Subscriber access provided by KEAN UNIV
Article
Tandem mass spectrometric quantification of 93 terpenoids in Cannabis using static headspace (SHS) injections Anna Shapira, Paula Berman, Kate Futoran, Ohad Guberman, and David Meiri Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b02844 • Publication Date (Web): 01 Aug 2019 Downloaded from pubs.acs.org on August 1, 2019
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
Tandem mass spectrometric quantification of 93 terpenoids in Cannabis using static headspace (SHS) injections Anna Shapiraa,‡, Paula Bermana,‡, Kate Futorana, Ohad Gubermana, and David Meiria,* a The
Laboratory of Cancer Biology and Cannabinoid Research, Department of Biology, Technion-Israel Institute of Technology, Haifa 320003, Israel. ABSTRACT: The therapeutic effect of Cannabis largely depends on the content of its pharmacologically active secondary metabolites, mainly phytocannabinoids, flavonoids and terpenoids. Recent studies suggest of therapeutic effects of specific terpenoids, as well as synergistic effects with other active compounds in the plant. Although Cannabis contains an overwhelming milieu of terpenoids, only a limited number are currently reported and used for metabolic analysis of Cannabis chemovars. In this study, we report the development and validation of a method for simultaneous quantification of 93 terpenoids in Cannabis air-driedinflorescences and extracts. This method employs the full evaporation technique via a static headspace sampler, followed by gas chromatography–mass spectrometry (SHS-GC–MS/MS). In the validation process, spiked terpenoids were quantified with acceptable repeatability, reproducibility, sensitivity and accuracy. Three medical Cannabis chemovars were used to study the effect of sample preparation and extraction methods on terpenoid profiles. This method was further applied for studying the terpenoid profiles of sixteen different chemovars acquired at different dates. Our results demonstrate that sample preparation methods may significantly impact the chemical fingerprint compared to the non-treated Cannabis. This emphasizes the importance of performing SHS extraction in order to study the natural terpenoid contents of chemovars. We also concluded that most inflorescences expressed relatively unique terpenoid profiles for the most pronounced terpenoids, even when sampled at different dates, although absolute concentrations may vary due to aging. The suggested method offer an ideal tool for terpenoid profiling of Cannabis and set the scene for more comprehensive works in the future.
Terpenes and terpenoids are naturally-occurring hydrocarbon biomolecules that constitute an extremely diverse family of compounds that is present in many species including plants, animals and microorganisms.1 Both compound families are volatile, natural, complex biomolecules that are often characterized by a strong odor. Terpenoids are modified terpenes, generally differing from terpenes by a varying oxygen rearrangement or oxidation state. Hence, some sources include both groups of terpenes and terpenoids under the general term “terpenoids”.2 These are classified according to the number of repeating isoprene units (five-carbon building blocks), e.g. mono-, sesqui-, di- and tri-terpenoids, all consist of skeletons of 10, 15, 20 and 30 carbons, respectively. Terpenoids are ubiquitously found throughout the plant kingdom, especially in aromatic plants.1 One such plant is Cannabis sativa (Cannabis) with an overwhelming milieu of approximately 200 identified terpenoids at various concentrations.2,3 These are biosynthesized as secondary metabolites mainly in glandular trichomes which are abundant on the surface of the female inflorescence.4,5 Terpenoids are not unique to Cannabis, and in fact are the same compounds as those expressed in other plants. Terpenoids significantly influence Cannabis inflorescence aroma and participate in protection from predation, attraction of pollinators, and many other roles.6 They have also been suggested to possess synergistic effects with other therapeutic compounds in Cannabis, as well as therapeutic effects of their own.2 In a recent publication, Russo and Marcu6 review the available literature on the known pharmacology of 27 of the identified
