Antioxidant Potential of Pine Needles: A Systematic Study on the Essential Oils and Extracts of 46 Species of the Genus Pinus †

The antioxidant activity of the essential oils, as well as of the organic and hydroethanolic extracts, of the fresh needles of 54 pine taxa was evaluated using the peroxy-oxalate and luminol chemiluminescence assays. Among all evaluated essential oils, P. canariensis and P. attenuata displayed the highest levels of activity. P. contorta var. murrayana, followed by P. nigra var. caramanica, exhibited the highest antioxidant capacity among the organic extracts, while the most active hydroethanolic extract was that of P. nigra subsp. nigra. Based on the overall levels of activity, the latter taxon was selected for phytochemical analysis targeting the isolation of the bioactive constituents. As such, the organic extract of P. nigra subsp. nigra was subjected to chromatographic separations to yield 11 secondary metabolites (1−11) that were evaluated for their antioxidant activity. Nonetheless, the isolated compounds were found to be less active than the crude extract, thus suggesting the potential role of synergism.


Introduction
In recent years, interest towards finding new antioxidant agents derived from natural sources has increased, since synthetic antioxidant compounds currently in use, such as butylated hydroxy-anisole (BHA) and tertiary butyl hydroquinone (TBHQ), may induce serious side effects (e.g., carcinogenesis) [1,2]. Apart from preventing food deterioration by militating against the activity of reactive oxygen species, natural antioxidant agents may also add nutritional value as functional food ingredients [3].
Pines are monoecious woody plants, mostly seen as tall trees and rarely as shrubs, with distinctive needle-shaped evergreen leaves, encountered in a variety of terrestrial environments and climatic zones in both hemispheres, mainly distributed over the northern hemisphere, while they also occur in subtropical and tropical areas of Central America and Asia [4,5]. The genus Pinus, including approximately 110 species [4,5], is important from an ecological point of view, since its representatives form extended forests either in pure stands or mixed with other conifers. Furthermore, from an economic point of view, pines are a valuable source of nuts and seeds, as well as resin, but also of pulp and paper, timber and construction materials.

Plant Material
Fresh needles of 54 taxa of genus Pinus, namely 37 taxa of subgenus Pinus and 17 taxa of subgenus Strobus, were collected from either well-documented wild localities or from botanical gardens, as previously described [21]. Voucher specimens of the taxa have been deposited at the Herbarium of the Section of Pharmacognosy and Chemistry of Natural Products, Department of Pharmacy, National and Kapodistrian University of Athens. The moisture content of the fresh needles ranged between 49-57%, as determined after incubation in an oven at 80 • C for 6 h.

Isolation of Essential Oils
Fresh needles of each sample (30-50 g) were cut into small pieces (0.5-1 cm) and separately subjected to hydro-distillation for 3 h using a modified Clevenger-type apparatus with a water-cooled receiver, in order to reduce overheating artifacts. The isolated essential oils were taken up in pentane, dried over anhydrous sodium sulfate and stored at 4 • C until analyzed.

Preparation of Extracts
Fresh needles of each sample were cut into small pieces (0.5-1 cm) and divided into two parts (A and B), of approx. 0.5 g each. Part A was macerated with 5 mL CH 2 Cl 2 /EtOH (2:1) to prepare the organic extract, while part B was macerated with 5 mL EtOH/H 2 O (1:2) to prepare the hydroethanolic extract. In both cases, extraction was repeated twice for 24 h each time at 25 • C. After evaporation of the solvents in vacuo, the extracts were weighted and stored at 4 • C until assayed.

Evaluation of Antioxidant Activity Using the Peroxy-Oxalate Chemiluminescence Assay
The antioxidant activity of the essential oils was evaluated using the peroxy-oxalate chemiluminescence (POCL) assay, based on the measurement of CL as a result of the oxidation of an aryl oxalate ester by H 2 O 2 in the presence of 9,10-DPA as a fluorophore (activator) and developed for assessing the hydrogen peroxide scavenging activity of low polarity hydrophobic samples [22]. Briefly, 0.2 mL of TCPO solution (0.45 mM) and 0.05 mL of the sample solution (at least three different concentrations were tested) or solvent (EtOAc) in the case of blank measurements were placed in a cuvette and immediately 1.8 mL 9,10-DPA solution (0.5 mM), 0.2 mL imidazole solution (4.5 mM) and 0.025 mL H 2 O 2 solution (2.25 mM) were added and mixed well for 5 s. All solutions were prepared in EtOAc/MeCN (9:1), with the exception of the sample solutions which were prepared in EtOAc. CL was continuously monitored in a JENWAY 6200 fluorimeter (Jenway Ltd., Essex, UK), keeping the lamp off and using only the photomultiplier of the apparatus, until the reaction reached a plateau and CL intensity was recorded.

