Astragalus mongholicus Bunge and Panax notoginseng formula is a traditional Chinese medicine formula used clinically to treat kidney diseases such as diabetic nephropathy, chronic kidney disease, and acute kidney disease (Wen et al., 2020). It consists of Astragalus mongholicus Bunge (3g), Panax notoginseng (1g), Angelica sinensis (3g), Achyranthes bidentata Blume (3g) and Ecklonia kurome Okamura (3g). The main component profile of A&P was analyzed by high performance liquid chromatography in our previous study (Hui et al., 2020). The formula was purchased from the affiliated traditional medicine hospital of Southwest Medical University, Luzhou City, Sichuan Province. The dosage for gavage is determined by referring to the conversion method of human and laboratory animals in “Pharmacological Experimental Methodology” edited by Professor Xu Shuyun. Mice daily dose = (human clinical daily dose × human defined body weight 70 kg × human and mouse body surface area conversion coefficient 0.0026) ÷ mouse defined body weight 20g. Therefore, the daily gavage dose of mice calculated by the formula is 1972 mg/kg/d. Bifidobacterium Lactobacillus triple live bacteria tablets (golden Bifidobacteria), its main components are Bifidobacterium, Lactobacillus bulgaricus and Streptococcus thermophilus. Bifidobacterium was purchased from Shenzhen Xinwanze Pharmaceutical Co., Ltd. The dose for gavage is also determined by referring to the conversion method of human and laboratory animals, and the daily gavage dose of Bifidobacterium in mice calculated by the formula is 303 mg/kg/d. Dissolve the powder of A&P and BB in distilled water, followed by intragastric administration of these drugs (200 μl) to mouse once a day. Mice in sham group were treated with gavage of equal volume of saline.

A&P (1972 mg/kg/d), Bifidobacterium (303 mg/kg/d), and their combination were administered to male C57BL/6 mice (8 weeks of age, 20–24g body weight) by gavage for seven consecutive days. On the 7th day, blood was taken from the heart 2 h after the gavage, and stored at 4°C overnight. The next day, the blood was centrifuged at 3000 rpm for 10 min at 4°C to extract serum and collect all serum for subsequent cellular experiments. In the cell experiment, we used the prepared medicated serum instead of ordinary fetal bovine serum (10%).

Add 200 ng/ml of Lipopolysaccharide (LPS) (L2630, Sigma-Aldrich, United States) to the complete medium of RAW264.7 cell to put the cells in an inflammatory stress state, then add indophenol sulfate (16926, Cayman, United States) 200 μM and trans aconitic acid (122750, Sigma, United States) 200 μM, respectively, to simulate the uremic environment.

Male C57BL/6 mice (8 weeks of age and 22–25 g body weight) were classified into 5 groups: Sham, CKD, CKD + BB, CKD + A&P and CKD + BB + A&P group with 8 mice in each group. The CKD mouse model was constructed by directly ligating the upper and lower poles of left kidney after removal the right kidney 1 week later (Tan et al., 2019). Briefly, the right kidney was exposed and removed. One week later, the left kidney was exposed and the upper and lower poles were ligated with 3–0 non-absorbable suture. CKD mouse model can be obtained after six to 8 weeks of normal feeding after surgery. In this study, we determined that the model was successfully built at 7 weeks post-surgery, subsequently, the treatments were performed on these modeled mice for 4 weeks. Urine and feces of mice were collected by metabolic cage. All animal experiments were carried out according to the guidelines approved by the Animal Ethics Committee of Southwest Medical University.

Cells were seeded in 96-well plates at a density of 3000 per well, with 6 replicate wells per group. After 24 h of starvation culture, cells were treated with a drug-containing medium for 24 h, then the medium was aspirated, 100 μl of 10% CCK-8 reagent was added to each well, and incubated in an incubator for 3 h. The absorbance value of each well was measured at 450 nm by BioTeK reader, and the cell survival rate was calculated according to the results.

According to the instructions of the kit: Eastep qPCR Master Mix kit (LS2068, Promega, United States) is used to detect the mRNA expression levels of cells and kidneys; TRIzol Reagent (11596026, Invitrogen, United States) was used to extract total RNA from RAW264.7 cells and kidneys; RevertAid first-strand cDNA synthesis kit (K1622, Thermo Scientific, MA, United States) was used to reverse transcribe total RNA to cDNA; RT-PCR was performed with the Mastercycler ep Realplex2 real-time PCR system (LightCycle480II, Roche, United States). The relative expression of a given gene is calculated by the threshold cycle (CT) method. Primer sequences are listed in Table 1.

