Cell culture medium and human primary renal PTECs were provided by ScienCell (San Diego, CA, United States). Anti-SLC2A9 (URAT1) antibodies, TSG101, CD63, caspase-1, GSDMD, IL-1β, caspase-3, caspase-8, caspase-9, cytochrome c, Bcl-2, and Bax were purchased from ABclonal (Wuhan, China). Oxonic acid (OA) and UA were provided by Sigma (St. Louis, MO, United States). CIAS1/NALP3, GAPDH, and anti-GLUT9 were provided by Abcam (Cambridge, United Kingdom). In addition, the CCK-8 assay kit was provided by Jiwei Biological Technology (Shanghai, China).

GGQLD, a famous decoction documented in the Treatise on Exogenous Febrile Disease around 1,900 years ago, was used and approved by the Huadong Hospital Affiliated to Fudan University. The dried herbs, including Pueraria montana var. lobata (Gegen), Scutellaria baicalensis Georgi (Huangqin), Glycyrrhiza uralensis Fisch. ex DC (Gancao), and Coptis chinensis Franch (Huanglian), in the ratio of 8:3:3:2 (w/w/w/w), were first soaked in distilled water with 15-fold volumes of herbs (v/w) for 30 min and then extracted by decoction twice, 2 h for the first time and 2 h for the second time, with 8-fold volumes of water to herbs (v/w). After filtration, the solution was evaporated under reduced pressure to a suspension with a final density of 1 g/ml and stored at 4°C for further use (Xie et al., 2006).

Puerarin, baicalin, liquiritin, and berberine (all purities >98%) were provided by Shanghai Yuanzhi Biotechnology Co., Ltd. (Shanghai, China). Supplementary Figure S1 presents the chemical structures of the ingredients. Acetonitrile, formic acid, and methanol of HPLC grade were provided by Merck Company (Darmstadt, Germany). The Milli-Q system (Millipore, Milford, MA, United States) was utilized to obtain deionized water. Other reagents of analytical grade were acquired from commercial sources. To identify the components of GGQLD, we performed HPLC analysis using an Agilent 1260 Infinity brand chromatographic chain with a C18 chromatographic column (250 × 4:6 mm, 5 μm, Thermo Fisher Scientific, NY, United States), along with a column oven, a binary pump, a diode array detector, and an autosampler (Agilent 1100 series). Thereafter, 20 ml ethanol extract was added to the column. Throughout the whole chromatographic analysis process, the column injector was maintained at 25°C. The mobile phase comprised the ammonium formate solution (10 mmol/l, pH 3.0, A) and acetonitrile. For the components, their detection wavelength was set at 270 nm. Subsequently, we identified each compound according to the pure standard absorbance spectra and retention time (Guo et al., 2019).

We searched the active ingredients of Gegen, Huangqin, Huanglian, and Gancao in GGQLD against the TCMSP database (http://tcmspw.com/index.php) by the thresholds of druglikeness (DL, also known as the similarity between the known medicines and the components) ≥0.18 and oral bioavailability (OB, medicine component oral availability) ≥30%. Based on the TCMSP database (Yao et al., 2013; Liu et al., 2020), we aligned the GGQLD active ingredients with the candidate targets separately, then retrieved targets of diverse origins against the UniProt database (http://www.uniport.org), and acquired the official target gene symbols to perform subsequent assays. A total of 196 HUA-associated targets were identified from the DisGeNET (http://www.disgenet.org) and NCBI databases. Cytoscape 3.8.0 software is the openly accessible bioinformatics platform used to visualize the molecular interaction networks (Deng et al., 2020). In the current study, we adopted Cytoscape to construct the active ingredient–target–disease–medicine interaction network. GO functional annotations and KEGG pathway analyses on the critical therapeutic targets in GGQLD against HUA were analyzed by adopting the DAVID 6.8 database (Komada et al., 2018).

From January 1st, 2020 to October 31st, 2020, a total of 30 male patients with asymptomatic hyperuricemia were enrolled from the Inpatient and Outpatient Departments of the Huadong Hospital Affiliated to Fudan University (Shanghai, China). The patient inclusion criteria are shown below, those aged 18–75 years with a diagnosis of HUA (serum uric acid, SUA > 420 μmol/l). For patients receiving anti-HUA treatment, they should receive washout for 2 weeks, and we only enrolled patients with SUA > 420 μmol/l post-washout. The patient exclusion criteria were as follows: 1) those with an allergic physique or previous allergic history to benzbromarone or TCM, 2) serum creatinine >1.5 mg/dl, 3) two-fold elevation of ALT compared with the normal upper limit, 4) serious stiffness or deformity due to gouty arthropathy, 5) clinically significant arrhythmia, and 6) alcohol abuse history. In addition, patients confirming to any one of the following conditions were also excluded: 1) those having serious concurrent diseases in the hematopoietic system, liver, cerebrovascular system, or kidney, mental diseases, or malignant cancers; 2) those taking salicylate or aspirin (>325 mg/d)-containing medications; 3) those taking hypouricemic medications, 6-mercaptopurine, or azathioprine; and 4) those who had been involved in additional clinical trials in the last 3 months.

