; Figure 1A). After rigorous quality control, dimensionality reduction, and uniform manifold approximation and projection (UMAP) clustering, renal cells were classified into several major cell populations, including macrophages (Mac), neutrophils (Neu), NK cells, tubular epithelial cells (TECs), collecting duct cells, endothelial cells and fibroblasts (Figure 1A, B). The proportions of Neu and Mac cells increased following injury (Figure 1C), indicating their prominent involvement in the inflammatory response. Cell‐cell communication analysis further revealed increased overall intercellular communication strength (Figure 1D, E). Further comparison of outgoing signalling from Neu cells to all recipient cell populations revealed that compared with other cell‐cell interactions (such as Neu‐NK cells and Neu‐endothelial cells), Neu‐Mac communication was relatively stronger on both Day 1 and Day 3 after injury (Figure 1F). In addition, further analysis showed increased numbers of pro‐inflammatory N1‐like Neu and M1‐like Mac following injury (Figures 1G, H and S1A, B). Consistently, KEGG pathway enrichment analysis of the upregulated differentially expressed genes (up‐DEGs) in these populations revealed significant enrichment of inflammation‐related signalling pathways, including the TNF, IL‐17 and NF‐κB signalling pathways (Figure 1I, J). Together, these findings indicate increased infiltration of Neu and Mac with predominantly pro‐inflammatory phenotypes during the early phase of injury, and further highlight prominent Neu‐Mac communication as a potentially important component of the inflammatory response following AKI.
FIGURE 1.

Evaluating the effects of Neu‐Mac crosstalk on uncontrolled inflammation. (A) Schematic illustration of the experimental design for the uIRI mouse model. (B) UMAP visualization of the major cell populations in kidney tissues. (C) Quantitative analysis of changes in major cell populations at different time points after uIRI. (D) CellChat analysis of intercellular communication among major cell populations in kidney tissues after uIRI. (E) Quantification of the overall interaction strength among major cell populations. (F) Relative interaction strength of outgoing signalling from Neu to different recipient cell populations. (G) Quantification of pro‐inflammatory N1‐like Neu after uIRI. (H) Quantification of pro‐inflammatory M1‐like Mac after uIRI. (I) KEGG pathway enrichment analysis of up‐DEGs in N1‐like Neu. (J) KEGG pathway enrichment analysis of up‐DEGs in M1‐like Mac. (K) Schematic diagram of ALI mice. (L) Representative IF images of the lung tissue from ALI mice showing the colocalization of CD68+ (green), Ly6G+ (red), and DAPI (blue) (right panel) (scale bar = 10 µm). (M) Schematic diagram of CKD mice. (N) Representative IF images of kidney tissue from CKD mice showing fluorescence colocalization of CD68+ (green), Ly6G+ (red), and DAPI (blue) (right panel) (scale bar = 10 µm).
To further substantiate these findings, we analysed the transcriptomic profile of M1 macrophages and revealed that multiple pathways, such as those involved in neutrophil degranulation and the innate immune system, were enriched in these cells (Figure S1C), which again suggests that Macs can interact with Neus to modulate local inflammation. We next detected changes in Neu‑Mac interactions in acute (acute lung injury, ALI) and chronic (chronic kidney disease, CKD) inflammatory disease models. In line with the multiomics data, increased numbers of infiltrated Ly6G+ Neu and F4/80+ Mac cells were observed in the lungs and kidneys of ALI and CKD mice, respectively (Figure S1D, E). Moreover, the spatial colocalization of Ly6G+ Neu and CD68+ Mac cells was greater in the injured lungs or injured kidneys than in the controls (Figure 1K–N). Consistently, dysregulated Neu‐Mac crosstalk has been reported in many inflammatory diseases, such as ALI, ischemic stroke and renal fibrosis (Doran et al. 2020; Mehrotra and Ravichandran 2022). Under these conditions, hyperactivated N1‑like Neu cells at injury sites can recruit and activate M1‑like Mac cells through the release of various proinflammatory mediators, and M1‑like Mac cells promote further Neu influx and activation, thereby creating a vicious cycle that amplifies inflammatory cascades and exacerbates tissue injury. Together, these results indicate that dysregulated Neu‐Mac communication might drive inflammatory tissue injury and that targeted modulation of this crosstalk is a potential therapeutic strategy for resolving inflammation.
3.2. Efferocytosis Itself Can Induce an Unconventional Immunoregulatory Phenotype in Mac
Having revealed the critical role of dysregulated Neu‐Mac communication in driving inflammatory tissue injury, we explored specific therapies for modulating this crosstalk. Notably, Mac plays a vital role in regulating inflammation resolution and tissue repair in vivo via a conserved efferocytosis process. In injured tissues, Macs can recognize dying tissue cells and immune cells (e.g., neutrophils) via surface markers (e.g., phosphatidylserine, PS), followed by binding and engulfment via efferocytic receptors (e.g., the TIM family, MerTK) (Greenlee‐Wacker 2016). The engulfment of apoptotic cells leads to the acquisition of excess cellular materials, which activate catabolic pathways of Mac to manage this metabolic burden. Efficient efferocytosis of Mac can trigger a coordinated resolution program that involves the secretion of a variety of mediators (e.g., IL‐10, TGF‐β and polyamines) to promote inflammation resolution and tissue repair (Doran et al. 2020; Gerlach et al. 2021).
Inspired by this intrinsic and robust immune‐modulating role, we speculated that efferocytotic Macs might serve as a source of cells for therapeutic applications. To this end, we induced efferocytotic Macs (MacEffero) in vitro (Figure S2A). Briefly, apoptotic cells (ACs) were induced with staurosporine (STS) and then identified via Annexin V/propidium iodide (PI) staining as previously reported (Wang et al. 2024) (Figure S2B). Annexin V/PI staining revealed that 96.1% of the STS‐treated cells were apoptotic (Figure S2C). The ACs were then labelled with dye (DiD) and added to the Mac, and flow cytometry assay (FCA) revealed that these ACs could be efficiently taken up by the macrophages after 4 h (Figure S2D, E), indicating the successful establishment of MacEffero. Next, we globally profiled the phenotype and potential functions of MacEffero via RNA‐seq (Figure 2A). Compared with native Mac cells without AC engulfment, MacEffero cells displayed distinct gene expression patterns (MacCon) (Figure S3A, B). In line with previous reports (Gerlach et al. 2021; Mehrotra and Ravichandran 2022), MacEffero highly expressed many key genes related to immunoregulation, such as arg1, havcr2, cd47 and lgals3 (Figure 2B). Further assays confirmed that the pro‐resolving and tissue repair‐associated genes (e.g., arg1, il‐10 and tgf‐β) were significantly upregulated in MacEffero compared with MacCon (Figure 2C, D). These findings further support the pro‐resolving phenotype of MacEffero.
FIGURE 2.

Evaluating the potential immunoregulatory effects of MacEffero. (A) Schematic diagram of the cell molecular investigation. (B) Heatmap of the immunoregulatory DEGs between MacCon and MacEffero (n = 3). (C) qPCR analysis of canonical M2 marker gene expression in MacEffero (n = 5). (D) Comparison of the M2 marker protein content between MacCon and MacEffero (n = 3). (E) KEGG pathway enrichment analysis of the up‐DEGs in MacEffero compared with MacCon. (F) Schematic diagram of the investigation of metabolic changes. (G) Seahorse analysis of OCR levels in MacCon and MacEffero cells (n = 5). (H) Quantitative analysis of mitochondrial respiratory parameters, including basal respiration, ATP‐linked respiration, maximal respiration, and spare respiratory capacity (n = 5). (I) Quantitative analysis of 2‐NBDG levels in MacCon and MacEffero cells determined via flow cytometry (n = 3). (J) Quantitative analysis of BODIPY C16 and BODIPY 493/503 levels in MacCon and MacEffero cells determined via flow cytometry (n = 3). (K) Visualization of the AMDI and MPI index analysis (left panel) and the density plot of the MPI (right panel) between MacEffero and MacIL4 cells.
