Experimental animals

The male ICR mice (8 weeks old, weighing 38 ±3 g) were purchased from Hunan SJA Laboratory Animal Co., Ltd (Changsha, China). All mice were housed in the animal center laboratory under controlled environmental conditions: temperature 24 ±1°C, humidity 60 ±5%, and a 12-hour light/dark cycle. Standard chow and water were provided ad libitum. All animal procedures were approved by the Animal Ethics Committee and conducted in accordance with the 3Rs principle (replacement, reduction, and refinement).

Bioinformatics analysis

The sepsis dataset GSE137342 was downloaded from the Gene Expression Omnibus database. The raw count matrix of GSE137342 was analyzed using the DESeq2 package (v1.38.3) in R (v4.3.1). Differential expression analysis was performed using the Wald test based on a negative binomial generalized linear model, and p values were adjusted for multiple testing using the Benjamini-Hochberg method. Differentially expressed genes (DEGs) were identified using the criteria P adj. < 0.01 and logFC > 2. RNA-binding protein (RBP) genes related to RNA stability were obtained from the CLIPdb database (http://111.198.139.65/RBP.html). DEGs were intersected with RBP genes to identify core RBPs. Subsequently, downstream targets of the core RBP were predicted using CLIPdb and further intersected with DEGs. The intersected genes were ranked in ascending order based on the adjusted p values (P adj.) using Excel software, and the most significantly altered key gene in sepsis was identified.

Establishment of a sepsis mouse model

A total of 32 mice were randomly divided into four groups (n = 32): sham, model, sh-NC, and sh-Ago2, with 8 mice in each group. The sepsis mouse model was established using CLP surgery. Mice were fasted for 12 hours prior to surgery and anesthetized with isoflurane (792632, Sigma-Aldrich, Shanghai, China). In the sham-operated group (sham), an abdominal incision was made to expose the distal cecum and mesentery without further intervention, followed by abdominal closure. In the model group (model), CLP surgery was performed. Specifically, after routine disinfection, a 2 cm midline incision was made in the abdomen to expose the cecum. The distal cecum was separated from the mesentery while avoiding vascular injury. The cecum was ligated three-fourths of the way from its distal end using sterile 4-0 sutures. A sterile 7-gauge needle was used to puncture the cecal wall between the ligation site and the cecal tip. After perforation, the abdominal cavity was closed layer by layer [20]. The model group was subdivided into three subgroups: model (untreated), sh-Ago2, and sh-NC. In the sh-Ago2 and sh-NC groups, shRNA AAV9 viral vectors targeting AGO2 (sh-Ago2) and control shRNA (sh-NC) were injected via the tail vein one week before the CLP procedure (shRNA lentiviruses were constructed by Shanghai GeneChem Co., Ltd.). At 48 hours after CLP, blood serum was collected, and mice were euthanized with an overdose of sodium pentobarbital (100 mg/kg). Lung and liver tissues were then harvested. Collected samples were used for subsequent experimental analyses.

Construction of a sepsis cell model

RAW264.7 cells (CL-0190, Procell, Wuhan, China) were cultured in Dulbecco’s modified Eagle medium (12491015, Gibco, Shanghai, China) supplemented with 10% fetal bovine serum (F0193, Sigma, Shanghai, China), 100 mg/ml streptomycin, and 100 µg/ml penicillin (15070063, Gibco, Shanghai, China). When cell confluence reached 70-80%, plasmids including sh-NC, sh-Ago2, sh-Atox1, oe-NC, and oe-Atox1 (synthesized by GeneChem Co., Ltd., Shanghai, China) were transfected into RAW264.7 cells using Lipofectamine 3000 (L3000001, Thermo Fisher, Shanghai, China). After 48 hours of transfection, AGO2 and ATOX1 expression levels were assessed using reverse transcription quantitative polymerase chain reaction (RT-qPCR) to confirm transfection efficiency. Subsequently, cells were stimulated with 1 µg/ml LPS for 48 hours to construct the sepsis cell model [20]. Cells were divided into 8 groups based on treatments: phosphate buffer saline (PBS; control), LPS, LPS + sh-NC, LPS + sh-Ago2, LPS + sh-Atox1, sh-NC + oe-NC, sh-Ago2 + oe-NC, and sh-Ago2 + oe-Atox1. After 48 hours, cells and supernatants were collected for subsequent analyses.

