Possible Protective Effect of Bone Marrow-Mesenchymal Stem Cells (BM-MSCs) Against the Remote Liver Injury Induced by Renal Ischemia Reperfusion in Male Albino Rats

Liver and kidney are both important regulator organs in our body and any liver or kidney injury may influence the other [1]. Acute kidney injury and ischemic renal diseases are common clinical entities [2]. Renal Ischemia-reperfusion injury (IRI) describes the tissue ischemia with an inadequate oxygen supply followed by reperfusion which initiates a wide array of inflammatory responses [3]. Remote organ injury is anoxidative damage that may be seen in various organs away from the tissue exposed to IRI [4]. Remote organ failure like liver, brain and lung commonly coexists with acute renal injury in the intensive care units and it increases the mortality risk [2,5,6]. Despite the advanced renal replacement therapy, mortality among patients who sustain acute renal injury complicated by multiorgan dysfunction, remained unchanged and was estimated at approximately 50% [2].

functional derangements and cell death ultimately [9]. This complex pathophysiology resulting from a number of contributing factors, so it is difficult to achieve effective treatment or protection by targeting only individual mediators or mechanisms despite the efforts made to deal with this problem [10]. The increasing mortality rate in cases with acute renal injury necessitates designing strategies to assess the effect of renal injury on distant organs [11]. Therefore, several studies [8,[1][2][3][4][5][6][7][8][9][10][11][12][13][14] tried to protect liver from remote injury induced by renal IRI for the prevention and attenuation of acute renal injury related morbidity.
Mesenchymal stem cells (MSCs) are emerging as a potential cell therapy for various inflammatory diseases giving much promise for kidney repair after damage [4]. Although the mechanism of MSCs affects remains unclear, their early protection is primarily mediated through complex paracrine actions as immunomodulation and growth factor secretion [15]. Recent studies have shown that transplantation of MSCs in renal IRI in animal models [16][17][18][19] or humans [20][21][22] with renal IRI promotes renal functional recovery and reduces renal tissue damages. This indicates that cell therapy could be a promising alternative treatment for tissue repair and preventing IR-induced renal damage [18]. However, its therapeutic effects on the remotely affected organs induced by renal IRI, was not yet investigated. Therefore, the aim of this study was to investigate the possible therapeutic effect of BM-MSCs therapy on the remotely affected liver in a rat model of renal IRI and the mechanisms involved in this protection which monitored by chemical histological and Immunohistochemical studies.

Ethics
All aspects of animal care and treatment were carried out according to the local guide-lines of the Ethical Committee of Faculty of Medicine of Minia University.

Bone marrow derived-mesenchymal stem cells (BM-MSCs)
Rat bone marrow derived, bromouridine-labeled, mesenchymal stem cells (BM-MSCs) were purchased from the Nile Center Laboratory Animal Unit of Mansoura Experimental Research (Mansoura, Egypt).

Preparation of BM-MSCs
After anesthetizing rats (6-weeks-old male albino rats) by halothane and skin sterilization, femurs and tibiae of rats were dissected, and washed in Petri dish by phosphate buffer saline (PBS) (Hyclone, USA). Bone marrow was collected by flushing femurs and tibiae with Dulbecco's modified Eagles medium (DMEM) (lonza, Belgium) supplemented with 10% fetal bovine serum (FBS) (lonza, Belgium) and 1% penicillin-streptomycin (lonza, USA). Nucleated cells were isolated using a Ficoll/Paque density gradient (Pharmacia) and were cultured in 20 ml complete media (lonza, USA) supplemented with 1% penicillin-streptomycin and incubated at 37°C in 5% humidified CO 2 incubator (shellab, USA) for 7-10 days until formation of large colonies. The cultured cells were examined daily using the inverted microscope (Olympus CKX41, USA) to follow up the growth of cells. The 2 nd exchange for media was done after 3-4 days. The cells took 3-4 weeks to be confluent (80∼90% confluence) and be ready for passaging. The media changed twice a week. The culture was washed with PBS and cells were trypsinized with 1-2 ml of 0.25% trypsin in 1 mM EDTA (lonza, Belgium) (5 min at 37°C). Then, 5 ml complete media was added to stop trypsinization, and to test cell. After centrifugation, the cells were resuspended with complete medium in 2 tissue culture flasks (Falcon) containing 15 ml complete media and incubated in humidified incubator at 37°C in 5% CO 2 . BM-MSCs in culture were characterized by their adhesiveness and fusiform shape [23]. Counting of cells were done and number of cells/ml=4 × 10 5 .

