Essential Oils from Bolivia. XV. Herzogole, an Original Monoterpene Benzodioxole from an Essential Oil from Pentacalia herzogii (Cabrera) Cuatrec

Over 15 years, with the support of a Canadian funding agency, the Universidad Mayor de San Simón, in Bolivia, undertook a large survey of aromatic plants of the South American country. More than a hundred species were studied under various aspects, including the production and characterization of essential oils. As part of this survey, the chemical composition of an essential oil sample obtained from Pentacalia herzogii (Asteraceae) growing wild in the High Valley region of the department of Cochabamba was determined by a combination of GC and GC-MS measurements. α-Pinene was the main constituent of this essential oil (34%), accompanied by limonene (22%) and germacrene D (7.5%) as well as an important fraction of methoxylated monoterpenoids. They were mainly isomers of thymol methyl ether, accounting for 13% of the chromatogram. A new quantitatively important compound (9%) was identified through NMR and chemical synthesis as 4-isopropyl-6-methylbenzo[d][1,3]dioxole, and designated herzogole, alongside the minor related compound 1-isopropyl-2,3-dimethoxy-5-methylbenzene. The monoterpene benzodioxole featured a distinctive green-phenolic aroma which could raise interest for fragrance use. Since these compounds were not known naturally, a biosynthetic mechanism of their formation was proposed and put in perspective to illustrate the metabolic originality of P. herzogii.


Introduction
In the late 1980s until the early 2000s, several research groups of the Universidad Mayor de San Simón, Cochabamba, Bolivia, have driven a large survey of the Bolivian flora. This effort was conducted with the support of the International Development Research Centre (IDRC) in Canada. In this regard, agronomists, biochemists, phytochemists, engineers and chemists joined efforts to gather knowledge about national plant species. One of the objectives of this multidisciplinary endeavor was to identify potential crops and vegetable resources that could in part replace the traditional coca plantations in Bolivia and provide revenue to local populations.
Within the phytochemical investigation efforts, one group was particularly devoted to the examination of essential oils. Given their expertise in the field, partners from the Université du Québec à Chicoutimi, Québec, Canada, joined the project to provide support

Results and Discussion
The composition of the P. herzogii essential oil sample appears in Table 1. This batch of essential oil is characterized by a high percentage of α-pinene (33.6%), limonene (21.6%) and germacrene D (7.5%). More interestingly, three oxygenated monoterpenes with molecular mass of 164 amu account for 13.2% of the essential oil. They are three isomers, namely 1-methoxy-4-methyl-2-(1-methylethyl)-benzene, 2-methoxy-4-methyl-1-(1-methylethyl)benzene and 2-methoxy-1-methyl-4-(1-methylethyl)-benzene, more commonly known as isothymol, thymol and carvacrol methyl ethers, respectively [12]. The first two molecules account for about the same percentages (6.0% and 6.6%). The carvacrol methyl ether percentage is less than 1%. Overall, this profile corresponds well with the preliminary observations on a previous P. herzogii essential oil [10], which featured 32% α-pinene, 12% isothymol methyl ether, 11% thymol methyl ether and 8% germacrene D; limonene was, however, not found in significant proportions back then (0.5%). In the Asteraceae family, the genus Senecio sensu lato occupies an important place (more than 1500 species). Despite this great diversity, only about one hundred and sixty papers devoted to the study of the essential oils and volatile compounds of Senecio species (which was the former genus of P. herzogii) are reported in the SciFinder ® database [13]. Less than ten of them show the presence of low percentages (<2%) of thymol methyl ether and only three the occurrence of carvacrol methyl ether ( Table 2) [14][15][16][17][18][19][20][21][22]. Surprisingly, one paper mentions the ambiguous o-methyl thymol as an identified compound in the S. scandens essential oil [19]. In South America, the reported studies indicate that hydrogenated monoterpenes constitute the most important part of the identified compounds in Senecio essential oils. For example, this is the case of essential oils obtained from Argentina [23,24], Chile [25] and Peru [26]. Three members of this genus Senecio stand apart. Essential oil of the leaves S. crassiflora from Brazil shows the presence of two quantitatively important sesquiterpenols, τ-muurolol and α-cadinol. Germacrene D is the most important compound of the stem of the same species [27]. Another hydrogen sesquiterpene, α-zingiberene, is also the main constituent of the S. selloi essential oil from Brazil [28]. Dehydrofukinone is by far the most important compound observed in the case of S. viridis from Argentina, since it constitutes more than 92% of the essential oil [29].