Metabolic reprogramming plays an important role in regulating the Mac phenotype and function (Zhang et al. 2019; Wculek et al. 2023; Lin et al. 2025). Transcriptomic analysis revealed that several metabolism‐related signalling pathways (e.g., FoxO, PI3K‐Akt and MAPK signalling) were upregulated in MacEffero (Figure 2E), suggesting metabolic remodelling following efferocytosis. We therefore further characterized the metabolic phenotype of MacEffero (Figure 2F). Seahorse analysis showed a significantly higher oxygen consumption rate (OCR) in MacEffero than in MacCon, and quantitative analysis further revealed significant increases in basal respiration, ATP‐linked respiration, maximal respiration and spare respiratory capacity (Figure 2G, H). These findings provide direct evidence of enhanced mitochondrial oxidative metabolism in MacEffero. In addition, MacEffero exhibited altered glucose and lipid metabolism, as indicated by reduced uptake of exogenous glucose (2‐NBDG) and fatty acids (BODIPY C16), and increased intracellular neutral lipid accumulation, as detected by BODIPY 493/503 staining (Figures 2I, J and S3C, D). Previous studies have similarly shown that efferocytosis induces metabolic reprogramming in phagocytes, including enhanced fatty acid oxidation, electron transport chain activity and OXPHOS, which can promote IL‐10 production and reinforce a pro‐resolving macrophage program (Schilperoort et al. 2023). Collectively, these findings indicate that efferocytosis induces metabolic reprogramming in MacEffero characterized by enhanced mitochondrial oxidative metabolism and altered glucose and lipid metabolism, which may contribute to their pro‐resolving phenotype.
In addition, changes in the metabolic state and phenotype of Mac cells during the efferocytosis of apoptotic Neu (apo‐Neu) cells were assessed. Seahorse test results revealed that MacEffero (apo‐Neu) exhibited significantly enhanced OXPHOS and mitochondrial respiration parameters, such as increased basal respiration, ATP‐linked respiration, maximal respiration and spare respiratory capacity (Figure S4A–C). These metabolic changes were highly consistent with those observed in Mac cells following the efferocytosis of apoptotic T cells (Figure 2G, H), indicating a similar enhancement of Mac mitochondrial function during efferocytosis of apoptotic Neu cells. Furthermore, the efferocytosis of apoptotic Neu by Mac (MacEffero(apo‐Neu)) also significantly increased the expression of pro‐resolving genes (e.g., arg1 and il‐10) in Mac cells (Figure S4D), whose expression pattern was comparable to that observed in Mac cells following the efferocytosis of apoptotic T cells. Together, these data indicate that Mac cells exhibit similar metabolic and phenotypic changes during the efferocytosis of apoptotic T cells and Neu cells.
Next, we used RNA‐seq to determine whether MacEffero cells have signatures similar to those of conventional cytokine (e.g., IL‐4) induced M2‐like cells (MacIL4). The PCA plot revealed a clear separation between MacEffero and MacIL4, suggesting substantial gene signatures despite their similar immunoregulatory features (Figure S4E). We further performed activation macrophage differentiation index (AMDI) and macrophage polarization index (MPI) analyses for full‐spectrum coverage and high‑resolution annotation of Mac subpopulations (Li et al. 2019). The ADMI quantifies the dynamic transition of Mac from early activation to terminal maturation to measure cellular maturity, and the MPI reflects the polarization state and can be used to discriminate the activation phenotype of Mac (Li et al. 2019). Hierarchical clustering analysis based on ADMI and MPI revealed a pronounced difference between MacEffero and MacIL4 (Figure 2K), indicating that compared with conventional MacIL4, MacEffero results in a distinct immunoregulatory phenotype. Together, these findings indicate that efferocytosis itself can induce Mac toward an unconventional phenotype with intrinsic immunoregulatory potential.
3.3. Efferocytic Mac‐Derived ACDVs Suppress Neu Overactivation by Broadly Inhibiting Proinflammatory Networks
As key components of the innate immune system, Neu can be rapidly recruited to sites of tissue injury or infection as first responders in the acute phase of inflammation. Activated Neu cells produce myeloperoxidase (MPO) and proinflammatory cytokines and chemokines (e.g., TNF‑α and IL‑1β), which amplify immune activation and recruit other immune cells (e.g., Mac). Upon completing their effector functions, the infiltrating Neu undergo spontaneously programmed apoptosis, which is driven by the induction of pro‐death signalling pathways and the suppression of survival signals (Rane and Klein 2009; Liew and Kubes 2019). However, the aberrant activation of Neu, coupled with prolonged apoptosis, can sustain the excessive production of proinflammatory mediators, thereby amplifying local inflammation, recruiting additional immune cells and fuelling a vicious cycle that exacerbates tissue injury (Liew and Kubes 2019; Singhal and Kumar 2022). Thus, balancing Neu secretion and timely apoptosis is essential for regulating immune homeostasis. The natural immunoregulatory potential of MacEffero and their intrinsic crosstalk with Neu make them promising therapeutic candidates for the modulation of Neu activity and subsequent Neu‐Mac crosstalk.
In recent years, cell‐derived artificial nanovesicles (ACDVs) have emerged as promising alternatives to conventional cell‑based therapies because of their multiple advantages, such as lower immunogenicity, no risk of tumorigenesis, ability to cross biological barriers and better tissue penetrability (Wen et al. 2022; Murphy et al. 2019; Lou et al. 2026). Moreover, ACDVs can exert similar biological effects on their parent cells, as they can regulate recipient cell functions by transferring diverse types of bioactive cargos (e.g., proteins, lipids and metabolites) inherited from their parent cells. Owing to these unique properties, ACDV‐based therapies have been widely utilized in the fields of tissue engineering, regenerative medicine and targeted drug delivery (Wen et al. 2022). As shown in Figure 3A, Mac‐derived ACDVs were prepared via a previously reported cell extrusion‐based method (Lv et al. 2024). The prepared ACDVs had a bilayer spherical structure with a mean size of ∼160 nm (Figure 3B, C). The effects of MacEffero‐derived ACDVs (ACDVEffero) on Neu activation were assessed both in vitro and in vivo. To this end, we isolated primary Neus from mouse bone marrow via a commercial kit, and these mouse Neus were highly pure (∼97.5%) (Figure S5A, B). FCA also revealed that Memglow (a membrane dye)‐labelled vesicles could be efficiently taken up by Neus in vitro (Figures 3D and S6A).
FIGURE 3.

Functional evaluation of the immunoregulatory effects of ACDVEffero. (A) Schematic diagram of the isolation of ACDV preparations. (B) Representative TEM images of the prepared ACDVs. (C) Size distribution of the isolated ACDVs. (D) Statistical analysis of the ability of Neu to take up ACDVs (n = 3). (E) qPCR analysis of proinflammatory cytokine gene expression in LPS‐induced Neus treated with ACDVcon or ACDVEffero (n = 5). (F) Quantitative analysis of the degree of apoptosis in Neu cells treated with ACDVcon or ACDVEffero via FCA (n = 3). (G) Experimental scheme of the ALI study. (H) Representative IHC images of Ly6G (up panel, scale bar = 100 µm) and MPO (down panel, scale bar = 40 µm) in the lung tissues of ALI mice. (I) Quantitative analysis of Ly6G and MPO expression (n = 6). (J) qPCR analysis of proinflammatory cytokine gene expression in the lung tissues of ALI mice (n = 6).
Next, proinflammatory Neu were induced by LPS stimulation as previously reported (Xie et al. 2024). In LPS‐activated Neu cells, ACDVEffero treatment significantly reduced the expression of proinflammatory cytokines (e.g., mpo and tnf‑α) (Figure 3E) and increased the number of apoptotic cells (Figures 3F and S6B), suggesting a robust inhibitory effect of ACDVEffero on Neu overactivation. Interestingly, native Mac‑derived ACDVs (ACDVCon) also inhibited Neu activation to some extent, which might be due to the natural role of Mac in maintaining immune homeostasis and needs to be revealed in future studies. A previous study revealed that IL‐4‐induced M2‑like Mac cells may have increased efferocytic capacity (Jha et al. 2024), and thus we sought to determine whether ACDVs from MacEffero with IL‐4 pretreatment (ACDVEffero+IL4) would have greater beneficial effects. However, ACDVEffero and ACDVEffero+IL4 had comparable capacities for inhibiting proinflammatory Neu at the same protein mass or particle number (Figure S7A–C). These results further suggest that efferocytosis as a conserved innate immune process can induce a robust immunoregulatory phenotype in Mac cells, whereas additional cytokine stimulation (e.g., IL‐4) does not appear to further enhance this phenotype. We next evaluated the therapeutic effects of ACDVEffero in vivo. In the ALI mouse model, ACDVEffero treatment significantly reduced Ly6G+ Neu infiltration and the expression of the Neu activation marker MPO in lung tissues (Figure 3G–I), accompanied by decreased levels of the pro‐inflammatory cytokines tnf‐α and il‐1β (Figure 3J). Consistently, histopathological analysis showed that ACDVEffero treatment ameliorated lung injury, as evidenced by reduced alveolar wall thickening, inflammatory cell infiltration, haemorrhage and tissue oedema (Figure S8A–D). Collectively, these findings demonstrate that ACDVEffero suppresses inflammatory activation and promotes the apoptosis of Neu cells in vitro while attenuating Neu‐associated inflammation and tissue injury in vivo.
