All animal experiments were performed under the approval of the Institutional Animal Care and Use Committee of Seoul National University Hospital (Seoul, Korea; approval no. 12-0132), according to the Guide for the Care and Use of Laboratory Animals of the National Research Council. Wild type and C-C chemokine receptor (CCR) type 5-deficient (CCR5−/−) mice were obtained from the Orient Company (Seoul, Korea) and the Jackson Laboratory (Bar Harbor, ME, USA) on a C57BL/6 background. Animals were maintained at 22±2°C with a relative humidity of 40–60% and a 12 h light/dark cycle. Animals had ad libitum access to a commercial rodent diet and tap water. A total of 72 adult male mice (7-8 weeks old) were used for the 5/6 nephrectomy model. To produce visible renal ischemia in one-third of the kidney, the lower branch of the right renal artery was suture ligated, and the upper branch of the right kidney was amputated using electrocoagulation. The upper branch of the left renal artery was then ligated, and the lower pole was placed back in the renal fossa by cauterization. Left nephrectomy was performed 1 week later, which indicated the onset of moderate-to-severe kidney injury (day 0). Parameters such as blood pressure, serum blood urea nitrogen, creatinine, and proteinuria were subsequently monitored.

The study protocol complied with the Declaration of Helsinki and received full approval from the Institutional Review Board at Seoul National University Hospital (approval no. 1607-133-777). Primary human podocytes were obtained in January 2017 from the unaffected pole of a nephrectomy specimen from a patient with renal cell carcinoma (45 years old, male) from the Seoul National University Hospital (Seoul, Korea). Kidney cortices were mechanically dissected. Glomerular cells were sieved using optimized media (17), and isolated by flow cytometry for a podocalyxin-positive (cat. no. FAB1658P; R&D Systems, Inc., Minneapolis, MN, USA) and CD90-negative (cat. no. 11-0909-41; eBioscience; Thermo Fisher Scientific, Inc., Waltham, MA, USA) population. The purities of sorted cells were >98%. Subsequent gating analyses were conducted with the FlowJo software (version 10; FlowJo LLC, Ashland, OR, USA). The overall procedure was performed as previously described (18).

For in vitro stimulation, cells were maintained at a concentration of 1×10⁵ cells per well in 6-well plates, and starved for 24 h. They were then treated with 1 μM retinoic acid (cat. no. R2625; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) or 2 ng/ml recombinant transforming growth factor-β (TGF-β) for 48 h (R&D Systems, Inc.). In other experiments, cyclosporine (Calbiochem; Merck KGaA) was used as a renoprotective agent. Small interfering RNA (siRNA) was transfected into cells in order to knockdown the KLF15 expression as suggested in instructor's protocol (cat. no. sc-45567; Santa Cruz Biotechnology, Inc., Dallas, TX, USA). For KLF15 silencing, 1 μmol human KLF15 siRNA or 1 μmol scrambled non-targeting siRNA (cat. no. sc-36869) was added.

Briefly, 24 h prior to transfection, cells were washed, detached with trypsin, diluted to 1:5 with fresh medium without antibiotics (2×10⁵ cells/ml), transferred to 6-well plates (1 ml/well) and grown to 60–80% confluency. A mixture of KLF15 siRNA with transfection reagent (cat. no. sc-29528; Santa Cruz Biotechnology) into siRNA transfection medium (cat. no. sc-36868; Santa Cruz Biotechnology) was prepared and incubated for 30 min at room temperature. Complete medium was subsequently replaced with 1 ml of the above transfection mixture. Approximately 6 h later, this was replaced with 1 ml normal complete medium. After 24 h, cells were treated with recombinant TGF-β (2 ng/ml) and cyclosporine (2 μmol) in fresh complete medium for 48 h. Specific silencing of KLF15 was confirmed using western blot analysis in three independent experiments.