terpenoids in Cannabis. Their physiological mechanisms are attributable to their lipophilicity, and include effects on ion channels, neurotransmitters, odorants, and tastant receptors, among others.6 Among the vast number of terpenoids in Cannabis, only approximately 20 to 30 terpenoids have been reported and used for chemotyping of Cannabis cultivars.7-15 Hundreds of additional terpenoids and other volatiles have been recently identified in Cannabis,16-18 but their absolute concentrations have never been determined using a validated analytical method. Due to their high volatility, terpenoids are usually analyzed by gas chromatography (GC) coupled to different detectors, mainly flame ionization detector (FID) or mass spectrometry (MS).19 In the case of terpenoids, a suitable sample preparation method should be carefully chosen, due to their high volatility. Extraction techniques often used for this purpose include static headspace (SHS), purge and trap, solid-phase microextraction (SPME), and solvent (liquid) extraction.20 Among these, liquid extraction by a variety of solvents including ethanol,7 npentane,8 petroleum ether21 and chloroform22 is most common. These extraction methods are multi-step, apply environmentally-hazardous solvents, and may include inducing heat, which can alter the terpenoid composition compared to the natural content in the plant. SHS, on the other hand, exhibits several advantages for extraction of terpenoids compared to other extraction methods. The theory of SHS extraction is thoroughly described,23 and was more recently summarized by Snow and Bullock.24 Among
ACS Paragon Plus Environment
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
others, SHS advantages include straightforward sample preparation (although, aliquots and products of liquid extraction may be sampled using SHS as well), minimal contamination of the GC-MS as a result of non-volatile components entering the inlet, and automation. It is effective for rapid analysis of bulk drug substances and herbal formulations.25 In combination with instrument sensitivity, these make SHS the technique of choice for the quantification of volatiles in many sample types, Cannabis included, and over a wide range of concentrations. The main disadvantage of SHS analysis is a possible matrix effect in the analysis of complex samples compared to calibration standards. Matrix effects can influence the established equilibrium between the condensed and vapor phases in a HS vial, giving rise to different response factors for a certain compound depending on the matrix. This can be resolved by applying the full evaporation technique (FET).26 This method involves dispensing a small amount of sample into a HS vial, and heating the vial to relatively high temperatures for short periods of time in order to fully evaporate the sample, hence removing the equilibration of the condensed and vapor phases. The most widely used method for Cannabis terpenoid analysis throughout the literature is liquid solvent extraction followed by GC-FID,7-9 primarily due to the low cost of instrumentation. GC-MS, on the other hand, may be more expensive and complex to operate, but provides a much needed improvement of identification, selectivity and sensitivity. This method of detection was used by Rossi and ElSohly27 and Ferioli et al.21 Others, like Hillig,22 performed most of their work using FID while relying on MS for verification of unknown compounds. Recently a GC/MS method was validated for quantification of the ten most abundant terpenoids in Cannabis.28 Also, several HS-SPME-GC-MS methods have been recently reported for comprehensive profiling of terpenoids in Cannabis.16-18 HS-SPME, however, is considered less robust and precise than SHS, as terpenoid concentrations injected may be affected by the nature of the fiber selected for the extraction.17 Also, SPME methods are often applied for improving signal intensity in diluted samples. This is clearly not the case in Cannabis inflorescences as reports suggest of a total terpenoid content of approximately 0.5 to 3.5% (w/w).9 In this study, we developed and fully validated for the first time, a novel SHS-GC-MS/MS method for the purpose of simultaneously quantifying 93 terpenoids in Cannabis inflorescences. Experimental section Reagents and materials Unless otherwise stated, pure analytical standards (>95 %) were purchased from Sigma-Aldrich (Germany). (+)Valencene, and α- and β-curcumene were 80 and 90 % pure, respectively. α- and (-)-β-pinene, camphene, β-myrcene, (-)-αbisabolol, 3-carene, β-caryophyllene, geraniol, (-)-guaiol, αhumulene, p-cymene, (-)-isopulegol, d-limonene, linalool, trans-nerolidol, α- and β-ocimene, - and γ-terpinene, terpinolene were purchased from Restek (U.S); (+)-2-carene, (R)-(-)-phellandrene, azulene, 1,4-cineole, (-)-citronellal, carvacryl acetate, α- and β-curcumene, eucalyptol, (+)longifolene, isolongifolene, sabinene, were purchased from Extrasynthese (France). Ethanol, 2-propanol, methanol and hexane for standard solutions and sample preparation were LCMS grade (Merck KGaA, Germany).
