Diterpenoids from the Buds of Pinus banksiana Lamb

Three new diterpenoids, namely 7α-hydroxyabieta-8,11,13,15-tetraen-18-oic acid, 7β,15,18-trihydroxyabieta-8,11,13-triene, 13,15-dihydroxypodocarpa-8,11,13-triene, and 12 other known compounds were isolated from buds of Pinus banksiana Lamb. All these compounds, except for 7-oxodehydroabietinol, were isolated for the first time from this plant. Their structures were elucidated by detailed spectroscopic studies and comparison with published data. All isolated compounds were tested for cytotoxic and antibacterial activities. Overall, two compounds, 7-oxodehydroabietinol and 18-nor-4,15-dihydroxyabieta-8,11,13-trien-7-one, showed moderate cytotoxicity against a human lung carcinoma cell line.


Introduction
Pinus banksiana Lamb (jack pine) is widely distributed in North American forests and particularly in Canada, where its presence extends from Cape Breton Island, Nova Scotia, up to the Mackenzie river in the Northwest Territories [1][2][3]. Although jack pine was essentially used in the wood industry as a source of pulpwood, lumber, and round timber [3], it was also used in traditional medicines. Gum, when chewed, can fight colds [4], inner bark, soaked and softened, has been used as a poultice to heal wounds [5] and leaves were used as a fuminant to revive comatose patient and to clear congested OPEN ACCESS lungs [6]. The pine oil and pine tar have been also used to make disinfectants, antiseptics and insecticides [7]. Numerous studies realized on bark, wood, needles, resin and essential oils of Pinus banksiana have reported the presence of monoterpenoids [8,9], diterpenoids [8][9][10][11], sesquiterpenoids [8,12], triterpenoids [11], phenylpropanes [13], flavonoids, lignans and stilbenes [10,14]. However, to this day, no study of the cytotoxic and antibacterial activities of the buds of jack pine was made.

