In this study, we examined the metabolites from different parts of Acanthopanax senticosus and their role in alleviating damage caused by oxidative stress. We used UHPLC-QTOF-MS to analyze the chemical components in the root, seed, and leaf extracts of A. senticosus. Two multivariate statistical analysis methods—namely, principal component analysis and partial least square discriminant analysis—were used to distinguish the samples obtained from different parts of the plant. Using univariate statistics, 130 different metabolites were screened out. Among these, the relative content of flavonoids and terpenoids was found to be highest in the leaves, the lignin and phenolic acid content was highest in the roots, and the amino acid and phenolic acid levels were highest in seeds. An MTT assay was used to test the anti-H2O2 oxidative damage to PC12 cells in different parts of the sample. Lastly, using Pearson’s correlation analysis, various metabolites from different parts of A. senticosus were correlated with their antioxidant effects from the corresponding parts. Fifty-two related different metabolites were found, of which 20 metabolites that were positively correlated to oxidative stress were present at a relatively higher level in the roots, whereas 32 metabolites that were negatively correlated were present at relatively higher levels in the seeds and leaves. The results of this study reveal the distribution characteristics and the antioxidant activity of different metabolites of A. senticosus and provide a reference for the rational development of its medicinal parts.

1. Introduction

With an increase in the aging phenomenon, millions of individuals worldwide are affected by neurodegenerative diseases, which is one of the most significant challenges faced by modern society [1]. In particular, Alzheimer’s disease (AD), Parkinson’s disease (PD), and amyotrophic lateral sclerosis (ALS) are rampant. AD affects nearly 50 million people worldwide and is expected to grow steadily in the next decade. Although the annual medical expenses for neurodegenerative diseases are high, there is currently no definitive cure [2]. Therefore, research on the prevention and treatment of neurodegenerative diseases is warranted owing to the huge medical and public health burden that is associated with neurodegenerative diseases.

Oxidative stress injury as one of the breakthrough points of neurodegenerative diseases has received extensive attention in recent years. Oxidative stress is a phenomenon attributed to an imbalance in the production and removal of oxygen-reactive substances in cells [3]. Oxidative stress injury produces a large number of free radicals, which damages cells and tissues by destroying the integrity of the cell membrane by activating apoptosis-related signaling pathways and eventually inducing apoptosis [4]. Free radicals produced because of oxidative stress can damage neurons and result in neurodegenerative disorders. Neuronal cells have a stronger dependence and demand for energy resulting from oxidative phosphorylation, and they themselves are easier to be oxidized because they contain more polyunsaturated fatty acids than other tissues. Moreover, the concentration of antioxidant enzymes is relatively low and other factors render the neurons being highly sensitive to oxidative stress [5]. Several studies have shown that oxidative stress is closely related to the pathogenesis of neurodegenerative diseases. Activation of glial cells and an excess of reactive oxygen species can cause protein misfolding and mitochondrial dysfunction, in turn leading to apoptosis [6, 7]. PC12 cells are widely used in the establishment of neuronal damage-related models owing to their high structural and functional similarity with dopaminergic neurons. These cells are, therefore, used to study neurodegenerative diseases such as AD and PD [8]. The primary methods used to develop this model include H2O2 damage or glutamate damage [9]. In our study, a model of oxidative stress injury was induced by H2O2 in PC12 cells and used for subsequent experiments.

Acanthopanax senticosus is a traditional Chinese medicine rich in saponins, flavonoids, and phenolic acids [10]. It plays an important role in the treatment of diseases of the central nervous system [11]. For example, A. senticosus is indicated as auxiliary treatment for vascular dementia, cerebral ischemia-reperfusion injury, depression, and PD among other conditions [12, 13]. Studies report that the polysaccharide of A. senticosus has protective effects on H2O2-injured hippocampal nerve cells. However, the primary compound of A. senticosus that is responsible for the neuroprotective effect has not yet been elucidated. Using UHPLC-Q-TOF-MS plant metabolomics technology and modern activity evaluation methods, and through correlation analysis of various data, the effective ingredients of traditional Chinese medicine are clarified. Compared with traditional research methods, this research method has more comprehensive and more efficient analysis results and plays a very good role in promoting basic research on the quality of traditional Chinese medicine. It has become a new trend and hot spot in the research of active ingredients of traditional Chinese medicine in recent years [14, 15]. We used ultra-performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS), an effective technique for the systematic analysis of complex metabolomes, to analyze and identify the chemical components of A. senticosus from different parts of the plant [16, 17]. The MTT assay, based on colorimetry, was used to determine the H2O2-induced oxidative damage to PC12 cells. Lastly, using Pearson’s correlation analysis, the differential marker results of different parts of A. senticosus and the anti-H2O2 oxidative damage results of the corresponding parts were correlated to find the material basis of antioxidative stress damage in different parts of A. senticosus.