In this study, we aimed to acquire and test the same types of products which were also distributed to patients. Therefore, airdried medical Cannabis female inflorescences grown in greenhouses were obtained from several licensed local medical Cannabis distributors in sealed bags with 12-14% w/w moisture. Standard and sample preparation Separate groups of terpenoid mixtures consisting of 17–20 analytes each were prepared by accurately weighing and mixing 10 mg of each standard in 1 ml ethanol. For construction of calibration curves, one complete stock with all the terpenoids was prepared by mixing 100 µl aliquots from each group to achieve a final concentration of about 200 µg/ml. Ten-point calibration curves were prepared in the range of approximately 0.2–100 μg/ml by diluting the stock standard mix in ethanol. 10 µl of each concentration were used for the calibration curves. Cannabis air dried inflorescences from different chemovars were grinded to a fine powder using an electrical grinder. For quantification of terpenoids, 5 mg of the powder were accurately weighed in duplicates. Sample preparation for comparison of decarboxylation, evaporation and extraction methods appear in Method S1. A schematic representation of the iso-butane extractor and operation mode appear in Figure S1. For SHS-GC-MS/MS analysis, 5 mg of each of the extracts were accurately weighed in duplicates. All SHS samples were prepared in 20 ml amber rounded bottom HS-vials sealed with a magnetic 32 mm PTFE septa cap by a crimper. Instruments and GC-MS/MS Analysis Injections for the optimization of the MS parameters were performed in liquid injection mode at a 1:10 split. The injector temperature was set at 250 ºC and the helium flow rate was 0.9 ml/min. Terpenoid standard mixtures were prepared at approximately 25 µg/ml for each analyte, and 1 µl from each mixture were injected into the liquid injection port. In SHSinjection mode, a CTC autosampler (Pal RTC, CTC analytics, Switzerland) was used with a Headspace static tool in splitless mode. The incubation parameters were set as follows: 40 min incubation time in agitator mode at a temperature of 140 ºC and speed of 250 rpm. The injection volume was 600 µl. Terpenoid analyses were performed on a Trace 1310 GC (Thermo scientific, Germany) coupled to a TSQ 8000 Evo triple-quadrupole MS (Thermo scientific, Germany) equipped with a DB-35MS UI capillary column (30 m × 0.25 mm x 0.25 μm, Agilent, US). The oven temperature program was set as described in Table S1. The MS was operated in electron impact (EI) ionization mode. The mass range analyzed by the MS was 35-250 amu. Identification of all terpenoids was performed by comparing the retention times (RTs) and protonated masses of the chromatographic peaks with those of the analytical standards analyzed under the same conditions. Kovats retention indexes (RIcalc) for all the standards were calculated as described by Babushok et al.,29 using a mixture of n-alkanes (C8-C20, Sigma R 8769, Saint Louis, MO, USA) run under the same chromatographic conditions. Quantification was performed by single reaction monitoring (SRM) mode. The collision gas was argon. Identification and
ACS Paragon Plus Environment
Page 2 of 9
Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry confirmation of terpenoid compounds was performed by spectral searching against the NIST library (version 2.2). Method validation Linear ranges of calibration curves were determined empirically according to the weighted least-squares linear regression method with a weighting factor of 1/X. Limits of detection (LODs) and quantification (LOQs) were calculated according to Eq. 1: Eq. 1
temperatures and times of incubation. Maximum intensities for each sample in all analyzed conditions were found ideal for the extraction. Equilibrium temperature and time of incubation were found to be 140 ºC and 40 min, respectively, as shown in Figure S2. According to this figure, no thermal degradation of terpenoids occurred in the chosen optimal parameters.
