D-glucose was obtained from AMRESCO (Ohio, United States). GPS used in animal experiments (purity>98%) was purchased from Nanjing Dilger Medical Technology Co., Ltd. (Nanjing, China). GPS used in in vitro experiment (purity>99%) was purchased from Nanjing ZheWeiKang Biological Technology Co., Ltd. (Nanjing, China). Dulbecco’s modified Eagle’s medium (DMEM) and fetal bovine serum (FBS) were purchased from Life Technologies (New York, United States). Furthermore, 10% goat serum was purchased from Boster Biological Technology Co., Ltd. (Wuhan, China). BCA Protein Assay Kit and ECL detection kit were purchased from Pierce (Illinois, United States). The secondary antibody used in immunofluorescence was purchased from Thermo (Massachusetts, United States). DAPI was purchased from the Beyotime Institute of Biotechnology (Shanghai, China). Secondary antibodies labeled with HRP were purchased from Promega (Wisconsin, United States). Polyvinylidene difluoride membrane was purchased from Millipore (Massachusetts, United States).

Antibodies against α-tubulin, FN, IκBα, phosphorylated IκBα (p-IκBα), and lamin B1 were purchased from Proteintech (Wuhan, China). Antibodies against p65 and CK2α were obtained from Santa Cruz (California, United States). Antibodies against α-SMA and E-cad were purchased from Cell Signaling Technology (Boston, United States). Antibody against vimentin was purchased from Boster Biological Technology Co., Ltd. (Wuhan, China). Antibody against ATR1 was purchased from Abcam (Massachusetts, United States). Antibody against phosphorylated CK2α (p-CK2α) was purchased from Affinity Biosciences (Ohio, United States). GFP antibody was purchased from the Beyotime Institute of Biotechnology (Shanghai, China).

Male db/db mice (6 weeks of age) were randomly divided into the diabetic model group, low-dose GPS treatment group (50 mg/kg), medium-dose GPS treatment group (100 mg/kg), high-dose GPS treatment group (200 mg/kg), and valsartan (Val) treatment group (10 mg/kg) based on the fasting blood glucose (FBG) value. FBG was measured by using a glucometer (Johnson, New Jersey, United States) weekly. The number of animals in each group was 8. GPS and Val were intragastrically administered for 10 weeks. The mice in both the control group and diabetic model group were given the same volume of normal saline by intragastric administration. At the end of the experiment, 24 h urine was collected. After being sacrificed, the serum samples were separated and stored at −20°C for biochemical analysis. Kidney samples were fixed in 4% paraformaldehyde or frozen at −80°C. The immunohistochemical staining experiment was conducted as the reported method in Sawai et al. (2006).

Healthy male C57/BL6 mice (6 weeks of age) were used as the control group. Male C57/BL6 mice (Animal Quality Certificate Number: 44008500019074) were obtained from the experimental animal center of Sun Yat-sen University (Guangzhou, China). Male db/db mice (Animal Quality Certificate Number: 201919681) were purchased from GemPharmatech Co., Ltd. (Nanjing, China). Animals were housed under specific pathogen-free conditions. Experiments were implemented in accord with China Animal Welfare Legislation and were approved by the Ethics Committee on the Care and Use of Laboratory Animals of Sun Yat-sen University (Guangzhou, China).

The kits used for biochemical analysis were provided by the Jiancheng Bioengineering Institute (Nanjing, China). The catalogs of serum creatinine (Src), blood urea nitrogen (BUN), and urine protein (Up) are C011-2-1, C013-2-1, and C035-2-1, respectively. The catalogs of glycated serum protein (GSP), hemoglobin A1c (HbA1c), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C) are A037-2, A056-1-1, A110-1-1, and A113-1-1, respectively. The aforementioned parameters were measured according to instructions.

NRK-52E cells were incubated in DMEM with 10% FBS at 37°C under an atmosphere of 5% CO₂. When NRK-52E cells reached about 50% confluence, cells were cultured in a serum-free medium for 12–16 h and then divided into different groups. Subsequently, cells were treated with GPS for a specified time. In this study, the control group was cultured with 5.6 mM glucose, while the model group and GPS treatment group were cultured with 30 mM high glucose (HG).

MTT assays were performed as the reported method in Xiao et al. (2020). The NRK-52E cells were treated with GPS for 36 h. A microplate reader (Omega Bio-Tek, Georgia, United States) was used for the determination of optical density at 570 nm.

