Printer Friendly

Combining metabolomic analysis and microarray gene expression analysis in the characterization of the medicinal plant Chelidonium majus L.


Background and objective: Even though herbal medicines have played an important role in disease management and health for many centuries, their present frequent use is challenged by the necessity to determine their complex composition and their multitarget mode of action. In the present study, modern methods were investigated towards their potential in the characterization of herbal substances. As a model the herbal substance Chelidonii herba was used, for which several reports on liver toxicities exist. Extracts of Chelidonii herba with different solvents were characterized phytochemically and functionally by experiments with HepG2 liver cells.

Methods: Chelidonii herba was extracted with four solvents of different polarity (dichloromethane, water, ethanol, and ethanol 50% (V/V); four replicates each). The different extracts were characterized metabolomically by [sup.1]H-NMR fingerprinting analysis and principal component analysis (PCA). The content of alkaloids was additionally determined by RP-HPLC. Functional characterization was achieved by the determination of cell proliferation and by transcriptomics techniques (Whole Genome Gene Expression Microarrays v2, Agilent Technologies) in HepG2 cells after exposure to the different extracts (four experimental replicates each).

Results: Based on data from [sup.1]H-NMR fingerprints and RP-HPLC analyses the different extracts showed a divergent composition of constituents depending on the solvent used. HepG2 liver cells responded differentially to the four extracts. Microarray analysis revealed a significant regulation of genes and signal cascades related to biotransformation. Also liver-toxic signal cascades were activated. Neither the activated genes nor the proliferation response could be clearly related to the differing alkaloid content of the extracts.

Conclusion: Different manufacturing processes lead to different herbal preparations. A systems biology approach combining a metabolomic plant analysis with a functional characterization by gene expression profiling in HepG2 cells is an appropriate strategy to characterize variations in plant extracts. Safety assessments of herbal substances may benefit from such complementary analyses.


Chelidonii herba

Herbal safety

Gene expression profiling

[sup.1]H-NMR fingerprint


Medicinal plants and products derived thereof have a long tradition of therapeutic use and are widely accepted in the population. The quality of medicinal plants is a basic requirement for the safety of herbal preparations. In terms of quality requirements. the authentication of herbal substances is indispensable. In this context molecular biological methods were tested successfully for their applicability in the identification of herbal substances (Kersten et al., 2008). For safety assessment, several established invitro and in-vivo "toxicological" methods are available: for example the comet assay determines DNA strand breakage or the T-cell-dependent antibody response assesses the antibody response to immunization. However, in the field of regulatory affairs there exists the necessity to investigate novel technologies, including functional genomics, proteomics, metabolomics, high-throughput screening and systems biology in order to replace current toxicology assays used for drug approval (Hamburg, 2011). In future investigations it is important to characterize the toxicant/s with additional complementary methods. Thus, the chemical profile of a plant-derived extract, together with data on its cellular responses, toxicological findings and extrapolated effects of dose-response investigations have to be taken into account for the compilation of an overall picture (Hartung and McBride, 2011; Krewski et al., 2010).

In contrast to single substances, plant extracts present a challenge, as they are complex mixtures. Often, plant extracts are characterized by marker substances used for the standardization of the respective herbal substances. But multi-component mixtures can have synergistic effects, which may arise due to the combination of many components of an herbal preparation (Wagner, 2006) not covered by the determination of marker substances alone.

Chelidonium majus L. (Papaveraceae) is a biennial or perennial plant native to Europe, North America and Western Asia and commonly known as swallow wort, rock poppy or greater celandine. C. majus has been used in traditional Chinese medicine, western phytotherapy, homeopathy and anthroposophy. The German Commission E indicated Chelidonii herba for the use in spastic discomfort of bile ducts and gastrointestinal tract. The fresh latex was externally used in the treatment of warts but also for other skin complaints such as corns, tinea infections, eczema and tumours of the skin. According to the German Comission D monograph on C. majus, the herbal substance is used in homeopathy for different disorders of the liver and the gallbladder, inflammation of the respiratory organs and the pleura and in rheumatism. According to the Chinese medicine, C. majus is mainly used to treat blood stasis, to relieve pain, to promote diuresis in oedema and ascites, to treat jaundice and to relieve cough (Gilca et al., 2010).

Several in-vivo and in-vitro studies suggest anti-viral, anti-microbial, anti-spasmodic, choleretic, anti-inflammatory, immunomodulatory, analgesic and anti-tumour effects for preparations containing C. majus or isolated constituents thereof (Gilca et al., 2010).

In this study Chelidonii herba was used as a model for our investigations. In literature, there are several reports on liver toxicity associated with application of herbal preparations derived from C. majus (Teschke et al., 2012; Stickel et al., 2003; De Smet, 2002; Strahl et al., 1998).

The secondary metabolites most abundant in C. majus L. are alkaloids, more than 20 of which are chemically identified. The yellow-orange alkaloid-containing latex is present throughout the entire plant. The most important alkaloids (Fig. 1) are the benzophenanthridine alkaloids (sanguinarine, chelerythrine and chelidonine) and protoberberines (berberine, coptisine). But also organic acids like chelidonic acid (Shen et al., 2001), citric acid, malic acid, or succinic acid were isolated from C. majus L. (Slavik, 1955), as well as saponins, choline and histamine (Kwasniewski, 1955), ferulic acid, caffeic acid and p-coumaric acid (Hahn and Nahrstedt, 1993) are present.

In the European Pharmacopoeia monograph, the aerial parts of C. majus L. are used. On Chelidonii herba there is a minimum content of 0.6% of alkaloids specified, calculated as chelidonine. For the investigations performed in the current study four extracts with solvents of different polarities were used. The complex mixture of phytochemical constituents was characterized by [sup.1]H-NMR fingerprint analysis, which is a comprehensive approach in the characterization of the plant metabolome (Daniel et al., 2008). The content of the alkaloids chelidonine, protopine, coptisine, berberine, sanguinarine and chelerythrine was quantified by RPHPLC-DAD. The chemical profile was correlated to effects on liver cells. Therefore, liver cell proliferation was investigated in response to the different extracts and effects were further characterized by a transcriptomics-based approach. Alterations in gene expression were monitored by microarrays, permitting the assessment of complete gene expression profiles induced by different compounds or extracts. By systems biological data evaluation it was possible to place our data into a biological context.

