Bone Marrow Derived Mesenchymal Stromal Cells Ameliorate Ischemia/Reperfusion Injury-Induced Acute Kidney Injury in Rats via Secreting Tumor Necrosis Factor-Inducible Gene 6 Protein

Aims To investigate whether bone marrow derived mesenchymal stromal cells (BMSC) have ameliorated ischemia/reperfusion injury-induced acute kidney injury (IRI-AKI) via tumor necrosis factor-inducible gene 6 protein (TSG-6) and how TSG-6 exerted this effect. Methods We used lentiviral vectors of short hairpin RNA (shRNA) targeting TSG-6 gene to silence TSG-6 in BMSC. And TSG-6-silenced BMSC were administrated into IRI-AKI rats. Then we analyzed serum creatinine (Scr) and renal histology of IRI-AKI rats treated with BMSC after different pretreatments. Furthermore, we explored the effect of TSG-6 on renal tubular epithelial cells proliferation in vivo and in vitro assays. Results The Scr levels of IRI-AKI rats treated with BMSC (73.5±7.8 μmol/L) significantly decreased compared to those of IRI-AKI rats treated without BMSC (392.5±24.8 μmol/L, P<0.05) or with DMEM (314.0±19.8 μmol/L, P<0.05). Meanwhile, the renal tissue injury in IRI-AKI rats treated with BMSC improved markedly. However, the Scr levels of IRI-AKI rats treated with TSG-6-silenced BMSC (265.1±21.2 μmol/L) significantly increased compared to those with BMSC (74.0±8.5 μmol/L, P<0.05). The proportion of Ki67-positive cells was reduced in IRI-AKI rats treated with TSG-6-silenced BMSC compared to that in IRI-AKI rats treated with BMSC (29.7±0.8% versus 43.4±3.0%, P<0.05). In vitro, the cell proliferation rate of TSG-6-stimulated NRK-52E cells under hypoxia (89.2±3.9%) increased significantly compared to that of NRK-52E cells alone under hypoxia (82.4±0.8%, P<0.05). Similarly, the proportion of Ki67-positive cells was significantly elevated in TSG-6-stimulated NRK-52E cells under hypoxia. Furthermore, TSG-6 could inhibit infiltration of neutrophils in kidney tissue of IRI-AKI. Conclusions TSG-6 plays a key role in the treatment of IRI-AKI with BMSC, which may be due to its effect on promoting renal tubular epithelial cells proliferation by modulating inflammation.


Introduction
Acute kidney injury (AKI) is a common clinical syndrome characterized by renal dysfunction, which is caused by a spectrum of etiologies in multiple settings. AKI occurs in about 13.3 million people per year, 85% of whom live in the developing world. Moreover, AKI is thought to contribute to about 1.7 million deaths every year [1]. In China, the epidemiological data shows that at least 2.9 million people suffer from AKI, which need to be treated in hospital each year, and 0.7 million of them die. About two-thirds of the survivors will develop chronic kidney disease (CKD) [2]. AKI is the main contributory factor for poor prognosis, which not only increases the incidence of CKD and causes end-stage renal disease but also increases mortality in critically ill patients [3]. AKI has become a global health hot topic. However, the treatment options for patients with AKI are limited, mainly 2 BioMed Research International including renal replacement therapy or waiting for renal selfrepair.
The pathological changes of AKI are characterized by renal tubular injury and renal interstitial inflammation [4]. The main etiologies of AKI include ischemia/reperfusion injury (IRI), nephrotoxic drugs, and sepsis. The damaged renal tubular epithelial cells participate in the chemotaxis and activation of inflammatory cells by releasing cytokines, which further amplifies the inflammatory response in the kidney. Meanwhile, the inflammatory immune response plays an important part in the process of acute kidney injury and repair. Although the damaged renal tubular epithelial cells have the ability to repair and regenerate, this ability cannot meet the urgent need for renal recovery in most cases. Therefore, it has become a research focus to promote AKI repair in the field of nephrology.
