Profiling of Terpene Metabolism in 13CO2-Labelled Thymus

ics: O om pe ol n b A a c t c e e

M s s Kutzner et al., Metabolomics 2014, 4:1 Metabolomics: Open Access DOI: 10.4172/2153-0769.1000129 ISSN: 2153-0769

Research Article Open Access

13 Profiling of Terpene Metabolism in CO2-Labelled Thymus transcaucasicus Kutzner E1#, Manukyan A2# and Eisenreich W1* 1Lehrstuhl für Biochemie, Technische Universität München, Lichtenbergstraße 4, 85748 Garching, Germany 2Center of Life and Food Sciences Weihenstephan, Technische Universität München, Dürnast 4, 85350, Freising-Weihenstephan, Germany #Both authors contributed equally to the paper

Abstract Metabolism is characterized by the functional pathways and fluxes connecting substrates, intermediates and products of a given organism. Pathways in carbon metabolism can be identified by incorporating 13C-labelled tracers. 13 In the present model study, we have evaluated potential benefits from a single CO2 pulse-chase experiment with the Caucasian endemic plant Thymus transcaucasicus for the analysis of terpene composition and biosynthesis. The study design was conducive of low 13C-enrichments (< 1%) in terpenes that were detected at enhanced NMR-sensitivities without hampering GC-MS-based methods for terpene profiling. From the specific 13C-labelling patterns, pathways of terpene biosynthesis could be gleaned as exemplified for the monoterpene thymol which is made via the non-mevalonate 13 route under the physiological in vivo conditions of the CO2 experiment.

Keywords: Transcaucasian thyme; Thymus transcaucasicus; DMAPP can be synthesized by the well-known mevalonate pathway 13 in the cytosolic compartment of the cell or by the more recently Lamiaceae; Biosynthesis; Pathway; CO2; Terpene discovered MEP pathway that is operative in the plastids (Figure 1A). Introduction Depending on the specific compartment where the final terpene is Non-targeted or semi-targeted metabolite profiling has become a made, the respective pathway leads to the product. Typically, mono- powerful method to determine the composition of small molecules in and diterpenes originate from the plastids and are therefore derived plant extracts using apolar or polar solvents. In principle, metabolite from the MEP pathway (for reviews, see [14-17]). However, mixed profiles are specific snapshots (fingerprints) of the metabolic processes biosynthetic patterns were also reported for certain plant terpenes [17- that have occurred in the plant under study. Using the tools of GC- 20]. 1 MS, LC-MS or H-NMR spectroscopy, crude mixtures can be analyzed Until now, there is no information available on monoterpene to provide information about metabolic profiles. Depending on the biosynthesis in endemic Thymus species, like T. transcaucasicus, specific method, dozens to hundreds (by1 H-NMR) or hundreds to while it is established knowledge that thymol from common thyme thousands (by GC-MS or LC-MS) of metabolites can be detected and (T. vulgaris) is biosynthesized by aromatization of γ-terpinene to assigned in a single run [1,2]. Whereas some of these methods (e.g. p-cymene followed by hydroxylation of p-cymene [21] (Figure 1B). GC-MS) typically need chemical transformation of the compounds More recently, it was also reported for cut shoots of T. vulgaris that under study, others (e.g. 1H-NMR) are qualified to deal with the crude mixtures without any prior chemical treatment. On the other hand, the two isoprene units of thymol are made via the MEP pathway [22]. the 1H-NMR approach is hampered by low sensitivity in comparison However, the experimental design of the former study [22] does not to MS-based methods, as well as by limited resolution due to the rule out effects on the metabolic network due to wounding or the usage 13 narrow chemical shift range of 1H-frequencies (i.e. approximately 10 of artificial carbon substrates which were used as C-labelled tracers in ppm). Due to the even lower sensitivity of 13C-NMR as compared to the earlier experiment. In sharp contrast, these intrinsic drawbacks can 1H-NMR, 13C-profiling is not established in metabolomics, although the be avoided by the experimental design of the current study using intact 13 C-NMR chemical shift range is much larger (i.e. about 200 ppm) than plants and the natural carbon substrate CO2 as tracer. in 1H-NMR and therefore less compromised by signal overlap. Only recently, 13C-NMR has been introduced as a tool for metabolomics, Methods 13 particularly for the analysis of CO2-grown photosynthetic bacteria Plant material and study design 13 [3]. Using special growth chambers, CO2-labelling experiments with intact plants revealed considerable details about metabolic pathways A one year old full blooming plant of T. transcaucasicus from under physiological conditions [4-10]. However, all of these studies have not addressed the possibility to combine benefits of13 C-labelling for pathway analysis and 13C-based metabolite profiling from a single experiment. Therefore, we have now performed a model study with *Corresponding author: Eisenreich W, Lehrstuhl für Biochemie, Technische 13CO -labelled Transcaucasian thyme (Thymus transcaucasicus Ronn.) Universität München, Lichtenbergstraße 4, 85748 Garching, Germany, Tel: +49 89 2 28913336; Fax: +49 89 28913363; E-mail: [email protected] (Lamiaceae) to assess both metabolite profiles (e.g. isoprenoids) as well as metabolic pathways from the same experiment. This Caucasian Received November 18, 2013; Accepted January 24, 2014; Published January endemic species was selected since it is known as a rich producer of 31, 2014 essential oils including terpenes [11-13]. Citation: Kutzner E, Manukyan A, Eisenreich W (2014) Profiling of Terpene 13 Metabolism in CO2-Labelled Thymus transcaucasicus. Metabolomics 4: 129. Indeed, terpenes (e.g. monoterpenes) constitute the major fraction doi:10.4172/2153-0769.1000129 of secondary metabolites in all thyme species. Generally, terpenes Copyright: © 2014 Kutzner E, et al. This is an open-access article distributed under are derived from the basic building units, isopentenyl diphosphate the terms of the Creative Commons Attribution License, which permits unrestricted (IPP) and dimethylallyl diphosphate (DMAPP). In all plants, IPP and use, distribution, and reproduction in any medium, provided the original author and source are credited.

