Germacranolides from Carpesium divaricatum: Some New Data on Cytotoxic and Anti-Inflammatory Activity

Carpesium divaricatum Sieb. & Zucc., a traditional medicinal plant used as an inflammation-relieving remedy, is a rich source of terpenoids. At least 40 germacrane-type sesquiterpene lactones, representatives of four different structural groups, were isolated from the plant. Cytotoxicity against cancer cells in vitro is the most frequently described biological activity of the compounds. However, little is known about the selectivity of the cytotoxic effect. The anti-inflammatory activity of the germacranolides is also poorly documented. The objective of the present study was to assess the cytotoxic activity of selected C. divaricatum germacranolides-derivatives of 4,5,8,9-tetrahydroxy-3-oxo-germacran-6,12-olide towards cancer and normal cell lines (including cells of different p53 status). Moreover, to assess the anti-inflammatory effect of the compounds, the release of four proinflammatory cytokines/chemokines (IL-1β, IL-8, TNF-α and CCL2) by lipopolysaccharide-stimulated human neutrophils was measured by ELISA. The investigated sesquiterpene lactones demonstrated nonselective activity towards prostate cancer (Du145 and PC3) and normal prostate epithelial cells (PNT2) as well as against melanoma cells (A375 and HTB140) and keratinocytes (HaCaT). Cytotoxic activity against osteosarcoma cells was independent of their p53 status. In sub-cytotoxic concentrations (0.5–2.5 µM) the studied compounds significantly decreased cytokine/chemokine release by lipopolysaccharide-stimulated human leukocytes.


Introduction
Plants from the genus Carpesium (Asteraceae, Inuleae, Inulineae), native to Asia and Europe, have long been used by the traditional systems of medicine in Far Eastern countries. Preparations of whole Carpesium plants have been applied as antipyretic, analgesic, antiparasitic and inflammation-relieving remedies. In China, Korea and Japan, C. divaricatum Sieb. & Zucc. (Jin Wa Er) found use in the treatment of common cold, fevers, toothache, sore throat, parasitic infestations, diarrhea and urinary tract infections [1,2].
Phytochemical studies on C. divaricatum started in the 1990s with the isolation of three germacranolide-type sesquiterpene lactones by Maruyama [3]. Since that time, over 40 compounds from this class of secondary metabolites have been isolated from plants of the species [4][5][6][7][8][9]. Moreover, acyclic diterpenes, monoterpenoid thymol derivatives, 12-oxophytodienoic acid and numerous caffeic acid derivatives were identified as constituents of

Cytotoxic
The examined compound demonstrated moderate cytotoxicity to both cancer a normal cells in vitro. It is worth noting that 3 was more active against PC3 cells th doxorubicin, the reference cytostatic drug. IC50 values of 3 ranged from 3.88 µ g/mL (C co-2 cells) to 16.39 µ g/mL (PNT2 cells). All cell lines used in the experiment were su ceptible to the studied compound.

Effects of 1 and 2 on Two Lines of Osteosarcoma with Different p53 Status
Two osteosarcoma cell lines: U-2 OS (with wild-type p53 status) and SAOS-2 (p null cell line) were used to evaluate cytotoxic activity of 1 and 2. Doxorubicin was e ployed as a reference drug. Viability of the cells treated by 1 or 2 (concentration ran 0.014-50.0 µ M) for 24 and 48 h was assessed by thiazolyl blue tetrazolium bromide (MT test. Results are shown in Figures 2 and 3.    normal cells in vitro. It is worth noting that 3 was more active against PC3 cells th doxorubicin, the reference cytostatic drug. IC50 values of 3 ranged from 3.88 µ g/mL (C co-2 cells) to 16.39 µ g/mL (PNT2 cells). All cell lines used in the experiment were su ceptible to the studied compound.

Effects of 1 and 2 on Two Lines of Osteosarcoma with Different p53 Status
Two osteosarcoma cell lines: U-2 OS (with wild-type p53 status) and SAOS-2 (p null cell line) were used to evaluate cytotoxic activity of 1 and 2. Doxorubicin was e ployed as a reference drug. Viability of the cells treated by 1 or 2 (concentration ran 0.014-50.0 µ M) for 24 and 48 h was assessed by thiazolyl blue tetrazolium bromide (MT test. Results are shown in Figures 2 and 3.

