SQ (Batch No. 210102; Cantonese medicine ratification No. Z20071155), compound cyclophosphamide tablets (Batch No. 180604, Country medicine ratification No. H22026738), and prednisone acetate tablets (Batch No. 1710150; Country medicine ratification No. H12020123) were provided by Guangdong Provincial Hospital of Chinese Medicine. Anti-Fx1A antiserum (#PTX-002S) was obtained from PROBETEX (San Antonio, United States). ADR (#HY-15142) and trigonelline (Trig, #HY-N0414) were obtained from MCE (New Jersey, United States). Primary antibodies against Caspase-3 (#14220), Bax (#2772), Nrf2 (#12721), GAPDH (#5174), and Histone H3 (#4499) and secondary antibodies against rabbit IgG (7074) and mouse IgG (7076) were purchased from Cell Signaling Technology Inc. (Beverly, MA, United States). Goat pAb to rat IgG (Alexa Fluor^® 488) (ab150165), primary antibodies against C3 (ab11887), Bcl-2 (ab196495), and HO-1 (ab13248) and secondary antibody against Goat pAb to Mouse IgG (Alexa Fluor^® 488) (ab150113) were purchased from ABCAM (Cambridge, MS, United States). Primary antibodies against C5b-9 (sc-66190) and synaptopodin (sc-515842) were purchased from Santa Cruz Biotechnology (Dallas, Texas, United States). One-step TUNEL apoptosis assay kit (C1089) was obtained Beyotime Biotechnology Institution (Jiangsu, Nanjing). 4′,6-Diamidino-2-phenylindole (DAPI) (BS097) was purchased from Biosharp Life Sciences (Hefei, China). Other reagents were of analytical grade from commercial suppliers.

SQ is authorized for clinical treatment by the Drug Administration of Guangdong Province and chemically characterized and manufactured by Guangdong Provincial Hospital of Chinese Medicine (Guangzhou, China). The extraction method of Chinese Pharmacopoeia was applied to get SQ from Radix Astragali and Radix Notoginseng with a concentration of 0.333 and 0.056 g/ml, respectively, and the botanical name of Radix Astragali and Radix Notoginseng could be confirmed on https://mpns.science.kew.org/mpns-portal/. The botanical samples of SQ have been stored in the Pharmaceutical Preparation Department of Guangdong Provincial Hospital of Chinese Medicine, and are available whenever needed. Sanqi oral solution lyophilized power (SQL) was prepared according to following steps: SQ was frozen in culture dish (diameter: 100 mm) at −80°C for 48 h. Then SQ was dried in vacuum condition at −60 to 70°C to obtain SQL, and SQL was stored at −80°C before use. The preparation and chemical components analysis of SQ were performed referring to our published study (Tian et al., 2020). The content and chemical profiles of SQ are shown in Supplementary Table 1 and Supplementary Figure 1.

Adult male Sprague-Dawley (SD) rats weighing between 180 and 220 g with a special pathogen-free (SPF) level were obtained by the Medical Experimental Animal Center of Guangdong Province (production certification No. SCXK 2019-0035, Guangdong). All the experimental rats were housed in Experimental Animal Center of Guangdong Provincial Hospital of Chinese Medicine (use certification No. SYXK 2018-0094, Guangdong) with 20 ± 2°C controlled temperature and 50 ± 10% humidity, and a 12-h light/dark cycle (lights on: 07:00–19:00), ventilation, more than 10 air exchanges per hour (all-fresh-air system), and allowed free access to standard laboratory water and diet ad libitum. All animal experiments were carried out as per the Regulations of Experimental Animal Administration issued by the State Committee of Science and Technology of the People’s Republic of China, with the approval of Animal Care and Use Committee in Guangdong Provincial Hospital of Chinese Medicine (Guangzhou, China).

The PHN rat model was induced as previously described with a few modification (Di Tu et al., 2020). After acclimatization of 3 days, healthy adult SD rats were randomly divided into four groups (n = 6 per group): 1) Control group (Control); 2) Model group (Model); 3) SQ group (SQ); and 4) CP (Compound cyclophosphamide + Prednisone acetate) group (CP). Combination of compound cyclophosphamide tablets and prednisone acetate tablets was chosen as the positive control drug, and the intragastric dosage of SQ, complex cyclophosphamide together with prednisone acetate were calculated according to the clinical dosage using human–rat conversion coefficient. The SD rats of the PHN group, SQ group, and CP group were given a tail vein injection with a single dose of anti-Fx1A antiserum (0.5 ml/100 g), while the SD rats of the SQ group and the CP group were administrated with either SQ (12.6 mg/kg) or compound cyclophosphamide tablets (12.6 mg/kg) and prednisone acetate tablets (6.3 mg/kg) daily by oral gavage. When modeling and administration, the Control group and the PHN group were given the same amount of saline or distilled water. (Figure 1A) On day 5, 10, 15, and 20, the 24-h urine samples were recorded and collected for the urine protein analysis. After 21 days, all SD rats were sacrificed and required samples were collected.