Increasing evidence indicates that ACDVs can influence recipient cell functions through the transfer of diverse bioactive cargos, such as proteins. To explore the molecular basis underlying the therapeutic effects of ACDVEffero, we performed proteomic profiling to compare its bioactive protein composition with that of ACDVCon (Figure 4A). PCA and heatmap analyses revealed clear separation between the two groups (Figure S9A, B). In total, 1456 proteins were identified, with 222 significantly upregulated and 179 significantly downregulated in ACDVEffero relative to those in ACDVCon (Figure 4B). Gene ontology (GO) enrichment analysis of the significantly enriched differentially expressed proteins (DEPs) in ACDVEffero (vs. ACDVCon) revealed a predominant association with metabolic and immune regulation (Figure 4C). Heatmap analysis revealed that ACDVEffero carried multiple bioactive molecules with immunoregulatory effects, such as Ptgs2, Zfp36, Thbd, Ier3 and Acod1 (Figure 4D), which is consistent with the findings that ACDVEffero can effectively suppress inflammation both in vivo and in vitro (Figure 3).
FIGURE 4.

Exploring the effective mechanism of the immunoregulatory ACDVEffero. (A) Schematic diagram of the ACDV content investigation. (B) Volcano plots showing the number of DEPs between ACDVCon and ACDVEffero (|FC| > 1.5 and p‐adjusted < 0.05, n = 3). (C) Biological process of GO enrichment analysis of the up‐DEPs in ACDVEffero compared with those in ACDVCon. (D) Heatmap of the top DEPs in the GO‐BP pathways. (E) Venn analysis of the repeated functional molecules in MacEffero and ACDVEffero. (F) Quantitative analysis of immunoregulatory molecule levels between ACDVCon and ACDVEffero. (G) Schematic diagram of the anti‐inflammatory mechanism of ACDV. (H) KEGG pathway enrichment of the DEGs downregulated by ACDVEffero in LPS‐induced Neu (|FC| > 2 and p‐adjust < 0.05, n = 3). (I) Heatmap of the top DEGs in these pathways.
To further validate these findings, an integrated transcriptomic‐proteomic analysis of MacEffero and ACDVEffero was conducted. Compared with ACDVCon, ACDVEffero was enriched in immunoregulatory molecules, such as Acod1, Zfp36, Hmgcr, Ier3 and Ptgs2 (Figures 4E, F and S9C). Acod1 (aconitate decarboxylase 1) catalyses the conversion of the TCA cycle intermediate cis‐aconitate into itaconate, a vital anti‐inflammatory metabolite that can suppress the transcription of various proinflammatory genes (Hooftman and O'Neill 2019; Wu et al. 2020), and elevated Acod1 expression in neutrophils promotes an immunosuppressive microenvironment (Zhao et al. 2023). Prostaglandin E2 (Ptgs2) catalyses the synthesis of prostaglandin E2 (PGE2) from arachidonic acid, which can drive immunosuppressive macrophage polarization and immune evasion by activating the cAMP/CREB pathway (Jiang and Dingledine 2013). Zfp36 (zinc finger protein 36) is an RNA‐binding protein that can play an anti‐inflammatory role by promoting the mRNA degradation of key proinflammatory cytokines, such as TNF‐α and IL‐6 (Snyder and Blackshear 2022). In addition, Ier3 (immediate early response 3) is a stress‐responsive gene that regulates apoptosis, differentiation and DNA repair (Arlt and Schafer 2011). Collectively, these data indicate that ACDVEffero is enriched in a repertoire of immunomodulatory proteins.
To elucidate the potential mechanisms by which multiple immunoregulatory molecules of ACDVEffero inhibit Neu overactivation, we performed RNA‐seq analysis of overactivated Neus with or without ACDVEffero treatment (Figure 4G). Compared with those in the control group, the number of DEGs (2906 upregulated genes and 1715 downregulated genes, |FC| > 2, p‐adj<0.05) in the LPS‐stimulated Neu group suggested that a complicated proinflammatory signalling network drives Neu overactivation (Figure S9D, E). In contrast, ACDVEffero treatment reversed the upregulation of many DEGs (185 genes) in the overactivated Neu (Figure S9F). Further pathway analysis revealed that the downregulated DEGs in the ACDVEffero group were enriched in multiple key proinflammatory signalling pathways, such as cytokine‐cytokine receptor interactions, the TNF signalling pathway, the NF‐κB signalling pathway, and the chemokine signalling pathway (Figure 4H). These pathway‐enriched DEGs were predominantly proinflammatory cytokines and chemokines, such as il12rb1, il1a, tnf, ccl2 and cxcl2 (Figure 4I), which can serve as central mediators of immune activation. The activation of proinflammatory networks (NF‐κB, TNF, chemokine, cytokine interaction pathways, etc.) in Neu cells might act as a critical “switch” for initiating and amplifying inflammatory responses, driving robust transcription of downstream effectors. Conversely, ACDVEffero carrying multiple immunoregulatory cargos can broadly suppress these pathways, thereby blocking the transcriptional cascade of inflammatory signalling in the Neu of injured sites. Compared with single‐factor blockade, targeted inhibition of key signalling nodes may provide a comprehensive and durable anti‐inflammatory strategy. Together, these results indicate that ACDVEffero harbours a repertoire of potent immunomodulatory proteins that can suppress Neu overactivation via the inhibition of proinflammatory signalling networks.
3.4. Bioengineered ACDVEffero Can Efficiently Target Neu at Injured Sites In Vivo
Having established the immune‑modulating potential of ACDVEffero, we next sought to enhance its Neu‑targeting capacity via surface modification, with the goal of improving its therapeutic efficacy while minimizing off‑target effects. As a major type of innate immune cell, Neu can actively and efficiently recognize microbes via their surface receptors (Silvestre‐Roig et al. 2019; Zhang et al. 2024), and nanoparticles with surface modifications of microbiota‐derived membrane contents have enhanced inflamed site‐targeting ability (Gao et al. 2022). Inspired by this natural property, we designed bioinspired nanovesicles (BNVs) via surface modification of ACDVEffero with probiotic‐derived membrane contents to increase the active targeting capacity of ACDVEffero to Neu. Escherichia coli Nissle 1917 (ECN1917) is a well‐characterized probiotic strain that lacks the major virulence factors of pathogenic E. coli, and many studies have reported its excellent biosafety and beneficial effects (e.g., anti‐inflammatory role) in vivo (Chen et al. 2023; Henker et al. 2007). The European Crohn's and Colitis Organization (ECCO) has recommended ECN1917 as an alternative therapy to mesalazine for patients with ulcerative colitis (Scaldaferri et al. 2016). These reports suggest that membrane components from ECN1917 may facilitate interactions with Neu while maintaining a favourable safety profile, which prompted us to further investigate their Neu‐associated uptake in subsequent experiments.