Metabolomics ISSN: 2153-0769 JOM an open access journal Volume 4 • Issue 1 • 1000129 13 Citation: Kutzner E, Manukyan A, Eisenreich W (2014) Profiling of Terpene Metabolism in CO2-Labelled Thymus transcaucasicus. Metabolomics 4: 129. doi:10.4172/2153-0769.1000129

Page 2 of 11

Terpenes were extracted from the leaves as well as from the flowers

A Mevalonate pathway Methylerythritol phosphate (MEP) pathway of T. transcaucasicus with cold chloroform-D (CDCl3) (Sigma-Aldrich, O O O Steinheim, Germany). In more detail, 400 mg of the plant material (fresh weight) were placed into a 10 ml test tube. 2 ml of CDCl were 1 SCoA 5 COO- H O P 3 OH 6 added and mixed in an attempt to cover all leaves with the solvent. The mixture was left at room temperature for several minutes. The solvent O O was then transferred into the next test tube containing another set of 2 SCoA OH 400 mg of plant material from the same plant. These leaves were treated O P by the same way as the ones in the first test tube. The procedure was O OH 7 repeated again with another portion of 400 mg of fresh material. The O HO O solvent was collected and 100 mg of dry MgSO4 were added to remove 3 HO SCoA HO water. The mixture was held for 30-90 minutes under periodically HO shaking. Flowers were in general subjected to the same procedure as O O P HO OH 8 the leaves. Without further preparation, 200 µl of the extract were used 4 HO OH for GC-MS measurements (see below). For NMR analysis, 600 µl of the extract were filled into a 5 mm NMR tube and measured. Metabolite and isotopologue profiling: Gas chromatography O P P 9 OH – mass spectrometry The gas chromatography – mass spectrometry (GC-MS) O P P 10 measurements were performed on a GC 2010 Gas Chromatograph and a GCMS-QP 2010 Plus mass spectrometer coupled to a QP-5000 mass selective detector (Shimadzu, Duisburg, Germany) working with 11 O P P electron impact (EI) ionisation at 70 eV. A Silica capillary column Equity TM-5 (30 m x 0.25 mm x 0.25 µm film thickness) from Supelco Inc. (Bellefonte, PA, USA) was used. An aliquot of the chloroform 12 O P P extract was injected in 1:10 split mode at 230°C and a helium inlet pressure of 82.8 kPa. The interface temperature was 260°C and the B helium column flow was 1.17 ml/min. The column was developed at

O P P 90°C for 2 minutes and then with a temperature gradient of 5°C/min

OH to a temperature of 150°C followed by a gradient of 50°C/min to a final temperature of 250°C that was held for 2 minutes. Each sample was

12 13 14 15 analyzed three times in order to acquire selected ion monitoring (SIM) data. The identification of the essential oil components was carried Figure 1: Monoterpene biosynthesis. A. The terpenoid precursors out by comparing the retention times and mass data of pure reference isopentenyl pyrophosphate (IPP, 10) and dimethylallyl pyrophosphate (DMAPP, 11) are synthesized in plants via two different pathways, the compounds, and by comparing data from the NIST05 and NIST05s mevalonate and the methylerythritol phosphate (MEP) pathway. DMAPP mass spectral reference library. The relative intensities of the standards (11) and IPP (10) are further condensed to geranyl pyrophosphate (GPP, and the samples obtained from GC-MS analysis (peak integration) were 12) which can be converted into monoterpenes such as thymol (15). The isoprene unit derived from IPP (10) is highlighted in red, the isoprene unit processed with an in-house Excel-based software package according to derived from DMAPP (11) in green. B. Biosynthesis of thymol (15) by [23-26]. This evaluation resulted in the molar13 C-excess of the carbon aromatization of γ-terpinene (13) to p-cymene (14) followed by hydroxylation of p-cymene (14). isotopologues in thymol. Metabolite and isotopologue profiling: NMR spectroscopy controlled greenhouse soilless culture was placed into a gas incubation For the measurement of 13C NMR and INADEQUATE spectra an chamber (Advance Optima Biobox AO2000, GWS, Berlin, Germany). Avance III 500 system (Bruker, Rheinstetten, Germany) with a cryo The plant was illuminated with white light. The temperature was 1 13 31 13 probe head (5 mm CPQNP, H/ C/ P/ 19F/29Si; Z-gradient) was adjusted to 25°C. Prior to the labelling period with CO (pulse phase), 1 2 used. H and ADEQUATE spectra were measured with an Avance I the chamber was flushed with synthetic air (Westfalen AG, Münster, 500 system (Bruker) and an inverse 1H-13C probe head. The resonance Germany) containing only oxygen (20.5 vol. %) and nitrogen (79.5 frequencies of 1H and 13C were 500.13 MHz and 125.82 MHz, vol. %) until the atmospheric unlabelled CO was almost completely 2 respectively. The temperature was 300 K. The data analysis was done removed from the chamber. The plant was then labelled using a 13 13 with the MestReNova Software Version 7.0.0 (Mestrelab Research, mixture of synthetic air and CO . The concentration of CO was held 13 2 2 Santiago de Compostela, Spain). The one-dimensional C NMR spectra constant at 700 ppm by permanently adding the gas from the reservoir as well as the one- and two-dimensional spectra 1H NMR spectra were (Sigma-Aldrich, Steinheim, Germany) for 4 hours. During this labelling 13 measured with standard Bruker parameter sets. period, the plant consumed about 120 ml of CO2. Subsequently, the plant was allowed to grow at room temperature under standard Results and Discussion greenhouse conditions and under a natural atmosphere (i.e. containing 13 12CO ) for 12 days. A GC chromatogram of a leave extract from CO2-labelled T. 2 transcaucasicus is shown in Figure 2A. About 20 - 30 peaks (> 2% Metabolite profiling: Extraction of terpenes in relation to the highest peak) were detected in typical runs. The