Labeling of stem cells
BM-MSCs were harvested during the 2 nd passage. Cells were centrifuged and washed twice in serum free medium. Cells were suspended in bromouridine (Sigma Aldrich, Egypt) fluorescent linker dye solution for labeling [23].

The Experimental Design
This study was carried out on 54 adult male Wistar albino rats, weighing 200-250 gm, and of 12 weeks of age. The animals were obtained from Minia University Laboratory Animal House, then housed in plastic cages at room temperature with normal light/dark cycles. Rats were fed a standard laboratory diet and water ad libitum. Animals were randomly divided into three equal groups (18 rats/ group): i. Group I; the sham-operated control group: rats underwent the same surgical procedure of renal IRI, except that the kidneys were exposed but not clamped. Rats received a single intravenous (iv) injection of saline as a vehicle [12].
ii. Group II; the renal IRI group: rats were subjected to bilateral renal IRI, then received a single iv injection of saline immediately after reperfusion.
iii. Group III; BM-MSCs injected group: rats were subjected to bilateral renal IRI, then received a single iv injection of BM-MSCs immediately after reperfusion.
Rats from each group were subdivided into 3 sub-groups (6 rats each) and were sacrificed as follows: 1. Sub-group a: 1 day after saline or BM-MSCs injection.
2. Sub-group b: 3 days after saline or MSCs injection.

Induction of experimental Renal IRI
Rats were placed on a warming pad and were anesthetized with pentobarbital sodium (60 mg/kg i.p). A midline abdominal incision was performed after disinfection, kidneys were exposed, and renal pedicles were bilaterally clamped by non-traumatic clamps for 45 minutes, followed by reperfusion by removal of the clamps [24]. The occlusion and reperfusion were verified visually by changes in color. The incisions were covered with gauze soaked with normal saline to prevent evaporation. Body temperature was maintained at 37 ± 1°C. The incision was closed in 2 layers and a topical application of garamycin cream was used to avoid wound infection.

Injection of BM-MSCs
Immediately after assurance of renal reflow, rats of group III received 0.5 ml bromouridine-labeled BM-MSCs (2 × 10 5 ) diluted with 1 ml saline, loaded in a 1-ml sterile syringe and injected intravenously in the tail vein for each rat [19].

Animal sacrifice and tissue obtaining
At the end of the experiment, 6 rats of each sub-group at each time point were sacrificed by decapitation after light ether anesthesia. Tail vein blood samples were collected shortly prior to scarification.
Livers were rapidly removed and were divided; some specimens were placed into liquid nitrogen and then stored at -70°C until assayed for oxidant and antioxidant parameters, and other specimens were used for paraffin embedding.

Determination of renal and liver functions
Tail vein blood samples were collected shortly prior to IRI induction (at day 0 for exclusion of renal or hepatic abnormal functions) and after intervention throughout the experiment at day 1, day 3 and day 5. Sera were used for estimation of: urea by Berthelot enzymatic colorimetric method, creatinine by Jaffé calorimetric-end point method [25] as renal function tests. Liver transaminases; alanine aminotransferase (ALT) and aspartate aminotransferase (AST), by the enzymatic colorimetric methods [26] using commercial kits (Biodiagnostic, Egypt).

Determination of lipid peroxidation; malondialdehyde (MDA), and the antioxidant enzyme; reduced glutathione (GSH)
Liver tissues were homogenized in 1:10 (wt:vol) phosphate buffer (pH 7.4) by the use of a Teflon headed homogenizer at a speed of 2500 rpm. Triton x100 and protease inhibitor cocktail were added. The homogenates were centrifuged at 6.000g for 10 min at 4°C. The resulting supernatant was used for colorimetric determination of hepatic MDA [27] and GSH [28] using commercially available kits (Biodiagnostic, Egypt) using the spectrophotometer (Jenway 7305, Staffordshire, UK).

