Modification of Natural Eudesmane Scaffolds via Mizoroki-Heck Reactions

The Mizoroki-Heck reaction was applied to substrates derived from isocostic and ilicic acids, important sesquiterpene components of Dittrichia viscosa L. Greuter that were extracted directly from plant material collected in Morocco. After optimization of the metallo-catalysis conditions, various aryl-groups were successfully introduced on the exocyclic double bond with an exclusive E-configuration and without racemization.


Introduction
Natural products are a well-known continuous source of inspiration for the design of new bioactive agents with applications in the therapeutic, cosmetic or agricultural areas [1]. They can serve as building blocks for the synthesis of more complex bioactive compounds [2]. Based on traditional use, numerous plants of medicinal interest have been identified and conventional drugs developed from their extracts [3,4]. As part of our program studying Moroccan plants [5,6], our interest is focused on Dittrichia viscosa L. Greuter [7,8], an invasive perennial plant particularly abundant in wasteland areas. Its extract is used in traditional Moroccan medicine for its antipyretic, antiseptic and anti-inflammatory properties [9,10], and also exhibits antifungal activity [11][12][13]. Recent studies reported that the aerial parts of the plant are a rich source of eudesmane sesquiterpenes, among which ilicic acid (1) and isocostic acid (2) (Figure 1) represent up to 2.5% and 2% of the aerial part dry weight, respectively [5]. This plant, like others, could represent a renewable source of enantiopure compounds to obtain diversified libraries of products of interest.

Optimization of the Mizoroki-Heck Reaction on Ilicic Acid
We first examined the behavior of ilicic acid methyl ester in a model Mizoroki-Heck reaction (Scheme 1). Unfortunately only starting material was recovered [20]. Protection of the alcohol was then considered and the methoxymethyl-ether (MOM) group was selected as the most efficient protecting group allowing the isolation of compound 3 in 60% yield. With 3 as starting material, Indeed, sesquiterpene derivatives have attracted considerable attention lately [14] due to their pharmacological and phytochemical activities, in particular sesquiterpene lactones [15][16][17] which

Optimization of the Mizoroki-Heck Reaction on Ilicic Acid
We first examined the behavior of ilicic acid methyl ester in a model Mizoroki-Heck reaction (Scheme 1). Unfortunately only starting material was recovered [20]. Protection of the alcohol was then considered and the methoxymethyl-ether (MOM) group was selected as the most efficient protecting group allowing the isolation of compound 3 in 60% yield. With 3 as starting material, compound 4b was then obtained in 30% yield in presence of Pd(OAc)2 (0.1 equiv), p-tolyl iodide (1.1 equiv) and Et3N (3 equiv) in N,N-dimethylformamide (DMF) ( Table 1, entry 1). The use of silver acetate as oxidizing agent and base offered no significant advantages (entry 2) [35,36]. Next a bulky electron-rich phosphine, tri(o-tolyl)phosphine was used, enabling the isolation of 4b in 65% yield, and acetonitrile was used to replace DMF as solvent, but a reduced yield was noted (entry 5) [21,22]. An increase in the catalytic system loading failed to improve this result (entry 6). The optimized conditions were as follows: 3 (1 equiv), aryl iodide (1.1 equiv), Et3N (3 equiv), Pd(OAc)2 (0.1 equiv) and P(o-Tol)3 (0.1 equiv) in DMF at 120 °C for 24 h. These conditions were extended to aryl bromides without any significant loss of reactivity (entry 8).

Substrate Scope and Deprotection
Various aryl iodides and bromides were then used to generate a library of eudesmane analogues using the optimized cross-coupling reaction conditions (Scheme 2 and Table 2). The reactions were clean and the expected products were synthesized in good yields. As expected, the palladium coupling reaction tolerated different aromatic halides bearing electron-donating (Me, OMe) and withdrawing groups (F, CO2Me, CHO) in the ortho, meta and para-positions (compounds 4a-i). Then, Scheme 1. Mizoroki-Heck reaction of ilicic acid.

