Volatile Organic Compounds from Centaurium erythraea Rafn (Croatia) and the Antimicrobial Potential of Its Essential Oil

GC and MS were used for the analysis of Croatian Centaurium erythraea Rafn essential oil (obtained by hydrodistillation) and headspace (applying headspace solid-phase microextraction). The headspace contained numerous monoterpene hydrocarbons (the major ones were terpinene-4-ol, methone, p-cymene, γ-terpinene and limonene). Oxygenated monoterpenes were present in the headspace and oil, while 1,8-cineole, bornyl acetate and verbenone were present only in the headspace. High headspace percentages of toluene and naphthalene were found, followed by hemimellitene. Lot of similarities were observed with Serbian C.erythraea oil [neophytadiene (1.4%), thymol (2.6%), carvacrol (6.1%) and hexadecanoic acid (5.7%)], but different features were also noted such as the presence of menthol, menthone and phytone. The oil fractionation enabled identification of other minor compounds not found in total oil such as norisoprenoides, alk-1-enes or chromolaenin. The essential oil showed antimicrobial potential on Escherichia coli, Salmonella enteritidis, Staphylococcus aureus and Bacillus cereus. On the other hand, no antibacterial activity of the oil was observed on Pseudomonas fluorescens and Lysteria monocytogenes.


The Headspace C. erythraea VOC Composition
Two fibers (PDMS/DVB and DVB/CAR/PDMS) were selected for HS-SPME after preliminary research with respect to overall number of isolated compounds. Both fibers showed qualitatively similar chemical profiles of the extracted compounds, but individual compound percentages varied (Table 1). A total of 52 VOCs were identified and reported for the first time in C. erythraea headspace. RI = retention indices on HP-5MS and HP-FFAP columns; A = solvent-free HS-SPME with the fiber PDMS/DVB; B = solvent-free HS-SPME with the fiber DVB/CARPDMS; -= compound not found; / = compound not identified on the column; * -correct isomer not identified.
Aliphatic and aromatic hydrocarbons were abundant, particularly toluene (4.5%; 18.0%) and naphthalene (8.2%; 1.8%) followed by o-and m-ethyltoluene (0.8-1.7%). The trimethylbenzenes pseudocumene (0.9%; 1.0%) and hemimellitene (1.8%; 2.5%) were also present, but were not found in the oil composition (Table 2). Although benzene derivatives are usually considered as compounds of possible anthropogenic origin, new results indicate possible other pathways of their natural biogenesis. Namely, the emission of toluene from different plants was observed, although no biochemical pathway for toluene production is known [11]. Trace naphthalene amounts are produced by magnolias and flower extracts from gynoecia of five taxa (Magnolia denudata, Magnolia liliiflora, Magnolia tomentosa, Magnolia praecocissima var. praecocissima and var. borealis) contained naphthalene as main component [12]. Furthermore, Formosan subterranean termite [13] and some strains of the fungus Muscodor albus naturally produce naphthalene, while Muscodor vitigenus produces naphthalene almost exclusively [14]. Therefore natural origin of the benzene derivatives found in C. erythraea oil could be similar, and they can be excluded as pollutants since the plant was collected from ecologically pure area. Aliphatic hydrocarbons and carbonyls up to C18 were present as minor constituents (probably originated from fatty acids catabolism) and those up to C6 were only found in C. erythraea headspace (most likely due to high volatility and solvent delay applied for the oil GC analysis).   RI = retention indices on HP-5MS and HP-FFAP columns; C = total essential oil; D = non-polar fraction of the oil; E = polar fraction of the oil; -= compound not found; / = compound not identified on the column; * -correct isomer not identified.

