Triptolide contributes to decrease in TLR4 expression by upregulating miR-224-3p to inhibit the inflammatory reaction in diabetic nephropathy

Triptolide has been shown to have anti-inflammatory properties. miRNAs have emerged as important regulators of DN. However, the role of miRNAs in inflammation regulation during DN development is poorly understood. In this study, we investigated the effects of the association between miRNAs and triptolide on inflammation in DN. We found that triptolide significantly inhibits the inflammatory reactions of DN in rats, and miR-224-3p is selectively upregulated more than 2 fold in NRK-52E cells treated with triptolide compared to the cells without treatment. miR-224-3p overexpression exerts similar effects to those of triptolide, while miR-224-3p downregulation increases all the inflammatory reactions above. In summary, we found that triptolide attenuates the inflammatory reactions in DN by upregulating miR-224-3p expression.


INTRODUCTION
Diabetic nephropathy (DN) has become an important cause of end-stage renal disease and death in diabetic patients [1 , 2] . Approximate 40% of patients with type 2 diabetes (T2D) will develop DN and even renal failure [3]. However, the pathogenesis of DN is complex, and inflammation theory and immune system activation have attracted the interest of scientists worldwide [4] .
TLRs represent the receptors of the innate immune system [5]. TLRs activate downstream inflammatory response, such as the NF-κB pathway, which initiates the acquired immune response [6]. TLR4 activation promotes tubulointerstitial inflammation and fibrosis, which deteriorated the renal function of DN [7].
Unfortunately, there are no effective treatments for DN [8 -10]. In recent years, treatment of DN with traditional Chinese medicines has achieved positive results [11 , 12]. Tripterygium glycosides can promote the production of anti-inflammatory mediators in many cells. Triptolide is the most active ingredient of tripterygium glycosides [13 -15]. Our previous study showed that triptolide could ameliorate tubulointerstitial fibrosis via TLR4/NF-κB [16]. miRNAs bind specific genes and facilitate mRNA cleavage or mRNA translation inhibition and thus participate in a variety of physiological and pathological processes [17]. Recent studies have shown that miRNAs correlate with DN [18 -20].
We studied the mechanism underlying the antiinflammatory effects of triptolide via biochip technology in DN rats. Triptolide attenuated inflammation by upregulating miR-224-3p expression and the plasma levels of miR-224-3p were significantly lower in the patients with DN.

R E T R A C
T E D www.impactjournals.com/oncotarget

Effects of triptolide treatment on blood biochemical parameters
The blood biochemical parameters were showed in Table 1. During the study period, the levels of blood glucose, TC and TG were significantly higher in both DN and TP-treated rats compared with the NC rats (P < 0.05). However, there were no significant differences between DN and TP-treated groups (P > 0.05), indicating that TP had no effects on blood glucose and blood lipid. In addition, the AST and ALT levels did not differ among the five groups (P > 0.05). Furthermore, Scr and BUN did not change in these groups by the end of the experiment (P > 0.05).

Effect of triptolide on urinary protein, NAG and KW/BW
The 24 h urine protein was significantly increased in the DN rats compared with that in the NC rats (P < 0.05), which was attenuated after TP administration for 8 weeks (P < 0.05). NAG, one of the urinary tubular injury biomarkers, was significantly increased in the DN rats compared with that in the NC rats, but was markedly reduced by TP treatment (P < 0.05). The kidney weight to body weight ratio (KW/BW) was significantly higher in the DN rats compared with the KW/BW in the NC rats (P < 0.05). By contrast, KW/BW was significantly reduced in the TP treatment group (P < 0.05), but still remained significantly higher in comparison with normal control (Table 1).

Triptolide attenuated renal histopathological changes in DN rats
The renal tubules in DN rats exhibited signs of atrophy and luminal dilatation, as well as evidence of inflammatory cell infiltration in HE staining, compared to those in the NC rats. In the triptolide group, the degree of inflammatory cell infiltration was significantly reduced ( Figure 1A). Masson staining results indicated that triptolide attenuates renal tubulointerstitial fibrosis ( Figure  1B).

Triptolide improved renal inflammation in DN rats
The expression levels of TLR4, NF-κB, TGF-β1 and ICAM-1 were significantly increased in the DN group, compared with those in the NC group. After treated with triptolide, the situation was significantly improved (

Triptolide attenuated high glucose-induced inflammation in NRK-52E cells
The mRNA and protein expression levels of TLR4, NF-κB, TGF-β1 and ICAM-1 were significantly higher in NRK-52E cells stimulated with high glucose concentrations for 48 h than those in NC cells. Following 48h triptolide, cytokines above were significantly decreased ( Figure 3). observed. Red indicates upregulated miRNAs (higher expression), and green indicates downregulated miRNAs (lower expression); the aggregation tree is above the sample, and the miRNA aggregation tree is on the right. B. Statistical analysis: differential expression of miR-224-3p in each group, according to the aggregation tree. C. Differential expression of miR-224-3p in each group. The level of miR-224-3p expression is expressed relative to that of U6 snRNA. NG, NRK-52E cells treated with normal glucose (5.56 mmol/L); HG, NRK-52E cells treated with high glucose (30 mmol/L); HG+TP, NRK-52E cells treated with high glucose + 5 ng/mL triptolide. * p < 0.05 versus NG group, # p < 0.05 versus HG group.