Materials and methods

Extraction of Chelidonii herba

The dried aerial parts, purchased from a local pharmacy were used. The herbal substance was complying with the monograph in Ph. Eur. 7.5/1861. Chelidonii herba was powdered and extracted with four different extraction solvents (ethanol, ethanol 50% (V/V), dichloromethane and water). Respectively, 1.0 g of plant material was extracted with 10 ml of the respective solvent, frequently stirring at room temperature for 10 min. After filtering, the extraction procedure was repeated. The solvents were eliminated under reduced pressure and the dried extract was resolved in 500 [micro]l deuterated dimethyl sulphoxide. Each extract was performed in four replicates. The sample material Chelidonii herba in analogy to voucher specimens was deposited at the Institute of Pharmaceutical Biology, University of Bonn.

Quantification of alkaloids in Chelidonii herba extracts with high performance liquid chromatography (HPLC)

Liquid/liquid extraction cartridges (Chem Elut CE 1010) were used for sample preparation. For separation of alkaloids a Synergy 4u Polar-RP (80 [Angstrom], 250 x 4.6 [mm.sup.2] column, Phenomenex) was used with a cartridge (SequrityGuard Cartidges, Polar RP 4 x 3.0 [mm.sup.2] ID, column, Phenomenex). Alkaloids were detected by DAD-UV at 285 nm. Two different mobile phases were used: solvent A, 90% [H.sub.2]O adjusted to a pH of 2.5 with phosphoric acid/10% methanol (V/V) and solvent B, 30% [H.sub.2]O adjusted to a pH of 2.5 with phosphoric acid/70% methanol (V/V). Alkaloids were eluted using a linear gradient from 30% B to 60% B for the first 6 min, followed by a linear gradient from 60% B to 100% B in 23 min. Subsequently, the column was washed for 6 min with 100% B and re-equilibrated to the starting conditions first by a linear gradient from 100% B to 30% B in 3 min and then an isocratic gradient (30% B) for 10 min. The solvent flow rate was adjusted to 1.0ml/min. Spectral data for all peaks were recorded in the range of 200-600 nm. The alkaloids were identified by the comparison of retention times and UV-spectra to the respective reference substances. Protopine was shown to elute first (17.3 min), followed by chelidonine (19.6 min) and coptisine (23.2 min). The peak obtained at 26.1 min was identified as berberine, subsequently sanguinarine (29.9 min) and chelerythrine (32.3 min) eluted.

[sup.1]H-Nuclear magnetic resonance (NMR) spectroscopy and PCA

[sup.1]H-NMR spectra of extracts were recorded with a Bruker advance 300 DPX instrument (temperature 25 [degrees]C, 64 scans). Calibration of spectra was performed according to signals of incomplete deuterated solvents. Data were processed using TOPSPIN software. Data evaluation by PCA was performed with TopSpin[TM] AMIX software (Bruker). Data from 0 to 10 ppm were included. The region from 2.2 to 5.5 ppm was excluded to eliminate the effects of signals from water and the solvent peak of DMSO (2.54 ppm) in the extract. The bucket width was 0.05 ppm and data were integrated according to the sum of intensities. Four replicates of each extract were recorded by [sup.1]H-NMR.

Preparation of total RNA

Total RNA was isolated from HepG2 cells using RNeasy Mini Kit (Qiagen, Germany) following the manufacturer's instructions.

Whole genome microarray

The quality of RNA was examined by the RIN-values (Agilent 2100 Bioanalyzer). The concentration and purity of RNA was measured using an UV-visible spectrometer (NanoDrop 1000; Thermo Fisher Scientific, Waltham, MA) by absorption at wavelengths of 260 and 280 nm. RNA samples with a 260/280 nm absorption ratio >1.8 and RIN >9.5 were used in subsequent microarray analysis. For microarray profiles, fluorescence labeled cRNA samples were prepared from 100 ng RNA using reverse transcriptase. The amplification reaction with simultaneous introduction of Cy3-dCTP to the amplified complementary RNA (cRNA) was performed using a Quick Amp Labeling Kit for One-Color (Agilent Technologies). The concentration of the purified samples and Cy3 dye incorporation efficiency was evaluated using a NanoDrop 1000 spectrophotometer. After fragmentation (60 [degrees]C, 30 min), single colour cRNA samples were hybridized on a DNA chip (Whole Genome Gene Expression Microarrays v2, Agilent Technologies, 8x60K, 27958 Gene RNAs, 7419 lincRNAs) at 65[degrees]C for 17h in a hybridization oven (Agilent Technologies). Four independent experiments were performed under each experimental condition.


After reverse transcription by random priming the resulting cDNA was used for qRT-PCR. Following initial denaturation (95 [degrees]C, 10 min), amplification was performed over 45 cycles (Light Cycler 480, Roche) with denaturation at 95 [degrees]C for 10 s and annealing with primers at temperatures shown in Table 2. Elongation was performed at 72 [degrees]C for 20 s. The size of PCR fragments was analyzed by agarose gel electrophoresis. Gene expression was evaluated using Light Cycler 480 Software 1.5. Cp-values were normalized to GAPDH.