Page 3 of 11

other hand, signals in the down-field region were well separated

(x100,000) l A o 9.0 TIC m and immediately allowed the assignments of some intense signals to y h 1 T thymol, γ-terpinene, and α-pinene (Figure 4) by comparison with H 8.0 NMR reference data (own library of 1H NMR spectra of terpenes (see 7.0 also Table 1) or data provided by the Spectral Database for Organic

l Compounds, SDBS). Moreover, the assignments of the signals were

6.0 o e n e i n

e confirmed by two-dimensional NMR experiments with the crude C e e n - n n e 8 i e , P n m 5.0 1 - i

y mixture (e.g. COSY, HMQC, HMBC) (Table 1 and Figure 5). Notably, α p l r C o e r 13 T c - a γ the weak C-enrichments (< 1%) did not disturb the assignments of the

4.0 v r

a 1 13 1 C

Relative detector response H-NMR signals since the C-induced H-satellite signals were small 3.0 1

e (< 1% in the overall signal intensity of a given H-atom). However, the n

l e e e l e l o

n n n y e e e e e e l e l h n n n n 2.0 n e o r h o p e e e o n l b o p e o h c e r e o a r n m B y p r i i o n n r s y i i u p L e a m j r m P u B M a - u e - - C A M - Valencene h

1.0 α β ot identified α T C - - α n T γ not identified - α α A 10 B 10 B D 5.0 7.5 10.0 12.5 15.0 17.5 Retention time [min] B 5 2 % 135 F 4 6 E 1 3 100.0 E J 9 7 A B/I C H + OH 1 6 75.0 7 7 H G 3 4 H G/H + OH 2 OH C6H5 C 8 5 D F 50.0 C 7 - CO 150 25.0 91 8 9 115 A A 107 136 77 105 65 121 63 74 82 94 128 0.0 147 154 60 70 80 90 100 110 120 130 140 150 m/z C 7 D 10 A C B % 135 1 5 100.0 6 2 4 6 E/D E/D D E

75.0 5 C 3 3 1 F/D 4 F/D E D 2 50.0 O 8 7 150 C 25.0 91 10 9 8 9 115 B B A A 77 107 65 79 63 103 109 121 132 0.0 60 70 80 90 100 110 120 130 140 150 m/z Figure 3: Chemical structures and numbering of the major monoterpenes detected in chloroform extracts of T. transcaucasicus. A, thymol. B, α-pinene. C, 1,8-cineole. D, γ-terpinene. Figure 2: GC-MS analysis of chloroform extract of T. transcaucasicus 13 flowers harvested 9 days after the labelling with CO2 for 4 hours. A. Gas chromatogram performed as described under Experimental. B.

Corresponding mass spectrum and fragmentation pattern of thymol (molar Thymol C-8/9 (d)

mass 150.22 g/mol) from the labelled sample. C. Corresponding mass spectrum of unlabelled thymol from the NIST05s library. 1,8-Cineole C-9/10 (s)

-Pinene C-8 (s) α identification of the essential oil components was performed by comparison with reference data from the NIST05 and NIST05s libraries. 1.32 1.30 1.28 1.26 1.24 13 Thymol It turned out that the weak C-enrichments (< 1%) did not disturb C-10 (s) the identification of most compounds in the mixture. As a result, the γ-Terpinene apolar fraction obtained from the leaves as well as from the flowers of C-3/6 (m)

T. transcaucasicus was predominantly (more than 90% in the overall α-Pinene intensity of all detected peaks) composed of monoterpenes including C-3 (m) thymol, γ-terpinene, α-pinene and 1,8-cineol (Figure 3). Sesqui- terpenes were only detected at minor amounts, e.g. caryophyllene, 5.50 5.40 5.30 5.20 Thymol C-3 (d) germacrene D, α-bisabolene and β-ocimene. In addition to these Thymol Thymol Thymol compounds, the flowers produced additional terpenes, e.g. borneol. C-4 (d) C-6 (s) C-7 (m) These results are in good accordance to previous investigations with T. transcaucasicus and other thyme species [12,13,27]. 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm Figure 4: 1H NMR spectrum of crude leaves extract of T. transcaucasicus. The same chloroform extracts were now also analysed by NMR Leaves were extracted (see section Experimental) after a chase period of spectroscopy. A typical 1H NMR spectrum is shown in Figure 4. 12 days. The number of scans in the 1H experiment accounted for 64 and the solvent was CDCl . The FID was multiplied with a Gaussian function (lb Not unexpectedly, the high-field NMR region was crowded due to 3 = -1.00; gb = 0.8 Hz). the presence of multiple compounds in the crude extract. On the