In this study, 30 patients were given GGQLD treatment (two doses daily for 4 weeks). SUA and urine uric acid (UUA) were adopted for evaluating the therapeutic efficacy (Table 1). The Local Ethical Committee of Huadong Hospital Affiliated to Fudan University approved our study protocols (No. 20190037).

The Local Ethical Committee of Huadong Hospital Affiliated to Fudan University approved our experimental protocols on human blood (No. 20190037). To be specific, whole blood was centrifuged at 500 × g for 5 min at room temperature to separate the plasma. Thereafter, the equivalent amount of 1× PBS (under ambient temperature) was used to dilute the rest blood samples, followed by Ficoll-Paque underlaying (under ambient temperature) and 30 min of centrifugation (2,000 rpm, 21°C) using a Heraeus Multifuge X3R (Thermo Fisher Scientific), with the deceleration and acceleration being set at 0 and 5, respectively. Afterward, we obtained PBMCs from the interface between plasma and Ficoll-Paque in the 15 ml tube and used 1× PBS to wash the cells twice for 10 min (1,500 rpm, 4°C).

We collected urinary exosomes from 8 controls and 8 GGQLD-intervened HUA patients. Later, 40 ml freshly prepared urine samples were collected to obtain urinary cell pellets. Thereafter, the supernatants were subjected to 30 min of centrifugation at 12,000 × g, followed by standing to isolate the urinary exosomes. ExoQuick-TC for tissue culture media and urine (Exiqon, Woburn, MA, United States) was used to isolate urinary exosomes in accordance with specific protocols. To validate the protocols for exosome purification, we performed cryo-transmission electron microscopy to analyze the size, shape, and morphology of urinary exosomes. The isolation of exosomes was characterized by WB using NanoSight.

A total of 15 male SPF SD rats (age, 8 weeks; weight, 200–250 g) were raised at the Animal Center of Shanghai Rat and Mouse Biotech Co., Ltd. Then, all animals were randomized into three groups, including the 1) Cont, 2) OA (8-week administration of OA at 750 mg/kg/day), and 3) OA + GGQLD (8-week administration of OA at 750 mg/kg/day + 4-week administration of GGQLD at 10 ml/kg/day initiating in week 5) groups, with five in every group. At the end of week 8, each animal was terminated. The 24-h urine was sampled from animals in metabolic cages. Subsequently, the levels of sUA, uUA, serum creatinine (Scr), blood urea nitrogen (BUN), and urinary creatinine (Ucr) were measured through the enzymatic colorimetric assay using a fully automatic chemistry analyzer (MODULAR D/P, Roche). Afterward, the freshly prepared renal cortical samples were obtained immediately after the animals were terminated to measure UA in accordance with specific protocols. Part of the tissues was stained with hematoxylin–eosin (HE) or Masson’s trichrome stain for light microscopy. Next, the dissected tissue samples were preserved at −80°C to carry out immunoblotting and RT-PCR assays. Each animal experiment was approved by the Animal Care and Use Committee of Shanghai Rat and Mouse Biotech Co., Ltd. following guidelines of the National Institutes of Health (NIH Pub. No. 85-23, revised 1996).

The epithelial cell medium, which was supplemented with basal medium (500 ml), fetal bovine serum (10 ml, FBS), penicillin/streptomycin (5 ml), and epithelial cell growth supplement (5 ml), was used to culture human primary PTECs. Then, we incubated cells under 37°C and 5% CO₂ conditions. A 24-h “growth arrest” period was observed within the serum-free medium before stimulation in each experiment.

Previously, we dissolved UA into 1M NaOH to the final concentration of 50 mg/ml (Xiao et al., 2015). Then, we examined the solution so as to ensure that there was no mycoplasma and also filtered the solution (pore size, 22 μm) prior to use. No detectable crystal was observed under polarizing microscopy or in the process of cell incubation.

A Cell Counting Kit-8 (CCK-8, Beijing Solarbio Science and Technology Co., Ltd.) assay was conducted to assess cell viability after GGQLD treatments at diverse doses. Cells (1 × 104 cells/well) were cultivated within the 96-well plate for 24–48 h. The microplate reader was used to measure absorbance (OD) at 450 nm. All experiments were conducted thrice.

