(+)-Podocarpic Acid as Chiral Template in the Synthesis of Aphidicolane, Stemodane and Stemarane Diterpenoids †

In this review the synthetic work in the field of aphidicolane, stemodane and stemarane diterpenoids, in which readily available (+)-podocarpic acid (4) was used as chiral template for the construction of their polycyclic structures, is described as it developed along the years. In the frame of this work (+)-podocarpic acid (4) was a very useful tool in a model study leading to the syntheses of tetracyclic ketones 7 and 8, models of key intermediates 5a and 6 in the syntheses of (+)-aphidicolin (1) and (+)-stemodin (2a), respectively. (+)-Podocarpic acid (4) was also converted into (+)-2-deoxystemodinone (2d), allowing confirmation of the stemodane diterpenoids absolute configuration, into (+)-aphidicol-15-ene (36) and into Stemodia chilensis tetracyclic diterpenoid (+)-19-acetoxystemodan-12-ol (2f), allowing confirmation of its structure. (+)-Podocarpic acid (4) was then extensively used in the work which led to the synthesis of (+)-stemar-13-ene (57) and (+)-18-deoxystemarin (3b). Finally, (+)-4 was converted into (+)-2-deoxyoryzalexin S (66), which made it possible to demonstrate that the structure of (+)-66 could not be attributed to a Chilean Calceolaria isolated diterpenoid to which this structure had been assigned.


1982 Synthesis of 17-Noraphidicolan-16-one and 17-Norstemodan-16-one from (+)-Podocarpic Acid via (−)-9(11)-Podocarpen-12-one (1982)
Between 1976 and 1978 one of us (RMB) had been involved, as a member of the Wiesner group at the University of New Brunswick (Fredericton, NB, Canada), in the synthesis of some diterpene alkaloids whose C/D ring system, constituted by a bicyclo [ [24][25][26]. The end of this fruitful and exciting experience coincided with the disclosure of the first two syntheses of (±)-aphidicolin (1) [7,8]. Thus, back in Rome, we desired to show that the approach developed by Wiesner for the construction of the C/D ring system of diterpene alkaloids [27][28][29][30] could be also particularly convenient for the synthesis of the aphidicolane and stemodane systems. In fact the above rearrangement would have not only produced the required bicyclo[3.2.1]octane system but also introduced the oxygenated function present at C-16 in 5a and at C-13 in 6. Our initial aim was, therefore, that of showing the usefulness of this strategy by obtaining the tetracyclic ketones 7 and 8 which differ from 5a and 6 only in the A ring substitution, respectively. This choice would permit adoption of (+)-podocarpic acid (4) as starting material. The first two formal total syntheses of (±)-aphidicolin (1) were published at the beginning of 1979 [7,8]. Both syntheses had as a target compound 5a, the acetonide of the periodic acid cleavage product of (+)-1, reconverted, though not diastereoselectively into (+)-1 [2]. This synthetic problem was overcome in 1988 when a solution to it was described [9,10]. Later syntheses of (±)-1 via 5a and 5b ( Figure 2) were published also by other groups [11][12][13][14][15][16][17][18][19]. Enantioselective total syntheses of (+)-1 via 5a and 5b were also described [20][21][22]. In 1980 the first total synthesis of (±)-stemodin 2a and (±)-stemodinone 2b via the tetracyclic ketone 6 ( Figure 2) was also published [23]. The first two formal total syntheses of (±)-aphidicolin (1) were published at the beginning of 1979 [7,8]. Both syntheses had as a target compound 5a, the acetonide of the periodic acid cleavage product of (+)-1, reconverted, though not diastereoselectively into (+)-1 [2]. This synthetic problem was overcome in 1988 when a solution to it was described [9,10]. Later syntheses of (±)-1 via 5a and 5b ( Figure 2) were published also by other groups [11][12][13][14][15][16][17][18][19]. Enantioselective total syntheses of (+)-1 via 5a and 5b were also described [20][21][22]. In 1980 the first total synthesis of (±)-stemodin 2a and (±)-stemodinone 2b via the tetracyclic ketone 6 ( Figure 2) was also published [23]. Between 1976 and 1978 one of us (RMB) had been involved, as a member of the Wiesner group at the University of New Brunswick (Fredericton, NB, Canada), in the synthesis of some diterpene alkaloids whose C/D ring system, constituted by a bicyclo [ [24][25][26]. The end of this fruitful and exciting experience coincided with the disclosure of the first two syntheses of (±)-aphidicolin (1) [7,8]. Thus, back in Rome, we desired to show that the approach developed by Wiesner for the construction of the C/D ring system of diterpene alkaloids [27][28][29][30] could be also particularly convenient for the synthesis of the aphidicolane and stemodane systems. In fact the above rearrangement would have not only produced the required bicyclo[3.2.1]octane system but also introduced the oxygenated function present at C-16 in 5a and at C-13 in 6. Our initial aim was, therefore, that of showing the usefulness of this strategy by obtaining the tetracyclic ketones 7 and 8 which differ from 5a and 6 only in the A ring substitution, respectively. This choice would permit adoption of (+)-podocarpic acid (4) as starting material. Between 1976 and 1978 one of us (RMB) had been involved, as a member of the Wiesner group at the University of New Brunswick (Fredericton, NB, Canada), in the synthesis of some diterpene alkaloids whose C/D ring system, constituted by a bicyclo [ [24][25][26]. The end of this fruitful and exciting experience coincided with the disclosure of the first two syntheses of (±)-aphidicolin (1) [7,8]. Thus, back in Rome, we desired to show that the approach developed by Wiesner for the construction of the C/D ring system of diterpene alkaloids [27][28][29][30] could be also particularly convenient for the synthesis of the aphidicolane and stemodane systems. In fact the above rearrangement would have not only produced the required bicyclo[3.2.1]octane system but also introduced the oxygenated function present at C-16 in 5a and at C-13 in 6. Our initial aim was, therefore, that of showing the usefulness of this strategy by obtaining the tetracyclic ketones 7 and 8 which differ from 5a and 6 only in the A ring substitution, respectively. This choice would permit adoption of (+)-podocarpic acid (4) as starting material.

