Electronegative low density lipoprotein induces renal apoptosis and fibrosis: STRA6 signaling involved[S]

Dyslipidemia has been proven to capably develop and aggravate chronic kidney disease. We also report that electronegative LDL (L5) is the most atherogenic LDL. On the other hand, retinoic acid (RA) and RA receptor (RAR) agonist are reported to be beneficial in some kidney diseases. “Stimulated by retinoic acid 6” (STRA6), one retinol-binding protein 4 receptor, was recently identified to regulate retinoid homeostasis. Here, we observed that L5 suppressed STRA6 cascades [STRA6, cellular retinol-binding protein 1 (CRBP1), RARs, retinoid X receptor α, and retinol, RA], but L5 simultaneously induced apoptosis and fibrosis (TGFβ1, Smad2, collagen 1, hydroxyproline, and trichrome) in kidneys of L5-injected mice and L5-treated renal tubular cells. These L5-induced changes of STRA6 cascades, renal apoptosis, and fibrosis were reversed in kidneys of LOX1−/− mice. LOX1 RNA silencing and inhibitor of c-Jun N-terminal kinase and p38MAPK rescued the suppression of STRA6 cascades and apoptosis and fibrosis in L5-treated renal tubular cells. Furthermore, crbp1 gene transfection reversed downregulation of STRA6 cascades, apoptosis, and fibrosis in L5-treated renal tubular cells. For mimicking STRA6 deficiency, efficient silencing of STRA6 RNA was performed and was found to repress STRA6 cascades and caused apoptosis and fibrosis in L1-treated renal tubular cells. In summary, this study reveals that electronegative L5 can cause kidney apoptosis and fibrosis via the suppression of STRA6 cascades, and implicates that STRA6 signaling may be involved in dyslipidemia-mediated kidney disease.

Hospital Institutional Review Board (KMUH-IRB-20130397). Plasma L1 and L5 were obtained from blood samples of 30 patients with metabolic syndrome (MetS) in Kaohsiung Medical University Hospital, and used for animal and cell experimentation. MetS was defined according to modification criteria for Asian populations from NCEP in Adults (ATP-III). Subjects with more than three of the following five criteria were diagnosed as MetS: 1) waist circumference 90 cm in men and 80 cm in women; 2) TG 150 mg/ dl; 3) HDL cholesterol <40 mg/dl in men and <50 mg/dl in women; 4) blood pressure >130/85 mmHg or taking antihypertensive medication; and 5) fasting plasma glucose 100 mg/dl and/ or taking antidiabetic agents. Human blood samples were added into 50 mU/ml aprotinin, 1% ampicillin/streptomycin, and 5 mmol/l EDTA immediately after collection to prevent contamination and experimental oxidation. LDL particles were isolated by sequential potassium bromide density centrifugation to remove chylomicron, VLDL, and IDL fractions, and yielded LDL at a final density of 1.019-1.063. LDL was equilibrated by dialysis in a column loaded with 20 mmol/l Tris-HCl (pH 8.0), 0.6 mmol EDTA, and 0.01% NaN 3 . LDL samples were injected into a UnoQ12 anion-exchange column by using anion-exchange columns (Uno-Q12; Bio-Rad Laboratories, Inc.) with the AKTA fast protein liquid chromatography system (GE Healthcare Life Sciences, Little Chalfont, UK) and eluted with multistep chloride gradient and then divided into subfractions L1 and L5. The proportion of L5 levels was 5.3 ± 6.9% in MetS subjects. The value was significantly higher than that (2.1 ± 1.4%) in healthy subjects.

Animal studies
Eight-week-old chow diet-fed C57B6/J male mice (n = 3) were injected with 150 l L5 or L1 (1 mg/kg) through the tail vein every day for 4 weeks. An equal volume of saline was also used as control. Eight-week-old chow diet-fed LOX1 knockout mice (LOX1 / , n = 3) were injected with 150 l L5 (1 mg/kg) daily for 4 weeks. At the end of the experiments, animals were euthanized with chloral hydrate by ip injection, and kidneys were harvested for further analyses. The Institutional Animal Care and Use Committee of Kaohsiung Medical University approved all animal experiments (IACUC number 102149). C57B6/J mice were purchased from BioLASCO Taiwan Co., Ltd. (Taipei, Taiwan). All animals were housed and cared for in a pathogen-free facility at Kaohsiung Medical University.