2. Experimental

2.1. Materials

The A. senticosus plant samples were collected from Yichun, Heilongjiang Province, China, and were identified by Professor Zhong Guoyue of the Jiangxi University of Traditional Chinese Medicine. The roots, seeds, and leaves of the samples were collected, dried in air, ground, dried to a constant weight, and stored in a desiccator until use. As a control, 2-chloro-L-phenylalanine was purchased from Shanghai Hengbo Biotechnology Co., Ltd. (99% purity; Shanghai, China). Methanol, acetonitrile, and formic acid were purchased from Shanghai Baiye Biotechnology Center (LC-MS grade; Shanghai, China). Highly differentiated rat pheochromocytoma cells (PC12) were purchased from the Institute of Basic Medicine, Chinese Academy of Medical Sciences (Beijing, China). Thiazolyl blue tetrazolium bromide (MTT) was purchased from Beijing Mengyimei Biotechnology Co., Ltd. (Beijing, China). The purifier was purchased from Merck Millipore (D24 UV; Massachusetts, USA). The ultrasonic instrument was purchased from Shenzhen Redbond Electronics Co., Ltd. (PS-60AL; Shenzhen, China). The centrifuge was purchased from Thermo Fisher Scientific (Heraeus Fresco 17; Massachusetts, USA). The grinder was purchased from Shanghai Jingxin Technology Co., Ltd. (JXFSTPRP-24; Shanghai, China). The balance was purchased from Sartorius (BSA124S-CW; Göttingen, Germany). Other equipment includes Nexera UHPLC LC-30A Shimadzu (Kyoto, Japan); high-resolution mass spectrometer Triple TOF 5600 AB Sciex (Foster City, CA, USA); and ACQUITY UPLC BEH C18 column (1.7 μm 2.1100 mm; Waters, USA).

2.2. Sample Preparation

2-Chloro-L-phenylalanine (5.03 mg) was precisely weighed, and the volume was made up with 70% methanol in a 25 mL Erlenmeyer flask. This solution was mixed and used as the mother liquor. It was diluted twentyfold (10.06 μg/mL) to yield the working solution to prepare test samples.

Ten parts of A. senticosus roots, leaves, and seeds was extracted individually from random batches, pulverized in a grinder, and passed through a No. 3 sieve. After sieving, 1 g of the sample was weighed, fixed it to 50 mL of the working solution, and extracted using ultrasonication (250 W) for 30 min at room temperature. After resting at room temperature for 1 h, the volume was made up using the working solution and centrifuged at 12000 rpm at 4°C for 15 min. The supernatant was filtered through a 0.22 μm microporous filter membrane into a 2 mL sample bottle and used as the test sample. To prepare the quality control sample (QC), all test samples were mixed together.

2.3. UPLC/QTOF-MS Conditions

A Shimadzu Nexera UHPLC LC-30A ultra-high-performance liquid chromatography system was used; the flow rate was set to 0.4 mL/min, and the volume of the sample injection was 5 μL. The mobile phase comprised 0.1% formic acid in water (A) and acetonitrile (B). The multistep linear elution gradient program was as follows: 0–3.5 min, 95–85% A; 3.5–6 min, 85–70% A; 6–6.5 min, 70–70% A; 6.5–12 min, 70–30% A; 12–12.5 min, 30–30% A; 12.5–18 min, 30–0% A; 18–22 min, 0–0% A. A UPLC BEH C18 column (1.7 μm2.1100 mm; Waters) was used.

The AB 5600 Triple TOF mass spectrometer can collect primary and secondary mass spectrometry data based on IDA (Information Dependant Acquisition) function (Analyst TF 1.7; AB Sciex). In each data collection cycle, the molecular ions with the strongest intensity and greater than 100 were selected to obtain the corresponding secondary mass spectrometry data. Bombardment energy: 40 eV, collision energy difference: 20 V, temperature: 550°C. To ensure the quality of the final collected data and method, the TOF-MS was calibrated after the analysis of every 4 samples. The relative standard deviations (RSDs) of the retention times and typical peak intensities (including internal standards) in the QC samples were used to evaluate data quality [18].

2.4. Data Analysis

Mass spectrometry data were collected in the positive and negative ion modes using UPLC-Q-TOF/MS, and Progenesis QI software was used to import the original mass spectra [19]. 2-Chloro-L-phenylalanine was used as an internal standard to normalize the data. Principal component analysis (PCA) and orthogonal projection potential structure discriminant analysis (OPLS-DA) were used to ensure data quality and model reliability. The default 7-fold cross-validation and 200 random permutation tests were performed using SIMCA-P to avoid overfitting of the OPLS-DA model. In the OPLS-DA model, the data with VIP scores >1 and value <0.05 were selected as differential metabolites. The existing components in A. senticosus were used to establish an MS/MS database and perform material identification on the data.

