The chemical structures and the applied method of synthesis for the production of sesquiterpenes (EA910a, EA910b, zedoarol, gweicurculactone, VDS71, VDS58, furanogermenone and EA1184) are shown in Fig. 1. The total chemical synthesis of natural substances furanogermenone, zedoarol and gweicurculactone was based on the unified synthetic protocol and methods in our previously published studies (18-21). All natural and synthetic substances were purified by flash column chromatography with Merck silica gel 60 (particle size, 0.040-0.063 mm; EMD Millipore, Billerica, MA, USA). The purity of all compounds was established by nuclear magnetic resonance on a Bruker 300 AM (Bruker Corporation, Billerica, MA, USA) and Agilent 500 MHz spectrometer and a high-resolution mass spectra recorder with an Agilent Electrospray ionisation time-of-flight mass spectrometer (Agilent Technologies, Inc., Santa Clara, CA, USA). All tested compounds had purity levels of >95%.

The previously established cancer U-87 MG, MCF-7, murine erythroleukemia FLC clone 745 (MEL-745) and K562 cell lines, and a normal MRC-5 cell line, were obtained, stored and used in a routine manner (22,23) in the Laboratory of Pharmacology, School of Pharmacy, Aristotle University of Thessaloniki (Greece) (<25 passages were applied). The murine erythroleukemia MEL-745 cells were obtained from Dr C. Friend (Division of Cytology, The Sloan-Kettering Institute for Cancer Research, New York, NY, USA) (24) and were cultured as described by our previous study (25). Regarding the dispute on the misidentification of U87 MG (26), the U87 MG cell line used in this study is a clone originating from that established at the University of Uppsala (Uppsala, Sweden). Malignant U-87 MG (human epithelial glioblastoma grade IV astrocytoma), MCF-7 (human breast cancer) and MEL-745 cells, along with the normal MRC-5 (human fetal lung fibroblast) cells, were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% v/v fetal bovine serum (FBS) and 100 µg/ml penicillin and streptomycin. K562 (human leukemic) cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium. Cells were maintained in culture (37°C in 5% v/v CO₂), and were passaged every 2-3 days when 75-80% confluence was achieved using trypsin-EDTA (0.25% w/v), for the attached cultures. Trypsin-EDTA, DMEM, RPMI and FBS were purchased from Thermo Fisher Scientific, Inc. (Waltham MA, USA).

In all experiments, the new synthesized sesquiterpene compounds were tested by dissolving them in DMSO prior to application to the various cell cultures. The final concentration of DMSO was ≤0.1%, which had no effect on cell proliferation (data not shown). The malignant (U-87 MG, MCF-7, K562 and MEL-745) and normal (MRC-5) cell lines were seeded in 24-well plates at an initial concentration of 1×10⁵ cells/ml. For the attached cultures (U-87 MG, MCF-7 and MRC-5), cells were allowed to attach for 3-4 h (at 37°C in 5% v/v CO₂) prior to the addition of the sesquiterpenoids. The specified concentrations used in the cultures were between 1×10⁻⁷ and 1×10⁻⁴ M. Cells were grown in the presence of the molecules for 48 h prior to the cell proliferation being measured with Neubauer counting chambers under an optical microscope (x10 magnification). Subsequently, the calculation of the half-maximal inhibitory concentration (IC₅₀) values of each compound for a specific cell line was estimated. Moreover, cell death within cell cultures was also determined using the Trypan blue dye-exclusion method, as previously described (23).

Cytotoxicity was further assessed through the re-application of pretreatment and washout experiments using human malignant U-87 MG and normal MRC-5 cell cultures, respectively. The cells were pretreated for 24 and 48 h with the previously calculated IC₅₀ concentrations of VDS58. VDS58 is a guaiane sesquiterpene that was synthesized by our group in the Laboratory of Organic Chemistry of the Aristotle University of Thessaloniki, starting from the commercially available (R)-carvone and following an 12-step route (20,21). The re-application of VDS58 to U-87 MG cells was achieved by adding fresh DMEM containing the IC₅₀ concentration of VDS58 to the cells. The replenishment of the MRC-5 cell population and the removal of VDS58 from the cultures were achieved by washing out the cells twice with PBS and then incubating them with fresh DMEM in the absence of drug treatment. MRC-5 and U-87 MG cells were then cultured for 96 and 120 h, respectively. The assessment of cell proliferation and death in the cultures was performed every 24 h using the aforementioned methods.