Effects of Triptolide on expression of miR-224-3p
miRNA microarray analysis showed that miR-224-3p expression was significantly decreased in the HG group, while returned to normal level in the HG+TP group ( Figure 4A, 4B). We subsequently increased our sample size and utilized qPCR to verify the differences in miR-224-3p expression in each group. Our results were consistent with the miRNA microarray analysis data ( Figure 4C).

miR-224-3p binding and suppression of TLR-4 expression in NRK-52E cells
To elucidate the molecular mechanisms underlying the anti-inflammatory effects of miR-224-3p, we next attempted to identify potential inflammatory pathway targets. TargetScan analysis predicted one putative miR-224-3p binding site within the 3'UTR of TLR4 ( Figure  7A). To determine if miR-224-3p effectively inhibits TLR-4 expression through its interaction with the TLR4 3'UTR, we utilized a dual-luciferase reporter system. These results confirmed that TLR4 is a direct target of miR-224-3p and miR-224-3p can reduce the expression of TLR-4 in NRK-52E cells ( Figure 7B).

The plasma miR-224-3p level was significantly decreased in DN patients
Interestingly, we also detected the plasma levels of miR-224-3p via qPCR in three groups (NC, DM, and DN). The plasma level of miR-224-3p expression in the DM patients was similar to that of NC. However, the level of miR-224-3p in the DN patients was significantly lower.

DISCUSSION
DN is a type of glomerular disease mainly characterized by vascular damage. In recent years, renal tubular lesions and interstitial fibrosis have been found to be unique in terms of pathogenesis [21] . We found that renal tubules of DN rats exhibited pathological changes, such as partial epithelial cell brush border detachment,

R E T R
A C T E D

R E T R
A C T E D as well as vacuolization and granular degeneration. Inflammatory cell infiltration was also noted.
More and more researchers have concluded that DN is a type of autoimmune disease [22]. Renal cells in the settings of hyperglycemia and hemodynamic alterations can produce many types of inflammatory cytokines, which trigger an inflammatory cascade reaction [23], resulting in the development of DN.
TLRs are important factors in inflammatory reactions, served as a bridge linking innate and acquired immunity [24]. TLR4 expressed on the surface of a variety of cells is one of the best studied TLR subtypes. Recently, TLR4 was reported to express in renal tubular epithelial cells in diabetic rats. TLR4 activation induces production of cytokines and chemokines and participates in various immune and inflammatory reactions [25 , 26]. TLR4 activates the NF-κB signaling pathway, resulting in inflammatory factor released. After NF-κB activation, IL-1β, TGF-β1 and TNF-α transcriptions and translations are initiated, which then activates the Smad signaling pathway. NF-κB also stimulates the transcription of several fiber-related factors, such as ICAM-1 and VCAM-1 [27], and promotes the proliferation and differentiation of renal tubular cells.
Triptolide has strong anti-inflammatory effects. In line with our study, Triptolide inhibits the expression of various cytokines and adhesion molecules, which inhibits the activation of inflammatory cells, such as neutrophils and macrophages, thereby inhibiting inflammation [28 , 29]. In this study, Triptolide reduces renal tubular epithelial cell vacuolization and granular degeneration, as well as inflammatory cell infiltration, and improves renal tubular expansion and contraction via downregulation of the expression of TLR4, NF-κB, TGF-β1 and ICAM-1 in vivo and in vitro. This might be one of the mechanisms underlying the effects of triptolide on renal tubular interstitial injury.
miRNAs play an important role in cell proliferation, growth, differentiation and apoptosis [30 , 31]. Several important miRNAs were found to highly express in  [32 -34]. Abnormal miRNA expression may lead to the development of renal disease. In this study, we analyzed the levels of miRNA expression using miRNA microarray technology. Our results showed that miR-224-3p expression was significantly downregulated in the high-glucose group, while triptolide reversed the inhibition of miR-224-3p. After the cells were transfected with miR-224-3p mimics, the expressions of TLR4, NF-κB, TGF-β1, ICAM-1 and other inflammatory factors were decreased significantly, indicating that miR-224-3p mediated the decrease in inflammation in DN by triptolide.