2.5. Protective Effects of Different Parts of A. senticosus in H2O2-Induced Oxidative Damage in PC12 Cells

All samples obtained from the metabolomic processing method were filtered through a microporous filter membrane, dried under a stream of nitrogen, weighed, and reconstituted with dimethyl sulfoxide (DMSO) to a concentration of 100 mg/mL. The experiment was divided into a control group, H2O2 injury model group, and drug group. The control group and the 6 administration groups were treated with 200, 160, 120, 80, 40, and 20 μg/mL, and the optimal dose was determined to be 80 μg/mL. Next, a H2O2-induced oxidative damage model of PC12 cells was established. We found that when the H2O2 concentration was 400 μm/L, the cell viability was closest to half of the control group. Thus, we chose 400 μm/L as the concentration for studies using the oxidative damage model. After the preliminary experiment, the concentration of all samples was set to 80 μg/mL, and the concentration of oxidative damage modeling was 400 μm/L. PC12 cells were planted in a 96-well plate with a seed plate density of 1.25104/well. After 24 hours of incubation, except for the control group and the model group, the other groups were given the above-mentioned sample at a concentration of 80 μg/mL in each group of cells. After continuing to incubate for 12 hours, except for the control group, the other groups added 100 μL of 400 μm/L H2O2 to damage PC12 cells. After 12 hours, add 10% MTT to react for 4 hours. Next, DMSO was added to dissolve the formazan crystals. Lastly, the absorbance was measured at 570 nm to determine the effect of the intervention of samples on the survival rate of PC12 cells in H2O2-induced oxidative stress.

3. Results

3.1. Multivariate Analysis of Plant Metabolism Data

A total of 130 compounds, including 93 in ESI+ mode and 37 in ESI− mode, were identified or tentatively characterized from three parts of A. senticosus. By matching the retention time, precise molecular mass, and fragment ions of the metabolites with the local database and combining with the literature, their structure was determined. The mass spectrum information is listed in Table S1.

To determine the effects of different metabolites of A. senticosus, the metabolic data of the three different parts of this plant were analyzed using multivariate statistical analysis. PCA is the most commonly used method in metabolomics research [20]. It shows the internal structure to make the data variables clearer, so that high-dimensional images can be converted into low-dimensional images with as little loss of spatial information as possible [21]. Sample classification information can be obtained from the graph. In the present study, the known data were analyzed using PCA and good results have been obtained. The PCA of all samples including the QC sample is represented as a scatter plot and a loading plot (Figure 1). The metabolites of the three different parts were significantly different, and the separation trend was obvious. The peak intensity and retention time of the ion peaks in the QC sample (including the internal standards) in the positive and negative ion modes had a high degree of overlap (Figures 1(a) and 1(b)). The data quality met the requirements of statistical analysis.

The signal responses of the different parts of A. senticosus in the ESI+ and ESI− modes were combined, and their distribution was analyzed. The results showed significant differences in the metabolites between the two modes. In Figure 2, the PCA and OPLS-DA plots show that the roots, leaves, and seeds have good clustering. Pareto diagrams show that the model is not overfitted. The quality of OPLS-DA is usually evaluated based on R2Y and Q2 and in our study, and these values were determined to be 0.989 and 0.983, respectively, in the positive mode and 0.994 and 0.987, respectively, in the negative mode. These results indicated that the OPLS-DA model was reliable.

3.2. Identification of Different Metabolites from the Different Parts of A. senticosus

After processing the data using multivariate statistical methods, univariate statistical analysis (UVA) was used to retain the data with VIP >1 and a significance of for the differential metabolites. Based on the literature review of the existing components of A. senticosus, using information from the secondary mass spectrometry database, and the application of the corresponding fragmentation law to identify the data containing the peaks corresponding to the secondary fragments, the relationship between the different components of the plant obtained from the different parts was established. The chromatograms of extracts from three different parts of A. senticosus showed that there were 130 different metabolites: 39 in roots, 51 in seeds, and 40 in leaves (Figures S3, Table S3). As seen in Figure 3, the accumulation of 130 metabolites is affected by the different parts and shows obvious changes in the heat map. In cluster 1, the flavonoids—including genkwanin, hyperoside, morin, rutin, quercitrin, and avicularin—and terpenoids—including hederagenin, ciwujianoside C1, ciwujianoside D2, and 3-coumaric acid—appear to be accumulated at high concentrations in the leaves. These levels were similar to the concentration of the compounds extracted from A. senticosus leaves [22, 23]. The lignans—namely, matairesinol, (+)-lirioresinol B, acanthoside B, and eleutheroside E—and the phenolic acids—namely, gentisic acid, caffeic acid, vanillin, and coniferyl aldehyde—in cluster 2 were present in high concentration in the roots. These findings were consistent with the reports of compounds isolated and identified from the roots of A. senticosus [24]. Amino acids—including L-leucine, L-2-aminoadipic acid, 3-aminopentanoic acid, and N-acetyl-L-glutamic acid—and phenolic acids—including p-hydroxyphenyllactic acid, L-tyrosine, 3-amino-2-naphthoic acid, and 2-furancarboxylic acid—were present in high concentration in the seeds in cluster 3. These phenolic acids have also been reported in previous studies [25, 26]. It is well known that plants containing high concentrations of phenols and flavonoids have various health benefits such as antioxidant effects [27, 28].

