Comparative Evaluation of Korean White Ginseng and Red Ginseng Ecacies in a Mouse Model of Ischemia/Reperfusion-Induced Stroke

Background: Stroke is a condition characterized by brain tissue damage owing to a decrease in the brain's oxygen supply due to blocked blood vessels, and 80% of all strokes are classied as cerebral infarction. Notably, the incidence rate tends to increase with increasing age. In this study, we compared the ecacy of white ginseng (WG) and red ginseng (RG) extracts (WGex and RGex, respectively) in an ischemic stroke mouse model and conrmed the underlying mechanisms of action. Methods: Mice were orally administered WGex or RGex 1 h before performing middle cerebral artery occlusion (MCAO) for 2 h; the size of the infarct area was measured 24 h after MCAO. The neurological decit score was evaluated, the ecacy of the two drugs was compared, and the mechanism of action was conrmed using methods such as tissue staining and protein quantication. Results: In the MCAO-induced ischemic stroke mouse model, WGex and RGex showed neuroprotective effects in the cortical region, with RGex demonstrating a generally stronger ecacy than WGex. Furthermore, it was conrmed that ginsenoside Rg1, a representative indicator substance, was not involved in mediating the effects of WGex and RGex. Conclusion: WGex and RGex inhibited brain injury attributed to ischemia/reperfusion, with RGex revealing a more potent effect. At 1,000 mg/kg body weight, only RGex reduced cerebral infarction and edema, and both anti-inammatory and anti-apoptotic pathways were involved in mediating these effects.

leading cause of morbidity and mortality in modern society [7][8][9][10][11]. Ischemic stroke can be characterized by arterial embolic or thrombotic occlusion, resulting in brain in ammation and cell death. Notably, white blood cells delivered via the blood are known to in ltrate the brain to activate in ammation, and then cell death signals are activated by the permeability of the outer mitochondrial membrane [12][13][14].
Thrombolytic drugs such as anticoagulants and antiplatelets are used as therapeutic agents to treat ischemic stroke, a condition that can cause morbidity and disability and impose a heavy burden on patients by increasing medical expenses and necessitating rehabilitation and long-term care [15,16].
Therefore, further research on the prevention and treatment of ischemic stroke is urgently needed [17,18].
Typically, there are two types of surgical procedures to induce ischemic stroke in rodents, by temporarily blocking the middle cerebral artery (MCA): Koizumi's and Longa's methods [19][20][21][22]. Koizumi's method inserts mono laments into the MCA via the common carotid artery (CCA), whereas Longa's method inserts laments via the external carotid artery (ECA) [23]; however, Koizumi's method is commonly employed to establish a stroke model in mice.
Although several studies have reported [24][25][26] that ginseng or its constituents have the potential to inhibit brain damage, its effects, as well as its mechanisms of action, in animal models of cerebral infarction have rarely been reported. In this study, we investigated the differences and mechanisms underlying the protective effects of WG and RG on cerebral injury induced by ischemia/reperfusion of MCA blood ow.

