PRC1 promotes immunosuppressive macrophages in sepsis via β-catenin/STAT3 signaling

Abstract

Background

Immunosuppression is a distinctive condition resulting from sepsis, marked by impaired immune response and immune dysregulation, with a poor prognosis. PRC1, a mitotic regulatory protein, is associated with immune suppression within the tumor microenvironment. However, the role of PRC1 in septic immunosuppression remains unclear. This research aimed to explore the implication and potential mechanism of PRC1 in septic immunosuppression.

Methods

Dataset GSE95233 and GSE65682 were used to validate the expression and prognostic value of PRC1 in sepsis patients. LPS was used to stimulate naïve or endotoxin-tolerant THP-1 and BMDMs. PRC1 expression was measured in by RT-qPCR and Western blot. Small interfering RNA was used for PRC1 knockdown in THP-1. The phosphorylated STAT3 and active β-catenin was detected by Western blot. The expression levels of cytokines and surface markers of macrophages were validated by RT-qPCR. β-catenin inhibitor MSAB and agonist SKL2001 were used to explore the functional relationship among relevant molecules.

Results

PRC1 expression was increased in sepsis non-survivors in both dataset GSE95233 and GSE65682, and increased PRC1 expression was associated with increased 28-days septic mortality. PRC1 expression was elevated in endotoxin-tolerant macrophages rather than naïve macrophages. Sustained phosphorylation of STAT3 was detected in endotoxin-tolerant macrophages. Increased PRC1 expression maintained the phosphorylated STAT3 level via a β-catenin-dependent mechanism, which was reversed by β-catenin inhibitor MSAB. PRC1 knockdown could reduce STAT3 phosphorylation and restore inflammatory responses in endotoxin-tolerant macrophages, while this effect was eliminated by β-catenin agonist SKL2001. Septic microenvironment promoted the expression of PRC1 in endotoxin-tolerant macrophages.

Conclusion

Our data demonstrated that PRC1 is upregulated in endotoxin-tolerant macrophages, and that increased PRC1 expression maintains STAT3 activation via a β-catenin-dependent mechanism and impairs inflammatory response of macrophages during septic immunosuppression. Targeting PRC1/β-catenin/ STAT3 could represent a novel strategy for the management of septic immunosuppression and restore the inflammatory response of endotoxin-tolerant macrophages.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-025-06070-4.

Introduction

Sepsis is a type of immune dysregulation and inflammatory response caused by pathogens, defined as infection accompanied by organ dysfunction, with high incidence and mortality rates [1]. It is estimated that there are approximately 50 million cases of sepsis worldwide in 2017, accounting for 19.7% of global deaths and representing a significant health burden [2]. In China, the estimated incidence rate of hospitalized sepsis are approximately 420 cases per 100,000 with a substantial burden and significant regional variations, presenting a persistent upward trend [3]. The lack of therapies to support and reverse host immune dysregulation is a significant factor in the high mortality and poor prognosis associated with sepsis [4]. In the early phases of sepsis, the immune system initiates a rapid immune response to invading pathogens, thereby triggering an inflammatory response that can lead to organs damage and disorder, ultimately resulting in death [5, 6]. Then the persistent inflammation and organ injury, abnormal catabolism, and immune dysregulation leading to a state of chronic immunosuppression and deterioration, rendering it unable to respond effectively to infected pathogens and eventually resulting in death [6–9]. Therefore, a comprehensive understanding of the potential mechanisms underlying septic immunosuppression is imperative for enhancing sepsis management.

Immunosuppression during sepsis is the result of multiple factors, involving metabolic reprogramming and epigenetic changes in monocytes, lymphocytes, neutrophils, and bone marrow-derived suppressor cells [10]. Macrophages, which function as key response cells in the antibacterial defense system, are a primary component of sepsis-associated immunosuppression [8, 11]. Macrophages generally function as the primary line of defense within immune system, phagocytizing and eradicating abnormal pathogens. When pathogens are recognized, pattern-recognition receptors on macrophages are activated by pathogen associated molecular patterns (PAMPs). Then these receptors rapidly initiate downstream signaling transduction, thereby triggering the immune response. However, prolonged exposure to PAMPs has been shown to induce a shift in the phenotype of macrophages towards an immunosuppressive state and presenting endotoxin tolerance, which is characterized by impaired glycolysis, M2-like polarization, defective activation of pro-inflammatory signaling pathways, and increased sensitivity to ferroptosis [12–15].

The immunosuppression of macrophages can exist as a protracted condition [16]. Therefore, the restoration of normal immune function in macrophages is of great significance. A variety of approaches have been attempted to reverse the sepsis-induced immunosuppression of macrophages. For example, exogenous supplementation of interferon-γ was used to enhance the glycolytic function of macrophages, or, using granulocyte-macrophage colony-stimulating factor to restore the expression of major histocompatibility complex II (MHC II) on immunotolerant monocytes [8, 17]. Several approved agents, such as artemether and metformin, also exert certain immune-modulating effects and alleviate immunosuppression following sepsis [18, 19]. However, most potential targets are not feasible in clinical care. The exploring of novel potential therapeutic targets for septic immunosuppression is of critical importance.

Protein regulator of cytokinesis 1 (PRC1), a mitotic regulatory protein, plays a pivotal role in the processes of cytokinesis and microtubule organization [20, 21]. PRC1 is distributed extensively in cell and the increased PRC1 expression is associated with various malignant tumors [22]. The aberrant expression of PRC1 has been demonstrated to promote chromosomal instability and result in a poor prognosis in malignant tumors [23]. However, the potential role of PRC1 in immune function remains unclear. Previous studies have indicated a correlation between high expression of PRC1 and immune suppression within the tumor microenvironment [22, 24, 25]. Microtubule networks can function not only in coordinating cellular behaviors such as deformation and movement, but also in facilitating intracellular transport and signal transduction [26–28]. Previously study has reported that PRC1 is involved in multiple different signaling pathways within cells, including RAS/RAF/MAPK, Wnt/β-catenin, EGFR/VEGFR, RHO GTPase-related pathway [22]. The data from Biogrid dataset (thebiogrid.org) also indicates that PRC1 interacts with multiple proteins, including cGAS, LTBR, IFI30, BTK, MYC, and other genes involved in immune regulation [29]. Therefore, PRC1 may potentially affect the immune function of macrophages. Nevertheless, the capacity of PRC1 to modulate the inflammatory response of the immune system remains to be elucidated.

β-catenin is well recognized for its role in coordinating immune function and inflammatory responses, which vary greatly depending on the disease or cell type [30–33]. It has been demonstrated that β-catenin induces the expression of PD-L1 and promotes immune evasion in tumor cells [34]. The activation of the Wnt/β-catenin signaling pathway can also regulate the process of macrophage polarization, thereby promoting the progression of renal fibrosis [35]. In colorectal cancer, the Wnt/β-catenin signaling pathway induces the polarization of macrophages to the M2 phenotype and promotes cancer metastasis [36]. Activation of the Wnt/β-catenin pathway has been demonstrated to inhibit the NF-κB signaling pathway, thus suppressing inflammation [31]. The association between β-catenin and tumor-associated immunosuppressive microenvironment has been reported [37]. However, the potential role of β-catenin in septic immunosuppression remains unclear.

The objective of this study was to explore the potential role of PRC1 in immunosuppression. In this study, we identify an association between elevated PRC1 expression and an elevated risk of mortality in sepsis datasets GSE65682 and GSE95233. In vitro experiments indicates that the PRC1 expression was increased in endotoxin-tolerant macrophages rather than naïve macrophages. Increased expression of PRC1 promotes active β-catenin accumulation and nuclear translocation, thereby maintaining the activation of signal transducer and activator of transcription 3 (STAT3) in endotoxin-tolerant macrophages and impairing the immune response of macrophages. The knockdown of PRC1 by siRNA in THP-1 reduced the phosphorylation of STAT3 in endotoxin-tolerant macrophages and restored the expression of inflammatory cytokines, such as IL-1β, IL-6 and TNF-α, and surface markers including CD86 and HLA-DR in endotoxin-tolerant macrophages. Our results indicate that PRC1/β-catenin/STAT3 may serve as promising therapeutic targets for sepsis-induced immunosuppression in macrophages to restore immune response.