Next, ECN1917‐derived double membrane vesicles (DMVs) were prepared via an ultrasonic disruption‐based method as previously reported (Chen et al. 2023). The DMV preparations exhibited a bilayer spherical structure with a mean size of ∼160 nm (Figure 5A). The uptake of ACDVEffero and DMVs by Neu cells was evaluated in vitro. Compared with ACDVEffero, FCA significantly increased the uptake of MemGlow‐labelled DMVs by Neu, indicating enhanced Neu‐associated uptake of DMVs (Figure 5B). We then fabricated BNVs by fusing ACDVEffero with DMVs via an ultrasonication/coextrusion method as previously reported (Figure 5C) (Chen et al. 2023). The prepared BNVs had a bilayer spherical structure with a mean size of ∼150 nm (Figure 5D). To verify and quantify the fusion efficacy of ACDVs with DMVs, ACDVs and DMVs were, respectively, labelled with DiD (APC) and MemGlow (PE), and BNVs were generated by the fusion of two differently fluorescently labelled vesicles. The fusion efficiency of BNVs was quantitatively assessed by FCA based on the vesicle proportions with dual‐positive signals, which represent vesicles containing components from both ACDVs and DMVs. The FCA results showed that ∼88% of the dual‐positive signals were displayed in the BNV preparations, indicating a relatively highly efficient and reproducible fusion process (Figure 5E, F). To enhance Neu targeting while preserving the therapeutic payload of ACDVEffero, we next optimized the fused BNVs by adjusting the DMV: ACDV protein mass ratios (1:100, 1:50 and 1:10). The FCA results revealed that increasing the proportion of DMVs in BNVs progressively enhanced their uptake by Neu (Figures 5G, H and S10A). A DMV ratio of 1:50 was therefore selected for subsequent experiments to balance enhanced Neu uptake with the retention of ACDVEffero components.
FIGURE 5.

Evaluation of the Neu‐targeting abilities of BNVs in vitro and in vivo. (A) Representative TEM images of the prepared DMVs (left panel) and their size distributions (right panel). (B) Comparison of the Neu uptake efficiency between the DMV and ACDV via the FCA (n = 3). (C) Schematic diagram of the BNV preparation. (D) Representative TEM images of the prepared BNVs (left panel) and their size distributions (right panel). (E) FCA analysis of BNVs generated by the fusion of DiD‐labelled ACDVs and MemGlow‐labelled DMVs. (F) Quantitative analysis of DiD+MemGlow+ double‐positive vesicles (n = 5). (G) Schematic illustration of the experimental design for evaluating the Neu uptake of BNVs prepared at different ratios by FCA. (H) Comparison of the Neu uptake efficiency of BNVs prepared with different fusion ratios via FCA. (I) Schematic illustration of the experimental design for evaluating the in vivo uptake of BNVs by major innate immune cell populations using FCA. (J) Comparison of Neu uptake efficiency between ACDVs and BNVs via the FCA in vivo (n = 5). (K) Comparison of the uptake efficiency of major innate immune cells for ACDVs and BNVs via the FCA. (L) Schematic illustration of the experimental design for evaluating the in vivo biodistribution of BNVs. (M) Representative IVIS images of injured tissues harvested from mice at 4 h after i.p. injection of Cy7‐labelled ACDVs or BNVs. (N) Quantitative analysis of the fluorescence intensity (n = 3). (O) Representative IF images of Cy7‐labelled ACDV/BNV (red) and Ly6G‐labelled Neu, and the red arrows indicate the colocalization of red and green signals (scale bar = 10 µm).
To further determine whether the molecular composition of ACDVEffero was preserved at this fusion ratio, we performed comparative proteomic profiling of ACDVEffero and BNVEffero. The proteomic profiles of the two vesicle populations were highly similar, with 99.75% of the identified proteins shared between ACDVEffero and BNVEffero (Figure S11A). Among these shared proteins, 98.6% remained at comparable abundance levels after fusion, whereas only 1.4% were significantly altered (Figure S11B). Functional enrichment analysis revealed no significantly enriched biological pathways among these limited DEPs (Figure S11C). Moreover, the unchanged proteins were predominantly associated with metabolic and immune regulatory processes (Figure S11D), whereas the abundance of candidate functional proteins characteristic of ACDVEffero remained largely unchanged following DMV incorporation (Figure S11E). Collectively, these findings demonstrate that fusion with DMVs at a 1:50 ratio largely preserves the proteomic composition and characteristic functional protein profile of ACDVEffero.
We next investigated whether the preferential Neu uptake associated with DMVs was preserved in BNVs following membrane fusion in vivo. The near‐infrared dye Cy7‐labelled‐DMVs or BNVs was intraperitoneally administered to LPS‐pretreated mice, and their uptake by major innate immune cell populations, including Ly6G+ Neu, Ly6C+ monocytes (Mono) and F4/80+ Mac, was assessed by FCA 1 h after administration (Figure S12A, B). The results revealed that the uptake of the DMV tended to be slightly higher than that of the BNV in Ly6G+ Neu, Ly6C+ Mono and F4/80+ Mac cells, but the differences between the DMV and BNV groups were not significant (Figure S12C). Notably, when compared across different immune cell populations, both the DMV and the BNV resulted in significantly higher uptake rates of Neu than those of Mono or Mac (Figure S12D). These results suggest that the preferential uptake of the DMV by the Neu observed in vitro is maintained in vivo and largely preserved in BNVs.
We further determined whether compared with the parental ACDVs, the DMVs could increase the Neu uptake and inflammatory‐site accumulation of BNVs. Cy7‐labelled ACDVs (ACDVCy7) or BNVs (BNVCy7) were systemically injected (i.p.) into LPS‐pretreated mice, followed by FCA analysis of vesicle uptake by Ly6G+ Neu, Ly6C+ Mono and F4/80+ Mac cells (Figure 5I). BNV uptake by Neu was significantly higher than that by ACDV (Figure 5J). Moreover, in the BNV‐treated group, Neu exhibited significantly greater vesicle uptake than Mono or Mac (Figure 5K), further supporting the preferential association of BNVs with Neu in vivo. We next evaluated BNV accumulation in injured lung tissues in an ALI mouse model. At 4 h after administration, BNVs elicited substantially stronger signals in injured lungs than ACDV did (Figure 5L–N). Consistently, fluorescence imaging of lung sections revealed prominent colocalization between BNVs and Ly6G+ Neu, whereas substantially less colocalization was observed for ACDV (Figure 5O). Together, these results indicate that BNVs can efficiently target Neus at injured sites, which enables specific modulation of dysregulated Neu‐Mac communication in disease states.
3.5. BNV‐Based Drug Delivery Systems Can Hitchhike the Apoptotic Neu Route to Regulate Mac Function
Neu‐Mac communication is a natural and dynamic immune process in vivo, with Neu engaging Macs either directly through efferocytosis or indirectly via paracrine signalling (Rosales 2020). During the inflammation resolution phase, the Neu recruited at the injury site can undergo apoptosis followed by clearance by Mac efferocytosis. However, excessive accumulation of dying Neu beyond the clearance capacity of Mac can lead to secondary necrosis and consequently amplify inflammatory tissue injury (Brostjan and Oehler 2020). In addition, the efferocytotic capacity of Macs can be impaired under harsh tissue injury conditions (Doran et al. 2020). Given that ACDVEffero can suppress inflammatory Neus and induce apoptosis, we hypothesized that enhancing Mac efferocytosis may further augment its therapeutic efficacy by enabling spatiotemporal regulation of Neu‐Mac crosstalk. Mitochondrial oxidative metabolism, such as fatty acid oxidation (FAO), plays a vital role in regulating the efferocytosis and phenotypic transitions of Macs because it can provide energy for supporting the cytoskeletal remodelling and phagosome maturation required for AC engulfment (Schilperoort et al. 2023). Moreover, the metabolites of lipid catabolism can modulate transcriptional networks to promote Mac toward an immunoregulatory phenotype (Gerlach et al. 2021; Zizzo and Cohen 2015). Peroxisome proliferator‐activated receptor γ (PPAR‑γ), the main member of the nuclear receptor superfamily, plays a major role in modulating lipid metabolic programs, and PPARγ activation can enhance the efferocytosis and anti‐inflammatory responses of Mac cells (A.G. et al. 2009; Heming et al. 2018).