Total cytoplasmic RNA isolation from U-87 MG and K562 cells, and the two-step RT-qPCR analysis (kit KK4602; Kapa Biosystems, Wilmington, MA, USA) were performed as previously described (22). The following primer sequences were used to amplify the indicated genes: β-actin (ACTB) forward, 5′-ttgctgacaggatgcagaag-3′ and reverse, 5′-tgatccacatctgctggaag-3′; BCL2 associated X apoptosis regulator (BAX) forward, 5′-tctgacggcaacttcaactg-3′ and reverse, 5′-gaggaagtccaatgtccagc-3′; BCL2 apoptosis regulator (BCL2) forward, 5′-acttcgccgagatgtcca-3′ and reverse, 5′-caaagaaggcc acaatcctc-3′; CDKN1 forward, 5′-gagcgatggaacttcgactt-3′ and reverse, 5′-gtgggaaggtagagcttggg-3′; CASP9 forward, 5′-tcgaagc caaccctagaaaa-3′ and reverse, 5′-cctccagaaccaatgtccac-3′; BAD forward, 5′-cagatcccagagtttgagcc-3′ and reverse, 5′-ctgctcctgctg gtgactg-3′; CASP3 forward, 5′-ggttcatccagtcgctttgt-3′ and reverse, 5′-aattctgttgccacctttcg-3′; CDK4 forward, 5′-accagatggcactta caccc-3′ and reverse, 5′-ccacagaagagaggctttcg-3′; CDK2 forward, 5′-ttgtcaagctgctggatgtc-3′ and reverse, 5′-tgatgaggggaagagga atg-3′; CDK6 forward, 5′-tgcacagtgtcacgaacaga-3′ and reverse, 5′-acctcg-gagaagctgaaaca-3′; cyclin D1 (CCND1) forward, 5′-ctgc gaagtggaaaccatc-3′ and reverse, 5′-ttgaagtaggacaccgaggg-3′; CASP8 forward, 5′-gatgacatgaacctgctgga-3′ and reverse, 5′-caggctcttgttgatttggg-3′; MYC forward, 5′-aggagaatgtcaagaggcga-3′ and reverse, 5′-ggccttttcattgttttcca-3′; catenin β1 (CTNNB1) forward, 5′-gctgggaccttgcataacctt-3′ and reverse, 5′-attttcac cagggcaggaatg-3′; RB transcriptional corepressor 1 (RB1) forward, 5′-tgtcagagagagagcttggt-3′ and reverse, 5′-ctcatctaggtcaactgctgc-3′; and transforming growth factor β1 (TGFB1) forward, 5′-actgcg gatctctgtgtcattg-3′ and reverse, 5′-acagtagtgttccccactggtc-3′. The ΔΔCq values of gene expression corresponding to the amplified mRNAs from each sample were calculated by normalizing the values to the internal control (ACTB). The fold-change in expression levels was calculated using the 2^−ΔΔCq method (27). All experiments were performed in triplicate.

Bioinformatic analysis was performed to estimate the levels of gene expression for the U-87 MG and K562 cell lines employed throughout this study to assess the cytotoxicity behavior of VDS58. In detail, processed gene expression values [transcript per million reads (TPM)] resulting from the analysis of RNA-Seq data obtained from the ENCODE project (28) were downloaded from the Gene Expression Omnibus repository for untreated samples of U-87 (GSE90176) and K562 (GSE78561) cell lines. Spike-in controls and genes with TPM values equal to zero were removed from the data, and the means of the TPM expression values for each gene were then calculated. Next, for each cell line, the mean TPM values were processed via logarithm (log2) transformation and normalized to z-scores. The density plots of the z-score gene expression values obtained for each cell line exhibited a great degree of similarity (data not shown). Using the z-scores for gene expression in U-87 MG and K562 cell lines and the Pathview R package (29), visualizations were created for: i) The Homo sapiens pathway of the cell cycle (hsa04110) using information from the Kyoto Encyclopedia of Genes and Genomes database (30-32); and ii) the expression of the genes that constitute this pathway.

