Effect of geographical origin on yield and composition of cone essential oils of Cedrus libani A. Rich. growing in Lebanese protected areas and variability assessment in comparison with literature survey

2.2.1 GC analyses

The GC analysis was carried out using a Thermo Electron Corporation apparatus (Agilent, Santa Clara, CA, USA) equipped with a flame ionization detector, a nonpolar HP-5MS (5% phenylmethylsiloxane) capillary column (30 m×0.25 mm i.d., film thickness 0.25 μm), and a polar fused-silica HP Innowax capillary column (polyethylene glycol, 50 m×0.20 mm i.d., film thickness 0.20 μm) in order to confirm the identification of the components. The carrier gas was He at a rate of 1 mL/min. The oven temperature followed a gradient rising from 40 °C to 300 °C at 5 °C/min and finally held isothermal at 300 °C for 5 min. Diluted 1-μL samples (1/100, vol/vol in pentane) were manually injected at 250 °C and in the splitless mode. Flame ionization detection was performed at 310 °C.

2.2.2 GC/MS analyses

The GC/MS analyses were achieved using an Agilent Thermo Electron Corporation (Agilent, Santa Clara, CA, USA) apparatus coupled with Mass Detector 5975 and equipped with 7683 B auto sampler [injection of 1 μL of each oil sample diluted in pentane (1/100 vol/vol)]. A fused-silica capillary column HP-5MS (30 m×0.25 mm i.d., film thickness 0.25 μm) and a fused-silica HP Innowax polyethylene glycol capillary column (50 m×0.20 mm i.d., film thickness 0.20 μm) were used. The oven temperature program was identical to that described above (cf. GC analyses). Helium was the carrier gas (1 mL/min). The mass spectra were recorded at 70 eV. Ion source and transfer line temperatures were 310 °C and 320 °C, respectively. The acquisition was recorded in full scan mode (50–400 amu).

2.2.3 Identification and quantification of the components

An n-alkane (C₇-C₂₅) mixture was analyzed under the same experimental GC/MS conditions to calculate the retention indices (RIs). Identification of EO components was performed by comparing their mass spectra on both columns with those listed in the commercial mass spectral libraries NIST and Wiley 275 computer libraries, our home-made library constructed with pure compounds, and those from literature [21], [22], allowing a reliable confirmation of the identity of each component. A further identification was accomplished by comparing their RIs on both polar and apolar columns relative to the retention times of the series of n-alkanes (C₇–C₂₅) with those from literature [21], [23] or with those of standard compounds available in our laboratories, obtained from Sigma-Aldrich (Germany). Standards of some EOs of known composition (such as the EO of Rosmarinus officinalis L. from Phytosun Aroms, Plélo, France) were also injected in similar conditions for the comparison of retention times (t_R) and mass spectra. In addition, the following standards were purchased from Sigma-Aldrich: α-pinene, camphene, β-pinene, myrcene, limonene, borneol, terpinen-4-ol, α-terpineol, verbenone, bornyl acetate, β-caryophyllene, γ-cadinene, δ-cadinene, caryophyllene oxide, α-cadinol, and β-bisabolol. Relative proportion of each compound was determined from the GC peak areas.

2.3.1 Analysis of variance

One-way analysis of variance (ANOVA) was applied in order to evaluate EO chemical variability according to geographical location and to identify where any difference may have existed between samples. Statistical analyses were performed using SPSS 16.0 software (SPSS Inc., Chicago, IL, USA). A probability value of p≤0.05 was considered to highlight a statistically significant difference between oil samples.

2.3.2 Principal components analysis

Principal component analysis (PCA), a multivariate statistical technique that aims at reducing the multivariate space in which objects (oil samples) are distributed, was applied. This processing was performed using XLSTAT 2014.5.03 (Addinsoft, Paris, France).

3.2.1 Essential oil chemical composition

The EOs extracted from the cones of C. libani constitute a mixture of 51 compounds, representing 95.2% to 99.9% of the total oil composition. Table 1 lists the average of the relative percentages of the constituents of the oils, collected in triplicate from each site of the five Lebanese locations during September 2016. Cedrus libani EO chemical compositions were distinguished by their wealth in α-pinene (25.1%–37.3%), β-pinene (6.4%–35.6%), myrcene (0%–30.6%), limonene (5.6%–14.1%), sclarene (1.1%–5.5%), abieta-8,11,13-triene (2.2%–10.2%), kaur-16-ene (3.5%–7.5%), and abieta-7,13-diene (1.9%–7.5%), representing more than 5% of the total oil content in at least a sample, with a predominance of α-pinene, β-pinene, and myrcene (Table 1).