Harvested mouse or biopsied human kidney tissue samples were fixed in 4% paraformaldehyde at 4°C overnight. Paraffin-embedded tissue sections (4 μm) were deparaffinized and rehydrated with xylene three times (5 min each) and a descending ethanol series (100, 100, 95, 90, 80 and 70%; 5 min each), respectively. Endogenous streptavidin activity was blocked in 0.3% hydrogen peroxide. Antigens were retrieved by heating paraffin-embedded sections to ~95°C in 10% citrate buffer in a microwave oven, five times for 5 min each. Blocking reagent (cat. no. X0909; Agilent; Santa Clara, CA, USA) was incubated with sections at room temperature for 1 h to block non-specific background staining. The sections were subsequently stained with periodic acid-Schiff and Sirius red (cat. no. ab150681; Abcam, Cambridge, UK), and antibodies against KLF15 (1:100; cat. no. AV32587; Sigma-Aldrich; Merck KGaA), and counterstained with hematoxylin at 4°C overnight. Morphometric parameters were captured using a microscope coupled to a computerized morphometry system (Leica Microsystems, Rijswijk, Netherlands). By the use of isotype controls for the anti-KLF15 antibody (1:100; cat. no. A0545; Sigma-Aldrich; Merck KGaA), non-specific staining could be excluded.

Confocal imaging was also performed following staining with DAPI (1:1,000; cat. no. 62248; Thermo Fisher Scientific, Inc.). Blocking reagent (cat. no. X0909; Agilent) was added for 1 h at room temperature. PBS-diluted antibodies against synaptopodin (1:100; cat. no. 61094; Progen Biotechnik, Heidelberg, Germany), nephrin (1:100; cat. no. PA5-25932; Thermo Fisher Scientific, Inc.), CD31 (1:50; cat. no. ab56299; Abcam), Wilms tumor-1 (1:100; cat. no. ab8990-1; WT-1; Abcam), and KLF15 (1:100; cat. no. AV32587; Sigma-Aldrich; Merck KGaA) were subsequently incubated with the sections at 4°C overnight. Alexa Fluor 488- (1:400; cat. no. A-10631; Thermo Fisher Scientific, Inc.), Alexa Fluor 555-(1:400; cat. no. A-21433; Thermo Fisher Scientific, Inc.), and Alexa Fluor 647-antibodies (1:400; cat. no. A-21445; Thermo Fisher Scientific, Inc.) were then added at room temperature for 30 min. Confocal microscopy was performed with a Leica TCS SP8 STED CW (Leica Microsystems).

Primary human podocytes were lysed with radioimmunoprecipitation assay lysis buffer (cat. no. 89900; Thermo Fisher Scientific, Inc.) supplemented with complete protease inhibitor (cat. no. 78420; Thermo Fisher Scientific, Inc.). The kidney homogenate was centrifugated at 12,000 × g and 4°C for 30 min. Protein concentration of the supernatant was determined with the Bradford method. Equal amounts (80 μg) of extracted protein were separated by 10% SDS-PAGE and transferred onto Immobilon-FL 0.4 μM poly-vinylidene difluoride membranes (Millipore, Bedford, MA, USA). Blocking was conducted at room temperature for 1 h with blocking buffer (5% skin milk and 2% bovine serum albumin (cat. no. 10735086001; Sigma-Aldrich; Merck KGaA) in PBS with 0.2% Tween 20). The membranes were then incubated with antibodies against fibronectin (1:300; cat. no. sc-6952; Santa Cruz Biotechnology, Inc.), zonula occludens-1 (ZO-1; 1:300; cat. no. 339100; Invitrogen; Thermo Fisher Scientific, Inc.), WT-1 (1:400; cat. no. sc-192; Santa Cruz Biotechnology, Inc.), β-actin (1:10,000; cat. no. a5316; Sigma-Aldrich; Merck KGaA), and KLF15 (1:200; cat. no. ab167192; Abcam) at 4°C overnight. Horseradish peroxidase-conjugated secondary anti-rabbit (1:4,000; cat. no. 7074S; Cell Signaling Technology, Inc. Danvers, MA, USA), anti-goat (1:4,000; cat. no. sc-2005; Santa Cruz Biotechnology, Inc.), and anti-mouse (1:4,000; cat. no. 7076S; Cell Signaling Technology, Inc.) immunoglobulin G antibodies were added for 1 h at room temperature. The blots were developed and enhanced using a Super Signal West Pico Chemiluminescence kit. Band intensity was assessed using ImageJ software (version 1.5; National Institutes of Health, Bethesda, MD, USA).

Total RNA was extracted from primary human cells and mouse renal tissue with an RNeasy Micro kit (Qiagen, GmbH, Hilden, Germany). Subsequently, RNA was converted to cDNA using a kit according to the manufacturer's instructions (cat. no. A3500; Promega Corporation, Madison, WI, USA). The TaqMan gene expression assay was used for human fibronectin and GAPDH (Applied Biosystems; Thermo Fisher Scientific, Inc.). The SYBR Green expression assay was used for mouse fibronectin, interleukin (IL)-6, monocyte chemotactic protein (MCP)-1, IL-10, CCR5, KLF15, and GAPDH (Applied Biosystems; Thermo Fisher Scientific, Inc.). Primer sequences were as follows: Mouse fibronectin, forward 5′-ATGCAACGATCAGGACACAA-3′ and reverse 5′-TGTGCCTCTCACACTTCCAC-3′; mouse IL-6, forward 5′-CCTCTGGTCTTCTGGAGTACC-3′ and reverse 5′-CGACCTCAGTGTCTTCCTCA-3′; mouse MCP-1, forward 5′-GCTCAGCCAGATGCAGTTAA-3′ and reverse 5′-TCAAAAACAGTGGTTCGAGTTCT-3′; mouse IL-10, forward 5′-ATAACTGCACCCACTTCCCA-3′ and reverse 5′-TGGACCATCTTCACTACGGG-3′; mouse CCR5, forward 5′-TTGTCTACTTTCTCTTCTGG-3′ and reverse 5′-ATCGGGTATAGACTGAGC-3′; mouse KLF15, forward 5′-CACCAAGAGCAGCCACCTCA-3′ and reverse 5′-CGGGACACTGGTACGGCTTC-3′; mouse GAPDH, forward 5′-CACCAAGAGCAGCCACCTCA-3′ and reverse 5′-CGGGACACTGGTACGGCTTC-3′. Thermocycling conditions were as follows: 95°C for 10 min followed by 40 cycles of dissociation (95°C for 10 sec), annealing (55°C for 30 sec), and elongation (72°C for 30 sec). Data were quantified using the comparative Cq (2^−∆∆Cq) method (19).

Biopsy samples from human kidney tissue were obtained from patients with primary membranous nephropathy (n=26; 27-76 years old; male, 69.2%) or diabetic nephropathy (n=21; age, 24-67 years old; male, 61.9%) from the Seoul National University Hospital (Seoul, Korea) between April 2013 and April 2015. Glomerular KLF15 expression was quantified microscopically, and its association with clinical outcomes was analyzed. The clinical outcomes that were analyzed for membranous nephropathy and diabetic nephropathy were spontaneous remission and end-stage renal disease, respectively.

All analyses and calculations were performed using SPSS (version 23.0; IBM Corp., Armonk, NY, USA) and GraphPad Prism (version 5.0; GraphPad Software, Inc., La Jolla, CA, USA). The results are expressed as mean ± standard error of mean. Differences between groups in the western blot, RT-qPCR, and immunohistochemistry analyses were evaluated using one-way analysis of variance, followed by the least significant difference post hoc test. Survival curves were drawn using the Kaplan-Meier method. To compare survival curves between groups, the log-rank test was applied. P<0.05 was considered to indicate a statistically significant difference.