This comparison with other closely related species of Asteraceae suggests that the P. herzogii specimens that yielded this essential oil were phytochemically original, based on the high proportions of monoterpenic ethers. Thymol or isothymol methyl ethers as a proportionally important essential oil constituent are more commonly encountered in Lamiaceae species, e.g., Origanum [30,31] and Thymus [32], or Apiaceae species, such as Cyclospermum [33], Ammoides [34] and Apium [35]. In these cases, the methyl ethers accumulate in conjunction with thymol, carvacrol or thymoquinone dimethyl ether. Another pattern is found in the Apiaceae Crithmum maritimum [36][37][38][39], where important proportions of thymol methyl ether correlate with large amounts of γ-terpinene, a direct metabolic precursor of thymol [40]. Of these four biosynthetically correlated constituents, only small amounts of γ-terpinene and traces of carvacrol are encountered in the studied P. herzogii essential oil accession, further highlighting the metabolic originality of those individuals among aromatic plants owing to their ability to accumulate methyl ethers selectively.
The percentage for this essential oil of a fourth and until now unidentified compound 1 is 9.2%, and its structural elucidation was one of the driving interests for the present study. Indeed, thoroughly characterizing essential oils is important from both a theoretical and practical perspective. On the one hand, it is an important endeavor to keep on expanding the knowledge assembled over decades of research on essential oils in an effort to characterize both major and minor compounds, which can then in turn refine the comprehension of biosynthetic routes involved or allow for easier analysis of other species later on (e.g., [41]). On the other hand, the quality control of essential oils relies on knowledge of both quantitatively important compounds but also minor signature compounds, and the continued expansion of the trade of essential oils worldwide calls for precise and thorough analyses, which in turn rely on expanding theoretical knowledge.
This unknown compound has a very similar mass spectrum to those of the three methoxy ethers just identified above, and it was hypothesized that it bore a metabolic relationship to the latter. The main difference lies in the fact that its main ions are 14 uma units higher. This value could have suggested the presence of a CH 2 group added to one of the three methoxy ethers. Taking thymol methyl ether as a reference, the unidentified compound has a retention index of about 100 units higher on the non-polar column, in agreement with the known additive rule for the retention indices [42,43]. However, this difference in the polar column is about 200 units higher and then does not follow the same additive rule, ruling out the ethoxyl substituent hypothesis. A sample of essential oil was therefore submitted to column chromatography to afford a fraction of pure compound 1 as a clear liquid. 1 D NMR experiments suggested the presence of an aromatic ring, three methyls, one methylene and one methine. Methylene C-11 was notably deshielded on both 1 H and 13 C spectra, with shifts characteristic of acetal groups. This, along with the HMBC correlations of H-11 with aromatic C-1 and C-2 ( Figure 1), indicated that the compound featured a benzodioxole moiety. A spin system comprising H-8 and the equivalent methyls H-9 and H-10 was visible on the DQF-COSY spectrum, with multiplicities characteristic of an isopropyl substituent. The HMBC cross-peak between H-8 and aromatic methine C-5 and quaternary aromatics C-1 and C-6 indicated that this substituent was vicinal to one of the methylenedioxy groups and that the other ortho position was unsubstituted. Another methyl singlet, H-7, was also observed on the 1 H spectrum, with clear HMBC correlations to quaternary C-4 and aromatic methines C-5 and C-3, suggesting that both ortho positions were unsubstituted. Therefore, the observed spectroscopic data were reasonably consistent with the structure of 4-isopropyl-6-methylbenzo[d][1,3]dioxole 1. Its mass spectrum is featured for reference in Figure 2. However, an unexpected HMBC cross-peak was found between methyl H-7 and the methylenedioxy attachment point C-1, beyond typical correlations expected. Furthermore, 1 H peaks for H-5 and H-3 exhibited the same chemical shift, appearing as an indistinct singlet that did not allow the observation the characteristic meta 2-3 Hz coupling constant that would have fully validated the proposed structure. These peculiarities prompted further investigation. The putative structure of 1 was therefore confirmed through chemical synthesis. Commercial 5-methyl-3-(propan-2-yl)benzene-1,2-diol 2 was reacted with diiodomethane in the presence of cesium carbonate in hot dimethylformamide [44], affording compound 1 (Scheme 1). Superimposition of 1 H and 13 C spectra of the isolated and synthesized compounds confirmed that both molecules were identical and that the proposed structure was correct, despite the unusual HMBC correlation. Examination of the observed fragmentation pattern was also consistent with the structure (see Supplementary Materials). By analogy with safrole, another molecule with a