Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

A strategy for quality evaluation of salt-treated Apocyni Veneti Folium and discovery of efficacy-associated markers by fingerprint-activity relationship modeling


In this study, a fingerprint-activity relationship between chemical fingerprints and hepatoprotective activity was established to evaluate the quality of salt-treated Apocyni Veneti Folium (AVF). Characteristic fingerprints of AVF samples exposed to different concentrations of salt were generated by ultrafast liquid chromatography tandem triple time-of-flight mass/mass spectrometry (UFLC-Triple TOF-MS/MS), and a similarity analysis was performed based on common characteristic peaks by hierarchical clustering analysis (HCA). Then, the hepatoprotective activity of AVF against CCl4-induced acute liver damage in mice was investigated by assessing biochemical markers and histopathology, which showed that a high dose of AVF exposed to low levels of salt stress produced a marked amelioration of hepatic damage compared with the other salt-treated AVF. Finally, fingerprint-activity relationship modeling, which was capable of discovering the bioactive markers used in the quality evaluation, was investigated by the chemical fingerprints and the hepatoprotective activities utilizing multivariate statistical analysis, gray correlation analysis (GCA) and bivariate correlation analysis (BCA). The results showed that the accumulation of polyphenols, such as flavonoids and phenolic acids, in AVF subjected to low levels of salt stress could result in the effective scavenging of free radicals. Therefore, the present study may provide a powerful strategy to holistically evaluate the quality of salt-treated AVF in combination with chemical fingerprint and bioactivity evaluation.


Apocynum venetum L., as one of the known medicinal halophytes, has attracted much attention in terms of its antioxidant properties associated with human well-being. This plant grows widely in China, especially in a variety of saline habitats, and therefore, it is characterized by a high physiological plasticity for salt tolerance. Salinity has adverse effects on all aspects of plant health by activating salinity-induced molecular networks associated with salt stress perception, ion and osmotic homeostasis, and regulation of stress-related genes, proteins and metabolic pathways1,2. However, halophytes develop a robust and sophisticated protective system and accumulate secondary metabolites under appropriate levels of salt stress, which may be responsible for maintaining homeostasis and protecting against excessive reactive oxygen species-induced oxidative stress3.

Apocyni Veneti Folium (AVF) contains abundant bioactive compounds, including the main and prominent antioxidative constituents of phenolic acid and flavonoids4. Modern pharmacological studies have demonstrated that AVF has anti-hypertension, anti-depressant, hepatoprotection, anti-anxiety, antioxidation and diuretic functions5,6. In the Chinese Pharmacopoeia (2015 edition), hyperoside, a stable active ingredient of AVF occurring at a high concentration, was selected as the marker for the quality control analysis of AVF7. In fact, most published reports quantify one or a limited number of components to achieve quality evaluation8,9. However, considering the fluctuations in chemical compositions and contents on the basis of many factors, such as abiotic stress, cultivation region and harvest time, it is difficult to achieve consistent quality in medicinal herbs3. Furthermore, it is generally recognized that Chinese medicines exert therapeutic efficacies holistically through a ‘multicomponent, multitargeted, and multipathway’ mode10. In addition, as mentioned above, AVF is composed of many bioactive components, and its therapeutic effects are not confined to the individual or simple effects of a single bioactive component. Therefore, it is necessary to establish a truly meaningful protocol to effectively and systematically achieve quality control of this plant.

The fingerprint test proved to be a useful tool in assessing the chemical consistency of traditional Chinese medicine (TCM) and has been widely accepted by the World Health Organization (WHO), China Food and Drug Administration (CFDA), the United States Food and Drug Administration (FDA), and European Medicines Agency (EMEA). Several chromatographic techniques, including liquid chromatography (LC)11, gas chromatography (GC)12, thin-layer chromatography (TLC)13, and high-performance capillary electrophoresis (HPCE)14, have been used to construct chemical fingerprints of AVF, which has allowed researchers to visualize and identify as many components of the plant as possible. Among these techniques, a high-performance liquid chromatography (HPLC) system is typically employed to establish the chemical fingerprints and quality assessment of AVF because of its ease of operation, high selectivity, and accuracy9,11,15. However, limits of this HPLC system include the identification of unknown components. Accurate masses and molecular formulae of untargeted compounds acquired from LC coupled with mass spectrometry/mass spectrometry (LC-MS/MS) have been used to predict and find such components5,16,17. In particular, triple time-of-flight mass spectrometry/mass spectrometry (Triple TOF-MS/MS) is considered to be the first accurate, high-throughput and high-resolution system of its kind and operates by means of information-dependent acquisition18,19. Therefore, the ultrafast liquid chromatography (UFLC)-Triple TOF-MS/MS system could be introduced to obtain AVF chemical fingerprints. Although similarity analysis of chromatographic fingerprints can directly reflect whether an analyzed sample is chemically similar to others, it alone cannot assess quality consistency. Many studies have proven that samples with high similarity values do not always exhibit the expected equivalent efficacies20,21. Hence, fingerprint-activity relationship modeling correlating major components to bioactivity is meaningful to achieve quality consistency and discover bioactive markers22,23,24,25.

Liver injury induced by carbon tetrachloride (CCl4) is one of the most widely used experimental models26. Increased serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and hepatic malondialdehyde (MDA) contents along with decreased superoxide dismutase (SOD), catalase (CAT) and peroxidase (POD) activities demonstrate hepatotoxicity induced by CCl4 in mice27. Many published studies have indicated that natural substances, such as phenolic acids and flavonoids from plants, exhibit strong antioxidative activity that could act against CCl4-induced liver damage28,29. References have shown that AVF has protective effects against CCl4-induced acute liver injury due to its antioxidative, anti-inflammatory and immunomodulatory constituents6.

The main objectives of the present study are thus to chemically and biologically characterize the quality of salt-stressed AVF by fingerprint-activity relationship modeling and discover efficacy-associated bioactive markers. First, the chromatogram fingerprints of AVF exposed to salt stress were established by UFLC-Triple TOF-MS/MS. Then, hierarchical clustering analysis (HCA) was used to discriminate fingerprints based on common characteristic peaks to support the chemical consistency of AVF. Second, CCl4-induced acute liver damage in mice was selected to evaluate the consistency of bioactivity according to the quantitative parameters of enzyme activities and histopathological assessment in vivo. Third, the fingerprint-activity relationship between common characteristic peaks and efficacy was modeled using multivariate statistical analysis, gray correlation analysis (GCA) and bivariate correlation analysis (BCA) to assess the quality of AVF in response to salt,  and screen out the efficacy-associated bioactive markers and the underlying biological pathways.

