Astragaloside IV was purchased from Shanghai Bogoo Biotechnology Company, Limited (purity at 98%, Shanghai, China). RGZ was purchased from Sigma-Aldrich (St. Louis, MO, USA).

db/db mouse exhibits clinical and histological features of DN resembling those found in human DN, such as hyperglycemia, hyperinsulinemia, hyperlipidemia, obesity, albuminuria, glomerular enlargement, and mesangial matrix expansion (Sharma et al., 2003; Tesch and Lim, 2011). Six-week-old male diabetic db/db (BKS.cg-m +/+ Leprdb/J) and age-matched non-diabetic db/m littermates were purchased from Model Animal Research Center of Nanjing University (Nanjing, China). All the work was carried out in accordance with the approved guidelines for the use of experimental animals in Putuo Hospital, Shanghai University of Traditional Chinese Medicine. The BKS.cg-m +/+ Leprdb/J mouse becomes obese and diabetic by 8 weeks of age. They were housed in Experimental Animal Facilities at Shanghai Putuo District Central Hospital under specific-pathogen-free (SPF) conditions. Animals were fed with standard diet and free to water. At 8 weeks of age, db/m and db/db mice were randomly assigned to seven groups (n = 10/each group): (1) normal db/m mice receiving vehicle (db/m-vehicle), (2) db/m mice receiving AS-IV at 18 mg kg^-1 day^-1 (db/m-18 mg kg^-1 day^-1 AS-IV), (3) diabetic control db/db mice receiving vehicle (db/db-vehicle), (4) db/db mice receiving AS-IV at 2 mg kg^-1 day^-1 (db/db-2 mg kg^-1 day^-1 AS-IV), (5) db/db mice receiving AS-IV at 6 mg kg^-1 day^-1 (db/db-6 mg kg^-1 day^-1 AS-IV), (6) db/db mice receiving AS-IV at 18 mg kg^-1 day^-1 (db/db-18 mg kg^-1 day^-1 AS-IV), and (7) db/db mice receiving RGZ as the positive control at 2 mg kg^-1 day^-1 (db/m-2 mg kg^-1 day^-1 RGZ). AS-IV and RGZ were suspended in 0.5% carboxymethyl cellulose as a vehicle for their administrations and were given to mice by oral gavage once daily for 8 weeks. Dose conversion from human to animal was using the body surface area normalization method (Reagan-Shaw et al., 2008).

Body weight, food and water intake, urine volume, fasting blood glucose, and systolic blood pressure were measured at 4-week intervals. Mice were housed in individual metabolic cages for 24 h urine collection. Fasting blood glucose levels were monitored with the Omron HEA-230 Glucometer using one drop of tail blood. Systolic blood pressure was obtained by tail-cuff plethysmography using the ALC-NIBP blood pressure measuring system (Shanghai Alcott Biotech Co., Ltd., Shanghai, China). At the end of the study, blood samples were drawn from orbit and centrifuged for serum collection. Then the animals were killed and both kidneys were removed immediately. A portion of the renal cortex was fixed in 4% paraformaldehyde for histological examination. The rest of the renal cortex was snap-frozen in liquid nitrogen and then stored at -80°C for protein and total RNA extraction, and for ER isolation. Urinary albumin (albuminuria) was measured using an ELISA Kit (Biovision, Milpitas, CA, USA) and urinary albumin excretion was expressed as urinary albumin to creatinine ratio (ACR). BUN was measured using Urea Nitrogen Colorimetric Detection Kit (Arbor Assays, Ann Arbor, MI, USA). Urinary and serum creatinine levels were determined by Creatinine Colorimetric/Fluorometric Assay Kit (Biovision, Milpitas, CA, USA). Serum HbA1c was measured using Mouse Glycated Hemoglobin (HbA1C) ELISA Kit (Wuxi Donglin Sci &Tech Development Co., Ltd., Wuxi, China). Serum insulin was measured using Rat/Mouse Insulin 96 Well Plate Assay Kit (Millipore, Billerica, MA, USA). All the biochemical parameters were determined according to the manufacturer’s instructions. Insulin resistance was determined by calculating the HOMA-IR. HOMA-IR index was calculated according to the formula: HOMA-IR (mmol/L × μU/mL) = [fasting glucose (mmol/l)] × [fasting insulin (μU/ml)]/22.5 (He et al., 2014).