Reactive Oxygen Species (ROS) Production
Activation of PMNs, either at a site of inflammation or induced in vitro by proinflammatory agents, results in an oxidative burst in these cells. The phenomenon is characterized by intense ROS generation and liberation of proteolytic enzymes from azurophilic granules. An effect of the sesquiterpene lactones 1 and 2 on ROS production in PMNs was assessed in response to N-formyl-Met-Leu-Phe (f-MLP) stimulation. The examined sesquiterpenoids significantly and dose-dependently reduced ROS generation at a concentration range of 1-2.5 µM ( Figure 5).    In response to stimulation with proinflammatory agents, e.g., LPS or f-MLP, human neutrophils excrete an array of cytokines and chemokines, including TNF-α, IL-1β, IL-8 and CCL2 [29,30]. In the present study, neutrophils were pretreated with the examined compounds (1 or 2) before their priming with LPS. Levels of IL-1β, IL-8, TNF-α and CCL2 were determined in the culture medium, using ELISA, 24 h after LPS stimulation. Preincubation of human neutrophils with each of the investigated compounds (at concentrations of 0.5, 1.0 and 2.5 µM) caused significant and dose-dependent inhibition of IL-1β, IL-8 and CCL2 production upon LPS stimulation (Figure 6a,b,d). The compounds were slightly less active as inhibitors of TNF-α excretion. Compound 1 caused a statistically significant reduction in the release of this cytokine only in the two higher concentrations (1.0 and 2.5 µM, see Figure 6c). In the case of compound 2, a statistically significant effect was observed for all tested concentrations. In response to stimulation with proinflammatory agents, e.g., LPS or f-MLP, human neutrophils excrete an array of cytokines and chemokines, including TNF-α, IL-1β, IL-8 and CCL2 [29,30]. In the present study, neutrophils were pretreated with the examined compounds (1 or 2) before their priming with LPS. Levels of IL-1β, IL-8, TNF-α and CCL2 were determined in the culture medium, using ELISA, 24 h after LPS stimulation. Preincubation of human neutrophils with each of the investigated compounds (at concentrations of 0.5, 1.0 and 2.5 µ M) caused significant and dose-dependent inhibition of IL-1β, IL-8 and CCL2 production upon LPS stimulation (Figure 6a,b,d). The compounds were slightly less active as inhibitors of TNF-α excretion. Compound 1 caused a statistically significant reduction in the release of this cytokine only in the two higher concentrations (1.0 and 2.5 μM, see Figure 6c). In the case of compound 2, a statistically significant effect was observed for all tested concentrations. (c) (d)

Discussion
In the past five years, a series of papers on sesquiterpene lactones from C. divaricatum has been published by Zhang and coworkers [6][7][8][9]. They isolated and described 20

