Synthetic Studies towards the Ent-labdane Diterpenoids: Rearrangement of Ent-halimanes

For the first time ent-labdanes have been synthetised starting from ent-halimic acid, following a route that is the reverse of the biosynthetic one leading to the former compounds.

Due to its functionality, ent-halimic acid can be viewed as an excellent synthon for the synthesis of new natural products.For example, the bicyclic system of ent-halimic acid can be visualized as the basis for the BC rings of tricyclic diterpenes or quassinoids with a picrasane skeleton (Scheme 1) whereby the C-4 quaternary carbon and the C-5 carbon of the starting material are incorporated with the correct configuration as C-10 and C-9, respectively, in the targets.However, to the best of our knowledge, a diterpene of the antipodal series has not been used previously as a starting material for the synthesis of diterpenes of the normal series or quassinoids with a picrasane skeleton.

Tricyclic diterpene
As can be observed in Scheme 1, the synthesis of the picrasane skeleton would require a rearrangement of the abeopicrasane skeleton, namely a 1,2 migration of the Me-18 group, in the same manner as seen in the biosynthetic pathway of euphane and tirucallane, biosynthetic precursors of the quassinoids [5].

ent-halimane ent-labdane
Scheme 2 shows the 1,2-shift of the Me-20 of ent-halimanes that leads to ent-labdanes.A rearrangement of this kind would be required to access the picrasane skeleton quassinoids.This rearrangement would be the opposite of the biosynthetic route, in which the ent-labdanes are the precursors of ent-halimanes [6].
In the present work we have explored the use of ent-halimic acid for the synthesis of biologically active compounds with different carbon skeletons.In particular we have studied the rearrangements with different Lewis acids of epoxides 3 and 4 (Figure 2), derived from ent-halimic acid.

Results and Discussion
Synthesis of the starting materials 3 and 4.
The starting materials 3 and 4 were obtained from compound 1, the methyl ester of ent-halimic acid, via two or four carbon atom side chain degradations, respectively, with methylketone 2 being the key intermediate in both syntheses (Scheme 3).Treatment of 2 with urea-hydrogen peroxide and trifluoroacetic acid (UHP-TFAA) [7] for one hour gives the epoxyacetates 3 and 5 with high stereoselectivity as a result of a Baeyer-Villiger reaction and epoxidation of the double bond.By contrast, the reaction of 2 with m-CPBA was slow, gave low yields and is non-stereoselective.The configurations of the epoxide groups were established by nOe experiments, as shown in Figure 3.While in compound 3 H-1 has an nOe with Me-17 and Me-20, in 5 a nOe can only be observed between H-1 and Me-20.In order to obtain compound 4 it was necessary to reduce the ketone in 2, followed by acetylation to give the corresponding acetate, which was treated with UHP-TFAA, under the same conditions as before, to give the corresponding epoxides 4 and 6 in excellent yield and with good stereoselectivity.

Treatment of epoxide 4 with Lewis acids
Corey et al. studied the rearrangement of epoxides in their synthesis of gicinoelepin A [8] and after using different Lewis acids and reaction conditions, achieved the rearrangement of an angular methyl with FeCl 3 -Ac 2 O.With this precedent in mind we started our study by treatment of epoxide 4 with FeCl 3 -Ac 2 O, and after several minutes a mixture of compounds 7 (20%) and 8 (56%) was obtained (Scheme 4).The less polar compound, 7, was identified as a heteroannular diene (U.V.: 238 nm) that could arise through the mechanism shown in Figure 4.The more polar compound 8 was a mixture of compounds which proved very difficult to separate by chromatography, so it was decided to simplify the analysis by carrying out the hydrolysis of the acetate group, followed by oxidation with TPAP/NMO [9], to thus obtain compounds 9 and 10 in good yield.The structures of these compounds were established by HMQC and HMBC experiments.The presence of an olefin and a methyl on a methine indicates that a skeletal rearrangement has taken place in the reaction of 4 with the Lewis acid.The observed correlation in the NMR experiments between Me-17 and C-10, tell us that B ring of ent-halimane has undergone a contraction to produce a five membered ring, giving a perhydroindene system.The mechanism that explains the formation of these compounds is shown in Figure 5.