In recent years, stem cell therapy has become a novel cellbased therapy for several inflammatory diseases. Experimental studies have shown that mesenchymal stromal cells (MSC) can promote AKI repair [5]. However, the research on MSC ameliorating AKI is mainly based on experimental animals. The clinical application of MSC is limited due to their unclear mechanisms. Several studies have suggested that MSC exert their effect on AKI via the paracrine mechanism [6]. In the damaged microenvironment, activated MSC can secrete a variety of anti-inflammatory factors to promote tissue repair, including tumor necrosis factor-inducible gene 6 protein (TSG-6), prostaglandin E2, and interleukin-1 receptor antagonist [7]. Among them, TSG-6 has displayed remarkable therapeutic effects in several models of acute organs injury [7]. However, there are very few studies on the effect of TSG-6 secreted by MSC in AKI.
In view of TSG-6 as a protective inflammatory response gene, we hypothesized that bone marrow derived MSC (BMSC) might exert their therapeutic effect by secreting TSG-6 in AKI. Then we tested our hypothesis in a series of in vivo and in vitro assays. The effect of TSG-6 was verified by administrating TSG-6 silenced-BMSC into IRI-induced AKI (IRI-AKI) rats and was then examined in rat renal tubular epithelial cells under hypoxia. Our results verified that TSG-6 was the key factor that allowed BMSC to treat IRI-AKI and we also discussed its possible mechanism.

Materials and Methods
. . Procedure and Protocol of IRI-AKI. Pathogen-free, adult male Sprague-Dawley (SD) rats (Shanghai Laboratory Animal Research Center, Shanghai, China) weighing 200±10 g were utilized in the present study. The protocol for the acute kidney ischemia/reperfusion procedure has been detailed in our previous reports [8]. Briefly, animals were anesthetized by sodium pentobarbital (40 mg/kg, intraperitoneally) and placed on a warming pad to maintain body temperature at 37 ∘ b for midline laparotomies. The sham control animals underwent laparotomy only. Acute IRI of both kidneys was induced in all IRI-AKI rats by clamping the renal pedicles for 45 min using nontraumatic vascular clips.
. . Rat BMSC Isolation and Identification. Rat BMSC were isolated and harvested as follows. Briefly, 3-to 4-week-old SD rats were sacrificed and soaked in 75% alcohol for 10 min. Under aseptic conditions, the femurs and tibias of SD rats were taken out and flushed with phosphate-buffered saline (PBS). By rinsing the bone marrow cavity, cell suspension was collected and cultured in 60 mm culture dish at 37 ∘ C in a humidified atmosphere of 5% CO 2 . The cell culture medium was Dulbecco modified Eagle's medium (DMEM) (Gibco, NY, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, NY, USA). The nonadherent cells were removed every 2 days and primary adherent cells were subcultivated 1:2 until the cells reached around 80% confluence. The typical markers (CD29, CD44, and CD90) of BMSC were detected in the cells of passage 3 by flow cytometry. Also, the cells were tested for their ability to differentiate into adipogenic, chondrogenic, and osteogenic lineages by a manufacturer of differentiation kits, including StemPro6 Adipogenesis Differentiation Kit (A1007001, Gibco, NY, USA), StemPro6 Chondrogenesis Differentiation Kit (A1007101, Gibco, NY, USA), and Stem-Pro6 Osteogenesis Differentiation Kit (A1007201, Gibco, NY, USA). The BMSC of passages 3-5 were used in animal experiments.
. . NRK-E Cells Culture and Grouping. NRK-52E cells, which were rat renal tubular epithelial cell line, were purchased from the cell bank of Chinese Academy of Sciences (Shanghai, China). Cells were cultured in DMEM (Gibco, NY, USA) supplemented with 5% FBS (Gibco, NY, USA). Cells were grown at 37 ∘ C in a humidified atmosphere with 5% CO 2 and changed with fresh growth medium every 2 days until confluence. Cells were isolated by trypsinization when near confluence. Serum-free medium with 150휇M cobalt dichloride (CoCl 2 ) was added to mimic hypoxic conditions [9]. The NRK-52E cells were divided into three groups (A, B, and C). In group A: control normoxic cells were maintained in normal atmosphere; in group B: serum-free medium with 150휇M CoCl 2 was added for a 48 h incubation after normoxic culture; in group C: serum-free medium with 150휇M CoCl 2 and 0.1휇g/ml recombinant human TSG-6 (rTSG-6, 2104-TS-050, R&D Systems) were added for a 48 h incubation after normoxic culture. Before the stimulation experiments, cells were growth-arrested in DMEM without FBS for 24 h. Each group was established in three holes and cultured in 6 times repeatedly.