Data processing and statistics

Hybridized DNA chip slides were scanned using an Agilent Scanner (Agilent Technologies) with Feature Extraction Software (Agilent Technologies). Fluorescence intensity data were imported to GeneSpring GX version 12.5 (Agilent Technologies) with the quantile scaling normalization. Data were filtered to exclude low-quality data to guarantee the accuracy of the statistical analysis. In the first step, spots with lower intensities than the threshold, which was determined based on the intensities of the Agilent RNA Spike-Mix, were filtered to exclude intensity data weaker than background noise. In the second step, spots with saturated intensities and near-background intensities were filtered using the 'flag' function of the Feature Extraction Software. In the third step, spots with large variance among the four repeated experiments (coefficient value >50%) were filtered. The final filtration step was conducted based on the fold increase. Statistical analysis of the genes remaining after the four filtration steps was performed with a t-test corrected using the Benjamini and Hochberg false discovery rate. Genes were considered as up- or down-regulated with a fold [greater than or equal to] 2 [less than or equal to] 2 and p<0.05.

These processed data were used for further analysis by Ingenuity Systems Inc. Reedwood City, USA (Qiagen). Ingenuity data base was used as reference set. In the networks interaction 70 molecules were set and all data sources were integrated in data analysis. Data were compared to the human reference data (tissues and primary cells not otherwise specified, cells not specified and other cells; organ systems liver and other cells not otherwise specified).

Treatment of HepG2 cells with plant extracts

HepG2 cells were purchased from the Leibniz institute DSMZ-German Collection of Microorganisms and Cell Cultures GmbH and were used between passages 3 and 12. The cells were maintained in T75 flasks in RPMI1640 medium supplemented with 10% foetal bovine serum, 1% penicillin/streptomycin solution (100U/ml) and 1% L-glutamine. The cells were constantly incubated in humidified atmosphere at 5% C[O.sub.2] and 37 [degrees]C.

When the cells were about 80% confluent they were treated with trypsin and harvested by centrifugation. Cells were counted and an equal number (1.0 x [10.sup.6] cells/5ml media) was transferred to each experimental plate (5 cm [empty set]) for treatment with the different extracts on the following day. Cells were treated with the extracts in a concentration of 1:501. The vehicle treated cells (1% DMSO) served as a control group. Four replicates were performed for each experimental condition.

Cell growth and proliferation assay using xCELLigence system

About 7500 HepG2 cells were seeded to each well. After 24 h, medium containing the extracts in a concentration of 1:501 adjusted to 1% DMSO or medium containing 1% DMSO as solvent control was added, respectively. The cell index was monitored every 15 min.


NMR fingerprint analysis

The spectra of the ethanolic, ethanolic 50% (V/V), dichloromethane and aqueous Chelidonii herba extracts were recorded by [sup.1]H-NMR in order to characterize the groups of constituents of Chelidonii herba. This method is a comprehensive approach, where the complexity of extracts can be shown, in contrast to methods, where just marker substances are analyzed. The data of the1 H-NMR spectra were statistically evaluated by PCA (Fig. 2). Discrimination between the four different extracts was achieved. Replicates of the different extracts were clustering with regards to their extraction solvent. The distribution of clusters demonstrated that the ethanolic and the dichloromethane extract as well as the aqueous and the ethanolic 50% (V/V) extract were similar to each other, respectively.

To identify classes of compounds responsible for the distribution, the loading plot (Fig. 3) was evaluated. The range between 2.2 and 5.5 ppm was excluded, because of water residues in the extracts. The main differences between the extracts were found in the region from 0 to 2.2 ppm, where most of the protons of terpenoids, steroids and organic acids resonate. In the region between 5.25 and 7.25 ppm protons attached to phenylpropanoids, flavonoids, phenolics and tannins cause resonance signals and in this region differences between the extracts can be observed. [sup.1]H-NMR resonances of alkaloids like chelidonine, protopine, coptisine, sanguinarine, chelerythrine and berberine are expected in the range of 7.5-10 ppm. As the loading plot shows, discrimination of PCA did not result from the different alkaloid content of the extracts.

Quantification of Chelidonii herba alkaloids by RP-HPLC

A more detailed characterization of the different Chelidonii herba extracts was done by quantifying chelidonine, protopine, sanguinarine, coptisine, chelerythrine and berberine of the Chelidonii herba extracts by RP-HPLC (Fig. 4). The dichloromethane extract showed the highest total alkaloid content (8.11 mg/ml extract; 4.06 mg/g dry weight) (Table 1). The content of the alkaloid chelidonine (7.15 mg/ml extract) was higher compared to the other Chelidonii herba extracts. The total alkaloid content of the ethanolic extract (5.93 mg/ml extract; 2.96 mg/g dry weight) and the ethanolic 50% (V/V) extract (4.74 mg/ml extract; 2.37 mg/g dry weight) was higher compared to the aqueous extract (0.94 mg/ml extract; 0.47 mg/g dry weight).

HepC2 cell proliferation

Effects of different Chelidonii herba extracts on liver cells were investigated. Each extract was applied to the cells in a dilution of 1:501 in DMSO, respectively. The dichloromethane and ethanolic extracts exhibited the highest content of alkaloids applied to HepG2 cells (16.2 p-g/ml and 11.8 [micro]g/ml, respectively). The alkaloid content of the 50% (V/V) extract and the aqueous extract applied to HepG2 cells was lower (9.5 [micro]g/ml and 1.9|xg/ml, respectively).

The aqueous extract did not exhibit antiproliferative effects on HepG2 cells compared to the solvent control. After 6h of exposure to the aqueous extract 105% of the HepG2 cells were viable in relation to DMSO control (Fig. 5). The strongest effects on proliferation of liver cells were observed for the ethanolic extracts (ethanol and ethanol 50% (V/V)) of Chelidonii herba. After 6h of exposure to the ethanolic extract 58% and 73% of the HepG2 cells were viable, respectively. Less cytotoxic effects were observed for the dichloromethane extract. 77% of the HepG2 cells were viable after 6 h of exposure to the dichloromethane extract.