Hematoxylin and eosin staining

Mouse lung and liver tissues were fixed in 4% paraformaldehyde solution (441244, Sigma, Shanghai, China), embedded in paraffin, and sectioned. The paraffin sections were cleared with xylene (534056, Sigma), followed by dehydration through a graded ethanol series and immersion in distilled water for 2 minutes. The sections were stained with ematoxylin and eosin (HE) staining solutions (G1120, Solarbio, Shanghai, China) for 3 minutes and 1 minute, respectively. After staining, the sections were dehydrated with graded ethanol and cleared with xylene. Finally, the sections were mounted and observed under an optical microscope (XDS-30, BINGYU, Shanghai, China).

Biochemical analysis for aspartate aminotransferase and alanine aminotransferase levels

Serum aspartate aminotransferase (AST) levels were measured using the AST Assay Kit (MAK467, Sigma). Detection reagents were prepared according to the manufacturer’s instructions. A 96-well plate was loaded with 20 µl of serum (with NADH standard and 20 µl of H₂O serving as calibration and blank controls, respectively) and 200 µl of detection reagent. After mixing, the plate was incubated at 37°C. Optical density (OD) values at 340 nm were recorded at the 5^th and 10^th minutes using a microplate reader (168-1130, Bio-Rad, USA).

Serum alanine aminotransferase (ALT) levels were measured using the ALT Activity Assay Kit (MAK052, Sigma). Briefly, 20 µl of serum was added to a 96-well plate, and reaction solutions were prepared according to the kit instructions. Subsequently, 100 µl of reaction solution was added to each sample, positive control, and standard well. After mixing, the plate was left to stand for 2 minutes. OD values were measured at 570 nm at 5-minute intervals until the sample well OD values exceeded the standard well values.

Enzyme-linked immunosorbent assay for TNF-α, IL-6, and IL-1β levels

Enzyme-linked immunosorbent assay (ELISA) kits specific for mouse TNF-α (88-7324-88, Thermo Fisher, Shanghai, China), IL-6 (88-7064-88, Thermo Fisher, Shanghai, China), and IL-1β (88-7013-22, Thermo Fisher, Shanghai, China) were used to measure the levels of these cytokines in serum or cell culture supernatants. In brief, 100 µl of serum or standard solution was added to ELISA plates pretreated with capture antibody buffer and ELISA/ELISPOT dilution buffer (1×). Plates were incubated at room temperature for 2 hours. After washing, 100 µl of detection antibody was added, followed by incubation at room temperature for 1 hour. After another washing step, 100 µl of Streptavidin-HRP was added, and the plates were sealed and incubated at room temperature for 30 minutes. Following a final wash, 100 µl of stop solution was added, and OD values were recorded at 450 nm using a microplate reader.

Western blot analysis of protein expression

Proteins from tissues and cells were extracted using radioimmunoprecipitation assay lysis buffer (R0278, Sigma, Shanghai, China), and protein concentration was measured using the BCA method (A55861, Thermo Fisher, Shanghai, China). After separating the proteins by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, they were transferred to polyvinylidene difluoride membranes and blocked with 5% non-fat milk for 1 hour. The membranes were then incubated overnight at 4°C with primary antibodies against AGO2 (1 : 1000, ab186733, Abcam, Shanghai, China), ATOX1 (1 : 500, 31090-1-AP, Proteintech, Wuhan, China), inducible nitric oxide synthase (iNOS; 1 : 1000, ab178945, Abcam, Shanghai, China), arginase 1 (Arg1; 1 : 1000; ab315110, Abcam), and β-actin (1 : 1000; ab8227, Abcam). After washing, the membranes were incubated with HRP-conjugated goat anti-rabbit immunoglobulin G (IgG; 1 : 2000; ab205718, Abcam) for 1 hour at room temperature. The protein bands were visualized using enhanced chemiluminescent reagent (MA0186, Meilunbio, Liaoning, China). Relative protein expression levels were quantified using ImageJ software (V1.8.0.112, NIH, Madison, WI, USA), with β-actin as the internal control. Western blot analysis was performed across all experimental groups to comprehensively evaluate the expression levels of various relevant proteins under different treatment conditions.