σ
LOD LOQ = F ∙ , S
where F is a factor of 3.3 and 10 for LOD and LOQ, respectively; σ is the standard deviation of the y-intercept; and S is the slope of the linear plot. All standard mix solutions were stored at 4-8 ºC in amber crimp-neck vials. The mixture consisting of all the terpenoid standards was analyzed once a week in two concentrations to verify its stability. Whenever a decrease in over 20% of any of the signals occurred the terpenoids mixture was freshly prepared. Selectivity of terpenoids was determined by calculation of resolution for adjacent terpenoids with similar fragment peaks in the SRM analysis. Acceptable values for the resolution were considered as ≥ 1.5.30 The precision of the whole method was assessed by calculating the relative standard deviations (RSDs) for repeatability and reproducibility on the same Cannabis sample. Repeatability was quantified by intra-day variation by analyzing three samples from the same chemovar on the same day (n=3); and reproducibility by inter-day variation, analyzing the same sample in triplicate on three different days (n=9). Precisions of less than 20% were considered acceptable. Accuracy was determined by spiking a Cannabis sample with three different concentration levels of the standard mixes within the linear range concentration of all the compounds. Each sample was prepared and injected in triplicates (n=3). Accuracy was determined as the percent difference between the mean concentration of the three analyses and the expected concentration. Results and discussion Development of the SHS-GC-MS/MS method The terpenoids included in this study (Figure 1) were according to the commercially available analytical standards of terpenoids that were either (a) previously identified in Cannabis16,17 or (b) were identified by spectral searching in several extracted Cannabis samples against the NIST library. The screen was performed using a single quadrupole MS on the total ion count (TIC) chromatogram. All the terpenoids in this research have one or more optical centers. Since we could not chromatographically separate diastereomers and/or enantiomers in the presented method, the molecules in Figure 1 and throughout this research appear as the sum of optical isomers. In order to avoid matrix effects, we applied in this study the FET method.26 To this end, we used a small amount of sample (10 µl and 5 mg of the standards and samples, respectively) and a high equilibration temperature in order to establish full evaporation of all compounds. Determination of the equilibrium conditions in the SHS system for terpenoids was performed by analyzing peak areas of 20 representative terpenoids in the standard mix and a characteristic Cannabis sample, at different
Figure 1. Molecular structures of all the terpenoid standards analyzed in this study. Compounds were classified according to functional groups.
Most of the GC columns reported in the literature for terpenoid analysis use non-polar stationary phase columns, which are considered best suited for many different types of compounds, and the standard for spectral matching and identification according to RI values in spectral libraries like NIST. These columns have also low bleed at high temperatures (approximately 320 ºC), which is usually required when performing liquid injections, in order to remove high boiling point compounds remaining on the column. Non-polar stationary phase columns, however, are not ideal for the analysis of terpenoids, which are unsaturated isobars with different types of isomerism that elute at very close RTs. We therefore chose for this method an intermediate-polar stationary phase column, in order to achieve the best possible resolution in a single run. In addition, the oven temperature program was optimized using a mixture of all the terpenoid standards. According to the observed RTs of the available terpenoid standards, the Kovats RIcalc values of each compound were calculated and summarized in Table S2. These were found to somewhat deviate from the reported values in the literature,29 probably due to the polar nature of the column used in this study (Table S2). The achieved peak separation for all the identified terpenoids in the standard mix and in a representative Cannabis sample
ACS Paragon Plus Environment
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
appear in the TIC chromatogram in Figures 2A and 2B, respectively.
Figure 2. TIC chromatogram of the identified terpenoids in the (A) standard mix and in (B) a representative Cannabis sample. Compounds in the Cannabis sample were identified by comparison of RTs, masses and MS/MS fragmentations as observed for the standards.
As shown, several additional unidentified peaks were observed in the TIC chromatogram of the Cannabis sample (Figure 2B) for which there were no analytical standards commercially available. These compounds and several others, which appeared also in numerous other Cannabis chemovars, were putatively identified by spectral searching against the NIST library, and by the closest RIcalc value as shown in Table S3. Some of these are noticeably large peaks, including selina3,7-diene, - and -eudesmol, β and γ-selinene, β-bisabolene, and others. However, since there were no analytical standards available for these compounds, we could not validate their extraction and quantification, and therefore they were not included as part of this research. According to the presented TIC, several peaks overlap (also for peaks for which there were no analytical standards commercially available). This emphasizes the uncertainty and inaccuracy in reported results when using FID and/or single quad MS detection. Therefore, in order to improve the selectivity for quantification of overlapping compounds and increase sensitivity by reducing the limits of detection and