. . Lentiviral Vectors of TSG-shRNA Silencing TSGin Cells. For RNA interference experiments, short hairpin RNA (shRNA) targeting rat TSG-6 gene (GenBank accession number: NM 053382.1) (TSG-6 shRNA) was designed and synthetized. Meanwhile, the scrambled control shRNA (sc-shRNA) was designed and synthetized. The sc-shRNA was nontargeting shRNA, which was demonstrated to have no homology to rat genes. The forward sequences of DNA corresponding to TSG-6 shRNAs and sc-shRNA were as follows: The lentiviral vectors of TSG-6 shRNAs were constructed by iCarTab Biomedical Inc. (Suzhou, China). The lentiviral vectors had reporter gene (green fluorescent protein, GFP). BMSC were transfected with TSG-6 shRNA or sc-shRNA with a commercial kit (iCarTab Biomedical Inc., Suzhou, China). Briefly, BMSC were seeded into 150 cm 2 flasks at the density of 3 × 10 6 /flask. Then, complete culture medium was added and cells were cultured overnight. Next, 100 휇l lentivirus was added to the 150 cm 2 flasks and blended with the pipette gently. The flasks were centrifuged at the speed of 800 g for 1 h. After that, the BMSC in flasks continued to be cultured in the incubator for 48 h. Under laser confocal microscope, the efficiencies of transfecting TSG-6 shRNA into BMSC were assessed. At last, shRNA-silenced BMSC were collected and total RNA was extracted. Quantitative real-time PCR was used to analyze the silencing efficiencies of TSG-6 in BMSC. The primer sequences of TSG-6 and GAPDH from the cDNA of BMSC in rats were as follows: GAPDH sense ACAGCAACAGGGTGGTGGAC GAPDH antisense TTTGAGGGTGCAGCGAACTT TSG-6 sense CCACGGCTTTGTAGGAAG-ATAC TSG-6 antisense GACGCATCACTCAGAAAC-TTCA TSG-6-silenced BMSC and BMSC transfected with sc-shRNA were passaged and cultured for animal experiments. Passages 3∼5 cells were used for the subsequent experiments.
. . Animal Grouping and Rationale for the erapeutic Regimen. The animals were equally categorized into groups (n=6/group): (1) sham group: laparotomy only without ischemia of bilateral renal pedicles; (2) IRI-AKI group: IRI-AKI rats without treatments; (3) DMEM group: IRI-AKI rats receiving 0.5 ml fresh DMEM by intravenous injection after ischemia; (4) BMSC group: IRI-AKI rats receiving the treatment of 5 × 10 6 [10] BMSC by intravenous injection after ischemia; (5) sc-shRNA BMSC group: IRI-AKI rats receiving the treatment of 5 × 10 6 BMSC transfected with sc-shRNA by intravenous injection after ischemia; (6) TSG-6 shRNA BMSC group: IRI-AKI rats receiving the treatment of 5 × 10 6 BMSC transfected with TSG-6 shRNA by intravenous injection after ischemia. Before injection, the BMSC were resuspended in 0.5 ml fresh DMEM. The kidneys were harvested at the designated time after the surgery. Blood was collected from the inferior vena cava for measurements of serum creatinine (Scr). Scr was determined quantitatively with picric acid method on Beckman Automatic Chemistry Analyzers according to the manufacturer's instructions. All reagents used were supplied by Beckman Inc.
. . Immunohistochemistry and Immunocytochemistry. Immunohistochemical staining was processed in 4 휇m paraffinized sections. Briefly, the sections were deparaffinized. Then, 3% H 2 O 2 was used to block endogenous peroxidase activity for 10 min at room temperature. And antigen retrieval was enhanced by microwave irradiation in citrate buffer. Nonspecific adsorption was minimized by incubating sections in normal goat serum in PBS for 20 min. Sections were incubated with the primary antibodies overnight at 4 ∘ C. The primary antibodies included rabbit anti-Ki67 (ab9260, Abcam, dilution1:100) and rabbit polyclonal to myeloperoxidase (MPO, ab45977, Abcam, dilution1:50). Then, Elivi-sion6 plus polymer HRP (Mouse/Rabbit) IHC kit (Kit-9902, Maixin Biotechnology Corp., Fuzhou, China) was used. The sections were incubated with universal mouse/rabbit polymers for 30 min. The color reaction was developed with 3,3diaminobenzidine and sections were counterstained with hematoxylin. The slides were assessed by an experienced renal pathologist who knew nothing about the origin of the slides. Digital photomicrograph analysis by Image-Pro Plus 5.02 (Media Cybernetics, Silver Spring, MD, USA) was chosen to quantitate the expression of MPO in each group.