To investigate the migration ability of NRK-52E cells, cell scratch tests were performed as the reported method in Pang et al. (2020).

The Western blot assay was conducted as the reported method in Xiao et al. (2020). Protein samples were extracted with RIPA buffer (Beyotime Institute of Biotechnology, Shanghai, China) containing protease and phosphatase inhibitors (Selleck Chemicals, Texas, United States). Nuclear and cytoplasmic proteins were extracted, according to the instruction of the protein extraction kit (Active Motif, Carlsbad, United States). The Tanon 5200 Chemiluminescent Imaging System (Shanghai Tanon Technology Co., Ltd., Shanghai, China) was applied to observe the target proteins. ImageJ software was used for quantitative analysis.

The immunofluorescence assay was conducted as the reported method in Xiao et al. (2020). A Zeiss LSM 510 laser confocal fluorescence microscope (Oberkochen, Germany) was used to collect images.

NRK-52E cells were lysed with immunoprecipitation (IP) lysate buffer (Beyotime Institute of Biotechnology, Shanghai, China) containing the protease inhibitor (Selleck Chemicals, Texas, United States). After that, the supernatants were collected. Next, the aforementioned cell lysates were incubated with 20 μL protein agarose A/G beads (Thermo, Massachusetts, United States), and the supernatants were obtained by transient centrifugation. Later, an antibody (against CK2α or IgG) was added with shaking for 24 h at 4°C. Subsequently, the aforementioned solutions were incubated with 20 μL protein agarose A/G beads for 4 h at 4°C. Beads were collected by transient centrifugation and washed with IP buffer 3 times. Finally, about 20 μL loading buffer (Fdbio, Hangzhou, China) was mixed with the beads and boiled for 5 min. The method of the Western blot was adopted to detect protein levels.

The DNA-binding activity of transcription factor NF-κB was measured by electrophoretic mobility shift assay (EMSA). After that, the procedure was conducted according to the instruction of the Light Shift Chemiluminescent EMSA Kit (Pierce, Illinois, United States). The sequence of the biotin-labeled oligonucleotide probes for NF-κB (Beyotime Institute of Biotechnology, Shanghai, China) was as follows:

Sense: 5′-ACT​GAG​GGT​GAC​TCA​GCA​AAA​TC-3′

The images were captured by using a Tanon 5200 Chemiluminescent Imaging System (Shanghai Tanon Technology Co., Ltd., Shanghai, China).

The pcDNA3.1-GFP-AT1R plasmid was obtained from GenePharma (Shanghai, China).

The plasmid was transfected to NRK-52E cells, according to the instruction of LTX & PLUS (Life Technologies, New York, United States). Briefly, NRK-52E cells were cultured in plates for 24 h prior to transfection, and then, 2 μg plasmids were transfected into cells by using LTX & PLUS. After 36 h of incubation, the medium was replaced with a serum-free medium for another 12 h. Then, the cells were harvested after further treatment.

NRK-52E cells were cultured on 96-well plates with various interventions, according to specific experimental requirements. After that, the cells were co-transfected with 0.2 µg pNF-κB-Luc (Beyotime Institute of Biotechnology, Shanghai, China) and 0.04 µg pRL-TK (Promega, Wisconsin, United States). Next, collecting the cells and a dual-luciferase reporter assay system kit (Promega, Wisconsin, United States) was adopted to detect the luciferase activity. Renilla luciferase activity was used as an internal reference in this assay.

All experimental data were analyzed by the software GraphPad Prism 5.0. Values were expressed as mean ± standard deviation (SD). The data between the two groups were analyzed by the unpaired Student’s t-test. As for multiple comparisons, one-way ANOVA with post hoc multiple comparisons was adopted. A value of p < 0.05 was regarded as statistically significant. Independent experiments were performed at least three times with similar results.

Hyperglycemia and dyslipidemia are closely related to the progress of diabetic renal fibrosis (Zoja et al., 2020). Therefore, the biochemical parameters related to glucose and lipid metabolism of db/db mice were determined after 10 weeks of GPS administration (Figure 1A). Compared to the control group, the FBG level was significantly increased in the diabetes group, which was effectively improved by GPS treatment (Figure 1B). Furthermore, GSP and HbA1c, other biomarkers for the assessment of long-term glycemic control, were also decreased after 10 weeks of GPS treatment (Figures 1C,D). In addition, GPS treatment notably decreased LDL-C and TG levels (Figures 1E,F). In addition to that, GPS reduced the body weight of db/db mice (Figure 1G). These results revealed GPS ameliorated lipid and glucose metabolism disorders in diabetic mice.