C. erythraea Essential Oil Composition
A total of 89 compounds were identified in the essential oil of C. erythraea Rafn (Table 2). In comparison with the sole previous report on this oil from Serbia [10], the overall number of identified compounds seems moderate, but it should be emphasized that the previously published oil composition predominated (ca. 50%) with the compounds in traces (<0.05%). This research was focused on detailed determination of non-trace compounds of the oil including the results of the oil fractionation to non-polar and polar fractions.
Total essential oil contained a low abundance of oxygenated monoterpenes, the major ones being menthol (7.0%), linalool (3.0%), borneol (1.4%) and methone (2.5%). Menthol and menthone were not found in Serbian oil (only traces of isomenthol were detected) as well as β-thujone (0.8%). Borneol (1.4%) and camphor (1.5%) were identified in Croatian oil, while only traces were reported in Serbian oil. Monoterpene phenols thymol (2.6%) and carvacrol (6.1%) were interesting features also reported among the major constituents in Serbian oil (thymol 7.9% and carvacrol 4.2%). Although these phenols were present in the oil, they were not identified in the headspace. On the contrary, their biosynthetic precursors γ-terpinene and p-cymene were found in the headspace, but not in the oil.
Neophytadiene (10.1%) was major compound of Serbian oil, while in this research it amounted to 1.4%. Neophytadiene is presumably a chlorophyll metabolite [15]. Chlorophyll chlorin rings can be constructed from several different side chains, usually including the long diterpene alcohol phytol (found in the oil at 1.9%). Isophytol (0.3% in the oil) is a phytol isomer and isomerisation of isophytol to phytol in plant leaf waxes is reported [16]. Oxidation of the phytol moiety of chlorophyll could lead, among others, to the methylated long chain fatty acid ketone hexahydroxyfarnesyl acetone (6,10,14trimethylpentadecan-2-one, phytone) that was found in the oil at 4.0%. The fourth major compound of Serbian oil was hexadecanoic acid (4.9%), while Croatian oil contained 5.7%, followed by linoleic acid (3.9%) and tetradecanoic acid (1.5%). High-molecular aliphatic hydrocarbons were present, particularly tricosane (6.8%). Their abundance was higher in comparison with Serbian oil. Sesquiterpenes were also found with minor percentages; the major ones were caryophyllene oxide (0.9%) and δ-cadinene (0.8%). As was expected, high-molecular compounds of the essential oil were not identified in the headspace due to low volatility.
Silicagel microcolumn fractionation enabled a more detailed analysis by the oil separation into non-polar and polar compounds. As was expected from total oil composition, the fraction of non-polar compounds (Figure 1a) dominated by higher aliphatic hydrocarbons such as tricosane (25.7%), heneicosane (8.7%), nonadecane (4.6%) and others. However, the pentane fraction revealed the presence of low-molecular aliphatic hydrocarbons (overlapped in total oil), and nonane (8.6%) was the major one. In addition, alk-1-enes were only found in this fraction, including hexadec-1-ene (1.7%), octadec-1-ene (2.6%) and tetradec-1-ene (0.5%) as well as eicos-3-ene (2.3%). According to their chain length and positions of double bonds they are derived from corresponding fatty acids [17]. It is possible that some of the hydrocarbons found were hydrodistillation artifacts. The plants were collected from an ecologically pure area and the possibility of contamination is thus excluded. Natural existence of hydrocarbons in plants is also known. Many higher members of the n-alkanes, n-alkan-1ols, n-alkanals, n-alkanoic acids and n-alkyl esters were identified in cuticular waxes [18].
The variety of identified sesquiterpene hydrocarbons in the non-polar fraction was mainly composed of δ-cadinene (1.9%), β-caryophyllene (1.1%) and trans-β-farnesene (1.0%). Their percentages were higher in comparison with the oil, and several were only identified in this fraction, and not in the oil. The rather unusual monooxygenated compound chromolaenin (2.1%), with a cadinane structure and previously reported in the essential oils of Baccharis salicifolia, B. latifolia and B. dracunculifolia [19], was also only detected in the pentane fraction.

