Essential oil composition of Salvia fruticosa Mill. populations from Balkan Peninsula

The aim of this study was to investigate the yield and chemical composition of the essential oil (EO) isolated from 19 different populations of Salvia fruticosa Mill. (Greek sage, Lamiaceae) from nine different regions of Albania and Greece. The EO yield ranged from 0.25% to 4.00%. Eighteen of the total analyzed populations met the Ph.Eur.8.0 minimal requirements concerning the essential oil yield. Performing GC/FID/MS analyses, a total of 75 components were detected, representing 79.15-97.83% of the oils. Thirteen components (α-pinene, camphene, β-pinene, myrcene, 1,8-cineole, γ-terpinene, cis-thujone, trans-thujone, camphor, terpinene-4-ol, trans-(E)-caryophyllene, aromadendrene and α-humulene) were identified in all samples, with 1,8-cineole as a predominant constituent. Statistical analysis showed that the geographical origin of plants did not have significant influence on the variation in chemical composition of the Greek sage essential oil.


Introduction
The genus Salvia L. from Lamiaceae is one of the largest genera in this family and includes around 1000 species that have almost cosmopolitan distribution (Kintzios, 2000). It is an important aromatic genus which is frequently used as herbal tea and as a source of essential oils and aroma chemicals (Karamanos, 2008). With significant economic importance are the pharmacopoeial herbs: S. fruticosa Mill. (Syn. S. triloba L. or Greek sage) and S. officinalis L. (Dalmatian, common or garden sage) (Kosar et al., 2005).
S. fruticosa is an endemic species of the Eastern Mediterranean basin (Ali-Shtayeh et al., 2000;Carmona et al., 2005;Elmann et al., 2009). Naturalized can be found in parts of the Western Mediterranean regions like Mal-ta, Spain and Portugal. S. fruticosa is the most widespread sage species in Greece, forming extended populations in littoral areas of the mainland, as well as the Ionian and Aegean islands (Kintzios, 2000).
The leaves of this herb have been used for treatment of various skin, blood, and infectious ailments as well as ailments of the digestive, circulatory and respiratory systems (Ali-Shtayeh et al., 2000;Carmona et al., 2005). Greek sage posses hypoglycemic effect and can be used against inflammations, hepatitis, and tuberculosis (Pitarokili et al., 2003). On the other hand the essential oil (EO) showed good antimicrobial activity against food borne bacteria (Longaray Delamare et al., 2007) and has antifungal activity (Pitarokili et al., 2003). Numerous investigations have been reported dealing with the essential oil composition and their biological activity, often referring to the S. officinalis species (Pierozan et al., 2009;Giweli et al., 2013).
However, to the best of our knowledge, there is limited information on the chemical composition of the essen-Ivana Cvetkovikj, Gjoshe Stefkov, Marija Karapandzova and Svetlana Kulevanova tial oil isolated from S. fruticosa from Balkan Peninsula. Therefore the aim of the present work was to assess and compare the oil yield and composition from Greece and Albania as native area of distribution of this herb.

Plant collection
Plant samples were collected from 19 different (18 indigenous and one non-indigenous) populations of Salvia fruticosa Mill., Lamiaceae, from nine different locations from Greece and Albania. The leaves were air dried, packed in paper bags and kept in a dark and cold place until analysis. Plant identity was verified and voucher specimens were deposited at the Institute of Pharmacognosy, Faculty of Pharmacy, Skopje, R. Macedonia (Table 1).

Essential oil isolation
The EOs were isolated from dried, cut leaves, by hydrodistillation in all-glass Clevenger apparatus for 2 hours according to pharmacopoeial method (Ph. Eur. 8.0., 2014). The oil yield was measured and presented in Table 1.

Analysis of essential oils' chemical composition
EO samples in hexane (1:1000) were analyzed on Agilent 7890А Gas Chromatography system equipped with FID detector and Agilent 5975C mass spectrometer. For that purpose, HP-5ms capillary column (30 m x 0.25 mm, film thickness 0.25 μm) was used. Analytical conditions were as follows: oven temperature at 60 °C (0 min), 3 °C/ min to 240 °C (1 min) and at the end increased to 280 °C at a rate of 10 °C/min (1 min); helium as carrier gas at a flow rate of 1 ml/min; injector temperature 220 °C and that of the FID detector 270 °C. One ml of each sample was injected at a split ratio of 1:1. The mass spectrometry conditions were: ionization voltage 70 eV, ion source temperature 230 °C, transfer line temperature 280 °C and mass range from 50 -550 Da. The MS was operated in scan mode.

Identification of the components
The compounds were identified on the basis of literature (Adams, 2007) and estimated Kovat′s (retention) indices that were determined using a mixture of homologous series of normal alkanes (C9-C25) analyzed under Automated Mass Spectral Deconvolution and Identification System (AMDIS)' conditions. Confirmation was made by comparing the mass spectra of the components present in the EOs with the reference spectra obtained from Nist, Wiley and Adams mass spectra libraries. Quantification of the EOs components was performed using the normalization method of the GC/FID peak areas without any correction factors.

Statistical analysis
The significance of differences between oil samples was tested by analysis of variance (ANOVA) using the package Excel for Windows 7 and represented by critical value of an F-test (F) and statistical significance (p). ANO-VA was performed on the geographical origin of plants as grouping factor.

Essential oil yield
The EO yield of 18 indigenous and one non-indigenous (commercial) population of Greek sage from nine different locations from Greece and Albania ranged from 2.50 ml/kg to 40.00 ml/kg (0.25-4.00 %) ( Table 1). The highest yield was recorded in two S. fruticosa populations, one from Dhermi, Albania (ALB 13) and the other population from Vrysses, Greece (GR 17), while the lowest was found in the Albanian population (ALB 5) from Llogora. Eighteen of the analyzed 19 populations met the Ph.Eur.8.0 minimal requirements concerning the essential oil yield.

EO composition
Data analysis of the EO chemical composition revealed six different classes of components: monoterpene hydrocarbons (MH), oxygen-containing monoterpenes (OMT), sesquiterpene hydrocarbons (SH), oxygen-containing sesquiterpenes (OST), diterpenes (D) and non-terpene components (NT). Generally, OMT was dominant fractions, present with more than 45% in all tested samples, followed by smaller amounts of MT or ST (Table 2). On the other hand, the diterpene chemical class was absent from two Greek populations (GR 14 and GR 17).
Our results are in full agreement with Giweli et al. (2013) who reported high amounts of 1,8-cineole, followed by camphor, β-pinene, myrcene and α-pinene in their samples of S. triloba provided from Libya. Additionally, Kosar et al. (2005), confirmed 1,8-cineole and camphor as predominant constituents in their Turkish sage samples and the same applies for S. fruticosa from Greece (Pitarokili et al., 2003;Pavlidou et al., 2004         ray Delamare et al. (2007) and Pierozan et al. (2009) found α-thujone as a major compound in their analyzed populations which is in contrast with our results and other available data.
Taking into concideration high cineole and camphor content, compounds with well established antibacterial properties, potential antimicrobial activity of the essential oil of S. fruticosa from Greece and Albania can be foreseen. Positive results of this activity could be used in the pharmaceutical industry, food production, and in the production of cosmetics or in any other purposes as substituted for S. officinalis.
Statistical analysis of variance (ANOVA) of the chemical composition of the essential oils for the complete data set of 19 samples revealed that there wasn't a statistically significant difference in the EO composition of S. fruticosa populations, regardless its origin.
The results obtained from the essential oils isolated from Greek sage correlate with the available literature data, and the statistical analysis showed that the oil composition is not influenced by the geographical locations.