Creatinine is the metabolite of muscle, and its amount is proportional to muscle mass. Physically, creatinine is cleared away from plasma by glomerular filtration and then excreted into the urine without tubular reabsorption. Therefore, the amount of Scr is constant under normal conditions, and an increased Scr level usually indicates the impaired renal function (Li et al., 2020). In addition to that, urea nitrogen, the main end product of protein metabolism, is another important parameter to evaluate the renal function. The elevated plasma urea nitrogen level is also a common clinical manifestation of DN (Wang et al., 2020). In addition, microalbuminuria is the early clinical sign of renal disease. Screening for microalbuminuria is crucially important for the diagnosis and management of DN (Kaburagi et al., 2019).

We assessed the effects of GPS on renal function in this assay. The results showed that the parameters related to renal function including Scr, BUN, and 24 Up were significantly increased in the model group. After the treatment of GPS, the levels of Scr, BUN, and 24 Up were remarkably reduced (Figures 1H–J). Thus, we concluded that GPS improved the renal function of db/db mice.

Additionally, the methods of hematoxylin-eosin (HE), periodic acid–Schiff (PAS), and Masson staining were applied to investigate the effects of GPS on TIF. Lumen expansion was observed in the renal tubular of db/db mice by HE staining. PAS staining results showed that glycogen deposition in the renal cortex of db/db mice was increased when compared with the control group. Masson staining was used to detect the extent of collagen accumulation. The positively stained area in the renal tubular of db/db mice was significantly increased, indicating an increase in collagen accumulation in renal tissue. However, after GPS intervention, the lumen expansion gradually converted to normal, and the degrees of inflammation and fibrosis in the renal tubule were ameliorated (Figure 2A). Considering the vital role of renal tubular EMT in TIF, we further detected the effects of GPS on EMT marker proteins in kidney tissues of diabetic mice. Western blot results showed that the EMT biomarkers including FN, α-SMA, and vimentin were significantly decreased after GPS treatment, while E-cadherin protein expression was increased (Figures 2B–D, Supplementary Figures S1A,B). Taken together, the aforementioned results indicated that GPS ameliorated TIF in db/db mice.

As mentioned earlier, highly activated AT1R may promote diabetic TIF by activating the CK2/NF-κB signaling pathway. Therefore, we speculated whether GPS suppressed the AT1R/CK2/NF-κB pathway.

First, the network pharmacology model was constructed. The result showed that interfering with the renin angiotensin system (RAS) was related to the treatment of DN with GPS (Supplementary Table S1). Next, we detected the influence of GPS on the AT1R/CK2/NF-κB pathway. The results showed that GPS treatment downregulated expressions of AT1R and CK2α in the kidney of db/db mice. Additionally, IκBα, the main endogenous inhibitory protein of NF-κB, was significantly upregulated by GPS treatment in kidney tissues of db/db mice, which indicates GPS could inhibit the activation of the NF-κB signaling pathway in kidneys of diabetic mice (Figures 3A,B). In summary, these results indicated that GPS inhibited the AT1R/CK2/NF-κB pathway in the renal tissue of db/db mice.

The results in vivo showed that GPS improved TIF and inhibited the AT1R/CK2/NF-κB pathway in the kidney of db/db mice. Chronic hyperglycemia is the key factor leading to TIF. In addition, growing evidence demonstrated that hyperglycemia was a vital contributor to the over-activation of RAS including AT1R in DN (Amorim et al., 2019). The results in vivo also showed that GPS markedly improved glycolipid metabolism disorder. Such a remarkable effect of GPS on glycolipid metabolism disorder in db/db mice may mainly contribute to the inhibition of TIF and the AT1R/CK2/NF-κB pathway. Therefore, the effects of GPS on TIF and the AT1R/CK2/NF-κB pathway in db/db mice may mainly be related to glucose lowering. However, whether GPS affects renal function directly was unclear. To further confirm the effects of GPS on TIF, renal tubular epithelial cell NRK-52E was applied to in vitro experiments.

Western blot results showed that E-cad expression was decreased, and the protein expressions of α-SMA, vimentin, and FN were increased in NRK-52E cells stimulated with HG for 36 h (Figures 4A,B). Additionally, the migration ability of NRK-52E cells was significantly increased after HG treatment for 36 h (Figure 4C, Supplementary Figure S1C). Therefore, the treatment of HG for 36 h was adopted for the forthcoming studies.