After 8 weeks of treatment, mice were fasted for 6 h (8:00 AM to 2:00 PM), and the oral glucose tolerance test (OGTT) and intraperitoneal insulin tolerance test (IPITT) experiments were performed as previously described (Chen et al., 2014b; Wang C. et al., 2014). Briefly, the animals were orally administered with glucose (2 g kg^-1 body weight) or intraperitoneally administrated with insulin (0.75 U kg^-1 body weight), and blood glucose levels were monitored over a 2 h period using Omron Glucometer. Quantification of AUC was achieved using Graphpad Prism software 5 software (GraphPad Software Inc., San Diego, CA, USA).

Renal cortex was processed, embedded in paraffin and cut into 5 μm-thick sections. Renal sections were stained with PAS. Semiquantitative scoring of glomerular sclerosis was performed in a blinded manner using a five-grade method described previously (Taneda et al., 2003). Twenty to thirty glomeruli randomly selected from per mouse were scored from six mice in each group. For immunohistochemistry, the sections were incubated with primary antibody against Wilms’ tumor-1 (WT-1), Podocin, monocyte chemotactic protein-1 (MCP-1), GRP78 (Abcam, Cambridge, MA, USA), TNF-α (Santa Cruz Biotechnology, Santa Cruz, CA, USA) overnight at 4°C, and then incubated with Biotinylated secondary antibody (Vector Laboratories, Burlingame, CA, USA) followed by VECTASTAIN ABC Reagent (Vector Laboratories, Burlingame, CA, USA) incubation for signal amplification. Color development was achieved using 3,3′-diaminobenzidine (DAB) (Vector Laboratories, Burlingame, CA, USA). The IOD was assessed by computer analysis with Image-Pro plus 6.0 (Media Cybernetics, MD, USA).

TUNEL staining was performed with the ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit (Millipore, Billerica, MA, USA) according to the manufacturer’s instructions. TUNEL-positive cells were semiquantified by randomly counting at least 30 glomeruli in each mouse.

Renal cortex or cultured podocytes were lysed with the buffer containing 50 mM Tris (pH7.5), 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 2 mM EDTA, 0.5 mM dithiothreitol, 1 mM PMSF, protease inhibitor cocktail and phosphatase inhibitor (both from Sigma-Aldrich, St. Louis, MO, USA). Proteins were separated by SDS-PAGE and western blotting was carried according to the standard procedures. The following primary antibodies were used: eIF2α, JNK, phospho-JNK^{Thr183/Tyr185}, AIF, cyclophilin D, and ANT from Santa Cruz Biotechnology (Santa Cruz, CA, USA), GAPDH, SERCA1, inositol-requiring enzyme 1α (IRE1α), TNF receptor associated factor 2 (TRAF2), PERK, phospho-PERK^Thr980, phospho-eIF2α^Ser51, activating transcription factor 4 (ATF4), caspase 12, caspase 9, caspase 3, CHOP, and cytochrome c from Cell Signaling Technology (Danvers, MA, USA), SERCA2, Podocin, Nephrin, GRP78, ATF6, XBP1, voltage dependent anion channel 1 (VDAC1), APAF1, and PPARγ from Abcam company (Cambridge, MA, USA), phospho-IRE1α^Ser724 from Novus Biologicals, SERCA3 from Proteintech Group (Chicago, IL, USA), ASK1 from ProSci Incorporated (San Diego, CA, USA). Secondary antibodies were horseradish peroxidase-conjugated goat anti-rabbit IgG (BOSTER, Wuhan, China) and goat anti-mouse IgG (BOSTER, Wuhan, China), and signals were detected using Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, MA, USA). Densitometric quantitation was performed using Image J 1.37 software (NIH, Bethesda, MD, USA). Protein expression was quantified as the ratio of specific band to GAPDH.