Materials and methods

Reagents and antibodies

LPS (Cat#L2630) and FITC-conjugated LPS (Cat#F3665) from Escherichia coli O111:B4 were purchased from Sigma-Aldrich (MO, USA). MSAB (Cat#S6901) and SKL2001 (Cat#S8320) were purchased from Selleck Chemicals (TX, USA). Protein A/G Magnetic Beads (Cat#HY-K0202), Mouse M-CSF (Cat#HY-P7085), human IFN-γ (Cat#HY-P7025) and mouse IFN-γ (Cat#HY-P7071) were purchased from MedChemExpress (NJ, USA). RBC lysis buffer (Cat#G2015), cell counting kit-8 assay kit (Cat#G4103), lactic acid assay kit (Cat#G4308), RIPA lysis buffer (Cat#G2002), IP lysis buffer (Cat# G2038), protease inhibitor cocktail (Cat#G2006), PMSF (Cat#G2008) and phosphatase inhibitor cocktail (Cat#G2007) were purchased from Servicebio (Wuhan, China). The BCA protein concentration assay kit (Cat#BL521A) was purchased from Biosharp (Hefei, China). Stattic (Cat#T6308) and SP600125 (Cat#T3109) were purchased from TargetMol (MA, USA).

FastPure Cell/Tissue Total RNA Isolation Kit V2 (Cat#RC112) purchased from Vazyme (Nanjing, China). Mouse IL-4 (Cat#RP01161), human IL-4 (Cat#RP01703), mouse IL-13 (Cat#RP01596), human IL-13 (Cat#RP01320), ABScript III RT Master Mix for qPCR with gDNA Remover (Cat#RK20429) and 2X Universal SYBR Green Fast qPCR Mix (Cat#RK21203) were purchased from ABclonal (Wuhan, China). Nuclear and Cytoplasmic Protein Extraction Kit (Cat#PC204) was purchased from Epizyme Biotech (Shanghai, China). Extracellular acidification rate (ECAR) Fluorometric Assay Kit (Cat#E-BC-F069) was purchased from Elabscience (Wuhan, China). The complete list of antibodies utilized can be found in Table S1.

Public data collection and analysis

The mRNA matrix and clinical information for sepsis patients used in this study were obtained from the Gene Expression Omnibus (GEO) datasets (dataset GSE95233 and GSE65682). Dataset GSE95233 contains 124 samples from 51 septic shock patients (including 34 survivors and 17 death) and 22 healthy volunteers. Samples from GSE95233 were divided into three distinct groups: healthy controls, sepsis survivors, and sepsis non-survivors, based on their survival status, and GEO2R tool was used for differential gene expression analysis. Differentially expressed genes (DEGs) with a |logFC| > 0.5 and an adj. p < 0.05 in the Healthy vs. Death group and Survivor vs. Death group were identified for further enrichment analysis. MCODE method in Metascape tool was then used for enrichment and interactome analysis of the common DEGs [38]. Dataset GSE65682, which contains 802 samples (including 479 sepsis samples and 323 controls) was used to validate the PRC1 expression level and prognostic value. CIBERSORTx, a useful tool to estimate the abundance of different types of immune cells in mixed samples based on transcriptomic data, was used to estimate the immune cell abundance in dataset GSE65682 [39]. Transcriptomic sequencing data of blood monocytes from healthy volunteers and sepsis patients in dataset GSE46955 was then used for mechanism exploration, which contains mRNA matrix of monocytes from 6 healthy donors and 8 sepsis donors.

Animal model

All animals were maintained according to the guidelines outlined in the Guide for the Care and Use of Laboratory Animals. All experiments involving animals were approved by the Animal Ethics Committee of Renmin Hospital, Wuhan University (WDRM202200167). To establish a CLP-induced sepsis model, 8- to 10-week-old C57BL/6 male mice were kept under specific pathogen-free conditions and used for sepsis model. A cecum ligation and puncture (CLP) -induced sepsis model was established according to the standard procedures [40]. In summary, C57BL/6 mice were first anaesthetized using isoflurane. A 1.5 cm incision was made along the midline of the abdomen, followed by an incision of the abdominal cavity to expose the cecum. The cecum was then ligated at the 30% to distal pole, followed by puncturing with a 21G needle and carefully extruding a small drop of feces. Then 1 ml of pre-warmed 37°C saline was administered via subcutaneous injection. To establish the subsequent infection model, fecal suspension intraperitoneal injection was performed as previously reported [41]. Briefly, 10 mg of fresh feces was collected and resuspended with 1 ml of saline. Then fecal suspension was filtered through a 70 μm cell filter. 200 µl fecal suspension was then injected intraperitoneally to induce a subsequent infection. The mice were sacrificed via inhalation of excess isoflurane at a predefined time point following CLP. The lungs were subsequently harvested and preserved with liquid nitrogen or 4% paraformaldehyde for further analysis. Blood samples were collected from the inferior vena cava.

Extraction and induction of BMDM

C57BL/6 mice were euthanized with excess isoflurane and then sacrificed for primary bone marrow-derived macrophages (BMDMs) extraction. In summary, the femur and tibia of C57BL/6 mice were meticulously extracted. Subsequently, the bones were immersed in pre-chilled 75% ethanol for 10 s. The femur and tibia were rinsed thrice in pre-chilled PBS containing 1% penicillin/streptomycin. The two extremities of femur and tibia were then severed. Thereafter, the bone marrow cavity was rinsed with 3 mL of serum-free Dulbecco’s modified Eagle medium (DMEM). A total of 12 mL of bone marrow rinse solution was obtained from both lower limbs. The rinse solution was filtered through a 70 μm cell filter and subjected to centrifugation at 4 °C and 1,000 rpm for 5 min, then the upper layer was discarded. Subsequently, RBC was lysed at room temperature for two minutes. The cells were then collected after centrifugation at 1,000 rpm for 5 min and cultured in complete DMEM supplemented with 10% fetal bovine serum, 1% penicillin/streptomycin and 10 ng/mL M-CSF.

Cell culture and treatment

The human monocytic leukemia cell line THP-1 (Cat#CL-0233) and human Peripheral Blood Macrophages (PBM) (Cat#CP-H264) were kindly provided by Wuhan Pricella Biotechnology Co.,Ltd. THP-1 was cultured in complete RPMI-1640 medium supplemented with 10% fetal bovine serum, 0.05 mM β-mercaptoethanol and 1% penicillin/streptomycin (Pricella, China). PBM was cultured in human complete medium (Pricella, China). Phorbol 12-myristate 13-acetate (PMA, 100 ng/ml) (MedChemExpress, USA) was applied to induce the differentiation of THP-1 cells. Lipofectamine 3000 Transfection Reagent (Invitrogen, USA) and Opti-MEM (Gibco, USA) were used for siRNA and plasmid transfection according to the manufacturer’s instructions. All cells were maintained at 37 °C in a 5% CO₂ humidified incubator. Unless otherwise stated, the cells were challenged with the 100 ng/ml LPS for 6 h.

Lactic acid concentration measurement

Cell culture medium was collected and centrifuged 3,000 rpm for 10 min, then using lactic acid assay kit to determine the lactic acid concentration in the culture medium according to the manufacturer’s instructions.

Cell viability assay

Cell viability was measured via the cell counting kit-8 assay according to the manufacturer’s instructions.

Real-time quantitative PCR

Total RNA was extracted from THP-1 and BMDMs with FastPure Cell/Tissue Total RNA Isolation Kit. Then 1 µg of total RNA was used for cDNA synthesis via ABScript III RT Master Mix for qPCR with gDNA Remover. The standard protocol for cDNA synthesis was as follow: 37℃ for 2 min, 55℃ for 15 min and 85℃ for 5 min. Then using 2X Universal SYBR Green Fast qPCR Mix for RT-qPCR, the standard protocol for RT-qPCR was as follow: initial denaturation at 95℃ for 3 min, then followed by 40 cycles of reaction (95℃ for 5 s and then 60℃ for 30 s). β-actin was used for housekeeping gene and normalization. The 2^−ΔΔCt method was used to measure the relative gene expression levels. Primer sequences were summarized in Table S2.

Separation of nuclear and cytoplasmic proteins

Nuclear and cytoplasmic proteins were separated and extracted using Nuclear and Cytoplasmic Protein Extraction Kit according to the manufacturer’s instructions.

Immunoprecipitation

The cells were lysed with IP lysis buffer containing cocktail, PMSF and phosphatase inhibitor cocktail followed by short-duration low-power sonication (15 W, 5s×3). Cell lysate was incubated overnight at 4℃ with the primary antibody in a rotating incubator, then incubated with protein A/G magnetic beads at room temperature for 1 h. The magnetic beads were then washed and collected for western blotting.

