Pharmacodynamic Effect of Ellagic Acid on Ameliorating Cerebral Ischemia/Reperfusion Injury

As a serious morbidity worldwide, ischemic stroke causes at least 6,671,000 deaths [1‒3] and more disabled patients every single year [4, 5]. The immediate ischemic stroke injury results from the vessel occlusion that reduces temporarily or permanently the cerebral blood flow [6, 7]. Such an injury is intensified at the cellular and structural levels by the subsequent ischemia/reperfusion (I/R), leading to the irreversible damage to the brain and other organs [8‒10]. However, fewer effective protocols are available to manage individually the I/R injury [11], since most if not all current therapeutics are largely based on a single factor and show limited efficacy in alleviating a diversity of I/R pathological outcomes such as inflammation and neuronal cell apoptosis [11‒14]. Therefore, there is an urgent need to develop alternative treatments for the I/R amelioration.

In the Tibetan area, the cerebrovascular diseases are highly incident owing to the high-altitude anoxic environment and the diet habit of consuming more wine and meat [15, 16]. As a vivid showcase of multiple component-based therapies, the traditional medicine practitioners in that region were smart enough to manage such diseases by developing decoctions composed of diverse natural medicines. According to the canonical collection of Tibetan medicines – called in Chinese “Sibuyidian,” the “Shanhu,” “Ruyi Zhenbao,” and “Chenxiang” – pills have been, since 700 BC, prescribed for the therapy of cerebrovascular diseases including cerebral ischemia. After ascertaining their protective effect on the I/R injured rats [17, 18], the 3 prescriptions were comparatively analyzed for the presence of “common active components” in plasma after the oral administration of each of the extracts derived from the 3 pills. Co-present therein were the 6 phytochemicals in the following order: ellagic acid (EA) ∼luteolin > crocin ∼ glycyrrhizin ∼ mangiferin ∼ glycyrrhetinic acid. Since luteolin has been intensively investigated [19], our attention was focused on the presumable alleviating effect of EA on the I/R injury in the brain as EA distributes widely in plants [16]. Gratifyingly, EA reduces substantially the percentage of cerebral infarction volume in permanent middle cerebral artery occlusion ­(pMCAO) rats primarily through a combination of its anti-inflammatory and anti-oxidative actions. The work showcases the effectiveness in identifying efficacious compounds by the prioritization of the constituents absorbed in plasma.

In this study, we first analyzed the compounds in these 3 prescriptions in vivo and in vitro. Then, the therapeutic effects of the major active component EA for I/R were evaluated by measuring the percentage of cerebral infarction volume in pMCAO models. Histopathological examination, western blot (WB) analysis, and biochemical analysis were used to detect the preventive effects of the inflammatory response and oxidative stress on neurons after the administration of different doses of EA. Finally, since EA has been identified to activate coagulation factor XII to trigger the intrinsic coagulation system, we further detected the activity of coagulation factor XII. All the results demonstrated that EA was a promising candidate for I/R.

“Shanhu” pill, “Ruyi Zhenbao” pill and “Chenxiang” pill, the nimodipine as the positive control and EA were supplied from Balaqushenshui Tibetan Pharmaceutical Company (Tibet, China), Tibetan Pharmaceutical Company (Tibet, China), Bayer medical care Co., Ltd. and Pusi Biotechnology Co., Ltd. (Chengdu, China) respectively. Pheochromocytoma cell (PC12 cell) line was obtained from the American Type Culture Collection (Rockville, MD, USA), and grown in DMEM. The cell culture was maintained in a 37°C incubator with a humidified 5% CO₂ atmosphere.