Rats in all groups were put into separate metabolic cages to collect urine for 24 h. During the collection of urine, rats were allowed free to drink water but were limited to diet. The volume of 24-h urine was recorded after collection, and the collected urine samples were centrifuged at 3,000 rpm for 15 min at room temperature, the supernatants were stored at −80°C until analysis. After the experiment, plasma samples in rats were collected by abdominal aorta and put into coagulation-promoting vacuum tubes. The collected plasma samples were allowed to stand for 1 h at 4°C and centrifuged at 12,000 rpm for 15 min at 4°C to collect serum, and then the serum was stored at −80°C for a biochemical analysis. The urine protein and serum albumin (ALB) analyzed by the clinical laboratory of Guangdong Provincial Hospital of Chinese Medicine (Guangzhou, China).

Examination of transmission electron microscopy (TEM) was conducted as previously described with a few modifications (Wang et al., 2020). Briefly, the sections of kidney tissues (1 mm³) were fixed in 5% glutaraldehyde for 2 h, followed by washing in 0.1 M phosphate buffer. After immersing in 1% osmic acid for 1.5–2 h, the kidney tissues were dehydrated through graded alcohols and immersed in embedding medium overnight, and then, the immersed sections were embedded in Epon 812 and dried in an oven. The dried kidney sections were cut into 50–70 nm slices and stained with uranyl acetate and lead citrate. A transmission electron microscope (JEM1400 PLUS, Japan) was used to examine the slices at 100 kV, and a CCD camera (EMSIS VELETA G3, Germany) was used to take micrograph.

For evaluating the foot process width per GBM length, images covering a glomerular cross section were captured by TEM. The peripheral length of GBM was measured, and the quantity of foot process overlying this part of GBM was counted using Adobe Photoshop (San Jose, California, United States). The arithmetic mean of the foot process width was calculated by the following formula (Chen et al., 2010) (1): T h e   f o o t   p r o c e s s   w i d t h   =   ( π / 4 )   × ∑ G B M   l e n g t h ∑ q u a n t i t y   o f   f o o t   p r o c e s s , (1) where ∑GBM length and ∑quantity of foot process represented the total GBM length measured in one glomerulus and the total quantity of foot process counted, respectively. The correction coefficient π/4 is set to correct the random orientation in which the foot processes are sectioned.

Basement membrane thickness in each ultrathin slice was measured by RADIUS software (Boston, MA, United States). Briefly, an open capillary loop was selected, and the value perpendicular to the GBM overlying this capillary loop length was metered.

Tissues and immunofluorescence staining were performed according to the method previously described with a few adjustments (Li et al., 2021). Briefly, renal tissues were embedded in Tissue-Tek OCT compound, snap frozen in liquid nitrogen, and stored at −80°C till examination. To inspect the deposition of IgG, C3, C5b-9, and synaptopodin in glomeruli, 5 µm cryosections were immersed in acetone and washed in cold PBS. After blocking with 5% bovine serum albumin in PBS for 30 min, the cryosections were stained with IgG, C3, C5b-9, and synaptopodin overnight at 4°C, followed by goat anti-mouse IgG H&L (Alexa Fluor^® 488). The cryosections were observed under confocal fluorescence microscope (Zeiss LSM710, Germany). Fluorescence intensity was measured in five randomly selected fields of six cryosections by ImageJ (Bio-Rad Laboratories, Hercules, CA, United States).

The genomic DNA in the nucleus cracks after apoptosis. With terminal deoxynucleotidyl transferase (TdT) as catalyst, exposed 3′- OH binds to fluorescent probe Cy3 labeled dUTP, which can be tested by fluorescence confocal microscopy. Cell apoptosis was detected by the TUNEL assay kit, referring to the manufacture’s protocol. Briefly, renal slices (3 μm) of different groups were dewaxed, followed by incubation with protease K (without DNase, 20 μg/ml) for 30 min at 37°C and washing with cold PBS, and then the kidney sections were incubated with 50 μL TUNEL reaction mixture for 1 h at 37°C in dark. Ultimately, stained slices were washed with PBS and examined by confocal fluorescence microscope (Zeiss LSM710, Germany). Five fields of six sections from each group were randomly selected for analysis.