quantification, we chose to quantify terpenoids using the SRM mode. This mode enables to quantify compounds by choosing product ions from specific precursors, hence two overlapping compounds with identical masses can be separated using dissimilar product ions. For optimization of the SRM parameters, 1 µl from each of five different terpenoid standard mixtures were injected into the liquid inlet. The mixtures were created after finding the RT of each of the terpenoids separately to avoid peaks overlapping, which can interfere with the selection of product ions. Optimization of the chromatographic peak resolution was performed by injecting a single mixture of all the terpenoid standards through the HS injection port (Table S4, values are reported only for pairs of compounds with similar parent and product ions, which elute in the same RT window, and that exhibited resolutions below 3). The total runtime of the established method was 74 min. The peak assignments, chemical formulas, molecular weights (MWs), RTs, and optimized parameters for the SRM transitions of each of the identified terpenoids are listed in Table S5. Terpenoids in this table are ordered by increasing RT. Optimal precursor (Q1) and product ions (Q2), and collision energies
(CEs), for the SRM detection mode were determined in the 1030 eV range in order to find the most intense product ions for each precursor. Three different transitions were selected accordingly; one used for quantification and the two others for qualification (see Table S5). Validation of the SHS-GC-MS/MS method The validation parameters for the SHS-GC-MS/MS method are summarized in Table S6. Calibration curves for each component were constructed in the concentration ranges described in Table S6. All calibration curves were linear with excellent fits (R2>0.99 for all terpenoids) and less than 20% deviation from expected values in the concentration ranges studied for each compound. Two additional calibration mixes were prepared and injected in different days to test for repeatability of preparations. RSDs of the slopes and intercepts of all three calibration curves showed excellent repeatability (less than 9.96 and 12.96% RSD, respectively, Table S7). The terpenoid standard mixtures were freshly prepared every month according to the stability test. All the calculated LODs and LOQs were between 0.001– 0.123 and 0.002–0.374 µg/ml, respectively (Table S6). The precision of the whole method calculated as repeatability and reproducibility according to the intra- and inter-day variations, respectively, was lower than 11.4 and 21.9%, respectively (Table S6). The measured mean concentrations for each compound in the analyzed samples appear in Table S8. Accuracy was determined by spiking a Cannabis sample with three different concentration levels of the standard mixes as presented in Table S6 for each compound. All the terpenoids showed accuracies within 81.2–119.6%. Similar accuracies were obtained for injection of three concentrations of the standard mixes on different days (Table S9). Given that the accuracy of the whole method depends on the natural variability of terpenoids in the inflorescences and precision of preparations (spiking a small volume of standards onto 5 mg sample), the reported accuracies were considered acceptable. Overall these results indicate the good efficiency of the SHS extraction protocol. Effect of decarboxylation, evaporation and extraction method on terpenoid profile The sample preparation method, extracting solvent and additional steps like evaporation and decarboxylation can affect
ACS Paragon Plus Environment
Page 4 of 9
Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry the natural terpenoid profile of a Cannabis sample. Evaporation of the extraction solvent is commonly performed in the process of sample preparation, prior to analysis, for increasing the sensitivity of the analytes, or for replacing the extracting solvent. In the preparation of medical Cannabis concentrates, both complete evaporation of the solvent and decarboxylation of the inflorescences or the extracts by heat are required in order to produce the bioactive neutral phytocannabinoids like cannabidiol (CBD) and (-)-Δ9-trans-tetrahydrocannabinol (Δ9THC). In order to study the effect of sample preparation on terpenoid concentrations, we compared terpenoid profiles of selected type I-III Cannabis inflorescences after decarboxylation and evaporation, and extraction pretreatment in Figures 3A-3C and 4A-4C respectively. The decarboxylation conditions appear in Method S1. Following the suggested classification of medical Cannabis chemovars by Lewis et al.,31 type I chemovars are Δ9-THC predominant, type II contain equal amounts of Δ9-THC and CBD, and type III are CBD predominant chemovars. Terpenoids in these heat maps were color coded in relation to the maximum value for each compound. Compounds were arranged by order of elution from the GC column, according to: monoterpenes monoterpenoids sesquiterpenes sesquiterpenoids. Terpenoids with absolute contents below 20 ppm in all the type I-III samples were not included in the comparison.
Figure 3. Terpenoid analysis of selected types (A) I, (B) II, and (C) III Cannabis inflorescences prepared as follows: (a) inflorescences without further pretreatment (control); (b) inflorescence concentrate extracted with ethanol without evaporation (EtOH); (c) inflorescences following decarboxylation (ADC); and (d) inflorescence concentrate extracted with ethanol following evaporation (EtOH & Evap). Terpenoids are color coded in relation to the maximum value for each compound. Compounds are arranged by order of elution from the GC column. The specified values are the average absolute concentrations in ppm of 4 different preparations (3 for the ethanol extracts following evaporation).
