A comparison between centrally and systemically administered erythropoietin on kidney protection in a model of fixed-volume hemorrhagic shock in male rats

In this study, a comparison between centrally and systemically administered erythropoietin (EPO) was performed on nephroprotection during hemorrhagic shock (HS) in male rats. Male rats were allocated into four experimental groups. (1) Sham; a guide cannula was inserted into the left lateral ventricle and other cannulas were placed into the left femoral artery and vein. (2) HS; stereotaxic surgery was done to insert a cannula in the left lateral ventricle and after a 7-day recovery; hemorrhagic shock and resuscitation were performed. (3) EPO-systemic; the procedure was the same as the HS group except that animals received 300 IU/kg erythropoietin into the femoral vein immediately before resuscitation. (4) EPO-central; animals was treated with erythropoietin (2 IU/rat) into the left lateral ventricle before resuscitation. Arterial oxygen saturation (SaO2) was measured during experiments. Urine and renal tissue samples were stored for ex-vivo indices assessments. Erythropoietin (systemically/centrally administered) significantly improved SaO2, renal functional and oxidative stress parameters and decreased renal inflammatory (TNF-α and IL-6) mRNA expression compared to the HS group. EPO-treated groups showed a decrease in active form of caspase-3 protein level and an increase in autophagy activity in comparison with the HS group. Considering the fact that the effective dose of systemic EPO (300 IU/kg) was roughly 50 times higher than that of central administration (2 IU/rat), centrally administered EPO was accompanied by more advantageous consequences than systemic way. EPO is likely to act as a neuro-modulator or neuro-mediator in the central protection of organs including the kidneys.


Introduction
Hemorrhagic shock (HS), as a complex outcome of systemic inflammatory response syndrome (SIRS), may lead to multiple organ dysfunction (MOD) and death [1,2]. Fluid resuscitation is known to have valuable effects, however, it is accompanied by some complications. If given in large amount, fluid resuscitation may deteriorate the acute traumatic coagulopathy because of dilution, hypothermia and dislodging recently formed clots [3].
The kidneys are among the prime target organs, whose protection during and after hemorrhagic shock and resuscitation might prevent SIRS and MODS. Following the fluid resuscitation, intra-renal inflammation induced by inflammatory cells infiltrating the kidney and subsequent generation of reactive oxygen species (ROS) may intensify kidney damage [4]. Moreover, HS is also known to induce cell apoptosis in different organs such as the kidneys [5].
Today, EPO is at the center of attention not only for its major role in erythropoiesis but also for its function on nonhematopoietic systems. The existence of EPO receptors in 1 3 non-hematopoietic tissues offers several biological functions for EPO [6]. The first observation of EPO receptor's existence in the CNS as well as hypoxia induced-local EPO production were reported in the early 1990s [7,8]. Studies have reported that EPO possess beneficial effects in several different organs and tissues, such as the kidneys [9], and CNS [10]. Extensive brain apoptosis was reported in the mice lacking the EPO receptors indicating that EPO receptors is likely to be effective in the brain development. Under physiological condition in the adult brain, constitutive EPO system's expression is low, whereas, expression of EPO and EPO receptors significantly increases during central nervous system damage [11]. The idea that EPO is likely to be produced in the brain and acts in the central nervous system as a neuro-protector or neuro-modulator has gained attraction in recent years [12,13].
Following increased production of reactive oxygen species and inflammation, programmed cell death pathways regulate cell survival through apoptosis and autophagy. Autophagy is a protective process that involves the degradation of damaged intracellular organs via a lysosomal pathway to help maintain cellular homeostasis and energy supply. Autophagy has been shown to play an anti-inflammatory role in many diseases [14]. Increasing autophagy activity in renal tubules during kidney transplantation, nephropathy, acute renal failure due to ischemic reperfusion has been reported. Several studies have shown the effect of EPO on activation of autophagy through regulation of ERK ,mTORC1, AMPK pathways [15]. Also, EPO is regarded to play a crucial role in the expression of essential proteins in the formation of autophagosomes during autophagy such as Beclin 1 [16].
We have already evaluated the effective role of systemically administered EPO on kidney protection during HS [2,9,13], so, in this study, we tried to compare the benefits of centrally and systemically administered erythropoietin on renal functional, oxidative, inflammatory and indices of autophagy activity in male rats subjected to hemorrhagic shock.