MTT results showed that GPS with concentrations less than 800 μM had no cytotoxicity to NRK-52E cells (Supplementary Figure S1D). Moreover, Western blot results showed that GPS suppressed the expressions of vimentin and α-SMA starting at 12.5 μM in HG-treated NRK-52E cells (Figure 4D). Therefore, 12.5, 25, and 50 μM GPS were adopted in subsequent experiments.

GPS reduced expressions of FN, vimentin, and α-SMA and promoted the E-cad expression in HG-induced NRK-52E cells (Figures 5A–C). Moreover, the migration ability was decreased after GPS treatment (Figure 5D, Supplementary Figure S1E). The aforementioned results indicated that GPS inhibited the HG-induced EMT process in NRK-52E cells.

After HG stimulation for 30 min, the NF-κB p65 expression in the nucleus was significantly increased (Supplementary Figure S1F). Thus, the effect of GPS on the CK2/NF-κB pathway was studied after HG stimulation for 30 min.

The result indicated that GPS effectively inhibited the nuclear expression of NF-κB p65, which was consistent with the effect of PDTC (NF-κB inhibitor) (Figures 6A,B). EMSA showed that HG stimulation increased the binding of NF-κB to target probes, but GPS intervention inhibited the aforementioned phenomena (Figure 6C). Furthermore, the dual luciferase reporter gene assay showed that the NF-κB p65 transcriptional activity was decreased after GPS treatment (Figure 6D).

Physiologically, IκBα, as the anchor protein of NF-κB p65, binds with NF-κB p65 in the cytoplasm. In response to HG stimulation, IκBα is phosphorylated by upstream kinase and then degraded after ubiquitination. Subsequently, NF-κB p65 translocates into the nucleus (Wu et al., 2020). Therefore, the levels of IκBα and phosphorylated IκB were measured in this assay. The decreased phosphorylation level and increased total protein expression of IκBα were observed after the intervention of GPS (Figure 6E). The phosphorylation of CK2α at the site of threonine 360 was an indicator of CK2α kinase activity (Ota et al., 2020). We further tested the effect of GPS on the phosphorylation of CK2α at the site of threonine 360. Western blot data showed that GPS significantly downregulated the p-CK2α expression as compared with HG-treated cells (Figure 6F).

In addition to that, the effect of GPS on AT1R was explored in this assay. Western blot data showed that the AT1R expression was upregulated under the HG condition (Figure 7A), while GPS significantly downregulated the AT1R expression (Figure 7B). In summary, GPS inhibited HG-induced activation of the AT1R/CK2/NF-κB pathway in NRK-52E cells.

In order to further clarify the significance of AT1R in this process, NRK-52E cells were transfected with the exogenous AT1R plasmids, and corresponding validation experiments were conducted.

Western blot results showed that the AT1R expression was significantly increased accompanied by a striking increase in the GFP expression after the NRKA-52E cells were transfected with over-expression plasmid targeting AT1R (Supplementary Figure S2A). Compared with the GPS intervention group, the protein expressions of FN, vimentin, and α-SMA were increased, and E-cad expression was reduced when GPS and the over-expression plasmid of AT1R were given at the same time (Figures 7C,D). Moreover, the scratch test also showed that the inhibitory effect of GPS on the migration ability of NRK-52E cells was reversed by the AT1R over-expression (Figure 7E, Supplementary Figure S2B). Therefore, these results indicated that GPS inhibited the HG-induced EMT process through AT1R.

Western blot results showed that the effects of GPS on NF-κB p65 nuclear translocation, NF-κB binding activation, and NF-κB transcriptional activity were abolished after the overexpression of AT1R (Figures 8A–C, Supplementary Figure S2C). In addition, the downregulatory effects of GPS on IκBα and p-CK2α levels were reversed by the overexpression of AT1R (Figures 8D,E). Furthermore, the IP assay showed that the interaction of CK2α with IκBα was increased in HG-induced NRK-52E cells. Importantly, GPS effectively inhibited the interaction of CK2α with IκBα in HG-induced cells. Notably, the downregulatory effect of GPS on the interaction was reversed after the overexpression of AT1R (Figure 8F, Supplementary Figure S2D).

Taken together, these observations further confirmed that inhibiting the CK2/NF-κB pathway via AT1R is one of the mechanisms for GPS to improve EMT in TIF.