2,3,5-triphenyltetrazolium chloride was obtained from Sigma-Aldrich (St. Louis, MO, USA). Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay kit, was bought from Boster Biological Technology Co., Ltd. (Wuhan, ­China), and the Rat interleukin 6 (IL-6) enzyme-linked immunosorbent assay (ELISA) Kit (Cusabio Co., Ltd.), Rat IL-6 (IL-1β) ELISA Kit (DuoSet®) and Rat IL-6 tumor necrosis factor-alpha (TNF-α) ELISA Kit (Affymetrix eBioscience) were used for detecting the inflammatory factors. For WB assay, the Mouse Anti-aquaporin 4 (AQP-4) Polyclonal Antibody, the Mouse Polyclonal matrix metalloprotein 9 (MMP-9) Antibody and the Mouse Polyclonal zonula occludens-1 (ZO-1) Antibody obtained from Abgent App Tec Co., Ltd. (Wuxi, China) were used as the Primary antibodies. All the materials used in this article were of the analytic grade.

Male Sprague-Dawley rats were purchased from Dashuo Bio-Technology. Co., Ltd. (Chengdu, China) throughout the experiment and the rats were housed at a temperature of 20 ± 2°C, the relative humidity of 50–60%, and with 12 h light-dark cycles. All rats were treated humanely throughout the experimental period, and all methods were carried out in accordance with the approved guideline (IACUC-S200904-P001).

All the rats were weighing around 280–300 g, and divided into 9 groups: sham group, I/R group, nimodipine group, “Shanhu” pill groups, “Ruyi Zhenbao” pill group, “Chenxiang” pill group, EA group (low dose, 10 mg/kg), EA group (30 mg/kg) and EA group (high dose, 50 mg/kg), randomly. All the drugs were prepared with 0.9% NS and orally administrated.

The active compounds in these 3 classical Tibetan prescriptions in vivo and in vitro were determined by a liquid chromatograph (LC)-mass spectrometer and high-performance liquid chromatography with C18 column (4.6 mm × 250 mm × 5 μm, Ultimate® XB-C₁₈). The methanol solution of 3 prescriptions was used for the in vitro analysis of compounds. The in vivo analysis of compounds was performed as follows: the rats were left to starve overnight prior to drug administration and randomly divided into 3 groups and the rats were respectively treated with “Shanhu” pill, “Ruyi Zhenbao” pill, “Chenxiang” pill in normal saline orally by gastric intubation. Two hours later, blood samples were collected and immediately centrifuged to obtain the plasma. The active compounds were extracted from plasma by methanol. The organic layer was evaporated and reconstituted in the mobile phase analysis.

Pre-treated serum samples were performed at 250 and 400 nm wavelength and each UV spectrum peak was compared to the standard compounds by an Agilent 1260 high-performance liquid chromatography system (Agilent Technologies, Santa Clara, CA, USA). Chromatographic analysis was performed at 35°C with Ultimate XB-C18 column and acetonitrile and water-formic acid (100:0.2, v/v) were used as the mobile phases A and B (the gradient elution ratio table of phase A and phase B is shown in Table 1) respectively. The mobile phase was delivered at a rate of 1.0 mL/min with a 10 µL injection volume. The data were analyzed by Lab Solutions software version C01.07 (Agilent Technologies, Santa Clara, CA, USA).

The cell viability of EA (dissolved by N-Methyl pyrrolidone) was evaluated by MTT assay on the PC12 cell line. In these tests, the cells were seeded in 96-well plates at a density was 2 × 10⁴ cells per well for 24 h incubation. The EA at different concentrations was added into the plates for 24 h co-incubation. Then, H₂O₂ was added into the plates for another 4 h of incubation. Next, 20 μL MTT solution (5 mg/mL) was added to each well, and the solution was replaced with DMSO (100 μL per well) after 3 h. The absorbance was measured by a 680-model microplate reader from an Infinite M200 microplate reader (Tecan, Durham, NC, USA).

Rats were fixed in a supine position after being anesthetized by choral hydrate (10%, 100 g/mL, ip). First, as the much previously reports, the one side of vena jugularis and bilateral carotid arteries was exposed. Middle cerebral artery was occluded for 2 h while the blood was taken from vena jugularis. And then, the reperfusion of the blood was followed by the removing of thread. The damages of neurocyte were evaluated according to Zea Longa. The same operation as above, except for the MCAO and hemospasia from the common vena jugularis, was performed on the rats in the sham group.