RT-qPCR for gene expression

Total RNA from RAW264.7 cells was extracted using TRIzol reagent (15596026CN, Thermo Fisher, Shanghai, China), and cDNA was synthesized from RNA using the Hifair AdvanceFast 1st Strand cDNA Synthesis SuperMix for qPCR kit (11156ES10, Yesen, Shanghai, China). Gene amplification was performed using the Hieff UNICON Universal Blue qPCR SYBR Green Master Mix kit (11184ES08, Yesen, Shanghai, China) on a LightCycler 480 real-time PCR instrument (Roche, Basel, Switzerland). The PCR conditions were as follows: 95°C for 30 s; 95°C for 3 s; 60°C for 20 s; 40 cycles, with a default dissociation curve setting. Relative gene expression was calculated using the 2^-ΔΔCT method with β-actin as the internal reference. The primers used are listed in Table 1. RT-qPCR was performed in LPS-induced RAW264.7 cells, including the AGO2-silenced groups (sh-NC, sh-Ago2-1, sh-Ago2-2) and the ATOX1-silenced groups (sh-NC, sh-Atox1-1, sh-Atox1-2), to assess gene expression.

RNA half-life assay

To assess the stability of ATOX1 mRNA under AGO2 knockdown, cells were treated with the transcription inhibitor actinomycin D (2 mg/ml; SBR00013, Sigma, Shanghai, China). RNA was extracted from cells at 0, 2, 4, and 6 hours after treatment and analyzed via RT-qPCR. ATOX1 mRNA levels were normalized to β-actin, and the relative expression at each time point was calculated.

RNA immunoprecipitation (RIP-qPCR) assay

RIP-qPCR analysis was performed using the Smart-RIP RIP Kit (RIP-1001, Engibody, Dover, America). Cell lysates were incubated overnight at 4°C with magnetic beads conjugated to either AGO2 or IgG antibodies. After magnetic separation and removal of the supernatant, beads were washed and resuspended in lysis buffer to collect RNA-protein complexes. RNA was extracted according to the kit instructions, and ATOX1 mRNA enrichment levels were quantified via RT-qPCR.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 9. Quantitative data were expressed as mean ± standard error of the mean (SEM). A t-test was used for comparisons between two groups, while one-way or two-way analysis of variance (ANOVA) followed by Tukey’s post hoc test with adjustment for multiple comparisons was applied for comparisons among three or more groups. In the case of nonparametric data, the Kruskal-Wallis test followed by Dunn’s multiple comparisons test with Bonferroni correction was used. A value of p < 0.05 was considered statistically significant.

AGO2 mediates macrophage polarization imbalance to exacerbate the sepsis-induced inflammatory response

A total of 194 DEGs were identified from the GSE137342 dataset (Fig. 1A). These DEGs were intersected with 71 RBP genes associated with RNA stability, revealing AGO2 (Fig. 1B), which was significantly upregulated in sepsis (Fig. 1C). To assess the role of AGO2 in the inflammatory response during sepsis, we established a sepsis mouse model. The results showed that, compared to the sham group, the model group exhibited damaged lung tissue with thickened alveolar septa. Liver tissue showed cell necrosis and increased infiltration of inflammatory cells (Fig. 1D), along with significantly elevated levels of the serum liver injury markers AST and ALT (Fig. 1E), indicating organ damage caused by sepsis. Additionally, levels of the serum pro-inflammatory cytokines TNF-α, IL-6, and IL-1β were significantly elevated, further exacerbating the inflammatory response (Fig. 1F). These findings confirmed the successful establishment of the sepsis mouse model.

Fig. 1

AGO2 mediates macrophage polarization imbalance to exacerbate sepsis-induced inflammatory response. A) Volcano plot of differential analysis from the GSE137342 dataset; B) Intersection of DEGs in sepsis with RBP genes; C) Expression of AGO2 in the sepsis dataset GSE137342; D) HE staining of lung and liver tissue showing damage in mice (100×); E) Biochemical analysis of AST and ALT levels in serum. N = 8, ****p < 0.0001

Fig. 1. Cont.

F) ELISA detection of serum TNF-α, IL-6, and IL-1β levels; G, H) Western blot analysis of AGO2, iNOS, and Arg1 expression in mouse lung (G) and liver tissues (H). N = 8, ****p < 0.0001

Further analysis revealed that in the model group, the expression of iNOS (a marker for M1 polarization) was significantly upregulated, while Arg1 (a marker for M2 polarization) was downregulated. Concurrently, AGO2 expression was significantly elevated in the lung and liver tissues of septic mice, consistent with bioinformatics analysis results (Fig. 1G, H). These findings suggest that AGO2 may contribute to the progression of sepsis by influencing macrophage polarization imbalance.