3.2.2 Essential oil chemical variability according to the geographical origin

For the purpose of detecting a potential impact of harvesting sites on the accumulation of the main terpene metabolites in EO of C. libani from cones, which may lead to specific oil chemical compositions, one-way ANOVA test (SPSS 16.0 software) was applied in order to evaluate the existence of significant correlations between EO main components of cedar cones and the geographical location of harvest and to estimate the optimal harvesting site of C. libani, providing the optimal EO mixture corresponding to a required activity according to user’s need.

One-way ANOVA test showed that EO chemical composition varied considerably according to the geographical location of harvest (p<0.05).

Our study revealed the dominance of α-pinene in all the samples collected from the different Lebanese sites. In Chouf reserve, we noted the highest α-pinene contents.

β-Pinene amounts significantly fluctuated among the regions and were higher in the samples collected from Ehden reserve, while the lowest values of β-pinene were noticed in Qartaba (p<0.05).

Limonene levels were significantly lower in Tannourine and Chouf, while this compound predominated in the EO of the cones harvested from Qartaba (p<0.05).

The region of harvest largely affected myrcene (p<0.05). A statistically significant difference in α-terpineol contents was noted between the regions (p<0.05); the highest values were noticed in Tannourine and Chouf, whereas we registered the lowest levels in Qartaba.

Bornyl acetate amounts significantly varied between the locations (p<0.05) and prevailed in the samples harvested from Bsharri and Chouf compared to the other sites.

Geographical location significantly affected sclarene, kaur-16-ene, abieta-8,11,13-triene, and abieta-7,13- diene levels (p<0.05). Sclarene and kaur-16-ene reached their minimum values in Ehden reserve and Qartaba. The samples collected from Chouf reserve and Qartaba were characterized by their lowest abieta-8,11,13-triene amounts, whereas the samples harvested from Tannourine were distinguished by their wealth in this component. The lowest values of abieta-7,13-diene were recorded in Ehden and the highest ones in Tannourine.

Geographical locations influenced the contents of β-farnesene and β-cembrenediol (p<0.05). Their amounts were significantly higher in Tannourine compared to the other provenances. The lowest values of β-farnesene were observed in Ehden and Chouf reserves and in Chouf reserve for β-cembrenediol.

3.2.3 Chemical variability of C. libani cones EOs according to geographical origin – comparison with the literature

In order to assess for the first time the existence of cedar cones EO chemical polymorphism within a group of samples collected from different geographical locations and to distinguish which constituent has signed the oil chemical identity in cedar cones EOs according to the provenance, PCA was conducted on all EO samples (Figure 1). Principal component analysis showed that the maximal amount of variance in the data and its direction are explained by the first two principal components PC1 and PC2, which extracted 72.01% and 10.92% of the total variance, respectively.

Figure 1:

Principal component projection plot of PC1 and PC2 scores and loadings based on GC/MS data of cones EOs of C. libani collected from several geographical locations (our results and data extracted from literature). The numbers represent the samples collected from the different studied locations (our samples and data previously published). (1): Bsharri (our study), (2): Ehden (our study), (3): Tannourine (our study), (4): Chouf (our study), (5): Qartaba (our study), (6): Hadath Eljebeh–Tannourine reserve [8], (7): Başkonuş District, Turkey [25], (8): Adana, Turkey [24], (9): Bsharri [27].

The first PC is mainly associated to α-pinene witnessing the predominance of this compound in all the oil studied samples and marking its lowest values in Adana, Turkey (8) [24]. β-Pinene, limonene, abieta-8,11,13-triene and abieta-7,13-diene contributed to the definition of the first PC.

Significant concentrations of β-pinene distinguished predominantly Ehden (2) (our study), Chouf (4) (our study), Bsharri (1) (our study), Tannourine (3) (our study), and Bsharri (9) [27] oil samples, defining consequently the predominance of α-pinene and β-pinene in these locations. Limonene, abieta-8,11,13-triene, and abieta-7,13-diene were identified in all samples at variable concentrations. It is noticeable that limonene, abieta-8,11,13-triene, and α-pinene were the major compounds of Adana, Turkey, oil sample (8) [24]. The oil composition of Başkonuş District, Turkey (7) [25], although dominated by α-pinene, exhibited the highest contents of abieta-7,13-diene and then characterized by α-pinene and abieta-7,13-diene majority.

The second PC is mainly correlated to myrcene, exclusively detected in Qartaba (5), Hadath Eljebeh–Tannourine reserve (6) [8] and then Adana, Turkey (8) [24], Bsharri (9) [27], and Başkonuş District, Turkey (7) [25]. This compound was checked as the most abundant in Qartaba samples (5) and was also found in significant concentration in Hadath Eljebeh–Tannourine reserve (6) [8]. The oils collected from these two regions were distinguished from the others by their remarkable content of myrcene. Therefore, the prevalence of myrcene, α-pinene, and limonene was marked in Qartaba (5), and α-pinene and myrcene in Hadath Eljebeh–Tannourine reserve (6) [8].