Results and Discussion
P. banksiana buds were extracted successively with hexanes, CH 2 Cl 2 and MeOH. After solvent evaporation, each extract was investigated for in vitro cytotoxic and antibacterial activities. Cytotoxic activity evaluations were carried out on lung cancer (A549), colorectal cancer (DLD-1) and normal skin fibroblasts (WS1) human cell lines using the Hoechst assay [26]. Antibacterial activity was evaluated against Escherichia coli and Staphylococcus aureus. Results displayed in Table 1 show that the hexanes extract exerted a moderate activity against A549 (IC 50 , 45 ± 4 µg/mL), DLD-1 (IC 50 , 44 ± 3 µg/mL) cell lines and its antibacterial activity against S. aureus was interesting (IC 50 , 29 ± 3 μg/mL). The CH 2 Cl 2 extract showed a significant cytotoxic activity against A549 (IC 50 , 26 ± 3 µg/mL), DLD-1 (IC 50 , 32 ± 3 µg/mL) cell lines and a weak activity on S. aureus (64 ± 5 µg/mL). Finally, the MeOH extract exhibited no activity. Therefore, following studies were focused on the CH 2 Cl 2 extract that was further separated over column chromatography on silica gel to afford eleven fractions (A-K). All fractions activities were tested against both cancer cell lines and bacterial strains. Only fraction F, G and H were found strongly cytotoxic activity against A549 with IC 50 values ranging from 5 to 7 µg/mL and against DLD-1 with IC 50 values ranging from 12 to 14 µg/mL. Interestingly, these fractions were found selective against cancer cell lines in comparison with normal cells (IC 50 , 39 to 79 µg/mL). Remaining fractions were found inactive with IC 50 values greater than 100 µM. Fraction F was active against S. aureus (IC 50 , 51 ± 2 µg/mL) while fractions G, H and K were found inactive (IC 50 > 100 µM). In a next step, fractions F, G, H and K were separated by a combination of chromatographic procedures to afford the three new compounds 1-3 together with 12 known compounds. The structures of the new compounds were determined as follows and the known products were identified by comparison of their spectroscopic data with values found in the literature.  Figure 2) observed between H 3 -19 and C-3, C-4, C-5 and C-18 (δ C 36.3, 46.9, 39.6, 182.3, respectively), together with the presence of an aromatic ring, suggested that the molecule was a dehydroabietane [17]. The hydroxyl group was assigned at position 7 because of its 1 H-1 H COSY correlation with H-6 (δ H 1.72 and 2.14, both overlapped) (see Figure 2).    The multiplicity of H-7 was a broad doublet (J = 3.5 Hz) suggesting that the configuration is α. The configurations of other stereocenters were assessed by comparing the 13 C chemical shifts of 1 with those of 7α-hydroxydehydroabietic acid [17]. Hence, compound 1 was established as 7α-hydroxyabieta-8,11,13,15-tetraen-18-oic acid.
Compound 2 was assigned the molecular formula C 20 H 30 O 3 , as established from its HRESIMS (m/z 341.2077 calcd for C 20 H 30 O 3 Na + 341.2087). The IR spectrum showed absorbances consistent with hydroxyl (3411 cm −1 ) and olefinic (1650 cm −1 ) groups. The 1 H-NMR spectrum of 2 (Table 2) Table 3) were observed at δ C 72.5 (C-15), 70.9 (C-7) and 71.6 (C-18) for tertiary, secondary and primary alcohol groups respectively, along with six other signals assigned to an aromatic ring. The methyl groups Me-16 and Me-17 showed HMBC correlations with C-15 and C-13 (δ C 146.5), indicative of a hydroxypropyl branched at C-13 (see Figure 2). Long-range correlations were observed between H 3 -19 and H 2 -18 with C-3, C-4 and C-5 (34.7, 37.5, 42.3, respectively) suggesting that the hydroxyl function of the primary alcohol was attached to C-18. The β-configuration of the hydroxyl group at C-7 was determined from the 1 H-NMR spectrum, in which a triplet at δ H 4.86 (J = 8.9 Hz) due to an axial proton was observed [18]. Cross peaks in the NOESY spectrum (see Figure 3) were observed between H 3 -19 and H 3 -20 and between H 2 -18 and H-5, supporting the relative stereochemistry depicted. Thus, compound 2 was identifed as 7β,15,18-trihydroxyabieta-8,11,13-triene.  Figure 2), indicating the position of the alcohol group. The 13 C signal at δ C 155.5 suggested the presence of a phenolic group. The location of this hydroxyl group at C-13 was confirmed by observation of C-H long range correlations from H-11 to C-8, C-10 and C-13 in the HMBC spectrum. The presence of the two alcohols was confirmed by the absorption bands at 3448 cm −1 in the FTIR spectrum of 3. These data, together with other results of COSY and HMBC analysis (see Figure 2), confirmed that compound 3 is 13,15-dihydroxypodocarpa-8,11,13-triene.
Cytotoxic and antibacterial activities of all compounds were evaluated against two human cancer cell lines, lung carcinoma (A549) and colorectal adenocarcinoma (DLD-1) and two bacterial strains, S. aureus and E. coli (Table 4). With the exception of compounds 12 and 13, all compounds were found inactive against both cancer cell lines tested, IC 50 > 100 µM. Compound 12 exhibited a moderate cytotoxicity against A549 and DLD-1 with IC 50 of 34 and 47 µM, respectively. Compound 13 was also found active against A549 (IC 50 , 46 µM) but not to DLD-1. These results are in good agreement with those reported by Barrero [27] for compound 12. As far as antibacterial activity was concerned, all compounds were inactive which is also in agreement with previously published results for compounds 7, 11, 12 and 15 [28][29][30]. Compounds 12 and 13 were in part responsible of the cytotoxic activity of the CH 2 Cl 2 extract of bud from P. banksiana. However, additional studies will be conducted to identify other active compounds.