Rats were sacrificed at 24 h after the last administration, and their brains isolated for the estimation of the infarct area (n = 5). The rats were sacrificed under deep anesthesia and their brains were removed immediately and rinsed by NS. Then, the 2 mm silence of the samples were placed in 2% 2,3,5-triphenyltetrazolium chloride stain for 15 min. The brain tissue was differentiated according to the white-colored infarct area and red-purple non-infarct area. The infarct volume was calculated using the following formula:

The infarct volume (%) = Weight of white area/Weight of whole brain × 100%

Evans blue (EB) leakage was used to investigate the brain-blood-barrier (BBB) disruption after I/R injury. 2% EB (v/v) ­solution in PBS was injected via tail vein at 4 h after the reperfusion (4 mL/kg). Five days after administration, mice were sacrificed, and the images of the EB leakage into the ischemic brain were obtained.

The brains of the rats were fixed with and post-fixation in 4% paraformaldehyde for at least 24 h. Then, the samples were embedded in paraffin and coronally sectioned around 7 μm for a series of detection. The thicknesses were stained with hematoxylin and eosin (H&E staining). The histomorphology of neurons was observed under the microscope. The injured neurons cells were dark staining, shrinkage or dysmorphic, and intact were distinct nucleus and nucleolus. For Nissl staining, sections were dehydrated in ethanol and chloroform, stained with 1% toluidine blue at 50°C for 10 min, and then which was rinsed, cleared in graded ethanol and xylene, and coverslipped. Furthermore, the TUNEL was used to evaluate the DNA fragmentation associated with apoptosis by a situ cell-death detection kit. The counts of cells were performed at a magnification microscope.

The activity of inflammation factor was investigated by evaluating the level of TNF-α, IL-6 and IL-1β respectively. Since the coagulation factor XII could be activated by EA, the content of coagulation factor XII in serum should be detected in this study. The samples were all measured by ELISA and all the methods were according to the manufacturer’s manual of the biochemistry assay kits. The content of malondialdehyde and the activities of superoxide dismutase (SOD) as the representation to be investigated for evaluating the anti-oxidative stress effect of EA. The samples were prepared to follow as described before, and then the contents were measured by the automatic biochemical analyzer (Rayto life and analytical sciences Co., Ltd. Chemray 24).

To investigate the promoting effect of EA in BBB repairing after I/R, MMP-9, AQP-4 and ZO-1 in the brain were used as the biomarkers to be detected by WB. First, the isolated brains in different groups were homogenized at –80°C. Next, the samples were lysed in RIPA, the supernate and equal amounts of protein were loaded on an SDS-PAGE gel. The gel was transferred to PVDF membranes during the night after SDS-PAGE and then blocked with 5% milk in TBS. The rat monoclonal antibody to MMP-9, AQP-4, and ZO-1 was used as the primary antibody and the goat anti-rat HRP labeled antibody was used as the secondary antibody. The protein expression on membranes was analyzed by an exposure meter.

The comparison of each group was evaluated by the statistical analysis using SPSS software with one-way analysis of variance. All data were expressed as the average value ± SD, and a pvalue <0.05 was considered statistically significant.

The work was initialized by developing an LC-MS method to recognize the main constituents that may contribute to the efficacy of the locally called “Shanhu” pill (25-herbs), “Ruyi Zhenbao” (72-herbs), and “Chenxiang” pills (20-herbs). Fourteen dominant compounds were detected in their methanol extracts including gallic acid, hydroxysafflor yellow A, corilagin, mangiferin, EA, anthocyanin, crocin I, crocin II, glycyrrhizin, liquiritin, Isoglucoside, luteolin, and glycyrrhiza chalcones A and B. As shown in Figure 1, the 6 compounds identifiable in vivo were EA, luteolin, mangiferin, crocin, glycyrrhizin, and glycyrrhetinic acid. Moreover, the abundance of crocin II, glycyrrhizin, and mangiferin were too low to be quantified in vivo. To our surprise, EA and luteolin were detected as common dominant phytochemicals in the 3 Tibetan formulations (Table 2). Luteolin is anti-inflammatory and anti-oxidative, and it has been used to ameliorate I/R injury in vivo [19]. But little information is available concerning the I/R injury alleviating effect of EA – another major compound common in the 3 prescriptions. To address this lack of information, follow-up work was carried out.