Animals
Male Wistar rats were obtained from Tehran University of Medical Sciences. Animals (n = 24, 285-300 g) were randomly assigned into different experimental groups. Animals were held under standard conditions of temperature and light-dark cycles with free access to food and water. Experimental protocols and animal care methods in the experiments were approved by the Animal Experimental Committee at Tehran University of Medical Sciences (Project number: 27721-30-04-93, Approval ID: 27,721).

Stereotaxic surgery
After anesthesia, using intraperitoneal ketamine (50 mg/kg) and xylazine (10 mg/kg), rats then placed in a stereotaxic apparatus. A guide cannula (Gauge 21) was inserted into the left lateral ventricle according to following stereotaxic coordinates: AP 0.8 mm, LR 1.4 mm, and H 3.6 mm [17]. All animals were allowed to have a 7-day recovery before the induction of hemorrhagic shock.

Hemorrhagic shock
Hemorrhagic shock was induced in the same way described in our previous study [2]. Briefly, after anesthesia, using intraperitoneal ketamine (50 mg/kg) and xylazine (10 mg/ kg), two catheters were placed into the left femoral artery and vein in order to blood withdrawal and blood resuscitation respectively. 50% of total blood volume was removed according to this formula [Weight (g) × 0.03 + 0.7 ml] [18] through the femoral artery catheter within 30 min using heparinized syringes. Blood was stored in plastic conical tubes at 37 °C. Two hours later, animals were resuscitated with the shed blood and equal volume of Ringer's lactate injection into the femoral vein catheter within 30 min. Rats were sacrificed after a 3-hour post-resuscitation period (Fig. 1). Fig. 1 Induction of hemorrhagic shock. Hemorrhagic shock stages were as follows: (1) blood withdrawal (50% of total blood volume) within 30 min, (2) the ischemia phase lasted for 2 h, (3) resuscita-tion phase was performed with the shed blood and equal volume of Ringer's lactate through the femoral vein catheter within 30 min. (4) Rats were sacrificed at the end of a 3-hour post-resuscitation period

EPO administration
In order to find the renal protective dose of EPO, it was injected centrally (into the left lateral ventricle) in the doses of 0.5, 1 and 2 IU in 5 µl normal saline just before resuscitation. The changes in renal functional indices were evaluated after the treatments (data is not reported here). In order to find the optimum systematically administered dose of EPO with the same level of protection, 100, 200 and 300 IU/kg in 0.5 ml normal saline were injected intravenously and renal functional indices were measured after the treatments (data is not reported here). Since, the most effective dose of centrally and systemically administered EPO appeared in 2 IU/rat and 300 IU/ kg respectively, these doses were selected for the subsequent experiments.
In brief, in EPO-central group, just before resuscitation, 2 IU EPO was injected into the left lateral ventricle by injection needle (27-Gauge) through guide cannula. EPO took 30 s to be injected and in order to prevent backflow, the injection needle was left in place for further 1 min.
When EPO administered systemically, rats were injected EPO (300 IU/kg) within 10 min before resuscitation in an intravenous way.

Experimental groups
Animals were randomly allocated to four groups (n = 6 in each). (1) Sham group, guide cannula insertion into the left lateral ventricle by stereotaxic surgery and cannulation into the left femoral artery and vein were performed in this group but HS was not induced. (2) HS group, in this group, first, a guide cannula was placed into the left lateral ventricle by stereotaxic surgery and after a 7-day recovery, hemorrhagic shock was induced as mentioned above. (3) EPO-systemic group, the protocol was the same as the HS group except that animals in this group received 300 IU/kg recombinant human erythropoietin into the left femoral vein within 10 min immediately before resuscitation. (4) EPO-central group, the procedure was the same as the HS group except that 2 IU recombinant human erythropoietin was microinjected into the left lateral ventricle before resuscitation. Arterial oxygen saturation (SaO 2 ) was measured by placing oximeter sensor on rat tail during experiments. After a 3-hour postresuscitation period, rats were anesthetized by the use of intraperitoneal ketamine (100 mg/kg) and xylazine (10 mg/kg). Urine samples were stored for biochemical studies and kidney tissue samples were collected for exvivo evaluation. At the end of the sampling, animals were euthanized by CO 2 inhalation.