All data represent at least 2 independent biological experiments and the data are expressed as the mean ± standard deviation. Comparisons were made using two-way repeated measures analysis of variance, followed by Dunnett’s test. All statistical analyses were performed using GraphPad Prism 6.0 (GraphPad Software, Inc. (La Jolla, CA, USA). P<0.05 was used to indicate a statistically significant difference.

The preliminary pharmacological evaluation of the obtained structural sesquiterpene derivatives shown in Fig. 1, including EA910a, EA910b, zedoarol, gweicurculactone, VDS71, VDS58, furanogermenone and EA1184, was performed in human glioblastoma U-87 MG cells by assessing their cytotoxicity profiles. The cultures were exposed to different concentrations (0.0001-0.1 mM; 10⁻⁷-10⁻⁴ M) of these compounds; cell proliferation and death were evaluated after 48 h of treatment. A small proportion of dead cells was observed in culture following treatment with all analogs (data not shown). Moreover, the majority of the synthetic sesquiterpene derivatives exhibited no substantial cytotoxicity based on their concentration-dependent inhibitory effects on the proliferation of U-87 MG cells; their IC₅₀ values were >10⁻⁴ M (data not shown). Notably, however, sesquiterpene VDS58 exhibited increased cytotoxicity in a concentration-dependent manner. The proliferation of U-87 MG cells was decreased ~70% following 48 h of exposure to VDS58 at 1×10⁻⁴ M, compared with untreated cells (Fig. 2A and B). The latter prompted further analysis of the behavior of VDS58 in cell cultures.

The initial VDS58-induced anti-proliferative activity exhibited by U-87 MG cells prompted the further investigation of its cytotoxicity profile in various cell lines (MCF-7, MRC-5, K562 and MEL-745). These cultures were exposed to increasing concentrations of VDS58 (10⁻⁷-10⁻⁴ M) for 48 h. As shown in Figs. 2 and 3, VDS58 reduced cellular proliferation of all cell lines in a concentration-dependent manner; and cells were grown either as attached monolayers (U-87 MG, MCF-7 and MRC-5) or as suspension (K562 and MEL-745) cultures. The estimated IC₅₀ values of VDS58 varied between the cell lines: 7×10⁻⁵ M for U-87 MG; 6.9×10⁻⁵ M for MCF-7; 9.2×10⁻⁵ M for MRC-5; 4.2×10⁻⁵ M for K562; and 2.9×10⁻⁵ M for MEL-745 (Table I). Notably, however, only in suspension, erythroleukemia cultures (K562 and MEL-745 cells) showed an increase in the proportion of dead cells, reaching ~80% at higher concentrations (1×10⁻⁴ M for K562, Fig. 3B; >7×10⁻⁵ M for MEL-745, Fig. 3D). Moreover, specific morphological changes have been recorded in the VDS58-treated cell cultures. As shown in Fig. 4, the treated U-87 MG cells lost their epithelial morphology and adherence properties. MRC-5 cultures exhibited a spindle-like phenotype and MCF-7 cells obtained a fibroblastic-mesenchymal type shape, whereas in suspension, K562 and MEL-745 cultures possessed large cells, an indicator of cytotoxicity. Thus, contrary to the limited effects observed for the other sesquiterpene molecules (EA910a, EA910b, zedoarol, gweicurculactone, VDS71, furanogermenone and EA1184), the VDS58 analog exhibited a notable concentration-dependent cytotoxicity profile, which was accompanied by cellular morphological changes in the malignant and normal (MRC-5) cells.