Unlocking Antimicrobial Potential of the Essential Oil
The essential oil of C. erythraea showed different antimicrobial potential towards the bacterial species tested. In general, Gram-negative bacteria (E. coli and S. enteritidis) were the most sensitive species to the oil, while there was no effect to P. fluorescens growth. Ampicilin (10 μg), norfloxacin (10 μg), ofloxacin (5 μg) and tetracycline (30.5 μg) were used as positive reference standards to determine the sensitivity of bacterial strains tested. Although applied antibiotics possessed in general strongest antibacterial activity, the activity of C. erythraea to E. coli was almost identical to the inhibition zones of ampicilin (Table 3). Antibiotics are molecules with selective activity towards the cells of microorganisms while essential oils are compounds with many different and, often, variable components that attack different parts of cell structures. It is extremely hard to compare zones of inhibition produced by antibiotics and bioactive compounds since they differ in nature and mechanism of the inhibition. Nevertheless, antibiotics are used as controls in any compound testing to ensure information about sensitivity of tested strains and internal control. Observed higher antimicrobial activity is presumably related to the oil principal constituents, although part of the activity could result from the presence of oil minor constituents. In general, the most active essential oils (principally composed of carvacrol, thymol, citral, eugenol and their precursors [20][21][22]) against the strains of E. coli are: the oil of oregano (Origanum vulgare), thyme (Thymus vulgaris), bay (Pimenta racemosa) and clove (Eugenia caryophyllata). The oil of C. erythraea contained thymol and carvacrol as major constituents and exhibited noticed activity against E. coli. Menthol is abundant in C. erythraea oil and the MICs demonstrated [23] that menthol is more toxic against E. coli than thymol. The antimicrobial activity mechanism of thymol, carvacrol and menthol is well known (alteration of membrane permeability [23]). In addition, strong antimicrobial activity of basil (Ocimum basilicum) oil (major constituents were linalool and 1,8-cineole) against S. enteritidis is known [24] and linalool was present in C. erythraea oil that exhibited activity against S. enteritidis. However, there is evidence that total essential oil is more strongly antimicrobial than is accounted for the additive effect of their major antimicrobial components; minor components appear, therefore, to play a significant role [25]. Gram-positive bacteria demonstrated increased resistance to C. erythraea oil, compared to Gram negative ones. To L. monocytogenes, the essential oil exhibited no inhibitory effect, while weak inhibitory zones (7 and 8 mm, respectively), were observed to B. cereus and S. aureus growth.
Compared to all tested species, E. coli and S. enteritidis were the most sensitive bacteria to C. erythraea essential oil. Of the Gram-negative bacteria, Pseudomonads, in particular P. aeruginosa and P. fluorescens, appear to be least sensitive to the action of essential oils (Table 3) [26,27].
The results (Table 3) are in contrast with previous reports indicating that Gram-positive bacteria are more susceptible to the essential oils than Gram-negative ones [28]. However, this contrast was already reported for the essential oils of three Greek Achillea species [29] and antimicrobial properties of the main constituents showed that caryophyllene oxide was the most efficient, followed by camphor and 1,8-cineole. The oil of C. erythraea contained caryophyllene oxide and camphor among minor constituents.
Kirbağ et al. [9] demonstrated activity of C. erythraea infusion to several bacterial species, where Bacillus megaterium exhibited highest sensitivity (13 mm) while E. coli and S. aureus showed no antibacterial activity. However, comparison with the results in Table 3 is not possible since different chemical classes of natural compounds were isolated by the plant infusion (mainly non-volatile compounds) and further tested on antibacterial activity.

Plant Material, Solvents and Isolation of the Essential Oil
The aerial parts (flower, leaves and steam) of C. erythraea Rafn were collected in the Podravina area (Croatia) near Đurđevac and Koprivnica (voucher specimen number RH-0078). Air-dried aerial parts of C. erythraea were subjected to hydrodistillation for 3 h, using a Clevenger-type apparatus [30] to produce a pale yellow highly fragrant essential oil. The obtained oil was separated, dried over anhydrous sodium sulfate and stored at 4 °C until the analysis. The yield was calculated based on dry weight of the sample. The solvents pentane and diethyl ether were purchased from Kemika (HR-Zagreb) and were distilled before usage.

Microcolumn Oil Fractionation
The essential oil (30 μL) was fractionated on a silica gel column (4 g; 30-60 mm) and two fractions were obtained. Pentane was used for the elution of non-polar compounds and diethyl ether for the elution of polar compounds. The separation was monitored by thin layer chromatography using Kieselgel 60 aluminum-backed sheets (Merck). The obtained fractions were concentrated by fractional distillation and analysed.

