The reno-protective effect of metformin against type 2 diabetic rats via up-regulating renal tissue pigment epithelium-derived factor expression

. Background. Diabetic nephropathy has become the primary cause of end-stage renal disease worldwide. Pigment epithelium-derived factor (PEDF) is an endogenous anti-inflammatory factor in vivo, which can inhibit the expression of pathogenic factors — transforming growth factor β 1 and connective tissue growth factor, and suppresses extracellular matrix protein production in diabetic kidney, suggesting an antifibrogenic activity. The aim of this study was to investigate the effect of metformin on renal tissue PEDF expression in type 2 diabetic rats and explore its possible underlying protective mechanisms in renal injury. Materials and methods. Ten clean male Sprague-Dawley rats were randomly selected as the normal control group. Type 2 diabetes model were induced by a high-fat diet and intraperitoneal injection of 30 mg/kg streptozotocin. A total of 30 type 2 diabetic rats were divided into 3 groups, which were treated with metformin 300 mg/kg/day (MET group, n = 10), glibenclamide 5 mg/kg/ day (GLY group, n = 10), or saline (DM group, n = 10) by gavage for 8 weeks. Various biochemical parameters, kidney histopathology and renal tissue PEDF expression levels were examined. Results . At the 8 th week, fasting blood glucose, glycated hemoglobin, triglyceride levels, urinary albumin and PEDF excretion, serum creatinine and blood urea nitrogen in MET group and GLY group decreased significantly compared to DM group, there were no significant differences in fasting blood glucose and glycated hemoglobin between MET group and GLY group. The histological examinations revealed amelioration of diabetes-induced glomerular pathological changes following treatment with MET and GLY when compared to DM group. In addition, urinary albumin and PEDF excretion were decreased, glomerular pathological changes was lightened and protein and mRNA expression of renal tissue PEDF were increased more in MET group compared with GLY group. Conclusions. Metformin reduced urinary albumin excretion in diabetic rats, and improved podocyte morphology and structural damage. The mechanism may be partly related to its role in restoring PEDF expression and inhibiting urinary excretion of PEDF.


Introduction
Diabetic nephropathy (DN) has become the primary cause of end-stage renal disease worldwide [1]. Pigment epithelium-derived factor (PEDF) is an endogenous anti-inflammatory factor in vivo [2], which can inhibit the expression of pathogenic factors TGF-β 1 and CTGF, and suppresses ECM protein production in diabetic kidney, suggesting an antifibrogenic activity [3].
In this study, we aim to observe its reno-protective effect and possible underlying mechanisms of metformin in diabetic rats so as to provide an experimental basic for the use of metformin in the clinical prevention and treatment of diabetic nephropathy.

Establishment of the experimental mode
We followed the experimental methods of Ye et al. (2015) [9]. Forty-four male SD rats (180-200 g, SPF grade) aged 8 weeks were purchased from the Experimental Animal Center of Anhui Medical University (Anhui, China). All the animals had free access to water under condition of relative humidity of 40-60 % and temperature of 20-30, and were maintained in a 12/12 h lightdark natural cycle during the experiment period. SD rats were fed a high-fat diet (basal diet +10 % lard compound +2 % cholesterol, 50 % calories from fat) for 4 weeks followed by a single intraperitoneal injection of a low dose of streptozotocin (STZ, 30 mg/kg in 10 mmol/L citrate buffer, pH 4.4) after an overnight fast. Normal control rats (n = 10) received an injection the same volume of citrate buffer alone. Only the animals with glucose levels higher than 16.7 mmol/L were considered diabetic and selected for further studies. One week later, diabetic rats were randomly assigned into three groups (n = 10): group DM treated with metformin 300 mg/(kg • d), group GLY injected with glyburide 5 mg/(kg • d) for 8 weeks.
Care, use, and treatment of all animals in this study were in strict agreement with the rules in the care and use of laboratory animals set forth by Anhui Provincial Hospital Affiliated to Anhui Medical University. Every effort was made to reduce the number of animals used and their suffering.

Biochemical detections
Biochemical detections were approached as described previously [9]. Urine samples were collected from the rats housed in metabolic cages for 24 h to measure urinary albumin (Ualb), urinary sediment PEDF (Upedf), and urinary creatinine (Ucr). Ualb was measured by radioimmunoassay. Radioimmunoassay was used to test Ualb. Enzyme-linked immunosorbent assay was used to test Upedf. Ucr was tested by Jaffe's assay. Urinary albumin and urinary sediment PEDF were expressed as urinary Ualb/Ucr (UACR) and Upedf/Ucr (UPCR) to eliminate the impact of urine volume, respectively.
The blood sample was preserved at -80 after centrifugation for the detection of FBG, HbA1c, serum BUN, TG, and LDL-C. High-performance liquid chromatography was used to detect HbA1c. Insulin (INS) was detected by the radioimmunoassay method. BUN was detected by the urease method. The serum TG, LDL and FBG were detected by automatic microplate reader.