Although PPARγ agonists (e.g., rosiglitazone, RSG) can promote efferocytosis, systemically administered small‐molecule drugs commonly have short half‐lives and nonspecific tissue distributions in vivo, and thus have limited efficacy in targeting Mac and may cause adverse side effects after repeated dosing. BNV efficiently targeted Neu at the injury sites (Figure 5), and ACDVEffero promoted overactivated Neu apoptosis (Figure 3). Based on these results, we speculated that BNVs may serve as targeted drug carriers that can hitchhike apoptotic Neu cells to regulate Mac cells. As shown in Figure 6A, drugs (RSG) were loaded into BNVs (RSG@BNV) via a coincubation method as previously reported (Liu et al. 2022). The prepared RSG@BNVs displayed a uniform spherical structure with a mean size of ∼150 nm (Figure 6B, C), and the loading capacity of RSG in BNVs determined by LC‐MS/MS was ∼11.3% (w/w) (Figure S13A). In vitro drug release tests revealed that the majority of the encapsulated RSG (∼80%) was slowly released from the RSG@BNV after 24 h (Figure 6D), which aligns with the finding that BNVs can rapidly (within 1 h) reach the Neu of the injured site after systemic injection (Figure 5J).
FIGURE 6.

RSG@BNVs hitchhike apoptotic Neu route to promote Mac efferocytosis. (A) Schematic diagram of the RSG@BNV construction. (B) Comparison of representative TEM images of BNVs and RSG@BNVs. (C) Size distribution of the prepared RSG@BNVs. (D) In vitro release profile of the RSG from RSG@BNVs (n = 3). (E) Schematic illustration of the in vitro assay for the Neu uptake of BNVs and Cur@BNVs. (F) Representative images of Neus treated with BNVs or Cur@BNVs (scale bar = 50 µm). (G) Quantitative analysis of the positive Cur signals in Neus treated with BNVs or Cur@BNVs via FCA (n = 3). (H) Schematic illustration of the in vitro assay for the efferocytosis‐mediated transfer of BNVs and Cur@BNVs from apoptotic Neu cells to Mac cells. (I) Representative images of Macs treated with apoptotic NeuBNV or NeuCur@BNVs (scale bar = 50 µm). (J) Quantitative analysis of the positive Cur signals in Macs treated with apoptotic NeuBNV or NeuCur@BNVs via FCA (n = 3). (K) Quantitative analysis of the RSG levels in Mac treated with apoptotic NeuBNV or NeuRSG@BNVs using LC‐MS/MS (n = 5). (L) FCA analysis of the efferocytosis index in Mac treated with RSG (n = 3). (M) qPCR analysis of proinflammatory cytokine gene expression in LPS‐induced M1‐like Mac cells treated with RSG (n = 3).
We next assessed whether BNV can transfer drugs into Mac cells via a two‐step cell test with a model drug (curcumin, Cur) (Figure 6E, H) because Cur has poor solubility similar to that of RSG and can be readily detected via its natural fluorescent signals (Xu et al. 2021). First, the prepared Cur‐loaded BNVs (Cur@BNVs) were added to Neu, and both imaging and FCA revealed significantly greater signals of Cur in the Neu in the Cur@BNV group than in the carrier alone (BNV) group at a short time (1 h) (Figures 6F, G and S13B), suggesting that BNVs have a high capacity to deliver drugs into the Neu. Second, apoptotic Neu cells pretreated with BNVs (NeuBNV) or Cur@BNV (NeuCur@BNV) were cocultured with Mac cells for 1 h (Figure 6H). Both imaging and FCA revealed clear signals of Cur in the Mac of the NeuCur@BNV group, whereas this effect was absent in the Mac of the NeuBNV group (Figures 6I‐J and S13C), indicating that RSG@BNV can efficiently deliver payloads to Mac through hitchhiking along the apoptotic Neu pathway, thereby providing the prerequisite for subsequent spatiotemporal regulation of Neu‐Mac crosstalk.
To directly validate RSG transfer between cells, apoptotic Neu cells were pretreated with RSG@BNVs or unloaded BNVs for 1 h, after which the free vesicles were removed, and the cells were cocultured with Mac cells. LC–MS/MS analysis revealed a strong RSG signal in Mac cells cocultured with NeuRSG@BNV, whereas no detectable RSG signal was observed in Mac cocultured with NeuBNV (Figure 6K). These results provide direct evidence for the transfer of RSG from RSG@BNV‐treated apoptotic Neu cells to Mac cells during efferocytosis. We next examined the functional effects of RSG on pro‐inflammatory Mac. In line with a previous report, LPS‐induced M1‐like Mac cells presented a decreased efferocytic capacity, which was partially restored by RSG treatment (Figure 6L). RSG treatment also reduced the expression of pro‐inflammatory cytokines, including IL‐1β, IL‐6 and TNF‐α (Figure 6M). Together, these findings support the feasibility of using apoptotic Neu as an intermediate cellular carrier for the efferocytosis‐mediated delivery of RSG to Mac, thereby enhancing Mac efferocytosis and promoting inflammation resolution.
3.6. RSG@BNVEffero Mitigated Dysregulated Neu‐Mac Interactions and Inflammatory Tissue Injury In Vivo
Next, the in vivo therapeutic effect of RSG@BNVEffero was evaluated in diverse types of tissue injury models, including acute inflammation (ALI) and chronic inflammation (CKD) (Figures 7, 8). In mouse ALI models, RSG@BNVEffero treatment via i.p. injection significantly reduced the number of infiltrating Ly6G+ Neus and the expression of activated Neu markers (e.g., mpo and cxcr2) (Figures 7A–D and G–H), suggesting that they can suppress Neu overactivation in vivo. RSG@BNVEffero treatment also decreased the numbers of infiltrated F4/80+ Mac, the spatial overlap between Neu and Mac cells, and the number of TUNEL+ apoptotic cells in the lungs of ALI mice (Figures 7E, I and S14A–D). Consequently, the expression of pro‐inflammatory cytokines (ICAM‐1 and IL‐1β) (Figure S15A–D) and the lung injury score of ALI mice could be alleviated by RSG@BNVEffero treatment (Figure 7F, J). Importantly, compared with BNVEffero or RSG alone, RSG@BNVEffero exhibited greater overall therapeutic efficacy in reducing lung inflammation and reducing the degree of tissue injury (Figures 7B–J and S15A–D).
FIGURE 7.

Evaluation of the therapeutic efficacy of RSG@BNVEffero in acute lung injury models. (A) Experimental scheme of the ALI study. (B) qPCR analysis of proinflammatory cytokine gene expression in the lung tissues of ALI mice (n = 6). (C‐E) Representative IHC images of Ly6G (C, scale bar = 100 µm), MPO (D, scale bar = 40 µm) and F4/80 (E, scale bar = 100 µm) in the lung tissues of ALI mice. (F) Representative H&E staining images of lung tissue from ALI mice (scale bar = 100 µm). (G–J) Quantitative analysis of the levels of Ly6G (G), MPO (H), and F4/80 (I) and the histology score (J) in lung tissue from ALI mice.
FIGURE 8.

Evaluation of the therapeutic efficacy of RSG@BNVEffero in chronic kidney disease models. (A) Experimental scheme of the CKD study. (B) qPCR analysis of proinflammatory cytokine gene expression in the kidney tissues of CKD mice (n = 6). (C) Representative IHC images of Ly6G in the kidney tissues in CKD mice (scale bar = 100 µm). (D) Quantitative analysis of Ly6G levels in CKD mice. (E) Representative IHC images of MPO in the kidney tissues in CKD mice (scale bar = 40 µm). (F) Quantitative analysis of MPO levels in CKD mice. (G) Representative IHC images of F4/80 in the kidney tissues in CKD mice (scale bar = 100 µm). (H) Quantitative analysis of F4/80 levels in CKD mice. (I) Representative picrosirius red‐stained images of kidneys on Day 3 after CKD (scale bar = 100 µm). (J) Quantitative analysis of the renal fibrosis area.
Given the preferential Neu uptake of DMVs, we further investigated whether DMV‐mediated Neu targeting and RSG delivery alone were sufficient to reproduce the therapeutic effects of RSG@BNVEffero. Although the DMVs were efficiently taken up by Neu, the DMV treatment did not significantly suppress the LPS‐induced inflammatory activation of Neu and induced only a modest increase in Neu apoptosis (Figure S16A–C). We therefore further prepared RSG‐loaded DMVs (RSG@DMV) as a control formulation (Figure S16D). In the ALI mouse model, RSG@DMV treatment failed to significantly reduce the expression of inflammatory markers (e.g., mpo, tnf‐α and il‐1β) or ameliorate histopathological lung injury (Figure S16E–G). These findings indicate that DMV‐mediated Neu targeting and RSG loading alone are insufficient to reproduce the therapeutic efficacy of RSG@BNVEffero, supporting the critical contribution of ACDVEffero in therapeutics. Collectively, the therapeutic effects of RSG@BNVEffero appear to arise from the coordinated functions of its components, including DMV‐mediated Neu targeting, ACDVEffero‐mediated regulation of Neu inflammation and apoptosis, and efferocytosis‐mediated delivery of RSG to Mac cells.