Renal functional indices assessments
Urine samples were collected from the urinary bladder and then centrifuged (3500 g for 10 min). Urine was used for N-acetyl-β-D-glucosaminidase (NAG) activity, creatinine (Cr) concentration and neutrophil gelatinase-associated lipocalin (NGAL) assessments.
Urinary NAG activity was determined by spectrophotometric assay [19]. Urinary Cr level was measured by using a clinical chemistry analyzer. To display less variability, urinary NAG level is suggested to be mentioned as a ratio to urinary Cr concentration [20].
Urinary NGAL concentration was measured by ELISA (Abcam, USA). Reactions were measured (at a wavelength of 450 nm) by the use of a microplate reader (BioTek Instrument, USA).

Renal oxidative stress assessments
Catalase (CAT) activity was evaluated according to Aebi method [21] in which H2O2 breakdown rate was accompanied by the diminished absorbance during 30 s at a wavelength of 240 nm. Reduced glutathione (GSH) level in the renal tissue samples was evaluated by spectrophotometric method which was based on the reaction of GSH with 5, 5′-dithiobis (2-nitrobenzoic acid). This reaction generates 2-nitro-5-thiobenzoic acid chromophore that is measurable at 412 nm [22].

Renal pro-inflammatory cytokines assessment
TNF-α and IL-6 mRNA expression were evaluated by Real-time PCR in the kidney tissue samples. Total RNA was extracted as instructed (RNeasy Mini Kit; Qiagen). Amount and purity of RNA were assayed by the use of the Nanodrop 2000 (Thermo-Scientific, USA). The cDNA synthesis was performed by 4 µg of total RNA using the PrimeScript RT Master Mix (Takara, Japan) based on the instructions and was stored at -80 °C. In the final step, Real-time PCR amplifications were performed using the ABI 7500 system.
The amount of mRNA for TNF-α and IL-6 genes were normalized by using HPRT-1. PCR primers for the three studied genes are as follows: TNF HPRT-1 (Anti-Sense Strand Sequence): AGC AAG TCT TTC AGT CCT GTCC.

Western blot analysis
Western blotting was performed in the way described in our previous study [2]. Briefly, PVDF membranes were incubated with rabbit anti active caspase-3 (Abcam), anti Beclin-1 (Cell Signaling Technology) anti LC3A/B (Cell Signaling Technology). After washing with TTBS, the membranes were individually incubated at room temperature with HRP conjugated secondary antibody (Abcam, 1:2000) for 1 h. The reactive bands were visualized using a chemiluminescence detection system (Image station 4000MM Pro, Kodak, USA). The blots were quantified using image j software.

Histological analysis
After fixing the kidney tissue samples in 10% paraformaldehyde, 4 μm paraffin-embedded kidney sections were stained using hematoxylin and eosin. At least 10 non-overlapping areas were randomly observed for the existence of tubular necrosis, tubular dilatations and infiltration of inflammatory cells.

Statistical analysis
The data were expressed as mean ± SEM. GenEX software was used to analyse the Real-time PCR data. The results were subjected to one-way ANOVA followed by Tukey's post hoc test using SPSS 22. P < 0.05 was considered as Significant.

Arterial oxygen saturation (SaO 2 )
A significant decrease in SaO 2 was seen in the HS animals in comparison with the Sham group (Table 1). EPO administration (systemic / central) significantly showed an increase in SaO 2 compared to the HS group.

Ratio of urinary NAG activity to urinary cr concentration
Urinary NAG-Cr ratio showed a significant rise in the HS group in comparison with the Sham animals (Table 1). EPO administration (systemic / central) just before resuscitation significantly decreased this ratio compared to the HS group.

Urinary NGAL concentration
HS led to a significant increase in urinary NGAL concentration compared to the Sham animals (Fig. 2). EPO administration in both route of systemic and central significantly prevented the rise in NGAL concentration compared to the HS rats.

Renal oxidative stress assessments
HS led to a significant decrease in renal CAT activity in comparison with the Sham group (Fig. 3A). EPO significantly prevented the reduction in renal CAT activity in both Table 1 The effect of centrally and systemically administered erythropoietin on NAG/Cr ratio and arterial O2 saturation in a model of fixed-volume hemorrhagic shock Data are presented as the mean ± standard error of mean for NAG/Cr ratio and as percent for arterial O2 saturation. *P < 0.05 compared to the Sham group. # P < 0.05 compared to the HS group HS led to a significant decrease in GSH levels compared to the Sham group (Fig. 3B). There was a significant rise in GSH amount in the EPO groups (systemic / central) compared to the HS group.