Whole genome microarrays

In order to explore cellular effects of the different Chelidonii herba extracts at the genetic level, expression profiling was carried out in HepG2 cells. The initial data analysis revealed that each treatment with respect to the type of extract showed a distinct expression profile. In total, 2861 genes were regulated due to the treatment with the dichloromethane Chelidonii herba extract, the regulation of 1871 genes was different from that of the control after exposure to the aqueous Chelidonii herba extract, 1850 genes were differently modulated when the cells were exposed to the ethanolic extract, and the expression of 695 genes was altered as a result of the treatment with the ethanolic 50% (V/V) extract.

Most significant regulations were observed for pathways related to biotransformation (Fig. 6). The signal pathway "Cytochrome P450 Panel-Substrate is a Xenobiotic" was significantly regulated by all four treatments, is. All other signal pathways related to biotransformation were activated at least by one extract or by two extracts. Genes involved in biotransformation that were regulated significantly in at least one condition are shown in Table 3. Exposure of cells to the dichloromethane and the ethanolic extracts led to significant regulation of 15 and 16 genes relevant to biotransformation, respectively. Relating to biotransformation, 12 differently regulated genes were found for the ethanolic 50% (V/V) extract, and the regulation of 11 genes was altered following exposure to the aqueous extract.

Data evaluation with IPA showed that processes related to hepatotoxicity were significantly activated (Fig. 6). Signal cascades involved in oxidative stress, fatty acid metabolism or liver damage were induced. An increase in the permeability transition or the transmembrane potential of mitochondria and mitochondrial membranes was activated significantly for at least one condition in HepG2 cells. Following the exposure to the ethanolic extract seven signal cascades were activated. Respectively six and two signal cascades were activated for the ethanolic 50% (V/V) and the dichloromethane extract. No signal cascades associated to toxicity were induced after exposure to the aqueous extract in HepG2 cells.


Data obtained by microarray were verified by qRT-PCR for selected genes (Fig. 7). Results of qRT-PCR displayed minor differences in fold changes, but overall confirmed the results obtained by microarray analysis.


The quality and safety of herbal substances are important aspects in drug approval. Clinical double-blinded studies are rare in the field of herbal medicinal products, compared to studies on approved chemically-defined substances. Technical development and the avoidance of animal experiments are driving forces to investigate the applicability of new methods in the characterization of plant extracts, hence, to get a first insight for their potential prospects in regulatory affairs of herbal substances/herbal preparations.

In the present study, the1H-NMR fingerprint of Chelidonii herba extracted with solvents of different polarities was analyzed and the alkaloid content was determined. Cell proliferation of HepG2 liver cells was investigated, as it is considered that Chelidonii herba may have toxic effects on the liver, as several reports in literature described (De Smet, 2002; Benninger et al., 1999; Stickel et al., 2003; Strahl et al., 1998).

Plant extracts are always multicomponent mixtures. [sup.1]H-NMR fingerprints of Chelidonii herba extracts showed that the latter differ in their composition regarding the groups of constituents. According to the [sup.1]H-NMR spectra and the loading plot (Fig. 3), discrimination of the different extracts resulted from plant constituents associated with a big range of metabolites suggested to include terpenoids, steroids, organic acids, phenylpropanoids, flavonoids, phenolics and tannins. Resonance signals corresponding to the above-named plant constituents were obtained in the 'H-NMR spectra of the dichloromethane and the ethanolic Chelidonii herba extracts.

For both ethanolic extracts the alkaloid content quantified by RP-HPLC (Table 1) differed only by about 1 mg/ml extract (ethanolic extract 5.93 mg/ml extract; 50% (V/V) ethanolic extract 4.74 mg/ml). The dichloromethane extract showed the highest content of the total alkaloids (8.11 mg/ml). Only a minor amount of alkaloids was extracted with water (0.94 mg/ml). As investigations of cell proliferation demonstrated, each extract of Chelidonii herba displayed different effects. Although the highest amount of total alkaloids was found in the dichloromethane extract, the strongest inhibition of proliferation of HepG2 cells was observed for the treatment with the ethanolic extract, suggesting that not exclusively alkaloids cause toxicity. This is also supported by the fact that relative proliferation rates of cells treated with the dichloromethane and the ethanolic 50% (V/V) extract were similar (77% and 73%, respectively), though the alkaloid content in the dichloromethane extract was nearly twice as high as in the ethanolic 50% (V/V) extract. Earlier studies reported in literature analyzed cytotoxic effects of the different alkaloids in HepG2 cells. [IC.sub.50] values of chelidonine were 34.50 [micro]M (12.19 [micro]g/ml) after 24 h (El-Readi et al., 2013), the corresponding IC50 values of berberine and coptisine were 129.56 p.M (48.17 [micro]g/ml) and 202.33 [micro]M (64.81 [micro]g/ml), respectively (Yi et al., 2013). Sanguinarine and chelerythrine showed [IC.sub.50] values between 5 and 10 [micro]M (24 h) (Zdarilova et al., 2006), while protopine did not show a significant change in relation to the control at a concentration of 75 [micro]M after 24 h (Vrba et al., 2011). Regarding the composition of investigated alkaloids, the dichloromethane extract showed the highest amount of chelidonine in relation to the other investigated Chelidonii herba extracts. Even with regard to the [IC.sub.50] values of the quantified alkaloids, the phytochemical characteristics indicate that the alkaloid content is not exclusively responsible for the antiproliferative effect of the extracts.

Microarray data evaluation of the present study showed that each extract caused a distinct expression pattern. For Equisetum arvense derived from different origins, discriminatory results in hierarchical cluster analysis of gene expression profiling in Saccharomyces cerevisiae were observed (Cook et al., 2013). Discriminatory expression profiles in blood samples and subsequent different mode of actions were also observed in rats after their treatment with different fractions of a Salix spec, extract (Ulrich-Merzenich et al., 2012). Moreover, methanolic preparations containing Anacardium occidental in different concentrations exposed to HepG2 cells showed discriminable expression profiles (Khaleghi et al., 2011). This indicates the specificity of transcriptome investigations.

