Antiviral and cytotoxic evaluation of coumarins from Prangos ferulacea

Abstract

Context: Prangos ferulacea (L.) Lindl. (Apiaceae) is a perennial plant found in the Middle-East, where it is commonly used as an antispasmodic and anti-inflammatory agent. It is a rich source of coumarins.

Objective: To purify several coumarins from P. ferulacea and to screen their cytotoxicity and anti-herpes activity.

Materials and methods: Acetone extract of roots of P. ferulacea was subjected to several chromatographic separations to render pure coumarins (1–8). Anti-herpes virus effects of 1–7 were evaluated at concentration 2.5, 5, and 10 µgmL⁻¹, on a confluent monolayer of Vero cells infected with 25 PFU of HSV1. Cytotoxic effects of 1 and 2 were evaluated on an A2780S cell line using the MTT assay. The cells were exposed to a series of concentrations of coumarins (0.01–2.5 mM, 37°C, 72 h).

Results: Compounds 1–8 were identified as osthole, isoimperatorin, oxypeucedanin, psoralen, oxypeucedanin hydrate, gosferol, oxypeucedanin methnolate, and pranferol. This is the first report of occurrence of 4 and 7 in this plant. Compound 1 showed a viability of 9.41% ± 2.4 at 2.5 mM on A2780S cells (IC₅₀ = 0.38 mM). The cell survival of 2 at 2.5 mM was 46.86% ± 5.5 with IC₅₀ equal to 1.1 mM.

Discussion and conclusion: Compound 1 shows cytotoxic effects on the A2780S cell line. Compound 2 is a cyclooxygenase-2 inhibitor and the A2780S cell line does not express COX-2 which may interpret the non-toxic effect of the compound on this cell line. None of the tested compounds showed an anti-HSV effect at non-toxic concentrations.

• A2780S
• furanocoumarin
• gosferol
• HSV-1
• oxypeucedanin methnolate
• oxypeucedanin hydrate

Introduction

The genus Prangos (Jashir in Persian) belongs to Apiaceae or Umbelliferae family and consists of 30 species (Evans, 1989). Fifteen species of the genus Prangos are found in Iran, of which five are endemic (Mozaffarian, 1996). They have been used in Asian medicine as emollient, carminative (Zargari, 1992), tonic, antiflatulent, anthelmintic, antifungal, and antibacterial agents (Bouaoun et al., 2007; Ulubelen et al., 1995). Phytochemical investigations on different species of Prangos have led to the isolation of coumarins (Eshbakova et al., 2006; Razavi et al., 2008; Tada et al., 2002) and volatile oils (Baser et al., 2000; Sajjadi & Mehregan, 2003; Sajjadi et al., 2009, 2011; Sefidkon et al., 1998) from different parts of the plants.

Prangos ferulacea (L.) Lindl. is a plant found in the Mediterranean and Middle East regions including Iran (Ghahreman, 1986), which is traditionally used as a flavoring additive in yogurt (Coruh et al., 2007) and widely used as a provender for mutton (Sajjadi et al., 2009). Previous studies showed antioxidant (Coruh et al., 2007; Mavi et al., 2004), antimicrobial (Durmaz et al., 2006; Massumi et al., 2007), and hepatoprotective (Farokhi et al., 2012, 2013) efficacies, and antispasmodic effects on rat uterus (Sadraei et al., 2012) and ileum (Kafash-Farkhad et al., 2013; Sadraei et al., 2012). In addition, the mutton nutritive value (Coşkun et al., 2004) and abortifacient effects (Kazerooni et al., 2006) have been previously evaluated.

Prangos ferulacea appeared as a promising source of furanocoumarins (Abyshev, 1974), a type of natural products with a wide range of pharmacological effects. Following our previous research on Prangos spp. (Sajjadi et al., 2009, 2011), several coumarins such as osthole (1), isoimperatorin (2), oxypeucedanin (3), psoralen (4), oxypeucedanin hydrate (5), gosferol (6), oxypeucedanin methnolate (7), and pranferol (8) have been isolated from the root extract in the current study, in which compounds 4 and 7 were isolated for the first time from this plant.

