Lipopolysaccharide was purchased from Sigma-Aldrich (St.Louis, MO, USA). Murine IL-4 was obtained from Peprotech Inc. (USA). The IL4, IL6, TNFα, TGFβ ELISA detection kit was gained from MultiSciences (Lianke) Biotech Co., Ltd. (Hangzhou, China). Mini-Extruder set (Avanti Polar Lipids), and Nuclepore polycarbonate Membranes (1.0μm, 200nm, 400nm) were gained from Sigma-Aldrich (St.Louis, MO, USA). Rabbit CD86, CD206, IL6, TGFβ and β-actin antibodies were obtained from Proteintech Group, Inc (Wuhan, China). Mouse IL4 and TNFα antibodies were purchased from Proteintech Group, Inc (Wuhan, China). Horseradish peroxidase (HRP) conjugated secondary antibody was purchased from Proteintech Group.

The murine RAW264.7 was purchased from the Institute of Biochemistry and Cell Biology, Academy of Science (Shanghai, China). Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), and 100 U/mL penicillin/streptomycin (Thermo Fisher Scientific, USA) at 37°C with 5% CO₂. The macrophage polarization experiment was designed as follows: RAW264.7 cells were exposed to LPS (100 ng/mL) and IL4 (20 ng/mL) for 24h in a serum-free medium to evaluate the cell morphology and polarization level of macrophage (Zhang et al., 2024).

M2NVs were produced through a continuous extrusion approach, as described previously (Pang et al., 2023). Firstly, 1×10⁶ macrophage cells were collected and then resuspended in 1mL PBS (pH 7.4) solution and centrifuged (3500xg, 15 min) to remove cells, apoptotic bodies and cell fragments. Secondly, to obtain the M2NVs, the cell suspensions were extruded through a mini-extruder (Avanti 610000-1EA), with a battery of Nuclepore polycarbonate membrane filters (pore sizes: 1.0 μm, 200nm, 400nm, Avanti 610010-1EA, Avanti 610007-1EA, Avanti 610006-1EA),. After the extrusion, the extruded suspension (1mL) was centrifugated at 10000 g for 10 min to purify M2NVs by removing cell debris and micro-vesicles, then, a centrifugal filter was used to concentrate and purify the M2NVs by centrifugation (1000g, 15 min) for three times. Finally, the filtrate was centrifuged at 12000g for 120 min to obtain M2-phenotype exosomes (M2exo), which were resuspended in PBS. All operations were performed under aseptic at 4°C. The macrophage repolarization experiment groups were designed as follows: Control group, LPS-induced M1 macrophage cells group, M2NVs (10⁶/mL) intervention group.

To observe the morphology of M2NVs, the precipitation after centrifugation was fixed in 2.5% glutaraldehyde overnight at 4°C, then washed with PBS immersion at least three times before being viewed via transmission electron microscope (JEOL Ltd., Tokyo, Japan).

The M2NVs’ particle size, distribution, and concentration were detected via the nanoparticle tracking analysis (NTA) as follows: the M2NVs suspension was diluted to 1 mL volume with distilled water, and then immediately transferred to a 1 mL volume syringe, subsequently, the particle size, distribution and concentration of M2NVs were detected by NTA Nano-Sight NS300 detectors (NanoSight, Malvern, United Kingdom).

SD rats were obtained from the Experimental Animal Center of Chengdu Dossy Experimental Animals Co., Ltd. (Sichuan, China). All animal experiments were approved by the Institutional Animal Care and Use Committee of Chengdu Women’s and Children’s Central Hospital (protocol number: 2022(145)-192). 3-day-old rat pups were fed by gavage with hyperosmolar formula (every 4 h) [Abbott Nutrition infant formula (USA): Esbilac (PetAg, USA) milk replacer, 2:1]. Rats were simultaneously exposed to at 4°C for 10 min and hypoxic conditions (5% O2, 95% N2) 3 times per day last four days. In addition, to establish the NEC model, the rats were also fed LPS (200μl, 100 ng/mL) daily, and the M2NVs intervention group was given intraperitoneal injection of M2NVs (300μl, 10⁶/mL) once a day until the day before death for building the NEC model. The untreated and age-matched neonatal rats lived with their mothers as the control group. On postnatal day 7, the intestines were harvested for further analysis.

