Erythropoietin Markedly Attenuates Brain Infarct Size and Improves Neurological Function in the Rat

Ethics

All animal experimental procedures were approved by the Institute of Animal Care and Use Committee at our hospital and performed in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication 85-23. Washington, DC: National Academy Press; revised 1996).

Animal Model of Acute Ischemic Stoke, Protocol, and Procedure

Pathogen-free, adult male Sprague-Dawley rats, weighing 250 to 300 g (Charles River Technology, BioLASCO Taiwan Co, Ltd, Taiwan) were used in this study (Fig. 1).

OPEN IN VIEWER

Figure 1

Induction of acute IS in the rat. A, Exposure of the brain after removing part of the skull. Note the nylon filament over the distal portion of left internal carotid artery (LICA) (black). B, Inserted nylon filament clearly identified in distal portion of LICA with selective occlusion of left middle cerebral artery (MCA). C, Clearly identifiable brain infarct region on gross inspection. D, Coronal sections showing infarct regions (white) after TTC staining.

After being anesthetized by intraperitoneal injections of chloral hydrate (35 mg/kg), the rats were placed in a supine position, with the ventral side of the neck shaved. The procedures were performed on a warming pad at 37°C, and the body temperature of each animal was measured by an anal probe. Under sterile conditions, the left common carotid artery (LCCA) was exposed.

After exposure of the LCCA through a vertical neck incision, a small incision was made on the LCCA through which a nylon filament (0.28 mm in diameter) was advanced into the distal left internal carotid artery for occlusion of left middle cerebral artery to induce brain infarction of its supplying region (Fig. 1). Four hours after occlusion, the nylon filament was removed, followed by closure of the muscle and skin in layers. This was a ischemia-reperfusion stroke model. The animals were allowed to remain on the warming pad for recovery under care.

In Vivo Experimental Protocol and Corner Test

Ten healthy rats served as normal controls (group 1). The rats with acute IS were divided into group 2 (acute IS treated with 1 mL intraperitoneal physiological saline at 0, 12, and 24 hours after IS induction, n = 15) and group 3 (acute IS plus intraperitoneal EPO 5000 IU/kg given at 0, 12, and 24 hours after IS induction, n = 15). The EPO dosage to be used was based on previous literature. ²⁶ The regimen and timing of EPO therapy in the present study were designed to determine the effect of such a treatment strategy on attenuating the BIA. The sensorimotor functional test (Corner test) was done for each rat on days 0, 7, 14, and 21 after acute IS induction as previously described. ²⁷ Briefly, the rat was allowed to walk through a tunnel and then into a corner, the angle of which was 60 degrees. To exit the corner, the rat could turn to either left or right. The results were recorded by a technician who was blind to treatment and nontreatment groups. This test was repeated 10 to 15 times with at least 30 seconds between each trial. We recorded the number of right and left turns from 10 successful trials for each animal and used the results for statistical analysis.

Cellular Proliferation Test

To evaluate whether EPO treatment promotes cellular proliferation in the BIA, 5-bromo-2-deoxyuridine (BrdU) was intravenously given in all 3 groups of animals on days 3, 5, 7, 9, and 12 after acute IS induction for labeling the proliferating cells.

Ex Vivo Determination of the Effects of EPO on Angiogenesis

For determining the effect of EPO on angiogenesis, 12 extra rats received bone marrow aspiration under general anesthesia. After careful separation of the ligament from the patella, a 0.2-mm-diameter electric rotablator was used to drill directly into the femoral bone from the distal end. A sterile 22-gauge needle syringe was then used to aspirate the bone marrow.

The bone marrow aspirates from the rats were divided into 2 groups (n = 6 for each group). The first group underwent centrifugation, and the bone marrow-derived mononuclear cell (BMDMNC) layer was harvested for study, whereas only the red blood cells were removed from the second group with the rest of the contents preserved.