Discussion
In the past five years, a series of papers on sesquiterpene lactones from C. divaricatum has been published by Zhang and coworkers [6][7][8][9]. They isolated and described 20 new highly oxygenated sesquiterpene lactones of germacrane type and elucidated the stereochemistry of some formerly known compounds. The germacranolides of C. divaricatum were divided into four structural types (I-IV). Compounds of type IV made up about 50% of the sesquiterpene lactones isolated from the plant [8,9].
Our previous work [12] described the isolation of 12-oxo-phytodienoic acid (12-OPDA) from aerial parts of C. divaricatum. By fractionation of a chloroform extract from the same plant material, several mixtures of sesquiterpene lactones were obtained in addition to the oxylipin. Further separation of the mixtures led to the isolation of individual compounds of different purity that were identified based on the spectral data available from the literature. Three of the compounds, all included in the structural type IV, were selected for further investigation. Compound 1, first isolated by Goswami et al. [27] from Inula cappa, and compound 3, first isolated by Gao et al. [28] from C. triste, were identified as constituents of C. divaricatum by Zhang et al. [8]. The compounds proved to be cytotoxic towards all selected cancer cell lines except for the MCF-7 cell line (resistant to compound 1) [8] and the A459 cells (resistant to compound 3) [7]. Cytotoxicity of 1 and 3 against normal cells had not been investigated. Compound 2 (cardivarolide G) was first described as a constituent of C. divaricatum by Zhang et al. [8], and its biological activity had not been evaluated previously. To assess the selectivity of the cytotoxic effect towards cancer cells, both normal and cancer cell lines were treated in vitro with different concentrations of compound 3. The investigated cell lines were divided into three panels: prostate (normal prostate epithelial cells and two prostate cancer cells with different metastatic potential), skin (normal keratinocytes and two melanoma cell lines) and gastrointestinal (two colon cancer cell lines). Both normal and cancer cells were susceptible to compound 3. Parthenolide, a frequently investigated germacranolide with known proapoptotic activity in cancer cells, was shown to induce p53 protein-the transcription factor that controls cell cycle progression and apoptosis [31,32]. Incaspitolide A (germacranolide of type III, not known as a constituent of C. divaricatum) also upregulated p53 expression [14]. Some cell lines with p53 deletion proved to be resistant to doxorubicin-induced apoptosis [33]. To establish whether or not p53 plays a crucial role in the cytotoxic activity of the examined germacranolides, two osteosarcoma cell lines (with wild-type p53 status and with p53 deletion) were treated with 1 and 2 (at a concentration range of 0.014-50 µM). Our results (Figures 2 and 3) demonstrated that both compounds exerted a cytotoxic effect towards the treated cells. IC 50 values calculated for cardivarolide G (2) were slightly lower than those estimated for 1. The examined lactones seem to have proapoptotic activity, as revealed by microscopic observation. Their cytotoxic effect could be observed already after 24 h treatment and did not increase significantly after another 24 h. Moreover, the cytotoxic activity was independent of the p53 status of the cancer cell line. The response of the treated cells to doxorubicin differed from the response to 1 and 2. Doxorubicin-treated cells seem to be first cell cycle-arrested (after 24 h) and then killed (48 h), as was evidenced by a strong drop in IC 50 value.
Polymorphonuclear leukocytes (PMNs) play a vital role in the human immune system. The cells, among others, are involved in the regulation of the inflammatory process and immune response via the capability to produce and to respond to a variety of small proteins engaged in cell signaling (cytokines) [29,30]. Some cytokines produced by human neutrophils have proinflammatory function and are implicated in the pathogenesis of inflammatory and autoimmune diseases in humans, such as psoriasis, rheumatoid arthritis, atherosclerosis or ulcerative colitis. Certain CNS dysfunctions characterized by neuronal degradation are also regarded to be connected with neuroinflammatory processes [34]. Secretion of four proinflammatory cytokines, i.e., IL-1β (leukocytic pyrogen), TNF-α and chemokines CCL2 (monocyte chemoattractant protein MCP-1) and IL-8 (neutrophil chemotactic factor CXCL8), by LPS-stimulated human neutrophils in the absence or presence of 1 or 2 was monitored by ELISA. Both tested compounds at a concentration range of 0.5-2.5 µM did not show a significant effect on the viability of LPS-treated PMNs (Figure 4). Preincubation of human neutrophils with the examined compounds significantly and in a concentration-dependent manner diminished f-MLP-induced ROS production by the cells (Figure 5). The effect was similar to that demonstrated for 12-OPDA, another C. divaricatum constituent investigated earlier [12]. The germacranolides 1 and 2 significantly and dose-dependently reduced secretion of proinflammatory cytokine IL-1β ( Figure 6a) and chemokines CCL2 and IL-8 (Figure 6b,d). Only at the lowest concentration of 1 (0.5 µM), a significant reduction in TNF-α secretion by LPS-stimulated neutrophils could not be achieved (Figure 6c). The effects of 1 and 2 on cytokine secretion were more pronounced than those displayed by 12-OPDA. The results suggest that C. divaricatum germacranolides of type IV can exert a potent anti-inflammatory effect in sub-cytotoxic concentrations.