. . Periodic Acid-Schiff Staining. Periodic Acid-Schiff (PAS) staining was processed in 4 휇m paraffinized sections. Briefly, the sections were deparaffinized and rinsed with distilled water. And the sections were immersed in the periodate alcohol solution for 10 min. Then, Schiff stain was added on the sections. After 10 min, the sections were washed with distilled water for 5 min. The nucleus was stained with Mayer hematoxylin solution. After flushing with running water, they were dehydrated and were transparent and sealed. PAS stain was used to observe the pathological changes of renal tubular basement membrane in this study.

Statistical Analysis
All data were expressed as the mean ± standard deviation (SD) and analyzed by using SPSS 17.0 (SPSS Inc., Chicago, IL, USA). Differences among groups were assessed by one-way ANOVA followed by Dunnett's tests. The significant statistical difference was defined as P<0.05.

. . Renal Function and Morphologic
Change on Different Days a er Ischemia. Rats were exposed to 45 min of bilateral renal ischemia and sacrificed on days 1, 2, 3, 5, and 7. After ischemia, rats showed marked deterioration of renal function with an increase in Scr levels on the 1st day (217.0±18.4 휇mol/L), 2nd day (333.5±9.2 휇mol/L), 3rd day (392.5±24.8 휇mol/L), and 5th day (198.4±16.5 휇mol/L) compared to basal levels of shamoperated rats (24.5±4.9 휇mol/L, all P<0.05). Furthermore, Scr reached its peak level on day 3. At 7 days, the Scr level (40.5±13.4 휇mol/L) was a little higher than the one in sham-operated rats, but there was no significant difference ( Figure 1). Light microscopy of H&E-stained and PASstained kidney sections showed varying degrees of tubular epithelial cell necrosis, naked basement membranes, and tubular dilation with proteinaceous or cellular casts on days 1, 2, 3, and 5 after ischemia (Figures 2(a) and 2(b)). By day 7 after ischemia, renal histology was near normal (Figures 2(a) and 2(b)). Moreover, the kidney injury score was the highest on day 3 after ischemia ( Figure 1). Accordingly, the third day was recognized as an observation point for subsequent studies.
. . Intravenous Transplantation of BMSC Attenuates IRI-AKI. To identify rat BMSC, their typical surface markers and ability to differentiate were tested. Flow cytometric analysis confirmed that CD29, CD44, and CD90 surface markers in BMSC were positive (Figure 3(a)). The cell matrix exhibited fat drops in some cell bodies following oil red staining (Figure 3(b)-(B)), mucopolysaccharide deposition following alcian blue staining after 2-week induction (Figure 3(b)-(C)), and calcium deposition following the alizarin red staining (Figure 3(b)-(D)). These suggested that the BMSC had the ability to differentiate into adipocytes, chondrocytes, and osteoblasts.
To assess the therapeutic effect of BMSC on AKI, IRI-AKI rats were treated with BMSC or medium control (DMEM) intravenously. As shown in Figure 4(a), the Scr levels of IRI-AKI rats treated with BMSC (73.5±7.8 휇mol/L) significantly decreased compared to the IRI-AKI rats treated without BMSC (392.5±24.8 휇mol/L, P<0.05) or with DMEM (314.0±19.8 휇mol/L, P<0.05) at 3 days after 45-minute renal ischemia. Meanwhile, the renal tissue injury in BMSC group improved markedly compared to that in IRI-AKI group or DMEM group (Figures 4(b) and 4(c)). These suggested that intravenous transplantation of BMSC could attenuate IRI-AKI in rats.

. . TSG-Inhibits Infiltration of Neutrophils in Kidney Tissue of IRI-AKI Rats.