Animals
Six-week-old C57BL/6 male mice (Samtako Bio, Osan, Korea), weighing 20-22 g, were housed in polypropylene cages at 24 ± 4℃, under a 12 h light/dark cycle for 1 week prior to the experiment. The experimental animals were fed a standard pellet diet and water ad libitum. All experimental procedures performed in the present study were approved by the Ethics Committee for Animal Care and Use at Pusan National University (Approval No. PNU 2019-2485), certi ed by the Korean Association of Laboratory Animal Care.
Preparation and high-pressure liquid chromatography (HPLC) analysis of WG and RG extracts Standardized WG and RG, manufactured from a 5-year-old ginseng sample, were obtained from Gwangmyung Natural Pharmaceutical Co. (Busan, Korea), and voucher specimens (No. 201KWG and 201KRG) were deposited at the Herbarium of School of Korean Medicine, Pusan National University. For both WG and RG, 200 g of each were nely ground and extracted with 10 times the volume of 99% methanol at 25℃ for 3 days; this was repeated three times. After ltration using lter paper (Advantec, USA), methanol was removed using a vacuum evaporator (Eyela, Japan) at 45℃, and then stored at -20℃ until use. The nal lyophilized WG extract (WGex) and RG extract (RGex) from WG and RG weighed 24.2 g and 25.3 g, yielding 12.1% and 12.7%, respectively.
For quality assurance of WGex and RGex, each extract was subjected to HPLC analysis. To obtain a ngerprint of extracts used in this experiment, we determined the presence of ginsenoside Rg1, a main component of ginseng, using an HPLC system (Shimadzu, Kyoto, Japan) consisting of an LC-20AD pump, a SIL-20A autosampler, an SPD-M20A detector, and a CTO-20A column oven. The samples were separated using a YMC-Triart C18 column. The mobile phase and elution systems are listed in Table 1.

Induction of transient middle cerebral artery occlusion (tMCAO)
In the present study, Koizumi's method was modi ed at our laboratory, and the modi ed method was performed in mice; detailed protocols and materials have been previously reported [27]. In brief, for left hemispheric MCAO, anesthesia was induced by inhalation of 1.5-2% iso urane in N 2 O/O 2 (70%/30%) until mice showed no response to mechanical stimuli. Following the induction of anesthesia, the mice were placed on a heating pad (Harvard Instruments, Boston, MA, USA). During the tMCAO surgery, the body temperature of mice was monitored using a rectal temperature probe and maintained at 37 ± 1℃. The surgical site, as well as the chest area of mice, were disinfected with povidone and 70% ethyl alcohol.
Then, the left side of the skull, near the ear, was exposed using surgical mesh, and an optic ber was attached to the skull to monitor the relative cerebral blood ow (rCBF). A midline incision was made on the neck, and soft tissues over the trachea were gently retracted. The ECA, distal CCA (dCCA), and proximal CCA (pCCA) were loosely knotted with 4-0 suture silk, and the ECA and CCA were carefully isolated from the vagus nerve. The ECA and dCCA were ligated before clamping of the internal carotid artery (ICA). Then, an 8-0 mono lament suture line (Ethicon, CA, USA) coated with silicon was inserted into the MCA through the perforated CCA. For the tMCAO operation, the mono lament was retracted when the predetermined occlusion time ended in the present experiment (2 h tMCAO). To complete the surgical process, the incised region was sutured and disinfected, and the mice were maintained in a recovery cage for 30 min. All surgeries were performed using laser Doppler owmetry (Moor Instruments Ltd., Devon, UK) to detect rCBF and con rm successful occlusion and reperfusion.

Pretreatment with WGex and RGex
Experimental mice were randomly divided into eight groups as follows: 100, 300, and 1000 mg/kg WGextreated; 100, 300, and 1000 RGex-treated; PBS-treated control; sham-operated sham group. Each group consisted of 15 mice ( Supplementary Fig. 1, Fig. S1). In Korean medicine, 30 mg/kg has been recommended as the daily dose of WG and RG extract in humans; in mice, the equivalent dose was determined as 400 mg/kg. Thus, the dose range varied between 100 and 1000 mg/kg. For each WGex and RGex concentration prepared in PBS, the solution was orally administered using an oral zonde 1 h before MCAO induction, with sham and control mice orally administered the same volume of PBS. After 24 h of MCAO induction, the mice were sacri ced to perform several assays. A schematic diagram of the experiment is shown in Fig. 1.