Retrosynthetic Analysis and Strategy
The approach adopted by us for obtaining 7 and 8 was based on the following retrosynthetic analysis: (a) obtaining the bicyclo [ [35][36][37]; (c) stereoselective addition of a -CH 2 CHO unit or synthetic equivalent from the α face to the C(9) of (−)-9(11)-podocarpen-12-one intermediate 10a. For the sake of precision the synthesis of (±)-1 described in ref. [12] also proceeds via a 9(11)-podocarpen-12-one and is also based on the bicyclo[2.2.2]octane → bicyclo[3.2.1]octane skeletal rearrangement; in this synthesis the bicyclo[2.2.2]octane system is obtained by Diels-Alder addition. Previously a synthesis of (±)-2c based on the same approach was also described [38].
As it can be observed (Scheme 1) the migrating bond is always the same one. Nevertheless, the different location of the leaving group causes its migration via the upper or lower face, leading to the stemodane and aphidicolane systems, respectively.

Retrosynthetic Analysis and Strategy
The approach adopted by us for obtaining 7 and 8 was based on the following retrosynthetic analysis: (a) obtaining the bicyclo [ [35][36][37]; (c) stereoselective addition of a -CH2CHO unit or synthetic equivalent from the α face to the C(9) of (−)-9(11)-podocarpen-12-one intermediate 10a. For the sake of precision the synthesis of (±)-1 described in ref. [12] also proceeds via a 9(11)-podocarpen-12-one and is also based on the bicyclo[2.2.2]octane → bicyclo[3.2.1]octane skeletal rearrangement; in this synthesis the bicyclo[2.2.2]octane system is obtained by Diels-Alder addition. Previously a synthesis of (±)-2c based on the same approach was also described [38].
As it can be observed (Scheme 1) the migrating bond is always the same one. Nevertheless, the different location of the leaving group causes its migration via the upper or lower face, leading to the stemodane and aphidicolane systems, respectively.
Besides, it was also known that the Birch reduction of 9, available in few steps from (+)-podocarpic acid (4), followed by acidic cleavage of the resulting dienol ether gives the chiron [39] (−)-9(11)podocarpen-12-one (10a) whose vinylogous H-C(8) is β-oriented as required [40] (Scheme 2). This is because the structure 10a is thermodynamically more stable than the epimeric one 10b in which the H-C(8) is α oriented [41].  The commercially available (Aldrich) chiral template [39] (+)-podocarpic acid (4) appeared to us a suitable starting material for a quick evaluation of our working hypothesis. In fact, it shows a good overlap of the skeleton and correctly set-up stereogenic centers at C(5) and C(10).
Besides, it was also known that the Birch reduction of 9, available in few steps from (+)-podocarpic acid (4), followed by acidic cleavage of the resulting dienol ether gives the chiron [39] (−)-9(11)podocarpen-12-one (10a) whose vinylogous H-C(8) is β-oriented as required [40] (Scheme 2). This is because the structure 10a is thermodynamically more stable than the epimeric one 10b in which the H-C(8) is α oriented [41]. Therefore the use of (+)-podocarpic acid (4) as starting material would have also resulted in the correct installment of the stereocenter at C(8) and generate a handle (the ring C α,β-unsaturated ketone) for the construction of the bicyclo[2.2.2]octane intermediate by the Wiesner two-carbon annulation sequence [42].