benzodioxole nucleus, the compound was given the trivial name herzogole. The presence of a benzodioxole moiety in a monoterpenoid is somewhat surprising, as this moiety is more characteristic of phenylpropanoids. To our knowledge, this compound has never been mentioned in the literature (no match in the SciFinder®database [13]). Thus, the question of its origin must be discussed to discard any possibility of an artifactual observation. The GC-MS analysis carried out in 1998 on a preliminary sample indicated its presence in a higher concentration, at about 15% of the essential oil [10]. The use of any undisclosed biocide in this almost virgin region is highly improbable, especially 20 years apart and in such high proportions of the essential oil. These observations lead us to believe that this unknown compound is not of exogenous origin. Conversely, by analogy with the proposed biosynthesis from caffeoyl coenzyme A (3) to eugenol (4) which is then converted either to methyleugenol (5) or safrole (6) in Asarum sieboldii [45], a plausible biosynthetic tree can be proposed from isothymol 7, arising from the monoterpene biosynthesis (Scheme 2). The existence of a terpenophenol O-methyltransferase is apparent in P. herzogii given the presence of abundant thymol methyl ether, carvacrol methyl ether and isothymol methyl ether 9 in the essential oil. Isothymol 7 or the methyl ether 9 could also be hypothetically converted to the monomethylated intermediate 8 (or its isomer where the methyl is bonded to the other oxygen), not readily observed in the essential oil. The latter could then either proceed towards herzogole 1 by the action of a cytochrome P450 or be further methylated by an O-methyltransferase to generate 1-isopropyl-2,3-dimethoxy-5-methylbenzene 10 (which was only reported once in the literature as a synthesis intermediate [46]), as for the biosynthetic trade-off that exists between safrole 6 and methyleugenol 5. Compound 10, with a mass of 194 amu, was interestingly consistent with a small unknown compound observed at RI DB-5 = 1387, whose mass spectrum showed not only this mass but also clear m/z at −15 and −30 amu, consistent with the abduction of one or two methyls (Figure 2). To verify whether compound 10 was indeed present in the essential oil, it was also prepared from a small remaining amount of catechol 2, using excess dimethyl sulfate to methylate both hydroxyls. Although an insufficient amount was obtained to attempt purification and characterize the obtained yield, comparison of the NMR of the crude product to published data [46] clearly showed the formation of the desired compound as the main reaction product. This allowed us to establish the identity of 10 by direct comparison with the main GC peak of this crude product. The simultaneous presence of isothymol methyl ether 9, herzogole 1 and compound 10 supports the plausibility of the proposed biosynthetic route (Scheme 2), which serves as an argument in favor of the endogenous nature of herzogole within the distilled accession of P. herzogii. This yet unique biosynthetic pattern leading to accumulation of a rare monoterpenic benzodioxole further stresses the metabolic originality of this species' sample and illustrates how the examination of previously unstudied plants can uncover new aspects of phytochemistry. pound 1. The isolated product has a strong green-phenolic scent, like that of the whole P. herzogii essential oil obtained in the course of this study. One could not rule out, however, that a minute impurity with high aroma potency from the essential oil would induce this scent. The synthesized product featured the same aroma, while being unlikely to contain the same hypothetical impurity as the essential oil-derived purified fraction. This suggests with some confidence that herzogole is one of the major aroma contributors of the P. herzogii essential oil and could by itself bear some potential valorization as a fragrant compound. Otherwise, a broader screening of P. herzogii, which is encountered as scattered individuals in the wild, should be conducted in the future to further assess potential variability of this metabolic pathway following, e.g., environmental or genetic factors. Although sesquiterpenes, to the exception of germacrene D, are not prominent constituents of the tested essential oil, some minor compounds need a brief justification of their identification. Both gynurenol and gynuradienol are rather rare compounds. They were identified as volatile constituents from the roots of Gynura bicolor DC, another plant belonging to the Asteraceae family. Their retention indexes on both non-polar and polar columns agree with the published values. The same is true for the recorded mass spectra [47]. 4,10-Diepi-guaiol was identified by comparison with an analysis of Boswellia occulta essential oil on the same chromatographic system, following published results [48]. Isoguaiols are also unusual compounds. They were described through their 1 H-NMR, 13 C-NMR and mass spectra. Their retention indices on both polar and non-polar columns were also published [41]. In fact, of eight possible optical isomers, four were characterized, The above synthesis of herzogole allowed us to confirm another property of compound 1.
