Cytotoxic effects of four Caryophyllaceae species extracts on macrophage cell lines

Abstract

Context: Saponins have been reported to possess antitumor properties, to inhibit angiogenesis and to induce tumor apoptosis.

Objective: To test the possible cytotoxic effect of crude extracts from four Caryophyllaceae species including Gypsophila paniculata L., Gypsophila trichotoma Wend., Saponaria officinalis L., and Dianthus sylvestris Wulffen on cultured monocyte/macrophage cell lines.

Materials and methods: After acid hydrolysis of the methanol-aqueous extracts, two representative prosaponins of the Caryophyllaceae, gypsogenin 3-O-glucuronide and quillaic acid 3-O-glucuronide were purified using solid-phase extraction (SPE), then identified by ultra-performance liquid chromatography–electrospray/mass spectrometry (UPLC-ESI/MS). Cytotoxic activity of the crude extracts at concentrations ranging from 0.1 to 200 µg/ml was evaluated on rat alveolar macrophage NR8383 and human monocytic THP-1 cell lines. Apoptosis was determined by measuring caspase-3 activity.

Results: Quantitative analysis by reversed-phase high-performance liquid chromatography (RP-HPLC) revealed a high content of gypsogenin 3-O-glucuronide in Gypsophila species roots (0.52–1.13% dry weight). At a concentration ≥10 µg/ml of crude extracts, a significant reduction of NR8383 and THP-1 cell lines viability was evidenced using the Trypan blue exclusion test. D. sylvestris extract exhibited the highest toxicity against THP-1 cells. Caspase-3 activation was evidenced after 4 and 24 h incubation of macrophages with 100 µg/ml of S. officinalis and G. trichotoma extracts, indicating apoptosis induction.

Discussion and conclusion: Crude extracts from the assayed species revealed cytotoxic effects toward macrophage cell lines. In Gypsophila species, gypsogenin 3-O-glucuronide derivatives could be responsible for the observed cytotoxicity. Therefore, crude extract of Caryophyllaceae is worth investigating for the potential development of agents against cancer cells.

• Apoptosis
• caspase-3
• Dianthus sylvestris Wulffen
• gypsogenin 3-O-glucuronide
• Gypsophila paniculata L.
• Gypsophila trichotoma Wend.
• Saponaria officinalis L.

Introduction

Saponins are plant glycosides with a steroid or triterpenoid aglycone and one or more sugar chains, covalently linked by glycosidic and (or) ester binding to the aglycone (Dinda et al., 2010). Because of the great variability of their structures, distinct saponins have been attributed diverse functions. For instance, saponins are thought to be potential antitumor agents due to their inhibitory effects on tumor cell growth (Bachran et al., 2008; Fuchs et al., 2009; Podolak et al., 2010). Other studies have reported that saponins play a role in the induction of apoptosis (Sun et al., 2009; Xu et al., 2009; Zhang et al., 2008) and in the reduction of invasiveness and multidrug resistance (Yanamandra et al., 2003), which suggests they could represent alternative agents in cancer treatment. In addition, it has been shown that saponins combined with some conventional chemotherapeutic agents exerted synergistic inhibitory effects on tumor growth in vitro and in vivo (Fuchs et al., 2009). Moreover, saponin-adjuvanted particulate vaccines were reported to have great potential in cancer immunotherapeutics (Skene & Sutton, 2006). Among other interesting properties of saponins is their role in permeabilization and hemolysis (Gauthier et al., 2009). However, there are only a few studies with respect to the toxic activity of saponins from caryophyllaceous species in cultured cells.

Caryophyllaceae is an extremely rich source of triterpene saponins belonging to the group of GOTCAB saponins (glucuronide oleanane-type triterpene carboxylic acid 3, 28-bidesmosides) (Bottger & Melzig, 2010; Gevrenova et al., 2006; Henry, 2005; Tan et al., 1999; Weng et al., 2010). Previous reports have shown that saponins from Gypsophila oldhamiana Miq. (Caryophyllaceae) exhibited cytotoxic activities against different human cancer cell lines (Bai et al., 2007). Gypsophila saponins are of interest in terms of their application as vaccine adjuvants (Press et al., 2000). It has also been shown that Saponinum album (Merck) enhanced 100 000-fold the cytotoxicity of the type I ribosome-inactivating protein saporin from Saponaria officinalis L. (Hebestreit et al., 2006).