General Methods
High-resolution mass spectra were obtained in the positive ion mode using Maldi-SYNAPT G2-S HDMS (Waters Corp., Milford, MA, USA) mass spectrometer equipped with an electrospray ion source and q-TOF type mass analyzer. 1 H NMR spectra were recorded either in CDCl 3 or in CD 3 OD on a Bruker AVANCE III HD 400 (resonance frequency 400.17 MHz) spectrometer (Bruker Corp., Billerica, MA, USA). Optical rotation was determined in CDCl 3 on a PolAAr31 polarimeter (Optical Activity Ltd., Huntingdon, England). RP-HPLC separations were performed using an Agilent 1200 Series HPLC system (Agilent Technologies Inc., Santa Clara, CA, USA) equipped with a column oven and a diode array detector. Analytical chromatographic separations were carried out on a Kinetex XB-C18 column (4.6 × 250 mm, 5 µm total particle size; Phenomenex, Torrance, CA, USA). Semipreparative RP-HPLC was conducted on a Synergi 4µ Fusion-RP, 80A, 250 × 10 mm column (Phenomenex), with an isocratic elution, using MeOH-H 2 O mixtures of different polarities. Conventional column chromatography was carried out on Silica gel 60 (0.063-0.2 mm, Merck, Darmstadt, Germany). TLC separations were performed using precoated plates (Silica gel 60, Art. No 5553, Merck, Darmstadt, Germany).
The examined compound (3) was diluted in the culture media from freshly made stock solution in MeOH (10 mg/mL) to the working concentrations (from 0 to 50 µg/mL). Cell viability was determined as described previously [36]. Cells suspended in the nutrient medium were transferred into 96-well microtiter plates (density 1.5 × 10 4 per well) and preincubated for 24 h (37 • C, 5% CO 2 ). Then, the culture medium was replaced with the medium containing different concentrations of the assessed compound (1-50 µg/mL). After 24 h of incubation, the viability of the cells was determined using colorimetric lactate dehydrogenase (LDH) assay in comparison to the controls to which corresponding aliquots of MeOH diluted with culture media were added. Cells grown in the medium without the tested compound were used as control I (negative), and the positive control (control II) was obtained by incubation of the cells in the medium containing 1% Triton X-100. LDH released from the damaged cells into the cell culture medium was quantified by measuring the absorbance at 490 nm using a Synergy II Biotek microplate reader. Cytotoxicities of the examined compounds were calculated as follows: ((absorbance of the tested sample−absorbance of control I)/(absorbance of control II−absorbance of control I)) × 100. Results were means of three independent measurements (± SD). Doxorubicin (Ebewe Pharma G.m.b.H., Unterach, Austria) was used as a reference cytostatic drug. The IC 50 values were determined by plotting the percentage viability of the cells versus concentration either in linear or in logarithmic scale, and the appropriate calculations were made using either Microsoft Office Excel 2003 or AAT Bioquest website program (https://www.aatbio.com/tools/ic50calculator, accessed on 30 June 2021), respectively.

Cytotoxicity Assessment of 1 and 2 against U-2 OS and SAOS-2 Cell Lines
Human osteosarcoma cell lines U-2 OS (ATCC HTB-96; wild-type p53 status) and SAOS-2 (ATCC HTB-85; p53 deletion) were cultured in McCoy's 5A medium containing L-glutamine and supplemented with 10% FBS at 37 • C and in a humidified atmosphere with 5% CO 2 . First, 50 mM stock solutions of compounds 1 and 2 were prepared in DMSO and then diluted with DMSO to 1000×-concentrated working solutions (from 0.014 to 50 mM). After that, working solutions were diluted with culture media to the final 1× concentrations (0.014-50 µM).
For the MTT assay, the osteosarcoma cells suspended in the nutrient medium were seeded into 96-well microtiter plates (density 1.0 × 10 4 per well) and preincubated for 24 h (37 • C, 5% CO 2 ). Subsequently, the cells were treated with 1 or 2 at different concentrations (0.014-50 µM) and further cultivated for another 24 or 48 h. After that, MTT was added for 1 h at the final concentration of 500 ng/mL. The culture medium was removed and the formazan crystals were dissolved in isopropanol containing 40 mM HCl. Absorbance was measured using Tecan microplate reader at 570 nm with the reference wavelength of 650 nm. Graphs (Figures 2 and 3) show data normalized to DMSO-treated controls (set as 100%). Data were analyzed using Origin software. IC 50 values were calculated using the AAT Bioquest website program (https://www.aatbio.com/tools/ic50-calculator, accessed on 30 June 2021). Peripheral venous blood was obtained from healthy human donors (18-35 years old) in the Warsaw Blood Donation Centre. Donors did not smoke or take any medications. They were clinically recognized to be healthy, and routine laboratory tests showed all values to be within the normal ranges. Neutrophils were isolated by dextran sedimentation and centrifugation in a Ficoll Hypaque gradient and then resuspended in (Ca 2+ )-free HBSS buffer or RPMI 1640 medium. Blood samples from three donors were used in each experiment.