To better characterize the cytotoxic effects of VDS58, the cell cultures were treated continuously with the corresponding IC₅₀ of VDS58 for 4-5 days. As shown in Fig. 5, U-87 MG cells exposed to the IC₅₀ of VDS58 grew at slower rates in culture; however, after 48-72 h, the cell numbers were comparable to those of the control cells, with no notable changes in the proportion of dead cells (Fig. 5A and B). Moreover, MCF-7 cells treated with the IC₅₀ of VDS58 showed similar behaviors, although a treatment time of 96-120 h was required to obtain cell numbers comparable with those of the untreated cells (Fig. 5C and D). Notably, normal MRC-5 cells incubated with the IC₅₀ of VDS58 did not proliferate notably, even after 120 h of treatment (Fig. 5E and F). The erythroleukemia K562 and MEL-745 cell lines, which were exposed to the corresponding IC₅₀ of VDS58, exhibited slow proliferation rates compared with those of the controls (Fig. 6).

Notably, in the attached VDS58-treated cultures, particularly in the malignant U-87 MG and MCF-7 cells, the cells survived and exhibited high proliferation rates; an effect that was more prominent in the U-87 MG cells compared with that observed in the normal attached MRC-5 cells (Fig. 5). Thus, to gain more insight into the cellular effects of VDS58, particularly in glioblastoma U-87 MG and normal MRC-5 cells, a complementary set-up of kinetic experiments was designed. For U-87 MG cells, the question to be addressed in these experiments was whether the cells were able to avoid the effects exerted by VDS58 conveyed by its intracellular metabolism leading to its cellular inactivity. For MRC-5, the current study investigated whether the cells regained their proliferation capacity following the removal of VDS58 from the culture. To this end, these cultures were initially exposed for 48 h to the corresponding IC₅₀ of VDS58 for each culture (U-87 MG, 7×10⁻⁵ M; MRC-5, 9.2×10⁻⁵ M). A washout step was subsequently introduced, (referred to as time-point ‘0’), followed by the addition of fresh medium containing VDS58 (7×10⁻⁵ M) for U-87 MG cells, whereas for MRC-5 cells, the culture medium was free of VDS58. In all cases, the control cultures were included in the experimental design to facilitate the interpretation of the results. The results of the current study showed that malignant U-87 MG cells have the ability to regain their proliferative potential subsequent to the initial pretreatment (Fig. 7A). At 48-72 h after the re-addition of VDS58, the cell number was similar to that of the culture continuously treated with VDS58, while the untreated control exhibited a higher cell accumulation rate (Fig. 7). These data suggest that the observed behavior of U-87 MG cells to escape and overcome the proliferation inhibitory effects of VDS58 may be attributed to an intrinsic capacity of U-87 MG cells rather than the intracellular metabolism of VDS58. The results of a typical washout experiment for MRC-5 cells pretreated for either 24 or 48 h with VDS58 are shown in Fig. 8. The initial inhibition of cell proliferation, as expected with exposure to VDS58 (Fig. 5E), was followed by a proliferative phase after the replenishment of cultures and transfer of cells into VDS58-free fresh medium (Fig. 8A). This implies that the intracellular actions of VDS58 may not cause any permanent harmful effects on the proliferation of normal MRC-5 cells; thus, cells are able to grow again, soon after the removal of VDS58 from the culture. Such behavior suggests the capability of MRC-5 cells to regain their potential to proliferate even subsequent to exposure to VDS58. Alternatively, it indicates the presence of inherent molecular machinery within MRC-5 cells, enabling them to regain their full proliferation potential, as VDS58-pretreated cells grow after the removal of the agent and attain cell numbers comparable with that of the control culture.

