Regioselective oxidative fragmentation of drimanic terpene alcohols: a short, easy and efficient access to natural and synthetic 11-nordrimane terpene derivatives

Using selective Cr(VI) oxidations with Jones reagent and PCC, a regioselective and direct access to different oxygenated 11-nordrimane skeletons was achieved, in good overall yield, starting from naturally occurring drimane terpenes.


Introduction
2][3][4][5][6] For that reason, many efforts have been devoted to the syntheses of drimanes, usually starting from abundant, naturally occurring compounds. 7n marked contrast, there is hardly any natural product with an 11-nordrimane skeleton known to date, being most of them those shown in Figure 1: isonordrimenone (1), obtained in 1973 from several species of tobacco, as for instance, Greek tobacco, Nicotiana tabacum L., 8 isopolygonal (2) and polygonone (3), obtained in 1985 from the leaves of the folk medicinal plant Polygonum hydropiper L., 9 and polygonal (4), also isolated from Polygonum hydropiper L. in 1979. 10Very recently, in 2000, aldehyde 5 was isolated from the stem bark of the conifer tree Thuja standishii (Gord.)Carr. 11nterestingly, transposed 9-oxo compounds like 6 and 7 were never obtained from a natural source.Nevertheless, both carbonyl compounds found use as synthetic key intermediates in the construction of more complex substances: α,β-unsaturated ketone 6 was recently used as a key intermediate in the synthesis of hispanone, 12 and cross-conjugated keto-aldehyde 7 (together with its synthetic equivalents) was used in the synthesis of the feeding-deterrent natural compound warburganal. 13,147][18][19][20] Both 8 and 10 were easily obtained from the stem bark of South American Winter's bark tree. 14,15

Results and Discussion
Compounds 8 and 10 were obtained by multigram preparative column chromatography (hexane: ethyl acetate gradient) from the crude hexane extract of the ground stem bark of Drimys winteri F., as previously described.We found that when an acetone solution of alcohol 8 was titrated with Jones reagent at 0°C, unsaturated ketone 6 was obtained as a single compound (71% yield, Scheme 1).On the other hand, when alcohol 8 was oxidized instead with pyridinium chlorochromate (PCC) suspended in dichloromethane at room temperature, unsaturated ketone 1 (65% isolated yield) was obtained as a single compound (Scheme 1).The identities of enones 1 8,21,22 and 6 12,21,23 were established by comparison with the previously reported physical data for the same compounds.
In order to have access to compounds 3 and 7, it was possible to perform an allylic oxidation of 1 and 6, but we felt that a better and more convenient way would be to start with natural dialdehyde 10 (which is available to us from the natural source in better yield than alcohol 8 itself).Reduction of 10 with sodium borohydride in methanol at room temperature produced diol 9 in an excellent yield (Scheme 2). 24,25It was previously reported that treatment of diol 9 with one equivalent of tert-butyl dimethyl chlorosilane monoprotected specifically the C-12 hydroxy position. 24In our case, protection of diol 9 with one equivalent of tert-butyl, diphenylsilyl chloride in DMF, in the presence of imidazole, produced silyl ether 11, with identical specificity towards the allylic alcohol, and excellent yield.

Conclusions
In summary, we described here a simple methodology for the degradation of drimane alcohols to the corresponding 11-nordrimane enones or ketoaldehydes, with total control of the regiospecificity of the reaction.Specifically, we prepared natural isonordrimenone (1) with a 71% overall yield and polygonone (3), with a 37% overall yield, and their synthetic regioisomers 6, obtained with a 65% overall yield, and 7, prepared with a 27% overall yield (all yields calculated from the natural drimanic starting material).Compounds 1 and 6 were previously isolated from natural sources, and compounds 3 and 7 were previously used as key intermediates in the total synthesis of complex natural compounds.Due to the easy availability of our starting materials from the chiral pool, and the good overall yields obtained in these short preparations, we believe that these compounds could find in future a good use in the stereoselective synthesis of complex bioactive molecules, even at the industrial scale.A study of their synthetic applications, is currently under way in our labs.

Experimental Section
General Procedures.All reactions were routinely run under a dry nitrogen atmosphere, with flame-dried glassware, and with magnetic stirring.All chemicals were used as purchased.
Melting points were determined in a Stuart Scientific Apparatus SMP3, and are uncorrected.Optical rotations were measured in CHCl 3 solutions, in a 0.1 dm cell, in an Optical Activity, Ltd instrument.Infrared spectra were recorded in a Bruker Vector-22 FTIR spectrometer.NMR Spectra were obtained on a Bruker AC 200P (200.13MHz for 1 H, 50.13 MHz for 13 C) or Avance 400 (400.13MHz for 1 H, 100.13 MHz for 13 C) spectrometer, in CDCl 3 solutions with TMS as internal standard.For 2D, COSY, NOE, HSQC and HMBC experiments, Bruker standard software was used.Assignments were done by a combination of 1D and 2D NMR techniques, in each case, as needed.The symbol * was used to denote signal pairs with interchangeable assignments.
For skeleton carbon numberings, please refer to Figure 1.Elemental analyses were obtained in a Fisons Instruments EA 1108 microanalyzer.Column chromatography was performed on silica gel 60 H, slurry packed, run under low pressure of air, and employing increasing amounts of ethyl acetate in hexane as solvent.Analytical TLC was carried out using Kieselgel Merck F 254 with thickness 0.20 mm.