Subjects and study design
We recruited patients with T2DM at the Department of Endocrinology from The First Affiliated Hospital of Soochow University. Patients who conforming to the following criteria were included: (1) patients between the age of 18–75 years; (2) patients who could undertake magnetic resonance imaging (MRI); (3) those who consented to participate in the present study; (4) patients under treatment with metformin or metformin combined with EMPA for at least 3 months. The patients with the following criteria were excluded: (1) those whose body mass index (BMI) < 18 kg/m2; (2) those with unstable physiological conditions or serious psychiatric disorder; (3) pregnant women.
This prospective work was approved by Local Ethics Committee of our hospital (Ethical No. 145). Each participant provided written informed consent before the examinations. We recruited 70 patients between November 2019 and October 2021. We eliminated patients who could not hold their breath in MRI examination (n = 2) or those who quit (n = 2). Finally, we enrolled 66 participants for data analyses. Among 66 subjects, 15 were normal volunteers with no history of diabetes and no abnormality showed up in the laboratory tests concerning serum creatinine (Scr), glycated hemoglobin (HbA1c), and fasting blood glucose (FBG) level. The rest 51 T2DM patients were assigned into two groups: those treated by metformin (n = 31, T2DM group) and those treated with a combination of metformin and EMPA (n = 20, T2DM + EMPA group).
Anthropometric and biochemical measurements
Patients were physically examined, including their heights and weights and BMI [weight (kg) divided by the square of height (m2)]. We also collected peripheral blood samples (in a fasting state) from every patient to analyze FPG, SCR, total cholesterol (TC), and total triglycerides (TG) with the automated biochemical analyzer (HITACHI 7600, HITACHI Company, Japan). We determined HbA1c level by the automated glycosylated hemoglobin analyzer (HLC-723G8, TOSOH) and human serum AGEs contents by the ELISA kit per specific protocols (CUSABIO, Wuhan, China).
Dixon MRI for renal fat measurements
The subjects underwent an MRI scan about 1 h after lunch. Each subject was asked to lie in the supine position during MRI scanning, with the 8-channel receiver coil being adopted for the concentration of standard torso phased-array coil in the liver at 1.5 T (SuperVan, Lonwin Medical System, China). Scanning protocols included the original localizer images and subsequent axial images by adopting the multi-echo liver-interpolated volume-excitation sequence using parameters below including 3 echoes of 2.25, 3.37, and 4.5 ms, separately, one 12° flip angle, forty 2.5 mm slices, a 256 × 205 mm matrix, a 400 × 320 mm field of view, and 38-s overall acquisition time (the initial 19-s scanning under free-breathing state, then 19-s scanning under breath-holding state). The participants held the breath in their final inspiration for ensuring data consistency. We adopted a plug-in algorithm to automatically generate MRI-FF maps using WinStation software (WinStation, Lonwin Medical System, China). An experienced radiologist, kept unaware of the study design to avoid any bias, reviewed the images by WinStation. We chose three central slices from bilateral renal sides. Thereafter, we manually put regions of interest (ROIs) onto every slice within the whole renal parenchyma. We determined the average renal fat fraction (RFF) by taking the average value of bilateral side measurements for every participant. The coefficient variations (CV) of repetition is 4%.
Animal experiments
The study with db/db (diabetic) as well as db/m (non-diabetic) male mice of strain C57BL/KsJ was conducted at National Model Animal Centre of Nanjing University (Nanjing, China). This work followed the updated Helsinki Manifesto. The Ethics Committee of Soochow University approved our study protocols. All mice were kept in polypropylene cages at 22 ± 2 °C, 60 ± 5% relative humidity (RH) and 12-h:12-h light/dark cycle. Animals were raised for 2 weeks. Specifically, we raised a group of 8-week-old db/db mice with the AGE-rich diet, and the other group was fed with AIN-76 basal diet (Xietong Biology Co., Ltd., Nanjing, China). AIN-76 basal diet contains 64% carbohydrate, 20% protein, 3.9 Kcal/g energy, and 7% fat; following 10-min heat treatment under 90 °C, diet-derived AGEs were produced from the AIN-76 basal diet [23]. We then classified animals into five groups: including Group 1 (db/m mice, control) raised on AIN-76 basal diet; Group 2 (db/db mice) raised on AIN-76 basal diet; Group 3 (db/db mice + EMPA) raised on AIN-76 basal diet plus intragastric administration of EMPA (10 mg/kg, MedChemExpress, China) once a day from week 6 to week 8; Group 4 (db/db mice + AGEs) raised on an AGE-rich diet; Group 5 (db/db + AGEs + EMPA) raised on an AGE-rich diet plus intragastric administration of EMPA (similar to Group 3). We raised mice in respective metabolic cages and collected their urine at 24-h. The week eight samples were used for subsequent analyses. Blood samples were subjected to biochemical tests, while renal tissue samples were processed for histological analysis.