By further data evaluation obtained with the different Chelidonii herba extracts biological processes associated with liver toxicities and biotransformation were analyzed. Most significant activated signal cascades were identified for biotransformation processes. HepG2 cells exposed to the dichloromethane (15 genes) or the ethanolic extract (16 genes) showed the highest number of regulated genes related to biotransformation processes. For the ethanolic 50% (V/V) extract 13 genes and for the aqueous extract 11 genes were induced, respectively. It is noticeable that the higher content of alkaloids was, but also the more plant constituents were extracted as suggested by 'H-NMR fingerprint, the more genes were regulated in microarray experiments that play a role in biotransformation.

As microarray results showed, mRNA levels of Cyp1A1 and Cyp1BI were induced in HepG2 cells. Protopine was shown to induce expression of mRNA of Cyp1A1 in HepG2 cells without changing the Cyp1A1 protein or activity (Vrba et al., 2011). But Cyp1A1 is inducible by various herbal substances, e.g. by the flavonoids quercetin and kaempferol (Ciolino et al., 1999), catechins like epigaIlocatechin-3-gallate (Palermo et al., 2005) or alkaloids like berberine (Vrzal et al., 2005; Zhou et al., 2011; Chatuphonprasert et al., 2011; Salminen et al., 2011; Lo et al., 2013). Cyp3A4 has a share of 30% of the relative abundance of cytochromes in the liver and plays a dominant role in drug metabolism (Rendic and Di Carlo, 1997). The exposure of HepG2 cells to all four Chelidonii herba extracts induced Cyp3A4 gene expression, which was shown earlier to be potentially mediated by alkaloids (Zhou et al., 2011; Salminen et al., 2011). Previously, gene expression profiles of Caco-2 cells exposed to chelidonine (20 [micro]M) and a methanolic extract containing C. majus L. (20 [micro]g/ml alkaloids) were investigated (El-Readi et al., 2013). Microarray data evaluation indicated the induction of various genes related to xenobiotic metabolism in Caco-2 cells including Cyp3A4 and GST (El-Readi et al., 2013).

Further data evaluation with IPA linked to liver toxic effects (Fig. 6) confirmed liver proliferation results. Most significantly regulated signal cascades and as cell proliferation showed most antiproliferative effects were observed for the ethanolic extract. According to microarray results, oxidative stress plays a role in hepatotoxicity. Oxidative stress can damage proteins, DNA and lipids in cells. Nrf2 is a transcription factor which is able to transactivate detoxifying enzymes and antioxidant enzymes (Itoh et al., 1997).

The most significantly activated cascade ("Increase in the Permeability Transition of Mitochondria and Mitochondrial Membranes") indicated that mitochondria were depolarized. This is a key event in necrotic and apoptotic cell death (Lemasters et al., 1998). The signal cascade "Increase in the Transmembrane Potential of Mitochondria and Mitochondrial Membranes" was observed for the exposure to the ethanolic extract. The transmembrane potential of mitochondria is essential for the ATP production of the cell. Mitochondrial dysfunction may lead to alterations in the transmembrane potential, subsequently this may lead to apoptosis.

Both ethanolic extracts of Chelidonii herba altered the fatty acid metabolism in HepG2 cells. The liver is involved in important aspects of lipid metabolism (fatty acid [beta]-oxidation, lipogenesis and lipoprotein uptake and secretion). Hepatic stellate cell (HSC) activation was significantly activated by the ethanolic extract. HSCs may be activated by hepatotoxins, initiating a cascade of proinflammatory events. Continued liver injury may lead to tissue fibrosis and liver cirrhosis. Significant results for "Hepatic Fibrosis" were observed for the dichloromethane extract. "Liver Proliferation" and "Increased Liver Damage" were activated after exposure to the ethanolic 50% (V/V) extract of Chelidonii herba.

The data evaluated with 1PA indicate that extracts of Chelidonii herba induce also nephrotoxic cascades. The signal cascade "Renal Glomerulus Panel" was activated for the ethanolic and dichloromethane extract of Chelidonii herba. The signalling cascade "Renal necrosis/Cell Death" was significantly regulated for the ethanolic and ethanolic 50% (V/V) extract of Chelidonii herba in HepG2 cells.

Overall, several signal cascades attributable to hepatotoxicity are regulated. But the relevance of in-vitro results obtained with cell culture models for human is difficult to estimate and still under debate. In previous studies, HepG2 cells indicated to be useful in the study of biotransformation properties (Wilkening et al., 2003). Nevertheless, the relevance of data obtained by microarray experiments should be further proven on the protein or cellular level.

Herbal preparations are defined according to the manufacturing process. Extracts obtained with the different solvents showed a different phytochemical composition and subsequently different biological activities. This emphasizes the importance of defined manufacturing processes in herbal drug preparation.

In conclusion, the results demonstrate that a combination of the metabolomic characterization and the analysis of the transcriptome by microarrays is an appropriate strategy to characterize plant extracts. The approach provides complementary information, which often may be more representative for the characterization of a complex mixture than analysis of single compound or groups of constituents. The analysis of the microarray experiments with bioinformatic tools can be differentially attributed to specific extracts. For safety assessment of extracts of Chelidonii herba, it is particularly interesting that signal cascades associated with biotransformation or toxicity have been identified. However, the current set of available data is limited and not permitting to draw any regulatory conclusion. Further studies are necessary to evaluate the suitability of different cell culture models and their relation to results obtained by liver proliferation studies.

Abbreviations: IPA, Ingenuity pathway analysis: PCA. Principal component analysis; qRT-PCR, quantitative real-time PCR.

