Chemical Composition and Cytotoxic Activity of Extracts from Carpesium divaricatum: In Vitro- versus Field-Grown Plants

Carpesium divaricatum Sieb. & Zucc. is a plant species rich in terpenoids of anti-inflammatory and cytotoxic activity, especially germacranolides of potential medicinal value. The present study describes in vitro multiplication of C. divaricatum, analysis of active constituents in the multiple shoots, and assessment of cytotoxic activities of extracts prepared from in vitro- and field-grown plants. The plant extracts were evaluated for cytotoxicity using two melanoma cell lines (HTB140 and A375); human keratinocytes (HaCaT); two colon cancer cell lines (Caco2 and HT29); human hepatocellular carcinoma cells (HepG2); two lines of prostate cancer cells (DU145 and PC3) and prostate epithelial cells (PNT2). Chemical compositions of the assayed extracts were analyzed by HPLC/DAD, in reference to isolated compounds. Maximum of 4.07 ± 1.61 shoots regenerated from a nodal explant of C. divaricatum, cultivated in a liquid MS medium supplemented with thidiazuron (1 μM). In vitro grown shoots and plantlets of C. divaricatum accumulated terpenoids that are known as active constituents of the intact plant. Cytotoxic activity of the extracts prepared from the in vitro cultured plants was like that demonstrated by the extracts prepared from field-grown plants and seemed to be more selective than cytotoxicities of the individual germacranolides.


Introduction
Except for the widespread C. cernuum L. and C. abrotanoides L., the genus Carpesium L. (Asteraceae, Inuleae-Inulinae) comprises ca. 20 Asian species, including six endemics in China [1,2]. Plants of Carpesium spp. are rich in sesquiterpene lactones and monoterpenoid thymol derivatives of well documented anti-inflammatory and cytotoxic activity [3,4]. Germacranolides isolated from the plants were frequently studied in respect of their cytotoxic activity against human cancer cells in vitro [5][6][7][8], but only a few studies dealt with the molecular mechanism of action of the compounds [9][10][11]. Recently, a paper reporting on poor selectivity of the cytotoxic effect exerted by Carpesium germacranolides has been published [12]. This may raise a question about the safety of internal use of the plant and its preparations. Despite a long-lasting traditional use of C. divaricatum Sieb. & Zucc. plants as both medicines and food, correlations between the chemical composition of the plant extracts and their biological activity are poorly investigated.
Though plant tissue culture offers a good experimental system for studying biosynthesis of biologically active plant metabolites and mechanisms of its regulation, there are no data on the production of biologically active terpenoids by in vitro cultures of Carpesium. The only study on in vitro culture of Carpesium spp. [13] described an effect of light quality on growth and selected physiological parameters (photosynthetic pigments, antioxidant enzymes, radical scavenging activity) of C. triste Maxim. plants cultivated in vitro and did not contain any data on biologically active mono-and sesquiterpenoids. Multiple shoots of Data from seven consecutive passages (n = 61), mean ± SD; 2 Data from five consecutive passages (n = 27), mean ± SD; 3 Mean of four independent measurements ± SD. Results denoted with the same letter were not statistically different (p > 0.05; one-way ANOVA).
Cultures grown on the media solidified with agar grew slower (up to 0.13 g DW after 8 weeks). Moreover, though significant differences in the number of shoots regenerated from one explant were not observed, the cultures on solidified media tended to produce fewer shoots. There was no visible difference in the leaf color of shoots grown in different illumination regimes.
Chromatographic analysis of the extracts from in vitro cultured plants revealed the presence of 10-isobutyryloxy-8,9-epoxythymyl isobutyrate, the monoterpenoid formerly found in the root cultures of Inula spp. and multiple shoots of Telekia speciosa [16,17,23]. The compound (for the structure, see Figure 1) is a known metabolite of C. divaricatum [24].
To assess the contents of the identified terpenoid metabolites in the acetone extracts prepared from the plant material under study, the RP-HPLC-DAD method previously used for quantification of sesquiterpene lactones and 10-isobutyryloxy-8,9-epoxythymyl isobutyrate in Telekia speciosa [25] was applied (Table 2, Figure 2). The gradient profile was modified (see Section 4.4.2) to improve the resolution. The results of the analysis are shown in Table 2. In the field-grown plants of C. divaricatum, germacranolides were major terpenoid constituents of aerial parts, whereas roots accumulated mainly 10-isobutyryloxy-8,9-epoxythym yl isobutyrate (Figure 2c,e). Extracts from the leaves collected in 2020 and in 2022 differed in their composition (Figure 2c,d). Extracts from the freshly collected leaves demonstrated higher contents of the compounds C and E, whereas extracts from the leaves collected in 2020 contained compound F (absent from the leaves collected in 2022). The cause of the difference may be disputable, but the effect of the storage period on the quality of the raw material is worth further studies. The extracts from roots of C. divaricatum collected in 2020 did not contain 10-isobutyryloxy-8,9-epoxythymyl isobutyrate (data not shown), probably due to the degradation of the compound. To assess the contents of the identified terpenoid metabolites in the acetone extracts prepared from the plant material under study, the RP-HPLC-DAD method previously used for quantification of sesquiterpene lactones and 10-isobutyryloxy-8,9-epoxythymyl isobutyrate in Telekia speciosa [25] was applied (Table 2, Figure 2). The gradient profile was modified (see Section 4.4.2) to improve the resolution. The results of the analysis are shown in Table 2. Table 2. Contents of the identified sesquiterpene lactones (A-F) and thymol derivative (G) semiquantitatively estimated in the plant material from in vitro culture and in the organs of field-grown plants.