Cytotoxicity Measurement
Cytotoxicity was assessed by a standard flow cytometric probe using propidium iodide (PI) staining. After 24 h of incubation in the absence or presence of the tested germacranolide (at concentrations of 0.5, 1.0, 2.5 and 5.0 µM), the neutrophils (3.5 × 10 5 ) were harvested and centrifuged (1500 r.p.m., 10 min, 4 • C), washed once with cold PBS and resuspended in 400 µL of PBS. A 5 µL aliquot of PI solution (50 µg/mL) was added to the cell suspension. After 15 min of incubation with PI at room temperature, cells were analyzed by BD FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA), and 10,000 events were recorded per sample. The number of cells that displayed high permeability to PI, expressed as a percentage of PI(+) cells, was determined.

Reactive Oxygen Species (ROS) Production by Neutrophils
ROS production was measured using luminol-dependent chemiluminescence test. A 70 µL aliquot of neutrophil suspension (3.5 × 10 5 ) in (Ca 2+ )-free HBSS buffer, 50 µL of the tested compound solution and 50 µL of luminol (100 µM) were added to a well in a 96-well plate. ROS production was initiated by the addition of f-MLP (30 µL of 0.1 µg/mL solution) to obtain a total volume of 200 µL per well. Chemiluminescence changes were measured for 40 min, at 2 min intervals, in a microplate reader (37 • C). The background chemiluminescence produced by nonstimulated cells was also determined. The percentage of inhibition was calculated by comparison to the stimulated control without the tested compound, at the maximum luminescence. 4.7.4. Proinflammatory Cytokine/Chemokine (IL-1β, IL-8, CCL-2 and TNFα) Production by LPS-Stimulated Neutrophils Neutrophils (2 × 10 6 ) were cultured in 24-well plates in RPMI 1640 medium with 10% FBS, 10 mM HEPES and 2 mM L-glutamine, in the presence or absence of LPS (100 ng/mL) and in the absence or presence of 1 or 2 (final concentrations of the examined germacranolide in a range of 0.5-2.5 µM), at 37 • C with 5% CO 2 . The neutrophils were harvested after 24 h and centrifuged (2000 r.p.m., 10 min, 4 • C). The amount of released cytokines was measured by enzyme-linked immunosorbent assay (ELISA) following the manufacturer's instructions (BD Biosciences, USA). The effects on IL-8, IL-1β, CCL2 and TNF-α production by the neutrophils were calculated by comparing the percentages of the released agents to the stimulated control without the tested compound.

Statistical Analysis
Results are expressed as the mean ± SEM of three independent experiments performed at least in duplicate. All analyses were performed using Statistica 13 software. The statistical significance of the differences between means was established by ANOVA with Dunnett's post hoc test p values.

Conclusions
Sesquiterpene lactones of C. divaricatum, derivatives of 4,5,8,9-tetrahydroxy-3-oxogermacran-6,12-olide, displayed moderate nonselective cytotoxic activity against the normal and cancer cell lines used in the study. The lack of selectivity makes them poor candidates for lead compounds in the development of antineoplastic drugs. The compounds in sub-cytotoxic concentrations significantly and dose-dependently reduced secretion of proinflammatory cytokines from LPS-stimulated human neutrophils. Based on the above results, traditional therapeutic use of Carpesium preparations, though supported by the anti-inflammatory activity of the germacranolides, should be treated with caution.

Data Availability Statement:
The data presented in this study are available on request from the authors.