In summary, we pointed out that the comparison of our results with those reported exhibited the occurrence of several EO types according to the region of harvest and provided a new insight into the chemical variability of cedar cones EOs. These results clearly indicate that biosynthesis of EO secondary metabolites is strongly influenced by geographical conditions: environmental factors of the different studied sites and cultivation conditions. Geographical location is associated to a number of environmental factors, such as precipitation, wind exposure, light intensity, UV radiation, high temperature and drought conditions, cold, frost, soil composition and fertility, and so on. The combination of all these factors puts pressure on the plants, with a need to develop mechanisms to be protected and to prevent damages. In fact, terpenoids are involved in specific adaptation strategies, such as defense against abiotic stresses and defense against pests and microorganisms [28]. Therefore, the variability observed in the accumulation of EO chemical compounds according to the region of harvest could be explained by the fact that specific compounds are mainly biosynthesized in a particular region in order to secure and protect the plant [13], [14], [18], [29]. Furthermore, this variability could be attributed to both abiotic and abiotic factors. These results stress the importance of selecting a harvesting region ensuring the optimal EO chemical composition that best fits the market demands.

3.2.4 Chemical variability according to cedar parts and extraction methods based on existing literature survey

According to previous reports, sesquiterpenes characterized mainly wood EO of the biblical cedar of Lebanon. Therefore, Saab et al. [30] in 2012 evidenced the predominant presence of β- (21.4%), α- (9.5%), and γ-himachalene (6%) isomers together with himachalol (5.6%), γ-(Z)-atlantone (5%), and allohimachalol (4.1%).

Başer and Demirçakmak in 1995 proved that C. libani wood EOs from Antalya and İçel were, respectively, distinguished by the dominance of β- (38.2%–34.3%), α- (12.8–11.5%), and γ-himachalene (7.6%–6.8%) isomers with trans-α-atlantone (7.8%–14.8%), himachalol (1.2%–8.8%), trans-β-atlantone (1.4%–2.4%), β-cedrene (1.8%–2.8%), allohimachalol (1.8%–2.3%), (E)-10,11-dihydroatlantone (3.7%–0.02%), cis-β-atlantone (1.1%–2.5%), and cis-α-atlantone (1.06%–2.1%) [5]. Saab et al. [3] in 2005 demonstrated the predominance of himachalol (43.1%) and γ- (11.9%), α- (7.12%), and β- (7%) isomers of himachalene and allohimachalol (3.92%) in C. libani wood EO. In addition, they determined smaller amounts of the (E) and (Z) isomers of α-, β-, and γ-atlantone. According to Loizzo et al. [9] in 2008, himachalol (22.5%), β-himachalene (21.9%), α-himachalene (10.5%), and γ-himachalene (9.1%) were considered as peculiar of cedar wood EO.

Cedrus libani root EOs from İçel were characterized by the presence of β- (27.5%), α- (10.3%), and γ-himachalene (7.1%) isomers with trans-α-atlantone (23.7%), himachalol (9.7%), trans-β-atlantone (2.3%), β-cedrene (1.1%), allohimachalol (2.2%), (E)-10,11-dihydroatlantone (0.1%), cis-β-atlantone (2.3%), and cis-α-atlantone (2.8%) [5].

Limonene (17.7%) was the main compound in the EO obtained from the cones of C. libani natively grown in Turkey, followed by abieta-8,11,13-triene (17%) and α-pinene (12.3%). Limonene is used as an antimicrobial inhibitor in the food industry [24]. The chemical profile of EO from the oleoresin of the cones of C. libani grown in Turkey was distinguished by the preponderance of α-pinene (24.8%), abieta-7,13-diene (16.7%), abieta-8,11,13-triene (6.9%), manool (5.8%) and terpinen-4-ol (3.7%), α-terpineol (3.4%), p-cymene (2.9%), and limonene (2.7%) [25]. The investigation of the phytochemical composition of C. libani ethanol extract obtained from cones illustrated the dominance of α-pinene (51%), β-myrcene (13%), 7,13-abietadiene (3.2%), terpinolene (3.1%), and limonene (2.3%) [9].

Manool (11.0%) predominated in the hydrodistilled EOs of the C. libani trunk-bark samples [31].

Saab et al. [32] in 2011 reported that the main components of C. libani seed chloroform extracts were α- and β-pinene (34.4%±1.22% and 33.3%±1.08%, respectively). The occurrence of fatty acids and ethers as most abundant compounds, in particular oleic acid (17.3%±1.34%), methyl oleate (7.8%±0.42%), and ethyl oleate (5.3%±0.48%), in C. libani seed ethanol extract has been evidenced. Diterpenes, as neoabietol (11.8%±0.98%), abieta-7,13-diene (8%±0.46%), abieta-8(14), 13(15)-diene (7.9%±0.66%), abietol (6%±0.55%), and abietal (2%±0.13%), were also detected [32], [33].