Evaluation of Antioxidant Activity Using the Luminol Chemiluminescence Assay
The antioxidant activity of the organic and hydroethanolic extracts, as well as of the isolated metabolites was evaluated using the luminol chemiluminescence (LCL) assay, based on the measurement of CL as a result of the oxidation of luminol by H 2 O 2 in the presence of cobalt (II) as a transition metal and EDTA as a metal chelator, and developed for assessing the hydroxyl free radical scavenging activity of medium and high polarity samples [23,24]. Briefly, 1 mL of Co(II)/EDTA solution (8.

Determination of Antioxidant Activity and Statistical Analysis
For both assays, an equation in the form I 0 /I = a × C ± b was obtained by plotting I 0 /I against C, where I 0 is the initial CL intensity recorded for the blank, I is the reduced CL intensity recorded after the addition of the sample and C is the concentration of the sample (in µg/mL). Correlations were established using linear regression analysis (with a coefficient R 2 > 0.98), employing Microsoft Office Excel 2007 software. Assignments a and b represent the gradient and the intercept of the equation, respectively. The concentration necessary to decrease the CL intensity by 50% (IC 50 ) was calculated by setting I 0 /I = 2. All measurements were performed at least in three independent experiments and data are presented as mean ± SEM (standard error of the mean).

Evaluation of the Antioxidant Activity of Essential Oils
The antioxidant activity of the essential oils obtained from the fresh needles of 46 pine species, including 37 and 17 taxa of the subgenera Pinus and Strobus, respectively, was evaluated using the POCL assay. According to the results of the evaluation (Table 1, Figure 1a), the IC 50 values of the pine needle essential oils ranged from 1.00 ± 0.08 (P. canariensis) to 20.03 ± 2.77 (P. cembroides var. monophylla). Besides P. canariensis oil which exhibited the most significant antioxidant activity, high levels of activity were also displayed by the essential oils of P. attenuata (1.30 ± 0.02), P. muricata (1.60 ± 0.09), P. sylvestris var. scotica (1.67 ± 0.05), P. halepensis (1.78 ± 0.17), P. mugo var. prostrata (1.79 ± 0.21), P. mugo (1.89 ± 0.16) and P. monticola (1.94 ± 0.09). As can be observed, with the exception of the latter needle oil derived from a species belonging to the subgenus Strobus, the most active essential oils were obtained from taxa of the subgenus Pinus. Analysis of the chemical composition of the essential oils evaluated for their antioxidant activity in the present study has shown that mono-and sesquiterpene derivatives characterize the majority of the essential oils [21]. In most cases, αand β-pinene were the major representatives of the monoterpene fraction. However, occasionally β-phellandrene and/or δ-3-carene were also present in high percentages. The sesquiterpene group was characterized by germacrene D, while the levels of diterpenes varied notably. Germacrene D was one of the common main metabolites among the three most active samples (P. canariensis 44.0%, P. attenuata 29.0%, and P. muricata 41.5%), and while it was detected in notably lower amounts in the essential oils of the following in activity order P. sylvestris var. scotica (5.1%) and P. mugo var. prostrata (2.8%), its oxygenated derivative germacrene D-4-ol reached a relatively higher percentage (10.0% and 6.0%, respectively). Instead, the major metabolite in P. halepensis needle oil was β-caryophyllene (19.0%). It should be noted though that no clear pattern correlating the antioxidant effect and the chemical composition of the investigated essential oils can be observed overall. Thus, according to our results and in agreement with the literature data [25], it can be deduced that the antioxidant activity exhibited by our samples may be a result of synergism, since pinenes, ubiquitous constituents of pine essential oils often appearing as major components, do not possess antioxidant properties [26]. On the other hand, terpene derivatives such as germacrene D, β-caryophyllene, and γ-terpinene have been reported to exert antioxidant activity [27]. needles of 54 pine taxa and evaluated for their antioxidant potential using the LCL assay.