Silencing AGO2 regulates macrophage polarization and inhibits the LPS-induced inflammatory response

To investigate whether AGO2 mediates the LPS-induced inflammatory response through the regulation of macrophage polarization, we constructed a sepsis cell model with sh-Ago2. RT-qPCR was used to verify the silencing efficiency of AGO2, and the sh-Ago2-1 construct with the highest silencing efficiency was selected for further experiments (Fig. 2A). The results showed that, compared to the PBS group, LPS stimulation significantly upregulated the expression of AGO2 and iNOS, while downregulating Arg1 expression.

Fig. 2

Silencing AGO2 regulates macrophage polarization and inhibits the LPS-induced inflammatory response. A) RT-qPCR to verify the silencing efficiency of sh-Ago2; B) RT-qPCR detection of AGO2 expression in cells; C) Western blot analysis of AGO2, iNOS, and Arg1 expression in cells; D) ELISA detection of TNF-α and IL-6 levels in cell culture supernatants. Cell experiments were performed in triplicate. **p < 0.01, ***p < 0.001, ****p < 0.0001

However, silencing AGO2 reversed these gene expression changes, promoting macrophage polarization towards the M2 phenotype (Fig. 2B, C). Furthermore, LPS treatment significantly increased the levels of TNF-α and IL-6 in the cell culture supernatant, while AGO2 knockdown effectively reduced these inflammatory cytokine levels (Fig. 2D).

These findings suggest that downregulation of AGO2 expression can correct macrophage polarization imbalance, thereby inhibiting the LPS-induced inflammatory response.

Silencing ATOX1 regulates macrophage polarization and inhibits the LPS-induced inflammatory response

To further explore the specific regulatory mechanism of AGO2, we intersected 15,487 downstream targets of AGO2 predicted from the CLIPdb database with the sepsis DEGs, resulting in the identification of 151 common genes (Fig. 3A). We then selected the top 10 significantly differentially expressed intersecting genes with the lowest P adj. values: MYL6, UBASH3B, TPD52L2, GSTO1, HIST1H2BK, NAPRT, ATOX1, TPST2, MRPL36, and TMEM256. Among these, ATOX1 was highly expressed in sepsis (Fig. 3B). As expected, ATOX1 expression was significantly upregulated in the lungs and livers of model group mice, further supporting our bioinformatics analysis results (Fig. 3C, D).

Fig. 3

Silencing ATOX1 regulates macrophage polarization and inhibits the LPS-induced inflammatory response. A) Intersection of sepsis DEGs and AGO2 downstream targets; B) Expression of ATOX1 in the sepsis dataset GSE137342; C, D) Western blot analysis of ATOX1 expression in lung (C) and liver tissues (D) from septic mice (N = 8); E) RT-qPCR detection of sh-Atox1 silencing efficiency; F, G) RT-qPCR (F) and Western blot (G) detection of ATOX1 expression in cells; H) Western blot detection of iNOS and Arg1 expression in cells; I) ELISA detection of TNF-α and IL-6 levels in cell culture supernatant. Cell experiments were performed in triplicate. ***p < 0.001, ****p < 0.0001

Fig. 3. Cont.

G) RT-qPCR (F) and Western blot (G) detection of ATOX1 expression in cells; H) Western blot detection of iNOS and Arg1 expression in cells; I) ELISA detection of IL-6 and TNF-α levels in cell culture supernatant. Cell experiments were performed in triplicate. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Subsequently, we constructed the sh-Atox1 sepsis cell model to investigate whether ATOX1 influences the inflammatory response through the regulation of macrophage polarization. RT-qPCR identified the sh-Atox1-2 construct with the highest silencing efficiency for subsequent experiments (Fig. 3E). The results showed that ATOX1 expression was significantly elevated in LPS-treated cells compared to the PBS group. Silencing ATOX1 dramatically reduced its expression levels (Fig. 3F, G). At the same time, silencing ATOX1 significantly inhibited iNOS expression and promoted Arg1 expression (Fig. 3H). Moreover, silencing ATOX1 also significantly reduced TNF-α and IL-6 levels in the cell culture supernatant, thus alleviating the inflammatory response (Fig. 3I). These findings indicate that ATOX1 plays an important role in the LPS-induced inflammatory response by mediating macrophage polarization imbalance.