Renal pro-inflammatory cytokines assessment
Renal tissue TNF-α and IL-6 mRNA expression significantly increased in the HS rats compared to the sham ones (Fig. 4). EPO administration (systemic / central) significantly prevented the renal tissues TNF-α and IL-6 mRNA expression from rising compared to the HS rats.
In the autophagy studies, we found a significant increase in beclin1/β-actin in the HS group as compared to the Sham group. Also, treatment with EPO (systemic / central) led to a significant increase in beclin1/β-actin compared to the Sham group (Fig. 5).
There was a significant increase in the Lc3II/I protein level in the HS and EPO-systemic groups compared to the Sham group. Centrally administered EPO caused a significant increase in the Lc3II/I protein level in comparison with the Sham, HS and EPO-systemic groups (Fig. 5).

Renal histology
No significant histological changes were seen in the renal sections of the sham group (Fig. 6A). In the HS group, severe histological changes were reported compared to the sham group. There were tubular necrosis, tubular dilatations and infiltration of inflammatory cells as well as tubular obstruction (Fig. 6B). EPO administration (systemic / central) reduced the renal structural damages. In these groups, less cast formation in the tubules and less leucocytes infiltration were observed ( Fig. 6C and D).

Discussion
In this study, in order to assess the benefits of EPO on the reperfusion type injuries, it was given just before resuscitation. This time was chosen based on the idea that resuscitation is probably the main component of the organ injury in the treatment of HS states.
Hypoxic condition during hemorrhagic shock leads to the raise in metabolic acid concentrations due to increase in anaerobic systems activity and a decline in aerobic metabolism [23]. In the HS group, SaO 2 showed a significant decrease in comparison with the Sham group. This could be related to the low arterial pH, leading to a shift in the oxyhemoglobin dissociation curve to the right. EPO administration (central and systemic) provided significantly higher tissue blood flow and increased SaO 2 , suggesting improved tissue perfusion following resuscitation through the pathways that affect the lungs.
Several reports indicated that EPO significantly improved the renal functional indices [24,25]. NAG/Cr ratio and NGAL concentration are considered as predictive biomarkers in early kidney tubular injury [26,27]. Similarly, in our study, EPO treatment (central and systemic) significantly prevented the HS-induced rises in urinary NAG/Cr ratio and NGAL concentration.
A growing body of evidence indicates that ROS formation and inflammatory cascade activation are the commonly conventional pattern of ischemia-reperfusion injuries. It is known that HS results in excessive formation of ROS, which causes damage to the antioxidant defense system and finally rendering cells and tissues more susceptible to oxidant mediated injury. Furthermore, ROS act as key agents in the activation of a number of inflammatory factors such as TNF-α and IL-6 (4). These cytokines, in turn, augment the inflammatory response causing the injury to the cells and organs and their surroundings [28]. In the present study, a significant decrease in antioxidant defense system (CAT activity and GSH level) and a significant rise in pro-inflammatory genes expressions (TNF-α and IL-6) were observed in the kidney samples of HS rats compared to the Sham ones. EPO (either central or systemic) showed almost similar effects on anti-oxidant factors and inflammatory cytokines confirming the antioxidative and anti-inflammatory roles of that [29]. EPO improves blood-brain barrier integrity and significantly augments the antioxidant system. It is known to protect endothelial cells from ischemic injury and may participate in the recruitment of inflammatory cells. Thus, EPO is considered to improve brain hemodynamics following ischemic conditions. Several pathways are suggested to be involved in the expression of apoptotic proteins such as caspases. In this Fig. 5 The effect of centrally and systemically administered erythropoietin on renal caspase 3/β-actin, beclin1/β-actin, Lc3II/I protein levels during hemorrhagic shock. Data are presented as the mean ± SEM (n = 4). *P < 0.05 compared to the Sham group. # P < 0.05 compared to the HS group. $ P < 0.05 compared to the EPO-systemic group. HS: Hemorrhagic shock, EPO-systemic: Intravenously administered erythropoietin, EPO-central: Erythropoietin injection into the left lateral ventricle study, we examined the apoptosis occurrence by measuring the cleaved caspase 3 protein levels. Caspases are produced as inactive form and then cleaved to active form if needed. Active form of caspase 3 begins to activate other caspases which are involved in the apoptosis process [30]. In the present study, a significant increase was reported in cleaved caspase 3 level in the kidneys of the HS group, whereas, EPO (central and systemic) reduced apoptosis by a decline in the cleaved caspase 3 level.