To explore the protective effect of EA in vitro, we first constructed an oxidative stress damage model in vitro through cultured PC12 cells with H₂O₂. Then, EA (dissolved by N-Methyl pyrrolidone) was added into the cell supernatants and incubated for another 4 h. Finally, the viable ratio of PC12 cells reflected the protective effect of EA. From Figure 2 it is clear that with increasing EA concentration to 20 μmol/L, the cell viability reached around 83%, and the EA group exhibited superior activity compared to the model group, suggesting that EA plays a protective role in preventing nerve damage.

In a comparative study including those of the I/R group, a 5-day oral administration of EA to I/R injured rats led to the infarct volume reduction in a dose-dependent manner (Fig. 3). With EA dose increasing from 10 to 50 mg/kg, the ratio of the infarct volume to the whole brain was decreased from 25.11 ± 0.85 to 19.99 ± 1.12%. These results corroborated that EA improves the I/R damage dose-dependently.

To reinforce the effect of EA on the treatment of I/R, the histopathology was studied with H&E staining, Nissl staining, and TUNEL staining. The brain tissue of the sham group was visualized by H&E staining (Fig. 4) as clear and integrated tissue structures with normal neuron cells that were clearly membranes, ordinarily shaped and with centered nuclei. In the experimental groups, the different degrees of interstitial edema could be found, and some neurons were swollen with disruption and disintegration of nuclei. Compared to the model group, the EA (50 mg/kg) and nimodipine groups had a relatively normal tissue structure. Next, the Nissl staining was applied to examine the magnitude of neuronal injury. As shown in Figure 5, the number of Nissl corpuscle of neuronal cells in experimental groups was much less than that of the sham group, again in a dose-dependent way (see above). Finally, the TUNEL staining was used to assess neuronal apoptosis. Compared to that of the sham group, an increased number of TUNEL-positive neurons was found in the EA-treated groups (Fig. 6). Collectively, this set of histopathological studies indicated that EA inhibits neuron apoptosis and promotes the tissue repair of the I/R injured brain.

It is reported that EA downregulates the expression of IL-6 and TNF-α on colon carcinogenesis in rats by suppressing nuclear transcription factor-kappa B [20]. We were therefore curious about whether EA exerts its neuroprotective effect via its reported anti-inflammatory action. Thus, TNF-α, IL-6, and IL-1β, the 3 important mediators of neuroinflammatory response, were used as the indicators to evaluate the anti-inflammatory effect of EA in pMCAO model. As illustrated in Figure 7, both in serum and brain, the inflammation response was apparent in the I/R injured group relative to the sham group. After the 5-day treatment with EA at doses of 10, 30, and 50 mg/kg day, the expression of the inflammation mediators in the brain was substantially reduced by approximately 50% in comparison to the untreated model group, with the magnitude close to that of nimodipine co-evaluated as a positive control. Thus, EA mitigates the inflammatory response in the brain as it does in serum [21, 22].

In the I/R injury, the increasingly released reactive oxygen radicals could lead to lipid peroxidation and cell membrane damage [23, 24]. As one of the normal lipid peroxidative products, the MDA could be used to reflect the extent of the tissue damage [25]. Moreover, as one of the important endogenous antioxidant enzymes, SOD could inhibit the lipid peroxidation damage by free radicals [26]. Therefore, to investigate the effect of EA on anti-oxidative stress, the content of MDA and the SOD activity in serum were measured. Compared to the sham group, the MDA content of the model group increased, whereas the SOD activity decreased (Fig. 8), indicating that the I/R damage is associated with the oxidative stress. As anticipated, the treatment of the I/R injured group with EA reduced the level of the oxidative stress as reflected by the dose-dependent reduction of the MDA content and increment in the SOD activity.