Based on all these studies, we aimed at investigating the cytotoxic effect of crude extracts from four Caryophyllaceae species on mammalian macrophage cell lines.

Materials and methods

Plant material

Gypsophila trichotoma Wend. roots were collected in August 2004 from the Black Sea (Kavarna region) (43°25′60′′N–28°19′60′′E) in Bulgaria and were identified by Dr. R. Gevrenova (Faculty of Pharmacy, Medical University-Sofia, Bulgaria). Gypsophila paniculata L. and S. officinalis roots were collected in 2009 from southwest of France (43° 33′43′′N–1°11′0′′E) while Dianthus sylvestris Wulffen shoots were collected from Grossio (Valtellina, Italy) (46°17′60′′N–10°16′0′′E). These plants were identified by Prof. Max Henry (Université de Lorraine, Nancy, France). Voucher specimens of plant material (HP101-4) were deposed in the Herbarium of the Faculty of Pharmacy of Nancy, Université de Lorraine, France.

Preparation of plant extracts and solutions for cell viability assay

Air-dried powdered parts of each plant (1 g) were treated with 50 ml of 10% methanol (×3) by sonication for 5 min at room temperature. The lyophilized extracts constituted the crude extracts: 317.9 mg (G. paniculata), 417.6 mg (G. trichotoma), 331.2 mg (S. officinalis), and 189.4 mg (D. sylvestris). The crude extracts were dissolved in phosphate-buffered saline (PBS) and their concentrations were adjusted at 10 mg/ml giving the stock solutions which were then filtered through a 0.22 µm filter (Millipore, Bedford, MA) and stored at 4 °C until use. For the cell viability assay, the stock solutions were diluted at 2 mg/ml, 1 mg/ml, 100 µg/ml, and 10 µg/ml.

Analysis of saponins in plant extracts

Acid hydrolysis of the extracts

Samples (50 mg) of plant extracts (three samples for each plant) and 30 mg of the saponin pure white Merck (Ref. N 7695, Merck, Darmstadt, Germany) were dissolved by sonication in 500 µl water. An acid hydrolysis to obtain prosaponins was performed according to Henry et al. (1989). Their purification was achieved by solid-phase extraction (SPE) on Bond Elut C18, 6 ml, 1 g cartridges (Varian, CA). After loading samples, and washing step with 5 ml of water, the prosaponins were eluted from the cartridges with three times 1 ml of methanol.

Analytical high performance liquid chromatography (HPLC) and ultra-performance liquid chromatography-electrospray/mass spectrometry (UPLC-ESI/MS)

Analytical HPLC and UPLC-ESI/MS were performed on a Shimadzu chromatographic system (Kyoto, Japan) at 210 nm and Waters ACQUITY UPLCTM system (Waters, Milford, MA), respectively. HPLC analysis was performed on a reversed phase column Symmetry Shield RP18, 5 µm, 250 × 4.0 mm I.D. (Waters, Milford, MA); UPLC-ESI/MS analyses were performed on a BEH Shield RP18 column (1.0 mm × 100 mm, 1.7 µm, Waters Corp, Ireland). All data were acquired and processed with Shimadzu CLASS-VP (Version 4.3) (HPLC) and Waters MssLynx software Version 4.1 chromatography data system (Milford, MA) (UPLC-ESI/MS). The mobile phase consisted of solvent A (sA): 25% acetonitrile with 0.1% ortho-phosphoric acid (HPLC) or 0.1% formic acid (UPLC) and solvent B (sB): 84% acetonitrile with 0.1% ortho-phosphoric acid (HPLC) or 0.1% formic acid (UPLC). The HPLC gradient program began at 100% sA followed by a linear gradient from 0 to 100% of sB for 15 min, an isocratic step (100% of sB) for 5 min (HPLC) or 1 min (UPLC) and then return to the initial conditions in 5 min (HPLC) or 3 min (UPLC). The flow rate was 1 ml/min (HPLC) or 190 µl/min (UPLC). The oven temperature was set at 40 °C (HPLC) and 50 °C (UPLC). About 10 μl (HPLC) and 5 μl (UPLC) aliquots of the above methanol eluate were used for analyses.