Numerous compounds exert their cytotoxic effects by deregulating the cell cycle or by inducing apoptotic signaling pathways. The kinetic analysis of cell proliferation prompted the further elucidation of the molecular mechanisms underlying the cellular responses in VDS58-treated cultures via RT-qPCR gene expression analysis in the present study. The assessment focused on 16 genes associated with the cell cycle, apoptosis and/or senescence (Fig. 9). U-87 MG cells continuously treated with the IC₅₀ concentration of VDS58 exhibited a substantial increase in gene expression levels of CDKN1 (~12-fold), CASP9 (~7-fold), BCL2 (~4-fold), TGFB1 (~3.5-fold) and CTNNB1 (~2-fold) (P<0.05) (Fig. 9A). However, the significantly upregulated expression of these genes returned to basal levels after 48 h of treatment, excluding that of TGFB1 and CTNNB1, in which upregulated levels were maintained even after 72 h. Such data indicate the transient induction of the CDKN1 signaling pathway and the ability of U-87 MG cultures to overcome the effects of VDS58. Notably, the re-addition of VDS58 to the U-87 MG cultures after an initial 48 h exposure caused a more profound activation of all genes, excluding BAX, CASP8 and RB1 (Fig. 9B). Highly increased expression levels of BAD (~13-fold), CASP9 (~12-fold), CASP3 (~9-fold), TP53 (~8-fold), CDK6 (~6-fold) and CDKN1 (~5-fold), as well as MYC, CCND1, CDK2 and CDK4 (~3-fold), and TGFB1 (~2.5-fold), were observed (P<0.05). In the latter case, the expression levels of all genes returned to their initial levels, excluding those for MYC and TGFB1 following 48 h of cell re-exposure to VDS58. By contrast, the gene expression profiles of VDS58-treated K562 cells revealed no notable alteration in gene expression levels, since only TGFB1 (~3-fold), CTNNB1 (~2-fold), CASP3 (~2-fold) and BAX (~1.5-fold) showed a time-dependent increase for at least 48 h (P<0.05) (Fig. 10). Similarly, RT-qPCR analysis was performed with MCF-7 and MRC-5 cultures exposed to VDS58; however, no alterations in gene expression levels were recorded between treated and untreated cells (data not shown). Notably, within the panel of genes analyzed in MRC-5 cells, only CDK2, CDK6, BAD, BAX and BCL2 appeared to have been amplified by RT-qPCR, as also reported in our recent study (22).

The diverse cytotoxic responses observed upon the exposure of U-87 MG and K562 cells to VDS58 were accompanied by a differential gene expression profile in the same cultures (Figs. 9 and 10). To improve our understanding of such cellular behavior, a bioinformatic pathway analysis of whole genome RNA-Seq expression data from the ENCODE project available for parental U-87 MG and K562 cells was performed, focusing on genes involved in the cell cycle, senescence and/or apoptosis signaling pathways (Fig. 11). Analysis of the associations between gene expression levels associated with these aforementioned signaling pathways in the two parental cell lines unveiled substantial differences (Fig. 11A and D). Notably, the comparison between the two cell lines showed that crucial genes involved in the cell cycle pathway, including TGFB1, MYC, TP53, RB1, CDKN2A, CDKN1B, CDKN1, CDK2 and CCND1, exhibited greater variation in expression levels. Consequently, this existing variability in cell cycle gene expression profiles could contribute toward the notable diversity in the cytotoxic response of cultures to VDS58 treatment. To further analyze the gene expression profile and detect the molecular profiles in VDS58-treated cell cultures over time, heatmaps were created from the generated RT-qPCR data (Figs. 9 and 10), and are presented in Fig. 11B, C and E. Evidently, the existing variability in gene expression levels of parental U-87 MG and K562 cells from the RNA-Seq data of the ENCODE project was further confirmed in the RT-qPCR data by comparing control cultures (Fig. 11B vs. E). Moreover, the kinetic analysis of the gene expression levels analyzed following VDS58 exposure exhibited a different pattern during the 24-72 h course of investigation. This may contribute toward the observed differential pharmacological response observed in the two cell lines. The re-addition of VDS58 to U-87 MG cultures, after initial exposure to the same agent, allowed cells to present a clear molecular profile, in which two main groups of genes could be identified. In the first group, the expression of the BAX, RB1, CTNNB1, TGFB1 and MYC genes was increased after 24 h following the re-addition of VDS58 and maintained at this level afterwards. In the second group representing the rest of the genes under investigation, a transient 24-h activation of expression was observed, which after 48 h returned to the initial level reaching that of control untreated cultures (Fig. 11C).