The isolated product has a strong green-phenolic scent, like that of the whole P. herzogii essential oil obtained in the course of this study. One could not rule out, however, that a minute impurity with high aroma potency from the essential oil would induce this scent. The synthesized product featured the same aroma, while being unlikely to contain the same hypothetical impurity as the essential oil-derived purified fraction. This suggests with some confidence that herzogole is one of the major aroma contributors of the P. herzogii essential oil and could by itself bear some potential valorization as a fragrant compound. Otherwise, a broader screening of P. herzogii, which is encountered as scattered individuals in the wild, should be conducted in the future to further assess potential variability of this metabolic pathway following, e.g., environmental or genetic factors.
Although sesquiterpenes, to the exception of germacrene D, are not prominent constituents of the tested essential oil, some minor compounds need a brief justification of their identification. Both gynurenol and gynuradienol are rather rare compounds. They were identified as volatile constituents from the roots of Gynura bicolor DC, another plant belonging to the Asteraceae family. Their retention indexes on both non-polar and polar columns agree with the published values. The same is true for the recorded mass spectra [47]. 4,10-Diepi-guaiol was identified by comparison with an analysis of Boswellia occulta essential oil on the same chromatographic system, following published results [48]. Isoguaiols are also unusual compounds. They were described through their 1 H-NMR, 13 C-NMR and mass spectra. Their retention indices on both polar and non-polar columns were also published [41]. In fact, of eight possible optical isomers, four were characterized, namely isoguaiol A, B, C and D: (1βH,7αH,10αH)-, (1αH,7αH,10βH)-, (1βH,7αH,10βH)and (1αH,7αH,10αH)-guai-4-en-11-ol, respectively; they all feature very similar mass spectra. Given this ambiguity and the differences in reported retention indexes between the two chromatographic systems, the observed peak could correspond either to one or a mixture of at least two of the available structures. Thus, the proposed identification as isoguaiol isomer(s) is not unreasonable. Bulnesoxide was tentatively identified based on the same study [41], and a liguloxide epimer also was conditionally labeled based on high mass spectral similarity, but a non-matching retention index with liguloxide itself.

Conclusions
The present study allowed us to clarify a two-decade-old unresolved result from a larger screening of Bolivian aromatic plants essential oils, conducted to provide valorization perspectives of the local flora and alternatives to the traditional coca culture. Previously, despite promising yield and aroma, an essential oil of Pentacalia herzogii could not be fully characterized earlier due to the presence of a major unknown constituent. Reexamination of a freshly distilled essential oil allowed us to identify herzogole, a previously unreported monoterpenic benzodioxole, as one of the main constituents of the essential oil, alongside thymol and isothymol methyl ethers. Rationalization of its probable biosynthesis and comparison with essential oils from related Senecio species and typical monoterpene methyl ether rich essential oils illustrate how those specimens of P. herzogii stand out as metabolically original among aromatic plants and allowed us to further identify a minor dimethoxylated monoterpene not previously reported in nature. Herzogole also bears a characteristic green-phenolic aroma, akin to that of the whole essential oil studied herein. This study overall allowed us to report 92 compounds of this essential oil and provided solid analytical ground for future studies and valorization of this plant in Bolivia.