Quantitative determinations of gypsogenin 3-O-glucuronide and quillaic acid 3-O-glucuronide

Gypsogenin 3-O-glucuronide was obtained from commercial saponin pure white as described by Henry et al. (1989). The quantification of gypsogenin 3-O-glucuronide was carried out using the external standard method. Triplicate HPLC analyses were performed for each concentration (2, 1, 0.75, 0.50, 0.25, 0.125, and 0.063 mg/ml) at 210 nm. The retention time of the compound was 10.27 ± 0.21 min. Calibration curve was constructed from peak areas versus analyte concentrations. The regression equation was y = 1.2014 × 10⁷x + 7.8022 × 10⁵, r²= 0.9964 (Analytik-Software STL; Leer, Germany). The concentration of quillaic acid 3-O-glucuronide was estimated using the gypsogenin 3-O-glucuronide calibration curve. The amounts of quillaic acid 3-O-glucuronide are thus relative and not absolute.

For each sample, the complete assay procedure was carried out in triplicate and standard deviation (SD) was calculated. The repeatability of the HPLC method was ascertained by injecting the standard solution of gypsogenin 3-O-glucuronide (0.1 mg/ml) six times. The reproducibility was determined over 10 d by three injections per day of the same solution. The limit of detection (LOD) and the limit of quantification (LOQ) were calculated according to ICH guidelines (International Conference on Harmonisation, 1995). They were based on SD of the regression line of specific calibration curves and their slope, using analyte concentrations in the range of their LOD and LOQ.

The recovery of analyte was evaluated by applying the entire SPE procedure to a G. paniculata hydrolysis mixture that had been spiked with a standard solution of gypsogenin 3-O-glucuronide (0.1 mg/ml) and measured in triplicate. The percentage recovery was determined by subtracting the values obtained for the control matrix preparation from those samples that had been prepared with the added standard, divided by the amount added of gypsogenin 3-O-glucuronide, and multiplied by 100. Using obtained data, the mean recovery and the standard deviation were derived.

Determination of cell viability and apoptosis

Cells and cell culture

The rat alveolar macrophage NR8383 (CRL-2192) and the human monocytic THP-1 (TIB-202) cell lines were from the American Type Culture Collection (ATCC, Manassas, VA). The NR8383 cells were grown as 50% adherent and 50% floating cells in Dulbecco’s modified Eagle’s medium (DMEM) (GIBCO, In-vitrogen, Cergy Pontoise, France) supplemented with 15% of fetal calf serum (Eurobio, Les Ullis, France), 200 mM l-glutamine, and a mixture of antibiotic/antimycotic (100 U/ml of penicillin, 100 μg/ml of streptomycin and 0.25 μg/ml of amphotericin B) (Sigma Aldrich, Saint Quentin Fallavier, France) at 37 °C under 5% CO₂. The THP-1 cells were grown in RPMI 1640 Glutamax (GIBCO Life, Cergy Pontoise, France) supplemented with 10% fetal calf serum and a mixture of antibiotic/antimycotic (100 U/ml of penicillin, 100 μg/ml of streptomycin and 0.25 μg/ml of amphotericin B) (Sigma Aldrich, Saint Quentin Fallavier, France), at 37 °C under 5% CO₂.

Cell viability assay

Trypan blue exclusion assays were carried out as previously described (Liu et al., 1999). Briefly, NR8383 cells (at 5 × 10⁵ cells/ml) or the THP-1 cells (at 2 × 10⁵ cells/ml) were dispensed into 24-well microplates, in 1 ml of medium/well and incubated overnight at 37 °C, under 5% CO₂. Cells were then treated with plant extracts for 24 h at the final concentrations of 200, 100, 50, 10, 1, and 0.1 μg/ml for NR8383 cells and 100 µg/ml for THP-1 cells. Control cultures were maintained under the same conditions. At the end of incubation time, seven enumerations per sample were performed by immediate microscopic observation, after blue Trypan staining; using a KOVA counting cell (Hycor, VWR, Strasbourg, France). Viability is expressed as the percentage of cells alive after contact with the different extracts.