3.3. Effects of Different Parts of A. senticosus in Protecting PC12 Cells from H2O2-Induced Oxidative Stress and Correlation Analysis of Its Different Metabolites

To evaluate the neuroprotective effects of the different parts of A. senticosus, we used a total of 30 samples from 3 different parts of the plant (roots, leaves, and seeds) to protect PC12 cells from the oxidative stress damage induced by H2O2. The results are shown in Figure 4, 10 batches of samples from roots all showed significant activity, 9 batches of samples from leaves showed significant activity, and 8 batches of samples from seeds showed significant activity.

We calculated the correlation coefficient between the data of the different metabolites of the different parts of A. senticosus (roots, seeds, leaves) and their protective effects in H2O2-induced oxidative stress damage of the samples of that part, using Pearson’s method. As shown in Figure 5, 52 compounds with significant correlations were found, of which 20 were positively correlated metabolites and 32 were negatively correlated metabolites. Using relative quantitative analysis of metabolomics, it was found that the relative content of the positively related substances in the roots was slightly higher than that in the leaves and seeds, whereas the relative content of negatively related substances in the seeds and leaves was slightly higher than that in roots. This is consistent with the findings that the root of A. senticosus could potentially be used in the management of neurodegenerative diseases [29].

3.4. Biological Significance of Related Metabolites

Our findings indicated that the relative contents of vanillin, sucrose, eleutheroside E, eleutheroside B, 2′-hydroxy-4′-methoxyacetophenone, dimethylfraxetin, and acanthoside B among others with significant positive correlation in the roots of A. senticosus were slightly higher than those in the leaves and seeds. Vanillin is an abundantly available and inexpensive natural product, which has significant antioxidant, anti-inflammatory, and neuroprotective effects [30, 31]. Sucrose is an energy carrier, which provides plants with the energy and carbon needed for growth and development and also improves their antioxidant activity [32, 33]. Previous studies report that eleutherosides E and B have significant antioxidant and neuroprotective effects and can delay the progression of AD [34, 35]. Acanthoside B can improve the symptoms caused by scopolamine in amnestic mice by regulating cholinergic function and restoring antioxidant status [36]. 2′-Hydroxy-4′-methoxyacetophenone is an important phenolic compound constituting Chinese herbal medicine. It can inhibit endoplasmic reticulum stress-mediated oxidative stress and improve the endothelial function in mice [37].

4. Conclusion

In this study, UPLC-Q/TOF-MS was used to analyze the plant metabolomics of different parts (roots, seeds, and leaves) of A. senticosus. Pearson’s correlation analysis was used to determine the effects of the different metabolites in H2O2-induced oxidative stress in PC12 cells. Our results showed that there were 130 different metabolites in the different parts of A. senticosus, of which 52 were significantly related to the H2O2-induced oxidative stress in PC12 cells. Among the 52 metabolites, 20 and 32 were found to be positively and negatively correlated, respectively. The results from our study indicated that the efficacy of the roots of A. senticosus was slightly stronger than that of the seeds and leaves. The method of this study based on plant metabolomics can provide a detailed metabolomics overview of different parts of a plant and prove to be useful in performing an overall metabolomics analysis. The findings from our study will help better understand the metabolism of A. senticosus, provide important data for the study of different parts of the plant, and promote the rational development of medicinal plants.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare no conflicts of interest regarding publication of this study.


This research was funded by the National Key Research and Development Project (2019YFC1712304), the Natural Science Foundation of Jiangxi Province (No. 20171ACB20039), Jiangxi Province Talent Project (No. [2016]332), Nanchang Innovative Talent Team (No. [2018]274), and Nanchang Key Laboratory of Active Ingredients of Traditional Chinese Medicine and Natural Medicine (2019-NCZDSY-011).

Supplementary Materials

Supplementary figure: TIC of all QC samples and TIC of different parts of A. senticosus. (Supplementary Materials)