OAc OAc
The Lewis acid coordination with acetic anhydride promotes a concerted rearrangement to produce olefins 8a and 8b.The stereochemistry for C-1, C-8 and C-10 is proposed in accordance with the shown mechanism.In a ROESY experiment of 10 (Figure 6) nOes can be observed between H-1 and Me-17, H-5 and Me-20, H-11 and H-2β and finally H-12 and Me-20, which corroborate the proposed configurations.Treatment of 4 with TiCl 4 gave similar results as those obtained with BF 3 .
Et 2 O, but in this case the corresponding hydroxyester 14 was obtained as well.As can be seen, the rearrangement of compound 4 with different Lewis acid did not give us the required results, so we decided to change the starting material to compound 3 which possesses a less demanding side chain.

Treatment of epoxide 3 with Lewis acids
Treatment of epoxide 3 with FeCl 3 /Ac 2 O takes place to give compounds 15 and 16 that were separated by column chromatography (Scheme 6).The less polar compound 15 corresponds to a diene similar to 7 obtained by the analogous treatment of 4. The more polar compound is 16, which is produced by a skeletal rearrangement similar to that of compounds 8.Besides diene 15 obtained in the previous treatment (Scheme 6), we obtained ketone 17 (analogous to 11), lactone 18, produced by transesterification of the hydroxy group in C-1 with the methoxycarbonyl at C-18 and, as the main component, a more polar mixture constituted by three compounds − 19, 20 and 21 − all of them ent-labdanes, as desired.In the 1 H-NMR spectra of 19 and 20 signals corresponding to a methyl group on a double bond can be observed in ring B instead of the methyl on a methine.In the minor compound, 21 a terminal double bond in ring B is observed instead.

Conclusions
The transformation of ent-halimanes to ent-labdanes has been achieved for the first time.This methodology open the way for the use of the available ent-halimanes for the synthesis of tri-and tetracyclic compounds of the normal series such as tricyclic diterpenes and degraded triterpenes like quassinoids with a picrasane skeleton.Future work will involve using this kind of rearrangement on tricyclic compounds with similar system to the tirucalanes and euphanes that are the biological precursors in which this rearrangement takes place.

General
Unless otherwise stated, all chemicals were purchased as the highest purity commercially available and were used without further purification.IR spectra were recorded on a BOMEM 100 FT-IR or an AVATAR 370 FT-IR Thermo Nicolet spectrophotometers. 1 H-and 13 C-NMR spectra were performed in CDCl 3 and referenced to the residual peak of CHCl 3 at δ 7.26 ppm and δ 77.0 ppm, for 1 H-and 13 C-, respectively, using Varian 200 VX and Bruker DRX 400 instruments.Chemical shifts are reported in δ ppm and coupling constants (J) are given in Hz.MS were performed at a VG-TS 250 spectrometer at 70 eV ionising voltage.Mass spectra are presented as m/z (% rel.int.).HRMS were recorded on a VG Platform (Fisons) spectrometer using chemical ionization (ammonia as gas) or Fast Atom Bombardment (FAB) technique.For some of the samples, a QSTAR XL spectrometer was employed for electrospray ionization (ESI).Optical rotations were determined on a Perkin-Elmer 241 polarimeter in 1 dm cells.Diethyl ether and THF were distilled from sodium, and dichloromethane was distilled from calcium hydride under an Ar atmosphere.

Figure 3 .
Figure 3. nOe experiments for determination of the configurations of epoxides 3 and 5.

Figure 4 .
Figure 4. Proposed mechanism for the formation of 7.

Figure 5 .
Figure 5. Proposed mechanism for the formation of 8.

Figure 6 .H
Figure 6.nOe experiments for determination of the configuration of 10 and a Chem 3D model to better understand the nOes.