Headspace Solid-Phase Microextraction (HS-SPME)
Two fibres (PDMS/DVB and DVB/CAR/PDMS) suitable for the extracting compounds with relatively wide range of polarities and volatilities were selected for HS-SPME after preliminary research. Other operating parameters for HS-SPME (such as extraction time and temperature) were also determined in preliminary research with respect to overall number of isolated compounds. The isolation of headspace volatiles was performed using manual SPME fibers with the layer of polydimethylsiloxane/divinylbenzene (PDMS/DVB) and divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS) obtained from Supelco Co (Bellefonte, PA, USA). The fibers were conditioned prior to use according to the manufacturer instructions. For HS-SPME extraction, 1 g of grounded plant material was placed in a 15 mL glass vial and hermetically sealed with PTFE/silicone septa. The vial was maintained in a water bath at 60 °C during equilibration (15 min) and extraction (45 min) and was partially submerged so that the plant material was below the water level. After sampling, the SPME fiber was withdrawn into the needle, removed from the vial, and inserted into the injector (250 °C) of the GC and GC-MS for 6 min where the extracted volatiles were thermally desorbed directly to the GC column [31].

Gas Chromatography and Mass Spectrometry
Gas chromatographic analysis was performed on an Agilent 7890 instrument (Agilent Technologies, Palo Alto, CA, USA) equipped with a flame ionization detector and an HP-5MS capillary column ((5%-phenyl)-methylpolysiloxane, Agilent J & W GC column, 30 m × 0.25 mm × 0.25 μm) or HP-FFAP column (nitroterephthalic acid modified polyethylene glycol, Agilent J &WGC column, 50 m × 0.32 mm × 0.50 μm). The compounds were identified on an Agilent Technologies 5975C mass spectrometer (MS conditions were: ionization voltage 70 eV; ion source temperature 280 °C; mass scan range: 30-300 mass units). The GC settings were as follows: the column HP-5MS (the initial oven temperature was held at 70 °C for 2 min and ramped at 3 °C min −1 to 200 °C.), the column HP-FFAP (the initial oven temperature was held at 70 °C for 2 min and ramped at 3 °C min −1 to 180 °C) and the injector temperature was maintained at 270 °C. The samples (1 μL) were injected with a split ratio of 1: 50. The carrier gas was helium at flow rate of 1.0 mL min −1 .
The constituents were identified by comparison of their mass spectra with those stored in Wiley 275 (Wiley, New York, USA) and NIST 05 (Gaithersburg, MD, USA) libraries or with mass spectra from the literature [32,33] as well as by comparison of their retention indices with those of the literature [32,33] or with those of available authentic compounds. The retention indices were determined in relation to a homologous series of n-alkanes (C 8 -C 24 ) under the same operating conditions. Component relative percentages were calculated based on GC peak areas without using correction factors. The component percentages (Tables 1 and 2) were calculated as mean values from duplicate GC and GC-MS analyses.

Conclusions
The present research has provided a more detail insight into the phytochemical composition of Centaurium erythraea Rafn essential oil and new data for the headspace VOCs composition were obtained. Molecular mass, polarity and volatility of VOCs as well as the type of used fiber resulted in significantly different headspace/oil chemical compositions. Differences in monoterpenes, sesquiterpenes and fatty acid-derived compounds in the headspace/oil distribution were expected, but the abundance of benzene derivatives in the headspace was a surprise, particularly the high percentages of toluene and naphthalene that could be connected with a non-anthropogenic origin. The oil fractionation enabled a more detailed oil analysis (a number of compounds were found only after fractionation). Lot of similarities were found with previous reports on this oil, but different features were also noted, indicating possible moderate geographical variability with structurally similar major compounds. The essential oil showed antimicrobial activities on Escherichia coli, Salmonella enteritidis, Staphylococcus aureus and Bacillus cereus, but more microorganisms should be tested in further research for detail evaluation of its antibacterial activity.