Materials and Methods

Chemicals and materials

The diagnostic kits for alanine aminotransferase (ALT; serial NO., C009-3-1), aspartate aminotransferase (AST; serial NO., C010-3-1), malondialdehyde (MDA; serial NO., A003-1-2), superoxide dismutase (SOD; serial NO., A001-1-2), catalase (CAT; serial NO., A007-2-1), peroxidase (POD; serial NO., A084-1-1) and protein (serial NO., W042) were purchased from the Institute of Biological Engineering of Nanjing Jiancheng (Nanjing, China). References of silymarin (UV ≥80%), hyperoside and quercetin were obtained from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China); isoquercitrin and chlorogenic acid were purchased from Baoji Chenguang Biotechnology Co., Ltd. (Baoji, China). Cryptochlorogenic acid was acquired from Chengdu Chroma Biotechnology (Chengdu, China). HPLC-grade acetonitrile was purchased from Merck (Darmstadt, Germany). CCl4 and olive oil were purchased from Aladdin (Shanghai, China).

Sample preparation

Salt stress experiments have been described in detail in our previous papers3. Briefly, 4 salt treatment concentrations were used: 0 (control), 100 (low stress), 200 (medium stress) and 300 (high stress) mM NaCl treatments were designed with 3 replicates and 3 pots per replicate. After 12 h of the last salt treatment experiment, AVF samples were harvested and dried at room temperature. Naturally dried, salt-treated samples were powdered and passed through a 60 mesh sieve. Four groups of dried samples (20 g) were extracted twice with 200 mL of 70% (v/v) ethanol for 2 h under reflux followed by centrifugation at 3,500 rpm for 10 min, and the supernatant was collected. After rotary evaporation and freeze-drying to a powder, a small portion was redissolved in 70% ethanol to form a 1 mL solution containing 0.05 g of AVF, and the solution was centrifuged at 12,000 rpm for 15 min. Then, the supernatant was stored at 4 °C and filtered through a 0.22 μm membrane before being subjected to UFLC-Triple TOF-MS/MS analysis. The rest of the freeze-dried powder was redissolved with 0.5% sodium carboxymethyl cellulose (CMC-Na) to form a 1 mL solution containing 0.5 g of AVF for the animal experiments. Mixed reference standard solutions were prepared by accurately weighing corresponding amounts of reference compounds dissolved in 70% ethanol to obtain a final concentration of approximately 1 μg/mL.

UFLC-Triple TOF-MS/MS fingerprint analysis and clustering analysis

MS data were recorded by a high-resolution quadrupole time-of-flight mass spectrometer (Triple TOFTM 5600 System-MS/MS, AB Sciex, Framingham, MA, USA) equipped with an electrospray ionization (ESI) source. Chemical fingerprints were obtained and peak identification was performed via the UFLC-Triple TOF-MS/MS system. An XBridge® C18 column (100 mm × 4.6 mm, 3.5 μm) was used for the analysis. The mobile phase consisted of water containing 0.1% formic acid (A) and acetonitrile containing 0.1% formic acid (B). The samples were eluted using a linear gradient program as follows: 0–3 min, 5% B; 3–8 min, 5–18% B; 8–12 min, 18–20% B; 12–15 min, 20% B; 15–17 min, 20–60% B; 17–18 min, 60–80% B; 18–18.5 min, 80–5% B; and 18.5–22.1 min, 5% B. The flow rate was set at 0.8 mL min−1 with the column maintained at 30 °C, and the injection volume was 5 μL. Runs under the positive and negative ion modes were set separately as follows: nebulizer pressure, 55 psi; drying gas pressure, 55 psi; curtain gas pressure, 40 psi; source temperature, 550 °C; and capillary voltage, 5500 V and −4500 V, respectively, for the runs under positive and negative ion modes. Data were acquired for each sample from 50 to 1,500 Da, and dynamic range enhancement was applied throughout the MS experiment to ensure accurate mass measurement.

Based on the established methods, precision and repeatability were assessed independently by successively analyzing 6 injections of sample 1 solution and 5 replicates of 100 mM NaCl-stressed AVF solution, respectively. Stability was assessed by analyzing one sample over a 48 h time period (0, 4, 8, 12, 24 and 48 h). “Chinese traditional medicine chromatographic fingerprint similarity evaluation system 2004, A Edition” was used to correct the retention time of each peak. The peak area was processed by equalization to obtain quantitative data. The reference atlas was generated with the median method from a general comparison of chromatograms obtained for the 12 samples. Similarity between the reference fingerprints and various chromatograms was determined by the software. Common characteristic peaks were identified by comparison with the reference compounds, while for compounds without a standard reference, the characteristic peaks were inferred using PeakView 1.2 software (AB SCIEX, USA), references on AVF5,6,9 and online resources, such as HMDB (, SciFinder ( and PubChem ( through comparing MS/MS fragment ions.

HCA was introduced to cluster samples based on common characteristic peaks by Cluster 3.0 and Java Treeview 3.0. Clustering was based on the Euclidean distance coefficient and average linkage method. The mean values of all parameters were taken from the measurements of three samples, and every sample had three replicates from which the standard deviations were calculated.

Assessment of liver function

SPF-grade ICR male mice (weighing 18–22 g) were obtained from Qinglong Mountain Animal Breeding Farm Limited Company (Jiangning District, Nanjing, China). All animals were maintained under standard laboratory conditions (temperature 22 ± 2 °C, relative humidity 50 ± 10%) with dark and light cycles (12/12 h). The mice were acclimatized to laboratory conditions for 3 days before the commencement of the experiment. All animal procedures were approved by the IACUC (Institutional Animal Care and Use Committee) of Nanjing University of Traditional Chinese Medicine and carried out in accordance with the Guidelines of Accommodation and Care for Animals formulated by the Chinese Convention for the Protection of Vertebrate Animals used for Experimental and Other Scientific Purposes. All efforts were made to minimize animal suffering as well as to reduce the number of animals required for experimentation.