In our study, after inducing hemorrhagic shock, beclin-1 level, as a main factor in autophagy, showed a significant increase compared to the Sham group. EPO administration caused a significant increase in beclin-1 level in both EPO-systemic and EPO-central groups compared to the Sham group. It seems that following EPO administration (systemic / central), Beclin-1 is separated from its inhibitor, Bcl-2, and it is now able to form autophagosomes and initiate autophagy process [31]. It should be noted that the activity and normal level of beclin-1 are important factors in tissue protection. It has been shown that excessive increase in beclin-1 level can cause autolysis by activating the Na + -K + -ATPase-regulated cell death [32]. In addition, in our study, LC3II/I level, as another factor of autophagy activity, in EPO-treated groups showed a significant increase compared to the HS group. LC3 proteins play a major role in the formation of phagophores and are considered to be autophagosome markers. The LC3I is combined with phosphatidylethanolamine to form LC3II, reflecting the formation of autophagosome in the autophagy process. Because LC3I level is almost constant under different conditions, it is considered as a control and LC3II level reflects autophagy activity [33]. Several studies have also shown improvement in the kidney damage after increased levels of LC3II. Xin et al. showed that excessive increase in blood glucose levels resulted in podocytes damage and a significant decrease in LC3II levels and autophagy activity [34]. In another study, following starvation and stress induction in AKI patients, autophagy activity increased by increasing LC3II/I level and resulted in improvement of kidney damage [35]. Since autophagy occurs in two selective and non-selective ways, it seems that hemorrhagic shock induction led cells to nonselective autophagy process in which cellular materials enter the lysosome in three forms of macroautophagy, microautophagy and chaperone-mediated autophagy. During these three steps, first, the damaged cellular components are separated by a membrane made by the endoplasmic reticulum, then, they enter the lysosome and are enclosed by a lysosome membrane. Next, substrate is identified by chaperones and, finally, the substrate is separated from its chaperone and used by the cell [36,37]. The EPO-systemic and EPO-central groups showed increases in autophagy activity which seems that erythropoietin is able to direct cells towards selective autophagy in which cell structure and function are protected.
Our study confirmed that EPO improved the kidney function, peripheral blood flow and acted as an anti-oxidative, anti-apoptotic agent and autophagy activator. We, then, tried to compare the systemically and centrally administered EPO in order to find out the optimum doses with similar effects. As mentioned above, the benefits of centrally administered EPO were appeared at the dose of 2 IU/rat, whereas, in the systemic route, the effective dose was 300 IU/kg which was roughly fifty times higher than that of central administration (2 IU/rat). It seems that the benefits of centrally microinjected EPO was consequently due to direct central mechanisms rather than its diffusion to the periphery because of very low dose EPO administration. Brain as a commander is able to detect and modulate any kind of environmental stress to keep living organism safe. Hence, one of these responses may be to secrete low doses of EPO to regulate remote organs such as the renal system. Some humoral and nervous mechanisms are considered to be involved in the crosstalk between the brain and remote organs [38]. EPO is likely to play a major role in this crosstalk as a neuromodulator or neuromediator. According to international neuromodulation society, neuromodulators are defined as the agents that are able to change the nerve activity through electrical stimulation or release of chemical agents to the specific neurological sites in the body. Neuromediators are chemical agents which are naturally produced in the body that transmit messages between neurons. Based on the above definitions, it should be noted that a neuromodulator or neuromediator exerts its effects through postsynaptic receptor configuration, G-protein coupling and a second messenger system depending on the brain region and it seems that EPO is not an exception. EPO may exert its protective effects on remote organs by activating second messenger systems. It seems that EPO binding to specific cell surface receptors results in activation of several downstream signaling pathways. In our recent study, Akt/eNOS/NO pathway was evaluated to be effective in the kidney protection by centrally injected erythropoietin [13]. The other probable mechanism of EPO protection may be the alteration of renal sympathetic nerve activity (RSNA). To maintain sufficient oxygenation of renal tissues in response to hemorrhage conditions, RSNA was shown to decrease [39]. It may be as a result of EPO modulation on RSNA and consequent hemodynamic changes to adapt renal oxygen availability. Further studies are needed to evaluate these assumptions.

Conclusions
Considering the fact that the effective dose of systemic erythropoietin (300 IU/kg) was roughly 50 times higher than that of central administration (2 IU/rat), centrally administered erythropoietin was accompanied by more advantageous consequences than that of systemic way. Erythropoietin is likely to act as a neuromodulator or neuromediator in the central protection of organs including the kidneys.