BBB is a specialized structure in the central nervous system, which keeps the neurotoxic substances away from the brain [27]. The I/R derived brain damage changes the BBB permeability, in which the inflammative and oxidative events play significant roles [28]. Therefore, EA was evaluated in the work for its protective effect on the integrity of BBB structure and function in the I/R injured animals. As a matter of fact, the major injurious proinflammatory factor called MMP-9 expresses immediately after the I/R injury to disrupt BBB via degrading the tight junction proteins (TJs) [29]. TJs preserve the integrity of microvascular endothelium in brain and are found to be composed of some TJ-related proteins, such as ZO-1 and Occludin [30]. As the main water protein in the central nervous system, AQP-4 regulates water homeostasis in BBB to help form the vasogenic edema associated with the I/R injury [31]. The observation motivated us to take the expressions of MMP-9, ZO-1, and AQP-4 as a “cross index” to signify the integrity of BBB structure and function. As illustrated in Figure 9a and b, the 5-day EA treatment of I/R injured animals led to the decreased expression of MMP-9 and AQP-4 accompanied by increased ZO-1 expression in comparison with counterparts in the untreated I/R group. It is noteworthy that these actions are similar in magnitude to those of nimodipine co-assayed as a positive control. It could be concluded that in the BBB repair, the EA could regulate the permeability of BBB by regulating the expression of relative proteins. Moreover, from Figure 8c, it could be easily observed that the Evens blue were accumulated in experimental groups, and the concentration of Evens blue in the nimodipine group and EA group was lower than that in the I/R group. These results suggested that the EA could promote the broken BBB recovery.

Previously, focal cerebral ischemia can be generated by the EA accumulation in the carotid artery, since EA triggers the intrinsic coagulation through the activation of coagulation factor XII [29]. Therefore, we were prompted to address the influence of EA (orally administered) on the expression of coagulation factor XII. To our surprise, the expression of coagulation factor XII was found to be independent of EA at the doses of 10, 30, and 50 mg/kg (Fig. 10).

I/R injury is the most common cause of serious morbidity, and 3 classical Tibetan prescriptions – including the “Shanhu” pill, “Ruyi Zhenbao” pill and “Chenxiang” pill – have been used for I/R injury treatment since 700 BC. The I/R injury alleviation mediated by EA, one of the common dominated compounds of 3 prescriptions, was investigated in this work. The vitro experiments showed that EA has protective effects on oxidative stress injury. The results of measured infarct area indicated that EA favors promoting the recovery of the brain. Furthermore, by histopathological study of the brain tissues in pMCAO, the results showed that the EA can inhibit the damage and apoptosis of neuron cells, which prevent the brain from injury. By measuring the expression of the inflammatory and oxidative factors, it demonstrates that the EA exhibits anti-inflammatory and anti-oxidative stress performance. The down-expression of MMP-9, AQP-4, and up-expression of ZO-1 in EA-treated groups further suggest the promoting effect of EA on BBB recovery. The alleviating effects of EA on I/R injury was dose-dependent. By detecting the activity of coagulation factor XII, it indicated that the clotting would not be activated at the therapeutic range of EA. The results demonstrate that the EA is a promising candidate for I/R injury.

This work was financially supported by the National Natural Science Foundation of China (81274184, 81473337, 81573563), Special funds for the national development and Reform Commission (ZYBZH-Y-SC-41-1) and China Postdoctoral Science Foundation (2018M631098).

The authors declare that they have no conflicts of interest to disclose.

L.T., Y. Wu, and R.T. designed the research. Y. Wang, Y. Wu, and C.L. conducted the experiments. L.T., Y. Wu, and R.T. participated in discussions. L.T., Y. Wu, Y. Wang, and R.T. wrote and revised the paper.

Y. Wang and Y. Wu co-first authors for this work.