Substrate Scope and Deprotection
Various aryl iodides and bromides were then used to generate a library of eudesmane analogues using the optimized cross-coupling reaction conditions (Scheme 2 and Table 2). The reactions were clean and the expected products were synthesized in good yields. As expected, the palladium coupling reaction tolerated different aromatic halides bearing electron-donating (Me, OMe) and withdrawing groups (F, CO 2 Me, CHO) in the ortho, meta and para-positions (compounds 4a-i). Then, sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9). sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9). Scheme 2. MOM deprotection conditions.  sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9). Scheme 2. MOM deprotection conditions. sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9).  sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9).  sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9).  sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9).  sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9).  sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9). sesquiterpenes 5a-i were rapidly generated under acidic conditions in good to excellent yields. When the fluoro substituent was in ortho position, deprotection occurred simultaneously with the Mizoroki-Heck reaction, compound 4g was never observed and 5g was isolated in one step in 62% yield (entry 7). The methodology was also extended to heterocyclic derivatives and the desired product 4i was isolated with good yield (entry 9).  A Nuclear Overhauser Effect Spectroscopy (NOESY)-NMR experiment on 5b allowed us to determine the double bond configuration. Interactions between H7-HAr and H7-H22 involved an E-olefin geometry emphasized by the absence of signals between H7 and H13 ( Figure 2). Our stereochemical result is in accordance with the work reported by Colby and co-workers [20]. 5i, 80% a Experimental conditions: 3 (1 mmol), aryl-halide (1.1 mmol) dissolved in DMF (2 mL), 120 • C, for 24 h; b Isolated yields; c Compound 4g was not observed during the reaction, 5g was directly obtained.

Scheme 2. MOM deprotection conditions.
A Nuclear Overhauser Effect Spectroscopy (NOESY)-NMR experiment on 5b allowed us to determine the double bond configuration. Interactions between H 7 -H Ar and H 7 -H 22 involved an E-olefin geometry emphasized by the absence of signals between H 7 and H 13 ( Figure 2). Our stereochemical result is in accordance with the work reported by Colby and co-workers [20].

Mizoroki-Heck Reaction on Esterified Isocostic Acid.
Next we focused on the other major constituent found on the acidified dichloromethane (DCM) extracts of Dittrichia viscosa L. Greuter. Isocostic acid (2) was esterified and submitted to our optimized Mizoroki-Heck conditions leading to a complex mixture of products. When an epoxidation was carried out on esterified compound 6 in order to prevent the migration of the endocyclic double bond, the epoxide 7 was obtained as a unique enantiomer (Scheme 3) [6].

Mizoroki-Heck Reaction on Esterified Isocostic Acid
Next we focused on the other major constituent found on the acidified dichloromethane (DCM) extracts of Dittrichia viscosa L. Greuter. Isocostic acid (2) was esterified and submitted to our optimized Mizoroki-Heck conditions leading to a complex mixture of products. When an epoxidation was carried out on esterified compound 6 in order to prevent the migration of the endocyclic double bond, the epoxide 7 was obtained as a unique enantiomer (Scheme 3) [6].

Mizoroki-Heck Reaction on Esterified Isocostic Acid.
Next we focused on the other major constituent found on the acidified dichloromethane (DCM) extracts of Dittrichia viscosa L. Greuter. Isocostic acid (2) was esterified and submitted to our optimized Mizoroki-Heck conditions leading to a complex mixture of products. When an epoxidation was carried out on esterified compound 6 in order to prevent the migration of the endocyclic double bond, the epoxide 7 was obtained as a unique enantiomer (Scheme 3) [6].

Mizoroki-Heck Reaction on Esterified Isocostic Acid.
Next we focused on the other major constituent found on the acidified dichloromethane (DCM) extracts of Dittrichia viscosa L. Greuter. Isocostic acid (2) was esterified and submitted to our optimized Mizoroki-Heck conditions leading to a complex mixture of products. When an epoxidation was carried out on esterified compound 6 in order to prevent the migration of the endocyclic double bond, the epoxide 7 was obtained as a unique enantiomer (Scheme 3) [6].   The reaction was successfully performed with ortho-, metaand para-aryl iodides substituted with electron donating groups (compounds 8a-e). However, the steric hindrance of aryl iodides substituted in the ortho-position had a negative effect on the transformation, as compound 8e was isolated only in 33% yield. When an electron-deficient aryl iodide was used, only traces of the expected product 8f were obtained. Unfortunately, aryl bromides proved incompatible with our optimized conditions as only the opening of the epoxide was observed (Scheme 4).
The reaction was successfully performed with ortho-, meta-and para-aryl iodides substituted with electron donating groups (compounds 8a-e). However, the steric hindrance of aryl iodides substituted in the ortho-position had a negative effect on the transformation, as compound 8e was isolated only in 33% yield. When an electron-deficient aryl iodide was used, only traces of the expected product 8f were obtained. Unfortunately, aryl bromides proved incompatible with our optimized conditions as only the opening of the epoxide was observed (Scheme 4).