Synthesis of 17-Noraphidicolan-16-one (1982)
The starting material for the synthesis of 7 (Scheme 3) was (−)-9(11)-podocarpen-12-one (10a). Allene photoaddition to 10a in THF at −78 °C gave the photoadduct (+)-17. The stereochemistry of the addition was hypothesized on the basis of the Wiesner empirical rule [43][44][45][46] and later confirmed by the preparation of the expected final product. Nowadays the regio-and stereochemistry of the allene photoaddition can be established by 2D-NMR experiments (see Section 7). The exocyclic methylene in acetal 18, obtained from (+)-17 by standard methods, was then converted into the cyclobutanol 19 by oxidative cleavage followed by reduction. Treatment of the latter with diluted HCl unmasked the carbonyl function and effected cyclobutanol ring opening by a retro-aldol reaction (see Section 8.1). Subsequent diluted NaOH treatment in MeOH afforded 12 as a mixture of epimers at C (12). In order to obtain the aphidicolane system three synthetic operations were required: a. protection of the HO-C(12); b. reduction of the carbonyl group to the corresponding 15α-alcohol; c. introduction of a double bond at C(11) in order to make the rearrangement quantitative.
This was achieved by transforming the HO-C(12) into the corresponding dithiocarbonate 20. Subsequent reduction of the carbonyl group with NaBH4 gave stereoselectively the desired alcohol 21 owing to the bulkiness of the dithiocarbonate group. The following Chugaev reaction gave finally the unsaturated alcohol 22 which was then transformed into the methanesulfonate 13. The latter, submitted to solvolysis in glacial acetic acid, gave the rearranged acetate 14 (see Section 8.2). Saponification of the latter followed by oxidation with PCC/Al2O3 afforded the enone 24 which was eventually reduced to the target ketone 7 [47], identical with an authentic sample kindly donated by Dr. B. Hesp (ICI Americas Inc., Wilmington, DE, USA) and Dr. A.H. Ratcliffe (ICI Ltd. Macclesfield, UK).

Synthesis of 17-Norstemodan-16-one (1982)
The starting material for the obtaining of 8 was the ketol 12a (vide supra). Thus 12a was converted (Scheme 4) into the tetrahydropyranyl derivative 25 by standard methods. The latter was then reduced with NaBH4 to give the alcohol 26 which was transformed into the dithiocarbonate 27. Pyrolysis of 27 gave then the unsaturated alcohol 28 which was converted into the corresponding mesylate 15. The latter on heating in glacial AcOH underwent rearrangement to give 16 (see Section 8.2) which was converted into 8 by saponification followed by oxidation and catalytic hydrogenation [47]. Its structure was confirmed by its conversion into (+)-stemod-12-ene (31), identical with an authentic sample kindly donated by Dr. P. S. Manchand (Hoffmann-La Roche Inc., Nutley, NJ, USA).
Therefore the use of (+)-podocarpic acid (4) as starting material would have also resulted in the correct installment of the stereocenter at C(8) and generate a handle (the ring C α,β-unsaturated ketone) for the construction of the bicyclo[2.2.2]octane intermediate by the Wiesner two-carbon annulation sequence [42].

Synthesis of 17-Noraphidicolan-16-one (1982)
The starting material for the synthesis of 7 (Scheme 3) was (−)-9(11)-podocarpen-12-one (10a). Allene photoaddition to 10a in THF at −78 • C gave the photoadduct (+)-17. The stereochemistry of the addition was hypothesized on the basis of the Wiesner empirical rule [43][44][45][46] and later confirmed by the preparation of the expected final product. Nowadays the regio-and stereochemistry of the allene photoaddition can be established by 2D-NMR experiments (see Section 7). The exocyclic methylene in acetal 18, obtained from (+)-17 by standard methods, was then converted into the cyclobutanol 19 by oxidative cleavage followed by reduction. Treatment of the latter with diluted HCl unmasked the carbonyl function and effected cyclobutanol ring opening by a retro-aldol reaction (see Section 8.1). Subsequent diluted NaOH treatment in MeOH afforded 12 as a mixture of epimers at C (12). In order to obtain the aphidicolane system three synthetic operations were required: a.
protection of the HO-C(12); b.
reduction of the carbonyl group to the corresponding 15α-alcohol; c.
introduction of a double bond at C(11) in order to make the rearrangement quantitative.
This was achieved by transforming the HO-C(12) into the corresponding dithiocarbonate 20. Subsequent reduction of the carbonyl group with NaBH 4 gave stereoselectively the desired alcohol 21 owing to the bulkiness of the dithiocarbonate group. The following Chugaev reaction gave finally the unsaturated alcohol 22 which was then transformed into the methanesulfonate 13. The latter, submitted to solvolysis in glacial acetic acid, gave the rearranged acetate 14 (see Section 8.2). Saponification of the latter followed by oxidation with PCC/Al 2 O 3 afforded the enone 24 which was eventually reduced to the target ketone 7 [47], identical with an authentic sample kindly donated by Dr. B. Hesp (ICI Americas Inc., Wilmington, DE, USA) and Dr. A.H. Ratcliffe (ICI Ltd. Macclesfield, UK).