Histological procedures
Liver specimens were fixed in 10% buffered formalin, followed by paraffin embedding. Some sections of 5∼7 μm thickness were cut, mounted on glass slides and stained with hematoxylin and eosin (H&E) for histological examination [29].

Immunohistochemistry
Other 5 μ-sections were used for immunohistochemical staining for the mouse monoclonal TNF-α (Sigma Aldrich, Egypt) according to the manufacture's guidelines. Briefly, sections were deparaffinized, rehydrated, incubated with 0.1% trypsin and 0.1% CaCl 2 2H 2 O in 50 mM Tris buffer at pH 7.4 at 37°C for 120 minutes, and soaked in absolute methanol containing 0.3% hydrogen peroxid for 30 min at room temperature. Sections were then incubated with 1.5% nonimmunized goat serum for 30 min, then incubated with the diluted TNF-α antibodies (5-10 μg/ml) for 30 minutes at room temperature, and washed 3 times with phosphate-buffered saline (PBS) for 30 min. Thereafter, sections were incubated for 60 min with biotinylated goat anti-mouse Ig serum. Sections were washed with PBS and incubated with avidin/biotin peroxidase complex (Vector, Burlingame, CA). Sites of peroxidase binding were detected using chromogenic 3,3'-diaminobenzidine (DAB) tetra hydrochloride substrate. Tissue sections were counterstained with haematoxylin. Positive cells showed brown cytoplasmic reaction [30].

Positive tissue control
Sections of human tonsil were immunostained for TNF-α.

Negative control
Other sections of liver in the sham-operated control group were processed in the same way but skipping the step of applying the primary antibody (figures not included).

Image capture
a) For detection of engraftment of bromouridine-labeled BM-MSCs in the liver tissue, H&E sections were examined using a fluorescence microscope (Leica M205 FCA, Germany). b) H&E and immunostained sections were examined using a light microscope (Olympus, Japan). Images were digitally captured using a hardware consisting of a high-resolution color digital camera mounted on the microscope, connected to a computer.

Morphometric study
The TNF-α immunopositive cells were counted in 10 adjacent nonoverlapping fields of the tissue sections of each rat in all sub-groups [30].

Data handling and statistics
Analysis of the quantitative data was carried out using SPSS version 20 (SPSS Inc., Chicago, Illinois, USA). Data were represented as mean ± standard deviation (mean ± SD). Significant difference between subgroups was done by one-way ANOVA followed by Tukey-Kramar post hoc test for multiple comparisons. p-value ≤0.05 considered statistically significant.

Effect of BM-MSCs on renal and hepatic function tests
Renal ischemia-reperfusion injury (IRI) resulted in a significant increase in serum urea, creatinine, aspartate aminotransferase (AST) and alanine aminotransferase (ALT), levels on days 1, 3 and 5 compared with those in sham-operated control rats (all p =0.000). Their levels in the BM-MSC group on day 1 showed no significant difference compared to the IRI group (all p>0.05). On days 3 and 5, the levels showed significant decrease compared with the IRI group (all p ≤ 0.05) but it was still showed a significant difference with the sham-operated control group (all p ≤ 0.05) ( Table 1).

Effect of BM-MSCs on renal IRI-induced oxidative stress in the liver evaluated by malondialdehyde (MDA) and reduced glutathione (GSH) levels
Hepatic tissue levels of MDA were significantly higher in the IRI group on days 1, 3 and 5 compared with those in sham-operated control rats (all p=0.000). BM-MSCs-injected group showed an early significant decrease in the MDA levels at day 1 and also at day 3 and 5 (all p=0.000) compared with the IRI sub-groups, but remained significantly elevated compared to the sham-operated control group (all p ≤ 0.05) at the corresponding time points.
Regarding the hepatic tissue levels of GSH there was a significant decrease in IRI sub-groups compared to the sham-operated control rats on days 1, 3 and 5 (all p =0.000). However, the BM-MSCs-injected sub-groups showed significantly higher levels on days 1, 3 and 5 (all p=0.000) compared to IRI sub-groups, but still there was a significant difference when compared with the sham-operated control group at the day 1 and 3 (all p ≤ 0.05) with an insignificant difference at day 5 (p= 0.26) ( Table 2).