Infarct and edema area measurements
TTC staining can be assumed as a function of mitochondria and is reportedly an important marker of ischemic volume following ischemic stroke. The mice were anesthetized by CO 2 respiration, and blood was collected to obtain serum (25℃ room temperature, 10 min, 13,000×g). Then, the mouse brain was harvested to measure the total infarct volume, and the olfactory bulb and cerebellum were isolated. Serial coronal sections (1 mm thickness) were stained with 2% TTC solution to assess infarct volume. Next, the sections were digitized using a digital camera and a tablet, and the ischemic area and non-infarct tissue were separated using an image analysis system (Digimizer, Ostend, Belgium).
Areas of brain edema were calculated by dividing the total infarction volumes in ipsilateral hemispheres by brain edema indices using TTC-stained brain sections. The formula used to calculate brain edema indices is as follows: Brain edema index = total volume of ipsilateral hemisphere / total volume of the contralateral hemisphere Neurological de cit score (NDS) measurement In mice, the behavioral changes were measured using a 5-scale assessment after surgery, and the characterization of each score was described as follows: Score 0, mouse presenting behavior similar to a normal mouse; Score 1, spontaneous exercise could be performed, but the behavior was slower; Score 2, slightly reduced temperature and reduced (2-3 g) body weight, and mice moved toward the contralateral direction; Score 3: reduced body temperature and weight (3-4 g), and mice demonstrated sensitivity to mechanical stimuli, including pain, walking, or circling to the right; Score 4, reduced body temperature and weight (4-5 g), with mice failing to exhibit spontaneous behavior or demonstrate extreme sensitivity to mechanical stimuli.
Sacri ce and cardiac perfusion for harvesting the brain The mouse abdomen was incised, and cardiac perfusion with PBS was performed. In brief, the pulmonary artery was blocked, the left ventricle was pierced with a 21 gauge needle; then, the needle was xed in the ascending aorta. Immediately after perfusion, the right atrium was incised using scissors. PBS and 4% paraformaldehyde (PFA) solution were used for perfusion and xation, respectively. Post-xation, the brain was soaked in 10% PFA with 10-30% sucrose at 4℃ for 3 days, followed by cryosectioning for appropriate staining.
Frozen sections of mice brain Mouse brains were sequentially placed in 10%, 20%, and 30% sucrose solutions before freezing (in OCT compound), and then stored at -80°C. For each mouse group, brain sections were obtained using a cryostat (Leica, Wetzlar, Germany) at a thickness of 30 μm, and the sections were then placed on glass slides for 12 h and stored at -80℃ until use.

Hematoxylin and eosin (H&E) staining
For each mouse group, the brain sections on slides were allowed to dry on a slide warmer; then, the slides were dipped into 80% ethyl alcohol. The slides were immersed in a hematoxylin solution for 5 min, followed by gently washing with distilled water. Next, the slides were soaked in acid alcohol (1%) two times, and immersed in a lithium carbonate solution. Then, the sections were washed and slides were dipped in eosin solution for 30 sec. The slides were immersed in graded ethyl alcohol concentrations (95% and 100% sequentially) for 1 min, placed in xylene, mounted, and then observed under a light microscope. The density of H&E-positive cells was measured in the cortical region of mouse brains using the ImageJ program (NIH, MD, USA).

Nissl staining
For each mouse group, the sections on glass slides were allowed to dry on a slide warmer. Then, the slides were immersed in an alcohol:chloroform solution overnight. The slides were incubated in 0.1% CV solution for 10 min at 40℃ in an incubator. Then, the slides were quickly washed with distilled water, placed in graded ethyl alcohol (95% and 100% sequentially) for 5 min, sealed with a coverslip and mounting solution, and then observed under a light microscope. The density of neuronal cells was measured in the cortical region of the mouse brains using the ImageJ program (NIH, MD, USA).