Total proteins from the intestine tissue or RAW264.7 cells were extracted using RIPA buffer added with protease inhibitors. Protein samples (20 μg) were quantified for analysis by BCA protein assay kit. Then, the protein samples were electrophoresed in 8–12% SDS-PAGE gels and transferred to PVDF membranes. After incubation with the respective primary antibodies and corresponding secondary antibodies, the target protein signal was visualized using an enhanced chemiluminescence (ECL) system and analyzed using the Image J software (NIH, Bethesda, MD, USA). The specific protein bands were quantified and normalized to β-actin.

After the corresponding treatment, the single-cell suspension made from the cultured cells was incubated with CD86 and CD206 according to the manufacturer’s protocol. The expression level of surface markers of macrophage cells was assessed using flow cytometry analysis using a BD FACS.

Each group of cell culture supernatant was collected and centrifugated to remove the cell debris, and then the concentrations of inflammatory cytokines, including IL-4, IL-6, TNFα, and TGF-β were measured in the purified culture supernatants using commercial ELISA kits according to the manufacturer’s protocols.

The intestinal section slices were immersed in hematoxylin for 15 minutes After dewaxing and hydration, and subsequently, immersed in eosin for 7 minutes after washed in distilled water. After being rewashed in distilled water, dehydrated, and mounted. The sections were observed, and digital images were obtained under light microscopy (Olympus). The control group was compared using the slices in equal levels.

The tissue slices and cultured cell samples were immersed in 4% paraformaldehyde for 20 min, then blocked in goat serum for 30min and incubated overnight with anti-TGFβ and anti-TNFα (1:200) primary antibody. The next day, cells were incubated for 1h with the corresponding secondary antibody (FITC-green, CY3-red) (1:250). Images were finally generated under fluorescence microscopy (Leica).

The weighed intestinal tissue was centrifuged at 14,000 g for 20 min at 4°C after being dried lyophilized and ground for 1 min at 65 Hz in a Grinding Mill and placed for 1 h ultrasonic shaking in ice baths with 1 mL precooled mixtures of methanol, acetonitrile and water (v/v/v, 2:2:1), then the supernatants were recovered and concentrated under vacuum until dry to extract metabolites. The extracted metabolites were analyzed using a UPLC-ESI-Q-Orbitrap-MS system (UHPLC, Shimadzu Nexera X2 LC-30AD, Shimadzu, Japan) coupled with Q-Exactive Plus (Thermo Scientific, San Jose, USA) for metabolomics profiling. For liquid chromatography (LC) separation, the samples were analyzed using a ACQUITY UPLC^® HSS T3 column (Waters, Milford, MA, USA) to detect liquid chromatography (LC) separation, the flow rate was 0.3 mL/min and the mobile phase contained: A: 0.1% FA in water and B: 100% acetonitrile (ACN). HESI (electrospray ionization)-MS data was acquired in positive/negative mode over m/z 70-1050, full MS resolution was 70,000 at m/z 200, and 17,500 at m/z 200 for MS/MS scan; MS/MS used a 2 m/z window, stepped NCE (20,30,40), and 17,500 resolution. Models were built on principal component analysis (PCA), the p value was calculated by one-way analysis of variance (ANOVA) for multiple groups analysis, metabolites with VIP values greater than 1.0 and p value less than 0.05 were considered to be statistically significant metabolites. The KEGG pathway analysis was used to identify the perturbed biological pathways via using the KEGG database (http://www.kegg.jp). Finally, multivariate data analysis is used for correlation data pre-processing and filtering.

Data are presented as the mean ± SD of at least four independent experiments. All statistical analyses were performed using the SPSS 21.0 software (IBM Corp., Armonk, NY, USA). An unpaired t-test was used for comparison of normal distribution and homogeneous variance samples and differences were determined using one-way analysis of variance (ANOVA) for more than two groups and followed by a Student-Newman-Keuls test for multiple comparisons. P < 0.05 or P < 0.01 was considered statistically significant.

