GUP (WH-0960) was gained from Shanghai Ronghe Medical Technology Development Co., LTD. Its purity was ≥95.0% (purity values are presented by the company) and the original materials of GUP (G. uralensis-20210,824-1) was gained from Shanghai Wanshicheng Sinopod Products Co., LTD. Anti-α-tubulin (66031-1-Ig, PTG), anti-Lamin B1 (66095-1-Ig, PTG), anti-IRF3 (11312-1-AP, PTG), anti-HSP90 (13171-1-AP, PTG), anti-hp-IRF3 (ab76493, abcam), anti-STING (19851-1-AP, PTG), anti-mp-IRF3 (GTX86691, GeneTex), anti-HA Tag (66006-2-Ig, PTG), anti-DYKDDDDK tag Polyclonal (20543-1-AP, PTG), HT-DNA (D6898, Sigma), 2′3′-cGAMP (HY-100564A, MCE), diABZI (HY112921B, MCE), DMXAA (HY-10964, MCE), CCK-8 (KTA1020-1000T), Hoechst (Boston, United States), opti-MEM (2427634, Gibco), Protease inhibitors Cocktail (C0001, TargetMol).

The process of GUP extraction: The G. uralensis would be washed, dried, sliced, and then G. uralensis with distilled water was put in a vessel, heated, and kept 80°C for 2 h and stirred regularly. The filtrates were combined and concentrated, extracted three times with ethyl acetate, recovered the aqueous layer and discarded the ethyl acetate. Subsequently, the water layer was placed in AB-8 microporous adsorbent resin, washed with distilled water and recycled water parts. Finally, the recovered water part was dried in a freeze dryer for 48 h to access purified GUP (company supplied).

C57BL/6 mice (female, 8 weeks old) were purchased from SPF Biotechnology Co., Ltd. (Beijing, China). Mouse cages were kept pathogen-free (SPF) and provided unrestricted access to food and water. The ambient humidity is 50% ± 5% and the temperature is 20°C–22°C. We try our best to minimize the number of experimental animals and the least possible suffering of mice during the experiment. The Fifth Medical Center of the Chinese People’s Liberation Army General Hospital approved all animal experimental procedures used in this study. These procedures were all carried out in accordance with guidelines for the care and use of laboratory animals (IACUC-2021-0008).

We employed Thermo Fisher ICS5000 ion chromatography to analyze the monosaccharide content of GUP. The chromatographic column used is DionexCarbopac^TMPA20 (3*150). The chromatographic mobile phase: solvent A, H₂O and solvent B, 15 mM NaOH and solvent C, 15 mM NaOH and 100 mM NaOAC. The flow rate was 0.3 mL/min, the injection volume was 5 μL, and the column temperature was set at 30°C. 16 monosaccharide standards including glucose, galactose and glucuronide were dissolved as controls. In the ampoule, the 10 mg sample was precisely weighed before 2 mL of 3M TFA was inserted and the sample was hydrolyzed at 120°C for 3 h. The acid hydrolysate was sucked out and blown dry with nitrogen. Add 5 mL of water vortexed to mix and draw 50 µL from it, add 950 µL of deionized water and centrifuge. Then, it is tested on an ion chromatograph.

Mouse bone marrow-derived macrophages (BMDMs) was isolated from bone marrow of 10-week-old female mice and incubated with DMEM medium containing murine macrophage colony-stimulating factor (MCE, Mon-mouth, NJ) with the addition of 1% penicillin/streptomycin (Macgene) and 10% fetal bovine serum (Gibco, Rockford, IL). Human primary monocytes (THP-1) were obtained from ATCC and cultured in RPMI 1640 medium. HEK-293T cells were provided by Tao Li (National Center of Biomedical Analysis) and was cultured in DMEM medium. Healthy volunteers provided the human PBMCs, which were cultivated in RPMI 1640 medium. The cell lines were developed at 37°C in an incubator with 5% CO₂ humidity. Human Ethics Number is KY-2023-2-10-1.

This experiment adopted the Cell Counting Kit 8 (CCK-8) to detect cell viability. The cells were exposed to varying amounts of GUP for a duration of 12 h. CCK-8 reagent was co-incubated with the cells in the incubator for a duration of 1 h. Finally, optical density values (O.D.) at 450 nm were measured. Based on the results of the CCK8 toxicity test for GUP, subsequent experiments were conducted at a non-toxic dose of GUP.

THP-1 cells (with PMA), human PBMCs, BMDMs and were inoculated overnight in well plates respectively at a density of 1.3 × 10⁶, 2.5 × 10⁶ and 1.2 × 10⁶ cells/mL. Next day, GUP was dissolved in opti-MEM and filtered through a non-pyrogenic filter membrane with a pore size of 0.22 μM. The cells were then treated with GUP. After 1 h, transfection of HT-DNA (2.5 μg/mL) or 2′3′-cGAMP (2 μg/mL), or addition of diABZI (10 mmol/mL) or DMXAA (2.5 μg/mL) was carried out in the cells. For Western blot analysis, we collected cell lysates after 2 h. After 4 h, lysates from cells lysed with TRIZOL were collected for qPCR.