Total RNAs were isolated using Trizol reagents (Invitrogen, Carlsbad, CA, USA). First-strand cDNA was synthesized from 1 μg total RNA in a 20 μl reaction volume using a random primer (Takara, Dalian, China) and Moloney murine leukemia virus reverse transcriptase (New England Biolabs, Ipswich, MA, USA). SYBR green based qRT-PCR was performed in an Applied Biosystems ViiA 7^TM real time PCR system using the following thermal cycle reaction protocol: 50°C for 2 min, 95°C for 10 min, then 40 cycles at 95°C for 15 s, and at 60°C for 1 min. Primers used to amplify SERCA genes were used as previously described (Kono et al., 2012) and synthesized by Sangon Biotech (Shanghai) Co., Ltd. (Shanghai, China). The relative mRNA amount was normalized to the invariant β-actin mRNA species.

Sarco/endoplasmic reticulum Ca²⁺-ATPase activity was measured based on the inorganic phosphate production using a commercially available kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instruction. ER fraction was obtained from the kidney using a commercial kit (ER isolation kit, Sigma Aldrich, St. Louis, MO, USA). Briefly, kidney tissues or podocytes were homogenized in isotonic extraction buffer by using a glass homogenizer. After a series of centrifugation (1,000 × g for 10 min, 12,000 × g for 15 min, and 100,000 × g for 1 h), the microsomes were obtained and then homogenized with isotonic extraction buffer. Protein concentration of the ER fraction was quantified and the SERCA activities were determined.

Conditionally immortalized mouse podocytes, kindly provided by Prof. Niansong Wang (Shanghai Sixth People’s Hospital, China) and originally provided by Dr. Peter Mundel (Division of Nephrology, Massachusetts General Hospital, Harvard University), were cultured as previously described (Mundel et al., 1997). Differentiated podocytes were cultured in RPMI 1640 containing 1% FBS for 24 h in six-well plates (apoptosis assay) and 10-cm dishes (protein isolation) before being exposed to different cultural conditions. The cells were allocated into the following groups: (1) normal control (BSA) group incubated in RPMI 1640 containing 1% FBS and BSA, (2) palmitate group incubated in RPMI 1640 with 1% FBS and 250 μM palmitate complexed to BSA (palmitate medium), (3) AS-IV-treated group pretreated with 20, 40 or 80 μM AS-IV for 12 h followed by incubation in palmitate medium for 24 h, and (4) RGZ group pretreated with 10 μM RGZ for 12 h followed by incubation in palmitate medium for 24 h. AS-IV and RGZ were dissolved in DMSO (AMRESCO, Solon, OH, USA) and the final DMSO concentration did not exceed 0.1% (v/v). Palmitate was prepared and podocyte apoptosis was determined by flow cytometry as previously described (Sieber et al., 2010). In brief, a 20 mmol/L solution of palmitic acid (Sigma, St. Louis, MO, USA) in 0.01 mol/l NaOH was complexed to 5% BSA in a molar ratio of 8:1. Then the mixture was sterile filtrated and was added to the serum-containing cell culture medium to achieve a palmitic acid concentration of 250 μM. The final palmitic acid concentration in the medium was measured using a commercial kit (Wako Chemicals, Richmond, VA, USA). BSA was used for control experiments and handled exactly the same as BSA complexed to palmitic acid. Apoptotic podocytes were defined as annexin V-positive/PI-negative (early apoptotic) and annexin V-positive/PI-positive (late apoptotic) cells.

Oligonucleotide siRNA duplex was synthesized by Invitrogen (Carlsbad, CA, USA). The sequence of mouse SERCA2b siRNA was 5′-GACUCUGCUUUGGAUUAUATT-3′. The sequence of scramble siRNA was 5′-UUCUCCGAACGUGUCACGUTT-3′. Mouse SERCA2b cDNA (GenBank accession no. AJ131821) was synthesized and cloned into pIRES2-EGFP vector (Addgene, Cambridge, MA, USA) to create pIRES2-EGFP-SERCA2b plasmid (protein isolation), and then subcloned into pcDNA3.0 vector (Invitrogen, Carlsbad, CA, USA) to create pcDNA3.0-SERCA2b plasmid (apoptosis assay). The transfection of siRNAs or plasmids in podocytes was carried out with Lipofectamine^® RNAiMAX Transfection Reagent or Lipofectamine^® 3000 Reagent (both from Invitrogen, Carlsbad, CA, USA), respectively, according to the manufacturer’s instruction. After being transfected with SERCA2b siRNA or pcDNA3.0-SERCA2b/pIRES2-EGFP-SERCA2b plasmid for 24 h, podocytes were pretreated with or without 80 μM AS-IV for 12 h followed by incubation with 250 μM palmitate medium for 24 h and then collected for apoptosis assay or western blotting.