Abbreviations: AhRR, Aryl-hydrocarbon receptor repressor; AKR1B10, Aldo-keto reductase family 1 member B10; AKR1C4, Aldo-keto reductase family 1 member C4; BAAT, Bile Acid-CoA:Amino Acid N-Acyltransferase; CHST10, Carbohydrate sulfotransferase 10; CHST14, Carbohydrate sulfotransferase 14; Cypl A1, Cytochrome P450 family 1 member Al; CYP19A1, Cytochrome P450 family 19 member Al; Cypl B1, Cytochrome P450 family 1 member Bl; CYP24A1, Cytochrome P450 family 24 member A1 (1,25dihydroxyvitamin D3 24-hydroxylase); CYP26A1, Cytochrome P450 family 26 member Al; CYP27C1, Cytochrome P450 family 27 member Cl; CYP2A7, Cytochrome P450 family 2 member A7; CYP2B6, Cytochrome P450 family 2 member B6; CYP2A7, Cytochrome P450 family 2 member A7; CYP2C18, Cytochrome P450 family 2 member C18; CYP3A4, Cytochrome P450 family 3 member A4; CYP3A7, Cytochrome P450 family 3 member A7; CYP4V2, Cytochrome P450 family 4 member V2; CYP8B1, Cytochrome P450 family 8 member Bl; EGR2, Early growth response protein 2; GLYATL1, Glycine-N-Acyltransferase-Like 1; GST02, Glutathione S-transferase omega 2; GSTP1, Glutathione S-transferase P; HMOX1, heme oxygenase (decycling) 1; HS3ST3A1, Heparan sulfate glucosamine 3-O-sulfotransferase 3A1; HTR3A, 5-hydroxytryptamine receptor 3A; IGFBP1, Insulin-like growth factor-binding protein 1; KCNAB3, Voltage-gated potassium channel subunit beta-3; SULT1C3, Sulfotransferase 1C3; SULT1E1, Sulfotransferase 1C3; SULT4A1, Sulfotransferase 1C3; UGT2A3, Uridine 5'-diphospho-glucuronosyltransferase 2A3; UGT2B11, Uridine 5'- diphospho-glucuronosyltransferase 2B11.

Conflict of interest

All authors declare they have no actual or potential competing financial interest.

The manuscript is reflecting the views of the authors on the results of a research project, the content shall not be quoted as representing the regulatory position of the Federal Institute for Drugs and Medical Devices (BfArM).


Authors are thankful to the Federal Institute for Drugs and Medical Devices (Bonn, Germany) (Grant nos. V-11788 and V-15132) for financial support to undertake these studies.


Benninger, J., Schneider, H.T., Schuppan, D., Kirchner, T., Hahn, E.G., 1999. Acute hepatitis induced by greater celandine (Chelidonium majus). Gastroenterology 117, 1234-1237.

Chatuphonprasert, W., Sangkawat, T., Nemoto, N., Jarukamjorn, K., 2011. Suppression of beta-naphthoflavone induced CYP1A expression and lipid-peroxidation by berberine. Fitoterapia 82, 889-895.

Ciolino, H.P., Daschner, P.J., Yeh, G.C., 1999. 15-6-Dietary flavonols quercetin and kaempferol are ligands of the aryl hydrocarbon receptor that affect CYP1A1 transcription differentially. Biochem. J. 340, 715-722.

Cook, R., Hennell, J.R., Lee, S., Khoo, C.S., Carles, M.C., Higgins, V.J., Govindaraghavan, S., Sucher, N.J., 2013. The Saccharomyces cerevisiae transcriptome as a mirror of phytochemical variation in complex extracts of Equisetum arvense from America, China, Europe and India. BMC Genomics 14, 445-463.

Daniel, C., Kersten, T., Kehraus, S., Konig, G.M., KnoK, W., 2008. Identifizierung und Charakterisierung von Arzneipflanzen mit "Metabolic fingerprinting". Z. Phytother. 29, 270-274.

De Smet, P.A., 2002. Sefety concerns about cava not unique. Lancet 360, 1336.

El-Readi, M.Z., Eid, S., Ashour. M.U Tahrani, A., Wink, M., 2013. Modulation of multidrug resistance in cancer cells by chelidonine and Chelidonium majus alkaloids. Phytomedicine 20, 282-294.

Gilca, M., Gaman, L, Panait, E., Stoian, L, Atanasiu, V., 2010. Chelidonium majus--an integrative review: traditional knowledge versus modern findings. Res. Complement. Med. 17, 241-248.

Hahn, R., Nahrstedt, A., 1993. High content of hydroxycinnamic acids esterified with (+)-D-malic acid in the upper parts of Fumaria officinalis!. Planta Med. 59, 189-190.

Hamburg, M.A., 2011. Advancing regulatory science. Science 331, 987.

Hartung, T., McBride, M., 2011. Food for thought. On mapping the human toxome. ALTEX 28, 83-93.

Itoh, K., Chiba, T., Takahashi, S., Ishii, T., Igarashi, K., Katoh, Y., Oyake, T., Hayashi, N., Satoh, K., Hatayama, L, Yamamoto, M., Nabeshima, Y., 1997. An Nrf2/small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through antioxidant response elements. Biochem. Biophys. Res. Commun. 236, 313-322.

Kersten, T., Daniel, C, Konig, G.M., Knoss, W., 2008. Das Potential PCR-basierter Markermethoden zur Identifizierung von Arzneipflanzen. Z. Phytother. 29, 122-128.

Khaleghi, S., Aziz, A., Razali, N., Junit, S., 2011. Microarray analysis revealed different gene expression patterns in HepG2 cells treated with low and high concentrations of the extracts of Anacardium occidentale shoots. Genes Nutr. 4, 413-427.

Krewski, D., Acosta Jr., D., Andersen, M., Anderson, H., Bailar III, J.C., Boekelheide, K., Brent, R., Charnley, G., Cheung. V.G., Green Jr., S., Kelsey, K.T., Kerkvliet, N.I., Li, A.A., McCray, L, Meyer, O., Patterson, R.D., Pennie, W., Scala, R.A., Solomon, G.M., Stephens, M., Yager, J., Zeise, L, 2010. Toxicity testing in the 21st century: a vision and a strategy. J. Toxicol. Environ. Health B Crit. Rev. 13, 51-138.