The harvested BMDMNCs with a quantity of 1.2 × 10⁶ from each animal were cultured in 10-cm culture dish using M199 medium with 20% fetal bovine serum. By day 7, fetal bovine serum content was reduced to 10%. Recombinant human EPO was used in this study (epoetin beta 5000 IU/amp, equivalent to 14.5 μg in 0.3 mL; Roche Co, Basel, Switzerland). Erythropoietin of increasing concentrations (0 [ie, vehicle of 0.1% dimethyl sulfoxide or phosphate-buffered saline [PBS], 10, and 30 IU/mL) (n = 6 for each concentration) was added into the culture dish on days 0, 7, 10, and 12.

After 14-day cell culturing, the BMDMNCs were plated in 96-well plates at 1.0 × 10⁴ cells/well in 150 μL serum-free M199 culture medium mixed with 50 μL cold Matrigel (Chemicon International, Billerica, MA) for 6-hour incubation at 37°C in 5% CO₂. Three random microscopic images (200×) were taken at each well for counting cluster, tube, and network formations, with the mean values obtained.

Specimen Collection

Rats in groups 1, 2, and 3 were killed on day 21 after IS, and brain in each rat was rapidly removed and immersed in cold saline. For immunohistofluorescence (IHF) study, the brain tissue was rinsed with PBS, embedded in OCT compound (Tissue-Tek; Sakura, Alphen aan den Rijn, the Netherlands) and snap-frozen in liquid nitrogen before being stored at −80°C. For immunohistochemical (IHC) staining, brain tissue was fixed in 4% formaldehyde and embedded in paraffin.

Measurement of Brain Infarct Area and Area at Risk

To evaluate the impact of EPO treatment on brain infarction, coronal sections of the brain were obtained from 5 extra animals in groups 2 and 3 (n = 5 for each group) as 2-mm slices. Each cross section of brain tissue was then stained with 2% TTC (3,5-triphenyl-2H-tetrazolium chloride) (Alfa Aesar, Ward Hill, MA) for BIA analysis. Briefly, all brain sections were placed on a tray (Brain Acrylic Matrix, RBS-A1-C; Stoelting, Wood Dale, IL) with a scaled vertical bar to which a digital camera was attached. The sections were photographed from directly above at a fixed height. The images obtained were then analyzed using Image Tool 3 image analysis software (version 3.0; University of Texas, Health Science Center, San Antonio, TX).

Infarct area was observed as either whitish or pale yellowish regions, whereas the area at risk was defined as pale pinkish discolored region surrounding the infarct tissue. Three to five 2-mm slices of the brain were obtained from each rat. The mean percentages of infarct area and area at risk were then obtained by dividing the area with total cross-sectional area of the brain.

TUNEL Assay for Apoptotic Nuclei

For each rat, 6 sections of BIA were analyzed by an in situ Cell Death Detection Kit, AP (Roche) according to the manufacturer's guidelines. Three randomly chosen high-power fields (HPFs) (×400) were observed for terminal deoxynucleotidyl transferase-mediated 2'-deoxyuridine 5'-triphosphate nick-end labeling (TUNEL)-positive cells. The mean number of apoptotic nuclei per HPF for each animal was obtained by dividing the total number of cells with 18.

Immunohistochemical Staining for Cellular Proliferation and Glial Fibrillary Acid Protein

Paraffin sections (5 μm thick) with BIA were obtained from each rat. To block the action of endogenous peroxidase, the sections were initially incubated with 3% hydrogen peroxide for 15 minutes and then further processed using Beat Blocker Kit (#50-300; Invitrogen, Carlsbad, CA) with immersion in solutions A and B for 30 and 10 minutes at room temperature, respectively. Rabbit polyclonal antibody (1:500 dilution at 4°C overnight) against glial fibrillary acid protein (GFAP) (Dako, Glostrup, Denmark) and monoclonal antibody (1:200 dilution at 4°C overnight) against BrdU (Sigma, St. Louis, MO) were used as primary antibodies. The anti-rabbit horseradish peroxidase (HRP; Zymed, Carlsbad, CA) (1:3 dilution at room temperature for 10 minutes) for GFAP and anti-mouse HRP (Zymed) (1:3 dilution at room temperature for 10 minutes) were used as secondary antibodies, followed by application of SuperPicTure Polymer Detection Kit (Zymed) for 10 minutes at room temperature. Finally, the sections were counterstained with hematoxylin. For negative control experiments, primary antibodies were omitted.