Western blotting

The cells were lysed with RIPA lysis buffer containing cocktail, PMSF and phosphatase inhibitor cocktail followed by sonication on ice. Protein samples were separated with 8–12% SDS-polyacrylamide gradient gels and transferred to PVDF blotting membranes (Cytiva, UK). The membranes were blocked with 5% skimmed milk and then incubated with primary antibodies overnight at 4 °C, followed by incubation with HRP-conjugated secondary antibody at 37 °C for 1 h. The protein expression levels were detected via chemiluminescence and analyzed via Fiji software. β-actin was used for normalization.

Hematoxylin & Eosin (HE) staining

The lungs, livers and kidneys were fixed with 4% paraformaldehyde, embedded in paraffin and sectioned into 5-µm sections. HE staining was subsequently performed as previously described [42].

Lung injury score

The severity of lung injury was evaluated using a semiquantitative scoring system as described previously [43]. Briefly, the lung injury score was assessed on the basis of inflammation, hemorrhage, edema and alveolar septal thickening, and was defined as absent (score = 0), mild (score = 1), moderate (score = 2), severe (score = 3) or very severe (score = 4). The final scores represented the overall assessment of all four parameters.

Immunohistochemical and Immunofluorescence staining

The lung sections were dewaxed with xylene and subsequently hydrated with gradient ethanol solutions. Then antigens were retrieved in citric acid buffer (pH 6.0). H₂O₂ (3%) and QuickBlock™ blocking buffer (Beyotime, Shanghai) were used for endogenous peroxidase inactivation and blocking. Then, the lung sections were incubated with primary antibodies at 4 °C overnight and then incubated with specific secondary antibodies at 37 °C for 1 h. The DAB Horseradish Peroxidase Color Development Kit (Beyotime, Shanghai) was used for immunohistochemical staining. Olympus SpinSR10 was used for immunofluorescence. Antibodies used for staining were summarized in Table S1.

Flow cytometry

Cells were collected and filtered with a 70 μm filter. RBCs were lysed as described above. Then cells were washed with PBS (supplemented with 3% fetal bovine serum) and centrifuged at 1500 rpm for 10 min at 4 °C. Cells were resuspended and stained with appropriate antibody at 4 °C for 30 min. Antibodies used for flow cytometry were summarized in Table S1. Beckman Coulter CytoFlex was applied for flow cytometry.

Statistical analysis

All analyses were performed in SPSS 23.0 and GraphPad 9. Unless otherwise specified, data were presented as mean ± standard deviations (SD). Two-tailed Student’s unpaired t-test was used for comparisons between two groups. One-way ANOVA with a post hoc Tukey test was used for comparisons between three or more groups. A p-value < 0.05 was considered statistically significant.

Results

Upregulated PRC1 is associated with increased mortality in sepsis patients

We first used the GSE95233 dataset to identify potential biomarkers for sepsis prognosis. According to the survival status at day 28, the samples were then divided into three distinct groups: healthy controls, sepsis survivors, and sepsis non-survivors. Then the DEGs with a |logFC| > 0.5 and an adj. p < 0.05 in the Healthy vs. Death group and Survivor vs. Death group were obtained. There were 228 common DEGs between the two groups, which were used for further analysis. (Fig. 1.A). The MCODE method identified the top four clusters with the highest significance in 228 common DEGs (Fig. 1.B), demonstrated the primary clustering of DEGs into mitosis-related clusters (red), kinesins-related clusters (blue), PD-1 signaling-related clusters (green), and condensin I complex-related clusters (purple).The enrichment of the kinesin-related cluster was noted, and subsequently, the expression levels of two core genes, PRC1 and Rac GTPase Activating Protein 1 (RACGAP1), were selected and compared in the kinesin-related cluster across three groups: healthy controls, sepsis survivors, and sepsis non-survivors (Fig. 1.C). The results demonstrated that PRC1 and RACGAP1 exhibited the highest expression levels of in sepsis non-survivors.

Fig. 1.

Upregulated PRC1 in macrophages is associated with increased mortality in sepsis patients. (A) Venn diagram showing the number of DEGs in dataset GSE95233. The common DEGs in Healthy vs. Death group and Survivor vs. Death were selected for further analysis. (B) The MCODE method identified the top four significant clusters in common DEGs. (C) The expression value of PRC1 and RACGAP1 in healthy volunteers, sepsis survivors and sepsis death group in dataset GSE95233. (D) The expression value of PRC1 and RACGAP1 in healthy volunteers, sepsis survivors and sepsis death group in dataset GSE65682. (E) Survival curves of PRC1-high and PRC1-low group, and RACGAP1-high and RACGAP1-low group, in sepsis patients in dataset GSE65682. (F) Survival curves of sham group and CLP group. Each group initially contained 10 mice. (G) HE and immunohistochemical staining for PRC1 in lung in CLP-survival mice, and the injury score and percentage of PRC1 positive area. (H) Cell composition of 22 immune cell types identified via CIBERSORTx. Data were presented with mean and 95% CI. (I) The pearson correlation coefficient between PRC1 expression and cell composition of T cells CD8, T cells CD4 naïve, T cells gamma delta, Monocytes and Macrophages M0 in sepsis patients in dataset GSE65682. (J) Cell composition of Macrophages M0 in healthy volunteers, sepsis survivors and sepsis death group. (K, L) Survival curves of PRC1-high and PRC1-low group in macrophage enriched and decreased subgroup in sepsis patients in dataset GSE65682. Scale bar = 100 μm

Then the GSE65682 dataset was used to further validate the correlation between disparate PRC1 and RACGAP1 expression levels and survival status of sepsis patient. Dataset GSE65682 is part of a prospective cohort study and includes patient information from various diagnoses, including community-acquired pneumonia and hospital-acquired pneumonia [44]. In the GSE65682 dataset, PRC1 expression levels exhibited an increasing trend across the healthy controls, sepsis survivors, and non-survivors, while RACGAP1 expression levels did not differ significantly between survivors and non-survivors (Fig. 1.D). Subsequently, all sepsis samples were categorized into high- or low-expression groups based on PRC1 or RACGAP1 expression levels, and 28-day survival were compared between high- or low-expression groups (Fig. 1.E). The survival rate was found to be significantly lower in the group with high PRC1 expression compared to the group with low PRC1 expression (p = 0.0006). However, no significant difference in survival rates was observed between the groups with high and low RACGAP1 expression (p = 0.2182). These results suggest that PRC1 expression levels are associated with the prognosis of sepsis.

PRC1 is upregulated in macrophages during sepsis

We then examined the expression of PRC1 in lungs in mouse CLP-induced sepsis model. Half of the mice survived on the fifth day after CLP (Fig. 1.F). As one of the most prevalent and critical organs affected during sepsis, lung samples from CLP survivals or sham group were harvested at day 5 after CLP. Sepsis-induced lung injury was observed in CLP survivals, characterized by localized edema, hemorrhage, and inflammatory infiltration. The positive PRC1 staining detected by immunohistochemistry in the lungs was also increased in CLP survivals (Fig. 1.G). Interestingly, the PRC1 positive staining was predominantly distributed in proximity to the bronchi or within the alveolar spaces. In contrast, little positive staining was observed in normal alveolar structures, suggesting that sepsis did not increase PRC1 expression in alveolar epithelial cells. Consequently, we hypothesize that the upregulation of PRC1 expression in sepsis may occur in immune cells, such as macrophages and neutrophils.

Inflammatory cytokine storms caused by excessive activation of the immune system and immunosuppression resulting from immune disorder are important causes of sepsis-related pathological damage and death. To clarify which immune cells are associated with the upregulation of PRC1 during sepsis, we used the CIBERSORTx tool to perform immune cell infiltration analysis on samples from the GSE65682 dataset. The results showed that the abundance of CD8⁺ T cells, naïve CD4⁺ T cells, and monocytes was decreased in sepsis samples compared to healthy controls, while the abundance of γδ T cells and M0 macrophages was increased in sepsis samples compared to healthy controls (Fig. 1.H). We then analyzed the correlation between PRC1 expression levels and the abundance of CD8⁺ T cells, naïve CD4⁺ T cells, monocytes, γδ T cells, and M0 macrophages in sepsis patients from GSE65682 (Fig. 1.I). Only naïve CD4⁺ T cells and M0 macrophages were positively correlated with PRC1. Since naïve CD4⁺ T cells were less abundant in sepsis samples, we examined the correlation between M0 macrophages and sepsis severity (Fig. 1.J). Consistent with PRC1 expression, M0 macrophage abundance was significantly higher in non-survivors than in survivors and healthy controls. Additionally, we divided sepsis patients into enriched and decreased groups based on M0 macrophage abundance, and then compared survival outcomes between patients with high or low PRC1 expression within each group. In patients with enriched M0 macrophage abundance, high PRC1 expression remained associated with increased 28-day mortality (p = 0.0229) (Fig. 1.K). However, decreased M0 macrophage abundance eliminated the effect of PRC1 expression levels on outcomes (p = 0.0823) (Fig. 1.L). These results suggested that high PRC1 expression in sepsis patient samples is associated with macrophages.