Confirmation of homing of BM-MSCs into the liver tissue:
Bromouridine-labeled BM-MSCs showed green auto fluorescence after transplantation into rats in the BM-MSCs-injected sub-groups, confirming that these cells were engrafted into liver tissue. Sections in livers of sub-groups IIIa and IIIb showed numerous BM-MSCs labeled with bromouridine with obvious decrease in sub-group IIIc (Figure 1).

Hematoxylin and eosin-stained sections
Liver sections of the sham-control group exhibited normal lobular architecture at the different time-points of the experiment (represented in Figure 2). Hepatocytes radiated from the central vein forming anastomosing fenestrated plates separated by the hepatic sinusoids. Portal tracts contained branches of portal vein, hepatic artery, and bile duct. The hepatocytes appeared polyhedral with acidophilic finely granular cytoplasm and large central vesicular nuclei with prominent nucleoli. Hepatocytes may be bi-nucleated.
Examination of sections of sub-groups IIa ( Figure 3) and IIIa ( Figure 4) showed disrupted lobular architecture and marked hepatic injury especially in the periportal areas characterized by marked hepatocellular ballooning and cytoplasmic vaculations. Their nuclei appeared malformed. Interestingly, sections of sub-group IIa exhibited marked recruitment of the inflammatory cells; neutrophils and lymphocytes, in the vicinity of the portal tract with prominent vascular congestion.
Sub-group IIb showed no obvious changes compared to sub-group IIa. Central vein appeared congested and engorged with inflammatory cells. The inflammatory cells infiltrated the hepatic tissue. Degenerated hepatocytes with vacuolated cytoplasm and ghosts of nuclei were numerous. Other cells showed apoptotic features in the form of dense nuclei and deeply acidophilic cytoplasm ( Figure 5). While sub-group IIIb showed some sort of regeneration e.g. restored lobular architecture    with less marked degenerative changes in cells, decreased congestion, mildly dilated blood sinusoids, and also minimal inflammatory cell infiltration. Most hepatocytes had more or less normal appearance, arranged in cords with minimal small vacuoles. Few cells with apoptotic features (dense nucleus and deeply acidophilic cytoplasm) were observed ( Figure 6).
Sections of sub-group IIc (Figure 7) showed no obvious differences compared to sub-groups IIa and IIb. While sub-group IIIc exhibited marked morphological restoration in the form of preserved general architecture and lacking evidence of major morphological injury. Most hepatocytes restored its cytoplasm without vaculations and most nuclei appeared normal vesicular with scarcely observed degenerated cells. Vascular congestion appeared to be remarkably reduced. Only mild dilatation of some blood sinusoids, and inflammatory cells were rarely seen ( Figure 8).

Immunohistochemical results
Examination of wide fields of sham-operated control group showed no TNF-α immunolabeled cells (Figures 9-14). Extensive immunolabeling for TNF-α was markedly observed in the liver sections of sub-groups IIa (Figure 10), IIb (Figure 12), IIc ( Figure  14), and IIIa ( Figure 11). Unexpectedly, sub-group IIIa showed numerous immunolabeled cells in some sections compared to subgroup IIa. Sub-group IIIb ( Figure 13) showed some decrease while sub-group IIIc ( Figure 15) showed marked decrease in the TNF-α expression when compared to the extensive immunolabeling seen in sub-group IIc. The immunoreactivity for TNF-α was observed in hepatocytes, macrophages (Von-kupffer cells), and endothelial cells. The immunolabeling was cytoplasmic.

Morphometric results
There was a significant increase in the number of TNF-α immunolabeled cells in the IRI and BM-MSCs-injected sub-groups compared to the corresponding sham-operated control group (all p =0.000). While there was a significant decrease in the BM-MSCsinjected sub-groups compared to the corresponding IRI sub-groups (all p<0.05) ( Table 3).