Podocytes cultured under indicated conditions were incubated with RPMI 1640 supplemented with 2 μM Fluo-4 AM (Invitrogen, Carlsbad, CA, USA) at 37°C for 30 min. Then the cells were harvested, pelleted, and suspended in ice-cold Ca²⁺-free Krebs-Ringer buffer and assessed by flow cytometry (BD FACSCalibur, Franklin Lakes, NJ, USA). The data were then analyzed with Flowjo 7.6.1 software (FlowJo LLC, Ashland, OR, USA).

Data are presented as means ± SEM, with n representing the number of animals for in vivo experiments and independent assays for in vitro experiments. Statistical comparisons were performed by using unpaired two-tailed t-test or one-way ANOVA followed by the Newman–Keuls multiple comparisons test, with P-values < 0.05 being considered significant. Statistical analyses were conducted with Graphpad Prism software 5 software (GraphPad Software, Inc., San Diego, CA, USA). The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015).

Diabetic db/db mice had a significantly higher body weight compared with normal db/m mice during the experiment. However, treatment with AS-IV at 18 mg kg^-1 day^-1 for 8 weeks significantly reduced body weight in db/db mice (Figure 1A). Moreover, AS-IV decreased food intake, water intake, 24 h urine volume and hypertension in db/db mice (Figures 1B–E). Db/db mice developed time-dependent, progressive albuminuria (expressed as urinary ACR) and exhibited a significant increase in BUN levels compared with nondiabetic db/m mice. Administration of AS-IV at 6 and 18 mg kg^-1 day^-1 significantly ameliorated the development of albuminuria (Figure 1F) and decreased BUN concentration in db/db mice (Figure 1G). RGZ had striking beneficial effects on the above parameters without altering body weight. No significant difference in serum creatinine levels was observed among all the groups (Figure 1H).

The db/db mice developed hyperglycemia as confirmed by significantly elevated fasting blood glucose (Figure 2A) and serum HbA1c (Figure 2B), which was markedly reduced by 6 and 18 mg kg^-1 day^-1 AS-IV (Figures 2A,B). OGTT experiment revealed that both 6 and 18 mg kg^-1 day^-1 AS-IV-treated db/db mice exhibited enhanced glucose clearance as demonstrated by the decreased AUC compared to db/db-vehicle group (Figures 2C,D). IPITT showed that insulin sensitivity was also improved significantly by 6 and 18 mg kg^-1 day^-1 AS-IV, confirmed by the decreases in AUC relative to db/db-vehicle group (Figures 2E,F). Serum insulin levels and HOMA-IR were also significantly decreased in db/db-6 mg kg^-1⋅day^-1 AS-IV group and db/db-18 mg kg^-1⋅day^-1 AS-IV group compared with db/db-vehicle group (Figures 2G,H). RGZ almost normalized the blood glucose and dramatically improved glucose tolerance and insulin sensitivity in db/db mice.

Diabetic nephropathy is a chronic renal complication accompanied by histopathologic changes in kidney (Cooper, 1998). Histological examination of the kidney revealed that the glomerular cross-sectional area was significantly greater in db/db mice than that in db/m mice at week 16. In addition, db/db mice developed more severe mesangial matrix expansion relative to their lean counterparts (Figure 3A). The 6 and 18 mg kg^-1 day^-1 AS-IV as well as RGZ markedly alleviated glomerular hypertrophy and mesangial matrix expansion (Figure 3A). Semiquantitative data further confirmed these observations (Figures 3B–D).