Mice were randomly divided into five groups of 11 animals each. The experimental process is summarized in Table S1. In the control and CCl4-intoxicated groups, animals were given a single dose of 0.5% CMC-Na orally using a gavage daily for 14 days. A positive control group was administered silymarin (100 mg/kg body weight (bw)) as a reference drug. The low dose (0.3 g/kg bw) and high-dose (3 g/kg bw) groups were given different treatments of AVF exposed to various levels of salt stress, named L0, L100, L200, L300, H0, H100, H200 and H300 at 0.1 mL/10 g bw and a 0.2 mL volume. After 2 h of the last treatment, the control group was given olive oil (10 mL/kg), while the other groups were simultaneously injected intraperitoneally (i.p.) with 10 mL/kg of 0.3% (v/v) CCl4 in olive oil. Then, all animals were fasted for 16 h and anesthetized using sodium pentobarbital (1% in physiological saline, 50 mg/kg; i.p.). Blood was collected immediately from the fundus vein and kept at 4 °C. After death, the liver was removed and placed in chilled saline solution, dried with filter paper and excised into two portions. One part was frozen quickly using liquid nitrogen and stored at −80 °C until the preparation of hepatic homogenates for the determination of biochemical parameters, while the other portion was preserved in 10% formalin for histopathological investigations.

During the whole experiment, the general conditions and bw of mice were monitored at 0, 7 and 14 days. For biochemical analysis, assays were conducted using kits obtained from the Institute of Biological Engineering of Nanjing Jiancheng (Nanjing, China) according to the manufacturer’s protocol. Blood was centrifuged at 3,000 rpm for 10 min at 4 °C, and then the serum levels of ALT and AST were determined by colorimetry30. Liver tissues (0.1 g) were excised and homogenized using a homogenizer. The sample livers were homogenized on ice with a 9× volume of homogenate prepared by precooled saline solution (0.9%) and centrifuged at 3,500 rpm for 15 min. The tissue homogenates were inspected for the determination of total soluble proteins by the Bradford method, and the calibration curve was prepared using solutions of bovine serum albumin (BSA)31. Absorbance was recorded at 562 nm via a UV-visible spectrophotometer (Shanghai Spectrum Instrument Co., Ltd.). The MDA content was detected by the thiobarbituric acid method32; SOD activity was measured at 550 nm following the nitro-blue tetrazolium method; and the basis for the evaluation CAT activity was decomposition of H2O2 to less reactive molecules, which was recorded at 240 nm and assessed by following the kit instructional protocol3,33,34. POD activity was estimated on the basis of guaiacol peroxidation measured at 420 nm35,36.

Histopathological observation was performed as follows: liver tissues preserved in 10% neutral formalin solution for at least 24 h were embedded in paraffin and cut into 5 μm thin sections, deparaffinized, dehydrated, and stained with hematoxylin-eosin for the estimation of hepatocyte necrosis and vacuolization. Histopathological changes and injuries were observed under light microscopy (Stereo Investigator, MBF Bioscience), and photographs were captured at 400× magnification.

Fingerprint-effect relationship modeling

GCA and BCA were performed to assess the fingerprint-activity relationship model with Excel 2010 for Windows 10 and SPSS 17.0 statistical software (SPSS Statistics 17 for Windows 10, SPSS Inc.), respectively. The correlation coefficient (ξ) derived from the GCA represents the degree to which the correlation factor sequence is close to the behavioral feature. The higher the value of ξ is, the closer the sequence of associated factors to the behavioral characteristics. Detailed steps can be found in the experimental section of the supporting information. The basic idea of BCA is to study the correlation between two sets of variations37, and BCA is widely used in the quality evaluation of TCMs38. Fingerprint-activity relationship analysis was performed by transferring common fingerprint peak area and pharmacological test data into SPSS software, and then BCA was used to assess the spectrum-effect relationships between the areas of common characteristic peaks in the fingerprint and the main hepatoprotective parameters. The statistical analysis was performed by one-way ANOVA followed by Duncan’s multiple-range test to compare the means with the significance level set as 0.05 by SPSS 17.0. Biological pathway analysis was constructed based on the bioactive markers according to KEGG and MetaboAnalyst 4.0. Colors varying from yellow to red indicate that the metabolite data have different levels of significance.

Results and Discussion

The many antioxidative phytochemicals of AVF have been extensively used for countering oxidative stress and oxidation-related impairments. Moreover, a considerable amount of AVF extract compounds have been investigated and confirmed to have hepatoprotective effects6,39. Studies have demonstrated that phenolic compounds can prevent CCl4-induced lipid peroxidation and hepatotoxicity due to antioxidative activities40. Therefore, in this study, fingerprint-activity relationship modeling by UFLC-Triple TOF-MS/MS analysis was performed to assess the quality of AVF exposed to different concentrations of salt and discover bioactive markers associated with efficacy of the AVF in vivo.

Fingerprints and similarity analysis of AVF samples

The representative base peak chromatograms obtained from the analysis under the positive and negative ion modes of UFLC-Triple TOF-MS/MS for the reference mixture and AVF were obtained under the optimal conditions (Fig. S1). A total of 12 peaks with large areas and good separation that were found in all chromatograms of the samples under the positive ion mode were regarded as “common characteristic peaks”, and they were labeled with P in the fingerprints; 14 common characteristic peaks detected under the negative ion mode, labeled with N and with the corresponding peak and relative peak areas, are displayed in Fig. 1 and Table S2. Thus, the present study established fingerprints of AVF exposed to salt stress. Comparisons were performed of the tR and m/z values of the reference compounds and MS/MS fragment ions with values from references5,6,9 and online databases for the compounds without standard chromatography data, and the information for these peaks is presented in Table 1.

Figure 1

Representative UFLC-Triple TOF-MS/MS fingerprints of AVF exposed to salt stress under positive (A) and negative (B) ion modes.

Table 1 Identification of the common characteristic peaks of AVF under salt stress.