According to the instructions, serum and intraperitoneal lavage liquid of mice were tested by mouse IL-6 (1210602, Dakewe), mouse TNF-α (1217202, Dakewe) and IFN-β (luex-mifnbv2, InvivoGen).

The conventional method of protein extraction was used (Wang Z. et al., 2019). Cell supernatant was collected, and centrifuged for 5 min, and whole cell lysate was prepared using 1× RIPA buffer. The supernatant of the centrifuged cell supernatant was then discarded and the fully lysed cell lysate was added. 10% SDS-PAGE was used to separate the protein samples at 100 V, and wet transfer was used to transfer the gels onto PVDF membranes. The PVDF membrane was then closed at room temperature in 10% skimmed milk for 1 h. We incubated PVDF overnight at 4°C with primary antibody, followed by 1 h with matched secondary antibody. Washing with TBST was required between each step. Detection of protein bands used enhanced chemiluminescence reagents (Millipore, Massachusetts, United States).

Total RNA was collected using the TRIZOL reagent (Sigma, 93289) following the guidelines provided by the manufacturer. The RT Master Mix was utilized to transcribe cDNA from RNA in equal quantities (A230, GenStar). The SYBR Green qPCR Master Mix (HY-K0501, MCE) was utilized for conducting real-time PCR. Table 1 and Table 2 display the primer sequences.

For immunoprecipitation assay, HEK-293T cells were inoculated overnight. Cells were transfected with Flag-IRF3 or Flag-TBKI and HA-STING or HA-IRF3 plasmids for 18 h. After the treatment with GUP for 4 h, cell pellets were collected and resuspended in lysis buffer (50 mM Tris [pH 7.45], 0.5% [v/v] Triton X-100, 150 mM NaCl, and 1% [m/v] containing protease inhibitors. In addition, the supernatant was incubated with M2 (A2220, Sigma) beads for 4 h at 4°C. The samples were washed three times with the lysis buffer, denaturalized by boiling with 1×RIPA loading buffer for 15 min and analyzed by immunoblotting.

Oligomerization experiments of STING was performed as previously reported (Li S. et al., 2018). In short, the cell lysates were loaded onto the native-PAGE gel using the native sample buffer, and electrophoresed for 50 min at 25 mA, and then analyzed by immunoblotting.

C57BL/6 wild Eighteen 8-week-old C57 mice were randomized and equally divided into three groups and administered vehicle or GUP (400 mg/kg) by gavage for five consecutive days. According to the relevant research reports and the exploration in the early stage of the experiment, we finally determined the dose selection of GUP (400 mg/kg) for subsequent animal experiments (Huang C. et al., 2022; Wu et al., 2022). On the sixth day, the mice were injected intraperitoneally injection (i.p.) with DMXAA (25 mg/mL). DMXAA dose selection as previously described (Khan et al., 2021). After 4 h, the mice were euthanized, intraperitoneal lavage liquid and serum were collected and DMXAA-induced cytokine levels were measured. GUP was dissolved in 0.9% saline and then administered to mice by gavage. DMXAA was dissolved in 1:1 with PEG300: normal saline.

CLP sepsis was induced with C57BL/6 female mice (8 weeks old) as mentioned earlier (n = 6 mice per group, 18 in total) (Zeng et al., 2017). In short, the mice were anesthetized and the abdominal surgical area was sterilized. Next, a minor incision was created in the middle of the abdomen, and the cecum was brought outside and secured with silk by tying it. We punctured the ligated portion of the cecum with a 22-gauge needle twice before returning it to the abdominal cavity. Finally, the abdomen was sutured into two layers and mice were immediately injected with saline at 50 mg/kg to prevent shock. Sham surgeries involve exteriorizing the cecum and reintroducing it into the abdominal cavity without ligating or puncturing it. Immediately after the operation, Ringer’s solution was injected subcutaneously into the mice.

After 72 h, the mice were executed through cervical dislocation and lung, heart, liver and kidney tissues were taken. One of the lung tissues was paraffin-embedded, followed by hematoxylin-eosin (H&E) staining using standard techniques. Heart, liver and kidney tissues were homogenized and then RNA was extracted for qPCR.

The software GraphPad Prism eight and SPSS statistics 26.0 were utilized for the analysis of all experimental data. The statistical evaluation was conducted by an unpaired Student’s t-test between two groups or one-way ANOVA with Dunnett’s post hoc test was used for multi groups. The results were shown as Mean ± Standard error of mean (SEM), and p values were presented by *p < 0.05, **p < 0.01 and ***p < 0.001.