The conditionally immortalized temperature sensitive mouse podocyte cell line used in this study was established by Professor Peter Mundel (Medical College of Harvard University, Boston, MA, United States). Briefly, podocytes were cultured in RPMI 1640 medium with 10% fetal bovine serum at 33°C in the presence of 10 U/ml recombinant mouse interferon-γ (Sigma, St. Louis, MO, United States). For inducing differentiation, podocytes were thermoshifted to 37°C and cultured in interferon-free medium for 10–14 days. SQL and Trig were dissolved in PBS and storage solutions were stored at −20°C. Podocytes were cultured overnight before experiments and treated with 400 ng/ml ADR with or without 25 μM Nrf2 inhibitor Trig or 600 μg/ml SQL intervention for 24 h.

Western blot was conducted as previously described (Wang XW. et al., 2019). Total proteins from renal tissues or podocyte cells were extracted using the T-PER™ Tissue Protein Extraction Reagent (Thermo Fisher Scientific, Rockford, IL, United States) with proteinase and phosphatase inhibitor tablets (Roche, Mannheim, Germany) or the RIPA Lysis Buffer (Beyotime, Shanghai, China). The nuclear proteins in kidneys and podocyte cells were prepared with the Nuclear Protein Extraction Kit (Beyotime, Shanghai, China), according to the protocols. Protein concentration was measured by the Pierce^® BCA Protein Assay Kit (Thermo Fisher Scientific, Rockford, IL, United States). Lysates were boiled with a 5 × loading buffer (CWBIO, Beijing, China), separated by 10–12.5% SDS–polyacrylamide gel electrophoresis, and transferred onto a PVDF membrane (Millipore, United States). After blocking with 5% non-fat milk in Tris-HCL buffer containing in Tween 20 (TBST), the membranes were incubated at 4°C with corresponding primary antibodies (Nrf2, HO-1, Bcl-2, Bax, Cleaved Caspase-3, Caspase-3, and GAPDH) overnight. Then the membranes were washed with TBST and incubated with horseradish peroxidase (HRP)–labeled anti-rabbit or anti-mouse IgG at room temperature. Blots were developed detected by Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, United States), referring to the manufacturer’s protocol, visualized by enhanced chemiluminescence detection system (Bio-Rad, Laboratories, Hercules, CA, United States), and quantified using Image Lab software 5.2.1 (Bio-Rad, Laboratories, Hercules, CA, United States).

All quantitative data were expressed as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) followed by Dunnett’s T3 test was used to compare group mean using SPSS 25.0 (SPSS, Inc., Chicago, IL, United States). p < 0.05 was considered statistically significant, and p < 0.01 was considered statistically highly significant.

Proteinuria, a general characteristic of IMN, was observed in rats of each group. Present data revealed that PHN rats developed severe proteinuria, and 24-h proteinuria was significantly elevated from Day 5. However, the secretion of proteinuria in PHN rats was found to be decreased remarkedly at all points of observation, including Day 5, Day 10, Day 15, and Day 20 (Figure 1B). Similarly, the urinary protein/creatinine ratio was increased in PHN rats from Day 5, and intervention of SQ and CP could reduce the ratio significantly from Day 10 (Figure 1C). Besides, the level of serum albumin was found to be reduced significantly in PHN rats (p < 0.01), whereas the serum albumin level was increased markedly with the treatment of SQ and CP (p < 0.05 and p < 0.01) (Figure 1F).

The changes of glomerular pathomorphology are shown in Figures 1D,E. Transmission electron microscopic examination showed subepithelial glomerular immune deposits and thickening of the GBM (p < 0.01) in glomeruli of rats with PHN, while treatment with SQ and CP, less thickened GBM were visible, compared to PHN rats without the treatment of SQ or CP (p < 0.01). The data above indicated that SQ showed therapeutic effect in anti-Fx1A antiserum induced PHN in rats.

According to the pathogenesis of the rat PHN model, the antibodies in rabbit anti-Fx1A antiserum occur in the kidney and recognize the rat autologous antigen megalin, which exists on podocyte (Zanetti and Druet, 1980; Raychowdhury et al., 1996), and then formed glomerular in situ immune deposits on the subepithelial layer of the glomerular capillary loops. Deposition of immune complexes stimulated the complement system and induced the assembling of C5b-9 membrane attack complex (MAC) in glomeruli. As in our study, PHN rats displayed pronounced granular autologous IgG deposition in glomeruli, dispersing along the capillary wall (p < 0.01) (Figures 2A,B), and compared to PHN rats, IgG deposition was diminished remarkably by SQ and CP intervention (p < 0.01). It is known that the complement system plays a major role in the course of IMN. In human IMN, the renal deposition of C3 and C5b-9 depositions is typical, and they are also found in PHN rats (Raychowdhury et al., 1996). Immunofluorescent staining of C3 and C5b-9 were also observed in glomeruli (Figures 2A,C,D). Compared with the control group, C3 and C5b-9 were particularly expressed along the capillary walls in glomerular of PHN rats (p < 0.01). In contrast, SQ and CP treatment significantly decreased the deposition of C3 and C5b-9 (p < 0.01), suggesting that SQ could alleviate the immune injury in PHN rats.