Inflammation is involved in DN development (Navarro-Gonzalez and Mora-Fernandez, 2008). The immunohistochemistry staining revealed that MCP-1 and TNF-α expression levels were dramatically elevated in renal tubular epithelial cells of db/db mice compared with db/m mice. However, AS-IV dose-dependently suppressed the induction of MCP-1 and TNF-α (Figure 3E). Quantification of IOD for MCP-1 and TNF-α further confirmed this observation (Figures 3F,G). RGZ also displayed powerful activity against renal inflammation. At week 16, kidney weight significantly increased and the percentage of kidney weight/body weight significantly decreased in db/db mice relative to nondiabetic db/m mice, whereas treatment with AS-IV or RGZ had no obvious effects on these changes (Figures 3H,I).

Podocyte apoptosis/loss is closely associated with the pathogenesis of albuminuria and glomerulosclerosis, thus being considered as a predictor for the progression of DN (Reidy et al., 2014). The number of apoptotic cells was dramatically increased in db/db mice relative to db/m mice as demonstrated by TUNEL staining, while AS-IV dose-dependently attenuated apoptotic cell number (Figures 4A,B). Immunohistochemistry staining of WT-1 (podocyte nuclei) and Podocin (podocyte foot processes) revealed marked podocyte loss in db/db mice, while AS-IV effectively prevented the loss of podocytes (Figure 4A), which was confirmed by semiquantitation of WT-1-positive cells and podocin-positive area in the glomeruli (Figures 4C,D). Western blot demonstrated that the expression levels of Podocin and Nephrin were reduced in renal cortex of db/db mice but was restored by AS-IV in a dose-dependent manner (Figures 4E,F). RGZ treatment significantly alleviated podocyte apoptosis in db/db mice (Figure 4).

To gain molecular mechanistic insights into the protective action of AS-IV against DN, alterations in SERCA expression were measured. qRT-PCR by using equally efficient primers adopted by Kono et al. showed that SERCA2b was the most prevalent mRNA species within the kidney and conditionally immortalized mouse podocyte cell line, while both SERCA2a and SERCA3 were expressed at extremely low levels (Figures 5A,B). In addition, western blot results showed that SERCA2 protein levels were significantly reduced by 50% in the renal cortex of db/db mice relative to that of db/m controls and there was no significant difference in SERCA3 protein levels between db/m and db/db mice (Figures 5C,D). SERCA1 expression could not be detected (Figure 5E). Given the qRT-PCR and western blot data, our following experiment focused on SERCA2b isoform. The decreased SERCA2 expression in db/db mouse kidney was restored by AS-IV in a dose-dependent manner (Figures 5F,G). In parallel to a marked depression in SERCA2 expression, SERCA activity was remarkably decreased by 50% in db/db mouse kidney compared to db/m mice. However, treatment with AS-IV at 6 and 18 mg kg^-1 day^-1 remarkably rescued SERCA activity in the kidneys of db/db mice (Figure 5H). RGZ significantly prevented the reductions in SERCA2 expression and activity (Figures 5F–H).

Aberrant expression or activity of SERCA2 leads to ER Ca²⁺ dysregulation and triggers ER stress (Lytton et al., 1991; Park et al., 2010; Fu et al., 2011). Immunohistochemistry staining revealed that the expression of ER chaperone GRP78 was dramatically up-regulated in the renal cortex of db/db mice compared with db/m mice, which indicated the activation of ER stress, but attenuated by AS-IV in a dose-dependent manner (Figures 6A,B). Western blot analysis further confirmed this observation (Figures 6C,D). Three ER stress master regulators ATF6, PERK, IREα and their downstream targets, such as eIF2α, ATF4, and XBP1 were activated, indicating the activation of all three branches of UPR signaling pathway. However, their activations were significantly suppressed by 6 and 18 mg kg^-1 day^-1 AS-IV treatment (Figures 6E–J). Three known mediators of ER stress-induced apoptosis, including ATF6/PERK downstream molecule CHOP, IRE1α downstream molecule phospho-JNK, and cleaved caspase-12, were all enhanced in db/db mice (Figures 6E–J). However, their expression was remarkably reduced by AS-IV, strongly manifesting that AS-IV abrogates ER stress and ER stress-induced apoptosis in db/db mice.