Western blot analysis
Proteins in the ipsilateral brain hemisphere were isolated using a protein extraction solution. Brain tissue lysates were obtained by centrifugation at 15,871×g for 10 min at 4℃. The total protein level in the supernatant was quantitated using the BSA method. In brief, 30 µm of the protein was developed and separated by SDS-PAGE, and then transferred onto PVDF membranes (Millipore, Darmstadt, Germany). The PVDF membranes were blocked with 5% skim milk in TBST buffer (Tris-buffered saline, 0.1% Tween 20) for 1 h at 25℃, and then incubated overnight at 4℃ with primary antibodies against inducible nitric oxide synthase (iNOS), p-IkBα, IkBα, p-JNK (1:1000), JNK (1:1000), Mn-SOD (1:1000), and β-actin (1:1000). After overnight incubation, secondary antibodies for HRP-conjugated goat anti-rabbit IgG, pAb (1:5000), and HRP-conjugated goat anti-mouse IgG pAb (1:3000) were added for 1 h at 25℃. After incubation, the membranes were treated with an ECL solution kit, and the expression levels of each protein were detected using a luminescent analyzer system (Amersham™ Imager 600, Buckinghamshire, UK). The densities of all detected bands were analyzed using the ImageJ program (NIH, MD, USA).

Statistical analysis
One-way analysis of variance (ANOVA) was performed, followed by the Holm-Sidak test to determine the signi cance of differences among groups using Sigmaplot 12.0 (Systat Software Inc., CA, USA). Data are expressed as the mean ± standard deviation (SD), and a p < 0.05 was deemed signi cant.

Analytical comparisons of WGex and RGex and their standard, ginsenoside Rg1
Ginsenoside Rg1 is the main derivative of radix ginseng, and its structure is shown in Figure 2a. The presence of ginsenoside Rg1 in WGex and RGex was con rmed by HPLC (Fig. 2b, 2c). Following quantitative measurement, a three-fold higher ginsenoside Rg1 concentration was detected in WGex than that in RGex ( Table 2).

Measurements of infarct volume, edema area, and NDS
To determine the effects of ginseng extracts on ischemic stroke in mice, pretreatment with WGex and RGex was performed 1 h before the induction of ischemic stroke, and the effective dose was selected. Single doses of 100, 300, and 1,000 mg/kg/day were administered (treatment on the day of MCAO).
Representative images of TTC-stained brain slices are shown in Fig. 3a. Among them, the most effective dosage selected for RGex was 1,000 mg/kg body weight (bw) (Fig. 3a). WGex had no effect on total infarct volume within the investigated dose range; however, the infarct volume tended to decrease in a dose-dependent manner (Fig. 3a). Following pretreatment with 1,000 mg/kg bw of RGex, the percentage of the total infarct volume in the ipsilateral hemisphere of the brain was 41.40 ± 5.28%, and that in the MCAO control group was 56.67 ± 10.05%.
For NDS, no signi cance was observed among the MCAO-induced groups (Fig. 3c); however, pretreatment with WGex and RGex (1,000 mg/kg bw) revealed lower scores than the low-dose treatment groups.
In anesthetized mice, we assessed rCBF during MCAO induction and reperfusion processes using a laser Doppler probe connected to an optic ber, attached to the temporal surface of the skull. Assuming that the entire blood ow before CCA ligation was 100%, CCA occlusion reduced CBF to approximately 59 ± 6% of baseline levels. On occluding the MCA, rCBF values decreased to approximately 10 ± 8%, and no signi cant difference was observed among groups ( Supplementary Fig. 2, Fig. S2).

