Chemical composition of the Brazilian native Cinnamomum stenophyllum (Meisn.) Vattimo-Gil essential oil by GC-qMS and GC × GC-TOFMS, and its cytotoxic activity

2.1 Materials

All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA). Leaves and stems from Cinnamomum stenophyllum (Meisn.) Vattimo-Gil were collected at Arujá (23°23′47″S and 46°19′17″W), São Paulo, Brazil, in the Atlantic Rainforest area. The plant material was identified by Dr. Sueli Nicolau (Instituto Botânico, São Paulo, Brazil). Voucher specimen (Cinammomum stenophyllum area 615 Arujá) were deposited in the Herbarium of the same institution. The leaves were separated and dried at room temperature until constant weight.

2.2 Oil extraction

Leaf essential oil was obtained by hydrodistillation using a Clevenger-type apparatus. The extraction was carried for 4 h and the oil was dried over anhydrous sodium sulfate and stored in a freezer (−20 °C) until further use (Machado et al., 2017). The essential oil yield (0.144%, w/v) was calculated based on the dry weight.

2.3 GC-qMS analysis

The leaf essential oil from C. stenophyllum was diluted with dichloromethane (1:10 v/v) and a single aliquot of 1 µL was analyzed using an Agilent 6890 Series GC apparatus (Agilent Technologies, Santa Clara, CA, USA) with a fused silica capillary column (DB-5, 5%-phenyl-95%-methylsiloxane, 30 m × 0.25 mm i.d. × 0.25 μm film thickness) hyphenated with an electron ionization system 5973 quadrupole MS detector (Agilent Technologies, Santa Clara, CA, USA) operating at 70 eV. The detector temperature was 250 °C, scan time of 0.1 scan/sec and acquisition mass range of m/z 35–500. Carrier gas helium (99.999% purity) was used at a flow rate of 1 mL/min. The sample was injected using split less mode. The injector temperature was set at 250 °C and the oven temperature was programmed from 40 °C (1 min) to 240 °C at 3 °C/min. The essential oil components were identified by comparing their retention index (RI), calculated about in a series of n-alkanes (C₆-C₄₀), and by comparison of their mass spectra with three commercial libraries: HPCH2205 (Adams, 2007), NIST 08/Wiley (2014). For peak assignment, only peaks with similar mass spectra higher than 80% and RI difference (RI calculated – RI literature) of ±10 unit were tentatively identified (Tranchida et al., 2013).

2.4 GC × GC-TOFMS analysis

For the GC × GC analysis of the leaf essential oil (1 μL) from C. stenophyllum a GC × GC-TOFMS system (Model Pegasus 4D, Leco, St. Joseph, MI, USA) was used. This equipment is composed of a gas chromatograph (Model 7890, Agilent Technologies, Palo Alto, CA, USA) equipped with a secondary oven, a non-moving quad-jet and dual-stage cryogenic modulator and a time-of-flight mass spectrometer (TOFMS) (Model Pegasus 4D, Leco, St. Joseph, MI, USA). The primary (1D) and secondary (2D) dimension columns were a non-polar (DB-5) and a short mid-polar (BPX 50, 50%-phenyl-50%-methylsiloxane), respectively. The 2D column and the TOFMS were connected by a 0.5 m × 0.25 mm i.d. uncoated deactivated fused silica capillary via SGE-mini-unions and SilTite^TM metal ferrules (0.1–0.25 mm i.d.) (SGE, Ringwood, VIC, Australia). A single sample was injected into GC × GC inlet using split mode (1:10). Carrier gas helium (99.9999% purity) was used at a flow rate of 1 mL/min. Modulation was carried out every 6 s, with the duration of the hot pulse of 2.5 s. For the chromatographic run, the primary oven was programmed as 40 °C (1 min), at 3 °C/min to 150 °C, then at 2 °C/min to 270 °C. The secondary oven was maintained at 10 °C above primary oven temperature. Mass spectrometer source and transfer line temperature were kept at 230 and 280 °C, respectively. The electron ionization was at 70 eV and the spectral acquisition rate was of 100 spectrum/s with full scan mode using a mass range of 40–450 Dalton.