An overall comparison of the IC50 values of the investigated organic extracts (Table 1, Figure 1b) revealed the superiority of P. contorta var. murrayana of section Trifoliae (subgenus Pinus), followed by P. nigra subsp. caramanica and P. nigra subsp. salzmanii of section Pinus (subgenus Pinus), along with P. monticola of section Quinquefoliae (subgenus Strobus), with the organic extracts of the four taxa exhibiting stronger antioxidant activity than quercetin.  The antioxidant activity of the Aleppo pine (P. halepensis) needle oils from Algeria was studied using four different assays, namely 2,2-diphenyl-1-picrylhydrazyl radical scavenging (DPPH), β-carotene bleaching (BCB), iron (II) chelating ability employing the Fe 2+ -ferrozine system (FICA) and potassium ferricyanide reducing power (PFRAP) assays, and high levels of activity, especially for a specific chemotype rich in caryophyllene oxide, were also observed, as in our case [28,29]. The high antioxidant potential of P. halepensis essential oil was further verified by Postu et al. who observed remarkable activity in the DPPH and 2,2-azino-bis-(3-ethylbenzothiazoline-6-sulphonate) radical cation scavenging (or Trolox equivalent antioxidant capacity, ABTS/TEAC) assays [30]. The antioxidant potential of P. mugo essential oil has been evaluated by Grassmann et al., employing a variety of biochemical tests in both aqueous (e.g., Fenton system, xanthine oxidaseinduced superoxide radical formation) and more lipophilic environments (e.g., ACCcleavage by activated neutrophils in whole blood, copper-induced oxidation of low-density lipoprotein), observing good antioxidant activity in more lipophilic rather than in aqueous environments [31]. In contrast, Kurti et al. observed a rather low to moderate DPPH radical scavenging activity for the needle essential oil of P. mugo from Kosovo [32]. High to moderate activity, as observed in the present study, for the needle oil of the Himalayan blue pine (P. wallichiana), was also noted by Dar et al. using the DPPH assay [26]. On the other hand, the essential oil of the Swiss stone pine (P. cembra), which showed high to moderate activity in our study, has previously exhibited rather weak DPPH radical scavenging activity [33]. The needle essential oil of the Japanese black pine (P. thunbergii) exerted a strong DPPH radical scavenging potential, as in our study, but insignificant nitrite radical scavenging activity [34]. P. tabuliformis, which exhibited moderate levels of activity in the present study, has also previously displayed moderate antioxidant activity when evaluated using the DPPH, ABTS/TEAC and ferric reducing antioxidant power (FRAP) assays [35]. In the study of Yener et al., the essential oil of P. brutia exhibited strong iron (II) chelating ability and relatively lower levels of activity in the DPPH and PFRAP assays, whereas the foliage essential oil of umbrella pine (P. pinea) displayed weak iron (II) chelating ability, as well as weak reducing power [36]. Moreover, in the study of Ustun et al., the essential oil of P. brutia exhibited weak activity in the PFRAP assay, while P. sylvestris essential oil showed moderate iron (II) chelating ability [37]. P. sylvestris essential oil and its fractions from Kosovo were also tested as DPPH radical scavenging agents, displaying a weak to moderate potential [32]. In the same study, the needle oils of P. nigra, P. peuce and P. heldreichii and their fractions were evaluated for their DPPH radical scavenging activity, which was proven rather weak [32]. Similarly, P. heldreichii var. leucodermis needle oil from central Herzegovina exhibited weak DPPH radical scavenging activity [38]. The red pine needle oil (P. densiflora) has exerted a rather weak DPPH radical scavenging potential, as well as nitrite radical scavenging ability [34]. The needle oil of the maritime pine (P. pinaster) has been evaluated by Tümen et al. for its antioxidant potential using the DPPH, ABTS/TEAC and FRAP assays, as well as for its hydroxyl radical scavenging activity, displaying a rather moderate potential [39]. The Monterey pine (P. radiata) needle oil, evaluated for its antioxidant capacity using the DPPH, BCB and LCL assays, exhibited a rather moderate to weak activity in all three tests [40], similarly to our results. The Japanese white pine (P. parviflora) needle oil has demonstrated weaker DPPH scavenging activity compared to thymol, but strong hydroxyl radical scavenging activity in reference to mannitol [27]. P. massoniana needle oil has exerted low to moderate antioxidant potential, as determined using the DPPH, ABTS/TEAC and FRAP assays [35], while the Chir pine (P. roxburghii) needle oil has showed weak DPPH radical scavenging activity [41].