Morphological changes in neuronal cells
To evaluate the neuroprotective effects of ginseng extracts, WGex and RGex, on ischemic stroke-induced neuronal damage, we investigated morphological changes in neuronal cells in MCAO-induced in the ipsilateral hemisphere of mouse brains. In the sham-operated group, H&E-stained neuronal cells were observed as normally intact and showed morphologically well-arranged cytoplasm and nucleus. The MCAO group showed aberrant morphology, displaying an H&E-negative area. Animal groups pretreated with ginseng extracts presented marginal recovery of neuronal damage (WGex, 74.33 ± 5.86%; RGex, 79.67% ± 5.03%), demonstrating similar cellular structure to that observed in the sham-operated group; however, on comparing the H&E color intensity between groups, ginseng extracts showed no signi cant change when compared with the MCAO control group (Fig. 4a).
For all groups, the cell density on CV staining in the ipsilateral hemisphere of the brain is shown in Figure   4b. The cell density of the sham-operated group was 100.00 ± 5.00% and that in the MCAO control group was 65.33 ± 7.02%. Following pretreatment with 1,000 mg/kg bw of WGex or RGex in MCAO-induced mice, the cell densities were 81.67 ± 3.06% and 85.33 ± 5.51%, respectively (Fig. 4b). The cell densities in mouse brains pretreated with ginseng extracts were signi cantly higher than those of the control group (Fig. 4b). The MCAO control group presented cellular damage, with shrinkage and pyknotic nuclei (red arrowheads, Fig. 5b), whereas the ginseng extract-treated groups revealed neuronal cell recovery (arrows, normal neurons; black arrowheads, glial cells). Overall, when stained with CV, the ipsilateral cell density of all ginseng treated groups was signi cantly recovered (Fig. 5c, 5d).

Expression of cell death related proteins in the ipsilateral cerebral cortex
Reportedly, iNOS promotes in ammation and exerts a synergistic effect with other in ammatory mediators; in the brain, increased iNOS protein levels after ischemia cause NO production and DNA damage. Thus, inhibiting the activity of iNOS could be a key target for regulating in ammatory responses in the ischemic brain [28][29][30]. In the current study, pretreatment with 1,000 mg/kg bw RGex decreased (1.44 ± 0.18-fold) the increased iNOS expression in brain damage induced by MCAO (2.16 ± 0.44-fold) (Fig. 6a).
NF-κB participates in in ammasome regulation, and NF-κB proteins are present in the cytoplasm in association with regulatory proteins called inhibitors of κB (IkBs). The IkB proteins are phosphorylated and subsequently degraded in the proteasome, and degradation of IkBs allows translocation of NF-κB into the nucleus, thus promoting in ammatory responses [36,37]. The increased expression of p-IkBα in the MCAO control group was inhibited by RGex administration. The ratio of p-IkBα/ IkBα expression was 1.00 ± 0.26 in the sham group, 2.06 ± 0.36 in the MCAO control group, and 1.48 ± 0.23 in the RGex-treated group (Fig. 6c).
Mn-SOD, also known as superoxide dismutase 2, is speci cally localized to the inner membrane of mitochondria and is reportedly related to antioxidant pathways [38][39][40], demonstrating neuroprotective effects against oxidative stress [41][42][43], thus presenting a potential target for determining herbal agents with activity against ischemic brain injury. In this study, MCAO signi cantly increased the protein expression of Mn-SOD, and pretreatment with 1,000 mg/kg bw RGex signi cantly inhibited this increased expression. The expression values were 1.00 ± 0.13 in the sham group, 1.54 ± 0.11 in the MCAO control group, and 1.08 ± 0.31 in the RGex pretreated group (Fig. 6d).