Plant Extract Contents of the Terpenoids (% DW) 1
0.016 ± 0.002 a 0.026 ± 0.002 a 0.036 ± 0.004 a 0.021 ± 0.003 a 0.027 ± 0.004 a 0.018 ± 0.003 a 0.064 ± 0.008 a 05TDZS 0 In the field-grown plants of C. divaricatum, germacranolides were major terpenoid constituents of aerial parts, whereas roots accumulated mainly 10-isobutyryloxy-8,9-epoxythymyl isobutyrate (Figure 2c,e). Extracts from the leaves collected in 2020 and in 2022 differed in their composition (Figure 2c,d). Extracts from the freshly collected leaves demonstrated higher contents of the compounds C and E, whereas extracts from the leaves collected in 2020 contained compound F (absent from the leaves collected in 2022). The cause of the difference may be disputable, but the effect of the storage period on the quality of the raw material is worth further studies. The extracts from roots of C. divaricatum collected in 2020 did not contain 10-isobutyryloxy-8,9-epoxythymyl isobutyrate (data not shown), probably due to the degradation of the compound.
C. divaricatum plantlets cultivated in vitro, in a liquid medium containing 0.5 μM TDZ (Figure 2a), accumulated lower amounts of germacranolides than the leaves of fieldgrown plants (Figure 2c,d) and the plantlets cultivated in vitro on solidified media supplemented with BA and NAA (Figure 2b), but contained a substantial amount of 10-isobutyryloxy-8,9-epoxythymyl isobutyrate. The plantlets grown on solidified media with addition of BA and NAA demonstrated a germacranolide accumulation pattern like that in leaves of the field-grown plants (Figure 2b-d).
Chromatographically analyzed extracts: 05TDZL, 20B01NS, CdN2020, CdN2022, and CdK2022 and the extracts used for cytotoxicity assays were prepared simultaneously, from the same batch of plant material and in the same manner. Chromatographically analyzed extracts: 05TDZL, 20B01NS, CdN2020, CdN2022, and CdK2022 and the extracts used for cytotoxicity assays were prepared simultaneously, from the same batch of plant material and in the same manner.