HSV-related opportunistic infections are involved in the development of various ailments, and there is a growing need to find new antiviral compounds to address the emergency of drug-resistant viral strains (Cheng et al., 2002). Coumarins have previously been shown to be potential antiviral (Abonyi et al., 2010; Domagala et al., 1996) and cytotoxic (Finn et al., 2002; Lee et al., 2003; Musa et al., 2008) agents, so it was decided to test the cytotoxicity and antiviral effects of the isolated coumarins.

Materials and methods

General experimental procedures

¹H (500 MHz) and ¹³C (125 MHz) NMR spectra were measured on a Bruker spectrometer. Chemical shifts were referenced to the residual solvent signal (CDCl₃: δH 7.26, δC 77.0). MS spectra were recorded on an Agilent 5975C mass spectrometer (Agilent Technologies, Palo Alto, CA). Open column chromatography was performed using a Silica gel (70–230 mesh) column; medium pressure liquid chromatography (MPLC) was performed on a Büchi apparatus (BUCHI Corporation, New Castle, DE) equipped with two pump modules 601/605, a pump manager (C-615), and a fraction collector C-660 using silica gel (70–230 mesh) columns. Separations were monitored by TLC on Merck 60 GF₂₅₄ (0.25 mm) plates and were visualized by UV inspection and/or staining with 5% H₂SO₄ in ethanol and heating; Preparative HPLC were achieved on a Waters® apparatus equipped with a UV detector (λ_max 254, 366) using an HPLC column (Normal phase, Shimpack, Merck, Darmstadt, Germany, 250 × 20 mm), with 10 mL min⁻¹ as a flow rate.

Plant material

Prangos ferulacea roots were collected in Dasht-Room village, Kohkiloye and BoyerAhmad Province, Iran, in April 2010. The plant material was identified by Dr. Azizollah Jafari, University of Yasouj, Yasouj, and a voucher specimen (No. 2408) has been deposited at the Herbarium of School of Pharmacy, Isfahan University of Medical Sciences, Isfahan, Iran.

Extraction and isolation

The air-dried roots of P. ferulacea (500 g) were exhaustively extracted with acetone at room temperature. After removing the solvent in vacuo, the residue (50 g) was dissolved in methanol, kept in −20 °C and underwent chill filtration to get rid of triglycerides. The defatted extract (30 g) was purified by a vacuum liquid chromatography using mixtures of heptane (H) and EtOAc (E) (10:0–0:10) to afford eight fractions (A–H). Fraction E (H:E, 8:2) rendered a mass of impure crystals in which recrystallization resulted in pure compound 1 crystals (1.8 g), and further purification of mother liquor resulted in pure crystals of compound 2 (600 mg).

Fractions F and G were mixed and purified using MPLC (Silica gel, Hep:EtOAc (7:3–5:5), flow rate = 20 mL min⁻¹) to get six fractions as FG1–6. FG-2 (H:E; 6:4) contained pure crystals of compound 3 (1.1 g).

Fraction H was purified on an open column using silica gel and Hep:EtOAc (6:4–2:8) to afford six fractions as H-1 to 6 in which fractions H-1 and H-6 were compounds 4 (200 mg) and 5 (215 mg), respectively. Fraction H-4 was further purified using a normal phase HPLC (Hep:EtOAc (6:4–4:6), flow rate = 10 cc min⁻¹) to afford six fractions in which H-4c to H-4e were pure coumarins 6 (4 mg), 7 (27 mg), and 8 (2 mg).

Osthole (1); 7-methoxy-8-isopentenyl coumarin, pale yellow crystals; EI-MS m/z 244[M]⁺, 229 [M − CH₃]⁺, 213 [M − OCH₃]⁺, ¹H-NMR (CDCl₃, 500 MHz): 7.64 (1H, d, J = 9.4, H-4), 7.32 (1H, d, J = 8.6, H-5), 6.86 (1H, d, J = 8.6, H-6), 6.26 (1H, d, J = 9.4, H-3), 5.26 (1H, t, J = 7.0, H-2′), 3.57 (2H, d, J = 7.0, H-1′), 3.95 (3H, s, 7-OCH₃), 1.87, 1.70 (each 3H, s, 3′-CH₃).