Electron microscopic observation
1 mm 3 renal cortex was fixed in 2.5 % glutaraldehyde to produce 4 μm thick sections. Ultrathin sections were observed by electron microscopy at a magnification of × 10,000. Three glomeruli were observed per section. Collected 10 photos from each glomerulus randomly and measure glomerular basement membrane (GBM) thickness, taking the average, was calculated as GBMT of the corresponding group.
Image processing and analysis program (Image Pro Plus 6.0) was used to assess the GBM and the part of fused foot process. Foot process fusion rate (FRFP) = the whole length of GBM/the grand total length of fused foot process covered in the corresponding GBM.FRFP was calculated and used to finally assess a mean FRFP per group [10].

Western blot analysis
Western blot analysis was performed as described previously [11]. Briefly, the renal tissue was lysed with the RIPA lysis buffer to obtain extracts of renal protein. Electrophoresis for the protein samples was performed on 10 % SDS-PAGE gels, and then the samples were transferred onto a polyvinylidene fluoride (PVDF) membrane for sealing for 2 h 50 ug of protein from each sample was blotted by an anti-PEDF antibody. The same PVDF membranes were stripped and incubated with an anti-β-actin antibody .Image acquisition system was used to obtain Images, the band intensities were quantified using Image J software.

Renal Tissue PEDF mRNA Expression
The levels of mRNA were analyzed in renal tissue using a real-time PCR approach as described previously [12]. Total RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. Primers specific for PEDF(forward5'-TC-GCATAGACCTTCAGG AGATTA-3',reverse5'-ATCAGAGTCCAAGCCATATCGTA-3'); Оригінальні дослідження /Original Researches/ PCR was performed using the reverse ranscription kit and SYBR Premix Ex Taq II. Results were normalized to the β-actin mRNA levels and represented using the comparative threshold cycle method. The cycle threshold (Ct) value of fluorescence units was used to analyze the mRNA levels by the 7500 software. The relative quantitative method (2 −ΔΔCt ) was used to calculate the relative changes in gene expression.

Statistical Analysis
Data processing was performed using SPSS 16.0 (SPSS Inc., USA). The results were calculated and expressed as group means ± SD. One-way analysis of variance was used to performed significant differences among groups and Student-Newman-Keuls test was used for comparison between individual groups. Individual differences among groups were analyzed by Dunnett's test. A p value of < 0.05 was considered statistically significant.

Effect of metformin on serum biochemical and urinary indices in diabetic rats
After 8 weeks, compared with NC group, the diabetic rats were showed a profound elevation in the level of FBG, HbA1c, TG, LDL-C, and BUN and INS (p < 0.05). All measured biochemical parameters above in MET group and GLY group decreased significantly (p < 0.05) as compared with group DM, there were no significant differences in FBG and HbA1c between group MET and group GLY. UACR and UPCR levels in DM group increased obviously compared with NC group. The indicators above in MET group and GLY group were significantly decreased compared to those in the DM group (p < 0.05), there were significantly decreased in group MET than those in group GLY (table 1).

Effect of metformin on renal histology in diabetic rats
The renal tissue obtained from NC group showed a normal mesangial matrix, the structure of glomerular podocytes remained clear, the foot processes showed no fusion. The glomerular basement membrane (GBM) was thickened diffusely and its the architecture became ambiguous, the foot processes were swol-len and fused significantly at the meanwhile, some foot processes were completely ruined and even disappeared in the DM group. Compare with DM group, both GBMT and FRFP of glomerular were alleviated in the MET and GLY groups (p < 0.05).Compared to GLY group, the pathological changes of glomerular were more alleviated in MET group (p < 0.05) (figure 1(a)).

Effect of metformin on the renal tissue PEDF protein expressions in diabetic rats
The PEDF protein expressions of the renal tissue in DM group decreased significantly compared to that in NC group (p < 0.05). After intervention with metformin or glimepiride, the expressions of PEDF protein in the renal tissue elevated significantly compared to those in the DM group, there were significantly increased in group MET than those in group GLY (p < 0.05) (figure 2(c)).

Effect of metformin on the renal tissue PEDF mRNA expressions in diabetic rats
As compared to that in NC group, the renal PEDF mRNA expressions was down-regulated significantly in DM group (p < 0.05). Administration of metformin or glyburide significantly inhibited renal PEDF mRNA expressions as compared to DM group, which was more prominent in group MET than GLY (p < 0.05) (figure 2(a) (b)).