General
Optical rotations were measured with an automatic polarimeter Rudolph Research Analytical Autopol IV. High resolution electrospray ionization mass spectrum was conducted in positive mode with an Applied Biosystems/MDS Sciex QSTARXL QqTOF MS system. FTIR spectra were recorded with a Perkin-Elmer SpectrumOne. The 1D and 2D NMR spectra ( 1 H-1 H COSY, HSQC and HMBC) were performed using an Avance 400 Bruker spectrometer equipped with a 5 mm QNP-probe. Chemical shifts were expressed in δ (ppm) units relative to TMS as an internal standard and coupling constants were given in Hertz. The analytical HPLC separations were performed using an Agilent 1100 series instrument fitted with a UV-Vis diode array detector and a MS detector Agilent G1946 VL together with an atmospheric pressure chemical ionization (APCI) source. Preparative HPLC was performed on an Agilent 1100 liquid chromatography system, equipped with a solvent delivery system, an autosampler and UV-MWD detector. The column configuration consisted of an Intertsil prep-ODS C18 column (6.0 × 250 mm; 10 µm) for analytical analysis and an Intertsil prep-ODS C18 column (20 × 250 mm; 10 µm) for preparative HPLC. Column chromatographic separations were carried out using silica gel (40-63 µm with indicator F 254 , Silicycle, Québec, Canada) and C 18 reversed phase silica gel (carbon 11%, 40-69 μm, Silicycle). High performance flash chromatography was performed using a HPFC-Analogix F12-40 system equipped with a silica gel column C18, 40 µM (silicycle, Québec, Canada). Analytical thin-layer chromatography was performed with silica gel 60 F 254 , 0.25 mm pre-coated TLC plates (Silicycle). Diterpene compounds were detected by spraying TLC plates with vanillin-sulfuric acid reagent followed by heating at 110 C. The yields were calculated from the weight of dry plant material.

Plant Material
Buds of Pinus banksiana were collected in the boreal forest of the Saguenay region (Quebec, Canada) in May 2007. The specimen was identified by Patrick Nadeau (Université du Québec à Chicoutimi) and a voucher specimen (QFA-0540468) was deposited at the Herbarium Louis-Marie of Université Laval, Québec, Canada.

Cytotoxicity Assay
Exponentially growing cells were plated on 96-well microplates (BD Falcon) at a density of 5 × 10 3 cells per well in 100 μL of culture medium (DMEM with 10% SVF) and were allowed to adhere for 24 h before treatment. Increasing concentrations of each compound in MeOH or DMSO were then added (100 μL per well) and cells were incubated for 48 h. The final concentration of MeOH or DMSO in the culture medium was maintained at 0.25% (v/v) to avoid solvent toxicity. Microplates were then emptied and stored at −80 °C for 24 h. In a next step, 100 μL of SDS (0.01%) were added and the microplates were incubated at room temperature during 3 h before being put back to the cold. After 24 h, cells were prepared for cellular DNA assay with 100 μL of Hoechst dye 33342. Measurements were performed on the same labsystems at 365 and 460 nm wavelengths. Survival percentage was defined as the fluorescene in experimental wells compared to the control wells after subtraction of the blank values. Etoposide was used as positive control. Each experiment was carried out two times in triplicata. IC 50 results were expressed as averaged values and the corresponding standard deviations were computed.

Antibacterial Assay
Antibacterial activity was evaluated using the microdilution method described by Banfi et al. 31 with some modifications. Exponentially growing bacteria were plated in 96-well flat bottom microplates (BD Flacon) at a density of 5 × 10 3 gram-negative E. coli (ATCC 25922) or 40 × 10 3 gram-positive S. aureus (ATCC 25923) per well in 100 μL nutrient broth (Difco). The concentration of ethanol in the culture medium was maintained at 0.25% (v/v) to avoid solvent toxicity. Thereafter, 50 μL of 4% resazurin was added to each well and the microplates were incubated for 6 h at 37 °C. Fluorescence was measured after 6 h with an automated 96-well Fluoroskan Ascent Fl™ plate reader (Labsystems) using 530 and 590 nm excitation and emission wavelengths.