To explore the anti-inflammatory effect of TSG-6 in IRI-AKI, the levels of expression of MPO protein in kidney were determined by quantitative immunohistochemistry (Q-IHC) (Figures 10(a) and 10(b)). It showed that the level of expression of MPO in IRI-AKI rats treated with TSG-6 shRNA BMSC was significantly higher than those in IRI-AKI rats treated with BMSC or sc-shRNA BMSC (P<0.05). This indicated that TSG-6 could inhibit infiltration of neutrophils in kidney tissue of IRI-AKI rats.

Discussion
This study investigated the mechanism of BMSC ameliorating IRI-AKI in rats. The present study first illustrated that TSG-6 was the key factor in BMSC which ameliorated IRI-AKI in rats. Moreover, we found that TSG-6 worked through  promoting renal proximal tubular epithelial cells to proliferate in vivo and in vitro. At present, MSC is an ideal target cell for stem cell-based therapy. MSC has the ability to regulate immune response. Previous studies suggested that MSC could ameliorate kidney damage induced by IRI [12]. Also in our study, administration of BMSC intravenously resulted in significant improvements of renal function and renal histology, which was consistent with previous studies. However, the underlying mechanism of MSC's therapeutic effect is still unclear. It was found that when MSC were administrated into animal models of AKI, only a small proportion (0.1%∼2.5%) of MSC could colonize and integrate into the damaged kidney [13]. Instead, the mechanism of MSC's effect is largely attributed to their unique anti-inflammatory and immune-modulatory properties via paracrine effects [14]. Among the secretory factors of activated MSC, TSG-6 has a powerful anti-inflammatory effect [7]. Its main functions include immunosuppressive modulation and extracellular matrix remodeling [15]. In particular, TSG-6 is used as a biomarker to predict efficacy of MSC in modulating sterile inflammation [16].
Previous studies have shown that the TSG-6 secreted by MSC is closely related to its therapeutic effects in several models of acute organs injury [17][18][19][20]. Lee  MSCs [18]. Although TSG-6 has immunomodulatory effects on several inflammatory diseases, the factors secreted by MSC have different effects in different environments. Currently, there are few studies about the effect of TSG-6 secreted by MSC on kidney injury. One previous study showed that BMSC might exert the effects of anti-inflammation and antifibrosis on renal tubular cells under albumin-overloaded conditions via hepatocyte growth factor (HGF) and TSG-6 [21]. However, there was a lack of studies in IRI-AKI. In the present study, stable animal models of IRI-AKI were constructed successfully. And our data showed that the levels of serum creatinine reached a peak at 3 days after renal ischemia, which was consistent with previous reports [22]. Then, the third day was chosen as the observational point for subsequent study. Furthermore, the therapeutic effects of BMSC on IRI-AKI were largely abrogated by silencing their TSG-6. Therefore, TSG-6 was the key factor in BMSC which ameliorated IRI-AKI in rats.
Our study found that TSG-6 promoted renal tubular epithelial cells to proliferate in IRI-AKI rats. Moreover, our in vitro data showed that TSG-6 had promoted the proliferation of rat renal proximal tubular epithelial cells under hypoxia. These findings suggested that TSG-6 exerts its effect by promoting renal tubular epithelial cells proliferation. However, the underlying molecular mechanism remains unknown. It was reported that, in animal model of acute liver injury, TSG-6 could promote liver regeneration by suppressing inflammation [23]. Several previous studies showed that TSG-6 was an inflammation-associated secreted protein that had been implicated as having important and diverse tissue protections. Accordingly, in the setting of IRI-AKI, the proliferative effect of TSG-6 might be exerted through modulating inflammation. In the present study, we observed that TSG-6 could inhibit infiltration of neutrophils in kidney tissue of IRI-AKI rats. This might indicate that TSG-6 contributed to promoting renal tubular epithelial cells proliferation through its anti-inflammatory effect. In addition, several studies indicated that MSC could promote normal macrophages to transit into M2 phenotype [24]. Therefore, further studies will be needed to identify the underlying mechanism of TSG-6 to promote proliferation and anti-inflammatory effect in AKI.
In summary, our data suggest that TSG-6 is the key factor that allows BMSC to treat IRI-AKI. Furthermore, TSG-6 might exert its effects on promoting renal tubular epithelial cells proliferation by modulating inflammation. This study not only reveals the possible mechanism of BMSC treating AKI but also provides a novel drug (TSG-6) as an alternative to BMSC therapy.