Discussion
The present study demonstrated that bilateral renal IRI led to damage of the liver as a remote organ. The significant increase in serum urea and creatinine confirmed the occurrence of renal dysfunction and the altered serum AST and ALT indicated remote liver injury. Renal IRI induces an inflammatory response, which results in the formation of reactive oxygen species (ROS) that augments local tissue damage or affects organs remote from the site of IRI [1,5,6,9]. Several previous studies [16][17][18]. suggested that stem cell therapy enhance renal IRI. Also in this study, administration of BM-MSCs-injected resulted in significant improvement of renal functions evidenced by serological results. Moreover, this study proved that administration of BM-MSC also rapidly and gradually restored liver functions evidenced by significant improvement of serum AST and ALT.
Hepatic oxidative stress is evidenced by increased liver MDA; a lipid peroxidation marker [31] and decreased levels of the antioxidant enzyme (GSH) [5,7]. In the present study, a significantly increased  MDA levels and decreased GSH levels reflected the overproduction of ROS generated during ischemia followed by reperfusion, which was then removed by GSH.
Hepatic GSH depletion occurred in the IRI sub-groups might be caused by inhibition of GSH biosynthesis through down-regulation of the trans-sulfuration pathway that limits the availability of cysteine; an essential precursor for GSH biosynthesis as mentioned by Shang et al. [32].
Light microscopic examination of this study showed that altered liver enzymes and changes in hepatic oxidative stress markers were in harmony with the hepatic morphological findings. Renal IRI rapidly caused marked hepatic vacuolization evident sinusoidal dilatation, and vascular congestion. These results were supported by other researchers [1,7] who stated that liver cell injury and hepatocyte death (necrotic and apoptotic) occurred following renal IRI as a result of oxidative stress and decreased its antioxidant capacity. In addition, some investigators attributed liver cell damage to the microvascular reorganization of hepatic parenchyma [33].
ROS causes oxidation of membrane lipids, essential cellular proteins, and DNA leading to release of proteolytic lysosomal enzymes and mitochondrial matrix enzymes into the cytoplasm ending finally in cellular damage [34].
Tissue sections also showed that renal IRI induced massive leukocytic recruitments in liver tissue. This was in a good agreement with several studies [5,7] who reported recruitment and activation of neutrophils as a key step in the development of local and systemic injury in renal IRI.
Another important finding in the current study was the expressions of TNF-α in liver tissues of IRI sub-groups detected by immunohistochemistry. There was marked increase in TNF-α expression in sections of IRI sub-groups compared to sham-operated control group. This was in agreement with those of previous studies [21,35]. TNF-α is a strong early pro-inflammatory mediator that released from a variety of cells (e.g. hepatocytes, Von-kupffer cells, extra) in response to injury [36]. Renal IRI results in uncontrolled expression of interleukin-17A in the small intestines (1). IL-17A causes recruitment of neutrophils, activates T cells, and induces expression of other cytokines and chemokines such as TNF-α and IL-6 in liver tissue [11,37]. Moreover, other studies showed that following experimental renal IRI in knockout mice for TNF-α or neutralized with TNF-α antibodies displayed reduced hepatic injury after renal IRI [1]. The overexpression of TNF-α in this study pointed out the severe damage observed in the hepatic parenchyma and impairment of liver functions in the IRI sub-groups.
In this study, the injected BM-MSCs migrated to the remotely injured liver in BM-MSCs-injected sub-groups, which was confirmed by localization of bromouridine-labeled BM-MSCs visualized by the fluorescent microscopy. Homing of stem cells to the injured sites could be attributed to certain substances; e.g. chemokine, released at sites of tissue damage at the inflammation site which directs MSCs migration to these sites [20]. MSCs express receptors for several chemokine's such as stromal-derived factor-1, platelet-derived growth factor and CD44 which are likely implicates in regulation of its homing [38]. The migrated MSCs might act by paracrine effects and/or its differentiation [20,39].
Several mechanisms of stem cell therapy which affect the injured tissues had been established which included anti-inflammatory reaction, anti-oxidative stress, stem cell homing and trans-differentiation, and immune-modulation. MSC transplantation limits cellular apoptosis and architectural damage [20,21,40].
Accumulating evidences demonstrated that MSCs contributes to the down-regulation of inflammatory reaction in ischemic condition [41]. In this study, serum ALT and AST levels were reduced in the BM-MSC sub-groups, corroborating the protective effect of BM-MSCs.