PRC1 expression is increased in endotoxin-tolerant macrophages

Previous results indicated that upregulated PRC1 is associated with increased sepsis mortality, and that PRC1 in mainly expressed in macrophages. Macrophages are usually classified into three main polarization phenotypes: M0, M1 (classically activated), and M2 (alternatively activated). Hence, we used in vitro THP-1 and BMDMs model to explore the PRC1 expression levels among different macrophage phenotypes. As described previously, primary BMDMs were isolated from C57/BL6 mice and cultured with 10 ng/ml M-CSF for 7 days (Fig. 2.A), and THP-1 monocytes were firstly treated with PMA (100ng/ml) for 24 h to differentiate into M0 macrophages. Then BMDMs and THP-1 were induced into M1 macrophages with LPS (100 ng/ml) + IFN-γ (20 ng/ml) or into M2 macrophages with IL-4 (10 ng/ml) + IL-13 (10 ng/ml). iNOS and CD86 were used as markers for differentiated mouse M1 BMDMs; CD80 and CD86 were used as markers for the differentiated human M1 THP-1 cells. Arg-1 and CD206 were used as markers for the differentiated M2 BMDMs and THP-1 cells. M0, M1, and M2 BMDMs and THP-1 were stimulated with 100 ng/ml LPS for 6 h, then PRC1 expression levels were detected (Fig. 2.B-E). Surprisingly, the expression levels of PRC1 in naive M0 and M2 BMDMs were not significantly different, while the PRC1 expression levels in both types of BMDMs decreased after LPS exposure compared to the naive state. The basal PRC1 expression in M1 BMDMs was lower compared to M0 and M2 BMDMs; however, the decrease in PRC1 expression after LPS stimulation was not significant. Consistent with the results in BMDMs, PRC1 mRNA levels decreased in M0 and M2 THP-1 cells after LPS exposure, while PRC1 mRNA levels in M1 macrophages remained unchanged after stimulation. This indicates that PRC1 expression is not increased in classical M0, M1, or M2 macrophages, and short-term LPS exposure cannot upregulate PRC1 expression.

Fig. 2.

PRC1 is upregulated in endotoxin-tolerant macrophages. (A) Induction and identification of primary bone marrow-derived macrophages (BMDMs). n = 3. (B) Relative mRNA expression of iNOS, CD86, Arg1, CD206 in M1 and M2 BMDMs. n = 6. (C) Relative mRNA expression of PRC1 in BMDMs with or without LPS stimulation for 6 h. n = 6. (D) Relative mRNA expression of CD80, CD86, Arg1, CD206 in M1 and M2 THP-1. n = 6. (E) Relative mRNA expression of PRC1 in THP-1 with or without LPS stimulation for 6 h. n = 6. (F) Relative fluorescence intensity of FITC-conjugated LPS (100ng/ml) in empty medium (left, without THP-1) or normal medium (right, with naive THP-1). =18. (G) Schematic diagram of inducing endotoxin-tolerant macrophages, and residual LPS concentration measured by FITC fluorescence intensity (right). n = 18. Macrophages were restimulated with 100ng/ml LPS 12 h after the first LPS stimulation, then cultured for 36 h. (H, I) Relative mRNA expression of PRC1 in endotoxin-tolerant BMDMs and THP-1 with or without LPS stimulation. n = 4. (J) Relative protein expression of PRC1 in naive or endotoxin-tolerant THP-1, with or without LPS stimulation for 6 h. n = 3. (K) Relative mRNA expression of PRC1 in endotoxin-tolerant PBMs. n = 3

During sepsis, stimuli such as PAMPs may persist in the body and circulation, leading to prolonged exposure of macrophages to inflammatory stimuli, which may subsequently result in endotoxin tolerance and immunosuppression. Therefore, we examined the PRC1 expression in endotoxin-tolerant macrophages after prolonged exposure of LPS. Since LPS was rapidly cleared within 3 to 6 h after added into culture medium (Fig. 2.F), we added a second LPS challenge (100 ng/ml) at 12 h to simulate prolonged LPS exposure (Fig. 2.G). After 48 h, the remaining fluorescence intensity of FITC-LPS in the medium was less than 10% (Fig. 2.G). After replacing the medium with fresh medium, LPS was administered at different concentrations (100 ng/ml, 1 µg/ml, and 10 µg/ml). Total RNA was then extracted at time gradients (3 h, 6 h, 12 h, 24 h). The prolonged LPS exposure increased the expression level of PRC1 in endotoxin-tolerant macrophages derived from BMDM or THP-1 compared to naive or M1 macrophages (Fig. 2.H, I), and re-stimulation with LPS did not alter the expression level of PRC1. The PRC1 protein expression level was increased in endotoxin-tolerant THP-1 cells compared to naïve THP-1 (p < 0.05), with or without LPS rechallenge, which was consistent with mRNA level (Fig. 2.J). We subsequently examined the mRNA expression level of PRC1 in resting PBMs, classical M1 PBMs (stimulated with 10 ng/ml LPS, then extracted RNA after 24 h) and endotoxin-tolerant PBMs (stimulated with 10 ng/ml LPS twice, then extracted RNA after 48 h) (Fig. 2.K). The results indicated that PRC1 was upregulated in endotoxin-tolerant PBMs. Based on these findings, we concluded that PRC1 is highly expressed in endotoxin-tolerant macrophages, rather than in classical M1 or M2 macrophages.

PRC1 knockdown restores macrophage response to LPS stimulation

Previous results indicate that endotoxin-tolerant macrophages exhibit elevated levels of PRC1 expression. To confirm whether endotoxin-tolerant macrophages with high PRC1 expression maintain normal immune responses, we measured the mRNA levels of several main cytokines after stimulation LPS for 6 h (Fig. 3.A). Interestingly, naive THP-1 cells exhibited varying degrees of increased expression of TNF-α, IL-6, TGF-β, and IL-10 after LPS stimuli, particularly TNF-α, IL-6, and IL-10. In contrast, endotoxin-tolerant THP-1 cells maintained higher mRNA levels of cytokines (especially TNF-α, IL-6, and TGF-β), indicating that endotoxin-tolerant THP-1 cells still maintain sustained cytokine expression. After re-stimulated with LPS, tolerant THP-1 cells exhibited impaired inflammatory responses, with expression levels of TNF-α (p = 0.0001), IL-6 (p = 0.0004), and IL-10 (p = 0.024) significantly lower than naïve THP-1, while TGF-β (p < 0.0001) mRNA levels were higher. We then examined the expression levels of CD86 and CD206, two specific markers commonly used to distinguish macrophage polarization phenotypes (Fig. 3.B). Endotoxin-tolerant THP-1 cells exhibited significantly decreased CD86 expression and upregulated CD206 expression level compared to naïve THP-1, whether stimulated with LPS or not. These results indicate that endotoxin-tolerant THP-1 exhibits characteristics different from naive cells, including impaired pro-inflammatory cytokine TNF-α and IL-6 expression after LPS stimulation, increased anti-inflammatory cytokine TGF-β expression, and M2-like macrophages surface markers expression (decreased CD86 and increased CD206).

Fig. 3.