GC × GC-TOFMS data acquisition and processing were performed using ChromaTOF® software (4.51.6.0, Leco, St. Joseph, MI, USA). After the data acquisition, the peaks that showed a signal-to-noise ratio above 1000: 1 were selected for processing. The areas and the mass spectra of the individual peaks were obtained. The compound identification was performed using the Mass Spectral Library NIST (NIST MS Search Program version 2.0) and Adams (2007), and the comparison of calculated 1D linear retention indices (related to the same C₆-C₄₀ series of n-alkanes used in GC–MS) with literature values. After the analysis, only peaks with similar mass spectra higher than 80% and RI difference (RI calculated – RI literature) of ± 10 unit were tentatively identified (Tranchida et al., 2013).

2.5 Cytotoxic assay

The evaluation of the cytotoxic activity from C. stenophyllum essential oil was obtained through the colorimetric MTT assay (Mosmann, 1983; Costa-Lotufo et al., 2010). Three cell lines were used in this study: Colon adenocarcinoma (HCT-116), Breast cancer carcinoma (MCF-7) and the non-tumor retinal pigment epithelial cells (RPE). They were maintained in RPMI 1640 (HCT-116 and RPE) or DMEM Glutamax (MCF-7) medium supplemented with 10% fetal bovine serum (v/v), 2 mmol/L glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin. Cells were cultured at 37 °C under a 5% CO₂ atmosphere. After incubation, they were seeded into a 96-well microplate at a density of 5 × 10⁴ cells/mL and cultured for 24 h in the proper culture medium. Cells were then treated with different concentrations of each sample and incubated for 72 h. Three hours before the end of treatment, 150 μL of MTT (5 mg/mL) was added into each well and then incubated for an additional 3 h. The absorbance intensity was measured at 595 nm using a multi-well scanning spectrophotometer (Multiskan FC, Fisher Scientific, USA).

The IC₅₀ experiments included triplicates for each 5-fold dilution and two independent assays. Final concentrations ranged from 0.002 to 50 μg/mL. Doxorubicin (0.001–5.43 μg/mL) was used as positive control. DMSO served as a negative control. The cell viability was calculated based on the control cells.

3.1 Chemical analysis

Initially, the essential oil was subjected to GC-qMS analysis. The GC–MS chromatogram showed a chromatogram with many coelutions resulting in a very low number of compounds identified. Of the 41 detected peaks, only 7 analytes could be identified with the described method, belonging to the following chemical groups: ketone (4-hydroxy-4-methyl-pentanone, 1), oxygenated monoterpenes (nopinone, 2; p-mentha-1,5-dien-8-ol, 5 and verbenone, 7), phenylpropanoid (methyl eugenol, 9) and oxygenated sesquiterpenes (spathulenol, 17; mustakone, 33) (Fig. 1, Table 1). In previous works with the leaf essential oils from Asian Cinnamomum species the main constituents identified were phenylpropanoids (eugenol, safrol and cinnamoyl derivatives), benzenoids (benzyl benzoate) and oxygenated monoterpenes (1,8-cineole and camphor) (Simic et al., 2004; Jantan et al., 2004, 2008). Although most of the detected compounds could not be identified in the Brazilian species, no cinnamoyl derivatives were observed. Despite this difference, results obtained with the GC–MS technique precluded further consideration, since most of the compounds remained unassigned probably due to the high co-elution rate.

Close

Fig. 1 Chromatographic profile of the C. stenophyllum leaf essential oil by GC-qMS (see Table 1 for peak assignment).

Export to PPT

Although separation of essential oil components is mainly accomplished by GC–MS, it is not always easy to obtain a reliable mass spectrum, without mutual interference of other compounds (Han et al., 2018), due to the large number of unresolved peaks, mainly when using quadrupole analyzer and with low resolution mass spectrometer. This fact explains the low rate of identification in the GC–MS analysis (Fig. 1). For this reason, GC × GC-TOFMS technique was used to obtain a more detailed characterization of the C. stenophyllum leaf essential oil. The use of two columns, on the basis of volatility (first dimension, DB-5) and polarity (second dimension, BPX 50) increased the chromatographic resolution, allowing a better separation of the analytes. In addition, the modulation used in the GC × GC system provided an increased sensibility and peak capacity, due to the accumulation and rapid release of the compounds (Lebanov et al., 2019).

The structured bidimensional GC × GC chromatogram allowed a better visualization of the compound classes present in the C. stenophyllum oil (Fig. 2A), indicating an abundance in oxygenated mono- and sesquiterpenes. Finally, the 3D chromatogram obtained allowed to observe the relative concentrations of the compounds and their better separation (Fig. 2B).