In Figures 3A-3C, type I-III Cannabis inflorescences, respectively, were prepared as follows: (a) inflorescence without further pretreatment (control); (b) inflorescence concentrate extracted with ethanol without evaporation (EtOH); (c) inflorescences following decarboxylation (ADC); and (d) inflorescence concentrate extracted with ethanol following evaporation (EtOH & evap). The specified values are the average absolute concentrations of 4 different preparations (3 for the ethanol extracts following evaporation) in ppm, relative to the untreated inflorescences. Concentrations of the ethanol extracts following evaporation were normalized to the weight of the untreated inflorescences, by multiplying each of the terpenoid concentrations by the extraction yield using Eq. 2 (Method S1). As shown, terpenoid concentrations of the non-treated inflorescences (control) varied between type I-III samples, which can be related to the natural biochemical diversity between Cannabis chemovars. Nonetheless, all three samples showed similar trends in relation to sample preparation. For all three samples, the non-treated inflorescences (control) contained the highest concentrations of all terpenoids. Ethanolic extracts without evaporation (insertion of 10 µl of the ethanolic sample into the HS vial) showed mostly non-significant differences in terpenoid concentrations compared to the control (Figure S3), especially for the sesquiterpenes. On the other hand, decarboxylation and evaporation, exhibited pronounced differences in terpenoid contents compared to the control, especially for monoterpenes which almost entirely disappeared from the analyzed samples (Figures 3A-3C). It should be noted that the terpenoid contents after decarboxylation may also depend on the general content of the non-terpenoid compounds in the inflorescences relative to the untreated ones. Interestingly, evaporation and decarboxylation similarly decreased the concentrations of most monoterpenes and monoterpenoids. In Figures 4A-4C, we compare the terpenoid profiles of type I-III Cannabis extracts prepared by (a) ethanol (EtOH); (b) supercritical CO2 (sCO2); and (c) iso-butane (Butane) extractions. Concentrations are presented in ppm per weight of extract. These three extraction methods are the most highly applied methods today for preparation of medical Cannabis concentrates for patients. According to Figures 4A-4C, the method of extraction may lead to distinct differences in the terpenoid profiles of the extracts. As shown, iso-butane extraction is best suited for monoterpene and monoterpenoid extractions, probably due to the very short extraction process which does not include any solvent evaporation. sCO2, on the other hand, has the highest affinity for sesquiterpenes and sesquiterpenoids. Ethanol extraction had the lowest concentrations of mono- and sesquiterpenes compared to the other two extraction methods, probably due to the long evaporation required to remove the solvent, and similar concentrations as sCO2 for the mono- and sesqui-terpenoids. The later may be due to the higher polarity of the hydroxyl group in ethanol which improves dissolving of mono- and sesqui-terpenoids. The variation in terpenoid contents as a result of using different sample preparation steps and/or solvents for the same chemovar, further demonstrate the importance of comprehensive terpenoid analysis. The type of solvent used in the sample preparation, evaporation, decarboxylation, and method of extraction can lead to very different terpenoid
ACS Paragon Plus Environment
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
concentrations compared to the non-treated inflorescences (Figures 3A-3C and 4A-4C). Monoterpenes and monoterpenoids, in particular, are susceptible to any kind of heating and/or evaporation of solvents. These results are extremely important and highlight the method of SHS as most suitable for metabolic analysis of highly volatile compounds, since no sample preparation is required prior to the analysis.
Figure 4. Terpenoid analysis of selected type I-III Cannabis extracts (A)-(III), respectively, prepared by (a) ethanol (EtOH); (b) sCO2; and (c) iso-butane (Butane) extractions. Terpenoids are color coded in relation to the maximum value for each compound. Compounds are arranged by order of elution from the GC column. Concentrations are presented in ppm per weight of extract.