The validation of the method showed that the experimental precision was less than 0.5% for tR and 0.9% for the peak area of common characteristic peaks. The method stability was less than 0.29% for tR and 1.9% for the peak area of common characteristic peaks. The repeatability was less than 0.3% for tR and 1.8% for the area of common characteristic peaks. The similarity values between the reference standard and chromatographic fingerprints for 12 samples ranged from 0.7 to 0.901 and from 0.839 to 0.962 under positive and negative ion modes, respectively. The results of precision, stability and repeatability analysis suggested the validity and suitability of the optimized method for analyzing samples.

A heat map derived from HCA graphically displays the changes in the accumulation of common characteristic compounds under salinity stress (Fig. 2) and the clustering of samples. Clear differentiation was observed in the positive and negative ion mode results, revealing that AVF samples could be classified into two clusters, that is, the 0/100 and 200/300 mM salt-treated samples formed clusters that were clearly separate from each other. The assessment of clustering based on common characteristic peaks was consistent with the similarity analysis both in the positive and negative ion modes and was also in agreement with our previous report on the metabolitesof AVF in response to salt stress3. A clear stress-induced trajectory for metabolomic changes was evident, indicating the dose-dependent responses of salt-stressed AVF. Thus, HCA could qualitatively compare salt-treated samples and effectively separate these samples.

Figure 2

Hierarchical clustering analysis based on the data of common characteristic peaks of AVF subjected to salt stress under positive (A) and negative (B) ion modes.

Hepatoprotective activity

The function of AVF extract in protecting against CCl4-induced hepatotoxicity was investigated to study its medicinal efficacy. Bw is a significant indicator of the negative effects of xenobiotics and is a determinant constraint in toxicity analysis. In the present study, no significant changes were observed in the coadministration of AVF to CCl4-treated mice (Fig. 3A). The outcomes were in agreement with the previous findings on AVF40.

Figure 3

Effects of AVF extract on body weight (A), ALT and AST activities (B), hepatic MDA content (C) and activities of antioxidant enzymes SOD (D), CAT (E) and POD (F) after CCl4 treatment in mice. Data are the mean ± SD (n = 6). Different letters following values (a, b, c, d, e, f and g) in the same row indicate significant differences among salt treatments using Duncan’s multiple-range test at p < 0.05.

Metabolic products of CCl4 not only affect the permeability of membranes, resulting in an increase in serum aminotransferase and lipid peroxidation but also decrease the activities of antioxidative enzymes41. ALT and AST are indications of hepatic impairment induced by CCl4 and are usually considered subtle markers of liver function. The hepatoprotective effects of AVF on the serum ALT and AST activities are shown in Fig. 3B and Table S3. In the CCl4-intoxicated model group, a significant increase in the ALT and AST activities was shown in the livers exposed to CCl4 toxicity, with values up to 130.67 and 99.77 U/L, respectively, whereas the values of the control group were only 25.4 and 24.63 U/L, respectively. Silymarin, which has been used to treat hepatotoxic diseases in clinical practice for decades42,43, significantly inhibited enzyme activities, with the ALT and AST activities reduced to 58.19% and 60.16%, respectively. Administration with a low dose (0.3 g/kg) of AVF stressed by four concentrations of NaCl for 14 days prior to CCl4-induced hepatotoxicity significantly decreased ALT activity by 29.16%, 25.89%, 20.97% and 24.16%, respectively, and AST activity by 31.14%, 31.47%, 29.84% and 23.72%, respectively, compared to the model group levels. Administration of a high dose (3 g/kg) of AVF prior to CCl4 injury markedly reduced hepatotoxicity, with the aminotransferase activities decreasing by approximately half, indicating promising effects of AVF protection against CCl4-induced damage by the regulation of enzymatic expression to a level similar to that of silymarin. Our results were in accordance with other findings in which treatment with a crude methanol extract of medicinal plants restored the levels of liver biochemical markers after CCl4 treatment35.

Pretreatment with AVF effectively inhibited CCl4-induced hepatotoxicity and significantly reduced the level of MDA, which was inversely proportional to the salt dose (Fig. 3C and Table S3). Briefly, an increase was shown in MDA content with increasing salt concentrations. At the same time, coadministration with AVF extract at a high dose significantly inhibited hepatic MDA formation, having nearly the same effect as silymarin, and coincidentally, there was no significant difference between the five groups, indicating that AVF could effectively inhibit MDA formation or eliminate excess free radicals.

SOD, CAT, and POD activities were also investigated to study the efficacy of AVF in the ability to restore and maintain activities in CCl4-injured livers (Fig. 3D–F and Table S3). Specifically, SOD catalyzes the dismutation of O2 into O2 and H2O2, whereas CAT and POD are responsible for the removal of H2O23,44. The increased production of free radicals was a major cause of the significantly reduced SOD activity in the model. In the model, SOD, CAT and POD activity was markedly reduced by nearly half compared with the control levels; however, mice administered a low dose of AVF showed varied responses. In detail, SOD activity increased significantly compared with that of the model group except in the L0 group, and the L100 group showed the highest value. The levels of SOD activity induced by salt-stressed AVF from high to low were found with H100, H200, H300 and H0, indicating that low and medium salt levels altered the plant constituents associated with liver-protecting activity. For CAT, no significant difference was shown among any group treated with a low dose of AVF exposed to salt stress, but each group showed higher CAT activity than the model group. In addition, the CAT activity in mice pretreated with watering AVF did not change significantly compared to that of the salt-treated groups at a high dose. This result suggested that salt-stressed AVF was superior to the control and that the low-salt treatment worked better at elevating antioxidant enzyme activity3. In addition, POD activity did not show marked changes among the low-dose groups except for L100, which showed significantly increased levels compared to those of the control; however, POD activity increased significantly in the high-dose groups compared with the control group, and POD activity decreased with increasing salt concentrations, revealing the failure of this enzyme to scavenge excessive free radicals.