Collect CKD mouse serum and RAW264.7 cell culture supernatant. The concentration of IL-1β (Neobioscience, EMC001b), IL-6 (Neobioscience, EMC004), and TNF-α (Neobioscience, 500850) in serum and culture medium were detected according to the instructions of the ELISA kit. The absorbance value of each well was measured at 450 nm by BioTeK reader.

After placing the frozen sections of the tissues at room temperature for 20 min, wash these sections for three times with PBS. Subsequently, block the samples with 10% goat serum for 30 min at 37°C, and then incubate the samples with rat anti-F4/80 primary antibody (Santa Cruz, 1:50, sc-52664) and mouse anti-Mincle primary antibody (Santa Cruz, 1:50, sc-390806) overnight at 4°C. The next day, wash these samples with PBS three times and conjugate with AlexaFluor488 conjugated anti-rabbit secondary antibody (CST, 1: 200, 4412S) and AlexaFluor647 conjugated anti-mouse secondary antibody (CST, 1: 200, 4410S) at room temperature for 1 h, followed by washing three times with PBS, then add diluted DAPI dropwise to stain the nuclei for 10 min in the dark. Finally, pictures were obtained with a confocal microscope (Nikon, Japan).

The total protein was extracted from cell and tissues using RIPA lysis buffer containing protease inhibitors and phosphatase inhibitors. Proteins (25 μg of each sample) were separated on a 10% SDS-PAGE gel and transferred to a PVDF membrane by wet blotting. After blocking in 5% milk, the membranes were incubated with the antibodies against p-p65 (Rabbit anti-mouse, sc-136548, 1:1000, Santa Cruz), p65 (Mouse anti-mouse, #6956, 1:1000, CST), Mincle (Mouse anti-mouse, sc-390806, 1:500, Santa Cruz), iNOS (Rabbit anti-mouse, 13120S, 1:1000, CST), ZO-1 (Rabbit anti-mouse, 61–7300, 1:1000, Life technologies), Occludin (Mouse anti-mouse, 33–1500, 1:1000, Life technologies), Claudin-1 (Rabbit anti-mouse, 71–7800, 1:1000, Life technologies) and GAPDH (Mouse anti-mouse, AB2000, 1:5000, Abiways) and β-actin (Rabbit anti-mouse, 1:50000, abcam) at 4°C overnight. The next day, membranes were subsequently incubated with the secondary antibody of peroxidase-conjugated goat anti-mouse IgG (ZB-2305, 1:3000) and goat anti-rabbit IgG (ZB-2301, 1:3000) for 1 h at room temperature. The band was analyzed by the gel imaging system. Gray intensity of the band was calculated by ImageJ software.

Fresh kidney tissue was digested with collagenase IV (RAW264.7 cells were digested with 0.25% trypsin-EDTA into cell suspension), then fixed at room temperature in 4% paraformaldehyde for 1 h. After washing with PBS for 2 times, cells were incubated with iNOS primary antibody (13120S, 1: 200, CST), CD206 primary antibody (ab64693, 1: 200, abcam, directly coupled antibody) and Mincle primary antibody (sc-390806, 1:50, Santa Cruz) at 4°C overnight. The next day, the cells were incubated with fluorescent secondary antibody for 1 h. After washing with PBS for 2 times, the cells were analyzed using a flow cytometer (FACSAria, BD Biosciences, United States), and the data were processed using FlowJo software (X 10.0.7).

The urine and the tail blood of the mice were collected to detect the serum creatinine (C011-1-1, Nanjing Jiancheng Bioengineering Institute), urea nitrogen (C013-2-1, Nanjing Jiancheng Bioengineering Institute) and urine protein (C035-1-1, Nanjing Jiancheng Bioengineering Institute) following the instructions of corresponding kits.

The feces from each group were collected and stored in a special feces storage reagent every week. All of the samples were sent to Shanghai Ouyi Biomedical Technology Co., Ltd., for 16s DNA detection.

The pre-treated tissue paraffin sections were stained in hematoxylin aqueous solution for 10 min, rinsed with running water for 5 min, then dehydrated in 100% alcohol, stained with Eosin (H&E, Beyotime, China) solution for 2–3 min, and finally transparent, and mounted. The tubulointerstitial injury index was graded as described previously (Radford et al., 1997). The degree of injury to the intestinal tissues was evaluated according to previous description (Chiu et al., 1970). Pictures were captured by the optical microscope (Eclipse 80i, Nikon, Japan).

The paraffin slices were stained with the Sirius red stain at 37°C for one and a half hour, rinsed with running water for 10 min, then stained with hematoxylin aqueous solution for 10 min, followed by rinsing with running water for 10 min, finally dehydrated, transparent, and mounted. Images were captured by the optical microscope (Eclipse 80i, Nikon, Japan).