PRC1 knockdown restores inflammatory response in endotoxin-tolerant macrophages. (A, B) Relative mRNA expression of TNF-α, IL-6, TGF-β, IL-10, CD86 and CD206 in naive and endotoxin-tolerant THP-1, with or without LPS stimulation (100ng/mL) for 6 h. n = 6. (C, D) Relative mRNA and protein expression of PRC1 in THP-1 after transfected with siRNA-358, siRNA-631 and siRNA-1001. n = 3–6. (E) Relative mRNA expression of PRC1 in endotoxin-tolerant THP-1 after transfected with PRC1-siRNA. n = 6. (F, G) Relative mRNA expression of IL-1β, TNF-α, IL-6, TGF-β, IL-10, CD86 and CD206 in resting or LPS-stimulated endotoxin-tolerant THP-1, with or without PRC1 knockdown. n = 3. (H) Relative mRNA expression of HLA-DRA and CD74 in resting or LPS-rechallenged endotoxin-tolerant THP-1, with or without PRC1 knockdown. n = 3. (I) HLA-DR expression levels identified by flow cytometry in LPS-rechallenged endotoxin-tolerant THP-1, with or without PRC1 knockdown. n = 3. (J) Extracellular acidification rate in LPS-stimulated naïve or endotoxin-tolerant THP-1. n = 8. (K) Relative protein expression of HIF-1α in LPS-stimulated endotoxin-tolerant THP-1, with or without PRC1 knockdown. n = 6

To further explore the effect of PRC1 in macrophage immune responses, we used Lipofectamine 3000 reagent and PRC1-siRNA (Table S3) to establish PRC1-knockdown THP-1 cells. PRC1 mRNA levels were reduced by three pairs of PRC1-siRNA (Fig. 3.C), but only siRNA-631 and siRNA-1001 successfully reduced PRC1 protein expression (Fig. 3.D). Hence, a pool of siRNA-631 and siRNA-1001 were used for PRC1 knockdown in THP-1 cells (Fig. 3.E). Subsequently, PRC1 knockdown THP-1 was induced to endotoxin-tolerant status as previously described. We examined whether PRC1 knockdown restored the inflammatory response of endotoxin-tolerant THP-1 cells to LPS re-stimulation. PRC1 knockdown restored the expression of TNF-α (p = 0.0229) and IL-6 (p = 0.0296) in LPS-stimulated tolerant THP-1 cells, and inhibited IL-10 (p = 0.0082) expression (Fig. 3.F). The TGF-β (p = 0.0461) expression level was also decreased by PRC1 knockdown. Additionally, PRC1 knockdown increased CD86 expression and downregulated CD206 expression in tolerant THP-1 cells, regardless of LPS stimulation (Fig. 3.G).

Septic immunosuppression is also characterized with impaired glycolysis and downregulation of MHC II molecule expression. Therefore, we examined the ECAR levels and expression levels of MHC II subunits in tolerant THP-1 cells. PRC1 knockdown upregulated the expression of HLA-DR in tolerant THP-1 cells after LPS stimulation (Fig. 3.H-I, Figure S1). However, PRC1 knockdown did not affect glycolysis and HIF-1α expression level in endotoxin-tolerant THP-1 cells (Fig. 3.J-K). These results indicated that high PRC1 expression is associated with impaired immune responses in macrophages during immunosuppression, while PRC1 knockdown can partially restore the inflammatory responses of tolerant THP-1 cells.

PRC1 knockdown inhibits STAT3 phosphorylation in endotoxin-tolerant macrophages

Next, we investigated the potential mechanism that PRC1 expression affect the inflammatory responses in endotoxin-tolerant macrophages. We first compared the transcriptomic sequencing data of basal blood monocytes from 6 healthy volunteers and 8 sepsis patients in database GSE46955 (Fig. 4.A). We selected DEGs with |logFC| >0.5 and adj.p < 0.05, followed by enrichment analysis with Metascape. Most of the differentially expressed genes were associated with immune activation and regulated by transcription factors such as NF-κB, STAT3, STAT1, and HIF-1α. Therefore, we examined the activation of NF-κB, STAT1, and STAT3 in endotoxin-tolerant macrophages. NF-κB, STAT1, and STAT3 are key signaling pathways activated by PAMPs that play crucial roles in macrophage activation and coordinate downstream immune responses during inflammation, and enriched in blood monocytes from sepsis patients (Fig. 4.A). Interestingly, tolerant THP-1 cells exhibited sustained STAT3 activation, accompanied by a slight increase in total STAT3 expression (Fig. 4.B, C). Total STAT1 expression levels were up-regulated while phosphorylated STAT1 levels were reduced (Fig. 4.B, Fig. 4.C). No significant differences were observed in total and phosphorylated NF-κB levels between naïve and tolerant THP-1 (Fig. 4.B, Fig. 4.C). These results suggested that sustained activation of the STAT3 signaling pathway may contribute to the impaired inflammatory response in tolerant THP-1 cells.

Fig. 4.

PRC1 knockdown inhibits STAT3 phosphorylation in endotoxin-tolerant macrophages. (A) Volcano plot of DEGs between healthy volunteers and sepsis patients in database GSE46955 (left), GO biological process (middle), and TRRUST enrichment (left) of DEGs. (B, C) Relative protein expression of p-NF-κB (S536), NF-κB, p-STAT1 (Y701), STAT1, p-STAT3 (Y705) and STAT3 in naïve and endotoxin-tolerant THP-1. n = 3. (D, E) Relative protein expression of p-NF-κB (S536), NF-κB, p-STAT1 (Y701), STAT1, p-STAT3 (Y705) and STAT3 in endotoxin-tolerant THP-1, with or without PRC1 knockdown. n = 3

We then investigated whether PRC1 knockdown affect NF-κB, STAT1, and STAT3 activation in tolerant THP-1 cells. PRC1 knockdown did not affect the expression levels and phosphorylation levels of NF-κB and STAT1 in tolerant THP-1 cells (Fig. 4.D, E). PRC1 knockdown led to a significant decrease in phosphorylated STAT3 levels in tolerant THP-1 cells (Fig. 4.D, E), indicating that PRC1 knockdown reversed the sustained activation of STAT3 in tolerant THP-1 cells. Next, we verified whether restoring PRC1 expression could reverse STAT3 activation and the inflammatory response in endotoxin-tolerant THP-1. pCMV-PRC1 plasmid was transfected into PRC-1 knockdown THP-1 to restore PRC1 expression. Thereafter, THP-1 cells with pCMV-PRC1 (PRC1-r THP-1) or blank vector (PRC1-kd THP-1) were induced to an endotoxin-tolerant status as previously described (Figure S1A, B). The phosphorylated STAT3 level increased in endotoxin-tolerant PRC1-r THP-1 (Figure S2C). In addition, PRC1-r THP-1 and PRC1-kd THP-1 were rechallenged with 100ng/ml LPS for 6 h, and total RNA was subsequently extracted. The restoration of PRC1 expression resulted in a decrease in the mRNA levels of TNF-α, IL-6 and CD86, while concurrently increasing the mRNA levels of IL-10 and CD206 in tolerant PRC1-r THP-1 (Figure S2D). Hence, our results suggested that increased PRC1 expression is associated with sustained STAT3 phosphorylation, thus impairs the inflammatory response in tolerant THP-1 cells.

PRC1 maintains STAT3 activation in a β-catenin-dependent manner

Previous results suggested that PRC1 maintains STAT3 phosphorylation in endotoxin-tolerant THP-1. STAT3 can be regulated by Wnt/β-catenin pathway [45], and induces immunosuppressive tumor microenvironment [46]. PRC1 promotes β-catenin nuclear translocation and activates Wnt/β-catenin pathway in hepatocellular carcinoma [47]. Therefore, we hypothesized that PRC1 regulates STAT3 phosphorylation by maintaining the activity of the Wnt/β-catenin pathway.