The protective effects of AVF against CCl4-induced hepatotoxicity were further confirmed by histopathological observations. Representative photographs of liver sections stained with hematoxylin-eosin in hepatic tissues under microscopy (400×) are shown in Fig. 4. In the control mice, normal histology was observed, with a well-preserved cytoplasm and prominent nucleus in hepatic cells. In CCl4-treated mice, extensive liver injuries were observed in hepatocytes, characterized by hepatocellular degeneration and necrosis around the central vein, inflammatory cell infiltration, ballooning degeneration, and the loss of cellular boundaries. However, coadministration with a low dosage of AVF prior to CCl4 treatment reduced the extent of histopathological injury. Furthermore, silymarin and high-dose AVF, especially 100 mM salt-stressed AVF, restored the altered parameters to the control levels. These results were in good agreement with the serum aminotransferase and hepatic antioxidant enzyme activity results, and similar results have been found for Brachychiton populneus extract against CCl4-induced liver injury35. Although the effects of salt-treated AVF manifested differently with respect to liver protection, the liver injury was remarkably ameliorated by pretreatment with H100, resulting in lower hepatocellular degeneration and less hydropic necrosis.

Figure 4

Representative photographs of histological liver damage after CCl4 treatment in mice. (A) Control group. (B) CCl4-intoxicated model group. (C) Shows mice tissue pretreated with silymarin prior to CCl4 treatment. (DG) Showed tissue pretreated with 0, 100, 200 and 300 mM salt-stressed AVF (0.3 g/kg) prior to CCl4 intoxication, respectively. (HK) Were treated with 0, 100, 200 and 300 mM salt-stressed AVF (3 g/kg), respectively, prior to CCl4 intoxication.

Fingerprint-activity relationship modeling analysis

The assessment of spectrum-effect relationships represents a powerful tool for the quality control of herbal medicines. However, the key difficulty in the application of spectral relationships to TCMs is how to associate complex chromatographic peaks with pharmacodynamic information24,45. Fingerprint-activity relationship modeling of TCM is based on the guidance of TCM theory. The fingerprints and pharmacodynamics of TCMs, which are correlated by chemometric models, establish a comprehensive evaluation system for revealing the material basis of the pharmacodynamics of TCMs46. This has been one of the approaches used for predicting active ingredients and performing quality control in natural products. In this study, multivariate statistical analysis was conducted to associate chemical composition with efficacy to evaluate AVF quality and, furthermore, to discover bioactive markers. GCA is often used to reveal quantitative comparisons of trends in dynamically changing systems and is suitable for solving problems with complicated interrelationships between multiple factors and variables47,48. GCA also provides a reliable method for the quality evaluation of TCMs. Compared with other analysis methods, such as regression analysis and canonical correlation analysis, GCA has several advantages, such as requiring only a small sample size and a small amount of calculations and providing intuitive results49,50. GCA can be used to assess the size of the relevance between the efficacy index and the chromatographic peak and offers a possibility for predicting the active components51. BCA is also known as canonical correlation analysis. The basic idea of BCA is to study the correlation between two sets of variations19,37, revealing much of the information about them and finding their linear combinations that have the highest correlation22. The Pearson correlation coefficient has a clear quantitative meaning and can reflect the magnitude and direction of the linear correlation between variables52. Although the outcomes analyzed by GCA and BCA were not completely consistent, there were either synergetic or antagonistic effects among these components52.

In the GCA, the correlation coefficient of peaks was between 0.581 and 0.963 (Table 2). Specifically, 7 peaks were closely correlated with efficacy, including peaks P1, P4, P5, N1, N9, N10 and N12, and the top five compounds with large correlation coefficient values for each indicator were selected as markers, so 17 peaks (P1, P2, P3, P4, P5, P8, P10, P11, N1, N3, N4, N6, N8, N9, N10, N11 and N12) were screened out. Furthermore, BCA clearly showed that P10, P12 and N6 were correlated with ALT with the absolute value of the Pearson correlation coefficient (r) being greater than 0.5 (Table 3). Similarly, N11 and N14 were negatively correlated with AST; for MDA, the same was true for N6 and P10. In addition, P10 appeared to have a negative correlation with SOD activity, while N9 and N14 were positively corrected with CAT activity and N8 and N11 were positively corrected with POD activity. Therefore, 7 peaks (P10, P12, N6, N8, N9, N11 and N14) were screened out. In other words, these peaks might represent the main hepatoprotective components in the AVF extract, but further study is needed to identify the chemical structures and to confirm their bioactivities. These results will be helpful for the study of AVF and other herbs to search for their effective components. In addition, the relative contents (peak areas) showed that the quality level was highest in samples treated with a low concentration of salt compared to that of other groups. This further implies that the spectrum-effect relationship method is efficient for quality evaluation.

Table 2 Gray correlation analysis between fingerprints and the efficacy of AVF exposed to salt stress.
Table 3 Bivariate correlation analysis between fingerprints and the efficacy of AVF exposed to salt stress.

Biomarker identification

Based on the GCA and BCA results, 19 peaks (P1, P2, P3, P4, P5, P8, P10, P11, P12, N1, N3, N4, N6, N8, N9, N10, N11, N12 and N14) were correlated with pharmacodynamic effects. P2/N3, P5/N6, P8/N8 and P11/N11 were identified as the same compound, respectively. Therefore, 15 bioactive markers were obtained, including organic acids (citric acid/ isocitric acid and shikimic acid), flavonoids (quercetin, hyperoside, isoquercitrin, procyanidin B2, quercetin 3-O-(6″-O-malonyl)-β-D-glucoside, acetylated isoquercitrin, and quercetin 3-O-(6″-O-malonyl)-β-D-galactoside), terpenoids (allamandin), phenylpropanoids (chlorogenic acid and cryptochlorogenic acid), quinones (hyperforin) and two unknown compounds (P3 and N4). This finding revealed that the defensive properties of AVF might be attributed to the presence of potent antioxidant constituents preventing tissue degeneration by inhibiting lipid peroxidation of the liver.

The metabolites citric acid/isocitric acid and shikimic acid are key metabolic components in the tricarboxylic acid cycle (TCA cycle), and the alteration of these metabolites might disrupt energy metabolism. Citric acid acted as a potential biomarker, showing hepatoprotective effects and contributing directly to the therapeutic effect of Angelica sinensis53, and shikimic acid was associated with the biosynthesis of phenylpropanoids. Published studies on chlorogenic acid have shown that it plays several important roles, such as displaying hepatoprotective activity by suppressing liver fibrogenesis and carcinogenesis, modulating lipid metabolism and scavenging free radicals54,55, while cryptochlorogenic acid isolated from Artemisia capillaris possesses potent activity against HBV DNA replication56.