In parallel, disturbance of ER Ca²⁺ homeostasis by ER stress also evokes mitochondria-mediated apoptosis. Western blot analysis showed a decrease in anti-apoptotic protein Bcl-2 expression and an increase in the protein levels of pro-apoptotic proteins, such as Bax, cytochrome c, APAF1, and AIF in renal cortex of db/db mice compared with db/m mice, while AS-IV restored their expression (Figures 7A,B). There was no significant difference in the expression levels of VDAC, ANT, and cyclophilin D, components of the permeability transition pore (PTP), between db/m mice and db/db mice (Figures 7A,C).

db/db mouse is an obese mouse model for type 2 diabetes, characterized by dyslipidemia and an increase in plasma levels of long-chain free fatty acids (FFAs) (Wahba and Mak, 2007; Mooradian, 2009). Saturated FFAs such as palmitate are pro-apoptotic factors and lead to podocyte apoptosis by inducing ER stress (Martinez et al., 2008; Sieber et al., 2010). We further confirmed the protective effect of AS-IV on ER stress-induced apoptosis using palmitate-induced podocyte injury model in vitro. Palmitate induced a significant podocyte apoptosis while treatment with AS-IV significantly decreased palmitate-induced podocyte apoptosis in a concentration-dependent manner, manifested by flow cytometric analysis (Figures 8A,B). Podocin expression was significantly downregulated in palmitate-stimulated podocytes, accompanied with a dramatic reduction in SERCA2 expression and activity and an increase in GRP78 expression. However, AS-IV markedly restored Podocin expression as well as SERCA2 expression and activity, and GPR78 expression was significantly attenuated at the same time (Figures 8C–E). Further study showed that AS-IV suppressed the activation of three UPR arms including ATF6, PERK-eIF-2α-ATF4, and IRE1α-XBP1 (Figures 8F–I). ER stress-mediated apoptotic pathway was activated by palmitate but inhibited by AS-IV, as manifested by the down-regulation of CHOP, TRAF2, ASK1, phospho-JNK and cleaved caspase 12 in AS-IV treated group relative to those in palmitate group (Figures 8F–K).

Consistent with a decrease in SERCA2 expression and activity, palmitate resulted in a rise in basal cytosolic Ca²⁺ levels. AS-IV restored basal cytosolic Ca²⁺ levels in a dose-dependent manner (Figures 9A,B). Meanwhile, palmitate induced the expression of cytochrome c, APAF1 and AIF in podocytes, suggesting the activation of mitochondria-mediated apoptotic pathway. However, the induction of cytochrome c, APAF1, and AIF was markedly suppressed by AS-IV in a dose-dependent manner. The expression levels of PTP proteins, including VDAC, Cyclophilin D, and ANT, were not significantly changed under different cultural conditions (Figures 9C–E).

To further confirm the role of SERCA2b in mediating the effect of AS-IV on ER stress-induced podocyte apoptosis, we knocked down endogenous SERCA2b expression using a SERCA2b-specific siRNA or overexpressed the exogenous SERCA2b by transfection with SERCA2b overexpression plasmid. SERCA2 expression was reduced by 70% after SERCA2b siRNA transfection (Figure 10A) and increased significantly after SERCA2b overexpression plasmid transfection (Figure 10B). We then transfected SERCA2b siRNA or SERCA2 overexpression plasmid into podocytes to investigate the role of SERCA2b in the action of AS-IV. Flow cytometric analysis showed that SERCA2b knockdown resulted in a significant increase in palmitate-induced apoptosis and abolished the protective effect of AS-IV on podocytes, while SERCA2b overexpression significantly decreased palmitate-induced apoptosis and achieved a similar beneficial effect as AS-IV on podocytes (Figures 10C,D). Further data showed that SERCA2b knockdown markedly increased palmitate-induced intracellular Ca²⁺ rise and ER stress, enhanced the activation of ER stress-mediated apoptotic pathway and blocked the inhibitory effect of AS-IV on palmitate-induced intracellular Ca²⁺ alteration and ER stress, while SERCA2b overexpression significantly attenuated palmitate-induced intracellular Ca²⁺ rise and ER stress, repressed the activation of ER stress-mediated apoptotic pathway, and achieved a similar potential as AS-IV on intracellular Ca²⁺ dysregulation and ER stress (Figures 10E–I).