Consistently, in the chronic tissue injury model (CKD), RSG@BNVEffero treatment significantly suppressed the expression of activated Neu markers (e.g., mpo, cxcr2, tnf‐α and il‐1β) and the infiltration of Ly6G+ Neu in the kidney tissues of CKD mice (Figure 8A‐F). Additionally, the RSG@BNVEffero treatment reduced the infiltration of F4/80+ Mac cells, the spatial overlap between Neu and Mac cells, and the number of TUNEL+ apoptotic cells in the kidney tissues of CKD mice (Figures 8G–H and S17A, B). RSG@BNVEffero treatment effectively reduced the excessive expression of proinflammatory cytokines/chemokines (IL‑1β and ICAM‐1) (Figure S18A, B). These results verified that RSG@BNVEffero can inhibit chronic inflammation via the regulation of Neu‐Mac crosstalk. In addition, persistent unresolved inflammation can elicit profibrotic pathways and result in organ fibrosis progression Pahwa et al. 2023. As a result, the RSG@BNVEffero treatments significantly reduced ECM deposition and kidney injury scores in the kidneys of CKD mice (Figure 8I, J). Taken together, these results indicate that RSG@BNVEffero effectively mitigates acute or chronic tissue injury via targeted spatiotemporal regulation of Neu‐Mac crosstalk, suggesting a promising approach for treating diverse inflammatory diseases.
4. Discussion
Efferocytosis of dying cells (e.g., apoptotic Neu or T cells) by Mac at tissue injury sites represents a conserved mechanism that coordinates inflammation resolution and tissue repair (Boada‐Romero et al. 2020; Rodriguez‐Morales and Franklin 2023). In this study, our results suggest that efferocytosis alone can induce a distinct immunoregulatory phenotype that unlike the conventional IL‐4‐induced M2‐like phenotype in Mac accompanied by metabolic reprogramming. MacEffero enhanced mitochondrial oxidative metabolism and increased expression of pro‐resolving and tissue repair program‐associated molecules. Importantly, ACDVEffero derived from these MacEffero retain their immunoregulatory properties and can effectively suppress the pro‐inflammatory activation of Neu while promoting their apoptosis. These findings provide a biological basis for exploiting the endogenous Neu‐Mac efferocytosis axis as a therapeutic route for inflammation resolution.
However, the therapeutic efficacy of ACDVEffero depends not only on its intrinsic biological activity but also on its efficient interaction with Neu in vivo. We therefore introduced ECN1917‐derived DMV membrane components (DMVs) into ACDVEffero to increase its targeting efficiency toward Neu. This engineering strategy was inspired by the intrinsic ability of Neu to rapidly recognize microbial components via diverse pattern‐recognition receptors (Gao et al. 2022). Consistent with these findings, the uptake of DMVs was preferential for Neu both in vitro and in vivo. Compared with unmodified ACDVEffero, the incorporation of DMV membrane components significantly enhanced the uptake of BNVEffero (fusion of DMV and ACDVEffero) by Neu. Importantly, proteomic analysis demonstrated that fusion with DMVs caused minimal alteration of the ACDVEffero proteome and efficiently preserved its characteristic functional protein profile. Thus, surface modification of ACDVEffero via membrane fusion provides an effective strategy for enhancing its Neu‐targeting ability while preserving its intrinsic biological functions.
Then, this BNV platform was subsequently engineered to leverage the Neu‐Mac efferocytosis process for sequential therapeutic delivery. Following BNV uptake, ACDVEffero suppresses the inflammatory activation of Neu and promotes their apoptosis. The clearance of apoptotic Neu relies on efferocytosis by Mac. However, previous studies and our findings have shown that the efferocytic capacity of Mac can be impaired under inflammatory conditions (Mehrotra and Ravichandran 2022; Zhou et al. 2025), which may compromise the efficient clearance of apoptotic Neu and subsequent inflammation resolution. To address this issue, RSG was selected as the therapeutic payload because of its ability to enhance macrophage efferocytic activity. Upon uptake of RSG@BNVs, apoptotic Neu can act as an intermediate carrier to transfer RSG into Mac cells during efferocytosis, thereby enhancing apoptotic Neu clearance and promoting inflammation resolution. This strategy exploits the endogenous efferocytosis process for the delivery of RSG to Mac cells rather than relying on direct Mac targeting. Importantly, LC‐MS/MS analysis directly confirmed that RSG was transferred from the RSG@BNV‐treated apoptotic Neu cells to the Mac cells, supporting the feasibility of this efferocytosis‐mediated delivery strategy.
Indeed, the rationale for this engineered platform is mainly based on the combined therapeutic requirements, and each individual component exerts distinct functional contributions. For instance, although the use of DMV alone resulted in efficient Neu uptake, it failed to significantly suppress the pro‐inflammatory activation of Neu and induce apoptosis. As a result, the RSG@DMV treatments did not significantly ameliorate pulmonary inflammation or tissue injury in the ALI model. These findings indicate that Neu‐targeting or RSG delivery alone is insufficient to reproduce the therapeutic efficacy of RSG@BNVEffero. In contrast, the advanced therapeutic capacity of RSG@BNVEffero arises from a coordinated process that integrates DMV‐mediated Neu‐targeting, ACDVEffero‐mediated regulation of Neu activation and apoptosis, and Neu‐to‐Mac RSG transfer‐boosted efferocytosis. This sequential strategy therefore differs from conventional drug delivery systems that primarily rely on direct targeting of a single cell type.
From a translational perspective, this bioinspired engineering strategy might offer several potential advantages. First, RSG@BNVEffero utilizes the intrinsic immunoregulatory properties acquired by Mac during efferocytosis and thereby harnesses an endogenous inflammation‐resolution program rather than broadly suppressing immune responses. Second, the incorporation of DMV alone enhances Neu uptake and largely preserves the functional proteomic profile of ACDVEffero. Furthermore, efferocytosis‐mediated RSG transfer enables sequential modulation of the Neu‐Mac axis by first regulating Neu activation and apoptosis and subsequently enhancing Mac efferocytosis and the resolution of inflammation. Consistent with this sequential strategy, RSG@BNV effectively attenuated inflammation and tissue injury in both acute and chronic injury models, highlighting its therapeutic potential across different inflammatory conditions.
However, there are several limitations in the current study. Although our results establish the functional treatment of ACDVEffero and demonstrate preferential Neu uptake of DMV‐modified BNVs, the molecular mechanisms responsible for DMV recognition by Neu remain to be determined. In addition, although our findings indicate the therapeutic efficacy and preliminary biosafety of the BNV platform in vivo, further studies addressing its pharmacokinetics, long‐term safety and scalable manufacturing will be needed before its clinical translation. Future studies integrating advanced approaches for tracking intercellular cargo transfer may provide a more detailed understanding of the dynamic Neu‐Mac communication underlying this therapeutic process. Nevertheless, this study highlights that bioinspired nanotherapeutics can harness intrinsic mechanisms to achieve spatiotemporal and targeted regulation of immune responses and resolution of inflammation, which may provide a conceptual strategy for developing next‐generation regenerative medicine.
5. Conclusion
In summary, we developed a bioinspired BNV (RSG@BNVEffero) platform with the ability to integrate Neu targeting, efferocytosis‐associated immunoregulation and Neu‐to‐Mac drug delivery to promote inflammation resolution and tissue repair. By combining the immunoregulatory properties of ACDVEffero with the Neu‐targeting property of probiotic‐derived membrane components, BNV can efficiently suppress pro‐inflammatory Neu and promote their apoptosis. The subsequent engulfment of BNVs‐containing apoptotic Neu by Mac enables the transfer of the pro‐efferocytic drug (RSG) into Mac, thereby enhancing Mac efferocytosis with dying cell clearance and inflammation resolution. This strategy highlights the potential of harnessing natural immune reactions for spatiotemporal regulation of the Neu‐Mac axis, providing a conceptual framework for the development of bioinspired regenerative therapies.