Biochemical assays
Each mouse was sacrificed after the experiments were over. We collected blood samples from the right ventricle of the heart for biomechanical analyses performed later. We measured the levels of blood urea nitrogen (BUN), SCR, FBG, TG, and TC by a fully-automated biochemical analyzer (Hitachi). The urinary neutrophil gelatinase-associated lipocalin (u-NGAL, the marker for renal tubular injury) as well as serum AGEs were measured with an ELISA kit (CUSABIO; Wuhan, China). Meanwhile, to detect protein content from the urine collected at 24-h, this work conducted Coomassie brilliant blue protein assay (Jiancheng Bioengineering Institute, Nanjing, Jiangsu).
Renal morphology
The renal cortex was embedded in paraffin and stored appropriately. The 3-µm cross-sections of tissue were kept on the gelatin-coated slides, followed by hematoxylin-eosin (HE) as well as periodic acid Schiff (PAS) staining.
Cell culture
This work obtained HK-2 cells in American Type Culture Collection (Manassas, VA, USA), then cultivated them following the method described elsewhere [11]. All assays were conducted within serum-free RPMI 1640 medium containing 0.2% bovine serum albumin (BSA, fatty acid free; SigmaAldrich; Merck KGaA). We obtained EMPA in MedChemExpress (Shanghai, China) and AGEs-BSA from Abcam (Cambridge, UK). HK-2 cells were cultivated in experimental medium including 200 µg/mL AGEs-BSA, 500 nM EMPA, or the combination of these two for 48-h.
Lipid deposition measurement
We used oil red O stain to observe lipid deposition within db/db mouse kidneys as well as HK-2 cells. In brief, four critical steps were needed: after fixation using 4% paraformaldehyde (PFA), samples were subject to 30-min oil red O staining, 5-min HE staining, and stained slides were observed under the optical microscope (Carl Zeiss, Hertfordshire, UK).
Intracellular cholesterol quantification
Enzymatic analysis (Applygen Technologies Inc., Beijing, China) was performed to quantify free cholesterol (FC) and TC content under both in vivo and in vitro conditions, as described earlier [9, 11]. Cholesterol ester (CE) was evaluated as per the following equation, CE = TC − FC.
Protein isolation and Western-blot (WB) assay
The nuclear and whole-cell proteins were extracted and denatured before WB assay following the protocol described elsewhere [11, 24]. We incubated the blots with antibodies specific for SREBP-2, SCAP, receptor of AGEs (RAGE), HMGCoAR, LDLr, and GAPDH (dilution, 1:200–1000, Abcam, Cambridge, UK). Incubation was carried out in Tween 20 (TBST), supplemented with 5% BSA, overnight at 4 °C. Blots were rinsed, and membranes were further probed for 1 h using secondary antibody (1:5000, within TBST, supplemented with 1% BSA; Santa Cruz Biotechnology) under ambient temperature. Finally, blots were exposed to X-ray films for detecting bands after the use of ECL (Pierce, Rockford, IL, USA). To quantify protein expression, we assessed band density using LabWorks software (UVP Laboratory Products, Upland, CA, USA); GAPDH was applied as the internal reference.
Confocal microscopy
As described previously, human SCAP cDNA was ligated with pEGFP-C1 vector at BstE-XbaI restriction sites (Genechem Co. Ltd., Shanghai, China) for creasing the green fluorescent protein (GFP)-SCAP expression construct [9, 11]. The clone, pEGFP-SCAP, was transfected into cells using the Effectene Transfection Reagent (Invitrogen, Paisley, UK) in line with specific protocols. Afterward, we inoculated HK-2 cells onto the chamber slides. At 48-h after diverse treatments, this work fixed cells for 30 min using 5% formalin, followed by 15-min permeabilization using 0.25% Triton X-100, and 2-h addition of Golgin-97 antibody (Molecular Probes, Inc., Eugene, OR, USA) at room temperature. Samples were then washed and incubated for 1 h with the secondary antibody. This work utilized Zeiss LSM 510 Meta confocal microscope (Carl Zeiss, Hertfordshire, UK) to visualize the slides.
Coimmunoprecipitation (Co-IP)
Co-IP was performed to analyze SCAP-SREBP-2 interaction within HK-2 cells. We initially isolated total proteins using the cell IP lysis buffer (Thermo Scientific Pierce, USA). After sample centrifugation, this work removed cell lysates using the protein A/G-agarose beads. In addition, supernatant was collected for immunoprecipitation with appropriate antibodies. Samples were further incubated with additional antibodies for immunoblotting assay.
Statistical analysis
Experimental data were analyzed by SPSS20.0 and presented as mean ± SD, mean ± S.E.M, or median (25th and 75th percentiles). One-way ANOVA or Bonferroni test was used for multiple data comparisons. The association of AGEs with RFF was assessed by Spearman’s correlation. p < 0.05 stood for statistical significance.