Podocyte, also called the visceral glomerular epithelial cell, is the key part of glomerular filtration barrier. Podocyte injury can induce proteinuria directly. Diffuse fusion of podocyte foot process was observed in PHN rats under TEM (Figure 3A). A semi-quantificative analysis of foot process width showed that glomerular podocyte food processes were ameliorated partially in PHN rats with the treatment of SQ or CP (Figure 3B). In addition, synaptopodin is an actin-associated protein that plays important role in the morphology and movement of podocytes based on actin, and is considered as a marker of differentiation and maturation. As shown in Figures 3A,C, the reduced expression of synaptopodin was observed in PHN rats (p < 0.01). However, intervention with SQ or CP significantly restored the downregulated glomerular synaptopodin (p < 0.01). The results demonstrated that SQ protects against podocyte injury in PHN rats.

Results of observation after TUNEL staining (Figure 4) showed that glomerular podocyte apoptosis exhibited in all groups. There was pronounced apoptosis of glomerular podocytes of PHN rats (p < 0.01), while, the accumulation of apoptotic podocytes of glomerulus was remarkedly inhibited with treatment of SQ and CP (p < 0.01). We further investigated apoptosis-related protein expressions of Bax, Bcl-2, Cleaved Caspase-3, Caspase-3 in kidney tissues of rats from all groups (Figure 5). Results of the western blot analysis showed that the apoptosis pathway was activated in PHN rats. However, SQ and CP administration significantly inhibited the activated apoptosis pathway by suppressing the elevation of protein levels of Bax together with Cleaved Caspase-3 and the reduction of Bcl-2 expression (p < 0.05 or p < 0.01).

ADR is a known injurious stimuli causing proteinuria, and the podocyte apoptosis induced by ADR is one of the main mechanisms of podocyte injury (Ni et al., 2018). In this study, our data indicated that 400 ng/ml ADR caused significant rise in Cleaved Caspase-3 protein expression (Figure 7), while SQL was given to ADR-treated podocyte and markedly reduced Cleaved Caspase-3 protein expression (p < 0.05).

The above results suggested that SQ could ameliorate podocyte injury in glomerulus of PHN rats by reducing apoptosis, and CP can exert better anti-apoptotic effect than SQ.

Numerous studies have reported that the Nrf2/HO-1 signaling pathway is involved in modulation of apoptosis, and HO-1 can directly inhibit podocyte apoptosis. So, we wished to detect the expressions of Nrf2 and HO-1 proteins in kidneys of PHN rats and ADR-treated podocyte cells. As indicated in Figure 6, there was a prominent suppression of HO-1 in PHN rat kidneys, which was elevated with SQ intervention (p < 0.01). In line with HO-1 expression, decreased total protein level of Nrf2 was found in PHN rats, and SQ intervention greatly upregulated the levels of total Nrf2 proteins (p < 0.05). Since the key to activation of the Nrf2/HO-1 signaling pathway is the nuclear translocation of Nrf2, the impact of SQ on nuclear Nrf2 was also detected. Nuclear Nrf2 was found to be suppressed in kidneys of PHN rats, and recovered with SQ and CP intervention (p < 0.05). Nrf2 protein levels in podocyte cells were further checked to confirm the effect of SQ on Nrf2 (Figure 7). The total Nrf2 expression recovered with SQL treatment in presence of ADR (p < 0.05), and ADR restricted the Nrf2 nuclear translocation, while SQL did the opposite (p < 0.01). To determine whether SQ stimulated Nrf2 activity exert leading role in mitigating podocyte apoptosis, an effective inhibitor of Nrf2, Trig, was added to podocyte. As shown in Figure 8, abrogated Nrf2 expression induced by Trig in ADR-injured podocyte led to higher Cleaved Caspase-3 protein expression (p < 0.05). However, SQL cotreatment substantially suppressed Nrf2 reduction and Cleaved Caspase-3 rise (p < 0.05). In a nutshell, SQ exerted therapeutic effect on PHN rats by mitigating podocyte apoptosis and proteinuria via the Nrf2/HO-1 signaling pathway.