After GGQLD treatment, we harvested PTECs and washed them twice with ice-cold PBS. We collected six samples altogether, among which three were from the GGQLD group while three were from the control group. Trizol reagent (Takara, Dalian, China) was employed to extract total RNA in accordance with specific protocols. Libraries were constructed and mRNA was sequenced by Nuohe Zhiyuan Technology Company (Beijing, China). In addition, DESeq (version 1.30.0) was utilized to identify differentially expressed genes (DEGs). The false discovery rate (FDR) was applied to adjust p value, and the significance thresholds of fold change (FC) ≥ 2 and p < 0.05 were applied.

Flow cytometric analysis was conducted with an Annexin V-FITC apoptosis detection kit (DOJINDO; Kyushu, Japan) to analyze cell apoptosis under specific instructions. After UA and GGQLD treatments, PTECs were harvested and rinsed with ice-cold PBS, followed by staining using the binding buffer that contained Annexin V-FITC/PI at 4°C for 15 min in the dark. Finally, flow cytometric analysis (Beckman Coulter, Fullerton, CA, United States) was performed to record cells.

Cells were cultured onto a laser confocal cell culture dish (Thermo Fisher Scientific, NY, United States) in accordance with specific protocols. 24 h later, the cells were washed with PBS and fixed by 4% paraformaldehyde in PBS. Thereafter, 10% BSA contained within PBS was utilized to block the cells for 30 min, whereas the GSDMD and NLRP3 antibodies in 10% BSA were used to incubate the cells at 4°C overnight. Then, the cells were rinsed with PBS thrice and incubated with goat anti-rabbit antibodies (Invitrogen, Grand Island, NY, United States) for 1 h in the dark. The cells were washed consecutively thrice, and a confocal imaging system (LSM 780;Carl Zeiss, Jena, Germany) was employed to examine the dishes.

Trizol (Life Technologies, United States) was employed to isolate total cellular and tissue RNA in accordance with specific protocols. The content and purity of RNA were detected through a NanoDrop 2000 (Thermo Fisher Scientific, United States). To carry out mRNA quantification, we applied a reverse transcription kit (Vazyme, China) for the reverse transcription of total RNA according to specific instructions. Afterward, SYBR Green Master Mix (Vazyme, China) was utilized to perform qPCR, whereas a Roche Light Cycler system (Roche, Switzerland) was adopted for analysis, with GAPDH being an endogenous control. Table 2 lists the primer sequences. Gene expression was normalized to the GAPDH level, which was presented in the manner of FC (2−ΔΔCT). All results were repeated thrice.

Supernatants of cell culture were collected according to the previous description and then the lysis buffer supplemented with protease inhibitor cocktails (Sigma, St Louis, MO, United States) was used to lyse the rest cells. After collecting total proteins, we carried out WB analysis according to the previous description (Xiao et al., 2015). In the present work, the following primary antibodies (all were diluted at 1:1,000) were adopted, including rabbit pAb Glut9, rabbit pAb URAT1, rabbit pAb caspase-1, rabbit pAb caspase-3, rabbit pAb NLRP3, rabbit pAb IL-1β, rabbit pAb GSDMD, rabbit pAb Bax, rabbit pAb caspase-8, rabbit pAb cytochrome c, rabbit pAb Bcl-2, and rabbit pAb caspase-9.

Data were presented in the manner of means ± SD, except as otherwise noted. SPSS19.0 (SPSS, Inc., Chicago, IL, United States) was applied for all statistical analyses. For continuous variables, multivariate ANOVA was utilized to assess the differences between the groups. p < 0.05 indicated that a difference was statistically significant. GraphPad Prism 5.0 was employed for drawing plots.

A total of 907 ingredients were discovered in GGQLD through the TCMSP database, including 62, 123, 178, and 278 in Gegen, Huangqin, Huanglian, and Gancao, respectively. The thresholds of DL ≥ 0.18 and OB ≥ 30% were used to screen the active ingredients. Finally, 372 candidate active ingredients conformed to our preset screening thresholds. A total of 196 genes in HUA were screened from the database (DisGeNET). In TCM, a medication displays diverse pharmacological effects via various targets. Therefore, it is greatly significant to determine the TCM medication mechanisms in treating complicated diseases based on network analysis. Previously, we acquired a total of 196 targets for HUA from the DisGeNET database. Then, targets were predicted by incorporating 372 active ingredients in the UniProt database. Consequently, 31 overlapped targets between GGQLD and HUA were obtained by VENN map (Figure 1A).