In order to clarify the evidence for the chemical characterization of GUP, ion chromatography was utilized to determine the monosaccharide composition. The monosaccharide component of GUP was measured by comparison with the reservation time and chromatograph peaks of standard monosaccharides including glucose, galactose and glucuronide acid. Data revealed that the molar ratio of glucose, glucuronide acid and galactose in the GUP sample was 0.902, 0.056, 0.043 respectively (Supplementary Figures S1A, B). The structural features of the GUP were analyzed by FT-IR and the IR spectra were scanned in the range of 4,000-400 cm^-1 and the results are shown in Supplementary Figure S1C. The infrared spectrogram shows the typical characteristics of GUP (Mowat and Agace, 2014; Nosalova et al., 2013), from which it can be seen that the broad and strong absorption peak at 3,388.50 cm^-1 is the O-H stretching vibration of the sugar, the smaller absorption peak at 2,929.65 cm^-1 is the stretching vibration of the methyl or methylenecyclohexane C-H, and the peaks between 1,400∼1,200 cm^-1 are supposed to be the sugar's C-H variable angle vibration, from the above it can be determined that GUP is polysaccharide. The peaks at 1,616.80 and 1,416.30 cm^-1 are caused by the asymmetric and symmetric telescopic vibration of C=O in saccharide (Pastorino et al., 2018; Ren et al., 2022a; Ren et al., 2022b), which indicates that GUP contains saccharide, and the results of the determination of the composition of monosaccharides containing glucuronide results are consistent with the results. The peaks between 1,200∼1,000 cm^-1 between 1,200 and 1,000 cm^-1 is the characteristic absorption peak of pyranose (Rizzato et al., 2017; Schulte et al., 2013). The above infrared test results indicate that GUP is acidic polysaccharides and contain pyranose rings. Methylation with GC–MS is widely used to determine the linkage types and ratios of polysaccharides. The methylated polysaccharide was analyzed by GC–MS, and the results were summarized in Table 3, which showed that GUP contained: t-linked Glcp and few t-linked Galp, 1,2-linked Glcp, 1,4-linked Glcp, 1,6- linked Glcp, 1,6-linked Galp. We examined the cytotoxicity of GUP in BMDMs and THP-1 cells to learn more about its impact on cGAS-STING pathway activation. A different concentration of GUP was applied to THP-1 and BMDMs cells for 12 h. As shown (Supplementary Figure S2A, B), GUP had no cytotoxicity at concentrations below 3 mg/mL in both cells. Subsequent experiments were carried out at non-toxic doses.

Researches have shown that HT-DNA enters the cytoplasm through transfection and can be sensed by cGAS and then activate STING (Chanut and Petrilli, 2019). Monocytes and macrophages, as secretory cells, are essential for the regulation of the immune response and the development of inflammation (Johnston, 1988). So, we used BMDMs and THP-1 cells to explore whether GUP affected the cGAS-STING pathway. Phosphorylation of IRF3 is a pivotal downstream signaling event in the cGAS-STING pathway. Using immunoblotting analysis to detect related proteins, we found that GUP treatment suppressed the phosphorylation of IRF3 induced by HT-DNA (Figures 1A, B). Meanwhile, we evaluated the effect of GUP on mRNA expression of downstream genes related to this pathway, including IFN-β, IL-6 and TNF-α. Surprisingly, GUP inhibited the mRNA expression of these genes in BMDMs (Figures 1C–E) and THP-1 (Figures 1F–H) cells. More importantly, GUP also inhibited HT-DNA-induced mRNA expression of IFN-β, IL-6 and TNF-α in human PBMCs (Figures 1I–K). In conclusion, these data demonstrate that GUP can suppress the activation of HT-DNA-induced cGAS-STING pathway in vitro.

We next investigated GUP affects the activation of the cGAS-STING pathway induced by multiple STING agonists. Some natural or enzymatically synthesized STING agonists, such as 2′3′-cGAMP, diABZI and DMXAA, can also induce the release of type Ⅰ IFNs and other pro-inflammatory factors (Jin et al., 2011; Messaoud-Nacer et al., 2022). We further study found that GUP inhibited the phosphorylation of IRF3 induced by various typical STING stimulators (HT-DNA, 2′3′-cGAMP, diABZI and DMXAA) in BMDMs (Figure 2A) and THP-1 (Figure 2F) cells. In addition, to further validate the above results, we also performed qPCR on its related downstream genes. Interestingly, GUP markedly inhibited the mRNA expression of IFN-β, IL-6, TNF-α and interferon-stimulated genes (CXCL10) activated by multiple STING stimulators in BMDMs (Figures 2B–E) and THP-1 (Figures 2G–J) cells. All in all, these results strongly demonstrate that GUP has an essential regulatory effect in cGAS-STING pathway activation triggered by multiple stimulus.