Plant Material
At the end of the flowering period, the aerial parts of P. herzogii were collected along the road going from Corani-Pampa to Tablas

Essential Oil Extraction
A total biomass of 0.5 kg of the aerial parts (frozen leaves and stems collected together from several individuals) was hydrodistilled for 3 h in a glass distillation apparatus (FIG-MAY, Argentina) set to maintain a constant level of 1.5 L of water through continuous water supply to yield 0.9% of the essential oil. Yield was calculated by dividing the mass of essential oil obtained by the mass of the frozen raw material, and multiplying by 100%. It was dried on anhydrous sodium sulfate and stored in amber glass vials at 4-5 • C until analysis. A single distillation was conducted.

General Experimental Procedures
The GC parameters are described in Table 3. Identification of the components was done by comparison of their retention indices (RIs) with normal alkanes from C7 to C40 and by comparison of their mass spectra with literature data [49][50][51] and with our own databases. Quantitative data were obtained electronically from GC-FID area percentages without correction (internal normalization). NMR spectra were recorded with a Bruker Avance 400 spectrometer at 400 MHz for 1 H nuclei and 101 MHz for 13 C nuclei, using deuterated chloroform (CDCl 3 ) as the solvent. Chemical shifts are reported in ppm relative to the solvent residual peak (δ = 7.26 ppm for 1 H and 77.16 ppm for 13 C). Reactions were monitored by thin-layer chromatography (TLC) with silica gel 60 F 254 0.25 mm pre-coated aluminum foil plates (MilliPore) and visualized under UV 254 . Chemical reactants were from Millipore Sigma (Burlington, MA, USA). [1,3]dioxole (herzogole; 1): Crude essential oil (1.49 g) was deposited on a chromatographic column comprising 200 g of silica gel 60 (15-40 µm, Silicycle, Québec, QC) previously wetted with n-hexane. The sample was eluted with pure n-hexane, then 9:1 and 8:2 mixtures of n-hexane and dichloromethane to afford compound 1 as a green-phenolic-scented light-yellow liquid (27 mg), alongside more fractions of 1 contaminated with thymol and isothymol methyl ethers. Compound 1 was also prepared by suspending 5-methyl-3-(propan-2-yl)benzene-1,2-diol (2; 63 mg, 0.38 mmol) in 3 mL dry dimethylformamide under argon atmosphere, then adding Cs 2 CO 3 (186 mg, 0.57 mmol) and diiodomethane (45 µL, 0.57 mmol). The solution was heated to 110 • C for one hour, cooled down, then quenched with saturated aqueous NH 4 Cl. The reaction product was extracted with 3 × 5 mL ethyl acetate, and the combined organic phases were dried over sodium sulfate and evaporated under reduced pressure. The dried residue was purified by flash chromatography over 4 g silica, eluting with 15 mL pure n-hexane then nhexane/dichloromethane (7:3) until the product was obtained as a green-phenolic-scented clear liquid (30.6 mg, 45% yield).

Isolation, Synthesis and Structural Elucidation
RI 1-Isopropyl-2,3-dimethoxy-5-methylbenzene (10): To a solution of diol 2 (14.9 mg, 89.6 µmol) in acetone (5 mL) was added potassium carbonate (131 mg, 948 µmol) and dimethyl sulfate (170 µL, 1.79 mmol). The mixture was stirred at room temperature for 18 h. The mixture was then cooled to 0 • C, quenched with an aqueous solution of sodium hydroxide (1 M, 10 mL) and stirred at room temperature for 3 h. It was then extracted with dichloromethane (3 × 10 mL), washed with a saturated aqueous solution of ammonium chloride (2 × 10 mL), water (2 × 10 mL) and brine (10 mL). The organic layer was dried over anhydrous sodium sulfate, filtered and the solvent was rotary evaporated. Comparison of the NMR data of the crude residue with the literature description [46] confirmed the formation of the expected dimethylation product. RI