Cytotoxic Activity of Extracts from In Vitro-and Field-Grown C. divaricatum
The assayed extracts, irrespectively of their chemical composition, exerted a cytotoxic effect on the cell lines used in the experiment (see Table 3 (Figure 2c,d). Leaves of the field-grown plants contained no detectable amount of 10-isobutyryloxy-8,9-epoxythymyl isobutyrate. The compound was a major terpenoid constituent of freshly collected C. divaricatum roots. Despite the various  (Figure 2c,e), both root and leaf extract were active in the assay, although their activities towards the individual cell lines differed. The extract from plantlets cultivated on a solidified nutrient medium, supplemented with BA and NAA, was especially active against PC3 and Caco2 cells (IC 50 = 14.07 µg/mL and 14.62 µg/mL, respectively, after 48 h treatment). The extract contained both sesquiterpene lactones and the monoterpenoid thymol derivative (compound G).

Discussion
C. divaricatum is a species native to East Asia. Plants of the species cultivated in Central Europe due to late flowering fail to produce seeds. In vitro multiplication of shoots may provide plantlets for further cultivation in the open field or may be a good starting point for developing an in vitro culture system to produce biologically active plant metabolites. The multiple shoots cultivated in vitro may be also helpful in elucidation of the mechanisms implicated in the regulation of secondary metabolism of the plant.
Literature data on micropropagation of plants from the Inuleae tribe are sparse. The only study on in vitro culture of Carpesium [13] described changes in several physiological parameters of C. triste plantlets, induced by the color of light used to illuminate the cultures. The plantlets were grown on MS nutrient medium deprived of growth regulators and the data on shoot proliferation were not given. Nodal explants of Inula royleana DC., cultivated for 6 weeks on a solidified MS medium containing 5.0 µM kinetin and 0.1 µM NAA, regenerated 5.1 ± 1.9 shoots per explant [26]. Shoot tip explants of Inula germanica L., after 4 weeks of culture on MS medium supplemented with 4.44 µM BA and 0.54 µM NAA, produced from 3.4 ± 0.7 to 4.9 ± 0.9 shoots per explant. The number of regenerated shoots varied depending on the number of subcultures [27]. Similar multiplication rate (ca. 4 shoots/explant), after 4-week culture, was achieved for T. speciosa, a species closely related to C. divaricatum, using the same medium composition [16]. By extending the culture period to 8 weeks, the multiplication rate of T. speciosa had been increased over two times (up to 10.9 ± 3.1). The nutrient medium containing 4.44 µM BA and 0.54 µM NAA was less effective when applied to C. divaricatum shoot regeneration from nodal explants (multiplication rate after 8-week culture: 3.49 ± 1.22, see Table 1). Replacement of BA and NAA by TDZ (0.5 µM) did not increase the multiplication rate, but significantly increased the dry weight of plantlets regenerated from one explant (from 0.081 ± 0.017 to 0.131 ± 0.023). Liquid media supplemented with TDZ (0.5 or 1.0 µM) were more advantageous for both multiplication of shoots (although the effect was not statistically significant) and the dry weight increment. As the achieved multiplication rates were low, further optimization of the shoot multiplication procedure was needed.
In vitro cultivated shoots of the Asteraceae plants are usually a reliable source of those sesquiterpene lactones, which are accumulated by aerial parts of intact plants [28][29][30][31]. Multiple shoots and plantlets of C. divaricatum cultivated on solidified MS media containing BA and NAA and leaves of the field-grown plants accumulated comparable amounts of germacranolides ( Figure 2, Table 2). Differences in terpenoid content observed in the in vitro cultivated plantlets (Table 2) may be caused by either different media composition or by the difference in the light regime. Biosynthesis of a pharmacologically active sesquiterpene lactone, artemisinin, by Artemisia annua L. tissue cultures was proven to be regulated by light [32,33]. In plant tissue cultures, fast growth is often accompanied with limited secondary metabolite production. The fast-growing C. divaricatum shoots, cultivated in liquid medium containing TDZ, produced less sesquiterpene lactones (as % DW) than the shoots grown on a solidified medium of the same composition.