Evaluation of the Antioxidant Activity of Extracts
In the framework of the present study, two extracts of different polarity, namely an organic extract resulting from maceration of the needles in CH 2 Cl 2 /EtOH (2:1) containing less polar constituents and a hydroethanolic extract resulting from maceration of the needles in EtOH/H 2 O (1:2) containing more polar constituents, were prepared from the fresh needles of 54 pine taxa and evaluated for their antioxidant potential using the LCL assay.
An overall comparison of the IC 50 values of the investigated organic extracts (Table 1, Figure 1b) revealed the superiority of P. contorta var. murrayana of section Trifoliae (subgenus Pinus), followed by P. nigra subsp. caramanica and P. nigra subsp. salzmanii of section Pinus (subgenus Pinus), along with P. monticola of section Quinquefoliae (subgenus Strobus), with the organic extracts of the four taxa exhibiting stronger antioxidant activity than quercetin.
The antioxidant activity evaluation of the hydroethanolic extracts (Table 1, Figure 1c) showed that only P. nigra subsp. nigra exhibited a lower IC 50 value than quercetin. However, significant levels of activity were also observed for the hydroethanolic extracts of P. brutia, P. canariensis, P. tabuliformis, P. contorta var. latifolia, P. mugo var. pumilio, P. pinaster, and P. ponderosa. All aforementioned taxa belong to the subgenus Pinus.
A number of studies employing different assays for the evaluation of the antioxidant activity of various pine needle extracts have been undertaken and their results are summarized in Table 2. Nonetheless, due to the different extraction protocols used in these investigations, direct comparison of the results obtained in the current study is not straightforward.       2 2,2-diphenyl-1-picrylhydrazyl radical scavenging, 3 N,N-dimethyl-p-phenylene diamine radical scavenging, 4 potassium ferricyanide reducing power, 5 iron (II) chelating ability employing the Fe 2+ -ferrozine system, 6 oxygen radical absorbance capacity, 7 2,2-azino-bis-(3-ethylbenzothiazoline-6-sulphonate) radical cation scavenging or Trolox equivalent antioxidant capacity, 8 ferric reducing antioxidant power, 9 expressed as IC 50 / EC 50 in µg/mL, 10 expressed as % of scavenging activity (at a given concentration), 11 expressed as absorbance at 700 nm (at a given concentration), 12 expressed as % of chelating ability (at a given concentration), 13 expressed as µM Trolox equivalents (TE) per g dry weight, 14 expressed as % of reducing capacity (at a given concentration).