Western Blot Analysis for CXCR4 and Stromal Cell-Derived Factor 1 in BIA

Equal amounts (60 kg) of protein extracts from BIA were loaded and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 12% to 13% acrylamide gradients. After electrophoresis, the separated proteins were transferred electrophoretically to a polyvinylidene difluoride membrane (Amersham Biosciences, Piscataway, NJ). Nonspecific proteins were blocked by incubating the membrane in blocking buffer (5% nonfat dry milk in T-TBS containing 0.05%Tween 20) overnight for CXCR4 and 1 hour for stromal cell-derived factor (SDF-1), respectively. The membranes were incubated with the indicated primary antibodies (CXCR4, 1:1000 [Abcam, Cambridge, MA]; actin, 1:10,000 [Chemicon]; SDF-1, 1:1000 [Cell Signaling, Danvers, MA]) for 1 hour at room temperature for CXCR4 and overnight at 4°C for SDF-1, respectively. Horseradish peroxidase-conjugated anti-rabbit immunoglobulin G (1:2000; Cell Signaling) was applied as the second antibody for 1 hour for CXCR4 and 45 minutes for SDF-1 at room temperature, respectively. The washing procedure was repeated 8 times within an hour, and immunoreactive bands were visualized by enhanced chemiluminescence (Amersham Biosciences) and exposure to Biomax L film (Kodak, Rochester, NY). For quantification, digitized enhanced chemiluminescence signals were analyzed using Labwork UVP software (Waltham, MA).

Real-Time Quantitative Polymerase Chain Reaction Analysis

Real-time polymerase chain reaction (PCR) was conducted using LighCycler TaqMan Master (Roche) in a single capillary tube, according to the manufacturer's guidelines for individual component concentrations. Forward and reverse primers were each designed based on individual exons of the target gene sequence to avoid amplifying genomic DNA.

During PCR, the probe was hybridized to its complementary single-strand DNA sequence within the PCR target. As amplification occurred, the probe was degraded because of the exonuclease activity of Taq DNA polymerase, thereby separating the quencher from reporter dye during extension. During the entire amplification cycle, light emission increased exponentially. A positive result was determined by identifying the threshold cycle value at which reporter dye emission appeared above background.

IHF Analysis for CXCR4, SDF-1, Doublecortin, and von Willebrand Factor

Serial cryosections (7 μm thick) with an average distance of 5 μm apart were collected from the BIA. The sections were fixed in acetone for 15 minutes at −20°C. For reducing the background, 200 μL of signal enhancer was used for blocking nonspecific signals at room temperature for 30 minutes. Immunohistofluorescence staining was performed using primary antibody (rabbit polyclonal antibody 1:200 dilution, at 4°C, overnight) (Santa Cruz, Santa Cruz, CA) for CXCR4, followed by the addition of anti-rabbit Alexa Fluor 488 fluorescein isothiocyanate (Molecular Probes, Carlsbad, CA) secondary antibody (1:200 dilution at room temperature for 30 minutes). Additionally, rabbit polyclonal antibody (1:500 dilution at 4°C overnight) (Santa Cruz) was used as primary antibody for SDF-1, followed by the addition of anti-rabbit Alexa Fluor 594 Rodamin (Molecular Probes) secondary antibody (1:200 dilution at room temperature for 30 minutes). Moreover, goat polyclonal antibody (1:50 dilution, at 4°C overnight) (Santa Cruz) was used as primary antibody to recognize doublecortin, followed by anti-goat Alexa Fluor 568 Rodamin (Molecular Probes) secondary antibody (1:200 dilution at room temperature for 30 minutes). Furthermore, rabbit polyclonal antibody (1:200 dilution at 4°C overnight) (Chemicon) was used as primary antibody against von Willebrand factor (vWF), followed by anti-rabbit Alexa Fluor 488 fluorescein isothiocyanate (Molecular Probes) secondary antibody (1:200 dilution at room temperature for 30 minutes). For negative control experiments, the primary antibodies were omitted. The sections were counterstained with DAPI (4',6-diamidino-2-phenylindole) (dilution 1/500) (Sigma) to identify cellular nuclei that represented the cell number.