Kwasniewski, V., 1955. Possible analytical determination of chelldonic acid in vegetable material. Arzneimittelforschung 5, 90-92.

Lemasters, J.J., Nieminen, A.L, Qian, T., Trost, L.C., Elmore, S.P., Nishimura, Y., Crowe, R.A., Cascio. W.E., Bradham, C.A., Brenner, D.A., Herman, B., 1998. The mitochondrial permeability transition in cell death: a common mechanism in necrosis, apoptosis and autophagy. Biochim. Biophys. Acta 1366, 177-196.

Lo, S.N., Chang, Y.P., Tsai, K.C., Chang, C.Y., Wu, T.S., Ueng, Y.F., 2013. Inhibition of CYP1 by berberine, palmatine, and jatrorrhizine: selectivity, kinetic characterization, and molecular modeling. Toxicol. Appl. Pharmacol. 272, 671-680.

Palermo, C.M., Westlake, C.A., Casiewicz, T.A., 2005. Epigallocatechin gallate inhibits aryl hydrocarbon receptor gene transcription through an indirect mechanism involving binding to a 90kDa heat shock protein. Biochemistry 44, 5041-5052.

Rendic, S., Di Carlo, F.J., 1997. Human cytochrome P450 enzymes: a status report summarizing their reactions, substrates, inducers, and inhibitors. Drug Metab. Rev. 29, 413-580.

Salminen, K.A., Meyer, A., Imming, P.. Raunio, H., 2011. CYP2C19 progress curve analysis and mechanism-based inactivation by three methylenedioxyphenyl compounds. Drug Metab. Dispos. 39, 2283-2289.

Shen, Z.-W., Fisinger, LL, Poulev, A., Eisenreich, W., Werner, L, Pleiner, E., Bacher, A., Zenk, M.H., 2001. Tracer studies with 13C-labeled carbohydrates in cultured plant cells. Retrobiosynthetic analysis of chelidonic acid biosynthesis. Phytochemistry 57, 33-42.

Slavik, J., 1955. Papaveraceae alkaloids, I. Chelidonium maius L substances. Cesk. Farm. 4, 15-17.

Stickel, F., Poschl, G., Seitz, H.K., Waldherr, R., Hahn, E.G., Schuppan, D., 2003. Acute hepatitis induced by Greater Celandine (Chelidonium majus). Scand. J. Gastroenterol. 38, 565-568.

Strahl, S., Ehret, V., Dahm, H.H., Maier, K.P., 1998. Necrotizing hepatitis after taking herbal remedies. Dtsch. Med. Wochenschr. 123, 1410-1414.

Teschke, R., Frenzel, C, Glass, X., Schulze, J., Eickhoff, A., 2012. Greater Celandine hepatotoxicity: a clinical review. Ann. Hepatol. 11, 838-848.

Ulrich-Merzenich, G., Kelber, O., Koptina, A., Freischmidt, A., Heilmann, J., Muller, J., Zeitler, H., Seidel, M.F., Ludwig, M., Heinrich, E.U., Winterhoff, H., 2012. Novel neurological and immunological targets for salicylate-based phytopharmaceuticals and for the anti-depressant imipramine. Phytomedicine 19, 930-939.

Vrba, J., Vrublova, E., Modriansky, M., Ulrichova, J., 2011. Protopine and allocryptopine increase mRNA levels of cytochromes P450 1A in human hepatocytes and HepG2 cells independently of AhR. Toxicol. Lett. 203, 135-141.

Vrzal, R., Zdarilova, A., Ulrichova, J., Blaha, L, Giesy, J.P., Dvorak, Z., 2005. Activation of the aryl hydrocarbon receptor by berberine in HepG2 and H4IIEcells: biphasic effect on CYP1A1. Biochem. Pharmacol. 70, 925-936.

Wagner, H., 2006. Multitarget therapy-the future of treatment for more than just functional dyspepsia. Phytomedicine 13, 122-129.

Wilkening, S., Stahl, F., Bader, A., 2003. Comparison of primary human hepatocytes and hepatoma cell line HepG2 with regard to their biotransformation properties. Drug Metab. Dispos. 31, 1035-1042.

Yi, J., Ye, X., Wang, D., He, K., Yang, Y., Liu, X., Li, X., 2013. Safety evaluation of main alkaloids from Rhizoma Coptidis. J. Ethnopharmacol. 145, 303-310.

Zdarilova, A., Vrzal, R., Rypka, M., Ulrichova, J., Dvorak. Z., 2006. Investigation of sanguinarine and chelerythrine effects on CYP1A1 expression and activity in human hepatoma cells. Food Chem. Toxicol. 44, 242-249.

Zhou, C, Li, X.C., Fang, W.H., Yang, X.L, Hu, LL, Zhou, S., Zhou, J.F.. 2011. Inhibition of CYP450 1A and 3 A by berberine in crucian carp Carassius auratus gibelio. Comp. Biochem. Physiol C. Toxicol. Pharmacol 154, 360-366.


Article history:

Received 19 November 2013

Received in revised form 12 June 2014

Accepted 21 July 2014

A. Orland (a,*), K. Knapp (b), G.M. Konig (b), G. Ulrich-Merzenich (c), W. Knoss (a)

(a) Federal Institute for Drugs and Medical Devices, D-53175 Bonn, Germany

(b) Institute of Pharmaceutical Biology, Rheinische Friedrich-Wilhelms-Universitat Bonn, D-53115 Bonn, Germany

(c) University Clinic Centre Bonn, Medical Clinic III, Centre for Internal Medicine, D-53127 Bonn, Germany

* Corresponding author. Tel.: +49 228 99 307 4961: fax: +49 228 99 307 5395.

E-mail addresses:, (A. Orland).

Table 1
Alkaloid composition of different extracts of Chelidonii herba.
Detection of alkaloids was obtained by RP-HPLC-DAD (285 nm).