Scheme 4. Control experiments.
The lack of reactivity of the aryl bromide was confirmed by heating 7 in DMF in the absence of the other reagents. This experiment showed the formation of 9a and 9b in almost the same proportions. The isolation of 9a and 9b was possible by column chromatography and each compound was fully characterized by comparing their spectroscopic data with the literature [6,37]. Further investigation was then conducted on the major product 9a resulting from the epoxide ring opening. The protection of the alcohol 9a was made as before with a MOM group leading to compound 10 then cross coupling conditions were applied and expected compound 11 was obtained in good yield. When 9a was submitted directly to the cross coupling reaction with a more reactive aryl iodide, only 27% of 8b was isolated along with degradation products, indicating that our cross-coupling conditions are not compatible with the presence of an unprotected alcohol. For this other major constituent found on the acidified DCM extracts of Dittrichia viscosa L. Greuter, the configuration of the trisubstituted double bond synthetized was examined though the correlations obtained by the NOESY-NMR experiment. A spatial correlation was observed between the aromatic proton and H7 when no signal was noticed between H7 and H13 led us conclude on the E-olefin geometry. The lack of reactivity of the aryl bromide was confirmed by heating 7 in DMF in the absence of the other reagents. This experiment showed the formation of 9a and 9b in almost the same proportions. The isolation of 9a and 9b was possible by column chromatography and each compound was fully characterized by comparing their spectroscopic data with the literature [6,37]. Further investigation was then conducted on the major product 9a resulting from the epoxide ring opening. The protection of the alcohol 9a was made as before with a MOM group leading to compound 10 then cross coupling conditions were applied and expected compound 11 was obtained in good yield. When 9a was submitted directly to the cross coupling reaction with a more reactive aryl iodide, only 27% of 8b was isolated along with degradation products, indicating that our cross-coupling conditions are not compatible with the presence of an unprotected alcohol. For this other major constituent found on the acidified DCM extracts of Dittrichia viscosa L. Greuter, the configuration of the trisubstituted double bond synthetized was examined though the correlations obtained by the NOESY-NMR experiment. A spatial correlation was observed between the aromatic proton and H 7 when no signal was noticed between H 7 and H 13 led us conclude on the E-olefin geometry.

General Methods
All reagents were purchased from commercial suppliers and were used without further purification except for DMF, which was stored under argon and activated molecular sieves. The reactions were monitored by thin-layer chromatography (TLC) analysis using silica gel (60 F254) plates. Compounds were visualized by UV irradiation. Flash column chromatography was performed on silica gel 60 (230-400 mesh, 0.040-0.063 mm). 1 H-NMR and 13 C-NMR spectra were recorded on Avance spectrometers (Bruker, France, SAS) at 250.13 MHz ( 13 C, 62.9 MHz) or 400.13 MHz ( 13 C, 100.62 MHz). Chemical shifts are given in parts per million from tetramethylsilane (TMS) as internal standard. The following abbreviations are used for the proton spectra multiplicities: s: singlet, d: doublet, t: triplet, q: quartet, qt: quintuplet, m: multiplet. Coupling constants (J) are reported in Hertz (Hz). Multiplicities were determined by DEPT-135 sequences. Attributions of protons and carbons were made with the help of Heteronuclear Single Quantum Correlation (HSQC) and Heteronuclear Multiple Bond Correlation (HMBC) 2D NMRs. Eudesmane numbering of carbons was used instead of the IUPAC numbering. High-resolution mass spectra (HRMS) were performed on a Maxis 4G instrument (Bruker, France, SAS).

General Procedure for Synthesis of Compounds 5a-5i
To a solution of the previous substrates 4 (1 equiv) in diethyl ether (10 mL) HCl solution in ether (2 equiv) was added. The reaction mixture was stirred for 30 min at room temperature then water (10 mL) was added and mixture was extracted with diethyl ether (3 × 10 mL). The organic layers were dried (MgSO 4 ), filtered and concentrated under reduce pressure leading to the desired products after purification by flash chromatography on silica gel (petroleum ether/EtOAc: 80/20). To a solution of ester 6 (315 mg, 1.3 mmol) in 10 mL of dichloromethane were added (220 mg, 1.3 mmol) of m-chloroperbenzoic acid. The reaction mixture was stirred at room temperature for 3 h then washed with a solution of sodium bisulfite (10%) (3 × 10 mL) then a solution of sodium hydrogen carbonate (5%) (10 mL). The aqueous layers were combined and extracted with DCM (3 × 10 mL). The organic layers were combined, washed with water (10 mL), dried with MgSO 4 , filtered and concentrated under reduced pressure. The resulting residue was purified by flash chromatography on silica gel.