Page 4 of 11

Chem. Shift Chem. Shift Coupling Compound 13C atom δ (ppm) 1H atom δ (ppm) constant (Hz) COSY NOESYa HMBCa Thymol C-8/9 22.7 A 1.28 (6 H, d) 6.9 C B, C, G, D(w) C-10 20.9 B 2.31 (3 H, s) - A, D, E, F G(w), C, B(w), A C-7 26.7 C 3.20 (1 H, m) 6.9 A A, D, G G(w), F, E -OH - D 4.65 (1 H, s) - E, C, B, A(w) G, E(w), A C-6 116.0 E 6.61 (1 H, s) - B, D, F, G - C-4 121.7 F 6.77 (1 H, d) 7.9 G B, E, G G(w), F, D, B C-3 126.2 G 7.12 (1 H, d) 7.7 F A, C, E, F G(w), E, B C-1 152.5 F(w), D(w), C, B(w) C-5 136.7 G, F(w), E(w), D, C, B(w) C-2 131.3 G, F(w), E(w), B G(w), F, E, D, C, B, A (+)-α-Pinene C-9 20.8 C-7 31.5 A 0.86 (3 H, s) - C, D, E(w), G/H, I, J C-8 26.4 B 1.18 (1 H, d) 8.5 I C, E(w), F(w), G/H, I, J(w) B, C C-10 23.0 C 1.29 (3 H, s) - A, B, D, E, F, G, H, I, J due to [28] C-1 47.0 D 1.68 (3 H, dd) 2.1 - A, C, E, F, G, H, I, J A, B, E 11.2 C-5 40.7 E 1.96 (1 H, t) 1.5 F, I A(w), B(w), C, D, F(w) E C-4 31.3 F 2.10 (1 H, m) E, I B(w), C, D, E(w), G, I(w), J(w) A, B, C, D, F, I, H(w) C-4 31.3 G 2.18 (1 H, m) 17.3 (dd) H A, B, C, D, F, H, I, J A, B, C, E, I, H(w) C-7 31.5 H 2.26 (1 H, m) 17.3 (dd) G A, B, C, D, G, I, J due to [28] 8.5 I 2.36 (1 H, m) B, E, F A, B, C, D, F, G, H, J C-3 116.0 11.2 due to [28] C-6 38.0 J 5.21 (1 H, m) - A, B, C, D, E, F, G, H, I due to [28] C-2 144.5 - A, B, C, E, I, F(w), G(w) B, D, E, I, C(w), G(w), H(w) 1,8-Cineole C-7 27.6 C-9/10 28.9 A 1.07 (3 H, s) - B, E(w) C-4 32.9 B 1.26 (6 H, s) - A, C(w), D(w), F D(w), E C-2/3/5/6 22.8 / 31.5 C 1.42 (1 H, s) - B(w), D, E(w), F B, D(w) C-2/6 31.5 D (endo)b 1.51 (4 H, m) E, F B(w), C, E, F B, D, E(w), F C-3/5 22.8 E (exo)b 1.68 (2 H, m) D, F D, C(w), F(w) C-1 69.8 F (exo)b 2.03 (2 H, m) D, E B, C, D, E(w) A, C, D, E, F C-8 73.7 A(w), C(w), D, E, F A, D(w), E, F B, D, E(w), F(w) γ-Terpinenec C-8/9 1.0 C-10 23.0 A 1.01 C-7 34.6 B 1.66 C-1/4 31.7 C 2.20 C-3/6 118.9 D 2.60 C-2 140.6 E 5.43 C-5 131.3 a’(w)’ indicates a weak correlation signal bdue to [29] call data due to Spectral Database for Organic Compounds SDBS (SDBS No. 23242HSP-06-383)

Table 1: NMR signal assignment of main compounds in essential oils of Thymus transcaucasicus. The solvent was CDCl3. 13C-enrichments significantly increased the sensitivity of HSQC and Nevertheless, it should be noted that the NMR sensitivity with the HMBC experiments, thus benefitting the quality of these spectra. More 13C-enriched sample was still lower than the sensitivity of the GC- than 100 well defined correlation peaks were observed in the full HMBC MS method. As a consequence, we could hardly detect by NMR some 13 spectrum of the extract from the CO2 labelled plant (Figure 6A), minor terpene compounds which were clearly observed by GC-MS. whereas ca. 50 peaks were detected in the spectrum from the unlabelled However, a major benefit of 13C-NMR analysis became obvious when plant (Figure 6B) (using the same amount of plant material and the analyzing the metabolic processes leading to terpenes. same NMR parameters). In accordance to the GC-MS data, thymol was As a model compound for pathway analysis, we have selected again identified by NMR as the most abundant metabolite in the crude thymol as the most prominent terpene in the chloroform extract of T. extracts of leaves and flowers. The minor terpenes 1,8-cineole, α-pinene transcaucasicus. We have started with a concise analysis of the isotope and γ-terpinene could also be clearly assigned in the 13C-enriched pattern detected in the MS spectrum of the 13CO -labelled compound. sample (Table 1 and Figure 4,5), but hardly in the unlabelled isolate. 2

Page 5 of 11

A γ-Terpinene C-3/6 (m) ? α-Pinene C-3 (m) ?

111 112 Thymol C-6 113 5.21 ppm / α-Pinene C-3 116.01 ppm 114 γ-Terpinene C-3 5.43 ppm / 115 116.06 ppm ) 116 m p p

117 ( 5.43 ppm / 118 118.94 ppm γ-Terpinene 119 C-6 120 121 (Thymol C-4)

122 C chemical shift

? 123 13 124 125 (Thymol C-3) ? 126 127 128

5.70 5.65 5.60 5.55 5.50 5.45 5.40 5.35 5.30 5.25 5.20 5.15 5.10 5.05 5.00 4.95 4.90 4.85 1H chemical shift (ppm) B 1,8-Cineole C-9/10 (s) α-Pinene C-8 (s)

67 68 15 69 α-Pinene

70 C-9 1.29 ppm/20.82 ppm 20 ) 71 ) m m p p

72 p 25 p ( ( 73 1,8-Cineole 30 C-8 74 75 35 1.26 ppm / 76 C-6 1.29 ppm/37.98 ppm C chemical shift 73.65 ppm C chemical shift 13 77 C-5 1.29 ppm/40.71 ppm 40 13 78 45 79 C-1 1.29 ppm/47.00 ppm 80 50 81

1.35 1.30 1.25 1.20 1.33 1.31 1.29 1.27 1.25 1.23 1H chemical shift (ppm) 1H chemical shift (ppm)

Figure 5: Expanded view of HSQC (A) and HMBC (B) spectra of crude leaves extract of T. transcaucasicus. Leaves were extracted (see section Experimental)

after a chase period of 12 days. The number of scans accounted for 8 and 16, respectively. The solvent was CDCl3.