Macrophages originating from the mouse mononuclear macrophage cell line RAW264.7 (M0) are a widely established and used system to estimate macrophage polarization mechanisms. Thus, to study the polarization of macrophages in vitro, we treated RAW264.7 cells with LPS to polarize into M1-type macrophages and IL4 into M2-type macrophages, which allows us to differentiate cell populations (M0/M1/M2) under the microscope, as shown in Figure 1A , under the resting conditions, M0 cells were round and small with clear edges, M1 cells showed the irregular polygonal shape such as spindle-shaped, enlarged or fibrous, whereas, M2 cells were small and rounded like those of M0 cells with low adhesion. In addition, to further verify types of macrophage polarization, we evaluated the expressions of M1 macrophage marker CD86 and M2 macrophage marker CD206 in vitro via flow cytometry, results have shown that LPS-induced macrophage polarization mostly clustered in CD86-positive regions (88.41%) and M2-type polarization enhanced by IL4 mostly concentrated in the CD206-positive region (85.06%) ( Figure 1B ). To further verify the type of macrophage polarization, we also conducted western blot experiments to identify the surface markers. We found that the CD86 expression (a marker for M1 macrophages) significantly increased after LPS induction (3.46-fold, p < 0.01), while the CD206 expression (a marker for M2 macrophages) was not pronounced. In contrast, the results were reversed (1.78-fold, p < 0.01) in the IL4-induced macrophage polarization group, as shown in Figure 1C .

Furthermore, we also examined the inflammatory cytokines expression levels in LPS/IL4-induced RAW264.7 cells. As anticipated, we found that the up-regulated expressions of pro-inflammatory cytokines (IL6, TNFα) in the LPS treatment group (p < 0.01) and the up-regulated expressions of anti-inflammatory cytokines (IL4, TGFβ) in the IL4 treatment group (p < 0.01) compared with the M0 group ( Figures 1D–G ) through ELISA assay. In addition, we also detected the expression of the proinflammatory factor TNFα and the anti-inflammatory factor TGFβ by cellular immunofluorescence assay and found that TNFα expression increased significantly and the expression of TGFβ was not obvious after LPS induction, while the result was opposite after IL4 induction as shown in Figure 1H . Taken together, these results implied that macrophage cells could polarize from M0 to M1 or M2, exhibit their characteristics, and release corresponding inflammatory factors.

To obtain M2NVs, a convenient and continuous extrusion system was established to execute and purify M2NVs from RAW264.7 macrophage cells. To detect the particle size, distribution and concentration, the NTA technique was applied and showed that the M2NVs are regular and granular, the average size of M2NVs was 149.5 nm, the percentage was 89.8%, and the concentration was 2.5*10⁶/mL( Figure 2A ) which suggested that our method could produce the homogeneous size and moderate concentration of M2NVs. Besides that, TEM images of M2NVs showed that the size of those vesicles ranged from 100 to 150nm and presented as a regular circle with intact spherical membrane structure ( Figure 2A ).

To further identify whether the M2NVs have an anti-inflammatory function and whether those vesicles could repolarize M1 macrophages into M2 macrophages, the M2NVs were co-cultured with LPS-induced M1 macrophages, then we detected the associated macrophage markers. As shown in Figures 2B,C , both flow cytometry and western blot results showed that the M2NVs could alleviate M1 cells’ surface marker CD86 and up-regulate the M2 surface markers CD206 (p < 0.01). Subsequently, we further examined the expression levels of inflammatory factors through ELISA assay and demonstrated that M2NVs could mitigate the expressions of pro-inflammatory cytokines (IL6, TNFα) and enhance the expressions of anti-inflammatory cytokines (IL4, TGFβ)( Figures 3A–D ). Meanwhile, we used cellular immunofluorescence assay to detect the expression of the proinflammatory factor TNFα and the anti-inflammatory factor TGFβ and found that TNFα expression decreased significantly and the expression of TGFβ was up-regulated after M2NVS intervention as shown in Figure 3E . Collectively, these results demonstrated that M2NVs could not only repolarize M1 macrophages into M2 macrophages but also have an anti-inflammatory function like M2 macrophage cells.