The most abundant bioactive flavonoid constituents correlated with hepatoprotective activity were quercetin, procyanidin B2, hyperoside, isoquercitrin, and derivatives of hyperoside/isoquercitrin. The potential mechanism of quercetin-induced hepatic protection is mainly mediated through its powerful antioxidative capacity, inhibition of hepatocyte apoptosis and suppression of inflammatory cytokines via signaling pathways57,58. Procyanidin B2 was investigated against hepatic oxidation in diabetic rats and exhibited a protective effect against CCl4-induced hepatic injury by elevating the antioxidative defense potential and consequently suppressing the inflammatory response59,60. The hyperoside and isoquercitrin in AVF could be potential natural hepatoprotective agents protecting the liver from injury though inhibition of oxidative stress as well as regulation of acetaminophen metabolism, and reports revealed that administration of these compounds effectively restored liver functions39,61. Although it has been speculated that the derivatives of hyperoside and isoquercitrin are biotransformed into corresponding glycosides62 or partly involved in the hepatoprotective function of AVF39, little research has been performed on their biological activities. In addition, as a phloroglucinol derivative, hyperforin, found in Hypericum perforatum L., provided protection against free radical-induced DNA damage based on its scavenging activity63; however, there is currently no related information on the hepatoprotective activity of hyperforin or allamandin.

The most significantly impacted biological pathways (Fig. 5 and Table S4), e.g., flavonoid biosynthesis and flavone and flavonol biosynthesis, were in line with our previous results3. In terms of the heat map, it seemed that a low concentration of salt stress could increase the accumulation of secondary metabolites, and AVF exposed to low levels of salt stress might be considered a source for dietary supplementation of polyphenols.

Figure 5

Pathway enrichment analysis showing the significantly changed biological pathways of AVF under salt stress.


In this study, fingerprint-activity relationship modeling was established to perform a quality evaluation of salt-tolerant AVF based on chemical fingerprints and efficacy. First, UFLC-Triple TOF-MS/MS fingerprints were established, and cluster analysis was performed based on the common characteristic peaks to evaluate the similarity between groups. Second, salt-stressed AVF protected against CCl4-induced acute liver injury in mice, and the effects of AVF were achieved via a reduction in serum transaminase activities, elevating hepatic antioxidative enzyme activities and ameliorating tissue lesions. Then, fingerprint-activity relationship modeling between the spectrum and the medicinal efficacy was examined by GCA and BCA to perform the quality evaluation and screen out bioactive markers for the AVF response to salt. According to the results, the abundant polyphenols might be responsible for the protective effects against acute liver damage in mice, and AVF with low-salt treatment was better in terms of efficacy than the other three groups. This strategy could serve as a useful reference for the quality evaluation, the discovery of bioactive markers and the future exploitation of salt-tolerant Chinese herbal medicines.


  1. 1.

    Neffati, M. et al. Salinity impact on fruit yield, essential oil composition and antioxidant activities of Coriandrum sativum fruit extracts. Food Chem 124, 221–225 (2011).

    CAS  Article  Google Scholar 

  2. 2.

    Dersch, L. M., Beckers, V. & Wittmann, C. Green pathways: metabolic network analysis of plant systems. Metab Eng 34, 1–24 (2016).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  3. 3.

    Chen, C. H. et al. Variations in Physiology and Multiple Bioactive Constituents under Salt Stress Provide Insight into the Quality Evaluation of Apocyni Veneti Folium. Int J Mol Sci 19, 3042 (2018).

    PubMed Central  Article  CAS  PubMed  Google Scholar 

  4. 4.

    Chen, C. H. et al. Quality Evaluation of Apocyni Veneti Folium from Different Habitats and Commercial Herbs Based on Simultaneous Determination of Multiple Bioactive Constituents Combined with Multivariate Statistical Analysis. Molecules 23, 573 (2018).

    CAS  PubMed Central  Article  PubMed  Google Scholar 

  5. 5.

    An, H. et al. Simultaneous qualitative and quantitative analysis of phenolic acids and flavonoids for the quality control of Apocynum venetum L. leaves by HPLC-DAD-ESI-IT-TOF-MS and HPLC-DAD. J Pharm Biomed Anal 85, 295–304 (2013).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  6. 6.

    Xie, W. et al. Botany, traditional uses, phytochemistry and pharmacology of Apocynum venetum L. (Luobuma): A review. J Ethnopharmacol 141, 1–8 (2012).

    CAS  PubMed  Article  Google Scholar 

  7. 7.

    Pharmacopoeia Commission of the Ministry of Health of the People’s Republic of China, Pharmacopoeia of the People’s Republic of China. Part I. Beijing: China Medical Science and Technology Press; pp 211–212 (2015).

  8. 8.

    Chen, C. H. et al. Site-specific accumulation and dynamic change of flavonoids in Apocyni Veneti Folium. Microsc Res Tech 8, 1315–1322 (2017).

    Article  CAS  Google Scholar 

  9. 9.

    Liang, T., Yue, W. & Li, Q. Comparison of the phenolic content and antioxidant activities of Apocynum venetum L. (Luo-Bu-Ma) and two of its alternative species. Int J Mol Sci 11, 4452 (2010).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  10. 10.

    Jiang, Y. et al. Recent analytical approaches in quality control of traditional Chinese medicines—a review. Anal Chim Acta 657, 9–18 (2010).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  11. 11.

    Yan, S. H. et al. Chemical composition and antioxidant activities of extracts from Apocyni Veneti Folium. Nat Prod Res 26, 600–608 (2012).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  12. 12.

    Wang, Y. Q. et al. Determination of aromatic ingredients of Apocynum venetum tea by headspace solid phase microextraction-gas chromatography-mass spectrometry. J Food Safety Quality 10, 4222–4226 (2019).

    Google Scholar 

  13. 13.

    Zhang, Y. C. et al. Identification on Folium Apocyni Veneti and its adulterant. Chin Med J Res Prac 22, 17–20 (2008).

  14. 14.

    Wang, M. et al. Studies on HPCE fingerprint of Folium Apocyni Veneti. J Chin Med Mater 33, 201–204 (2010).