Page 6 of 11

0 10 20 30 40 50 60 70 80 C (ppm)

90 13 100 110 120 130 140 150 160 170

B 0 10 20 30 40 50 60 70 80 C (ppm)

90 13 100 110 120 130 140 150 160 170

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 1H (ppm)

13 Figure 6: Two-dimensional HMBC spectra of crude leaves extracts of T. transcaucasicus. A. Leaves (1.7 g) of the CO2-labelled plant (4 hour pulse period) harvested after a chase period of 12 days. B. Leaves (1.7 g) of an unlabelled T. transcaucasicus plant (same developmental stage as the labelled plant). Both spectra were measured with the same standard parameter sets and 16 scans per increment.

The mass spectrum is shown in Figure 2B with a peak at m/z = 150 (M+3) can be explained by the MEP pathway which contributes reflecting the molar mass of thymol. The peaks at m/z = 151, 152, three 13C atoms to the terpene precursors, IPP and DMAPP, via 13C- 153, etc. are due to thymol molecules carrying one, two, three, etc. 3 labelled GAP (Figure 8). In contrast, the mevalonate pathway can 13C-atoms, respectively. By comparison of the mass intensities with only contribute two 13C atoms to a given isoprene precursor via the the corresponding ones of a thymol sample at natural 13C-abundance, two-carbon moiety in acetyl-CoA. Nevertheless, one has to take into the 13C-excess values of each isotopologue due to incorporation of consideration that M+2 isotopologues not only originate from the 13 CO2 were calculated. The normalized ratios are displayed in Figure MVA pathway (starting from acetyl-CoA) but are also transferred to 7. It turned out that thymol was a complex mixture of isotopologues IPP/DMAPP via the pyruvate building block in the MEP pathway (one with M+1, M+2, M+3, and M+4 as the most prominent specimens. carbon of pyruvate is lost as CO2 during the biosynthetic pathway). One Isotopologues with more than four 13C-atoms were less abundant (< also has to take into account that the GAP precursor in the MEP pathway 13 13 20% in all C-isotopologues). is probably not only C3 labelled, but also carries a substantial amount of 13C-2 labelled isotopologues that subsequently result in additional Tentatively, isotopologues in thymol comprising three 13C atoms

Page 7 of 11

signals due to the following transfer paths: 13C-8 → 13C-7 → H-7, 13C- 13 13 13 13 13 13 100% 7 → C-8 → H-8, C-2 → C-3 → H-3, C-1 → C-6 → H-6, C-5 → M + 10 13 M + 9 C-10 → H-10. This clearly indicates that the biosynthetic process 80% 13 M + 8 afforded four pairs of directly adjacent C-atoms, i.e. 8-7, 1-6, 5-10 13 M + 7 and 2-3 (shown in Figure 11B as bars connecting C-atoms). Carbon 60% M + 6 atoms C-4 and C-9 were not incorporated in form of a directly-bound 13 13 13 M + 5 C-pair, but only showed long-range C- C couplings (see Figure 10). 40% M + 4

20% M + 3 M + 2 13CO / 12CO

Relative Relative isotopologue distribution 2 2 0% M + 1 0 1 2 5 8 912912

Leaves Flowers OH Chase time [d] O - Figure 7: Relative isotopologue distribution (excess) of thymol analyzed P COO by GC-MS. Thymol was extracted with chloroform from T. transcaucasicus 3-Phosphoglycerate 13 leaves and flowers at different points of time after the labelling with CO2 for 4 hours (see Experimental). Excess values were obtained by subtracting the natural 13C abundance of 1.11%. The M+1 bar in the diagram represents 13 molecules carrying only one C atom (position unknown), the M+2 bar stands OH for molecules carrying two 13C atoms, and so on. O H P M+2 isotopologues of thymol via the MEP pathway. Therefore, it is not O possible to assess the absolute rates of thymol formation via the MEP Glyceraldehyde 3-phosphate and/or the mevalonate pathway on the basis of these MS data only. Another weakness of the GC-MS thymol data is the lack of positional label information. Unfortunately, the smaller fragments detected in the O O