To verify whether M2NVs have the function of therapeutic repair effect in vivo, we constructed a neonatal rat NEC model. As shown in Figure 4A , we found that rats in the control group were well-developed with intact intestinal tissue, and the intestinal tissue was well-elastic and non-congestion with non-bubbles, while in the NEC group, almost all intestinal tissues were black and red, the intestinal tubes were fragile and some intestinal segments were dilated, as expected, intestinal tissue damage was significantly reduced in the M2NVs treatment group, and the intestinal color was close to normal, with only local discoloration with no obvious dilatation on the intestinal wall. In addition, HE staining and pathological examination were used to determine intestinal microscopic changes and showed that the villous intestinal epithelial cells in the NEC group were disordered with submucosal edema and local villi shedding, however, in the M2NVs treatment group, the disorder of villous epithelial, submucosal edema and villi shedding showed improvement ( Figure 4B ).

Subsequently, we examined whether M2NVs have the function of repolarizing M1 into M2-type macrophages, the M1 cells’ surface marker CD86 and M2 surface markers CD206 were detected, and results showed that M2NVs could reduce CD86 expression effectively (decline rate was 42.5%, p < 0.01) and intensify CD206 expression after M2NVs intervention (4.34-fold, p < 0.01) ( Figure 4C ). Furthermore, to further verify whether M2NVs have the role in reducing inflammatory symptoms, we also examined the expression of the proinflammatory factor TNFα and the anti-inflammatory factor TGFβ via cellular immunofluorescence assay and found that TNFα expression decreased significantly and the expression of TGFβ was up-regulated after M2NVS intervention as shown in Figure 5A . Also, the associated intestinal inflammatory factors were determined by ELISA kit and found that M2NVs intervention alleviated the pro-inflammatory cytokines expressions (IL6, TNFα) and strengthened the anti-inflammatory cytokines expressions (IL4, TGFβ) in vivo (p < 0.01, Figures 5B–E ). These data also confirmed that M2NVs could effectively repolarize M1 into M2-type macrophages, slow or improve NEC progression, reduce inflammatory response, and alleviate intestinal damage and histological changes.

To analyze the changes in the intestinal microenvironment underlying the biological effects of M2NVs, differential analysis of intestinal metabolites was performed in the intestinal tissues of WT and NEC rats. In our animal model, the variation of the sample concentration within the group is small but the variation between the groups is obvious, indicating that our model is stable, reliable and has better predictability ( Figure 6A ). 423 differentially expressed metabolites (P.value < 0.05) were produced into a volcano map to show the up- and down-regulated differential metabolites ( Figure 6B ). Metabolites that had the most significant differences (top 300) associated with microenvironment change were then selected and analyzed in the heatmap and a complex heat map was produced based on these target metabolites (top 50) shown in Figures 6C, D and found that there were indeed differences in gut metabolites between normal and NEC rat models. Then, KEGG signaling pathway analyses were carried out on the metabolites with the most obvious differences and found that the membrane transport system and digestive system play an important role in the change of intestinal microenvironment in NEC ( Figure 6E ).

Meanwhile, we also screened for differences in metabolites among WT, NEC and M2NVs-treated groups, quality analysis showed that our models are stable and predictive ( Figure 7A ). In Figures 7B, C showed that the metabolites that had the most significant differences (top 300) and target metabolites (top 50) associated with microenvironment change were selected and analyzed in the heatmap and complex heatmap and found that there were differences in gut metabolites among these groups. Therewith, KEGG enrichment analysis was performed to identify the most significantly altered signaling pathways after the NEC and M2NVs-treated group, the similarity among these comparisons was the membrane transport system and the digestive system included in the intestinal microenvironment changes ( Figure 7D ). Differential metabolite abundance analysis revealed that 16-Hydroxy-l0-oxohexadecanoic acid, N-Acetylneuraminic acid, 1-methyl-4-[(4-methyl phenyl)amino]-3-nitrohydroquinolin-2-one, Vincristine have significant differences in intestinal differential metabolites (p < 0.01, Figures 7E–H ), which also indicated that M2NVs not only have anti-inflammatory effects but also participate in NEC intestinal environment changes. Generally, the metabolomics analysis results implied that M2NVs intervention could influence the inflammation and membrane transport-related pathways in the intestinal tissues of NEC rats.