Author Contributions
Yizhuo Wang: conceptualization, methodology, software, data curation, formal analysis, validation, investigation, visualization, project administration, resources, writing – original draft, writing – review and editing. Peng Lou: conceptualization, methodology, software, data curation, resources, formal analysis, validation, visualization, writing – review and editing, investigation. Xiyue Zhou: conceptualization, investigation, methodology, validation, visualization, writing – review and editing, software, formal analysis, data curation, resources. Zehuan Lin: methodology, software, formal analysis, visualization, resources, data curation, investigation. Shuyun Liu: methodology, software, data curation, writing – review and editing, investigation, funding acquisition. Lan Li: methodology, software, data curation, formal analysis, validation, investigation. Younan Chen: validation, formal analysis, data curation, visualization, writing – review and editing. Dan Du: visualization, writing – review and editing, formal analysis, methodology. Hao Yang: visualization, formal analysis, writing – review and editing, data curation. Yanrong Lu: writing – review and editing, visualization, formal analysis, resources. Meihua Wan: writing – review and editing, resources, data curation. Jingping Liu: conceptualization, investigation, funding acquisition, writing – original draft, methodology, validation, visualization, writing – review and editing, software, formal analysis, project administration, data curation, supervision, resources.
Conflicts of Interest
The authors declare that they have no competing interests in this work.
Supporting information
Supporting Information: jev270406‐sup‐0001‐SuppMat.docx
Supporting Information: jev270406‐sup‐0001‐FigureS5.png
Acknowledgements
This study was partly supported by the National Natural Science Foundation of China (32271438 to J.L., 82472129 to J.L., and 82570811 to S.L.), the Sichuan Science and Technology Program (2024NSFSC0586 to J.L., and 2024YFFK0199 to S.L.), and the 1.3.5 Project for Disciplines of Excellence of West China Hospital, Sichuan University. (ZYYC23001 to J.L.) We thank Xijing Yang and Xiaoting Chen from the Animal Experimental Center of West China Hospital for assisting with the animal experiments, Na Jiang from the Advanced Mass Spectrometry Center, Research Core Facility of West China Hospital for assisting with technical support on LC–MS/MS experiments.
Data Availability Statement
All the data needed to evaluate the conclusions in the paper are presented in the paper and/or the Supplementary Materials.
References
- A‐Gonzalez, N. , Bensinger S. J., Hong C., et al. 2009. “Apoptotic Cells Promote Their Own Clearance and Immune Tolerance Through Activation of the Nuclear Receptor LXR.” Immunity 31, no. 2: 245–258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Arlt, A. , and Schafer H.. 2011. “Role of the Immediate Early Response 3 (IER3) Gene in Cellular Stress Response, Inflammation and Tumorigenesis.” European Journal of Cell Biology 90, no. 6‐7: 545–552. [DOI] [PubMed] [Google Scholar]
- Boada‐Romero, E. , Martinez J., Heckmann B. L., and Green D. R.. 2020. “The Clearance of Dead Cells by Efferocytosis.” Nature Reviews Molecular Cell Biology 21, no. 7: 398–414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brostjan, C. , and Oehler R.. 2020. “The Role of Neutrophil Death in Chronic Inflammation and Cancer.” Cell Death Discovery 6: 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen, H. , Lei P., Ji H., et al. 2023. “Advances in Escherichia coli Nissle 1917 as a Customizable Drug Delivery System for Disease Treatment and Diagnosis Strategies.” Mater Today Bio 18: 100543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen, W. , Song Y., Bai S., et al. 2023. “Cloaking Mesoporous Polydopamine With Bacterial Membrane Vesicles to Amplify Local and Systemic Antitumor Immunity.” ACS Nano 17, no. 8: 7733–7749. [DOI] [PubMed] [Google Scholar]
- Doran, A. C. , Yurdagul A. Jr., and Tabas I.. 2020. “Efferocytosis in Health and Disease.” Nature Reviews Immunology 20, no. 4: 254–267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gao, J. , Su Y., and Wang Z.. 2022. “Engineering Bacterial Membrane Nanovesicles for Improved Therapies in Infectious Diseases and Cancer.” Advanced Drug Delivery Reviews 186: 114340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gerlach, B. D. , Ampomah P. B., Yurdagul A. Jr., Liu C., et al. 2021. “Efferocytosis Induces Macrophage Proliferation to Help Resolve Tissue Injury.” Cell Metabolism 33, no. 12: 2445–2463.e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Greenlee‐Wacker, M. C. 2016. “Clearance of Apoptotic Neutrophils and Resolution of Inflammation.” Immunological Reviews 273, no. 1: 357–370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Harirforoosh, S. , Asghar W., and Jamali F.. 2013. “Adverse Effects of Nonsteroidal Antiinflammatory Drugs: An Update of Gastrointestinal, Cardiovascular and Renal Complications.” Journal of Pharmacy & Pharmaceutical Sciences 16, no. 5: 821–847. [DOI] [PubMed] [Google Scholar]
- Heming, M. , Gran S., Jauch S. L., et al. 2018. “Peroxisome Proliferator‐Activated Receptor‐Gamma Modulates the Response of Macrophages to Lipopolysaccharide and Glucocorticoids.” Frontiers in Immunology 9: 893. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Henker, J. , Laass M., Blokhin B. M., et al. 2007. “The Probiotic Escherichia coli Strain Nissle 1917 (EcN) Stops Acute Diarrhoea in Infants and Toddlers.” European Journal of Pediatrics 166, no. 4: 311–318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hooftman, A. , and O'Neill L. A. J.. 2019. “The Immunomodulatory Potential of the Metabolite Itaconate.” Trends in Immunology 40, no. 8: 687–698. [DOI] [PubMed] [Google Scholar]
- Jha, P. K. , Aikawa M., and Aikawa E.. 2024. “Macrophage Heterogeneity and Efferocytosis: Beyond the M1/M2 Dichotomy.” Circulation Research 134, no. 2: 186–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jiang, J. , and Dingledine R.. 2013. “Prostaglandin Receptor EP2 in the Crosshairs of Anti‐Inflammation, Anti‐Cancer, and Neuroprotection.” Trends in Pharmacological Sciences 34, no. 7: 413–423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karagiannis, S. N. , and Arnold J. N.. 2022. “Immune Cell‐Antibody Interactions in Health and Disease.” Clinical and Experimental Immunology 209, no. 1: 1–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kelly, P. J. , Lemmens R., and Tsivgoulis G.. 2021. “Inflammation and Stroke Risk: A New Target for Prevention.” Stroke; A Journal of Cerebral Circulation 52, no. 8: 2697–2706. [DOI] [PubMed] [Google Scholar]
- Leuti, A. , Fazio D., Fava M., Piccoli A., Oddi S., and Maccarrone M.. 2020. “Bioactive Lipids, Inflammation and Chronic Diseases.” Advanced Drug Delivery Reviews 159: 133–169. [DOI] [PubMed] [Google Scholar]
- Li, C. , Menoret A., Farragher C., et al. 2019. “Single Cell Transcriptomics Based‐MacSpectrum Reveals Novel Macrophage Activation Signatures in Diseases.” JCI Insight 5, no. 10: e126453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liew, P. X. , and Kubes P.. 2019. “The Neutrophil's Role During Health and Disease.” Physiological Reviews 99, no. 2: 1223–1248. [DOI] [PubMed] [Google Scholar]
- Lin, J. , Yang F., Zheng Y., Wang X., Fan X., and Yang L.. 2025. “Glycometabolic Reprogramming of Immune Cells in Autoimmune Diseases.” Autoimmunity Reviews 24, no. 10: 103876. [DOI] [PubMed] [Google Scholar]
- Lin, Z. , Zhou X., Liu S., Wan M., and Liu J.. 2026. “Advancements in Nanomedicine for the Therapeutic Regulation of Efferocytosis: Opportunities and Challenges.” Theranostics 16, no. 10: 5537–5570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu, S. , Lv K., Wang Y., et al. 2022. “Improving the Circulation Time and Renal Therapeutic Potency of Extracellular Vesicles Using an Endogenous Ligand Binding Strategy.” Journal of Controlled Release 352: 1009–1023. [DOI] [PubMed] [Google Scholar]