                                    Ethanolic extract    50% (V/V)

                                            [+ or -]           [+ or -]
                                               SD                 SD

Protopine (mg/ml)                   0.44      0.06     0.72      0.09
Chelidonine (mg/ml)                 3.67      1.40     2.16      0.16
Coptisine (mg/ml)                   1.32      0.40     1.43      0.24
Berberine (mg/ml)                   0.14      0.01     0.15      0.00
Sanguinarine (mg/ml)                0.33      0.04     0.25      0.05
Chelerythrine (mg/ml)               0.03      0.00     0.02      0.01
Total alkaloids (mg/ml)             5.93               4.74
Total alkaloids (mg/g dry weight)   2.96               2.37

                                     Dichloro-methane       Aqueous
                                         extract            extract

                                            [+ or -]           [+ or -]
                                               SD                 SD

Protopine (mg/ml)                   0.10      0.02     0.14      0.08
Chelidonine (mg/ml)                 7.15      2.36     0.31      0.17
Coptisine (mg/ml)                   0.40      0.07     0.17      0.11
Berberine (mg/ml)                   0.13      0.00     0.13      0.01
Sanguinarine (mg/ml)                0.31      0.07     0.18      0.02
Chelerythrine (mg/ml)               0.02      0.01     0.00      0.01
Total alkaloids (mg/ml)             8.11               0.94
Total alkaloids (mg/g dry weight)   4.06               0.47

Table 2
Primer used for qRT-PCR for validation of microarray results.

name          Accession code   Direction

HMOXI         NM_002133.2      Forward
HMOXI         NM_002133.2      Reverse
EGR2          NM_000399.3      Forward
EGR2          NM_000399.3      Reverse
IGFBP1        NM_000596.2      Forward
IGFBP1        NM_000596.2      Reverse
AhRR          NM_020731        Forward
AhRR          NM_020731        Reverse
              NM_000499.3      Forward
CyplA1        NM_000499.3      Reverse
CyplBl        NM_000104        Forward
CyplBl        NM_000104        Reverse
HTR3A         NM_213621.3      Forward
HTR3A         NM_213621.3      Reverse
GAPDH         NM_002046.4      Forward
GAPDH         NM_002046.4      Reverse

Primer/Cene                                    Annealing temperature
name          Primer sequence                  ([degree]C)

EGR2          5'-GCTGGCACCAGGGTACTGA-3'        64
AhRR          5'-CAGTTACCTCCGGGTGAAGA-3'       59
AhRR          5'-CCAGAGCAAAGCCATTAAGA-3'       59
              5'-AACCTTTGACAAGGGCCACA-3'       55
CyplA1        5'-GACCTGCCAATCACTGTGTC-3'       55
CyplBl        5'-CACTGCCAACACCTCTGTCTT-3'      53
CyplBl        5'-CAAGGAGCTCCATGGACTCT-3'       53

Table 3
Fold changes of genes related to biotransformation after exposure to
the different Chelidonii herba extracts on HepG2 cells. Fold changes
of genes, which are related to biotransformation were obtained from
microarray experiments by data evaluation (Genespring 12.5). Values
in bold indicate statistical significance by at least 2 fold and p
[less than or equal to] 0.05.

                 FC                         FC Ethanolic
           Dichloromethane   FC Ethanolic     extract      FC Aqueous
Name           extract         extract       50% (V/V)      extract

AKR1B10         1.5822          2.0252#        1.9929        1.2447
AKR1C4         -1.161           2.7486#        2.1811#      -1.2813
BAAT           -5.5898#        -2.4609#       -2.0001#      -2.994#
CHST10         -2.7535#        -1.5089        -1.2222       -1.8048
CHST14         -2.2993#        -1.4761        -1.072        -1.6347
CYP19A1         1.1568          2.2508#        2.4009#       1.1475
CYP1A1          1.5768          2.9473#        3.2499#       1.4064
CYP1B1          2.9442#         3.2531#        4.1636#       2.2513#
CYP24A1         2.31#           3.9019#        2.7159#      -1.244
CYP26A1         3.5888#         3.5292#        1.856         1.04
CYP27C1         1.1677         -1.0281         2.8781#       1.223
CYP2A7          1.4945          1.4648         1.5285        2.5271#
CYP2B6          1.5528          2.2208#        1.4477        1.6054
CYP2C18         1.1694          1.5978         3.0097#       1.4221
CYP3A4          5.4536#         6.4891#        5.4454#       3.2177#
CYP3A7         -2.6684#        -1.1983        -1.1357       -2.8465#
CYP4V2         -2.8117#         1.1533         1.1997       -1.654
CYP8B1         -1.7735         -3.4139#       -2.1441#       1.1904
GLYATL1        -3.0002#        -4.1781#       -2.4801#      -2.4961#
GST02          -1.7483         -1.5308        -1.3509       -2.1757#
GSTP1           3.8747#         2.8508#        2.8556#       2.5285#
HS3ST3A1       -1.0795         -1.0437         1.2999        2.3288#
KCNAB3         -2.3195#        -1.7664        -1.5293       -1.7743
SULT1C3        -1.7304         -2.0568#       -1.9581        1.1393
SULT1 El       -5.7103#        -2.9995#       -2.6536#      -3.2872#
SULT4A1         2.00#           1.9249         1.6726        1.5822
UGT2A3         -2.2794#        -1.4233        -1.1904       -1.9758
UCT2B11        -2.6812#        -2.4336#       -1.8786       -2.1003#

Note: Values in bold indicate statistical significance by at least 2
fold and p [less than or equal to] 0.05 is indicated with #.
COPYRIGHT 2014 Urban & Fischer Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Orland, A.; Knapp, K.; Konig, G.M.; Ulrich-Merzenich, G.; Knoss, W.
Publication:Phytomedicine: International Journal of Phytotherapy & Phytopharmacology
Date:Oct 15, 2014
Previous Article:The vasodilating effect of a Hintonia latiflora extract with antidiabetic action.
Next Article:Biotransformation of monoterpenoids and their antimicrobial activities.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters