Cordyceps sinensis extract protects against acute kidney injury by inhibiting perforin expression in NK cells via the STING/IRF3 pathway

RNA-seq transcriptome analysis

Screening and processing of the original sequence data

Using the GEO datasets of the National Center for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov/), datasets GSE145085 and GSE145085 and the biological project PRJNA605973 related to AKI were obtained. Eight public FASTQ files in PRJNA605973 were downloaded: SRR11066831, SRR11066835, SRR11066839, SRR11066843, SRR11066847, SRR11066851, SRR11066855 and SRR11066859. Public FASTQ files were downloaded from the European Nucleotide Archive (https://www.ebi.ac.uk/ena). Trim-galore (v0.6.5-1) was used to trim the original sequence. FastQC and MultiQC were used to obtain gene quality control reports. Raw sequence reads were aligned with the human genome hg19 using Subjunc (v2.0.1). Gene expression levels were quantified using featureCounts (v2.0.1) based on the GENCODE human gene model version 19 (GRCh37.p13), and the expression matrix was obtained.

Differential gene expression and enrichment analyses

The expression matrix was assessed using intergroup and principal component analyses (PCA). Differential gene expressions between the cisplatin and untreated groups were predicted by a linear model using the Bioconductor package “DESeq2” [25]. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic pathway enrichment analyses were performed using the Bioconductor package “clusterProfiler” to obtain the key pathway information and other related information, visualized using R language [26].

Protein-protein interaction (PPI) network analysis

The PPI network was analyzed through the STRING pathway (https://string-db.org).

Isolation and purification of CS

CS (trade name Corbrin Capsule), also known as Bailing Jiaonang in the Chinese Pharmacopoeia (Lot number: 2001103), was obtained from Hangzhou Zhongmei Huadong Pharmaceutical Co., Ltd. (Hangzhou, China). 2’-deoxyadenosine was purchased from MedChemExpress (HY-W040329, USA). CS powder contains a variety of water- and alcohol-soluble nucleoproteins, including proteins, amino acids, mannitol, nucleosides (such as adenosine and cordycepin), nucleobases, sterols, saturated and unsaturated fatty acids, purines, and pyrimidines. Therefore, alcohol, hydrolysis, and enzymatic hydrolysis have been used for CS extraction. Alcohol extraction: CS powder (10 kg) was soaked in 80% medicinal alcohol for 2 days, and the alcohol extract was obtained by filtration and concentration under reduced pressure. Hydrolysis: Ten times the volume of pure water was added to the residue, which was then boiled for 2 h. This step was repeated to obtain the aqueous extract, which was then concentrated under reduced pressure, heated, and concentrated in the upper compartment of an electric furnace. Trypsin was added to the residue after water extraction, and enzymolysis was performed. The supernatant was separated by centrifugation, and the enzyme was eliminated. Subsequently, the three concentrated solutions were combined and homogenized to obtain 9.8 L of bacterial solution (1.02 g/mL), which was sterilized using UV light. Owing to the incomplete extraction process, CS was lost during the operation. If 3% was used to compensate for the loss, the concentration of the CS extract was calculated as 1.05 g/mL. Nucleosides and nucleobases in CS were analyzed using a high-performance liquid chromatography (HPLC) system (Agilent Technologies, Santa Clara, CA, USA) [27]. An ULtimate AQ-C18 (4.6 mm × 250 mm, 5 μm) column was employed to purify nucleosides and nucleobases. The mobile phase was composed of 0.1% (v/v) phosphoric acid solution as eluent A and methanol as eluent B. Gradient elution was conducted as follows: 0–5 min, 0% B; 5–10 min, 20% B; 10–15 min, 50% B; 15–16 min, 0% B. The column temperature was 40°C, the flow rate was 1.0 mL/min, and the injection volume was 5°L. UV/vis spectra were measured over a wavelength range of 200–400 nm, and the peaks of the nucleosides and nucleobases were captured at 260 nm. Among the quantified nucleoside and nucleobase compounds, adenosine was the most abundant in CS (2136.96 g/mL, followed by 2’-deoxyadenosine (203.36 g/mL) (Figure 1) (Tables 1 and 2).

Figure 1. HPLC chromatogram of CS. A representative HPLC chromatogram was acquired at 260 nm of (A) component-rich extract obtained from CS and (B) standards. Peaks were tentatively identified as: 1, adenine 2, uracil; 3, hypoxanthine; 4, uridine; 5, adenosine; 6, 2’-deoxyadenosine; 7, guanosine hydrate; 8, thymidine.

Herb formulation ingredient collection and active ingredient-associated target prediction

The TCM Systems Pharmacology Database and Analysis Platform (TCMSP) is a unique platform that provides pharmacological information on traditional Chinese medicine, including herbs. According to the literature recommendations of TCMSP, OB ≥ 30% and DL ≥ 0.18 were selected as screening thresholds. DrugBank data, a network-based tool for predicting the most likely targets of small molecules, was used to screen the relevant targets of each chemical component [28–31].

Animals and treatments

A total of 30 male C57BL/6 mice (Charles River, Wilmington, MA, USA, approved by the Laboratory Animal Welfare and Ethical Review Board of Shanghai Jiao Tong University School (Number: XHEC-STCSM-2022-051), weighing 25–30 g and aged 8–10 weeks, were randomly and equally divided into five groups (n = 6 per group): normal control (NC) group, folic acid (FA) group, and low-dose 2’-deoxyadenosine-treated (L-Deo) group, medium dose 2’-deoxyadenosine-treated (M-Deo) group, and high-dose 2’-deoxyadenosine-treated group (H-Deo). Mice in the FA group were intraperitoneally injected with a single dose of folic acid (100 mg/kg dissolved in 0.9% NaHCO₃) to induce kidney injury. Renal injury was also induced in the FA + CS group (n = 10), similar to that in the FA group. Nevertheless, the mice were immediately treated with CS extract (150 mg/kg), and CS injection lasted for two consecutive days following folic acid injection. Mice in the NC group (n = 10) were injected with equal amounts of saline. The mice were then placed on a thermostat to maintain the body temperature at 36.5–37°C and sacrificed 48 h after disposal. The left kidneys of the mice were harvested and snap-frozen for the isolation of RNA or proteins, and the right kidneys were perfused with phosphate-buffered saline (PBS) and fixed in 10% buffered paraffin for histological examination.

Cells and culture conditions

NK92 cells (ATCC) were stimulated for 4–8 days with or without IL-2 (1000 IU/mL) in RPMI 1640 medium supplemented with 10% fetal calf serum, penicillin (100 U/mL), streptomycin (100 g/mL), glutamine (2 mM), sodium pyruvate (1 mM), HEPES (10 mM), and 2-ME (0.5 mM) under a 5% CO₂ atmosphere at 37°C. Stimulated NK cells were used as effector cells.

Biochemical assay and enzyme-linked immunosorbent assay (ELISA)

Mouse blood was collected in a tube containing heparin using the tail-clamping method, and plasma was collected by centrifugation. According to the manufacturer’s protocol, SCr and BUN plasma concentrations were determined using kits from the Nanjing Jiancheng Institute of Biological Engineering (C011-2-1, C013-2-1). Kidney injury molecule-1 (Kim-1) and neutrophil gelatinase-associated lipocalin (NGAL) levels were measured using an ELISA Kit (Abcam, ab213477, ab199083, USA).

Histology and immunohistochemistry (IHC)

Mouse kidney tissue was fixed in 4% paraformaldehyde for 3 days and dehydrated using an alcohol gradient. The tissue was then immersed in an ethanol–xylene (1:1) mixture in xylene for 15–20 min until the tissue became transparent. Transparent tissue was placed in a mixture of paraffin and xylene (1:1) for 1 h, followed by paraffin embedding. Paraffin sections were cut into 5-μm-thick sections and dried in a 37°C incubator. The immunohistochemistry kit was purchased from MXB Biotechnologies (KIT-9710 and DAB-0031, China). Renal tissue paraffin slices were washed with water and subjected to antigen retrieval. The slices were then incubated with primary antibodies against Kim-1 (Abcam, 1:200, ab78494), NGAL (Abcam, 1:200, ab70287), and the appropriate biotin-conjugated secondary immunoglobulin G (1:1000, S0001, Affinity, USA). After mounting the slices, co-stained images were captured using a Nikon Eclipse 90i fluorescence microscope.

Immunofluorescence staining

Fluorescein-labeled antibodies were used against the corresponding antigens. PBS (0.01 mol/L, pH 7.4) was replaced every 10 min to maintain specimen humidity. The specimens were covered with fluorescent-labeled antibodies and stored in an enamel box for 30 min. The glass slide held on the glass stand was rinsed with PBS solution (0.01 mol/L, pH 7.4) in three cylinders, each washed for 3–5 min under oscillation. Excess moisture was absorbed from the glass slide, and a drop of glycerol was added to cover the glass. The slides were observed under a fluorescence microscope. The primary perforin antibody was purchased from Thermo Fisher Scientific (1:500, #MA5-12469, USA), and secondary fluorescent immunoglobulin G was purchased from Abmart (1:1000, M21014S, China).

Western blot analysis

Kidney tissues were lysed using radioimmunoprecipitation assay (RIPA) buffer containing a protease inhibitor cocktail (Sigma-Aldrich, PE0230, USA). Samples were vortexed and centrifuged at 15,000 × g, 4°C for 25 min. The whole-cell lysate was collected and separated using SDS-PAGE (100 V, 1.5 h), transferred to a PVDF membrane (20 V, 30 min), and blocked for 2 h. The membrane was then probed with primary antibodies for 1 h at 37°C and with the corresponding secondary antibody at room temperature for 30 min. Immunoreactive bands were detected using enhanced chemiluminescence and quantified using ImageJ software. Primary antibodies against STING (1:500; #13647), IRF3 (1:500; #4302), and P-IRF3 (1:500; #29047) were purchased from Cell Signaling Technology (USA). Anti-perforin antibody (1:500, #MA5-12469) was obtained from Thermo Fisher Scientific. Antibodies against Kim-1 (1:1000, ab233720) and NGAL (1:1000, ab63929, ab188551) were obtained from Abcam.

qRT-PCR analysis

Total RNA was extracted from the kidneys using TRIzol and Zymogen RNA extraction kits (Takara, RR820A, China). cDNA was synthesized using the PrimeScript RT Reagent kit (Takara, RR047A, China) in accordance with the manufacturer’s instructions. Gene-specific primers (Life Technologies, Carlsbad, CA, USA) were designed, and quantitative real-time PCR was performed on an ABI 7500 system. A comparison threshold was used to calculate the absolute mRNA number. Quantitative PCR mRNA data were normalized to that of the GAPDH signal used as an internal control (Table 3). The normalized delta threshold cycle value was calculated according to the manufacturer’s instructions.

Statistical analysis

Data are expressed as mean ± standard error (SE). Comparisons between time points were performed using one-way analysis of variance (ANOVA), followed by Tukey’s honest significant difference (HSD) test. Differences were considered statistically significant at P < 0.05.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article.

Transcriptome analysis of AKI revealed that AKI was affected by immune regulation

Quality control of the pruned gene sequence data revealed standard quality (Supplementary Figure 1). After comparing and enumerating the pruned gene data, the differentially expressed genes (DEGs) with statistical significance were screened using the Bioconductor package DESeq2 (P < 0.05) (Figure 2A). In the volcano plot, 399 DEGs were upregulated, and 1126 DEGs were downregulated (Figure 2B). A heat plot of the top 30 upregulated and top 30 downregulated genes was constructed (Figure 2C). The upregulated DEGs were selectively enriched by GO using P < 0.05 as screening criteria. Our data showed that GO cell components were mainly enriched in the secretory granule lumen, apical plasma membrane, cytoplasmic vesicle lumen, and vesicle lumen (Table 4, Figure 2D–2F). The KEGG analysis showed enrichment of cytokines and cytokine receptors, inflammatory factors, tumor necrosis factor, apoptosis, and other pathways (Table 5, Figure 3A, 3B). Notably, NK cell-mediated cytotoxicity pathways were also enriched. However, the association between AKI and NK cell-mediated cytotoxicity has rarely been investigated. The correlation between NK cell-mediated signaling and AKI requires further investigation (Figure 3C).

Figure 2. Differential transcriptome analysis of AKI samples. (A) Correlation of heatplot of differences between groups: a significant difference was observed between cisplatin and untreated groups in all the differently expressed genes (P < 0.05). (B) Differential gene volcano plot: 399 differentially expressed genes were up-regulated and 1126 differentially expressed genes were downregulated. The cutoff for logFC is 2.957. (C) Heatplot of the top 30 up-regulated genes and the top 30 downregulated genes. (D) Top 30 GO enriched analysis results in the dotplot (P < 0.05). (E) Correlation of the GO enriched analysis results in the goplot (P < 0.05). (F) Correlation of differently expressed genes in the GO enriched analysis in the cnetplot.

Figure 3. Natural killer cells participate in immune regulation of AKI. (A) Top 30 KEGG enriched analysis results in the dotplot (P < 0.05). (B) Heatplot of the regulated genes. (C) NK cell-mediated cytotoxicity pathways.

2’-deoxyadenosine administration attenuated tubular injury in AKI mice induced by FA

To explore whether 2’-deoxyadenosine can reduce the injury level by regulating the immune response in vivo, we established a mouse model of AKI induced by FA. We treated it with different doses of 2’-deoxyadenosine. Two days after FA administration, the average SCr and BUN levels were significantly higher in the FA group than in the NC group. The SCr and BUN levels in the FA+H+Deo group were significantly lower than those in the FA group (Figure 4A, 4B). Similar features were observed for NGAL and Kim-1 in the FA-induced AKI mouse model (Figure 4C, 4D). Therefore, we selected high doses of 2’-deoxyadenosine for subsequent experiments.

Figure 4. 2’-deoxyadenosine administration attenuated the development of FA-induced AKI in mice. (A, B) Renal function of all mice was assessed by SCr and BUN. (C, D) Kim-1 and NGAL results of the mice were measured by ELISA. (E, F) Representative H&E and PAS of renal tissues obtained from specimens in different group. (G, H) Expression of Kim-1 and NGAL was determined by IHC. (I–K) Protein levels of NGAL and Kim-1 protein expression in the kidney were evaluated by western blot analysis. ^*P indicates a significant difference between the NC group and the FA group. ^#P represents the difference between the FA+Deo group and the FA group. P < 0.05. Values are shown as means ± SD from three independent experiments. Representative images are shown from a total of six animals per group.

To illustrate the therapeutic effect of high-dose 2’-deoxyadenosine in AKI mice, we stained kidney samples from mice with H&E and PAS. Morphological changes in FA-induced AKI mice primarily manifested in the proximal tubules of the subcapsular region of the renal cortex. Obvious edema and vacuoles in the renal proximal tubule epithelial cells dilated the renal tubule lumen, and swelling of the mesangial spaces and cells in some glomeruli suggested kidney damage. The specimens in the FA+H-Deo group exhibited milder pathological damage than those in the FA group, suggesting that the substantial damage to the kidney caused by FA injection was markedly prevented with the high doses of 2’-deoxyadenosine treatment (Figure 4E, 4F). IHC staining of the specimens confirmed that the AKI markers NGAL and Kim-1 expression levels were remarkably higher in the AKI group. This upregulation was blocked by high doses of 2’-deoxyadenosine treatment (Figure 4G, 4H). The western blotting analysis consistently indicated the preventive function of high doses of 2’-deoxyadenosine treatment against AKI (Figure 4I–4K).

2’-deoxyadenosine reduced perforin expression through the STING/IRF3 signaling pathway in the FA-induced AKI mouse model

Perforin plays a major role in NK cell-mediated TEC injury. NK cell infiltration in kidney tissues was detected by immunofluorescence staining for the NK cell marker perforin (Figure 5A). These observations were further confirmed at the protein and mRNA levels by WB and qRT-PCR, respectively. FA mice showed significantly increased perforin expression compared with the NC group. These effects were alleviated by high doses of 2’-deoxyadenosine treatment (Figure 5B–5D). Bioinformatic analysis of PPI revealed an interaction between STING (TMEM173) and its downstream signaling molecules, TBK1, IRF3, and type I IFN (Figure 5E). IFN-γ (type II IFN), an important inflammatory indicator of kidney injury, was mainly released from NK and NKT cells. Through qRT-PCR analyses, we observed high expression of STING in AKI mice, and found that high doses of 2’-deoxyadenosine can reduce the expression level (Figure 5F). Treatment can also effectively reduce the release of inflammatory factors IFN-γ released by NK cells (Figure 5G). To verify the in vivo results of the inhibitory effect of high doses of 2’-deoxyadenosine on the STING pathway, we evaluated the levels of STING and IRF3 phosphorylation in kidney homogenates and found that they were downregulated upon high doses of 2’-deoxyadenosine injections by western blotting (Figure 5H–5J). These results confirmed the suppressive effect of high doses of 2’-deoxyadenosine on perforin and IFN-γ release via inhibiting STING/TBK1/IRF3 expression.

Figure 5. 2’-deoxyadenosine protected against AKI by inhibiting perforin expression in NK cells by regulating STING/IRF3 signaling pathway in vivo. (A) Immunofluorescence studies of perforin expression on NK cells after FA treatment. DAPI staining was used to determine the number of nuclei. Anti-perforin staining was used to determine perforin expression (×400). (B, C) Protein levels of perforin in the kidney were evaluated by western blot analysis. (D) The mRNA levels of perforin in the kidney were evaluated by qRT-PCR analysis. (E) Bioinformatics of STING (TMEM173) and its regulated IRF3 from STRING PPI database. (F, G) The mRNA levels of STING and IFN-γ in kidney were evaluated by qRT-PCR analysis. (H–J) Protein levels of STING, p-IRF3/IRF3 expression in kidney were evaluated by western blot analysis. ^*P indicates a significant difference between the NC group and the FA group. ^#P represents the difference between the FA+Deo group and the FA group. P < 0.05. Values are shown as means ± SD from three independent experiments. Representative images are shown from a total of six animals per group.

2’-deoxyadenosine protects against AKI by inhibiting perforin expression in NK cells via the STING/IRF3 pathway in vitro

To confirm the mechanism by which high doses of 2’-deoxyadenosine inhibited perforin and IFN-γ expressions in NK cells via STING/IRF3 signaling to attenuate AKI, we performed in vitro experiments. TCMSP analysis demonstrated that the composition of CS was consistent with the purification results. Based on the DrugBank database, the related targets of 2’-deoxyadenosine were identified as PTGS1 (COX-1), PTGS2 (COX-2), and MTAP (Figure 6A). PPI analysis revealed a strong correlation between PTGS1 (COX-1), PTGS2 (COX-2), STAT3, STING, TBK1, and IRF3 (Figure 6B). Furthermore, we explored whether 2’-deoxyadenosine could inhibit NK cell activation by inhibiting STING and IRF3 by targeting the inhibition of COX-2 and downstream STAT3.

Figure 6. 2’-deoxyadenosine inhibited the activity of NK cells by regulating the STING/IRF3 signaling pathway in vitro. (A) Related target of 2’-deoxyadenosine with PTGS1 (COX-1), PTGS2 (COX-2), and MTAP based on DrugBank and TCMSP analysis. (B) Bioinformatics of PTGS1 (COX-1), PTGS2 (COX-2), STAT3 and its regulated STING/IRF3 from STRING PPI database. (C–H) The protein levels of COX-2, STAT3, STING, IRF3, and perforin in NK cells were evaluated by western blot analysis. (NK: naive NK cells. NK+IL-2: NK cells+IL-2 group. NK+IL-2+deo: NK cells+IL-2+2’-deoxyadenosine group. NK+deo: naive NK cells+2’-deoxyadenosine group). (I–K) The mRNAs levels of IFN-γ, TNF-α and CCL2 expression in NK cells were evaluated by qRT-PCR analysis. (NK: naive NK cells. NK+IL-2: NK cells+IL-2 group. NK+IL-2+deo: NK cells+IL-2+2’-deoxyadenosine group. NK+deo: naive NK cells+2’-deoxyadenosine group.) ^*P indicates a significant difference compared with the NK group. ^#P represents the difference compared with the NK+IL-2 group. P < 0.05. Values are shown as means ± SD from three independent experiments.

Since IL-2 secreted by T lymphocytes is responsible for stimulating the growth and differentiation of T cells, B cells, and NK cells [32], we used 1000 IU/mL IL-2 to treat NK cells in vitro to determine the source of perforin and IFN-γ release. Strikingly, activation of NK cells resulted in significantly increased expressions of COX-2, STING, p-IRF3, and perforin and decreased STAT3 by western blotting, whereas concurrent exposure to 2’-deoxyadenosine inhibited these protein expressions and elevated STAT3 expressions (Figure 6C–6H). Next, we examined the mRNA expression of inflammatory factors by qRT-PCR. As indicated in Figure 6I–K, the mRNA levels of IFN-γ, TNF-α, and CCL2 (chemokine monocyte chemoattractant protein-1, MCP-1/CCL2) were significantly elevated in IL-2-treated NK cells compared with those in the naïve NK group and were remarkably relieved by treatment with 2’-deoxyadenosine.