Caspase-3 activity determinations

Caspase-3 activity in cell lysates was measured using the “EnzChek” Caspase-3 assay kit (Molecular Probes, Eugene, OR). Briefly, NR8383 cells (1 × 10⁶) were treated with 30 μl of plant extract stock solutions (10 mg/ml) dissolved in DMEM medium (final concentration in the cultures 100 µg/ml) for 4 and 24 h. Untreated cells were used as negative control. Positive control consisted of cells treated with 8 µg/ml camptothecin for 4 h at 37 °C. In addition, 1 mM of Ac-DEVD-CHO, an aldehyde caspase inhibitor, was added to cell treated with 100 mg/ml of S. officinalis extract for 10 min to ascertain that the observed fluorescence signal in induced cells was due to the activity of caspase-3. The fluorescence was read using a Fluoromark plate reader (Biotek, Colmar, France) with excitation at 360 nm and emission at 465 nm. Three independent assays were performed for each culture condition.

Statistical analysis

Data are expressed as mean ± SD. Student's t-test was used to evaluate the differences between experimental and control groups. p Values less than 0.05 were considered significant. IC₅₀ is calculated by Graph Pad Prism data analysis (Graph Pad software, La Jolla, CA).

Results

Quantitative determination of prosaponins

After a mild acid hydrolysis of the crude plant extracts, two representative prosaponins of the Caryophyllaceae were purified by SPE and identified by UPLC-ESI/MS (Figure 1A). These compounds were characterized by the presence of deprotonated molecular ions [М–Н]⁻ at m/z 645 and 661 (Figure 1A). HPLC-ESI/MS/MS of the [М–Н]⁻ ions was used to identify the aglycones and sugar moieties of the prosaponins (Lee et al., 1996). The product ions at m/z 469 (A₁) (Figure 1B) and 485 (B₁) (Figure 1C) are obtained because of the cleavage between the sugar ring and the aglycones of the prosaponins A and B, respectively. These ions correspond to [М–Н-176]⁻ formed by the elimination of a glucuronic acid residue. The loss of one water molecule from these ions yield fragment ions at m/z values 451 (A₂) (Figure 1B) and 467 (B₂) (Figure 1C). The mass difference of 176 between the daughter ions and the parent ions at m/z 645 and 661 is in agreement with sapogenins gypsogenin (A₁) (Figure 1B) and quillaic acid (B₁) (Figure 1C). Thus, in our experiment, peak A (Figure 1B) was identified as gypsogenin 3-O-glucuronide, and peak B (Figure 1C) as quillaic acid 3-O-glucuronide.

Figure 1. Identification of prosaponins. (A) UPLC-ESI/MS analysis of prosaponins – full-scan ESI mass spectrum of the prosaponins from G. paniculata roots in negative ion mode (A – gypsogenin 3-O-glucuronide; B – quillaic acid 3-O-glucuronide). (B) UPLC-ESI/MS/MS spectra of gypsogenin 3-O-glucuronide: A-[M − H]^−; A1-[M − Н–(GlcA–Н₂О)]⁻; A2-[(M − H)–GlcA]⁻. (C) UPLC-ESI/MS/MS spectra of quillaic acid 3-O-glucuronide: B-[M − H]⁻; B1-[M − Н–(GlcA–Н₂О)]⁻; B2-[(M − H)–GlcA] ⁻.

A RP-HPLC method involving gradient elution and UV detection for the quantification of these two prosaponins was applied following pretreatment of the samples by SPE. The detection of saponins at 200–210 nm limits the selection of solvents and the gradients that can be used. At this range of wavelength, acetonitrile gives much lower absorption than methanol, thus the selection of acetonitrile–water gradients seems to be the mode of choice (Oleszek & Bialy, 2007). Peaks of gypsogenin 3-O-glucuronide were assigned in the HPLC chromatograms by comparing individual peak retention times with these of the authentic reference standard, as well as by spiking techniques. The applied HPLC method gave linear results in the range of 0.06–2.00 mg/ml. RSDs of the repeatability and reproducibility were found to be 1% and 1.64%, respectively. LOD and LOQ were 0.0207 and 0.0627 mg/ml, respectively. The mean recovery of gypsogenin 3-O-glucuronide in the spiked control sample was 94.7%. The precision of the overall analytical procedure was estimated by measuring the within-day repeatability; the obtained relative standard deviation of the parallel results (n = 3) was 3.1%.

In both G. paniculata and G. trichotoma, the content of gypsogenin 3-O-glucuronide was higher than the one of the quillaic acid 3-O-glucuronides (Table 1). The gypsogenin 3-O-glucuronide represented 63% of the total root prosaponin content in G. paniculata, and 96% in G. trichotoma. The concentration of prosaponins was substantially lower in S. officinalis roots but with a relative level of the two prosaponins more balanced of 56% of gypsogenin 3-O-glucuronide and 44% of quillaic acid 3-O-glucuronide. These two compounds were not evidenced in D. syslvestris.