MS spectrum of the monoterpene are not useful to further elaborate COO- SCoA 13 the positional C-distribution in the molecule. Nevertheless, the Pyruvate Acetyl-CoA significant amounts of (M+3) isotopologues in the labelled thymol at least suggest its predominant formation via the MEP pathway, as shown earlier for thymol from cut shoots of T. vulgaris [22]. P P PP NH + In order to further support this hypothesis, the monoterpene O 3 O 1 13 was subjected to a detailed NMR analysis. H- and C-NMR signals COO- of thymol (detected in the chloroform extract, Fig. 9B) could be Alanine unequivocally assigned on the basis of literature data and on the basis of two-dimensional experiments (COSY, HSQC, HMBC, ADEQUATE) Geranyl pyrophosphate Geranyl pyrophosphate (Tables 1 and 2) (Figure 9B). For comparison, a 13C-NMR spectrum of an authentic thymol reference is shown in Figure 9A. A closer inspection of the signals of the labelled thymol sample revealed satellite signals due to 13C-13C couplings (Figure 10). To improve the spectral quality, the central and the satellite signals were fitted using the GSD algorithm implemented in the MestReNova software. Mathematically, OH OH OH this module applies a de-convolution of complex spectra transforming them into individual lines. The shape of the lines can further be approximated by using combinations of Lorentzian and Gaussian Thymol functions. Whereas for most of the signals sharp satellites were observed Figure 8: Incorporation of 13C atoms derived from labelling experiments 13 13 The due to coupling with one neighboured C-atom, the satellite signals of with CO2 into various metabolites during the plant metabolism. possible positions of 13C atoms are highlighted as colored bars and small C-1 and C-3 were more complex with an additional fine splitting (8.3 boxes, respectively; the color depends on the metabolic pathway which 13 Hz and 2.5 Hz, respectively) indicating the presence of C-3-motifs leads to a certain metabolite. 3-Phosphoglycerate (PGA) and glyceraldehyde in the molecule. Interestingly, corresponding coupling constants were 3-phosphate (GAP) are built up in the Calvin cycle which generate either molecules with two 13C atoms (pink bar) or molecules carrying three 13C atoms observed for satellites close to the central signals for C-4 and C-8/9, (red bars). A label at position 1 exclusively is also possible. The synthesis of respectively (see also Table 2 with all detected coupling constants). By monoterpenes is possible via the MEP pathway (left side) based on pyruvate 13 comparing these coupling constants, pairs or triples of C-atoms in and GAP (see Figure 1) which leads to a labelled C2 block originating from pyruvate (dark red bar and dark green bar, respectively) and/or a C or C thymol can be assigned as indicated in Table 2. More direct experimental 2 3 block originating from GAP carrying either two 13C atoms (pink and light evidence could be provided by ADEQUATE experiments that are green bar, respectively) or three 13C atoms (red/green bar and red/green box, based on magnetization transfer via one-bond 13C-13C couplings and a respectively), respectively. Red color marks carbon atoms coming from IPP subsequent transfer to the attached proton of the 13C-pair. Due to the and green color marks carbon atoms coming from DMAPP. On the right side 13 the predicted labelling pattern of thymol arising from the mevalonate pathway specific labelling pattern in the CO2 labelled thymol samples, only a based on acetyl-CoA is shown. This path does not generate monoterpenes few (but highly significant) signals were detected in these experiments. carrying three 13C atoms. Boxed thymol is displayed with the labelling pattern As shown in Figure 11B, the 1,1-ADEQUATE experiment displayed observed by NMR spectroscopy.

Page 8 of 11

Chemical shift Coupling constanta 13C enrichment 13 13 b 13 13 c Position C δ (ppm) JCC (Hz) % Crel. % C C

1 152.50 66.9 (6) 1.10 28.57 8.3 (4) 2 131.34 59.9 (3) 1.24 27.01 3 126.23 59.8 (2) 1.12 22.48 2.5 (9) 4 121.67 8.4 (1) 1.24 16.67 5 136.66 44.4 (10) 1.17 19.35 6 116.00 66.6 (1) 1.34 21.26 7 26.70 34.6 (8) 1.24 23.66 8/9 22.71 34.5 (7) 1.47 13.79 2.6 (3) 10 20.91 44.2 (5) 1.16 22.48 aThe numbers in parentheses indicate the coupling partners bThe signal with the lowest 13C enrichment is referenced to 1.10% 13C cFraction of the indexed coupling pair in the overall signal intensity of the respective 13C NMR signal

13 13 Table 2: C-NMR data of thymol obtained from T. transcaucasicus flowers labelled with CO2 for four hours and extracted with chloroform 12 days after the labelling pulse. The solvent was CDCl3.

A 10 8/9 5 4 6

1 3 2 OH 3 4 6 7 7 10 1 5 2 8 9

B 8/9

3 4 6 7 10 1 5 2

ppm 150 145 140 135 130 125 120 115 110 55 50 45 40 35 30 25 20 15 10

13 13 Figure 9: One-dimensional C NMR spectra of thymol. A. Standard sample (with natural C abundance of 1.1%) of thymol solved in CDCl3. Signals were assigned 13 to the corresponding carbon atoms of thymol. B. Sample of T. transcaucasicus flowers harvested and extracted with chloroform 12 days after the labelling with CO2

(4 hours). The solvent was CDCl3.

Page 9 of 11

A C-1 C-6 C-4

C-6 66.9 Hz C-1 66.6 Hz C-1 8.4 Hz

C4 8.3 Hz

fitted

original

ppm 152.8 152.7 152.6 152.5 152.4 152.3 152.2 116.3 116.2 116.1 116.0 115.9 115.8 115.7 121.8 121.7 121.6 121.5

C-8/9 B C-2 C-3 C-7 34.5 Hz C-3 59.9 Hz C-2 59.8 Hz C-3 2.6 Hz

C-9 2.5 Hz

fitted

original

ppm 131.6 131.5 131.4 131.3 131.2 131.1 131.0 126.4 126.3 126.2 126.1 126.0 125.9 22.9 22.8 22.7 22.6 22.5

C C-7 C-10 C-5

C-8 34.6 Hz C-5 44.2 Hz C-10 44.4 Hz

fitted

original

ppm 26.9 26.8 26.7 26.6 26.5 21.1 21.0 20.9 20.8 20.7 20.6 136.8 136.7 136.6 136.5 136.4

Figure 10: Thymol signals from the one-dimensional 13C NMR spectrum of T. transcaucasicus flowers. Flowers were harvested and extracted with chloroform 13 12 days after the labelling with CO2 (4 hours). The solvent was CDCl3. The spectrum was processed with MestReNova Software (zero filling 256k). The original spectrum was multiplied by a Gaussian window function of 0.7 Hz except for signals C-8/9 and C-5. For signal C-8/9 and signal C-5 a function of 0.5 Hz and of 1.2 was applied. The numbering of the carbon atoms is displayed in Figure 3.