We performed GO functional annotations of the 31 identified genes, suggesting their involvements in cell components (CCs), biological processes (BPs), and molecular functions (MFs). To be specific, these genes were mainly involved in various BPs, including inflammatory response and negative regulation of the apoptotic process (Figures 1B–D). In addition, according to KEGG pathway analysis, the intersected genes were mainly associated with 40 pathways. Among them, 35 pathways with the most significant p-values are shown in Figure 2A including cancers, apoptosis, and inflammatory signaling pathways. In addition, we also constructed the active ingredient–target–disease network by adopting Cytoscape 3.8.0 software, obtaining the interactions among drugs, compounds, diseases, and targets. The results are shown in Figure 3. Obviously, 17 targets were related to the apoptosis pathways, 14 to the metabolism pathways, 13 to the inflammation pathways, 13 to the immunity pathways, and 14 to the signal transduction pathways. As illustrated by the VENN map, we acquired 8 intersected targets in the 5 pathways (Figure 2B). Moreover, the STRING platform was adopted to establish the protein–protein interaction (PPI) (Figure 2C).

We found that the SUA and UUA levels (Figures 4A,B) in HUA patients were reduced after GGQLD treatment. Besides, the levels of, IL-6, and IL-8 (Figures 4C–E) also significantly decreased in the GGQLD group. According to our previous study, soluble UA induced NLRP3 inflammasome production. To define the possible therapeutic approaches of HUA-induced renal tubular injury, the present study analyzed the expression of NLRP3 in PBMCs and that of urinary exosome protein in HUA patients before and after GGQLD treatment. By isolating PBMCs and extracting urinary exosomes from HUA patients before and after treatment, this study conducted flow cytometry and WB assays to analyze two groups of PBMCs and urinary exosomes. As a result, GGQLD downregulated the expression of NLRP3 in HUA patients (Figure 4F,G,K).

We obtained GGQLD extract through the purification of chromatographic grade methanol, since saponin and flavone were the major bioactive ingredients in GGQLD. Besides, HPLC analysis was performed to investigate the chemical features of GGQLD extract, with methanol being the eluent. To obtain the superb efficiency and favorable separation selectivity of HPLC analysis, we optimized the composition of the mobile phase (10 mmol/l acetonitrile–water contained within the ammonium formate solution). According to Supplementary Figure S1, the major bioactive ingredient puerarin was identified in Gegen, baicalin in Huangqin, berberine in Huanglian, and liquiritin in Gancao, which were later adopted to construct a combined standard approach. Accordingly, we detected four peaks of GQT extract from the HPLC chromatogram and adopted related chemical standards to quantify levels of puerarin, liquiritin, berberine, and baicalin (Peaks A–D), respectively. To be specific, the contents of the above four active ingredients were 3.9, 0.3, 2.9, and 1.0 mg/ml in GGQLD extract, respectively (Table 3).

Furthermore, the current work verified the protection of GGQLD against HUA in rat models. In brief, gastric OA at 750 mg/kg/day was given to male SD rats for eight consecutive weeks. Meanwhile, GGQLD at 10 ml/kg/day initiating on week 5 was given for four consecutive weeks. As a result, NLRP3 staining and mRNA expression increased in renal cortical tissues of OA rats (Figures 5C,D). GGQLD reduced the NLRP3 expression in OA-treated rats (Figures 5C,D). GGQLD significantly attenuated the mRNA expression and staining of IL-1β and caspase-3 in renal tissues of UA rats (Figures 5C,G,J). Additionally, it was found that GGQLD significantly alleviated the mRNA expression of TNF, caspase-3, caspase-8, Bax, Bcl-2, and CYCS in renal tissues of HUA rats (Figures 5F,H–L).

According to the CCK-8 results, different doses of GGQLD (20, 15, 10, and 5) were used to treat the UA-induced PTECs (Supplementary Figure S2). Based on our RNA-seq data, GGQLD was found to suppress BPs related to mitochondrial apoptosis and inflammatory response (Figures 6A–C). Moreover, GGQLD significantly reduced the expression of NLRP3 and GSDMD compared with UA (Figure 6D). UA remarkably promoted apoptosis at the early and late stages (Figure 6E), while GGQLD had the opposite effect (Figure 6E). Besides, we also investigated the effect of GGQLD on the UA-induced expression of Bcl-2 and caspase gene family proteins in PTECs. Consequently, as further confirmed by WB assays, GGQLD significantly reduced the UA-induced NLRP3 expression (Figures 6F,G).

UA is mainly regulated by urate transporters. To explore the effect of GGQLD on HUA, this study investigated its effects on URAT1 and GLUT9. As a result, GGQLD significantly downregulated the expression of URAT1 and GLUT9 (Figure 7) in comparison with the HUA group.