Close

Fig 2 GC × GC-TOFMS total ion chromatogram (TIC) of the C. stenophyllum leaf essential oil represented as a topographic map, highlighting the main classes identified (A) and a tridimensional view (B).

Export to PPT

The GC × GC-TOFMS analysis allowed an increase in the number of peaks detected, 213 in total. However, despite the high chromatographic resolution power of the comprehensive two-dimensional gas chromatography, only 80 compounds could be identified (Table 1). There were five main chemical classes identified by GC × GC-TOFMS system in the essential oil: ketones (8 compounds), monoterpene hydrocarbons (7 compounds), oxygenated monoterpenes (30 compounds), sesquiterpene hydrocarbons (4 compounds) and oxygenated sesquiterpenes (23 compounds) (Table 1).

It is interesting to note that, due to the increase of the GC × GC chromatographic resolution, α-terpineol (43) was identified with 93% of similarity while in the one-dimensional GC analysis it was assigned with 58% of similarity. Background interferences makes GC–MS analysis less reliable. Furthermore, some compounds, myrtenol (44, ¹t_R 1680 s, ²t_R 2.63 s) and myrtenal (45, ¹t_R 1680 s, ²t_R 2.91 s) for example, were resolved only in the second dimension (Fig. 3).

Close

Fig 3 Oxygenated monoterpenes identified in the total ion chromatogram and their deconvoluted mass spectra: α-terpineol (43), myrtenol (44), and myrtenal (45).

Export to PPT

Despite the better resolution and many compounds detected and identified, there was still some compound coelution, where many of them could be resolved using the mass spectra deconvolution. In the present case, a particularly problematic coelution was observed in the elution region for oxygenated sesquiterpenes (Fig. 4). These compounds are known to have identical or nearly identical mass spectra (Shellie et al., 2004; Zoccali et al., 2015). As displayed in Fig. 4, GC × GC-TOFMS analysis showed that three compounds overlapped in the first dimension and even after modulation and the second-dimension elution these peaks were not resolved. With the mass spectra deconvolution, the identification of the oxygenated sesquiterpenes was still not clear due to the similarities in the MS spectra of 146 and 148. Therefore, the individual tentatively identification of 10-epi-γ-eudesmol (146), selin-6 -en-4-ol (147) and γ-eudesmol (148) was possible only by comparing their RI with libraries reference data.

Close

Fig 4 Separation of the oxygenated sesquiterpene region of leaf essential oil from C. stenophyllum using GC × GC technique. The amplified region still shows co-elution of compounds for first and second dimensions.

Export to PPT

The composition of the essential oils from Asian Cinnamomum species has been widely studied, but to our knowledge this is the first analysis of a native Brazilian one. For those species [C. zeylanicum, C. cassia, C. tamala (Buch.-Ham.) T.Nees & Eberm., C. burmannii, C. pauciflorum Nees, C. rhynchophyllum Miq., C. cordatum Kosterm., C. microphyllum Ridl., C. scortechinii Gamble, C. pubescens Kochummen, C. impressicostatum Kosterm., C. mollissimum Hook.f., C. camphora (L.) J.Presl], the leaf and bark oils contained mainly cinnamoyl derivatives, benzyl benzoates, oxygenated monoterpenes and other phenylpropranoids (Jantan et al., 2008; Wang et al., 2009; Unlu et al., 2010; Jiang et al., 2016; Barros et al, 2016), while in C. stenophyllum oil the majority were oxygenated mono and sesquiterpenes.

Considering first the cinnamoyl derivatives, it is interesting to note that trans-cinnamaldehyde, cinnamyl alcohol or methyl-cinnamate are found in almost all the leaf oils of the above-mentioned Asian species, with the exception of C. microphyllum, C. mollissimum and C. camphora, in which benzyl-benzoate and camphor were the major compounds (Jantan et al., 2008; Jiang et al., 2016). However, it is important to notice that none of these compounds was found in the C. stenophyllum leaf oil. The only neotropical Cinnamomum species studied [C. amoenum (Nees & Mart.) Kosterm.] also did not present the cinnamoyl derivatives, benzyl-benzoate and camphor in leaf oil (Maciel et al., 2019).