AGO2 enhances ATOX1 mRNA stability

To elucidate the regulatory mechanism between AGO2 and ATOX1, we conducted several experiments. The results showed that, compared to the LPS + sh-NC group, the expression of ATOX1 was significantly downregulated in the LPS + sh-Ago2 group (Fig. 4A). We then treated the cells with actinomycin D and found that silencing AGO2 significantly shortened the half-life of ATOX1 mRNA, reducing its stability (Fig. 4B). Additionally, we performed RIP-qPCR experiments to further examine the interaction between AGO2 and ATOX1 mRNA. The results indicated that AGO2 directly binds to ATOX1 mRNA (Fig. 4C). This finding not only confirms the direct interaction between AGO2 and ATOX1 but also reveals the molecular mechanism by which AGO2 enhances ATOX1 mRNA stability to upregulate its expression.

Fig. 4

AGO2 enhances ATOX1 mRNA stability. A) Western blot analysis of ATOX1 expression in cells; B) Actinomycin D treatment to measure the half-life of ATOX1 mRNA; C) RIP-qPCR analysis of AGO2 binding to ATOX1 mRNA. Cell experiments were performed in triplicate. ***p < 0.001, ****p < 0.0001

AGO2 promotes macrophage polarization imbalance and aggravates the inflammatory response by enhancing ATOX1 mRNA stability

We further assessed whether AGO2 regulates macrophage polarization and inflammation through ATOX1. To this end, we overexpressed ATOX1 in the AGO2 knockdown septic cell model. The results showed that AGO2 knockdown significantly inhibited the expression of ATOX1 and iNOS, while upregulating the expression of Arg1. However, overexpression of ATOX1 restored the levels of ATOX1, iNOS, and Arg1, without significantly affecting AGO2 expression (Fig. 5A). Additionally, AGO2 silencing reduced levels of the inflammatory factors TNF-α and IL-6, and ATOX1 overexpression counteracted the inhibitory effect of AGO2 silencing (Fig. 5B). These results suggest that AGO2 enhances ATOX1 mRNA stability and upregulates ATOX1 expression, promoting M1 macrophage polarization and aggravating the inflammatory response.

Fig. 5

AGO2 promotes macrophage polarization imbalance and aggravates the inflammatory response by enhancing ATOX1 mRNA stability. A) Western blot analysis of AGO2, ATOX1, iNOS, and Arg1 expression in cells; B) ELISA measurement of TNF-α and IL-6 levels in cell culture supernatant. Cell experiments were performed in triplicate. **p < 0.01, ***p < 0.001, ****p < 0.0001

AGO2 aggravates the sepsis-induced inflammatory response by mediating macrophage polarization imbalance through ATOX1

To further validate the impact of the AGO2–ATOX1 axis on macrophage polarization and inflammation in the septic mouse model, we administered sh-Ago2 lentivirus via tail vein injection. The experimental results showed that, compared to the sh-NC group, the sh-Ago2 group exhibited reduced lung and liver tissue damage, decreased inflammatory cell infiltration, and significantly lower serum levels of AST and ALT (Fig. 6A, B), indicating that silencing AGO2 alleviates organ damage caused by sepsis. Moreover, AGO2 knockdown significantly reduced the levels of TNF-α, IL-6, and IL-1β in the serum, effectively inhibiting the inflammatory response in septic mice (Fig. 6C). Further analysis revealed that the expression of iNOS, AGO2, and ATOX1 in the tissues of the sh-Ago2 group was significantly reduced, while the expression of Arg1 was significantly upregulated, suggesting a shift of macrophages toward M2 polarization (Fig. 6D, E). Taken together, AGO2 upregulates ATOX1 expression, promotes M1 macrophage polarization, and mediates polarization imbalance, thereby exacerbating the inflammatory response in septic mice.

Fig. 6

AGO2 aggravates the sepsis-induced inflammatory response by mediating macrophage polarization imbalance through ATOX1. A) HE staining to observe pathological changes in mouse lung and liver tissues (100×); B) Biochemical analysis of serum AST and ALT levels; C) ELISA detection of serum TNF-α, IL-6, and IL-1β levels. N = 8, ****p < 0.0001

Fig. 6. Cont.

AGO2 aggravates the sepsis-induced inflammatory response by mediating macrophage polarization imbalance through ATOX1. D, E) Western blot analysis of iNOS, Arg1, AGO2, and ATOX1 expression in mouse lung (D) and liver tissues (E). N = 8, ***p < 0.001, ****p < 0.0001