Phytochemical Analysis of P. nigra subsp. nigra and Evaluation of the Antioxidant Activity of the Isolated Metabolites
In the current study, both organic and hydroethanolic extracts, as well as the essential oil of the black pine (P. nigra subsp. nigra), were constantly among the most active samples tested, with IC 50 values of 0.17 ± 0.01, 0.14 ± 0.02, and 2.05 ± 0.20, respectively. Therefore, phytochemical analysis of the black pine needle extract was undertaken, aiming at the isolation of the metabolites responsible for the observed antioxidant activity.
A series of chromatographic separations of the organic extract of the fresh needles of P. nigra subsp. nigra led to the isolation of compounds 1−11 (Figure 2), which were identified as dehydroabietic acid (1) (9) [51], 5,4 -dihydroxy-3,6,7-trimethoxy-8-C-methylflavone (10) [51,52], (-)-catechin (11) [51], a rare stereoisomer of catechin, and β-sitosterol [51] by comparison of their spectroscopic and physical characteristics with those reported in the literature. Among them, compounds 2-4, 6, 11 and β-sitosterol are reported for the first time from black pine, whereas metabolite 10 is reported for the first time in Gymnospermae. It is worth noting that the chemical structure of 10, which has been reported from the leaves of three Vellozia species and the fungus Colletotrichum dematium f.sp. epilobii, has been so far only tentatively assigned [52,53]. that the chemical structure of 10, which has been reported from the leaves of three Vellozia species and the fungus Colletotrichum dematium f.sp. epilobii, has been so far only tentatively assigned [52,53]. The structure of compound 10 was elucidated after thorough analysis of its spectroscopic data. Specifically, according to the NMR and MS spectra, metabolite 10 was identified as a flavonol with a para-substituted ring B, bearing one aromatic methyl, two hydroxy and three methoxy groups. The positions of the functional groups were determined after analysis of a standard set of six UV spectra [54]. In particular, in the presence of NaOMe, band Ib exhibited a bathochromic shift of 56 nm with no decrease in intensity, typical of the presence of a free hydroxy group at C-4′. Moreover, no small additional peak or shoulder at 330 nm was observed, indicating the absence of a free hydroxy group at C-7. With AlCl3 and AlCl3-HCl, bathochromic shifts of 25 nm and 24 nm, respectively, were observed, diagnostic for the presence of 5-OH and 6-OMe in 3-O-substituted flavonols. No shift was observed in band II in the presence of NaOAc, verifying the presence of 6-OMe, as well as of a methyl group at C-8, also confirming a 7-O-substitution. The presence of 7-OMe was confirmed by the fact that no shift was observed in the presence of NaOAc and H3BO3. The proposed structure was further supported by the heteronuclear correlations observed in the HMBC spectrum of metabolite 10. The 1 H and 13 C NMR chemical shifts for compound 10 are reported herein for the first time, complementing the relevant literature.
Metabolites 1-11 were subjected to evaluation of their antioxidant potential using the The structure of compound 10 was elucidated after thorough analysis of its spectroscopic data. Specifically, according to the NMR and MS spectra, metabolite 10 was identified as a flavonol with a para-substituted ring B, bearing one aromatic methyl, two hydroxy and three methoxy groups. The positions of the functional groups were determined after analysis of a standard set of six UV spectra [54]. In particular, in the presence of NaOMe, band Ib exhibited a bathochromic shift of 56 nm with no decrease in intensity, typical of the presence of a free hydroxy group at C-4 . Moreover, no small additional peak or shoulder at 330 nm was observed, indicating the absence of a free hydroxy group at C-7. With AlCl 3 and AlCl 3 -HCl, bathochromic shifts of 25 nm and 24 nm, respectively, were observed, diagnostic for the presence of 5-OH and 6-OMe in 3-O-substituted flavonols. No shift was observed in band II in the presence of NaOAc, verifying the presence of 6-OMe, as well as of a methyl group at C-8, also confirming a 7-O-substitution. The presence of 7-OMe was confirmed by the fact that no shift was observed in the presence of NaOAc and H 3 BO 3 . The proposed structure was further supported by the heteronuclear correlations observed in the HMBC spectrum of metabolite 10. The 1 H and 13 C NMR chemical shifts for compound 10 are reported herein for the first time, complementing the relevant literature.
Metabolites 1-11 were subjected to evaluation of their antioxidant potential using the LCL assay (Table 3). Phenolic compounds 10 and 11 displayed significant levels of activity with IC 50 values of 1.95 ± 0.21 and 1.34 ± 0.16 µg/mL, respectively, whereas the isolated diterpenes showed moderate levels of activity (1 and 3) or were proven inactive (2 and 4-9). The fact that both extracts of the black pine needles showed higher antioxidant activity compared to that of the isolated compounds indicates that the higher antioxidant potential of the extracts may be the result of synergism. Table 3. Antioxidant activity (expressed as IC 50 in µg/mL) of compounds 1-11 isolated from the organic extract of the fresh needles of Pinus nigra subsp. nigra. 1.95 ± 0.21 11

Conclusions
The antioxidant activity of the essential oils, as well as of the organic (CH 2 Cl 2 /EtOH 2:1) and hydroethanolic (EtOH/H 2 O 1:2) extracts of the fresh needles, from 54 pine taxa was evaluated using the POCL and LCL assays. The extracts showed overall higher I 0 inhibition in comparison to the essential oils. Two samples from subgenus Pinus were proven to be the most potent among the investigated essential oils, namely P. canariensis (section Pinus) followed by P. attenuata oil (section Trifoliae), albeit with observed IC 50 values higher than that of the reference (β-carotene). The organic extracts of P. contorta var. murrayana (section Trifoliae), followed by P. nigra subsp. caramanica (section Pinus), P. nigra subsp. salzmanii (section Pinus), P. monticola (section Quinquefoliae), P. mugo var. prostrata (section Pinus) and P. sylvestris subsp. scotica (section Pinus), exhibited the same or higher levels of activity compared to the reference (quercetin). Among the hydroethanolic extracts, however, only P. nigra subsp. nigra (section Pinus) demonstrated stronger antioxidant activity than that of the reference (quercetin), albeit with several other taxa of subgenus Pinus displaying significant levels of activity.
Based on the overall levels of activity, P. nigra subsp. nigra was selected for phytochemical analysis targeting the isolation of the bioactive constituents. Among the secondary metabolites isolated from the organic extract of the black pine needles, the abietane and labdane diterpenes 1-9 were not active, whereas the two phenolic compounds 10 and 11 showed noteworthy levels of antioxidant activity. To the best of our knowledge, this is the first report on the evaluation of the antioxidant activity of the needle essential oils and extracts from 37 and 41 pine taxa, respectively.