Oxidative Stress of BIA

The Oxyblot Oxidized Protein Detection Kit was purchased from Chemicon (S7150). Western blot analysis was performed according to the manufacturer's guidelines for individual procedures. On each gel, a standard control was loaded.

Small-Vessel Density in BIA

Immunohistochemical staining of small blood vessels (ie, diameters ≤15 μm) was performed with anti-α-SMA (1:400) as primary antibody at room temperature for 1 hour, followed by washing with PBS thrice. The anti-mouse HRP-conjugated secondary antibody was then added and incubated for 10 minutes, followed by washing with PBS thrice. Then DAB (3,3-diaminobenzidine) (0.7 g/tablet) (Sigma) was added and incubated for 1 minute, followed by washing with PBS thrice. Finally, after hematoxylin treatment for 1 minute as a counterstain for nuclei, the sections were washed twice. Three coronal sections of the brain were analyzed in each rat. For quantification, 3 randomly selected HPFs (200×) were analyzed in each section. The mean number of per-HPF for each animal was then determined by summation of all numbers divided by 9.

Statistical Analysis

Data were expressed as mean (SD). Statistical analysis was adequately performed by unpaired Student t test or analysis of variance, followed by Scheffé multiple-comparisons post hoc test. SAS statistical software for Windows version 8.2 was used (SAS institute, Cary, NC). P < 0.05 was considered statistically significant.

EPO Enhanced in Vitro Angiogenesis

Figure 2 (A-I) shows the effect of EPO on angiogenesis. Surprisingly, the results demonstrated no significant difference in the number of cluster, tubular, or network formation from isolated BMDMNCs among concentrations of 0, 10, and 30 IU/mL of EPO treatment on Matrigel assay after 6-hour cell culturing when other bone marrow-derived contents were not included (Fig. 2A-C). On the other hand, not only was angiogenesis remarkably enhanced in the presence of whole aspirated bone marrow-derived contents (Fig. 2D-F), but also dose-dependent effects of EPO were observed in tubular and network formations (Fig. 2G-I).

OPEN IN VIEWER

Figure 2

Comparison of angiogenesis on Matrigel (100×) after 6-hour cell culture (1.0 × 10⁴ cells) after prior 14-day culture of BMDMNCs in endothelial growth medium containing EPO without (A-C) and with (D-F) bone marrow-derived contents (n = 6 in each group). No significant difference in the number of cluster, tubular, or network formation from isolated BMDMNCs among concentrations of (A) 0 IU/mL, (B) 10 IU/mL, and (C) 30 IU/mL of EPO treatment on Matrigel assay after 6-hour cell culturing in the absence of other bone marrow-derived contents. However, remarkably enhanced angiogenesis noted in the presence of whole aspirated bone marrow-derived contents (D-F) with dose-dependent effects of EPO observed in tubular and network formations (G-I).

Mortality in the Study Period

The numbers of deaths that occurred during the study period in groups 1, 2, and 3 were 0, 3, and 3, respectively. Most fatalities occurred within 24 hours, and the dead animals were not included in the study. Because a minimum of 10 rats in each group was needed to meet the requirement for significant statistical analysis, in total 10, 18, and 18 rats were used in groups 1, 2, and 3, respectively, in the current study.

EPO Limited Brain Infarct Size and Improved Recovery of Neurological Functional

TTC staining on day 21 after acute IS showed that the BIA was markedly larger, whereas the area at risk was remarkably lower in IS animals without treatment (group 2) compared with those having received EPO therapy (group 3) (Fig. 3A-D). Moreover, the summation of the damage areas (ie, BIA + area at risk) was notably higher in group 2 than in group 3 (Fig. 3E). These findings imply that not only did EPO significantly reduce BIA after IS, but it also enhanced the viability of ischemic brain tissue (ie, area at risk), which can potentially recover after further effective treatment.