Nuclear translocation is an important function in transcription. IRF3 is phosphorylated to form a dimer that translocate to the nucleus and thereby regulates the expression of type I IFNs (Zhang et al., 2019). Here, we explored whether GUP could inhibit the nuclear translocation of IRF3, thereby suppressing transcriptional activity. As shown in the picture (Figure 3B), GUP inhibited the nuclear accumulation of IRF3. To verify these results, we performed an immunofluorescence assay. Our results showed (Figure 3A), GUP significantly reduces diABZI induced nuclear aggregation of IRF3. The present study demonstrated that GUP blocked IRF3 transcriptional activity and thereby suppressed the expression of type I IFNs.

Next, we aim to explore the mechanisms by which GUP influences cGAS-STING pathway activation. Our previous research demonstrated that GUP could suppress the phosphorylation and nuclear translocation of IRF3, a key downstream signaling event in this signaling pathway. Therefore, we will further explore the role of GUP on the upstream of this pathway. Studies have shown that cGAMP causes a conformational change in STING, resulting in oligomerization of STING and subsequently recruits TBK1 to activate downstream pathways. However, our results indicated that GUP did not inhibit 2′3′-cGAMP-induced STING oligomerization (Figure 3C). STING is a scaffold protein that, upon binding to CDNs, recruits downstream signaling proteins (Li S. et al., 2018). After STING recruits TBK1, IRF-3 transcription factors are subsequently recruited to the signaling complex and phosphorylated by TBK1, which then regulates downstream pathways (Zhao et al., 2019; Ren Y. et al., 2022). Thus, binding between STING, TBK1 and IRF3 is necessary for activating cGAS-STING pathway. To this end, we examined whether GUP inhibited the interaction between three key proteins in this pathway, STING, TBK1 and IRF3. Notably, our results showed that in exogenous scenarios, GUP inhibited STING interactions with TBK1 (Figure 3D) and TBK1 interactions with IRF3 (Figure 3E), but not STING-IRF3 interactions (Figure 3F). In conclusion, these results suggest that GUP inhibits the activation of the cGAS-STING pathway by inhibiting the interaction of STING with TBK1 and TBK1 with IRF3.

Our next step was to investigate how GUP affects the activation of the cGAS-STING pathway in vivo. DMXAA, a mouse interferon gene (STING) stimulator (Xu et al., 2021), was also a potent inducer of type I IFNs and inflammatory cytokines, so we chose DMXAA for our in vivo study. Mouse were given GUP (400 mg/kg/day) for 5 days and DMXAA was injected intraperitoneally on the sixth day. After 4 h, serum and intraperitoneal lavage liquid were gathered from the mouse and the expression of IFN-β, IL-6 and TNF-α in the serum and intraperitoneal lavage fluid were measured by Elisa kits. Our results showed that DMXAA stimulation raised the expression of IFN-β, IL-6 and TNF-α in the intraperitoneal lavage liquid (Figures 4B–D) and serum (Figures 4E–G) of mouse. Notably, GUP markedly decreased DMXAA-mediated expression of IFN-β, IL-6 and TNF-α. Thus, GUP can inhibit the activation of the cGAS-STING pathway in vivo.

Study showed significant activation of cGAS-STING pathway in Cecal Ligation and Puncture (CLP)-induced sepsis (Li J. et al., 2022). In recent years, various models of sepsis have been studied, and CLP model is the most associated animal model for clinical sepsis (Buras et al., 2005; Dejager et al., 2011). To determine the effect of GUP on sepsis, CLP was conducted following the previous description (Zeng et al., 2017). GUP (400 mg/kg/day) was treated to mouse for 5 days before CLP and repeated oral administration to mouse at 12, 24, 48 and 72 h after CLP. Through the analysis of H&E staining of lung and intestinal tissues of mouse, we found that CLP-induced lung and intestine increased leukocyte infiltration and significant accumulation of inflammatory cells in the alveolar space, in accordance with previous researches (Zeng et al., 2017). Our results showed that GUP reduces damage to lung and intestinal tissues (Figure 5A). Furthermore, in CLP-induced heart, liver and kidney tissues, the mRNA expression of IFN-β and TNF-α was increased at 72 h, whereas GUP decreased the mRNA expression of IFN-β (Figures 5B–D) and TNF-α (Figures 5E–G). In general, these data suggest a protective effect of GUP against CLP sepsis. Combining the results of previous experiments and the action of the cGAS-STING pathway in the inflammatory response (Decout et al., 2021), we speculate that GUP may exert its anti-inflammatory effects by down-regulating the cGAS-STING pathway.