Also there was a significant decrease of MDA and a significant increase of reduced GSH. This means that BM-MSC suppressed the oxidative stress and lipid peroxidation induced by renal IRI. The lower level of MDA in the liver tissue at day 1 was indicative of an early protection of BM-MSCs to the liver by reducing the oxidative stress markers and preservation of the antioxidative enzymes with a consequent anti-inflammatory response.
Light microscopic examination of the BM-MSCs-injected subgroups revealed that during the 1 st 24 hours, there was hepatic tissue injury as observed in the IRI sub-groups, but an important observation was the less marked inflammatory cell recruitment. Subsequently, rapid improvement in the histological morphology of the liver observed at day 3 and became more or less as normal by the 5 th day. The process of trans-differentiation of administered MSC requires more than 3 days and the early improvement is due to the paracrine effects of stem cells [24]. In this study, the early effects of BM-MSCs were noted within 3 days of injection; before the sufficient time required for cellular growth, division and differentiation. Therefore, direct hepatic parenchymal repopulation by exogenous BM-MSCs is probably a less significant mechanism of BM-MSCs induced hepatic repair. However, the prolonged presence of BM-MSCs in liver and its trans-differentiation remains to be elucidated.
Inflammatory pathway was implicated in the kidney-liver cross-   talk [6]. Giving the importance of inflammation in the pathogenesis of remote liver injury induced by renal IRI, was very important to consider the immunomodulatory properties of BM-MSCs and its role in hepatic protection. In this study, BM-MSCs blocked the hepatic inflammation by inhibiting inflammatory cell recruitments, where minimal inflammatory cells were observed infiltrating the hepatic tissue if compared to massive infiltration of these cells in the IRI subgroup at the different time points of the study. Therefore, this study suggested that BM-MSCs, by its potent anti-inflammatory effects, were protective against renal IRI with subsequent reduced the remote hepatic injury in rats.
Stem cells were able to ameliorate renal IRI in a mouse model via the suppression of TNF-α (18,22). In the current study, administration of BM-MSCs normalized the overexpression of TNF-α in liver tissues. As Von-Kupffer cells play an important role as an initial cytotoxic cell type and are likely to be a source of ROS [42,43] conversely, it may protect the tissues by secretion of anti-inflammatory cytokines [44]. The unexpected apparent increase in the immunolabeled cells in some sections of sub-group IIIa compared to sub-group IIa could be due to the massive vaculations in cells of sub-group IIIa.
Taken together the findings of the present study beside the advantages of treatment of renal IRI by MSCs which confirmed by several studies [16][17][18][19]21,22,40], it could be assumed that treatment of renal IRI with BM-MSCs did not prevent the acute remote liver injury induced by renal IRI but ameliorated this injury and enhanced rapid liver recovery both morphologically and functionally which likely mediated by various paracrine mechanisms distinct from transdifferentiation. This was, at least in part, through early inhibition of inflammatory cell recruitment, inflammatory reactions, suppressing oxidative stress and lipid peroxidation, rapid restoration of cellular and architectural integrity of the liver. Therefore, BM-MSCs therapy might have the upper hand in treatment of renal IRI to prevent sever remote injury that could affect severity and mortality associated with renal IRI. However, the exact mechanisms underlying the better organ recovery of liver treated with BM-MSCs in renal IRI remain to be investigated.

Study Limitations
Even though our findings are promising, molecular studies of the liver tissues were not conducted and the underlying mechanisms involved in the therapeutic effect of BM-MSCs remain descriptive. The exact role of Von-kupffer cells also remained unclear. Stem cell therapy may be promising provided assurance of safe transition to use in humans by further studies to explore whether BM-MSCs have the same hepato-protective effect in humans. Follow up after IRI looking into the chronic influence of treatment of BM-MSCs.

Conclusion
This study suggested that renal IRI initiated remote liver dysfunction and BM-MSCs enhanced morphological and functional recovery of liver as a remote organ from the renal IRI. BM-MSCs protects against remote liver injury through early inhibition of inflammatory cell recruitment, inflammatory reactions, suppressing oxidative stress and lipid peroxidation, and rapid restoration of cellular and architectural integrity of liver. This provides another important basis for the therapeutic concept of BM-MSCs in treatment of renal IRI by adding a beneficial effect on decreasing the remote liver injury besides considering it as a promising treatment in renal IRI.

Conflicts of Interest
There is no conflict of interest to declare.