OPEN IN VIEWER

Figure 3

Ratios of infarct area and area at risk to total coronal section area using TTC staining. A and B, Identification of gross infarct area (red arrows) and area at risk (blue arrows) without and with EPO treatment, respectively. C, Significantly lower ratio of infarct area to total coronal section area in EPO-treated group compared with the diseased controls. D, Conversely, notably higher area at risk in EPO-treated group than the group without EPO treatment. E, Notably lower total ischemic area (ie, infarct plus at-risk areas) in animals with EPO treatment than in those without (n = 5 per group).

Corner test showed the attainment of a steady state of neurological functional impairment on day 3 after acute IS in both groups 2 and 3 (Fig. 3F). Significant improvement in neurological function became apparent in group 3, but not in group 2, compared with group 1 on day 14. Further substantial improvement in group 3 was noted on day 21 after acute IS.

EPO Attenuated Inflammatory Response, Apoptosis, and Oxidative Stress

On day 21 after acute IS induction, Bcl-2 mRNA expression, an index of antiapoptosis, was significantly higher in group 3 than in group 2 (Fig. 4D), whereas mRNA expressions of Bax (Fig. 4E) and caspase 3 (Fig. 4F) were significantly lower in group 3 than in group 2. Additionally, TUNEL assay showed that the number of apoptotic nuclei was remarkably lower in group 3 than in group 2 (Fig. 4G-J).

OPEN IN VIEWER

Figure 4

mRNA expressions and apoptotic nuclei in BIA (n = 10 per group). Significantly higher mRNA expressions of (A) IL-18, (B) TLR-4, and (C) PAI-1 in group 2 than in group 3 and normal controls (group 1), and notably higher in group 3 than in group 1.

On day 21 after acute IS induction, mRNA expressions of interleukin 18 (IL-18), toll-like receptor 4 (TLR-4), and plasminogen activator inhibitor 1 (PAI-1) in BIA, indexes of inflammation, were significantly lower in group 3 than in group 2. These findings indicate that EPO therapy attenuated inflammatory response.

On day 21, Western blotting (Fig. 5) demonstrated no significant difference in oxidative stress index in mitochondria between group 1 (normal control) and group 3, whereas it was significantly higher in group 2 than in group 3. These findings indicate that EPO exerted both antiapoptotic and anti-inflammatory actions in the brain after IS.

OPEN IN VIEWER

Figure 5

Western blotting showing notably increased oxidative index, protein carbonyls, in BIA of group 2 compared with groups 1 and 3 on day 21 after acute IS (upper panel), with quantification results of each group (n = 10) shown (lower panel).

EPO Enhanced in Vivo Angiogenesis and Neurogenesis

Immunohistochemical staining demonstrated that the number of cells positive for CXCR4 (Fig. 6A-D), a surface cell marker of EPCs, and SDF-1, a chemokine for attraction of EPCs having CXCR4 receptor (Fig. 6E-H), was significantly higher in group 3 than in group 2, suggesting an enhancement of circulating EPCs homing to ischemic area of the brain after EPO treatment. Consistently, Western blot analysis revealed significantly higher protein expressions of CXCR4 (Fig. 6I) and SDF-1 (Fig. 7J) in group 3 than in group 2.

OPEN IN VIEWER

Figure 6

Results of IHF staining (400×) showing significantly lower number of CXCR4-positive cells (red arrow) in group 1 (A) than in groups 2 (B) and 3 (C), and notably lower in group 2 than in group 3 (D). Immunohistofluorescence staining (400×) also demonstrating significantly lower number of SDF-1-positive cells (yellow arrows) in group 1 (E) than in groups 2 (F) and 3 (G), and notably lower in group 2 than in group 3 (H). Western blotting showing notably lower CXCR4 (I) and SDF-1 (J) protein expressions in group 1 than in groups 2 and 3, and notably lower in group 2 than in group 3. n = 10 per group. Scale bars in right lower corner represent 20 μm.