Table 1. Content of gypsogenin 3-O-glucuronide and quillaic acid 3-O-glucuronide (% dry weight) in the roots of the studied Caryophyllaceae species (n = 3).

Prosaponins	G. paniculata	G. trichotoma	S. officinalis	D. sylvestris
Gypsogenin 3-O-glucuronide	0.52 ± 0.01	1.13 ± 0.10	0.10 ± 0.02	–
Quillaic acid 3-O-glucuronide^a	0.30 ± 0.07	0.04 ± 0.006	0.08 ± 0.008	–

^aAmounts of quillaic acid 3-O-glucuronide are relative and not absolute.

Cytotoxicity and apoptosis

After 24 h of incubation of macrophage cell line NR8383 with 100 µg/ml of different extracts, the Trypan blue exclusion test showed a cell viability of 40 and 60% when S. officinalis and G. trichotoma extracts were used, respectively (Figure 2). S. officinalis extract seems to be the most cytotoxic (p < 0.0001) to cells while G. paniculata extract exhibited only a moderate effect. Cytotoxicity of extracts at 100 µg/ml decreased in the following order: S. officinalis > G. trichotoma > D. sylvestris > G. paniculata. An IC₅₀ of 97.1 µg/ml was calculated for S. officinalis extract. For G. paniculata and D. sylvestris extracts, the viability was above 50% at the studied concentration range (Figure 2). At concentrations of 0.1 and 10 µg/ml of each extract, cell viability was not significantly different from that of the control.

Figure 2. Cell viability (NR8383 macrophages) assayed by the Trypan blue exclusion test. Cells were treated with extracts of Gypsophila paniculata, G. trichotoma, Dianthus sylvestris, and Saponaria officinalis in the range from 200 to 0.1 µg/ml for 24 h as described in the Methods section. Viability was expressed as the percentage of cells alive after contact with the different extracts. Values are from seven independent experiments and represent the mean viable cells (±standard deviation). S. officinalis extract at 100 µg/ml has the highest cytotoxicity on macrophage cells. *Values significantly different from the control p < 0.05.

Similar experiments were conducted to test individually the effect of isolated extracts on THP-1 cells after 24 h incubation. D. sylvestris extract showed the highest cytotoxicity against these cells (p < 0.001), while S. officinalis, G. trichotoma, and G. paniculata extracts had the same effect on THP-1 cells but significant (p < 0.05) (Figure 3).

Figure 3. Cell viability (THP-1 cells) assayed by the Trypan blue exclusion test. Cells were incubated with 100 µg/ml of assayed plant extracts for 24 h. Results are expressed as mean ± standard deviation (n = 3). D. sylvestris extract shows the highest cytotoxic activity. *Values significantly different from the control p < 0.05.

Next, we assayed S. officinalis, G. trichotoma, G. paniculata, and D. sylvestris extracts for their possible apoptosis effect on NR8383 cell line. Our data demonstrated that S. officinalis, G. trichotoma, and D. sylvestris extracts were associated with caspase-3 activity in a time-dependant manner in NR8383 cell line (Figure 4). After a 24 h incubation of cells, either with S. officinalis or G. trichotoma extract, caspase-3 activity increased significantly (p < 0.0001) as compared to that of the control, whereas incubation for a period of 4 h had little or no effect. Interestingly, D. sylvestris displayed a significant activity of caspase-3 at 4 h (p < 0.001) but no activity at 24 h. However, G. paniculata extract did not show any induction of caspase-3 activity neither at 4 h nor at 24 h incubation. These latest data are consistent with the observed weak effect of G. paniculata extract on cell viability (Figure 3). When the inhibitor of caspase-3 Ac-DEVD-CHO was added to cells treated with saponin extract, no fluorescence was observed which confirms the involvement of caspase-3 in response to treatments with extracts (Figure 4).

Figure 4. Detection of caspase-3 activity in macrophage NR8383 cell line treated with extracts of Saponaria officinalis, Gypsophila trichotoma, Dianthus sylvestris, and Gypsophila paniculata. Cells were treated with 100 µg/ml of plant extracts for 4 and 24 h. Untreated cells were used as a control. Cells were treated with 8 µg/ml camptothecin for 4 h to induce apoptosis and were used as positive control. In addition, an aldehyde inhibitor (Saponin extract-Inhibitor) was used to confirm that the observed fluorescence signal is due to the activity of caspase-3. Saponaria officinalis and Gypsophila trichotoma extracts induced a time-dependent apoptosis in the macrophage cell line. Each bar represented the percentage (mean ± SD of triplicate determinations). Asterisks represent values significantly different from the control: *p < 0.01. **p < 0.05 versus control.

Discussion

In the present study, we show for the first time that extracts from species of Caryophyllaceae including S. officinalis, G. trichotoma, and D. sylvestris affect mammalian monocytes/macrophages cell lines viability and induce apoptosis through caspase-3 activation.

Plant extracts are characterized by means of their prosaponins content

Due to the fact that GOTCAB saponins isolated from the Caryophyllaceae species usually are a mixture of structurally related forms with very similar polarities, their separation remains a challenge (Henry, 2005; Oleszek & Bialy, 2007). Earlier, we developed a method for the determination of Gypsophila saponins by means of their prosaponin gypsogenin 3-O-glucuronide, which is much more stable than the aglycone gypsogenin (Henry et al., 1989). It is also known that determination of triterpene saponins with liquid chromatography creates analytical problems, due to the lack of appropriate chromophores (Oleszek & Bialy, 2007). Therefore, the development of hyphenated techniques combining UPLC with mass detection could offer an opportunity to obtain fast analysis and a better resolution. In our study, the UPLC-ESI/MS method was used for the identification of two prosaponins obtained from Caryophyllaceae species, gypsogenin 3-O-glucuronide and quillaic acid 3-O-glucuronide. The validated SPE-HPLC method enabled the determination of assayed prosaponins with advantages in terms of recovery and retention time (about 10 min) in comparison with that reported by Henry et al. (1989). Here, we noticed that S. officinalis, which contains high amounts of detergent saponins, gave a low yield of prosaponins compared to Gypsophila species. This could be explained by the fact that among the 15 saponins of S. officinalis previously described, 11 did not contain gypsogenin 3-O-glucuronide and quillaic acid 3-O-glucuronide (Koike et al., 1999; Tan et al., 1999). S. officinalis saponins are derived from different sapogenins, namely gypsogenic acid, hydroxygypsogenic acid, and derivatives (Bottger & Melzig, 2010). Usually, they possess a xylose at C-3 and branched sugar chains in position C-28 of the aglycone. Interestingly, assayed prosaponins were not found in D. sylvestris. The genus Dianthus L. seems to be an exception to the GOTCAB saponins of Caryophyllaceae.

Plant extracts have a cytotoxic effect on macrophage cell lines and induce activation of caspase-3

To test the biological effects of obtained extracts, we used NR8383 and THP-1 cell lines as a cell model for toxicity studies because in the body, they are circulating cells representing one of the first defense lines in the body (Eidi et al., 2010). Here, we show for the first time that the methanol-aqueous extracts from S. officinalis, G. paniculata, G. trichotoma, and D. sylvestris can alter the growth and cell viability of mammalian monocytes/macrophage cell lines. Indeed, saponins from the roots of Gypsophila oldhamiana and Gypsophila pilulifera Boiss. & Heldr. (Caryophyllaceae) have been reported to display cytotoxic activity against human cancer cell lines (Arslan et al., 2012; Bai et al., 2007). It has also been shown that Saponinum album, G. paniculata, and Gypsophila arrostii Guss. var. nebulosa (Caryophyllaceae) saponins enhance the cytotoxicity of the plant toxin saporin which is an inhibitor of protein synthesis through a RNA cleaving (Arslan et al., 2013; Gilabert-Oriol et al., 2013; Hebestreit et al., 2006; Weng et al., 2008, 2010). Recently, many chemotherapeutic agents have been shown to activate apoptotic mechanisms, and the activation of caspases plays a crucial role in the biological events associated with apoptosis (Thornberry & Lazebnik, 1998). Studies reported that caspase-3 is both necessary and sufficient to trigger apoptosis. Evidence came from its dramatic induction in K562 cells by steroidal sapogenin (Liu et al., 2005) and triterpene saponins (Podolak et al., 2010; Sun et al., 2009; Xu et al., 2009; Zhang et al., 2008) or by its inhibition by specific inhibitors (Zhang et al., 2008).

Our study showed that S. officinalis and G. trichotoma extracts revealed the highest cytotoxicity on NR8383 cells, while G. paniculata extract displayed very weak activity. The inhibition rate of assayed extracts against THP-1 cell line correlated with that against NR8383 cells, with a stronger toxicity of D. sylvestris on human cells.

To understand how Caryophyllaceae extracts can be toxic to cell lines, we examined the caspase-3 activity in the plant extracts-treated macrophage cells using a caspase-3-specific substrate, Z-DEVD-AMC, which is cleaved to produce a fluorescent product. Our results showed that caspase-3 activity in the cells treated with S. officinalis and G. trichotoma root extracts was markedly elevated, which was consistent with the observed high effect of these extracts on cell viability. This clearly shows that the cytotoxic activity of these species can be attributed to apoptosis induction through caspase-3 activation.

With respect to the prosaponins specificity, only G. trichotoma exerts a gypsogenin 3-O-glucuronide-specific cytotoxicity and, at the same time, induces caspase-3 activity. G. paniculata yielded lower amount of the assayed prosaponins. This suggests that the effects of G. trichotoma extract on cell viability, shown in our study, could be due to predominant gypsogenin 3-O-glucuronide, to the relative ratio of the assayed prosaponins, and to their derivatives, as well. The saponins of the assayed Gypsophila species are bisdesmosides of gypsogenin and quillaic acid belonging to GOTCAB group; the linking points for sugar moieties to the triterpenoid skeletons are at C-3 and C-28 (Arslan et al., 2012, 2013; Frechet et al., 1991; Weng et al., 2010). Our previous study of G. trichotoma led to the isolation of two new aminoacyl triterpene saponins (Gevrenova et al., 2006). In contrast, we showed that new bisdesmosides from G. trichotoma, derivatives of gypsogenin 3-O-glucuronide, have an acetyl and (or) sulfate groups at the oligosaccharide attached to C-28 of the gypsogenin (unpublished data). The saponin mixture of these bisdesmosides had a lower cytotoxicity against the chronic myelogenous leukemic line BV-173 (IC₅₀ 67.11 µg/ml) compared to that of the crude extract (IC₅₀ 38.08 µg/ml) but it was more potent than the purified aminoacyl saponins (IC₅₀ 261.4 µg/ml) (data not shown). This is consistent with the results from the present study, where the relative ratio of the derivatives of gypsogenin 3-O-glucuronide and quillaic acid 3-O-glucuronide is a feature responsible for cytotoxicity of Gypsophila species.

Structures of both aglycone part and oligosaccharide chains are very important for the cytotoxic activity of saponins (Podolak et al., 2010). The cytotoxicity of oleanane-type saponins seems to be related to both the aldehyde group at C-4 and branched trisaccharide at C-3 of the aglycone (Bai et al., 2007; Podolak et al., 2010). In addition, it has been reported that monosulfated monosaccharide residues in triterpene saponins increase the cytotoxicity and caspase activation (Jin et al., 2009). Our study also indicated that the aminoacyl, acetylated, and sulfated saponins occurring in the studied extract may play a significant role in the cytotoxicity of G. trichotoma and caspase activation. S. officinalis saponins contain xylose in C-3 carbohydrate chain (instead of glucuronic acid in Gypsophila species) (Koike et al., 1999; Tan et al., 1999), but difference in the sugar moieties may not play a significant role in the cytotoxicity since S. officinalis and G. trichotoma show similar effects in terms of caspase-3 activation. To our knowledge, no data are available on saponins occurring in D. sylvestris yet. Gypsogenic acid is the predominant aglycone in the Dianthus species (Bottger & Melzig, 2010).

In conclusion, a rapid validated SPE-HPLC method was developed for quantitative determination of two prosaponins in Caryophyllaceae species. In addition, we showed for the first time that root extracts from S. officinalis and G. trichotoma inhibit growth in two macrophage cell lines and induce caspase-3 activity. These results suggest that gypsogenin 3-O-glucuronide derivatives in Gypsophila species could be responsible for the toxicity against monocyte/macrophage cell lines. Therefore, further investigations are necessary to bring insights into the structural requirements of saponins from G. trichotoma and S. officinalis which may lead to possible application of extracts as cytotoxic agents.

Declaration of interest

The authors report no declarations of interest

Acknowledgements

The authors thank Pr. A. Le Faou for English revisions.