Several dozen of sesquiterpene lactones were isolated from C. divaricatum plants, and germacranolides were the most frequently found metabolites of the group. Due to the co-occurrence of many structurally related compounds in the plant extract, development of the analytical method for quantification of Carpesium terpenoids in the raw extract is difficult. For a semi-quantitative estimation of terpenoid contents in the assayed plant extracts, the analytical method formerly used for quantification of sesquiterpene lactones and 10-isobutyryloxy-8,9-epoxythymyl isobutyrate in T. speciosa extracts [25] was adapted. The peak resolution was affected by the composition of the plant material under study and, to some extent, was improved by dilution of the sample.
The cell lines used for cytotoxicity evaluation were grouped in three panels, namely prostate, skin and gastrointestinal panels. Each panel comprised two cancer cell lines, differing in their metastatic potential, accompanied by non-neoplastic cell line of similar origin. The use of a non-neoplastic cells in the experiments gives the information on the selectivity of the tested samples. In the case of the gastrointestinal panel, hepatoma cell line (HepG2) was used instead of the appropriate non-neoplastic cells. This cell line is widely used in hepatotoxicity studies, as it reveals several phenotypic characteristics and many functional properties of normal liver cells [34,35].
The most susceptible cell line among all used in the study was colon adenocarcinoma Caco2, while the other colon adenocarcinoma cells (HT29) were more resistant. IC 50 values estimated for the two cell lines were two-or threefold different. What is important, the HepG2 cells were almost unaffected, what indicates the selectivity of the extracts. As far as the prostate panel is concerned, the tested extracts were generally more active towards highly metastatic PC3 prostate carcinoma, compared to less metastatic DU145 prostate carcinoma. However, only low selectivity of the extracts was noted, as in most cases, non-neoplastic PNT2 prostate epithelial cells were also strongly affected. In the case of skin panel, the tested extracts were much more selective and revealed varied cytotoxic impact to two melanoma cell lines, while normal keratinocytes were affected to a lesser extent (IC 50 > 75.6 µg/mL). Interestingly, highly metastatic HTB140 cells were more susceptible to the tested extracts compared to the other melanoma cell line (A375) used in the study that had less metastatic potential.
Sesquiterpene lactones are regarded as the biologically active metabolites, responsible for cytotoxic and anti-inflammatory activity of Carpesium plants [3]. Cytotoxic activity of the germacranolides isolated from C. divaricatum towards the human cancer cells in vitro was confirmed in several assays [3,8,12,21,22,36,37]. The compounds are accumulated in aerial parts of the plant and the plant roots seem to be almost deprived of them (Figure 2c-e). The plantlets cultivated in vitro, in MS liquid medium (containing 0.5 µM TDZ), were also poor producers of sesquiterpene lactones (Figure 2a). However, as it was shown in the present study, cytotoxic activity of the plant extracts and the sesquiterpene lactone content are not closely linked. Interestingly, traditional medicinal systems used whole herb of C. divaricatum as a remedy for various ailments [3]. The results of RP-HPLC-DAD analysis (Figure 2e) showed that 10-isobutyryloxy-8,9-epoxy-thymyl isobutyrate is a major terpenoid constituent of the plant roots. The compound has been recently described as an inhibitor of aberrant proliferative signaling in melanoma cells [38] and antileishmanial agent [39] and is known as a secondary metabolite synthesized by numerous species from the Asteraceae [40].
Acetone extracts from both in vitro-and field-grown plants of C. divaricatum are complex mixtures of active constituents that demonstrate diverse profiles of activity. Detailed knowledge of biochemistry of the plant, stability of the active metabolites and activity of their degradation products, new data on molecular mechanisms of action of isolated compounds combined with a development of a dedicated analytical method may lead to the better understanding of traditional use of C. divaricatum as a medicine and to potential future applications.