    Google Scholar 

  15. 15.

    Hao, X. L. et al. HPLC fingerprint of total flavonoids of Folium Apocyni Veneti. Chin J Chinese Mater Med 33, 1968 (2008).

    CAS  Google Scholar 

  16. 16.

    Xua, J. et al. Development of a metabolic pathway-based pseudo-targeted metabolomics method using liquid chromatography coupled with mass spectrometry. Talanta 192, 160–168 (2019).

    Article  CAS  Google Scholar 

  17. 17.

    Want, E. J. et al. Global metabolic profiling of animal and human tissues via UPLC-MS. Nat Protoc 8, 17–32 (2013).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  18. 18.

    Genna, L. et al. Performance Characteristics of a New Hybrid Quadrupole Time-of-Flight Tandem Mass Spectrometer (TripleTOF 5600). Anal Chem 83, 5442–5446 (2011).

    Article  CAS  Google Scholar 

  19. 19.

    Liu, X. et al. Investigation on the spectrum-effect relationships of Da-Huang-Fu-Zi-Tang in rats by UHPLC-ESI-Q-TOF-MS method. J Ethnopharmacol 154, 606–612 (2014).

    PubMed  Article  PubMed Central  Google Scholar 

  20. 20.

    Yu, H. X., Sun, L. Q. & Qi, J. The integrated quality assessment of Chinese commercial dry red wine based on a method of online HPLC-DAD-CL combined with HPLC-ESI–MS. Chin J Nat Med 12, 517–524 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  21. 21.

    Chang, Y. X. et al. The antioxidant-activity-integrated fingerprint: an advantageous tool for the evaluation of quality of herbal medicines. J Chromatogr A 1208, 76–82 (2008).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  22. 22.

    Kong, W. J. et al. Spectrum–effect relationships between ultra performance liquid chromatography fingerprints and anti-bacterial activities of Rhizoma coptidis. Anal Chim Acta 634, 279–285 (2009).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  23. 23.

    Sun, L. Q. et al. Antioxidant anthocyanins screening through spectrum-effect relationships and DPPH-HPLC-DAD analysis on nine cultivars of introduced rabbiteye blueberry in China. Food Chem 132, 759–765 (2012).

    CAS  Article  Google Scholar 

  24. 24.

    Zhang, Y. J. et al. A strategy for qualitative and quantitative profiling of glycyrrhiza extract and discovery of potential markers by fingerprint-activity relationship modeling. Sci Rep 9, 1309 (2019).

    ADS  PubMed  PubMed Central  Article  CAS  Google Scholar 

  25. 25.

    Zheng, Q. F. et al. Spectrum-effect relationships between UPLC fingerprints and bioactivities of crude secondary roots of Aconitum carmichaelii Debeaux (Fuzi) and its three processed products on mitochondrial growth coupled with canonical correlation analysis. J Ethnopharmacol 153, 615–623 (2014).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  26. 26.

    Tsuchida, T. & Friedman, S. L. Mechanisms of hepatic stellate cell activation. Nat Rev Gastro Hepat 14, 397–411 (2017).

    CAS  Article  Google Scholar 

  27. 27.

    Puche, J. E., Saiman, Y. & Friedman, S. L. Hepatic Stellate Cells and Liver Fibrosis. Compr Physiol 3, 1473–1492 (2013).

    PubMed  Article  PubMed Central  Google Scholar 

  28. 28.

    Cheng, N. et al. Antioxidant and hepatoprotective effects of Schisandra chinensis pollen extract on CCl4-induced acute liver damage in mice. Food Chem Toxicol 55, 234–240 (2013).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  29. 29.

    Acero, N. et al. Comparison of phenolic compounds profile and antioxidant properties of different sweet cherry (Prunus avium L.) varieties. Food Chem 279, 260–271 (2019).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  30. 30.

    Reitman, S. & Frankel, S. A colorimetric method for the determination of serum glutamic oxalacetic and glutamic pyruvic transaminases. Am J Clin Pathol 28, 56–63 (1957).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  31. 31.

    Bradford, M. M. A rapid and sensitive method for the quantification of microgram amounts of protein utilising the principle of protein dye binding. Anal Biochem 72, 248–254 (1976).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  32. 32.

    Wu, F. B., Zhang, G. P. & Dominy, P. Four barley genotypes respond differently to cadmium: Lipid peroxidation and activities of antioxidant capacity. Environ Exp Bot 50, 67–78 (2003).

    CAS  Article  Google Scholar 

  33. 33.

    Sun, Y., Oberley, L. W. & Li, Y. A simple method for clinical assay of superoxide dismutase. Clin Chem 34, 497–500 (1988).

    CAS  PubMed  PubMed Central  Google Scholar 

  34. 34.

    Luck, H. Catalase, in: H. U. Bergmeyer (Ed.), Methods of Enzymatic Analysis, Academic Press, New York,; pp 855–888 (1965).

  35. 35.

    Riffat, B. et al. Brachychiton populneus (Schott & Endl.) R.Br. ameliorate carbon tetrachloride induced oxidative stress through regulation of endoplasmic reticulum stress markers and inflammatory mediators in Sprague-Dawley male rats. Biomed Pharmacother 107, 1601–1610 (2018).

    Article  Google Scholar 

  36. 36.

    Wang, X. H. & Huang, J. L. Principles and Techniques of Plant Physiological Biochemical Experiment, 3rd ed.; Higher Education Press: Beijing, China, (2015).

  37. 37.

    Gumus, E. et al. Application of canonical correlation analysis for identifying viral integration preferences. Bioinformatics 28, 651–655 (2012).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  38. 38.

    Huang, X. et al. Studies on principal components and antioxidant activity of different Radix Astragali samples using high-performance liquid chromatography/electrospray ionization multiple-stage tandem mass spectrometry. Talanta 78, 1090–1101 (2009).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  39. 39.

    Xie, W. et al. Apocynum venetum Attenuates Acetaminophen-Induced Liver Injury in Mice. Am J Chin Med 43, 457–476 (2015).

    PubMed  Article  PubMed Central  Google Scholar 

  40. 40.

    Xiong, Q. et al. Hepatoprotective effect of Apocynum venetum and its active constituents. Planta Med 66, 127–133 (2000).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  41. 41.