Discussion
P. ginseng, also known as Korean ginseng, is well-known as a functional health food for vitalizing energy and eliminating chronic fatigue while improving health and has been used as a dietary supplement in Asian countries for over 2000 years [44]. P. ginseng, both red and white preparations (RG and WG, respectively), are most commonly used in traditional Korean herbal medicine, with some differences noted among them, including the ginsenoside content and pharmacological effects. RG is manufactured when WG undergoes steaming and drying [4]. During this manufacturing process, the chemical composition of ginsenosides is altered, producing bioactive compounds that have the potential to induce special physiological activities in vivo [44]. Currently, more than 30 different ginsenosides from P. ginseng have been isolated and characterized, and these ginsenosides are known to have different pharmacological effects [44]. However, in terms of pharmaceutical differences, comparative studies on WG and RG using animal models have not been extensively investigated. By employing network pharmacology research tools, we have recently reported that WG and RG could be e cacious in ischemic brain diseases [26], but comparative studies on the e cacy and underlying mechanisms remain scarce in ischemic stroke animal models.
Stroke is the second most common cause of death worldwide, and the most common type is ischemic stroke. Ischemic stroke is a major neurological disorder that causes physical and psychological disabilities [7][8][9][10][11]. As the regeneration of cerebral tissue at the time of injury is challenging, cerebrovascular damage remains intact and there is a high possibility of recurrence. Therefore, patients suffering from post-ischemic cerebral dysfunction should be prescribed suitable therapeutic agents to adjust for ongoing risk factors. Ultimately, from a pharmacological perspective, it is crucial to develop a therapeutic agent to prevent neuronal cell death, excitotoxicity, and oxidative stress, all pathways of brain cell damage [45][46][47]. Ischemic stroke is primarily due to metabolic diseases such as atherosclerosis, and the pathogenesis of ischemic stroke includes oxidative stress, in ammatory response, apoptosis, and autophagy [7][8][9][10][11]. Owing to the various pathogenic processes of ischemic stroke, multicomponent and multitarget agents, including traditional herbal medicine, can be effectively used to treat ischemic stroke at various stages.
Several studies have reported that treatment with ginseng or ginsenosides improves general symptoms, physical exercise capacity, and uid metabolism in patients with coronary and myocardial ischemia and reperfusion diseases [48,49]. Furthermore, ginseng is reportedly used for various neuronal diseases such as neurodevelopmental, neurodegenerative, and neuropsychiatric disorders [50]. Zheng et al. have reported that ginseng total saponins improve neurological function de cits after focal cerebral ischemia by inducing endogenous neural stem cell activation, and enhance central nervous system regeneration in an MCAO rat model [51] In rats with ischemia/reperfusion-induced brain injury, Ban et al. have demonstrated that Korean RG extract showed a neuroprotective effect by reducing the level of lipid peroxidation and increasing the endogenous antioxidant enzymatic activity [52]. Recently, Liu et al. have reported that ginseng pretreatment affords protection against acute sensorimotor de cits, promoting its long-term recovery after proximal-distal MCAO in mice through nuclear factor erythroid 2-related factor 2 (Nrf2) activation [53]. Therefore, it can be assumed that WG, RG, and ginsenosides protect the brain against ischemic brain injury; however, no previous study has compared the effects of WG and RG using an animal model of ischemic stroke. Thus, in this study, the MCAO mouse model was used to compare the different effects and measure the content of ginsenoside Rg1, a representative component of ginseng, to determine whether the difference in e cacy of WGex and RGex was ginsenoside Rg1 dependent.