Isoimperatorin (2); (5-[(3-Methyl-2-butenyl)oxy] psoralen; white cubic crystals; EI-MS m/z 270 [M]⁺, 255 [M − CH₃], 227 [M − (CH₃)2CH]⁺. For ¹H-NMR data, see Table 1.

Table 1. ¹H-NMR (500 MHz) spectral data of compounds 1–8 (CD₃OD, δ in ppm).

Pos.	2	3	4	5	6	7	8
3	6.30, d, J = 9.6	6.34, d, J = 9.8	6.42, d, J = 9.6	6.35, d, J = 9.8	6.33, d, J = 9.8	6.30, d, J = 9.8	6.34, d, J = 9.5
4	8.19, d, J = 9.6	8.23, d, J = 9.8	7.83, d, J = 9.6	8.21, d, J = 9.8	8.23, d, J = 9.8	8.25, d, J = 9.8	8.21, d, J = 9.5
5	–	–	7.72, s	–	–	–	–
8	7.19, s	7.23, s	7.52, s	7.23, s	7.23, s	7.18, s	7.22, s
2′	7.62, d, J = 2.3	7.64, d, J = 2.2	7.73, d, J = 2.2	7.65, d, J = 2.3	7.65, d, J = 2.3	7.62, d, J = 2.4	7.64, d, J = 2.5
3′	6.99, d, J = 2.3	6.98, d, J = 2.2	6.87, d, J = 2.2	7.03, d, J = 2.3	7.01, d, J = 2.3	7.03, d, J = 2.4	7.00, d, J = 2.5
1″	4.95, d, J = 6.8	4.62, dd, J = 4.4, 11.5, 4.48, dd, J = 7.9, 11.5	–	4.57, dd, J = 3.0, 9.6, 4.48, dd, J = 7.8, 9.6	4.50, dd, J = 3.6, 9.7, 4.43, dd, J = 9.7, 7.2	4.60, dd, J = 3.1, 9.9, 4.42, dd, J = 7.8, 9.9	4.47, dd, J = 3.0, 9.8, 4.41, dd, J = 7.8, 9.8
2″	5.75, t, J = 6.8	3.25, dd, J = 4.4, 7.9	–	3.95, dd, J = 3.0, 7.8	4.59, dd, J = 3.6, 7.2	3.97, dd, J = 3.0, 8.0	3.88, m, J = 7.8, 3.0
3″	–	–	–	–	–	–	1.98, m
CH3	1.84, s, 1.73, s	1.44, s, 1.37, s	–	1.44, s, 1.37, s	= CH2, 5.24,s, 5.11, s, CH3 1.87, s	1.30, s, 1.27, s	1.10, d, J = 7, 1.07, d, J = 7
3″ OCH3						3.30, s

Oxypeucedanin (3); 5-(2′,3′-epoxy) dimethyl allyloxy psoralen; white crystals; EI-MS m/z 286[M]⁺, 202 [M-5-subst.]⁺, 85 [5-subst.]. For ¹H-NMR data see Table 1.

Psoralen (4); pale yellow powder; EI-MS m/z 186[M]⁺, 130 [M-CO]⁺. For ¹H-NMR data see Table 1.

Oxypeucedanin hydrate (5); 5-(2,3-dihydroxy-3-methyl-butoxy) psoralen; white powder; 304 [M]⁺, 202 [bergaptol]⁺. For ¹H-NMR data, see Table 1.

Gosferol (6); Pabulenol; (5-[(2-hydroxy-3-methyl-3-butenyl)-oxylpsoralen; pale yellow powder; 286 [M]⁺, 202 [bergaptol]⁺. For ¹H-NMR data, see Table 1.

Oxypeucedanin methnolate (7); alatol; ter-O-methylprangol; 318 [M]⁺, 202 [bergaptol]⁺. For ¹H-NMR data, see Table 1.

Pranferol (8); 5-(2-hydroxy-3-methyl-butoxy) psoralen; pale yellow powder; 288 [M]⁺. For ¹H-NMR data, see Table 1.

Cell and virus culture

African green monkey kidney cells (Vero cell line CCL-81-ATCC) were grown in Eagle's minimum essential medium (EMEM) supplemented with 10% (v/v) fetal calf serum (FCS) [Gibco], 100 U mL⁻¹ penicillin (Gibco) and 100 µg mL⁻¹ streptomycin (Gibco), 2 mM l-glutamine (Gibco) and 1 mM sodium pyruvate (Gibco, Invitrogen, Carlsbad, CA).

A2780S cells (human ovarian carcinoma cell line) were maintained in RPMI-1640 supplemented with 10% (v/v) FCS and penicillin/streptomycin (50 IU mL⁻¹, 50 µg mL⁻¹) at 37 °C in a humidified atmosphere containing 5% CO₂. Cells were subcultured regularly using trypsin/ EDTA.

A virus stock of herpes simplex virus type I, strain KOS (University of Isfahan/ Iran) was prepared as follows; Vero cells infected at a low multiplicity of infection, incubated for 4 d and virus containing supernatant was harvested every day after infection until 4 d.

Cytotoxicity assay

To assess the cytotoxicity effect of compounds 1–7 on uninfected Vero cells, dilutions ranging from 2.5, 5, and 10 µg mL⁻¹ in the maintenance medium, were added to Vero monolayers (using a 96-well microplate with 4.0 × 10⁴ cells per well). The solutions were prepared by dissolving the compounds in DMSO at sub-toxic concentration (maximum of 0.019%).

Cytotoxicity of compounds 1 and 2 was also investigated on A2780S cells, and doxorubicin in a concentration of 20 µg mL⁻¹ was used as a positive control. So, 5 × 10⁴ A2780S cells were plated in 96-well plates, and grown for 24 h. The cells were exposed to a series of concentrations of chemical compounds (0.01–2.5 mM), at 37 °C for 72 h (Sadeghi-Aliabadi et al., 2012). The MTT assay was performed to evaluate in vitro cytotoxicity. At the end of incubation, 20 µL of MTT solution with the concentration of 5 mgmL⁻¹ was added and incubated for a further 3 h at 37 °C. The medium containing unreacted MTT was removed, and 150 µL of DMSO was added to each well to dissolve the formazan crystals. Finally, the absorbance of dissolved formazan was measured at 540 nm in an ELISA reader (Bio-Rad, Model 680, Hercules, CA). All the experiments were performed in triplicate.

Antiviral activity

Anti-HSV activity was investigated using a plaque-forming assay. Compounds 1–7 solutions at concentrations of 10, 5, and 2.5 µg mL⁻¹ were added to a confluent 24 h old monolayer of Vero cells grown in microtitre tissue culture plates just before virus inoculation. The cell monolayer was infected with 25 PFU of HSV1 and incubated at 37 °C for 2 h. The infected cells were washed and overlaid with medium supplemented with 2.5% methylcellulose and different concentrations of compounds. About 0.1% DMSO was used as a negative control. After 4 d, the overlay medium was removed and the cell monolayer was stained at room temperature (RT). Finally, cell monolayer was fixed with 3.7% formalin for 5 min and visible plaques were counted after staining with 1% crystal violet. The antiviral activity was determined by the following formula: The minimal concentration of extracts required to suppress the formation of virus plaque number by 50% (IC₅₀) was calculated by regression analysis of the dose–response curve generated from the data (Cheng et al., 2002).

Statistical analysis

Data collected were analyzed by using a one-way analysis of variance (ANOVA) (SPSS, version 17, SPSS Inc., Chicago, IL). Means were further classified using the Tukey as a post test. p Values of <0.05 were considered significant.

Results

Structure elucidation of pure compounds

By a combination of gravity column chromatography and repeated HPLC purifications, one prenylated coumarin (1) and seven furanocoumarins (2–8) were isolated and structures were determined based on the spectroscopic data including NMR and mass spectra, melting points and comparing to the literature. Therefore, the structures of compounds were determined as osthole (1) (Sajjadi et al., 2009), isoimperatorin (2) (Liu et al., 2004), oxypeucedanin (3) (Ivie, 1978), psoralen (4) (Lin, 2007), oxypeucedanin hydrate (5) (Ivie, 1978), gosferol (6) (Adebajo & Reisch, 2000), oxypeucedanin methnolate (7) (Abyshev et al., 1973), and pranferol (8) (Kuznetsova et al., 1966) (Figure 1).

Figure 1. Structures of compounds isolated from Prangos ferulacea.

Regarding ¹H-NMR spectra, the chemical shifts assigned to the C3, C4, C2′, and C3′ protons are characteristic of the linear furanocoumarins with the H-3/H-4 coupling constant ca. 10 and H-2′/H-3′ ca. 2 Hz (Lee & Soine, 1969). The upfield position of H-8 in contrast to H-5 aromatic proton is related to the diamagnetic shift of an aromatic proton adjacent to an oxygen atom (Lee & Soine, 1969). Gem-dimethyls of C-5 substitutes usually show different chemical shifts because of the anisotropic effect due to the 3,4 double bond (Lee & Soine, 1969). ¹H-NMR spectra showed the number of methoxy groups appearing ca. 4 ppm, the number of methoxy groups, allylic or aliphatic and existence and double-bond position.

According to the mass spectra, the elimination of C = O readily occurred in both furano and simple coumarins to render benzofuran. Besides, changing of 5-substitute with OH could get bergaptol ion at m/z 202 which usually produces a peak at m/z 174 after expulsion of C = O. Ion fragments after losing one CH₃ or side chain are common as well (Kutney et al., 1971).

Cytotoxicity and antiviral evaluation

Osthole (1) and isoimperatorin (2) were subjected to MTT assays in A2780S cells. Osthole at a concentration of 2.5 mM showed a high percentage of cytotoxicity (viability of 9.41% ± 2.4) with IC₅₀ equal to 0.38 mM (Figure 2a). The cell survival of isoimperatorin at 2.5 mM was 46.86% ± 5.5, and its IC₅₀ on A2780S cell line was 1.1 mM (Figure 2b). None of the tested compounds 1–7 could show anti-HSV effect at non-toxic concentration on Vero cells (data not shown).

Figure 2. Relative cell viability of (a) osthole and (b) isoimperatorin on the A2780S cell line. Concentrations of 0.01, 0.05, 0.1, 1, and 2.5 mM of compounds were prepared in RPMI-1640 containing 1% DMSO. The relative cell viability read for the control (the medium containing 1% DMSO) after 72 h of incubation was taken as the reference (100%). Significance was calculated by ANOVA (*p ≤ 0). Error bars show standard deviation.

Discussion

Coumarins are of chemotaxonomic value in Apiaceae family (Razavi, 2012), and isolated coumarins in this study are mostly previously reported from Prangos spp. Osthole has been isolated from Prangos spp. (Abyshev & Denisenko, 1970; Kuznetsova et al., 1979b; Sajjadi et al., 2009) and has exerted many pharmacological effects like blood triglyceride and cholesterol lowering effects via suppression of HMG-CoA reductase gene expression (Ogawa et al., 2007), antispasmodic (Sadraei et al., 2012, 2013), and anti-osteoporosis (Li et al., 2002).

Isoimperatorin has been previously isolated from Prangos spp. (Danchul et al., 1979; Kuznetsova et al., 1979b), and has shown antitumor necrosis factor and antioxidant effects (Zhou et al., 2011). Oxypeucedanin from Prangos spp. (Kuznetsova et al., 1979b; Razavi et al., 2008) has shown anti-tumor effects (Kim et al., 2007). Psoralen is a widely distributed coumarin also present in Prangos species (Kuznetsova et al., 1979a) and is mostly used in the treatment of psoriasis, Lichen Planus and vitiligo (Morison, 1999). Oxypeucedanin hydrate has been isolated from Prangos lamellata Korovin, Prangos tschimganica B. Fedtsch, Prangos acris-romanae Boiss, Prangos lophoptera Boiss., and Prangos acualis (Danchul et al., 1979; Kuznetsova et al., 1979a,b) and showed anti-HIV effects (Shikishima et al., 2001). Gosferol has been isolated from P. lophoptera (Abyshev, 1974) and showed antiplatelet (Xiao et al., 2007) effects. Oxypeucedanin methnolate has been isolated from Ferula sumbul and showed anti-HIV effects (Zhou et al., 2000). Pranferol has been isolated from P. acris-romanae (Kuznetsova et al., 1979b).

Besides furanocoumarins, a coumarin glycoside called celereoside has been previously isolated from P. ferulacea (Razavi, 2012).

Phenolics have exerted notable antiviral effects. Coumarins have shown anti-HIV effects through inhibiting reverse transcriptase, integrase and protease enzymes, in which chirality, changing aliphatic substitutes and rings orientation affect the efficacy (Domagala et al., 1996; Kostova et al., 2006). Besides, anti-influenza (Hsieh et al., 2012), anti-H1N1 (Lee et al., 2009) and anti-HBV (Hsieh et al., 2012) effects have been reported from different coumarins. For example, oxypeucedanin hydrate and oxypeucedanin methnolate have shown anti-HIV effects (Shikishima et al., 2001; Zhou et al., 2000).

Anti-HSV activity has been reported from phenolics such as flavonoids, stilbenoids (Likhitwitayawuid et al., 2005), tannins (Yarnell & Abascal, 2005), cinnamic acid derivatives (Chattopadhyay et al., 2008), phenyl propanoids (Sajjadi et al., 2012a,b), and sesquiterpene coumarins (Ghannadi et al., 2014). Resveratrol, a stilbenoid, inhibited HSV-1 and HSV-2 replication by suppressing NF-kappa B activation, an essential step of its lifecycle during infection (Chattopadhyay et al., 2008). The SAR studies revealed that the site(s) and the number of OH groups on phenols are responsible for their antiviral activity (Chattopadhyay et al., 2008). However, there is not any structure activity relationship defined for coumarins that would explain the null effect of tested compounds on HSV-1. Although angular furanocoumarins showed anti-HSV effects (Hsieh et al., 2012), none of the tested linear furanocoumarins (2–7) was as effective as virucidal in sub-toxic concentration. So, we propose that angularity could be engaged in their anti-HSV effect. In addition, the non-toxic concentration of some of the compounds on normal Vero cells was very high making them not good candidates as antiviral agents.

Cytotoxicity of osthole and isoimperatorin was investigated with the MTT assay (Figure 2) on the A2780S cell line. Osthole with IC₅₀ equal to 0.54 mM has shown a cytotoxic effect on this cell line, a finding which is in agreement with other studies which clarified that osthole is toxic on breast cancer cell lines, e.g., MDA-MB 435 and MCF-7 (Wanga et al., 2012; You et al., 2010), and cervix cancer cell line, e.g., HeLa (Chou et al., 2007). Interestingly, osthole is not toxic on normal cells and its effect is specific for cancerous cells. Previous studies have shown that osthole was not toxic on HEK-293 (You et al., 2010), and our results also indicated no toxicity on normal Vero cell line by osthole. Comparing with osthole, isoimperatorin was less toxic on the A2780S cell line (Figure 2), and even in 1 mM, it could not kill more than 50% of cells. Isoimperatorin is one of the well-known cyclooxygenase-2 (COX-2) inhibitors (Moon et al., 2008), while A2780S is an ovarian cell line which does not express COX-2. Different mechanisms have been identified for anticancer activity of COX-2 inhibitors including induction of apoptosis, inhibition of tumor vascularization, stimulation of anti-tumor immune responses, and inhibition of cellular protein synthesis (Sadeghi-Aliabadi et al., 2012). Probably, COX-2 inhibition is the main cytotoxic mechanism of isoimperatorin, and lack of COX-2 in A2780S caused less toxicity on this cell line.

Conclusion

Eight coumarins (osthole, isoimperatorin, oxypeucedanin, psoralen, oxypeucedanin hydrate, gosferol, oxypeucedanin methnolate, and pranferol) have been isolated from P. ferulacea roots, some in large amounts. Therefore, P. ferulacea could be a good source of coumarins. The first seven coumarins did not show anti-HSV effect at non-toxic concentrations. Osthole has shown a moderate cytotoxicity effect on the A2780S cancerous cell line while no cytotoxicity was observed on normal Vero cells. Isoimperatorin was toxic neither on cancerous cells nor normal cells.

Declaration of interest

The authors report no declarations of interest. Financial support on behalf of Isfahan and Kermanshah Universities of Medical Sciences is acknowledged.