Page 10 of 11

10 8/9

A 3

-OH CDCl 3 4 6 7

10 20 10 8/9/7 8/9 7 7/8/9 CH3 30

H H 40 5 50 4 6 3 1 60 2 H OH CDCl 70 3 H 80 CH 7 3 CH3 90 8 9 (ppm) C 100 13 110 6 4 4/3 120 3 3/4 2 2/3 2/7 130 5 5/4 5/6 5/10 140 150 1 1/6 160

1.01.52.02.53.03.54.04.55.05.56.06.57.07.5 1 H (ppm) 8/9 B 3 10 CDCl 3 4 6 7 10

108/9 8/9/7 20 7 10 7/8/9 30 CH3 40 H H 5 50 4 6 60 3 1 2 70 CDCl3 H OH H 80

7 90 (ppm) C H3C CH3 13 8 9 100 110 6 4 120 3 2 2/3 130 5 5/10 140 150 1 1/6 160

1.01.52.02.53.03.54.04.55.05.56.06.57.07.5 1H (ppm) 8/9 10 C 3 CDCl 3 4 6 7 10 CH3 10 8/9 10 20 7 H H 30 5 4 6 40 3 1 50 2 H OH 60

CDCl3 H 70 80 H C 7 3 CH3 90 8 9 100 C (ppm) C

110 13 6 4 120 3 2 130 5 3/9 140 1 150 160 1/4 170 180 0.00.51.01.52.02.53.03.54.04.55.05.56.06.57.07.58.0 1H (ppm)

Figure 11: 1,1- and 1,n-ADEQUATE spectra of crude leaves extract of Thymus transcaucasicus. A. 1,1-ADEQUATE (65 Hz) of thymol standard solution in CDCl3 (44 mM). B. 1,1-ADEQUATE (65 Hz) of crude leaves extract. Leaves were extracted (see section Experimental) after a chase period of 12 days. The solvent was

CDCl3. C. 1,n-ADEQUATE (8 Hz) of crude leaves extract. Leaves were extracted (see section Experimental) after a chase period of 12 days. The solvent was CDCl3.

Page 11 of 11

Indeed, the expected long-range couplings between C-4 and C-1, and 6. Römisch-Margl W, Schramek N, Radykewicz T, Ettenhuber C, Eylert E, et 13 al. (2007) CO2 as a universal metabolic tracer in isotopologue perturbation C-9 and C-3, respectively, could be confirmed by 1,n-ADEQUATE experiments. Phytochemistry 68: 2273-2289. experiments which are based on long-range 13C-13C transfer (i.e. via two, or three bonds) followed by direct 13C-1H coupling (i.e. via one 7. Ostrozhenkova E, Schramek N, Winzenhörlein B, Bacher A, Eisenreich W (2009) Precursor strategies for studies of isoprenoid biosynthesis in plants. In: 13 13 bond). As shown in Figure 11C, peaks due to C-1 → C-4 → H-4 and Palazon J, Cusido RM (eds) Plant secondary terpenoids. Research Signpost, 13C-3 → 13C-9 → H-9 were observed. These correlations reflected that Kerala, pp 75–95 13 the carbon atoms 1, 6, 4 and 3, 2, 9 were incorporated as C-3-moieties 8. Schramek N, Wang H, Römisch-Margl W, Keil B, Radykewicz T, et al. (2010) 13 13 (indicated in Figure 11C as arrows connecting the outlier C with a Artemisinin biosynthesis in growing plants of Artemisia annua. A CO2 study. pair of 13C-atoms) during the biosynthetic process. Indeed, these 13C-3 Phytochemistry 71: 179-187. species can be predicted for thymol from 13C-3-labelled GAP via the 9. Eisenreich W, Huber C, Kutzner E, Knispel N, Schramek N (2013) Isotopologue MEP pathway. profiling - towards a better understanding of metabolic pathways. In: Weckwerth W, Kahl G (eds) The handbook of plant metabolomics. Wiley VCH, Weinheim, In summary, the NMR-based pathway analysis confirmed that the pp 25–56. monoterpene thymol from T. transcaucasicus was made predominantly 10. Szecowka M, Heise R, Tohge T, Nunes-Nesi A, Vosloh D, et al. (2013) or exclusively via the MEP route of isoprenoid biosynthesis under the Metabolic fluxes in an illuminated Arabidopsis rosette. Plant Cell 25: 694-714. physiological conditions of the CO2 experiment. It should also be 11. Kasumov FY, Gadzhieva TG (1982) Components of Thymus transcaucasicus. emphasized that thymol from the flowers showed the same labelling Chem Nat Compd 18(5):637–638 patterns as that from the leaves. Obviously, thymol was either made in 12. Kasumov F, Komarova V (1983) Essential oils of Thymus transcaucasicusRonn. both plant organs via the same pathway or was made in one organ and and Thymus eriophorusRonn. Maslo-Zhir. Prom.-st.:29 then rapidly distributed over the whole plant. The labelling pattern also 13. Manukyan A (2012) Bioactive compounds and their health-promoting capacity reflected that the linear precursor geranyl diphosphate was converted of some Caucasian endemic and rare medicinal plants. Acta Hort. (ISHS) into the cyclic compound via the known mechanisms of thymol (955):339–346. biosynthesis. 14. Eisenreich W, Bacher A, Arigoni D, Rohdich F (2004) Biosynthesis of isoprenoids via the non-mevalonate pathway. Cell Mol Life Sci 61: 1401-1426.