Extracting pcDNA3.1-Mincle and pcDNA3.1-vector plasmids from two Escherichia coli (purchased from GeneChem company) by using plasmid extraction kit (NucleoBond Xtra MiDi EF, 740420.50). Then, the plasmid and DNA transfection reagent (ZETA life) were mixed with complete medium and added to RAW264.7 cells for 48 h.

This study used SPSS22.0 and Prism 6 software for analysis. Data were presented as mean ± standard error. One-way analysis of variance (One-Way ANOVA) was used for multigroup comparison, followed by the Student-Newman-Keuls post-test. P < 0.05 was considered statistically significant.

During the 7 weeks of modeling, compared with the sham group, the blood creatinine, urea nitrogen and urine protein in the 5/6 kidney ligation model group were all increased, and the body weight was reduced. In contrast, the blood creatinine, urea nitrogen and urine protein in A&P + Bifidobacterium treated mice were significantly decreased, and their body weight were increased. However, mice using Bifidobacteria alone did not exhibit significant improvement on renal function (Figure 1A). H&E and Sirius red stainings were employed to detect the protective effect of A&P on kidney of CKD model. The H&E staining results showed that compared with the sham group, the kidney in CKD model mice present glomerular atrophy and deformation, renal tubular expansion as well as necrosis and shedding of renal tubular epithelial cells. After treatment with A&P + Bifidobacteria, the expansion of renal tubules in CKD mice was significantly reduced, and the glomerular morphology was restored (Figure 1B). The result of Sirius red staining demonstrated that, compared with the sham group, the positive area of sirius red staining in kidney of CKD mice was increased, indicating severe renal fibrosis, accompanied by glomerular atrophy and renal tubular expansion. However, compared with CKD group, kidney in A&P and Bifidobacterium combination treatment group showed significantly reduction of fibrosis and other pathological damages (Figure 1C). The tubulointerstitial injury index was graded as described previously (Figure 1D). Figure 1E shows the morphology of each group of kidneys. These results suggested that A&P combined with bifidobacterium treatment can significantly improve the body weight and renal function in CKD mice.

To investigate the effects of A&P + Bifidobacterium on inflammation in CKD mice, we performed Real-time PCR and ELISA to exam the expression and secretion of inflammatory cytokines. The PCR results showed that, compared with CKD group, A&P and Bifidobacterium combination significantly reduced mRNA expression of inflammatory indicators (IL-1β, IL-6, and TNF-α) in kidney of CKD mice (Figure 2A), whose inhibitory effect was better than using Bifidobacterium alone. Moreover, the ELISA results demonstrated that, compared with CKD group, A&P and Bifidobacterium combination strongly decreased secretion of inflammation indicators (IL-1β, IL-6, and TNF-α) in serum of CKD mice (Figure 2B). These results suggested that the combination of A&P and Bifidobacterium can effectively improve the renal and systemic inflammation in CKD mice.

To test whether A&P combined with Bifidobacterium improved the intestinal flora of CKD mice, we collected mouse feces and subjected to microbiological 16s DNA detection. Among them, alpha diversity is related to boxplot analysis. This method calculates the diversity index of the samples between each group and within the group on the basis of a uniform data depth, and draws a boxplot diagram of the diversity index of each group. Microbial 16s DNA test results showed that compared with CKD group, A&P combined with Bifidobacteria treatment recovered the intestinal flora to normal in CKD mice, which was better than using A&P or Bifidobacteria alone (Figure 4). These results showed that A&P combined with Bifidobacterium treatment has a positive effect on intestinal flora of CKD mice than treating with A&P or Bifidobacterium alone.

To investigate the underlying mechanism of A&P and Bifidobacterium combination in protection of kidney in CKD, we detected the Mincle/NF-κB signaling pathway in the kidney of each group. The immunofluorescence results showed that the macrophage (marker F4/80: green spots) largely infiltrated in the CKD kidney, and Mincle (red spots) was also strongly increased in the CKD kidney. Interestingly, the green and red spots were largely merged together, which means that Mincle is mainly expressed on macrophage. Compared with CKD group, the macrophage infiltration in the kidney of A&P + Bifidobacterium treated mice was significantly reduced, and the expression of Mincle was also down-regulated (Figure 5A). Moreover, Real-time PCR results showed that A&P and Bifidobacterium combination effectively inhibited the mRNA expression of Mincle and iNOS (Figure 5B). Western blot results also demonstrated that the protein levels of Mincle and iNOS as well as the phosphorylated NF-κB were significantly reduced in A&P and Bifidobacterium treatment group, which was consistent with the results of immunofluorescence (Figure 5C), but levels of these protein in mice treated with Bifidobacterium alone did not changed significantly. Subsequently, the flow cytometry was employed to exam the changes of macrophage polarization in each group, and the results showed that the expression of Mincle and M1 macrophage markers (iNOS) in kidney of CKD mice were increased, while A&P combined with Bifidobacterium treatment significantly decreased Mincle and iNOS. On the other hand, the expression of M2 macrophage surface marker (CD206) in kidney of CKD mice was decreased, which was recovered by A&P and Bifidobacterium combination (Figure 5D). These above results suggested that A&P and Bifidobacterium combination can strongly inhibit the activity of M1 macrophages and increase the activity of M2 macrophages by suppressing the Mincle/NF-κB pathway.