General Experimental Procedures
1 NMR spectra were recorded in MeOD on a Bruker AVANCE III HD 400 (resonance frequency 400.17 MHz) spectrometer (Bruker Corp., Billerica, MA, USA). RP-HPLC separations were performed using an Agilent 1200 Series HPLC system (Agilent Technologies Inc., Santa Clara, CA, USA) equipped with a diode array detector. Analytical chromatographic separations were conducted on a Kinetex XB-C18 column (4.6 × 250 mm, 5 µm total particle size; Phenomenex Inc., Torrance, CA, USA). Semipreparative RP-HPLC was conducted on a Vertex Plus Eurospher II 100-5 C18 column (8 × 250 mm; Knauer GmbH, Berlin, Germany), with an isocratic elution, using methanol-water ( NAA (1-naphtaleneacetic acid) or (3) 4.44 µM BA and 0.54 NAA. Alternatively, the nodal explants were inoculated in liquid MS medium supplemented with either 0.5 µM or 1 µM of TDZ. All nutrient media used in the experiments contained 3% sucrose and their pH was adjusted to 5.8, before autoclaving. The explants were kept at 28 • C, those on solidified media were maintained under continuous illumination (40 µmol m −2 s −1 , cool white fluorescent tubes) and those in liquid media were cultivated under the 16/8 (light/dark) photoperiod (20 µmol m −2 s −1 , cool white fluorescent tubes) with shaking (100 r.p.m.), to induce multiplication of shoots. After 12 weeks of culture, secondary explants (i.e., nodal explants excised from the regenerated shoots) were transferred to the fresh culture medium of the same composition as the induction medium, for further growth. Temperature and illumination conditions remained unaltered. The regenerated shoots were subcultured every eight weeks by inoculating nodal explants (ca. 0.1 g fresh weight) into the fresh nutrient medium. The biomass, containing regenerated shoots and plantlets, was collected at the end of the growth cycle and dried at room temperature prior to the phytochemical examination.

Extraction and Isolation of Sesquiterpene Lactones
Shoots and plantlets, collected from the culture on solidified MS medium containing 0.5 µM TDZ, were dried and powdered. The dry plant material (25.7 g) was extracted with CHCl 3 (5 × 0.3 L). The organic extracts were combined and concentrated in vacuo to provide 1.07 g of an oily residue. The residue was fractionated by conventional column chromatography (CC) over silica gel (80 g) using an n-hexane-EtOAc gradient solvent system (up to 100% EtOAc). Collected fractions (50 mL each) were combined as shown by TLC.
Isolated compounds were identified based on their 1 H NMR spectra and comparison of the experimentally obtained data with those found in the literature [5,8,[19][20][21][22].

Sample Preparation
The dry and pulverized plant material (0.1 g) was extracted twice with 10 mL of acetone at room temperature for 3 h on a rotary shaker (100 r.p.m.). The extracts from the two consecutive extractions were combined and evaporated to dryness under reduced pressure to give a residue that was redissolved in 1 mL of MeOH and centrifuged (11,340× g, 5 min) prior to HPLC analysis.

Preparation of Plant Extracts
The dry and pulverized plant material (1 g) was extracted two times (3 h) with 100 mL of acetone, at room temperature, with shaking. The combined extracts were concentrated in vacuo to remove the solvent. From the dry residues, the stock solutions in MeOH (10 mg/mL) were prepared. The working concentrations of the extracts (from 5 to 100 µg/mL) were obtained by dilution of the stock solutions with the culture media.

Statistical Analysis
Data were analyzed using the Statistica software (v.13.3). The analysis of variance (one-way ANOVA) and post hoc Tukey test were used to show statistical significance with assumed p < 0.05.

Conclusions
In vitro cultivated shoots and plantlets of C. divaricatum produced terpenoids characteristic of the intact plant and the extracts from the in vitro cultured plant material demonstrated cytotoxic activity comparable to that of extracts from the field-grown plant. Taking into consideration the results of RP-HPLC-DAD analysis of the assayed extracts and results of the cytotoxicity assessment, the activity of extracts from both in vitro-and field-grown plants can not be unambiguously assigned to the presence of sesquiterpene lactones. Extracts of different origin demonstrated various specificity towards the cancer cells used in the study. The complicated correlation of the extract composition and cytotoxic activity should be a subject of further research.