Conclusion 15. Gräwert T, Groll M, Rohdich F, Bacher A, Eisenreich W (2011) Biochemistry 13 of the non-mevalonateisoprenoid pathway. Cell Mol Life Sci 68: 3797-3814. The CO2 pulse-chase experiment was useful for pathway analysis in T. transcaucasicus. It should be noted that the experimental method 16. Nagegowda DA (2010) Plant volatile terpenoid metabolism: biosynthetic is not restricted to the analysis of terpene metabolism but can also be genes, transcriptional regulation and subcellular compartmentation. FEBS Lett 584: 2965-2973. 13 applied to virtually any metabolic (end)-product found in CO2-labelled plants as shown by many earlier studies [6-10]. The model study with 17. Vranová E, Coman D, Gruissem W (2012) Structure and dynamics of the 13 isoprenoid pathway network. Mol Plant 5: 318-333. T. transcaucasicus shows that a single CO2 pulse-chase experiment can not only provide valuable data for pathway analysis, but also can 18. Aharoni A, Jongsma MA, Bouwmeester HJ (2005) Volatile science? Metabolic engineering of terpenoids in plants. Trends Plant Sci 10: 594-602. enable 13C-based metabolite profiling at significantly enhanced NMR sensitivity without hampering GC-MS based compound assignment. 19. Flügge UI, Gao W (2005) Transport of isoprenoid intermediates across chloroplast envelope membranes. Plant Biol (Stuttg) 7: 91-97. However, the NMR sensitivity is still lower than in GC-MS analysis, mainly due to the low 13C-enrichments afforded by13 CO pulse- 20. Hemmerlin A, Harwood JL, Bach TJ (2012) A raison d’être for two distinct 2 pathways in the early steps of plant isoprenoid biosynthesis? Prog Lipid Res chase experiments. On this basis, the study design combines benefits 51: 95-148. of isotope labelling for pathway analysis as well as for metabolite 21. Poulose AJ, Croteau R (1978) Biosynthesis of aromatic monoterpenes: detection. Thus, the experimental settings appear of general value in conversion of gamma-terpinene to p-cymene and thymol in Thymus vulgaris L. metabolomics studies with intact plants. Arch Biochem Biophys 187: 307-314. Acknowledgments 22. Eisenreich W, Sagner S, Zenk MH, Bacher A (1997) Monoterpenoid essential oils are not of mevalonoid origin. Tetrahedron Letters 38(22):3889–3892. We thank the Hans-Fischer-Gesellschaft (München) and the Deutsche Forschungsgemeinschaft (EI 384/8-1) for generous sponsoring of this research 23. Brauman JI (1966) Least squares analysis and simplification of multi-isotope work. We also thank the DAAD for supporting A.M. mass spectra. Anal. Chem 38(4):607–610 References 24. Korzekwa K, Howald WN, Trager WF (1990) The use of Brauman’s least squares approach for the quantification of deuterated chlorophenols. Biomed 1. Fernie AR, Aharoni A, Willmitzer L, Stitt M, Tohge T, et al. (2011) Environ Mass Spectrom 19: 211-217. Recommendations for reporting metabolite data. Plant Cell 23: 2477-2482. 25. Lee WN, Byerley LO, Bergner EA, Edmond J (1991) Mass isotopomer analysis: 2. Giavalisco P, Köhl K, Hummel J, Seiwert B, Willmitzer L (2009) 13C isotope- theoretical and practical considerations. Biol Mass Spectrom 20: 451-458. labeled metabolomes allowing for improved compound annotation and relative quantification in liquid chromatography-mass spectrometry-based metabolomic 26. Pickup JF, McPherson K (1976) Theoretical considerations in stable isotope research. Anal Chem 81: 6546-6551. dilution mass spectrometry for organic analysis. Anal Chem 48(13):1885–1890.

3. Sekiyama Y, Chikayama E, Kikuchi J (2011) Evaluation of a semipolar 27. Stahl-Biskup E (2002) Essential oil chemistry of the genus Thymus - A global solvent system as a step toward heteronuclear multidimensional NMR-based view. In: Stahl-Biskup ESF (ed) Medicinal and aromatic plants - Industrial metabolomics for 13C-labeled bacteria, plants, and animals. Anal Chem 83: profiles 24 (Thyme), vol. 24. Taylor & Francis, London, pp 75–124. 719-726. 28. Laihia K, Kolehmainen E, Malkavaara P, Korvola J, Mänttäri P, et al. (1992) 4. Daunicht H, Richter C, Resemann J (1995) An autonomous plant cuvette for 1H, 13C and 17O NMR spectroscopic study of four bicyclo[3.1.1]heptenes advanced climate control and gas exchange measurement. SAE International (derivatives of a-pinene) and four bicyclo[3.1.1]heptanes (derivatives of Technical paper 95, 1536, pp7 ICES, San Diego (CA), Warrendale, PA ß-pinene). Magn Reson Chem 30(8):754–759.

5. Krug H, Fink M (1987) Gas-exchange measuring system with adaptive climate 29. Abraham RJ, Warne MA, Griffiths L (1998) Proton chemical shifts in NMR. Part control. Biotronics (16):71–74 12.1 Steric, electric field and conformational effects in acyclic and cyclic ethers. J Chem Soc, Perkin Trans. 2(8):1751–1758.

Metabolomics ISSN: 2153-0769 JOM an open access journal Volume 4 • Issue 1 • 1000129