Animals
SD male rats were purchased from HFK Bioscience Company (Beijing, China). All experimental rats were maintained under specific pathogen-free conditions in the animal facility of Tianjin Medical University (China) and used for experimentation at 6 wk. The Tianjin Medical University Experimental Animal Ethics Committee approved all animal procedures.
Rats were randomly assigned to 3 groups. The rats were fed a standard diet (8% fat) or a high-fat diet (HFD) containing 40% fat for 8 weeks. The HFD animals were intraperitoneally injected with a single dose of STZ (30 mg/kg) in 50 mM citrate buffer (pH 4.0), while the control animals received an equal volume of citrate buffer. The hyperglycemia noted in the HFD rats after STZ injection was not due to declining beta-cell function [35]. Animals with blood glucose levels ≥ 16.7 mmol/L were deemed diabetes. 24h UMA levels ≥30 mg were considered to

Gene chip technology analysis
Total RNA was isolated using TRIzol (Invitrogen, USA) and purified with an RNeasy mini kit (QIAGEN, GER), according to manufacturer's instructions. Shanghai Biotechnology Co., Ltd conducted the MicroRNA microarray gene expression experiments and data analysis.

Transfection of the miRNA mimic and inhibitors
An miR224-3p mimic, miR224-3p inhibitor, and appropriate negative controls (NC) of the miRNA inhibitor (miR-iNC) and miRNA mimic (miR-NC) were purchased from GenePharma (Sunzhou, China) and transfected at final concentrations ranging from 50-100 nM using Lipofectamine 2000 transfection reagent (Invitrogen, USA), in accordance with the manufacturer's recommendations. The transfected cells were growtharrested via overnight incubation with serum-free medium prior to incubation with 25 mM glucose for 48h.

Luciferase reporter assay for miR224-3p binding the TLR4 3′-UTR
A luciferase reporter assay was conducted using a Dual-Luciferase Reporter Assay System (Promega, USA), according to the manufacturer's instructions. 293T cells (1.5×10 4 /well) were seeded into 96-well plates. After 12 h, the cells were transiently co-transfected with 0.1 μg/ well pRL-TK plasmid (Promega) containing a Renilla luciferase gene for internal normalization and various constructs containing pMIR-TLR4 and pMIR-TLR4-mut. Cells were lysed and assayed for luciferase activity 36h after transfection. One hundred microliters of protein extracts were analyzed in a luminometer. All experiments were performed at least three times.

Histological analysis
Kidney tissues were fixed in formalin, embedded in paraffin, and cut into 5 μm thick sections. The sections were stained with a hematoxylin and eosin (HE) staining kit (Baibo Biotechnology Co., Ltd., Shandong, China), periodic acid-Schiff (PAS) staining kit and Masson's trichrome (MT) staining kit (Nanjing Jiancheng Bioengineering Institute, Jiangsu, China) according to the manufacturer's protocols. For each slide, 10 randomly selected non-overlapping fields of renal cortex without glomeruli and large vessels were measured, and the averages for each group were analyzed.

Immunohistochemistry staining of renal tissues
Formalin-fixed, paraffin-embedded sections (4μm thick) were deparaffinized by xylene and hydrated by graded ethanol. Heat-induced antigen retrieval was conducted at 95 o C by microwave in 10 mmol/L sodium citrate buffer (pH 6.0) for 5 min. Then, the sections were blocked with 5% BSA for 30 min and incubated overnight at 4°C with the following antibodies diluted with PBS: anti-NF-κB (Cell Signaling Technology, 1:100), anti-TGF-β1 (Abcam, 1:200), anti-ICAM-1 (Abcam, 1:100) and anti-VCAM-1 (Abcam, 1:100). After washed three times in PBS, the sections were incubated for 45 min with the appropriate secondary antibody. The sections were subsequently visualized with a Diaminobenzidine (DAB) kit and counterstained with hematoxylin.

RNA isolation and analysis with qPCR
Total RNA was isolated by using Trizol reagent (Invitrogen, Carlsbad, CA, USA). This total RNA (1 μg) was then used for cDNA synthesis with a Transcriptor First Stand cDNA Synthesis Kit (Roche, IN, USA). Realtime quantitative PCR (qPCR) was performed using a quantitative SYBR Green PCR kit (TaKaRa, Dalian, China). The SYBR Green fluorescence absorption values in each tube were detected at the end of each cycle. miRNA and RNA expression levels were normalized to those of a specific small nuclear RNA (RNU6) or housekeeping gene (GAPDH). All primers are listed in Table 2.

Patients and specimens
The participants were recruited in our hospital from January to July 2015 in Tianjin, China (Table 3). DN was diagnosed based on a urinary albumin excretion ratio greater than or equal to 30 mg/24 h, a serum creatinine greater than 167 μmol/L or a blood urea nitrogen greater than 7.8 mmol/L. Individuals were excluded if they were pregnant or lactating, had type 1 diabetes or another type of diabetes, had another secondary renal disease or renal dysfunction caused by primary renal disease or hypertension, had active liver disease or abnormal liver function, had a urinary tract infection or had severe hypertension. Serum and urine samples were collected from 12 patients. The control group is comprised of 12 healthy subjects.

Statistical analysis
All data are expressed as the mean ± standard error (SE). Comparisons between two parameters were analyzed by Student's unpaired t-test. P < 0.05 was considered statistically significant. Between-group differences in latency were analyzed via analysis of variance for repeated measurements (ANOVAR), followed by Fisher's post hoc test for all groups.