OPEN IN VIEWER

Figure 7

Results of IHF staining (D) (400×) showing significantly lower number of doublecortin-positive cells (yellow arrows) in group 1 (A) than in groups 2 (B) and 3 (C), and notably lower in group 2 than in group 3. Immunohistofluorescence staining (H) (400×) demonstrating significantly lower number of vWF-positive cells (yellow arrows) in group 1 (E) than in groups 2 (F) and 3 (G), and notably lower in group 2 than in group 3 (H). n = 10 in each study group. Scale bars in right lower corner represent 20 μm.

The expression of doublecortin, a marker of migrating neuroblasts, was significantly upregulated in group 3 compared with group 2 (Fig. 7A-D). Additionally, IHC staining showed that the expression of vWF, a marker of endothelial cells of cerebral blood vessels, was significantly increased in group 3 than in group 2 (Fig. 7E-H). Moreover, IHC staining also revealed an increased number of BrdU-positive cells (Fig. 8A-D) and small blood vessels (Fig. 8E) in group 3 than in group 2, implying an increased cellular differentiation and proliferation after EPO treatment. Furthermore, the number of arterioles (≤15 μm in diameter) in BIA was substantially lower in group 2 than in groups 1 and 3 on IHC staining (Fig. 9A-D). All of these findings indicate an EPO-induced enhancement in neurogenesis and vasculogenesis after acute IS.

OPEN IN VIEWER

Figure 8

Immunohistochemical staining (D) (400×) showing markedly lower number of BrdU-positive cells (yellow arrows) in group 1 (A) than in groups 2 (B) and 3 (C). Also note the significantly lower number of BrdU-positive small vessels (red arrows) (E) in groups 1 and 2 than in group 3. n = 10 in each group. Scale bars in right lower corner represent 20 μm.

OPEN IN VIEWER

Figure 9

Small vessels (diameters ≤15 μm) (red arrows) quantification for each group (n = 10) on day 21 after acute IS (D). Identification of blood vessel distribution in BIA using α-SMA IHC staining, showing notably higher small-vessel number in groups 1 (A) and 3 (C) than in group 2 (B) (200×), and significantly higher in group 3 than in group 1. Scale bars in right lower corner represent 50 μm.

EPO Reduced GFAP Expression in Infarcted Brain

Immunohistochemical staining showed that GFAP expression, the principal intermediate filament of mature astrocytes, was notably lower in group 3 than in group 2 (Fig. 10A-D), suggesting reduced IS-induced gliosis after EPO treatment.

OPEN IN VIEWER

Figure 10

Immunohistochemical staining (D) (200×) showing significantly higher number of GFAP-positive cells (red arrows) in group 2 (B) than in groups 1 (A) and 3 (C), and notably higher in group 3 than in group 1. n = 10 per group. Scale bars in right lower corner represent 50 μm.

EPO Therapy Improved Neurological Function through Enhancement of Angiogenesis

The present study demonstrated that EPO failed in enhancing angiogenesis of isolated BMDMNCs in vitro (Fig. 2). Conversely, EPO can enhance angiogenesis from bone marrow-derived cells in the presence of whole bone marrow aspirates. These findings suggest that angiogenesis enhancement by EPO could be, at least in part, attributed to the presence of hematopoietic stem cells and bone marrow-derived factors rather than BMDMNCs alone. The results are consistent with those of a recent study demonstrating enhancement of angiogenesis by EPO only in the presence of vascular endothelial growth factor and erythroid cells. ²⁸ Therefore, the current study further underscores the essential role of hematopoietic stem cells in angiogenesis. Of importance in the present study is the fact that our in vitro study may shed light on the importance of a bone marrow environment for the participation of EPO in angiogenesis^19–22 in mammalian tissue.