First, we examined the activation of the Wnt/β-catenin pathway in endotoxin-tolerant THP-1. Upregulated active β-catenin was detected in the nuclear protein extraction of endotoxin-tolerant THP-1 cells compared to naïve THP-1 (p = 0.0144) (Fig. 5.A). SKL2001 is a novel Wnt/β-catenin agonist that exerts its function by disrupting the Axin/β-catenin interaction [48]. Knockdown of PRC1 reduced active β-catenin levels in endotoxin-tolerant THP-1 cells, and the administration of SKL2001 increased the active β-catenin level in PRC1-knockdown endotoxin-tolerant THP-1 compared to PRC1 knockdown endotoxin-tolerant THP-1 (p = 0.0018) (Fig. 5.B, C). Immunofluorescence confirmed that SKL2001 restored the reduced nuclear translocation of active β-catenin caused by PRC1 knockdown (Fig. 5. D). Previous studies have demonstrated that PRC1 disrupts the phosphorylation of β-catenin by APC, thereby maintaining β-catenin activity [47]. Hence, we examined the levels of phosphorylated β-catenin that bind to APC in endotoxin-tolerant THP-1 cells with or without PRC1 knockdown (Fig. 5. E). As expected, the knockdown of PRC1 resulted in an augmentation of the phosphorylation of β-catenin by APC (p < 0.0001), hereby suggesting that PRC1 attenuates the phosphorylation of β-catenin by APC and maintains β-catenin activity in endotoxin-tolerant THP-1. We then investigated the phosphorylated STAT3 level in PRC1-knockdown endotoxin-tolerant THP-1 with or without SKL2001. SKL2001 increased the phosphorylated STAT3 level in PRC1-knockdown endotoxin-tolerant THP-1 (p = 0.0002) (Fig. 5.F), while the same dose of SKL2001 did not induce STAT3 activation in naive THP-1 cells (Figure S3). Additionally, SKL2001 reduced the expression levels of inflammatory cytokines IL-1β (p < 0.0001), TNF-α (p < 0.0001), and M1 marker CD86 (p = 0.0003) (Fig. 5.G) in PRC1-knockdown endotoxin-tolerant THP-1 after LPS rechallenge. Interestingly, SKL2001 also inhibited the expression of TGF-β (p = 0.0044) and IL-10 (p = 0.0005) but not affected the expression levels of IL-6 (p = 0.2601) and CD206 (p = 0.6901) in PRC1-knockdown endotoxin-tolerant THP-1 after LPS rechallenge. In summary, these results indicate that PRC1 knockdown reduces STAT3 phosphorylation in tolerant THP-1 cells, while the β-catenin agonist SKL2001 restores STAT3 phosphorylation in tolerant THP-1 cells. These results suggest that PRC1 maintains sustained STAT3 activation via a β-catenin-dependent mechanism and participates in suppressing inflammatory responses in tolerant macrophages upon re-exposure to LPS.

Fig. 5.

PRC1 maintains STAT3 activation in a β-catenin-dependent manner. (A) Relative protein expression of active β-Catenin in cytoplasmic and nuclear extraction from naïve and endotoxin-tolerant THP-1. n = 3. (B) Cell viability evaluated by CCK-8 in endotoxin-tolerant THP-1. n = 8. (C) Relative protein expression of active β-Catenin in whole cell lysate from endotoxin-tolerant THP-1. PRC1 knockdown endotoxin-tolerant THP-1 was treated with DMSO or SKL2001 (10µM) for 24 h. n = 3. (D) Typical image of immunofluorescence of PRC1 (Red) in endotoxin-tolerant THP-1. n = 3. (E) Immunoprecipitation assays of p-β-catenin that binding to APC in endotoxin-tolerant THP-1, with or without PRC1 knockdown. Anti-APC antibody was used as bait. n = 4. (F) Relative protein expression of p-STAT3 (Y705) and STAT3 in endotoxin-tolerant THP-1. n = 3. (G) Relative mRNA expression of IL-1β, TNF-α, IL-6, TGF-β, IL-10, CD86 and CD206 in resting or LPS-stimulated (pretreated with DMSO or 10µM SKL2001) PRC1 knockdown endotoxin-tolerant THP-1. n = 3. Scale bar = 20 μm

Pharmacological ablation of β-catenin restores the inflammatory response of endotoxin-tolerant macrophages

Since the high expression of PRC1 suppresses the inflammatory response in tolerant macrophages through a β-catenin-dependent mechanism, we investigated whether pharmacological ablation of β-catenin could restore the normal immune response of tolerant macrophages. MSAB is a selective inhibitor of the Wnt/β-catenin pathway that promotes β-catenin degradation via a proteasome-dependent mechanism. Pretreatment with 5µM MSAB for 16 h successfully reduced the active β-catenin level in tolerant THP-1 (p = 0.0109) (Fig. 6. A, B; Figure S4) and also decreased the phosphorylated STAT3 levels in tolerant THP-1 (p = 0.0266) (Fig. 6. C). We then examined whether MSAB pretreatment could restore the expression of inflammatory factors and surface markers in tolerant THP-1 (Fig. 6.D-F). MSAB increased the expression of TNF-α (p = 0.0009), IL-6 (p = 0.0303), CD86 (p = 0.0218), IL-1β (p = 0.0424), and HLA-DRA (p = 0.0301) in tolerant THP-1 cells upon rechallenge with LPS and downregulated the expression of CD206 (p = 0.0005). Flow cytometry confirmed that MSAB increased HLA-DR expression in tolerant THP-1 (p < 0.0001) (Fig. 6.G). These results suggest that pharmacological ablation of β-catenin by MSAB can restore the CD86 and HLA-DR expression and increase the expression of pro-inflammatory cytokines in endotoxin-tolerant THP-1.

Fig. 6.

Pharmacological ablation of β-catenin restores the inflammatory response of endotoxin-tolerant macrophages. (A) Relative protein expression of active β-Catenin in whole cell lysate from endotoxin-tolerant THP-1, with or without MSAB pretreatment. n = 3. (B) Typical image of immunofluorescence of PRC1 (Red) in endotoxin-tolerant THP-1, with or without MSAB pretreatment. (C) Relative protein expression of p-STAT3 (Y705) and STAT3 in endotoxin-tolerant THP-1, with or without MSAB pretreatment. n = 3. (D-F) Relative mRNA expression of TNF-α, IL-6, TGF-β, IL-10, CD86, CD206, HLA-DRA and CD74 in resting or LPS-stimulated (pretreated with DMSO or 5µM MSAB). n = 3. (G) HLA-DR expression levels identified by flow cytometry in LPS-stimulated endotoxin-tolerant THP-1, with or without 5µM MSAB pretreatment. n = 3. (H) Relative protein expression of p-STAT3 (Y705) and STAT3 in endotoxin-tolerant BMDMs with or without MSAB pretreatment. n = 4. (I, J) Relative mRNA expression of TNF-α, IL-6, TGF-β, IL-10, CD86 and CD206 in resting or LPS-stimulated (pretreated with DMSO or 10µM MSAB) BMDMs. n = 3. Scale bar = 20 μm

We then confirmed whether pharmacological ablation of β-catenin via MSAB could restore the inflammatory response of endotoxin-tolerant BMDMs. Similar to THP-1, endotoxin-tolerant BMDMs exhibited upregulation of active β-catenin and phosphorylated STAT3 levels (Figure S5, Figure S6), while 10µM SAB treatment reduced STAT3 phosphorylation levels in endotoxin-tolerant BMDMs (p = 0.0452) (Fig. 6.H), indicating that sustained activation of STAT3 in endotoxin-tolerant BMDMs is also associated with β-catenin. Furthermore, pretreatment of tolerant BMDM with 10µM MSAB increased TNF-α (p = 0.0128) and IL-6 (p < 0.0001) expression and reduced IL-10 (p < 0.0064) expression when rechallenged with LPS (Fig. 6.I). The CD86 expression was also increased by MSAB (p < 0.0001) (Fig. 6.J). Subsequently, we examined whether MSAB administration restored the inflammatory response of endotoxin-tolerant PBMs. The upregulation of active β-catenin and STAT3 phosphorylation was detected in endotoxin-tolerant PBMs, and the expression of APC decreased in endotoxin-tolerant PBMs (Figure S7.A). In addition, 5µM MSAB effectively reduced the levels of active β-catenin and p-STAT3 (Figure S7.B). MSAB administration also upregulated the expression of TNF-α (p = 0.0158), IL-6 (p = 0.0006) and CD86 (p = 0.0056), while decreasing CD206 expression (p = 0.0127) in endotoxin-tolerant PBMs exposed to LPS (Figure S7.C). These results suggest that pharmacological ablation of β-catenin can restore the expression of inflammatory cytokines and M1-phenotype surface markers in tolerant macrophages.

Lactic acid accumulation promotes the expression of PRC1 in endotoxin-tolerant macrophages

Since PRC1 upregulation did not occur in the early stages of LPS exposure or in classical M1 macrophages, we speculated that PRC1 upregulation might be a result of microenvironmental changes. Therefore, we added the medium supernatant from endotoxin-tolerant THP-1 to naive THP-1 and examined the PRC1 expression levels. After cultured with medium supernatant from tolerant THP-1 for 24 h, the expression levels of PRC1 were upregulated in naive THP-1 (Fig. 7. A, B). The medium from tolerant THP-1 also decreased the expression of CD86 and increased the expression of CD206 in naive THP-1 cells (Fig. 7.C). Then we stimulated the naïve THP-1 cells cultured with or without medium from tolerant THP-1. Medium from tolerant THP-1 decreased the expression of IL-1β, TNF-α, and IL-6 while upregulating TGF-β and IL-10 expression levels in naïve THP-1 (Fig. 7.D). Subsequently, the medium of the tolerant THP-1 cells was replaced with fresh medium and cultured for 24 h. Then tolerant THP-1 with original or fresh medium was stimulated with LPS again. Replacing the culture medium did not restore the expression of CD86, CD206, and PRC1 in the tolerant THP-1 cells (Fig. 7. E). These results suggested that the upregulation of PRC1 is attributed to microenvironmental changes following inflammatory stimulation, and this endotoxin-tolerant state in macrophages can be maintained after environment change.