Discussion
PEDF is a member of the serpin superfamily, which is a neurotrophic factor secreted by adipocytes, and it could block the development and progression of experimental diabetic nephropathy [2]. Initially identified in the conditioned medium of human retinal pigment epithelial cells, it is currently believed that PEDF is abundantly expressed in kidney, liver, and adipose tissue [13].
PEDF has been shown to be a highly effective inhibitor of angiogenesis in cell culture and animal models, not only reduces advanced glycation end product, angiotensin II or leptin-induced vascular endothelial growth factor (VEGF) expression, but also inhibits the biological actions of other factors, thus exerting an important  Оригінальні дослідження /Original Researches/ role in protecting blood vessels, inhibiting inflammation and oxidative stress [4,14,15]. It was reported that PEDF suppresses AGE-elicited endothelial cell da mage through its anti-oxidative properties and blocks the progression of experimental diabetic retinopathy (DR) [5,16]. However, the protective effect of PEDF in early experimental diabetic nephropathy has not been demonstrated in vivo. Joshua [3] provided the first evidence that a PEDF gene delivery ameliorates proteinuria in STZ-induced diabetic rats in vivo. AGE-induced RAGE gene expression, ROS generation, inflammatory and fibrogenic gene expression (MCP-1, TGF-β, fibronectin and type IV collagen mRNA levels) were significantly increased in diabetic kidney, which were suppressed by administration of PEDF, it suggests that PEDF could play a protective role against diabetic nephropathy by attenuating the deleterious effects of AGEs via the downregulation of RAGE expression [5,[18][19][20].
Previous study showed that metformin treatment increased PEDF expression in both prostate cancer cells and tumor tissue [21]. Turan et al demonstrated that metformin is effective in preventing ovarian hyperstimulation syndrome (OHSS) and its severity (VEGF expression and vascular permeability) by up-regulating PEDF expression [22]. Akın S. et al. reported that patients with newly diagnosed diabetes is associated with obvious increase in serum PEDF levels after metformin treatment [23]. In this study, the levels of serum BUN, UACR, FRFP and GBMT in DM rats were higher than those in group NC. These indexes were significantly decreased after metformin or glyburide treatment for 8 weeks, and the former is superior than the latter, but the value of FBG, HbA1c in MET group had no significant differences compared with GLY group, which indicated that metformin has a remarkable protective effect on the kidney independent of its hypoglycemic effect.
And the underlying mechanisms of metformin renoprotection are not clear completely, which may be related to its effects in activating AMP, activating protein kinase, inhibiting inflammation and oxidative stress, lightening lipid deposition, and so forth [25].
In present study, the UPCR, protein and mRNA expressions of PEDF in group DM were obviously decreased compared with group NC, which were upregulated by MET or GLY treatment, and that was greater in group MET than that in group GLY. It suggests that metformin can up-regulate the expressions of PEDF protein and mRNA in renal tissue of DM rats partially independent of hypoglycemia, which maybe also contribute to its reno-protection in some extent. At pre sent, the mechanism is not clear that metformin increasing the expression of PEDF in renal tissue, and there is no report about the direct effect of metformin on the expression and secretion of PEDF in diabetic kidney tissue [23].
In the early stage of diabetes, the expression of PEDF is decreased, hyperglycemia maybe is responsible for the decrease of PEDF in the diabetic kidneys [3,24]. Many studies have demonstrated that activating PPAR-γ signaling pathway played pivotal role in PEDF's beneficial effects [25][26][27][28][29]. PPAR-γ knockdown could reduce PEDF expression [30]. The decrease of PEDF was significantly down-regulated and the increase of NF-κB and MMP-9 was down-regulated by systemically administration of PPAR-γ agonist [31].
Our observations indicate that PEDF induces macrophage apoptosis and necrosis through the signaling of PPAR-γ In addition, metformin modulates CIS-induced hepatic damage by mitigating oxidative, and decreased caspase-3, MAPK activity and NF-kB level via a PPAR-γdependent pathway [40]. These are proposed that metformin increases the PEDF express by up-regulating PPAR-γ pathway, attenuates the deleterious effects of oxidative and fibrogenic, and exerting certain renal protective effects.

Conclusion
Taken together, metformin significantly alleviated proteinuria and renal podocyte lesions in diabetic rats induced by STZ, which may be due to that metformin increases the PEDF express by up-regulating PPAR-γ pathway, attenuates the deleterious effects of oxidative and fibrogenic, and exerting certain renal protective effects.

Conflicts of interests.
Authors declare the absence of any conflicts of interests and their own financial interest that might be construed to influence the results or interpretation of their manuscript.

Availability of data and materials
The data's and models used to support the findings of this study are available from the corresponding author upon request.