As shown, medical Cannabis extracts can have very different terpenoid contents than those naturally identified in inflorescences, as a result of the extraction method and/or solvent used (Figures 4A-4C). Currently, patients may choose an extract preparation over air-dried inflorescences, but the method and solvent of extraction are determined by the medical Cannabis distributors. To the best of our knowledge, there is no published scientific information currently available on the relationship between extraction solvent and/or method, and terpenoid contents. Further experiments might reveal ways of optimizing the content of a medical Cannabis preparation by choosing the appropriate extraction method and solvent for a desired content and/or ratio of terpenoids and other biomolecules, as well as directing the selection of suitable extract preparations for different medical conditions. Comprehensive analysis of terpenoid variations in medical Cannabis chemovars Several studies have suggested that while the concentrations and ratios of phytocannabinoids in Cannabis are relatively
reproducible for different chemovars,3-5,9,22,31 and some correlations have been suggested for specific phytocannabinoid types,9-12,32 terpenoid profiles are often unknown and/or unpredictable.9,22,33 In addition, the composition of terpenoids in Cannabis varies substantially as a result of genetic, environmental, and developmental factors.33 All these factors may lead to very different pharmacological effects for different, but also similar chemovars, in terms of therapeutic efficacy and side effects. Therefore, in order to explore the variability and stability of terpenoid profiles for different Cannabis chemovars, we collected preliminary terpenoid data using the developed SHS-GC-MS/MS method. To this end, we analyzed sixteen different medical Cannabis chemovars obtained from several licensed local distributors, at two/three different dates. The classification of the different samples into Cannabis chemovars was according to the reported information given by the licensed distributors. Inflorescences in this analysis were analyzed within several days after their arrival. In Figure 5 we compare the SHS-GC-MS/MS concentrations of the main terpenoids (terpenoids with absolute contents below 30 ppm appear in Figure S4). As suggested in previous studies,14,15 we also observed that the absolute terpenoid concentrations were less stable over time for the same cultivar, compared to their profiles. Therefore, for chemotyping purpose, each terpenoid was normalized first by the sum of all the terpenoids in the sample, and then by the highest terpenoid value in all the samples. The normalized values appear in Figure S5. Terpenoids in this heat map were arranged by decreasing absolute content in all of the samples as detailed by the maximum concentration column (Max conc), and chemovars according to the group assignments observed by Richins et al.15 The groups were according to predominant terpenoids as follows: (I) β-myrcene, (II) α- and β-pinene, (III) βcaryophyllene and limonene, (IV) β-caryophyllene, and (V) terpinolene. According to this figure, terpenoid concentrations vary greatly depending on the chemovar and/or type of compound. As shown, most chemovars displayed repeatable terpenoid profiles for the major components in each sample over the time periods analyzed (less than 20% difference between the dates for the same terpenoid). Some samples, however, showed very different terpenoid profiles. CAN1-a, for example, had 50% less β-myrcene, and considerably higher contents (>50% difference) of trans-β-farnesene, guaiol, α-bisabolol, transnerolidol and others, compared to CAN1-b and CAN1-c. CAN6-a and CAN6-b showed also very distinct differences for α-pinene, β-pinene, α-terpineol, sabinene, guaiol, and others. These marked differences in terpenoid profiles for the same chemovars at different dates, may be the result of mislabeling by the Cannabis distributor, age of inflorescences, environmental and developmental differences that occurred during the growing of the plants, or other non-predictable effects. Regardless of the reason for this variability, this data emphasize the need for comprehensive analysis of Cannabis samples intended for medical treatment, since all of the mentioned terpenoids and many others (marked in blue in Figure 5) have proven to produce biological/therapeutic effects, even at low concentrations, as summarized in a recent review.6 Overall, the terpenoids in Figure 5 and Figure S4 are either naturally biosynthesized in Cannabis plants, or are degradation products of the natural ones. Due to their unsaturated nature, terpenoids are highly prone to photo-oxidation.17,18 Much like
ACS Paragon Plus Environment
Page 6 of 9
Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry in triglyceride oils and other polyunsaturated molecules, photooxidation leads to the formation of allylic hydroperoxides.17,18 These are highly reactive species, that may undergo further decomposition into oxygenated products including alcohols, ketones, and aldehydes.34 Several oxidation products from terpenoids have been previously identified.17,18 As could be expected, almost all of the aldehydes and ketones in Figure S4 (marked in green in Figure S4; see Figure 1 for the group assignments), had considerably lower concentrations (