Fig. 7.

Lactic acid accumulation promotes the expression of PRC1 in endotoxin-tolerant macrophages. (A) Schematic diagram. Naive THP-1 was cultured with medium from endotoxin-tolerant THP-1, then total RNA was extracted at different time point. (B) Relative mRNA expression of PRC1 in naive THP-1 cultured with medium from endotoxin-tolerant THP-1. n = 3. (C) Relative mRNA expression of CD86 and CD206 in naive THP-1 cultured with or without medium from endotoxin-tolerant THP-1. n = 3. (D) Naive THP-1 cultured with or without medium from endotoxin-tolerant THP-1 for 24 h, then stimulated with LPS (100ng/ml) for 6 h. Then examined the relative mRNA expression of IL-1β, TNF-α, IL-6, TGF-β and IL-10. n = 3. (E) Endotoxin-tolerant THP-1 was washed with PBS and then cultured with fresh medium or endotoxin-tolerant medium for 24 h, then examined the relative mRNA expression of CD86, CD206 and PRC1. n = 3. (F) Lactic acid concentration in medium from naive or endotoxin-tolerant THP-1. n = 8. (G) Naive THP-1 was cultured in different lactic acid concentration for 24–48 h, then examined the relative mRNA expression of PRC1. n = 3. (H) Naive THP-1 was cultured in different lactate sodium or hydrochloric acid concentration for 24 h, then examined the relative mRNA expression of PRC1. n = 3. (I, J) Relative mRNA and protein expression of KLF4 in THP-1 after transfected with siRNA-878, siRNA-1335 and siRNA-1989. n = 3. (K) Relative mRNA expression of PRC1 in THP-1 after cultured with or without additional lactic acid (10mM). n = 3. (L) Naive BMDM was cultured in different lactic acid concentration for 24 h, then examined the relative mRNA expression of PRC1. n = 3. (M) Relative mRNA expression of PRC1 in BMDM after cultured with or without additional lactic acid (20mM). n = 3. (N) The motif of STAT3 (MA0144.3), and top 3 predicted binding sequence in PRC1 promoter domain

We then investigated whether lactate accumulation in the environment contributes to PRC1 upregulation in tolerant macrophages, because macrophages maintain highly active glycolytic levels to coordinate inflammation responses during sepsis. First, lactate levels were measured in the culture medium of tolerant and naive THP-1 cells. Lactate concentration was significantly elevated in the medium of tolerant THP-1 with a mean concentration of 13mM, while the mean lactate concentration in naïve THP-1 was 3.3mM (Fig. 7.F). We further stimulated resting THP-1 cells with different concentrations of lactic acid for 24–48 h and then measured PRC1 expression levels (Fig. 7.G). At lower lactate concentrations (below 5 mM), THP-1 PRC1 expression levels slightly decreased, while exposure to 10 mM and higher lactate concentrations significantly upregulated PRC1 expression levels in naive THP-1 cells. To investigate whether the upregulation of PRC1 is associated with an acidic environment, we treated THP-1 cells with equal concentrations of hydrochloric acid or sodium lactate (10mM and 20mM) for 24 h, then examined the PRC1 expression. Treatment with 10mM and 20mM sodium lactate upregulated PRC1 expression in naïve THP-1, similar to the results of lactic acid treatment at the same concentrations (Fig. 7.H). However, hydrochloric acid slightly reduced the expression levels of PRC1 (Fig. 7.H). The expression of PRC1 can be upregulated by multiple transcription factors, including KLF4, STAT3 and JUN (Figure S8), all of which were also activated in tolerant macrophages (Fig. 4.A). We then explored the potential transcription factors of PRC1 in tolerant macrophages. Since no inhibitors of KLF4 were available, si-KLF4-1989 was used for KLF4 knockdown (Fig. 7.I, J; Table S3). 5µM Stattic was used for STAT3 inhibition, and 10µM SP600125 was used for JUN inhibition. Interestingly, only Stattic decreased the PRC1 expression in both naive and 10mM lactic acid-treated THP-1 (Fig. 7.K). In addition, Stattic could also suppress PRC1 expression in both naive or 20mM lactic acid-treated BMDM (Fig. 7.L, M). The JASPAR database predicted multiple potential STAT3 binding sites within the PRC1 promoter region (NC_000015.10:c 90994436–90996536) (Fig. 7.N). These results suggest that STAT3 is involved in the upregulation of PRC1 expression in tolerant macrophages. In summary, the accumulation of lactate in the environment promotes PRC1 expression, rather than hydrogen ions.

MSAB enhances the survival rate of CLP-survival mice during subsequent infection

Finally, we investigated the potential clinical value of MSAB in a mouse model of sepsis. A dose of 10 mg/kg MSAB did not result in impaired liver or kidney function or pathological changes (Fig. 8.A, B). To validate the impact of MSAB on the prognosis of sepsis, 10 mg/kg of MSAB or an equivalent volume of vehicle was injected intraperitoneally into healthy or CLP-survival mice 24 h in advance (Fig. 8.C). The CLP procedure was performed on healthy mice. Given that commercially available purified LPS is devoid of live pathogens, an intraperitoneal injection of 200 µl of fecal suspension was used to induce a subsequent infection model in CLP-survival mice [41]. MSAB did not affect the baseline levels of TNF-α, IL-6 or IL-10 in the serum of healthy mice (Fig. 8.D). Additionally, MSAB administration did not increase the mortality rate during the initial CLP-induced sepsis (Fig. 8.E). In CLP-survival mice, MSAB increased both the concentrations of TNF-α and IL-6 in the serum 12 h after the injection of fecal suspension (Fig. 8.F), and improved the 5-day survival rate (Fig. 8.G).

Fig. 8.

MSAB enhances the survival rate of CLP-survival mice during subsequent infection. (A) Healthy C57BL/6 male mice were administered MSAB or vehicle via intraperitoneal injection, then sacrificed after 24 h. The concentrations of ALT, AST, creatinine and BUN in murine serum were subsequently measured. n = 4. (B) Typical images of HE and staining of murine liver and kidney. (C) Healthy or CLP-survival mice were administered MSAB (10 mg/kg) or vehicle. Then CLP-induced sepsis model was established in healthy mice, and fecal suspension intraperitoneal injection-induced subsequent infection model was established in CLP-survival mice. (D) The concentrations of TNF-α, IL-6 and IL-10 in serum in healthy mice before CLP, treated with or without MSAB (10 mg/kg). n = 4. (E) Survival curves of CLP group with or without MSAB. Each group initially contained 10 mice. (F) The concentrations of TNF-α, IL-6 and IL-10 in serum in CLP-survival mice after fecal suspension intraperitoneal injection-induced subsequent infection. Serum samples were collected 12 h after injection. n = 6. (G) Survival curves of CLP-survival mice with or without MSAB after fecal suspension intraperitoneal injection. Each group initially contained 16 mice. (H) The graphical abstract of this study. Scale bar = 100 μm

In summary, our results indicate that macrophages upregulate PRC1 following prolonged exposure to a septic environment, thereby stabilizing β-catenin and maintaining sustained activation of STAT3, which exacerbates immunosuppression in macrophage. Pharmacological ablation of β-catenin restores macrophage immune responses. Additionally, we confirmed that the upregulation of PRC1 is associated with lactate accumulation in the environment (Fig. 8.H).

Discussion

In this study, we aimed to identify the potential predictors for sepsis prognosis and explore the potential mechanisms. Our main finding was that increased PRC1 expression in endotoxin tolerant macrophages is associated with septic immunosuppression. We reported that increased PRC1 promotes active β-catenin stabilization and nuclear transposition, thus maintaining STAT3 activation. The knockdown of PRC1 or pharmacological ablation of β-catenin can reduce the phosphorylation level of STAT3 and restore the inflammatory response in endotoxin-tolerant macrophages, increasing the expression of pro-inflammatory cytokines TNF-α, IL-6, IL-1β and the expression of M1 markers CD86 and HLA-DR.