In contrast, the number of oxygenated monoterpenes identified for the C. stenophyllum oil (30 compounds) by the GC × GC technique was higher than that reported for those species, previously studied only by GC–MS, with only 3–13 compounds within this class (Jantan et al., 2008; Wang et al., 2009; Jiang et al., 2016). Among these, 1,8-cineole, also detected in C. stenophyllum oil, was also found in the leaf oils from C. rhyncophyllum, C. mollissimum, C. camphora and C. glanduliferum (Wall.) Meisn., even as the major component for the latter (Jantan et al., 2008; Singh et al., 2014).

Previous studies on Cinnamomum spp. essential oils indicated a smaller variety of oxygenated sesquiterpenes, with only 1–14 compounds for the Asian species (Jantan et al., 2008; Wang et al., 2009; Jiang et al., 2016), compared to the current study where 23 of these compounds were found after the GC × GC-TOFMS. The predominance of oxygenated terpenes in C. stenophyllum essential oil suggests cytotoxic effects since these compounds are known for their pro-oxidant effects at the cellular level (Sharma et al., 2013).

3.2 Cytotoxic assay

Cytotoxicity studies have shown that essential oils can be active, generally without causing mutagenicity in the treated organisms (Bakkali et al., 2008). Considering that selective cytotoxicity for tumor cells is a characteristic of interest for potential antitumor agents, the cytotoxic activity of C. stenophyllum oil was evaluated for two cancer cells, MCF-7 (breast) and HCT-116 (colon), and compared to that observed using the non-tumor retinal pigment epithelial cells (RPE). The two tumor cell lines are part of our initial panel for investigation of cytotoxic activity because those cells have some characteristics that make them useful as experimental models, little aggressive tumors, noninvasive and low metastatic potential in MCF-7 and the invasive ability and high sensitivity of HCT-116 (Rajput et al., 2008; Comşa et al., 2015).

As shown in Table 2, the leaf essential oil was toxic against the two tumor cell lines, with IC₅₀ values lower than 20 μg/mL, threshold to be considered as active (Bézivin et al., 2003). On the other hand, towards non-tumor cells (RPE), the oil presented IC₅₀ > 50 μg/mL, which demonstrates a selectivity for the cancer cells, i.e. some of its components might be acting in specific metabolic pathways from these cells (Unlu et al., 2010).

C. cassia and C. zeylanicum bark oils have been previously described to be cytotoxic against leukemia, colon, prostate, liver cancer among others. However, their mechanism of action is not fully elucidated (Dutta and Chakraborty, 2018). There is strong evidence that C. zeylanicum bark essential oil promotes apoptosis in cancer cell lines through different mechanisms. Especially in skin cancer models, the downregulation of EGFR tyrosinase kinase, an important signaling way for the growth and differentiation of the cells, was reported after treatment with this oil (Han and Parker, 2017). The bark extract from C. cassia, containing trans-cinnamic acid and cinnamaldehyde, decreased levels of tumor-associated growth factors such as EGF, FGF and VEGF, in a mouse melanoma model system (Kwon et al., 2009). Additionally, it promoted inhibition of the factors NFkappaB and AP1, key in tumorigenesis, and their target genes such as Bcl-2, BcL-xL and survivin, anti-apoptotic conductors (Kwon et al., 2010).

This cytotoxic activity of the bark essential oil from C. zeylanicum was associated with the presence of some components such as limonene, geraniol and 2′-benzyloxy-cinnamaldehyde (Unlu et al., 2010). Furthermore, cinnamaldehyde, cinnamic acid, 2-hydroxycinnamaldehyde and eugenol were also evaluated in different cancer models displaying cytotoxicity, and their mechanism of action involves apoptosis induction as evidenced by the increased levels of cleaved PARP and cleaved caspase-3. Other mechanisms such as induction Nrf2 activity, DNA damage and ROS production has also been reported for these compounds (Larasati and Meiyanto, 2018). Although, the abovementioned compounds were not found in the C. stenophyllum oil and could not be accounted for the observed cytotoxicity. For instance, some of detected oxygenated monoterpenes have also been implicated with antitumor activities, such as linalyl acetate, α-terpineol, borneol, carvacrol, 1,8-cineole, carvone (Sobral et al., 2014). Additionally, two of the oxygenated sesquiterpenes, β-eudesmol and spathulenol, were also reported as having cytotoxic activity (Bomfim et al., 2016; Kotawong et al., 2018). Thus, the presence of this compounds could explain the observed cytotoxicity for C. stenophyllum leaf oil.