For what concerns cedar leaves ethanol extract, germacrene D was checked as the most abundant compound (29.4%) [9]. It was found that fenchone (14.2%), β-thujone (28%), α-thujone (20.3%), camphor (8.7%), 1-borneol (4.7%), and endobornyl acetate (4%) were the main constituents in cedar leaf EOs [34].

Based on a PCA, a comparison of the main components (> 5%) of C. libani extracts, extracted from different parts (wood, roots, leaves, trunk-bark, seeds, and cones) and by different extraction methods (hydrodistillation, steam distillation, chloroform, and ethanol extraction), assessing our results and those previously published, is illustrated in Figure 2. Two PCs were chosen representing more than 62.29% of the variability. Principal component analysis results revealed the presence of three groups.

Figure 2:

Principal component projection plot of PC1 and PC2 scores and loadings based on GC/MS data of C. libani extracts extracted from different parts (our results and data previously published). The numbers represent the samples extracted from the different studied C. libani parts (our samples and data previously published). (1): C. libani wood EO from Antalya, steam hydrodistillation [5]; (2): C. libani wood EO from Tarsus, steam hydrodistillation [5]; (3): C. libani wood EO, hydrodistillation [30]; (4): C. libani wood EO, hydrodistillation [3]; (5): C. libani wood EO, hydrodistillation [9]; (6): C. libani root EO from Tarsus, steam hydrodistillation [5]; (7): C. libani leaves ethanol extract [9]; (8): C. libani leaves EO [34]; (9): C. libani trunk-bark EO [31]; (10): C. libani seed chloroform extract [32]; (11): C. libani seed ethanol extract [32]; (12): C. libani cones EO from Bsharri, hydrodistillation (our study); (13): C. libani cones EO from Ehden, hydrodistillation (our study); (14): C. libani cones EO from Tannourine, hydrodistillation (our study); (15): C. libani cones EO from Chouf, hydrodistillation (our study); (16): C. libani cones EO from Qartaba, hydrodistillation (our study); (17): C. libani cones EO, hydrodistillation [8]; (18): C. libani cones EO, hydrodistillation [25]; (19): C. libani cones EO, hydrodistillation [24]; (20): C. libani cones EO, hydrodistillation [27].

Cedar cones EOs, extracted by hydrodistillation (12, 13, 14, 15, 16) (our study), (17) [8], (18) [25], (19) [24], and (20) [27], and seed chloroform extract samples (10) [32] distributed in a group were clearly separated from the two other groups by the first principal component (PC1). These samples constituting the first cluster possessed in common high relative amounts of α-pinene, β-pinene, and then abieta-8,11,13-triene, myrcene, limonene, sclarene, palustradiene, and kaur-16-ene. α-Pinene and β-pinene are known to have antibacterial, antioxidant, anti-inflammatory, and anticarcinogenic activities [10], [18], [35], [36], [37], [38], [39], [40], [41].

A clear separation of the second cluster representing oils extracted from cedar wood (1, 2) [5], (3) [30], (4) [3], and (5) [9] and roots (6) [5] from the first cluster (cedar cones EOs and seed chloroform extract samples) is shown along PC1. This second cluster is separated from the third one by PC2 and characterized by high amounts of β-himachalene, α-himachalene, γ-himachalene, trans-α-atlantone, and himachalol. Principal component analysis allowed us to draw the following conclusions: cedar wood and root EOs had identical main components, and the relative concentration of these components was similar by both extraction methods: steam distillation and hydrodistillation.

The third cluster was separated from the group of seed chloroform extract and cedar cones EO samples by PC1 and from the group of wood and root EO samples by PC2. This cluster contained cedar leaves ethanol extract (7) [9], cedar leaves EO (8) [34], cedar EO from trunk-bark (9) [31], and cedar seed ethanol extract (11) [32] samples and was dominated by β-thujone, α-thujone, camphor, fenchone, β-caryophyllene, germacrene D, δ-cadinene, trans-α-bisabolene, 1-epi-cubenol, isolongifolene, manool, methyl oleate, oleic acid, ethyl oleate, abietol, neoabietol, abieta-7,13-diene, and abieta-8(14),13(15) diene.

According to PCA, seed ethanol extraction and seed chloroform extraction affected the chemical composition of seed extracts given the fact that different compounds were recovered after these two extraction methods.

These findings shed light on the cedar extract chemical polymorphism. Given the fact that each part of the same plant does not possess the same enzymatic equipment, the metabolomic profile is strongly dependent on plant parts or organs [42]. Extraction methods, extrinsic conditions, and interspecific and intraspecific factors also influenced plant extract chemical composition [14], [15], [16], [43], [44], [45].