    Liu, J. et al. SIRT3 protects hepatocytes from oxidative injury by enhancing ROS scavenging and mitochondrial integrity. Cell Death Dis 8, e3158 (2017).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  42. 42.

    Shaker, E., Mahmoud, H. & Mnaa, S. Silymarin, the antioxidant component and Silybum marianum extracts prevent liver damage. Food Chem Toxico 146, 803–806 (2010).

    Article  CAS  Google Scholar 

  43. 43.

    Wellington, K. & Jarvis, B. Silymarin: A review of its clinical properties in the management of hepatic disorders. Biodrugs 15, 465–489 (2001).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  44. 44.

    Tasgin, E. et al. Effects of salicylic acid and cold treatments on protein levels and on the activities of antioxidant enzymes in the apoplast of winter wheat leaves. Phytochemistry 67, 710–715 (2006).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  45. 45.

    Wang, F. et al. From chemical consistency to effective consistency in precise quality discrimination of Sophora flower-bud and Sophora flower: Discovering efficacy-associated markers by fingerprint-activity relationship modeling. J Pharm Biomed Anal 132, 7–16 (2017).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  46. 46.

    Kong, W. J. et al. Combination of chemical fingerprint and bioactivity evaluation to explore the antibacterial components of Salvia miltiorrhizae. Sci Rep 7, 8112 (2017).

    ADS  PubMed  PubMed Central  Article  CAS  Google Scholar 

  47. 47.

    Lü, S. et al. Spectrum-Effect Relationships between Fingerprints of Caulophyllum robustum Maxim and Inhabited Pro-Inflammation Cytokine Effects. Molecules 22, 1826 (2017).

    PubMed Central  Article  CAS  Google Scholar 

  48. 48.

    Zhou, S. et al. Monitoring chip fatigue in an IGBT module based on grey relational analysis. Microelectron Reliab 56, 49–52 (2016).

    Article  Google Scholar 

  49. 49.

    Zhang, X. F. et al. UPLC-MS/MS analysis for antioxidant components of Lycii Fructus based on spectrum-effect relationship. Talanta 180, 389–395 (2018).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  50. 50.

    Shen, C. H. et al. Evaluation of analgesic and anti-inflammatory activities of Rubia cordifolia L. by spectrum-effect relationships. J Chromatogr B Analyt Technol Biomed Life Sci 1090, 73–80 (2018).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  51. 51.

    Shen, Y. W. & Liu, J. J. Application of the Gray Relevancy Method to College Teachers’ Research Performance Assessment of Humanities and Social Sciences. Sci Technol Manag Res 30, 237–240 (2010).

    Google Scholar 

  52. 52.

    Bai, G. Y. et al. Spectrum-effect Relationship of Anti-inflammatory Effect of Different Polar Parts of Unripe Forsythiae Fructus. Chin J Exp Trad Med Form 23, 1–6 (2017).

    Google Scholar 

  53. 53.

    Hua, Y. et al. Metabonomics study on the hepatoprotective effect of polysaccharides from different preparations of Angelica sinensis. J Ethnopharmacol 151, 1090–1099 (2014).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  54. 54.

    Salomone, F., Galvano, F. & Volti, G. Li. Molecular Bases Underlying the Hepatoprotective Effects of Coffee. Nutrients 9, 85 (2017).

    PubMed Central  Article  CAS  Google Scholar 

  55. 55.

    Muhammad, N. et al. Chlorogenic acid (CGA): A pharmacological review and call for further research. Biomed Pharmacother 97, 67–74 (2018).

    Article  CAS  Google Scholar 

  56. 56.

    Zhao, Y. et al. UFLC/MS-IT-TOF guided isolation of anti-HBV active chlorogenic acid analogues from Artemisia capillaris as a traditional Chinese herb for the treatment of hepatitis. J Ethnopharmacol 156, 147–154 (2014).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  57. 57.

    Peng, Z. et al. Hepatoprotective effect of quercetin against LPS/d-GalN induced acute liver injury in mice by inhibiting the IKK/NF-κB and MAPK signal pathways. Int Immunopharmacol 52, 281–289 (2017).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  58. 58.

    Wu, L. et al. Hepatoprotective effect of quercetin via TRAF6/JNK pathway in acute hepatitis. Biomed Pharmacother 96, 1137–1146 (2017).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  59. 59.

    Justino, A. B. et al. Hepatoprotective Properties of a Polyphenol-Enriched Fraction from Annona crassiflora Mart. Fruit Peel against Diabetes-Induced Oxidative and Nitrosative Stress. J Agric Food Chem 65, 4428–4438 (2017).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  60. 60.

    Yang, B. Y. et al. Protective Effect of Procyanidin B2 against CCl4-Induced Acute Liver Injury in Mice. Molecules 20, 12250–12265 (2015).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  61. 61.

    Xie, W. et al. Protective effect of hyperoside against acetaminophen (APAP) induced liver injury through enhancement of APAP clearance. Chem Biol Interact 246, 11–19 (2016).

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  62. 62.

    Valentová, K. et al. Isoquercitrin: pharmacology, toxicology, and metabolism. Food Chem Toxicol 68, 267–282 (2014).

    PubMed  Article  CAS  PubMed Central  Google Scholar 

  63. 63.

    Ševčovičová, A. et al. DNA-protective activities of hyperforin and aristoforin, Toxicol. In Vitro 29, 631–637 (2015).

    Article  CAS  Google Scholar 

Download references


This research was supported by the Priority Academic Program Development of Jiangsu Higher Education Institutions of China (NO. ysxk-2014) and Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX18_1606).

Author information




X.L. designed the experiments, helped to write the manuscript and critically revised it. C.C., J.C., J.S., S.C., H.Z. and Y.Y. conducted the experiments. C.C., Y.J., L.G. and F.C. performed data analysis. C.C. provided research idea, supervised the experiments, performed numerical analyses, and wrote the article.

Corresponding author

Correspondence to Xunhong Liu.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Chen, C., Chen, J., Shi, J. et al. A strategy for quality evaluation of salt-treated Apocyni Veneti Folium and discovery of efficacy-associated markers by fingerprint-activity relationship modeling. Sci Rep 9, 16666 (2019).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing