Introduction

The genus Heracleum L. (Apiaceae) comprises about 125 aromatic taxa predominantly distributed in the temperate Northern Hemisphere, mainly in Eurasia (Pimenov and Leonov 2004). Heracleum pyrenaicum Lam. is edible wild plant that is often mentioned in survival handbooks in Serbia. The different parts of this plant are consumed as food. The young leaves are used to prepare salad, soup, stew and mash, their petioles are eaten after peeling and marinating, young stems are consumed fried, while roots, which contain significant amounts of sugar and starch, are used to make syrup and flour (Rašić 2002; Vračarić et al. 1977). The roots and aerial parts of this plant are also applied in folk medicine for the curing of intestinal catarrh and as aperitifs and antidiarrheal drugs (Rašić 2002). Similarly, in many other countries, several Heracleum taxa are used as food or traditional herbal medicines. For example, in India, H. candicans Wall. ex DC. fruits are added to food as spice and are known to be agents for the treatment of indigestion, sexual weakness and “nervous” disorders (Chauhan et al. 2014). In Iran, the fruits of H. persicum Desf. are used as carminative and spice for food, as well as pain killer herbal drug, and the young stems of this plant are consumed pickled (Hajhashemi et al. 2009).

The chemical composition and biological activities of Heracleum essential oils have been intensively investigated in the last decade (Chauhan et al. 2014; Hajhashemi et al. 2009; Karuppusamy and Muthuraja 2011; Tkachenko 2009). Heracleum pyrenaicum subsp. orsinii (Guss.) F. Pedrotti and Pignatti (Pignatti 1982) is a perennial plant, with simple, 5–7 lobed leaves and greenish flowers. This taxon inhabits the limestone screes of the mountains of the Balkan Peninsula, as well as Central and South Apennines (Brummitt 1968; Pignatti 1982). It belongs to widely circumscribed H. sphondylium L. group (Tonascia 1992). In this work, the chemical composition of H. pyrenaicum subsp. orsinii root, leaf and fruit essential oils, their antimicrobial activity against standard strains, clinical and food isolates, as well as cytotoxic activity on malignant and normal cell lines were investigated for the first time.

Materials and methods

Plant material

Plant material was collected on Mt. Durmitor in the north-western part of Montenegro. The fruits were collected in August 2011, and the roots and leaves in August 2013. Voucher specimen is deposited in the Herbarium of the Natural History Museum, Belgrade (BEO) under collector number 20110804/BEO. The plant was identified by Dr. Marjan Niketić, curator/botanist of the BEO.

Isolation of the essential oils

Air-dried material was powdered (roots and fruits) or crashed (leaves) and hydrodistilled using Clevenger-type apparatus for 2.5 h. Collecting solvent was n-hexane. The oils were dried over anhydrous sodium sulfate and kept at 4 °C until analysis.

Essential oils analysis

The chemical composition of isolated oils was analyzed by Gas chromatography (GC) and Gas chromatography-Mass spectrometry (GC–MS). GC analysis was carried out using an Agilent 6890N gas chromatograph (Agilent Technologies, USA), equipped with a split/splitless injector (200 °C), attached to a HP-5MS capillary column (Agilent Technologies; 30 m × 0.25 mm; film thickness 0.25 μm) and connected to a flame ionization detector (FID). The FID and transfer line temperatures were set at 300 and 250 °C, respectively. Split ratio was 1:10 and the injected volume was 1 μL of 3% solution of oil in 99.9% (v/v) ethanol. The carrier gas was He (1.0 mL/min). The thermal program was set from 60 to 280 °C at a rate of 3 °C/min. GC–MS analysis was performed on an Agilent 6890–5975 GC–MS system, operating in the electron ionization (EI) mode at 70 eV, equipped with a split/splitless injector (200 °C) and attached to a HP-5MS capillary column (30 m × 0.25 mm; film thickness 0.25 μm). The analytical conditions were the same as that used for the GC analysis. The identification of the compounds was based on the comparison of their retention indices (RI), retention times (Rt) and mass spectra to those from the NIST/NBS 05, Wiley libraries 8th edition and the literature (Adams 2007; Nitz et al. 1990). The linear RIs were determined in relation to homologue series of n-alkanes (C8–C40) ran under the same operating conditions. Relative percentages of the compounds were calculated based on the peak areas from the FID data.

Antimicrobial activity

Microbial strains

The Gram-positive bacteria Staphylococcus aureus (ATCC 6538), Bacillus cereus (clinical isolate), Listeria monocytogenes (NCTC 7973) and Micrococcus flavus (ATCC 10240), and the Gram-negative bacteria Pseudomonas aeruginosa (ATCC 27853), Escherichia coli (ATCC 35210), Salmonella typhimurium (ATCC 13311) and Enterobacter cloacae (human isolate) were used. The fungi Aspergillus fumigatus (human isolate), A. versicolor (ATCC 11730), A. ochraceus (ATCC 12066), A. niger (ATCC 6275), Trichoderma viride (IAM 5061), Penicillium funiculosum (ATCC 36839), P. ochrochloron (ATCC 9112) and P. verrucosum var. cyclopium (food isolate) were tested. The micromycetes were maintained on malt agar, the cultures stored at 4 °C and sub-cultured once a month.

Antibacterial activity

Minimum inhibitory and minimum bactericidal concentrations (MICs and MBCs) were determined by the microdilution method in 96-well microtitre plates (CLSI 2009; Tsukatani et al. 2012). Bacterial suspensions were adjusted with sterile saline to a concentration of 1.00 × 105 CFU/mL. The oils were dissolved in 5% dimethylsulfoxide (DMSO) solution that contained 0.10% Tween 80 (v/v) (10 mg/mL) and added to Tryptic Soy broth (TSB) medium (100 µL) with bacterial inoculum (1.00 × 104 CFU per well), to achieve concentrations from 0.21 to 8.30 mg/mL. The MICs were defined as the lowest concentrations without visible bacterial growth (determined at binocular microscope). The MICs were also determined by the colorimetric microbial viability assay that is based on the reduction of p-iodonitrotetrazolium violet (INT) color. Results were compared to the positive controls. The MBCs were determined by serial sub-cultivations of 2 µL of tested oils (dissolved in medium and inoculated for 24 h) into microtitre plates that contained 100 µL of broth per well, after further incubation for 24 h. The lowest concentration without visible bacterial growth was defined as the MBC, indicating that 99.5% of the original inoculum was killed. The optical density of each well was measured by Microplate manager 4.0 (Bio-Rad Laboratories, USA) at the wavelength of 655 nm and compared to the blank and positive controls. Streptomycin, Sigma P 7794 (0.04–0.52 mg/mL) and ampicillin, Panfarma, Serbia (0.25–1.24 mg/mL) were used as the positive controls. 5% DMSO was used as the negative control.

Antifungal activity

In order to investigate the antifungal activity of the oils, modified microdilution technique was used (Espinel-Ingroff 2001; Hänel and Raether 1988). Fungal spores were washed off from the surface of agar plates with 0.85% sterile saline that contained 0.10% Tween 80 (v/v). Spore suspensions were adjusted with sterile saline to a concentration of 1.00 × 105 in the final volume of 100 µL per well. The oils were dissolved in 5% DMSO solution that contained 0.10% Tween 80 (v/v) (10 mg/mL) and added to broth Malt medium with the inoculum (to achieve concentrations 0.12–8.26 mg/mL). The lowest concentrations without visible growth (at the binocular microscope) were defined as MICs. The minimum fungicidal concentrations (MFCs) were determined by serial sub-cultivations of a 2 µL of the tested oils (dissolved in medium and inoculated for 72 h) into microtitre plates that contained 100 µL of broth per well, after further incubation for 72 h at 28 °C. The MFC was defined as the lowest concentration without visible growth, indicating that 99.5% of the original inoculum was killed. Commercial fungicides bifonazole, Srbolek, Serbia (0.10–0.25 mg/mL) and ketoconazole, Zorkapharma, Serbia (0.20–3.50 mg/mL) were used as the positive controls. 5% DMSO was used as the negative control.

Statistical analysis

All the tests were carried out in triplicate. The results were expressed as mean values ± standard deviation (SD), and analyzed by one-way analysis of variance (ANOVA), followed by Tukey’s HSD test with α = 0.05, to determine whether there is a statistically significant difference between them. The analysis was carried out by Statistical Package for the Social Sciences (SPSS) version 18.0.

Cytotoxic activity

Cell cultures

Cervix adenocarcinoma HeLa, human colon carcinoma LS174, non-small cell lung carcinoma A549, as well as human normal fetal lung fibroblast MRC-5 cell lines (ATCC) were cultured as a monolayer in the RPMI 1640 nutrient medium, supplemented with heat inactivated (at 56 °C) 10% fetal bovine serum (FBS), 3 mmol/L of l-glutamine and antibiotics, at 37 °C, in a humidified air atmosphere with 5% CO2.

Treatment of cell lines

In vitro assay for the cytotoxic activity of the oils was performed when the cells reached 70–80% of confluence. The stock solution (100 mg/mL) of each oil was dissolved in RPMI 1640 medium to obtain required concentrations. Neoplastic HeLa (2000 cells per well), LS174 (7000 cells per well), A549 (5000 cells per well) and normal MRC-5 cells (5000 cells per well) were seeded into 96-well microtitre plates and 24 h later, after the cell adhesion, five different, double diluted concentrations of the oils were added to the wells. The final concentrations of the oils were 12.5, 25, 50, 100 and 200 µg/mL. Control wells contained only nutrient medium that was made of RPMI 1640 medium, supplemented with 3 mmol/L l-glutamine, 100 mg/mL streptomycin, 100 IU/mL penicillin, 10% heat inactivated (56 °C) FBS and 25 mmol/L HEPES (2-[4-(2-hydroxyethyl)piperazinyl] ethanesulfonic acid). The pH of the medium was adjusted to 7.2 with bicarbonate solution. The cultures were incubated for 72 h.

Determination of cell survival (MTT test)

The effect of the oils on cell survival was determined by the MTT test (microculture tetrazolium test), according to Mosmann (1983) with modification by Ohno and Abe (1991), 72 h after the addition of the oils. Briefly, 20 µL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution (5 mg/mL phosphate-buffered saline, PBS) was added to each well. The samples were incubated for further 4 h, at 37 °C, in 5% CO2 humidified air atmosphere. During this period MTT dye was converted to insoluble product, formazan by viable cells. This precipitate was then dissolved by adding 100 µL of 10% sodium dodecylsulfate (SDS). The number of viable cells in each well was proportional to the intensity of the light absorbance (A) that was measured 24 h later by an ELISA plate reader (Thermo Fisher Scientific Inc., Australia) at 570 nm. To calculate cell survival (%), the A of a sample with cells grown in the presence of various concentrations of the oils were divided with control optical density (the A of control cells grown only in nutrient medium) and multiplied by 100. The A of the blank was always subtracted from the A of the corresponding sample with target cells. The IC50 value was defined as the concentration of an agent that inhibits the survival of 50% cells, compared to the vehicle treated control. Cisplatin was used as the positive control. The IC50 values were expressed as mean values ± SD that were determined on the basis of the results of three independent experiments.

Results and discussion

Chemical composition of the essential oils

The hydrodistillation of H. pyrenaicum subsp. orsinii roots, leaves and fruits yielded 0.13, 0.20 and 0.61% (w/w) of the yellow essential oils, respectively. Seventy-five components were identified in the root oil, eighty-nine in the leaf oil and seventy-eight in the fruit oil (representing 91.3, 94.2 and 90.2% of the total oils, respectively) (Table 1).

Table 1 Chemical composition of H. pyrenaicum subsp. orsinii essential oils (%)
Full size table

The dominant monoterpene fraction of the root oil (70.2%) was characterized by non-oxygenated components (67.5%), with β-pinene being the most abundant (38.6%). Similarly, Tkachenko (2009) reported the prevalence of monoterpenes, mostly β-pinene (17.6–39.0%), in the root oils of seven other Heracleum taxa grown at experimental station of V. L. Komarov Botanical Institute in Leningrad Oblast′ (Russia), with the highest amounts being present in the root oils of Caucasian species, H. wilhelmsii Fisch. and C. A. Mey. and H. ponticum (Lipsky) Schischk. ex Grossh. The monoterpene fraction of H. pyrenaicum subsp. orsinii leaf oil was less abundant (20.3%), but qualitatively very similar to the monoterpene fraction of the root oil. The leaf oil was characterized by the high percentage of sesquiterpenes (65.4%), and among them oxygenated (33.6%) and non-oxygenated (31.8%) constituents were present in similar amounts, with (E)-nerolidol (20.5%) and (E)-caryophyllene (17.0%) being the most prominent. In the oil obtained by microdissection of widespread H. sphondylium L. subsp. sphondylium leaf companion canals, sesquiterpenes were also dominant; (E)-caryophyllene (28.0%) prevailed, while (E)-nerolidol was not identified (Bicchi et al. 1990). In contrast to H. pyrenaicum subsp. orsinii root and leaf oils, the fruit oil of this taxon contained significantly lower quantity of terpenes. This oil was mainly composed of aliphatic esters (78.7%), with octyl acetate (36.8%) and octyl hexanoate (22.1%) being the principal constituents. This is in accordance with the conclusion of Başer (2002) that aliphatic esters, e.g. octyl esters, can be considered as marker compounds of Heracleum fruit oils. For example, octyl acetate (65.3%) also dominated in the oil of H. siamicum Craib fruits, which are used as spice in Thailand (Kuljanabhagavad et al. 2011).

Antimicrobial activity of the essential oils

The antimicrobial activity of H. pyrenaicum subsp. orsinii essential oils was determined by microdilution method, and expressed as minimal inhibitory concentrations (MICs) and minimal bactericidal/fungicidal concentrations (MBCs/MFCs) (Tables 2, 3).

Table 2 Antibacterial activity of H. pyrenaicum subsp. orsinii essential oils and antibiotics (mg/mL)
Full size table
Table 3 Antifungal activity of H. pyrenaicum subsp. orsinii essential oils and antibiotics (mg/mL)
Full size table

The leaf essential oil exhibited the best antibacterial activity against the clinical isolate of B. cereus (MIC = 0.21 mg/mL, MBC = 0.53 mg/mL). This effect was comparable to the effect of ampicillin. The best activity against all other tested bacteria was shown by the root oil. The effect of this oil against S. typhimurium, E. coli and P. aeruginosa (MICs = 0.23 mg/mL, MBCs = 0.47 mg/mL) was comparable with the activity of streptomycin and even better than the effect of ampicillin, while the effect against Staphylococcus aureus (MIC = 0.23 mg/mL, MBC = 0.47 mg/mL) was similar to the activity of ampicillin. These bacteria cause a wide variety of diseases. For example, B. cereus, S. typhimurium, E. coli and S. aureus are the source of foodborne diseases. Moreover, E. coli causes urinary infections, while S. aureus causes respiratory, urinary, skin and eye infections. They are both the significant source of hospital-acquired infections. Similarly, P. aeruginosa causes nosocomial respiratory, urinary and wound infections (Pommerville 2011).

The most significant antifungal activity was exhibited by H. pyrenaicum subsp. orsinii leaf (MIC = 0.12 mg/mL, MFC = 0.25 mg/mL) and root (MIC = 0.23 mg/mL, MFC = 0.46 mg/mL) oils against T. viride, and the root oil (MIC = 0.46 mg/mL, MFC = 1.88 mg/mL) against Aspergillus ochraceus. These effects were more pronounced than the effects of ketoconazole. Additionally, the activity of the leaf oil against T. viride was comparable with the activity of bifonazole. Although the members of Trichoderma genus rarely infect humans, some species can cause infections in immunocompromised patients (De Miguel et al. 2005). Aspergillus ochraceus, on the other hand is a food contaminant that produces nephrotoxic, hepatotoxic, teratogenic and immunosuppressive ochratoxin A (Basílico and Basílico 1999).

The effect comparable with the effect of ketoconazole and slightly weaker than the effect of bifonazole was shown by H. pyrenaicum subsp. orsinii root oil (MIC = 0.23 mg/mL, MFC = 0.46 mg/mL) against the human isolate of A. fumigatus. This result is particularly interesting because A. fumigatus is an airborne pathogen that causes a usually fatal invasive aspergillosis in immunosuppressed hosts (Pommerville 2011).

Among analyzed essential oils, the fruit oil had the weakest antimicrobial activity. This is in accordance with its chemical composition, i.e. the fruit oil was dominated by aliphatic esters, compounds with lower antimicrobial potential than terpenic constituents identified in the root and leaf oils (Maggi et al. 2014). Demonstrated activity of the root and leaf oils can be at least partly explained by the presence of terpenes, antimicrobial potential of which was established previously. Among them are not only the major components of these oils [namely β-pinene in the root oil, and (E)-nerolidol and (E)-caryophyllene in the leaf oil], but also some of their minor compounds (such as α-pinene, caryophyllene oxide, limonene, germacrene D and α-humulene) (Setzer et al. 2006; Soković et al. 2010; Tao et al. 2013). Stronger antimicrobial activity of the root oil in contrast to the leaf oil could be justified by its different chemical composition and appropriate synergism between its components—a phenomenon described previously for different essential oils (Bakkali et al. 2008; Burt 2004).

Cytotoxic activity of the essential oils

The cytotoxic activity of isolated H. pyrenaicum subsp. orsinii essential oils was determined by MTT test and expressed as the concentrations of the oils that inhibited the growth of 50% cells (IC50). Analyzed oils showed significant effect against all the tested malignant cells (Table 4): human malignant cervix adenocarcinoma HeLa, colon carcinoma LS174 and non-small cell lung carcinoma A549 cells (IC50 = 6.49–14.56 μg/mL), satisfying the criterion of the National Cancer Institute (NCI) for cytotoxicity (IC50 < 30.00 μg/mL) (Suffness and Pezzuto 1991). The strongest effect was exhibited by the fruit oil against A549 cells. The effect of cisplatin, used as positive control, was more prominent (IC50 = 0.84–4.16 μg/mL). However, cisplatin also exhibited strong toxicity against human normal fetal lung fibroblast MRC-5 cells (IC50 = 15.22 μg/mL), in contrast to all the investigated H. pyrenaicum subsp. orsinii oils, which were not toxic against these normal cells at tested concentration (IC50 > 200.00 μg/mL).

Table 4 Cytotoxic activity of H. pyrenaicum subsp. orsinii essential oils and cisplatin (μg/mL)
Full size table

As in the case of antimicrobial activity, exhibited cytotoxic effect of analyzed H. pyrenaicum subsp. orsinii essential oils can be at least partly explained by the presence of some dominant and minor components. For example, cytotoxicity of β-pinene, (E)-nerolidol, (E)-caryophyllene, α-pinene, caryophyllene oxide, n-octanol and α-humulene against the some of the malignant cells used in the present study was previously shown (Bourgou et al. 2010; Da Silva et al. 2007; Jun et al. 2011; Kubo and Morimitsu 1995).

Conclusion

The present work reveals H. pyrenaicum subsp. orsinii root, leaf and fruit essential oils composition for the first time, and thus contributes to a better knowledge of the volatile constituents of Heracleum taxa. Demonstrated biological activity represents a good starting point for further investigations of these oils as potential new herbal raw materials. Namely, against sixteen pathogens including food contaminants, the oils exhibited antimicrobial activity, which was in some cases comparable or even better than the activity of the standard antibiotics. Also, all the oils showed significant cytotoxic activity that was weaker but more selective than the effect of cisplatin.