siRNA transfection
For cell experiments of LOX1 and STRA6 knockdown, HK-2 cells were seeded in 6-well plates at a density of 2 × 10 5 cells per well in 2 ml antibiotic-free medium and then cells were cultured at be more electronegative than LDL from control subjects (23). Therefore, it is worthy of investigating whether L5 can cause kidney damage.
Vitamin A (retinol) and its derivatives regulate a wide range of crucial biological functions, such as development, differentiation, metabolism, and immunity. Circulating retinol is bound to retinol-binding protein 4 (RBP4) and connects with transthyretin to produce a retinol-RBP4transthyretin complex (24,25). This complex is recognized by one RBP4 receptor, termed "stimulated by retinoic acid 6" (STRA6), which transports retinol into cells from the retinol complex (26,27). The STRA6-mediated translocation of retinol requires the participation of cellular retinol-binding protein 1 (CRBP1), an intracellular retinol acceptor, as well as retinol-metabolizing enzymes, such as lecithin:retinol acyltransferase (28)(29)(30). Retinol in cells is thereafter metabolized into retinoic acid (RA). RA exerts its function by binding to cytosolic nuclear receptors, including RA receptors (RARs) and retinoid X receptor (RXR), which can activate the transcription of numerous target genes (24,25). RA has also been reported to capably inhibit inflammation, apoptosis, and proliferation (31,32). Furthermore, RA and RAR agonist have been reported to have beneficial effects in several experimental models of kidney diseases (33)(34)(35)(36). Accordingly, it is reasonable to hypothesize that STRA6 and its cascades could be altered and involved in pathogenesis of kidney diseases. Therefore, the main goals of this study were to explore: 1) whether L5 treatment can cause kidney apoptosis and fibrosis in L5-treated mice and renal tubule cells; 2) whether L5 treatment can simultaneously alter STRA6, CRBP1, retinol, RA, RARs, and RXR; and 3) whether the alteration of STRA6-mediated cascades are involved in kidney apoptosis and fibrosis caused by L5.

L5 of human subjects
The treatment of all human subjects in this study was performed in accordance with the ethical guidelines of the 1975 Declaration of Helsinki and approved by the Kaohsiung Medical University 37°C and 5% CO 2 until the cell growth covered 80% of the area of the dish. After overnight incubation, negative control scramble siRNA (Santa Cruz Biotechnology Inc.), LOX1 siRNA (Santa Cruz Biotechnology Inc.) or STRA6 siRNA (Santa Cruz Biotechnology Inc.) was mixed into transfection reagent (Santa Cruz Biotechnology Inc.). The transfection medium (Santa Cruz Biotechnology Inc.), and the mixture were added into cells and incubated for 7 h. These cells were placed in fresh medium for 24 h, and then stimulated with PBS, native L1 (50 g/ml), or L5 (50 g/ml) for 24 h.

CRBP1 cDNA transfection
For cell experiments of CRBP1 overexpression, the crbp1 genetransfected HK-2 cells were established. The pCMV6-GFP vector and human crbp1 cDNA (gene number NM_002899) were purchased from OriGene Technologies Inc. (Rockville, MD). The CRBP1 cDNA was inserted into the Sgf 1/Mlu 1 site of the pCMV6-GFP expression vector plasmid (OriGene Technologies Inc.). HK-2 cells were transfected by using pCMV6-crbp1-I-GFP or pCMV6-GFP vector with Lipofectamine 2000 (Invitrogen). Cells were incubated in Opti-MEM (Invitrogen) at 37°C for 5 h and then placed in freshly changed culture medium for experiments. These cells were treated with PBS, native L1 (50 g/ml), or L5 (50 g/ml) for 24 h.

Western blot
The protein of kidney, HK-2, or RTECs was extracted with M-PER mammalian protein extraction reagent (Pierce Biotechnology, Rockford, IL). The protein of samples was separated with SDS-PAGE. The separated proteins on SDS-PAGE were transferred onto PVDF membrane (Millipore) with electrophoresis. Then, the PVDF membrane was blocked with TBS with 0.2% Tween 20 (TBS-T) containing 5% skim milk at 4°C overnight. To detect LOX1, STRA6, CRBP1, RAR, RAR, RXR, pJNK, JNK, p-p38MAPK, p38MAPK, pSmad2, Smad2, TGF 1 , caspase 3, or collagen 1 protein expression, the PVDF membrane was incubated with diluted primary antibodies in TBS-T containing 5% skim milk. After washing the membrane with TBS-T, the PVDF membrane was incubated with a 1:10,000 dilution of HRP-conjugated secondary antibody in TBS-T containing 5% skim milk. Western blots were detected by ECL detection kit (Millipore) to induce the chemiluminescence signal, which was captured by a luminescence imaging system.

RT-PCR analysis
Total RNA from the kidneys of saline-, L1-, or L5-injected mice, and L5-injected LOX1 / mice was extracted with Trizol reagent (Invitrogen). Total RNA was synthesized to first-strand cDNA by employing the AccessQuick RT-PCR system (Promega, Madison, WI). Then, the first-strand cDNA was processed to amplify cDNA fragments of STRA6, CRBP1, RAR, or RXR mRNA with the PCR core kit (Invitrogen). The primers of mouse STRA6, CRBP1, RAR, and RXR mRNA were purchased from Santa Cruz Biotechnology Inc. The cDNA fragments of PCR product were electrophoresed on agarose gels in a 100 V constant voltage field. Gels were photographed by using a gel 1000 UV documentation system and analyzed by densitometry. All mRNA levels were normalized by the corresponding actin mRNA level. +L5) after injection for 4 weeks. A: Western blots showed that L5 reduced STRA6, CRBP1, RAR, RAR, and RXR, but increased LOX1, pJNK, p-p38MAPK (p-p38), TGF 1 , pSmad2, caspase 3 (Casp3), and collagen 1 (Col1) in L5-injected mice. In LOX1 / mice, these changes caused by L5 were diminished. B: Bar graphs show that LOX1 increased and STRA6, CRBP1, RAR, RAR, and RXR decreased in L5-injected mice. These changes caused by L5 were attenuated in L5-injected LOX1 / mice. C: Bar graphs indicate that pJNK/JNK, p-p38/p38, and pSmad2/Smad2 ratios increased in L5-injected mice. These changes caused by L5 were attenuated in L5-injected LOX1 / mice. D: TGF 1 , caspase 3, and collagen 1 levels of L5-injected mice increased. These changes caused by L5 were recovered in L5injected LOX1 / mice. E: RT-PCR analysis shows that L5 decreased PCR products of STRA6, CRBP1, RAR, and RXR in L5-injected mice, but there was no effect in L5-injected LOX1 / mice. All results are represented as mean ± SE; *P < 0.05 versus Ctl and L1; # P < 0.05 versus L5.
anti-STRA6 antibody (ABGENT), anti-collagen 1 antibody (Santa Cruz Biotechnology Inc.), or anti-vitamin A antibody (MyBioSource, San Diego, CA) diluted with antibody diluent (Dako, Carpentaria, CA). After washing in PBS, endogenous peroxidase activity was blocked by incubation in 0.3% H 2 O 2 in methanol for 20 min, followed by sequential 10 min incubations with biotinylated link antibody and peroxidase-labeled streptavidin (Dako). Staining was completed after incubation with 3,3'-diaminobenzidine substrate-chromogen solution (Dako), and then counterstained with hematoxylin. Images from similar regions of sections of kidney were captured by bright field microscopy at 400× microscopic magnification.

Hydroxyproline analysis
Protein samples of kidneys of saline-, L1-, L5-injected mice, and L5-injected LOX1 / mice were hydrolyzed by 6 M HCl at 110°C for 18 h, and then centrifuged at 18,000 g for 2 min, and finally dried by freezing. Hydroxyproline of renal samples was analyzed with a hydroxyproline analysis kit (Sigma-Aldrich).

Immunofluorescence
Cells were plated in eight-chamber glass slides. After treatment, cells were fixed in 4% paraformaldehyde. After PBS washing, blocking reagent (Dako) was used to block nonspecific background staining for 24 h. After PBS washing, cells were incubated with the anti-STRA6 antibody (Santa Cruz Biotechnology Inc.) at 4°C overnight. After PBS washing, TRITC-or FITC-conjugated secondary antibody (Lonza, Walkersville, MD) was added to cells. Cell membranes were counterstained with green fluorescence-membrane stain (Invitrogen) or nuclei stain, DAPI (Lonza). Under a fluorescence microscope (400×), the red fluorescence images of STRA6 on cell membranes were visualized.

Histochemistry
After deparaffinization and rehydration, kidney sections of saline-, L1-, or L5-injected mice, or L5-injected LOX1 / mice were placed in 0.01 M sodium citrate buffer (pH 6.0) and heated in a microwave oven for 2.5 min at 720 W. For Mason's trichrome stain, sections were stained according to the protocol of the manufacturer (Sigma-Aldrich). For IHC, sections were washed in PBS and incubated with 1% BSA for 30 min to block nonspecific staining. Sections were drained and incubated for 3 h at room temperature in a humidity chamber with respective antibody for IHC, including D: Quantitative analysis showed that STRA6 immunostaining intensity increased but collagen 1 intensity decreased in the PCT, CCD, MTAL, and IMCD in L5injected mice, as compared with Ctl and L1 mice or L5-injected LOX1 / mice. Intensity of trichrome stain increased in the cortex, outer medulla (OM), and inner medulla (IM) of L5-injected mice. E: Bar graph shows that hydroxyproline concentration increased in the kidneys of L5-injected mice. All results are represented as mean ± SE. *P < 0.05 versus Ctl and L1; # P < 0.05 versus L5.  Terminal transferase-mediated deoxyuridine triphosphate nick end-labeling assay Cells were plated in eight-chamber glass slides. After treatment of L1 (50 g/ml) or L5 (50 g/ml) for 24 h, cells were fixed in 4% paraformaldehyde and then analyzed by using an in situ apoptosis detection kit (R&D Systems, Minneapolis, MN) in accordance with the manual of the manufacturer. In the immunofluorescence image of HK-2 cells, deoxyuridine triphosphate nick ends were counterstained with FITC and nuclei were counterstained with DAPI. The incorporation of apoptotic cells (green fluorescence) was visualized under fluorescence microscope (400×). In the histochemistry image of RTECs, deoxyuridine triphosphate nick ends were counterstained with HRP-developed 3,3'-diaminobenzidine solution and cells were counterstained with hematoxylin. The incorporation of apoptotic cells (deep brown) was visualized under optical microscope (200×). To quantify the apoptotic cells, cells were counted for the percentage of positive cells with image analysis software in each well. All scoring was performed blinded to the calculator on coded slides. mean ± SE. Differences in measured variables between experimental and control groups were assessed by one-way ANOVA with Bonferroni test. Probability values of P < 0.05 were considered significant.

ELISA
Analysis of TGF- 1 ELISA. Cells were grown in plates for experiments. After treatment, culture medium of cells was collected to measure TGF 1 concentrations with TGF 1 Emax® Immuno-Assay Systems (Promega) for HK-2 cells or LEGEND MAX™ Total TGF- 1 ELISA kit (BioLegend, San Diego, CA) for RTECs.

Analysis of vitamin A and RA ELISA. Kidneys of saline-, L1-, and L5-injected mice, and L5-injected LOX1
/ mice were homogenized and processed with tissue lysis buffer (Promega), and then lysates of kidney were collected to measure vitamin A or RA concentrations with ELISA kits. HK-2 and RETC cell experiments were collected and processed with cell lysis buffer (Promega), and then cell lysates were collected to measure concentrations of vitamin A or RA with ELISA kits. The vitamin A and RA ELISA kits were both purchased from MyBioSource.

DISCUSSION
In the current study, we indicate that L5 treatment can cause renal apoptosis and fibrosis via LOX1 in L5-injected mice and L5-treated HK-2 cells, while L1 treatment cannot. Especially, this study first reveals the mechanism that L5 simultaneously suppresses the STRA6/CRBP1/retinol/RA/RARs/RXR cascade via activating the LOX1/ JNK pathway (Fig. 8). Particularly, we furthermore show that the suppression of STRA6 cascades is involved in renal apoptosis and fibrosis caused by L5 treatment.
oxLDL has been commonly utilized to provide direct evidence for the causal effect of dyslipoproteinemia-related kidney damage. However, oxLDL was artificially produced by oxLDL in the presence of CuSO 4 in most investigations (9)(10)(11)(12)(13)(14). Here we showed that native L5 injection could markedly increase caspase 3, collagen 1 in kidneys of C57B6/J mice, but not in LOX1 / mice. Additionally, L5 treatment also increased these proteins and apoptotic cells in HK-2 cells, but not in LOX1 siRNA HK-2 cells. The immunostaining intensity of collagen 1 and trichrome in the kidneys of L5-injected mice was strongly expressed. Altogether, this study provides new evidence that dyslipoproteinemia can cause renal apoptosis and fibrosis by utilizing native L5.
L5 is not recognized by the normal LDL receptor, but is internalized by LOX1. LOX1 is one of several receptors for oxLDL, and its activation can result in oxLDL-induced SB203580 (supplemental Fig. S1A-H) in L5-treated RTECs. In optical microscopy images of terminal transferasemediated deoxyuridine triphosphate nick end-labeling analysis, L5-induced increase of apoptotic cell nuclei (thick dark stain) was recovered by SP600125 and by SB203580 (Fig. 4I, supplemental Fig. S1I). Higher TGF 1 concentration in cultured medium of L5-treated RTECs was reversed by the addition of SP600125 (Fig. 4J) and SB203580 (supplemental Fig. S1J). In transcription level, decreases of STRA6, CRBP1, RAR, and RXR mRNA were also reversed by SP600125 (Fig. 4K).

CRBP1 transfection reverses the suppression of STRA6 cascades and cell damage in L5-treated renal epithelial cells
Because STRA6-mediated retinol transport requires the presence of CRBP1, and we established stable crbp1 genetransfected HK-2 cells to investigate whether crbp1 transfection could affect STRA6 cascades in cells cultured with native L1 and L5. We found that crbp1 gene transfection could significantly reverse L5 treatment-induced increase of LOX1, L5-induced suppression of STRA6, CRBP1, RAR, and RXR, or L5-induced increases of pJNK, pSmad2, caspase 3, and collagen 1 expression in L5-stimulated HK-2 cells (Fig. 5A-J). Elevated TGF 1 concentration in cultured medium of L5-treated HK-2 cells was also suppressed by crbp1 transfection (Fig. 5K). The decrease of STRA6 immunofluorescence staining (green) was also recovered by crbp1 transfection in L5-cultured HK-2 cells (Fig. 5L).

Retinoid concentration is reduced by L5 treatment in vivo and in vitro
Immunohistochemistry by using anti-vitamin A antibody showed a considerable decrease of vitamin A staining density in the renal cortex (Fig. 6A), outer medulla (Fig. 6B), and inner medulla (Fig. 6C) of L5-injected mice in comparison with saline-injected, L1-injected, and L5-injected LOX1 / mice. Vitamin A and RA concentrations were diminished in the kidneys of L5-injected mice, as compared with saline-injected, L1-injected, and L5-injected LOX1 / mice (Fig. 6D). LOX1 siRNA transfection reversed the L5induced decrease of vitamin A and RA concentrations in renal cells (Fig. 6E). SP600125 (Fig. 6F) and SB203580 (supplemental Fig. S1K) could recover the reduction of vitamin A and RA concentration in L5-treated RTECs.

STRA6 RNA silence decreases STRA6 cascades and causes cell damage
To investigate whether suppression of STRA6 cascades was involved in kidney damage, we silenced STRA6 expression in HK-2 cells with siRNA. The efficiency of dosedependent STRA6 siRNA was adapted to significantly block STRA6 mRNA and protein expression in HK-2 cells. In this experiment, STRA6 siRNA transfection significantly increased LOX1, suppressed CRBP1 and RAR, but increased pJNK/JNK, pSmad2/Smad2, and collagen 1 expression in PBS-treated and L1-treated HK-2 cells (Fig. 7A,  C-I). It markedly increased TGF 1 concentration in cultured medium of PBS-treated and L1-treated HK-2 cells (Fig. 7J). Furthermore, silencing stra6 also increased Fig. 8. The schematic mechanism of L5 on STRA6 cascade suppression and kidney injury. L5-activated LOX1/JNK and p38MAPK pathway suppresses STRA6/CRBP1/retinol/RA/RARs/RXRs cascade and, thereafter, contributes to apoptosis and fibrosis of the kidney. RXR and then significantly contribute to high glucoseinduced cardiomyocyte apoptosis (39,40). In this study, JNK and p38MAPK phosphorylation was activated in the kidneys of L5-injected mice and L5-treated renal cells, but was not activated in LOX1 knockdown models in vivo and vitro. JNK and p38MAPK inhibitors can reverse the suppression of the STRA6/CRBP1/retinol/RA/RAR/RXR cascade, as well as the induction of renal apoptosis and fibrosis. These results indicate that JNK and p38MAPK signaling is activated by L5 to process kidney apoptosis and fibrosis after the activation of LOX1.
In summary, we demonstrate that electronegative L5 can cause kidney apoptosis and fibrosis and simultaneously suppress STRA6, CRBP1, retinol, RA, RARs, and RXR. The alteration of STRA6 cascades is involved in kidney apoptosis and fibrosis caused by L5 treatment. These findings implicate that the suppression of STRA6 signaling may participate in dyslipidemia-mediated kidney disease. endothelial dysfunction (37,38). In our previous studies, L5 induced endothelial cell apoptosis and platelet aggregation in a LOX1-dependent manner (16,17). LOX1 overexpression is also reported in renal tubular cells and capillaries in the kidneys of rats with dyslipidemia and diabetes (9,10). Intravenous injection of LOX-1 antibody could prevent the renal injury in these rats (10). Consistently, this study furthermore shows overexpression of LOX1 in kidneys of native L5-injected mice and L5-treated HK-2 cells. Additionally, LOX1 gene knockdown and LOX1 RNA silencing can attenuate renal apoptosis and fibrosis induced by L5 in mice and renal cells. These results elucidate that native L5 causes kidney damage via LOX1.
Although RA and RAR agonist have been demonstrated to provide protection in several experimental models of kidney disease (31)(32)(33)(34)(35), there are very few studies to report the change of STRA6/CRBP1 cascades in kidney diseases. In mice with Adriamycin nephropathy, it was demonstrated that albumin prevented podocyte differentiation from human renal progenitors in vitro by sequestering RA (34). A great reduction in RA and RAR levels was reported in the glomeruli of human immunodeficiency-associated kidney disease (35). Reduced retinol dehydrogenase 1 and 9 levels were also observed in human immunodeficiency transgenic mice (35,36). In this study, retinol and RA values in the kidneys of L5-injected mice and in L5-treated renal cells considerably decreased in comparison with L1-treated mice and renal cells. To our knowledge, this study is the first to indicate that native L5 can remarkably reduce STRA6, CRBP1, retinol, RA, RAR, and RXR expression in kidneys of mice and in renal cells. LOX1 knockdown and LOX1 RNA silencing reversed these alterations. Additionally, immunostaining intensity of STRA6 was markedly repressed, while collagen 1 immunostaining and trichrome stain were reciprocally increased by L5 injection in renal tubules and collect ducts of L5-injected mice. These changes were attenuated in L5-injected LOX1 / mice. Therefore, these data reveal that retinoid homeostasis in the kidney is suppressed by L5 treatment via repressing STRA6/CRBP1/retinol/RA/RARs signaling in the pathological process of kidney disease. STRA6-mediated retinol transport into cells requires CRBP1, an intracellular retinol acceptor (25,29). To investigate whether the suppression of STRA6 and CRBP1 can activate fibrotic process and apoptosis, we performed crbp1 gene transfection and STRA6 siRNA transfection. Notably, transfection of the crbp1 gene could significantly reverse increases of fibrotic and apoptotic proteins, while it concurrently reversed the suppression of STRA6/RAR/ RXR in L5-stimulated HK-2 cells. Conversely, STRA6 siRNA transfection significantly increased pJNK, pSmad2, collagen 1, and TGF 1 levels and apoptotic cells, whereas it negatively regulated CRBP1, RAR, and RXR in PBStreated and L1-treated HK-2 cells. These results reveal that L5-mediated suppression of STRA6/CRBP1/retinol/RA/ RARs/RXR cascades participates in L5-mediated kidney apoptosis and fibrosis.
Some studies propose that the high glucose-activated JNK pathway could suppress the expression of RAR and