- Lou, P. , Wang Y., Zhang Y., et al. 2026. “Native Cell Niche‐Inspired Multiorder Nanoassemblies Orchestrate Integrated Interaction Networks to Guide Muscle Regeneration.” Molecular Therapy 34, no. 4: 2081–2097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lou, P. , Zhou X., Zhang Y., et al. 2025. “Harnessing Tissue‐Derived Mitochondria‐Rich Extracellular Vesicles (Ti‐mitoEVs) to Boost Mitochondrial Biogenesis for Regenerative Medicine.” Science Advances 11, no. 29: eadt1318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lv, K. , Lou P., Liu S., et al. 2024. “Injectable Multifunctional Composite Hydrogel as a Combination Therapy for Preventing Postsurgical Adhesion.” Small 20, no. 1: e2303425. [DOI] [PubMed] [Google Scholar]
- Mao, R. , Wang C., Zhang F., et al. 2020. “Peritoneal M2 Macrophage Transplantation as a Potential Cell Therapy for Enhancing Renal Repair in Acute Kidney Injury.” Journal of Cellular and Molecular Medicine 24, no. 6: 3314–3327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mehrotra, P. , and Ravichandran K. S.. 2022. “Drugging the Efferocytosis Process: Concepts and Opportunities.” Nature Reviews Drug Discovery 21, no. 8: 601–620. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Murphy, D. E. , de Jong O. G., Brouwer M., et al. 2019. “Extracellular Vesicle‐Based Therapeutics: Natural Versus Engineered Targeting and Trafficking.” Experimental & Molecular Medicine 51, no. 3: 1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pahwa, R. , Goyal A., and Jialal I.. Chronic Inflammation, StatPearls, Treasure Island (FL) Ineligible Companies. Disclosure: Amandeep Goyal declares no relevant financial relationships with ineligible companies. Disclosure: Ishwarlal Jialal declares no relevant financial relationships with ineligible companies. 2023. [Google Scholar]
- Rane, M. J. , and Klein J. B.. 2009. “Regulation of Neutrophil Apoptosis by Modulation of PKB/Akt Activation.” Frontiers in Bioscience (Landmark Ed) 14, no. 7: 2400–2412. [DOI] [PubMed] [Google Scholar]
- Rayamajhi, S. , Nguyen T. D. T., Marasini R., and Aryal S.. 2019. “Macrophage‐Derived Exosome‐Mimetic Hybrid Vesicles for Tumor Targeted Drug Delivery.” Acta Biomaterialia 94: 482–494. [DOI] [PubMed] [Google Scholar]
- Rodriguez‐Morales, P. , and Franklin R. A.. 2023. “Macrophage Phenotypes and Functions: Resolving Inflammation and Restoring Homeostasis.” Trends in Immunology 44, no. 12: 986–998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rosales, C. 2020. “Neutrophils at the Crossroads of Innate and Adaptive Immunity.” Journal of Leukocyte Biology 108, no. 1: 377–396. [DOI] [PubMed] [Google Scholar]
- Scaldaferri, F. , Gerardi V., Mangiola F., et al. 2016. “Role and Mechanisms of Action of Escherichia coli Nissle 1917 in the Maintenance of Remission in Ulcerative Colitis Patients: An Update.” World Journal of Gastroenterology 22, no. 24: 5505–5511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schilperoort, M. , Ngai D., Sukka S. R., Avrampou K., Shi H., and Tabas I.. 2023. “The Role of Efferocytosis‐Fueled Macrophage Metabolism in the Resolution of Inflammation.” Immunological Reviews 319, no. 1: 65–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Silvestre‐Roig, C. , Fridlender Z. G., Glogauer M., and Scapini P.. 2019. “Neutrophil Diversity in Health and Disease.” Trends in Immunology 40, no. 7: 565–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Singer, M. , Deutschman C. S., Seymour C. W., et al. 2016. “The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis‐3).” Jama 315, no. 8: 801–810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Singhal, A. , and Kumar S.. 2022. “Neutrophil and Remnant Clearance in Immunity and Inflammation.” Immunology 165, no. 1: 22–43. [DOI] [PubMed] [Google Scholar]
- Snyder, B. L. , and Blackshear P. J.. 2022. “Clinical Implications of Tristetraprolin (TTP) Modulation in the Treatment of Inflammatory Diseases.” Pharmacology & Therapeutics 239: 108198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang, H. , Zhang Y., Zhang Y., et al. 2024. “Activating Macrophage Continual Efferocytosis via Microenvironment Biomimetic Short Fibers for Reversing Inflammation in Bone Repair.” Advanced Materials 36, no. 30: e2402968. [DOI] [PubMed] [Google Scholar]
- Wang, Y. , Lou P., Xie Y., et al. 2024. “Nutrient Availability Regulates the Secretion and Function of Immune Cell‐Derived Extracellular Vesicles Through Metabolic Rewiring.” Science Advances 10, no. 7: eadj1290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang, Y. , Zhao M., Liu S., et al. 2020. “Macrophage‐Derived Extracellular Vesicles: Diverse Mediators of Pathology and Therapeutics in Multiple Diseases.” Cell Death & Disease 11, no. 10: 924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wculek, S. K. , Heras‐Murillo I., Mastrangelo A., et al. 2023. “Oxidative Phosphorylation Selectively Orchestrates Tissue Macrophage Homeostasis.” Immunity 56, no. 3: 516–530.e9. [DOI] [PubMed] [Google Scholar]
- Wen, Y. , Fu Q., Soliwoda A., et al. 2022. “Cell‐Derived Nanovesicles Prepared by Membrane Extrusion Are Good Substitutes for Natural Extracellular Vesicles.” Extracellular Vesicle 1: 100004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu, R. , Chen F., Wang N., Tang D., and Kang R.. 2020. “ACOD1 in Immunometabolism and Disease.” Cellular and Molecular Immunology 17, no. 8: 822–833. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xie, X. , Wang P., Jin M., et al. 2024. “IL‐1beta‐Induced Epithelial Cell and Fibroblast Transdifferentiation Promotes Neutrophil Recruitment in Chronic Rhinosinusitis With Nasal Polyps.” Nature Communications 15, no. 1: 9101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xu, X. L. , Zhang N. N., Shu G. F., et al. 2021. “A Luminol‐Based Self‐Illuminating Nanocage as a Reactive Oxygen Species Amplifier to Enhance Deep Tumor Penetration and Synergistic Therapy.” ACS Nano 15, no. 12: 19394–19408. [DOI] [PubMed] [Google Scholar]
- Yao, W. , Chen Y., Li Z., et al. 2022. “Single Cell RNA Sequencing Identifies a Unique Inflammatory Macrophage Subset as a Druggable Target for Alleviating Acute Kidney Injury.” Advanced Science (Weinh) 9, no. 12: e2103675. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang, F. , Xia Y., Su J., et al. 2024. “Neutrophil Diversity and Function in Health and Disease.” Signal Transduction and Targeted Therapy 9, no. 1: 343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang, S. , Weinberg S., DeBerge M., et al. 2019. “Efferocytosis Fuels Requirements of Fatty Acid Oxidation and the Electron Transport Chain to Polarize Macrophages for Tissue Repair.” Cell Metabolism 29, no. 2: 443–456.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhao, Y. , Liu Z., Liu G., et al. 2023. “Neutrophils Resist Ferroptosis and Promote Breast Cancer Metastasis Through Aconitate Decarboxylase 1.” Cell Metabolism 35, no. 10: 1688–1703.e10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhou, H. , Xiao Y., Xue X., et al. 2025. “Hyperglycemia Exacerbates Osteoarthritis by Impairing Macrophage Efferocytosis Through Modulation of CD11b Lactylation.” Nature Communications 17, no. 1: 724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zizzo, G. , and Cohen P. L.. 2015. “The PPAR‐Gamma Antagonist GW9662 Elicits Differentiation of M2c‐Like Cells and Upregulation of the MerTK/Gas6 Axis: A Key Role for PPAR‐Gamma in Human Macrophage Polarization.” Journal of Inflammation (London) 12: 36. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information: jev270406‐sup‐0001‐SuppMat.docx
Supporting Information: jev270406‐sup‐0001‐FigureS5.png
Data Availability Statement
All the data needed to evaluate the conclusions in the paper are presented in the paper and/or the Supplementary Materials.