We discovered that PRC1, a microtubule-associated protein, is highly expressed in macrophages with endotoxin tolerance. PRC1 is primarily involved in mitosis. Previous studies have shown that PRC1 is a promising therapeutic target for tumors, particularly when combined with other anti-tumor therapies [49–51]. However, few reports have addressed the role of PRC1 in the immune system. Our results suggest that increased PRC1 expression is elevated in macrophages with decreased immune responses, reduced expression of CD86 and HLA-DR, as well as increased expression of CD206. Similar results have been observed in some solid tumors, that the upregulation of PRC1 predicts the formation of an immunosuppressive microenvironment, as well as an increased risk of tumor metastasis and invasion [22, 52, 53]. It is noteworthy that PRC1 is a mitosis-related protein. Therefore, the upregulation of PRC1 found in tumor tissues may be primarily attributed to the active division of tumor cells rather than increased infiltration of tumor-associated immune cells.

Our results linked the PRC1 and β-catenin in endotoxin-tolerant macrophages. We reported that the nuclear accumulation of active β-catenin was increased in endotoxin-tolerant macrophages. The treatment of β-catenin inhibitor MSAB can both restore the expression of CD86 and HLA-DR in endotoxin-tolerant macrophages, and also increase the pro-inflammatory cytokines expression. Previous studies have reported that the genetic or pharmacological ablation of β-catenin enhance the production of inflammatory cytokines [54, 55], which is similar to our results. Wnt/β-catenin signaling is activated in M2-polarized macrophages, leading to immunosuppressive tumor microenvironment [56]. We find that knockdown of PRC1 reduces the active β-catenin level in endotoxin-tolerant macrophages. The PRC1 can stabilize β-catenin by suppressing the function of β-catenin destruction complex [47]. The β-catenin agonist SKL2001 can disrupting the interaction between β-catenin and β-catenin destruction complex, thereby stabilizing β-catenin in PRC1 knockdown endotoxin-tolerant macrophages and inhibit the inflammatory responses. These results indicate that the pharmacological ablation of β-catenin provides a potential target for restoring macrophage immune function, given that there are no available targeted agents for PRC1.

Furthermore, we reported the sustained activation of STAT3 in endotoxin-tolerant macrophages. Our results indicate that the level of p-STAT3 in endotoxin-tolerant macrophages can be reduced by β-catenin ablation. Previous studies reported that β-catenin can increase STAT3 activation [57]. The ablation of β-catenin in macrophages can reduce M2 polarization and increase the expression of M1 macrophage markers [58], which is consistent with our results. STAT3, a pivotal signaling molecule that orchestrates the immune response of macrophages, is activated in the early stages following LPS intervention and regulates inflammatory responses during sepsis [59]. Previous studies have reported that the sustained activation of STAT3 signaling can lead to immune cell dysregulation in sepsis, and the inhibition of STAT3 phosphorylation can reduce the mortality in late-stage sepsis mouse models [60, 61]. The deficiency of STAT3 can inhibit the M2 polarization of macrophages and increase the pro-inflammatory mediator expression [46]. Given that sepsis is a prolonged and complex pathological state with intricate mechanisms, the role of STAT3 in sepsis is multifaceted, precisely regulating cellular processes such as cell survival, immune function, and autophagy at different stages of sepsis and in different cell types. Our findings support that targeting PRC1/β-catenin can inhibit the activation of STAT3 endotoxin-tolerant macrophages, and restore the expression of M1 macrophage markers and pro-inflammatory cytokines.

It is noteworthy that PRC1 expression in macrophages declines during the initial stages of LPS exposure; however, the PRC1 expression increases following prolonged exposure to a septic environment. It has been demonstrated that PRC1 plays a critical role in the process of cellular mitosis, thereby indicating a plausible mechanism for monocyte regeneration. Furthermore, emergency granulopoiesis during sepsis produces newly generated immature neutrophils that exhibit the similar immunosuppressive phenotype and STAT3 activation [62]. The emergency myelopoiesis that occurs during the acute infection is an adaptive change to replenish rapidly depleted innate immune cells, such as monocytes or granulocytes [63]. The accumulation of active β-catenin and the activation of STAT3/5 signaling contribute to the generation of ectopic myeloid cells [64, 65]. This suggests that PRC1 upregulation may indicate potential emergency myelopoiesis during acute infection, while impairing the proinflammatory response of macrophages via β-catenin and STAT3. However, caution is warranted in interpreting whether the potential mechanism by which PRC1 drives tumor-associated immunosuppressive microenvironments similarly relies on the β-catenin/STAT3 pathway. On the one hand, increased PRC1 expression in solid tumors may not be entirely attributable to the infiltration of immunosuppressive cells. Furthermore, despite STAT3 activation promoting the development of myeloid-derived suppressor cells (MDSC) [63, 65], the function of β-catenin in the development of the tumor immunosuppressive microenvironment is multifaceted. Overactivation of β-catenin in tumor cells promotes the infiltration of MDSC via CXCR3 and DKK1 [37, 66], while the downregulation of β-catenin in MDSC has been shown to further enhance the immunosuppressive effects [67]. For tumor-associated macrophages, β-catenin signaling is imperative for sustaining their immunosuppressive characteristics and M2 phenotype [68, 69]. There may be partially analogous mechanisms between immune suppression in macrophages during sepsis and immune suppression in the tumor microenvironment; however, further validation is necessary to confirm this hypothesis.

Our study suggests that the alterations in the microenvironment leads to the endotoxin-tolerant state of macrophages during immunosuppression. This increases the complexity of septic immunosuppression. Interestingly, the culture medium from endotoxin-tolerant macrophages also altered the immune response of naive macrophages, leading to a phenotype similar to tolerogenic cells, with increased CD206 expression and decreased CD86 expression. This finding is similar to how the tumor microenvironment shapes macrophage behavior. As levels of residual LPS are very low in tolerant macrophages, the transition of naive macrophages to a tolerant state is likely to be the result of multiple microenvironmental changes, such as high concentrations of lactate, accumulated cytokines, and other metabolites [70, 71]. Our results indicate that lactate accumulation drives PRC1 upregulation in tolerant macrophages. As an important metabolite in innate and adaptive immunity, lactate plays an important role in regulating macrophage functions [72]. Prolonged exposure to lactic acid in the immunosuppressive tumor microenvironment leads to a shift towards M2 polarization in macrophages [73]. Elevated lactate concentrations in the environment reduce the expression of the rate-limiting enzyme in glycolysis and delay the activation of downstream signal transduction by LPS stimuli [70]. As the renewal of the culture medium failed to restore expression of CD86 and CD206 in tolerant macrophages, it is likely that the influence of the microenvironment on macrophage function is persistent. The accumulation of lactate leads to the acetylation of histone H3K27, which alters cellular transcriptional activity and promotes the transcription suppression associated with inflammatory responses [74].

In summary, our results suggest that sepsis leads to lactate accumulation in the microenvironment of macrophages, thereby promoting the upregulation of PRC1 and activating β-catenin/STAT3 pathway. This leads to immunosuppression of macrophages and impaired inflammatory responses. PRC1 knockdown and the pharmacological ablation of β-catenin can restore impaired immune function and the production of inflammatory cytokines in macrophages.

There are some limitations in this study that require further investigation. While our preliminary results suggest that pharmacological ablation of β-catenin by MSAB can restore inflammatory response and improve survival rates of CLP-survival mice during secondary infection, systemic suppression of β-catenin may result in other potential impacts. Further research is necessary to confirm the safety and efficacy of β-catenin suppression during sepsis. Another limitation is that macrophage-specific PRC1-deficient mice were not applied in this study, which requires further exploration.

Supplementary Information

Below is the link to the electronic supplementary material.

Authors’ contributions

Yifan Zuo: Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft; Shishi Zou: Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft; Zhiwei Wang: Conceptualization, Supervision, Funding acquisition, Project administration, Writing – review & editing; Yi Liu: Formal analysis, Investigation, Validation; Xiaoping Xie: Formal analysis, Investigation, Validation; Bolai Shen: Formal analysis, Investigation, Validation; Guoqing Luo: Software, Formal analysis, Validation; Xiao Lu: Formal analysis, Investigation, Validation; Ning Li: Conceptualization, Methodology, Writing – review & editing; Wanli Jiang: Conceptualization, Project administration, Writing – review & editing.

Funding

This research project was supported by the natural science foundation of China (grant No.82070481) and the department of science and technology of Hubei province (grant No.2022BCA035).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics Approval

All animals were maintained according to the guidelines outlined in the Guide for the Care and Use of Laboratory Animals. All experiments involving animals were approved by the Animal Ethics Committee of Renmin Hospital, Wuhan University (WDRM202200167).

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.