Stromal cell-derived factor 1 is an EPC chemokine participating in the mobilization, incorporation, homing, survival, proliferation, and differentiation of stem cells.^29–31 Recently, SDF-1 and its receptor CXCR4 are proven crucial in bone marrow retention of hematopoietic stem cells, cardiogenesis, angiogenesis, and recruitment of EPCs into ischemic tissue.^31–34 Moreover, the expression of this chemokine is markedly upregulated in animal models early after myocardial infarction.^35,36 In the present study, both Western blot and IHC staining identified that both CXCR4 and SDF-1 expressions were significantly increased in animals with acute IS as compared with the normal controls. Importantly, CXCR4 and SDF-1 in BIA were found to be markedly increased after EPO treatment. These findings, therefore, in addition to supporting the interaction between SDF-1 and CXCR4 in angiogenesis,^29–36 further clarified the crucial role of EPO in enhancing the formation of SDF-1 and CXCR4 in ischemia area. Study has recently demonstrated that administration of SDF-1 to an animal model of critical limb ischemia enhances the concentrations of EPCs within the ischemic tissue and augments tissue reperfusion. ³³ Taking this finding ³³ into consideration, our results suggest that enhancement of the number of CXCR4-positive cells in BIA by EPO may be, at least in part, through strengthening SDF-1 chemokine expression in the BIA.

Besides the findings of upregulated expressions of CXCR4 and SDF-1 in the BIA, another principal finding was that EPO therapy also markedly increased the cellular expression of vWF, which is a marker of endothelial cells. Importantly, EPO therapy also increased the number of small vessels, an index of neovascularization, in BIA. Taken together, the improved neurological function and reduced BIA in the present study could be explained, at least in part, by the positive impact of angiogenesis.

EPO Therapy Improved Neurological Function through Enhancement of Neurogenesis and Vasculogenesis

The current study demonstrated that EPO therapy significantly increased the number of doublecortin-positive cells in BIA. Additionally, BrdU uptake in BIA, an index of cellular differentiation and proliferation, was substantially promoted after EPO therapy. Of interest is that BrdU uptake by vascular endothelial cells (Fig. 5) in BIA, an index of neovascularization, was frequently observed after EPO therapy. Accordingly, the results of the present study suggest that EPO therapy enhances both neurogenesis and vasculogenesis. These findings could be another explanation for the reduction in BIA and improvement in neurological function. In fact, a previous study has also demonstrated that treatment of stroke with EPO enhances neurogenesis and angiogenesis and improves neurological function in rats. ²⁶ Therefore, our findings reinforce the findings of that study. ²⁶

EPO Therapy Improved Neurological Function through Attenuating Inflammatory Response, Oxidative Stress, and Apoptosis

Studies have previously shown that EPO exerts both anti-inflammatory^19,23 and antioxidant^24,25 effects via inhibiting the production of proinflammatory mediators. In the present study, the mRNA expressions of IL-18, TLR-4, and PAI-1, indexes of inflammatory reactions, were markedly upregulated in rats after acute IS. In addition, the mRNA expressions of Bax and caspase 3, 2 apoptosis markers, were notably increased, whereas mRNA expression of Bcl-2, an index of antiapoptosis, was notably reduced in rats after acute IS. Furthermore, TUNEL assay and IHC staining demonstrated markedly increased apoptotic nuclei and GFAP, respectively, after acute IS. Moreover, Western blot showed remarkably upregulated oxidative stress in rats after acute IS. However, these biomarkers were significantly reversed by EPO therapy. Therefore, our findings, in addition to collaborating with those of previous studies,^19,23–25 further account for the improvement in neurological outcome after EPO treatment in rats after acute IS through attenuating inflammatory response, oxidative stress, and apoptosis.

The present study has limitations. First, although the mechanisms underlying the therapeutic potential of EPO in attenuating BIA and enhancing sensorimotor functional recovery have been carefully elucidated, the precise mechanistic basis of EPO treatment for acute IS may be more complex. Second, although BIA was significantly attenuated, the area at risk remained prominent after EPO therapy. Recently, stem cell therapy has seemed to be beneficial in improving ischemia-induced organ dysfunction.^37,38 Taking these findings and the results of our study into consideration, we suggest that a more promising therapeutic regimen of combining EPO with stem cell therapy may be another potential option for acute IS in the future. Finally, absence of physiological monitoring (cranial temperature, blood gases, etc) is a major weakness of the current study.

In conclusion, EPO therapy limited brain infarct size and improved neurological function in rats after acute IS through enhancement of angiogenesis/vasculogenesis and neurogenesis as well as its anti-inflammatory and antiapoptotic effects.