We further detected the Mincle/NF-κB signaling pathway in intestine of each group. The immunofluorescence results showed that the levels of F4/80 and Mincle were significantly increased in intestine of CKD mice, however, A&P and Bifidobacteria combination largely reduced macrophage infiltration and Mincle expression (Figure 6A). Western blot results showed that compared with the CKD group, the protein levels of Mincle, iNOS and phosphorylated NF-κB were significantly down-regulated in A&P and Bifidobacterium combined group, which is better than using A&P or Bifidobacterium alone (Figure 6B). These data suggested that A&P combined with Bifidobacterium can inhibit the Mincel/NF-κB signaling in intestine of CKD mice.

To simulate the uremic environment, we used LPS and indophenol sulfate to stimulate the RAW264.7 macrophage. Morphological observation revealed that normal RAW264.7 cells were smooth and round, with high density and distinct particles. Compared with the LPS group or LPS + IS group, A&P and Bifidobacterium containing serum significantly reduced protrusions and pseudopods on cells, and the overall cell morphology was improved (Figure 7A). Real-time PCR results showed that, compared with the LPS or LPS + IS groups, A&P and Bifidobacterium containing serum largely decreased the mRNA expression of inflammatory indicators (IL-1β, IL-6, TNF-α, MCP-1), Mincle and iNOS, however, the cells treated with Bifidobacterium containing serum alone did not inhibited the mRNA expression of these indicators (Figures 7B–G). ELISA results demonstrated that, compared with the LPS group, administration of LPS plus different concentration of IS significantly up-regulated the secretion of inflammatory indicators (IL-1β, IL-6, TNF-α) in supernatant (Figures 7H–J), and the IS concentration of 200 μM was used in the subsequent experiments. Then, we found that, compared with the LPS + IS group, A&P and Bifidobacterium containing serum strongly reduced the secretion of inflammatory indicators, but the cells treated with Bifidobacterium containing serum alone did not inhibited the secretion of these indicators (Figures 7K–M). These results suggested that A&P and bifidobacterium containing serum can improve the macrophage morphology and inflammation under indophenol sulfate induced uremic environment.

We used another urotoxin (trans aconitic acid) to simulate uremic environment. We found that, compared with the LPS group or LPS + TA group, A&P and Bifidobacterium containing serum largely decreased protrusions and pseudopods on cells, and the overall cell morphology was improved (Figure 8A). Real-time PCR results demonstrated that, compared with the LPS or LPS + TA groups, A&P and Bifidobacterium containing serum significantly reduced the mRNA expression of inflammatory indicators, Mincle and iNOS, however, the cells treated with Bifidobacterium containing serum alone did not inhibited the mRNA expression of these indicators (Figures 8B–G). ELISA results determined the 200 μM concentration of TA was used in the subsequent experiments (Figures 8H–J). We then found that, compared with the LPS + TA group, A&P and Bifidobacterium containing serum strongly reduced the secretion of inflammatory indicators, but the cells treated with Bifidobacterium containing serum alone did not inhibited the secretion of these indicators (Figures 8K–M). These results also suggested that A&P and bifidobacterium containing serum can improve the macrophage morphology and inflammation under trans aconitic acid induced uremic environment.

To further explore the relationship between Mincle and inflammatory response in macrophage, we overexpressed Mincle in A&P combined Bifidobacterium treatment group by plasmid transfection, and set up a positive control group (resveratrol (RV) 30 μM). Real-time PCR and ELISA results showed that after overexpression of Mincle in treatment group, the expression and secretion of inflammatory cytokines were significantly increased (Figures 9A–H). Western blot results demonstrated that the protein level of Mincle in the plasmid-transfected group (Mincle-OE) was strongly increased, which activated the Mincle/NF-κB signaling and M1 macrophage in A&P and Bifidobacterium containing serum treated inflammatory RAW264.7 cells (Figure 9I). These results suggested that the anti-inflammatory effect of A&P and Bifidobacterium in macrophage is dependent on the inhibition of the Mincle signaling pathway.