The 5/6 nephrectomy model was used to induce chronic podocyte injury. Additionally, CCR5-/- mice were used to induce podocyte injury with increased severity through alteration of macrophage composition (20). Twenty weeks post-injury, kidney markers such as blood urea nitrogen, creatinine, and the urine protein-to-creatinine ratio were increased in nephrectomized mice compared with sham-operated mice (Fig. 1A). Other surrogate markers, such as systolic blood pressure and body weight exhibited significant differences between nephrectomized and sham-operated mice (Fig. 1A). When these markers were compared between nephrectomized wild type and CCR5−/− mice, aggravated injury was observed in the CCR5−/− mice. In specific, the measurements were: Body weight (g), 26.1±0.44 vs. 24.5±0.33; blood urea nitrogen (mg/dl), 51.8±3.10 vs. 70.0±4.21; creatinine (mg/dl), 0.58±0.03 vs. 0.77±0.04; and urine protein-to-creatinine ratio (mg/mg), 13.9±0.97 vs. 17.6±1.00 in wild type and CCR5−/− mice, respectively (Fig. 1A). Fibronectin, a fibrosis marker, was expressed in higher levels in nephrectomized mice, but podocyte KLF15 expression was decreased and correlated with the degree of injury (Fig. 1B–1D). Production of inflammatory cytokines, IL-6 and MCP-1, was significantly increased (Fig. 1D), while IL-10 production was decreased depending on the degree of podocyte injury (wide type vs. CCR5−/− mice; Fig. 1D). These results suggest that chronic podocyte injury is associated with low KLF15 expression.

Next, we aimed to confirm that low KLF15 expression is associated with chronic podocyte injury using an in vitro model of primary human cells. Primary human podocytes were isolated and demonstrated to express synaptopodin and nephrin as podocyte markers, while they were negative for the endothelial cell marker, CD31 (data not shown). TGF-β is a potent inducer of fibrosis (21), and therefore it was used to induce podocyte fibrosis in vitro. Results from western blot analysis confirmed that fibronectin levels as a fibrosis marker increased when cells were treated with TGF-β for 48 h (Fig. 2).

Retinoic acid and glucocorticosteroids are known to induce the expression of KLF15 in podocytes (11,14). These observations suggest other renoprotective agents may regulate KLF15 expression as part of their protective function. Cyclosporine, a calcineurin inhibitor, was evaluated in the present study, because it is routinely administered to patients with podocytopathy in clinical practice (22). Compared with vehicle-treated cells, cells treated with cyclosporine displayed increased levels of KLF15 and decreased levels of fibronectin that were comparable with the positive control group treated with retinoic acid (Fig. 2A–C). siRNA was then used to knockdown expression of KLF15 in the cells. Following siRNA transfection, the protein expression levels of KLF15 were successfully decreased (Fig. 2D). The siRNA-treated cells exhibited lower ZO-1 and higher fibronectin expressions compared with scramble-treated cells, following TGF-β treatment (Fig. 2E and F). The renoprotective effect of cyclosporine was not evident in the KLF15-knockdown cells (Fig. 2E and F). Cyclosporine could maintain podocytes in a healthier state throughout enrichment of KLF15, in addition to the previous knowledge regarding cytoskeleton stabilization (16).

Previous studies have demonstrated that the decreased KLF15 expression in biopsied kidney tissue from patients with minimal change disease and focal segmental glomerulonephritis was associated with low response to glucocorticosteroids (14). In the present study, immunohistochemistry results demonstrated that decreased KLF15 expression was predominant in segmental sclerotic glomeruli (Fig. 3A). Next, we aimed to explore the relationship between KLF15 expression and renal outcomes in other podocytopathic diseases, such as membranous nephropathy and diabetic nephropathy. Regarding membranous nephropathy, cases with spontaneous remission (i.e. no use of immunosuppressive drugs) exhibited higher KLF15 expression compared with cases without spontaneous remission (Fig. 3B and C). However, KLF15 expression did not differ between the remission and non-remission cases following administration of immunosuppressive drugs (Fig. 3C). Regarding diabetic nephropathy, the patients were divided into three groups according to KLF15 expression levels (i.e. tertiles), and their renal function was followed for 30 months. The third tertile group (highest expression) did not exhibit doubling of serum creatinine (P=0.020) or end-stage renal disease (P=0.039; Fig. 3D and E). These results indicate that KLF15 expression is associated with renal outcomes commonly observed in several podocytopathic diseases.