To observe the effects of WGex and RGex following 2 h of MCAO in mice, pretreatment with 1,000 mg/kg bw of RGex showed a signi cant reduction in % infarct volume, but WGex showed no signi cant change in the dose range investigated (100 to 1,000 mg/kg bw) (Fig. 3a). Following pretreatment with WGex or RGex, the infarction volumes were reduced depending on the concentrations; although a high concentration of WGex demonstrated an effect, the effect of RGex at the same concentration (ranging from 100 to 1,000 mg/kg bw) was superior to that of WGex. On measuring the ginsenoside Rg1 content (Table 2), it was observed that WGex contained 3 times more ginsenoside Rg1 than RGex, indicating that ginsenoside Rg1 is not involved in the neuroprotective effect of RGex in the MCAO mouse model. A previous study has shown that total ginsenosides protect the brain from ischemic injury [51]; however, in this study, speci c ginsenosides were considered to play a key role in speci c pharmacological effects. Thus, there is a need to further investigate pharmacological activities using individual constituents. RGex pretreatment was shown to suppress brain edema but did not affect NDS changes; conversely, WGex did not affect brain edema, NDS, and rCBF changes (Fig. 3b, 3c, and Fig. S1).
To evaluate the neuroprotective effects of WGex and RGex in the cortical region of brain tissues, at a microscopic level, we used H&E-or CV-stained frozen sections. In tissues stained with H&E, the effects of WGex and RGex could not be con rmed (Fig. 4a); however, when stained with CV, both WGex and RGex showed signi cant neuroprotective effects in ischemic ipsilateral cortical neuronal cells (Fig. 4b). These results can be attributed to the characteristic of CV capable of speci cally staining neuronal cells; on magni cation, it was observed that RGex signi cantly inhibited cellular changes related to the in ammatory response (Fig. 5).
In this study, western blotting experiments were conducted to determine the molecular mechanism underlying the neuroprotective effects of WGex and RGex in the MCAO mouse model. iNOS plays a critical role in in ammation and apoptosis, and MAPK is an intracellular signaling molecule involved in cell migration, proliferation, and differentiation. Moreover, the MAPK pathway plays an essential role in the induction of iNOS expression [54][55][56]. JNK activation is associated with growth factor-mediated pathways [33][34][35]. In this study, the expression level of JNK was increased by MCAO induction, and pretreatment with both WGex and RGex reduced the expression level of JNK (Fig. 6b); the expression level of iNOS was regulated by RGex pretreatment (Fig. 6a). Overall, MCAO promoted the iNOS and MAPK pathways, and pretreatment with 1,000 mg/kg bw RGex altered these expressions signi cantly.
Phosphorylated IκB proteins allow translocation of NF-κB into the nucleus, promoting in ammatory responses [36,37]. The increased expression of p-IkBα in the MCAO control group was inhibited by RGex administration (Fig. 6c). Mn-SOD is an important potential target for identifying neuroprotective agents with e cacy against ischemic brain injury [41]. In this study, MCAO signi cantly increased the protein expression of Mn-SOD, and pretreatment with 1,000 mg/kg bw/day RGex signi cantly inhibited this increase (Fig. 6d); this indicated that antioxidative signals were activated by MCAO induction, and RGex reduced cerebral damage, thus inhibiting Mn-SOD protein expression.
In summary, both WGex and RGex protected neuronal cells in the MCAO mouse model, but the e cacy of RGex was stronger than that of WGex, and ginsenoside Rg1 was not involved in the e cacy of RGex. Mice treated with 1,000 mg/kg bw/day of RGex 1 h before MCAO induction revealed reduced neuronal cell death in the MCAO-induced mouse brain, and this effect was identi ed to be associated with antiin ammatory and anti-apoptotic pathways.

Conclusions
WGex and RGex showed neuroprotective effects in the cortical region of the MCAO-induced ischemic stroke mouse model, but the overall e cacy of RGex was stronger than that of WGex. Pretreatment with 1,000 mg/kg bw RGex signi cantly reduced the total infarction volume and edema area; these effects were not observed with WGex. Ginsenoside Rg1 was not considered to be involved in the effects of WGex and RGex. The protective effect of RGex in the MCAO mouse model was found to be related to antiin ammatory and anti-apoptotic pathways.

Figure 3
Measurements of infarct volume, edema area, and neuronal de cit score (NDS). a, representative photographs of TTC-stained brain slices (1 mm) showing the infarct area 24 h after MCAO (upper column) and quantitative analysis of the total infarct volume (lower column). b, quantitative analysis of the edema area, calculated by dividing total infarction volumes in ipsilateral hemispheres by brain edema indices using TTC-stained brain slice. c, quantitative analysis of NDS. All data expressed as means ± standard deviation (SD) (n=5). ### p < 0.001 vs. sham group and * p < 0.05 vs. MCAO control group. TTC, 2,3,5-triphenyl-tetrazolium chloride; MCAO, middle cerebral artery occlusion; NDS, neurological de cit scores.