Synthesis of 17-Norstemodan-16-one (1982)
The starting material for the obtaining of 8 was the ketol 12a (vide supra). Thus 12a was converted (Scheme 4) into the tetrahydropyranyl derivative 25 by standard methods. The latter was then reduced with NaBH 4 to give the alcohol 26 which was transformed into the dithiocarbonate 27. Pyrolysis of 27 gave then the unsaturated alcohol 28 which was converted into the corresponding mesylate 15. The latter on heating in glacial AcOH underwent rearrangement to give 16 (see Section 8.2) which was converted into 8 by saponification followed by oxidation and catalytic hydrogenation [47]. Its structure was confirmed by its conversion into (+)-stemod-12-ene (31), identical with an authentic sample kindly donated by Dr. P. S. Manchand (Hoffmann-La Roche Inc., Nutley, NJ, USA).
The strategy adopted for the obtaining of tetracyclic ketones 7 and 8, models of 5a and 6 respectively, and the methodologies set-up during the model work, was so efficient that the following year formal total syntheses of (±)-1 and (±)-2a, (±)-2b and total syntheses of (±)-2c and (±)-2d were achieved by a very small group of enthusiastic young people [13,48]. Later a diastereoselective total synthesis of (±)-1 was also described [14].

Synthesis of (+)-Stemod-12-ene and (+)-2-Deoxystemodinone (1983)
The starting material was the photoadduct 17 [47] (vide supra). Methylation at C(13) of the latter gave 37 which was acetalized to 38 by standard methods (Scheme 6). The exocyclic methylene in 38 was then cleaved to the cyclobutanone 39 and the latter reduced to the cyclobutanol 40 which on refluxing in a THF/HCl mixture underwent deacetalization followed by a retro-aldol and an aldol reaction to give 33 as an epimeric mixture at C(12) (see Section 8.1). The major epimer 33a was then converted into the corresponding dithioacetal derivative 41a and the latter desulphurized with Raney-Ni in EtOH to give 42a. The alcohol 42a was then transformed into the corresponding tosylate 34 which on heating in DMSO in the presence of the methyl sulfinyl carbanion gave (+)-stemod-12-ene (31) (see Section 8.3). (+)-2-Deoxystemodinone (2d) was eventually obtained from (+)-31 by epoxidation followed by hydride epoxide reduction [50]. The use of (+)-podocarpic acid (4) as starting material also allowed confirmation of the absolute configuration attributed to the stemodane diterpenoids on the basis of ORD and CD studies.

Synthesis of (+)-Stemod-12-ene and (+)-2-Deoxystemodinone (1983)
The starting material was the photoadduct 17 [47] (vide supra). Methylation at C(13) of the latter gave 37 which was acetalized to 38 by standard methods (Scheme 6). The exocyclic methylene in 38 was then cleaved to the cyclobutanone 39 and the latter reduced to the cyclobutanol 40 which on refluxing in a THF/HCl mixture underwent deacetalization followed by a retro-aldol and an aldol reaction to give 33 as an epimeric mixture at C(12) (see Section 8.1). The major epimer 33a was then converted into the corresponding dithioacetal derivative 41a and the latter desulphurized with Raney-Ni in EtOH to give 42a. The alcohol 42a was then transformed into the corresponding tosylate 34 which on heating in DMSO in the presence of the methyl sulfinyl carbanion gave (+)-stemod-12-ene (31) (see Section 8.3). (+)-2-Deoxystemodinone (2d) was eventually obtained from (+)-31 by epoxidation followed by hydride epoxide reduction [50]. The use of (+)-podocarpic acid (4) as starting material also allowed confirmation of the absolute configuration attributed to the stemodane diterpenoids on the basis of ORD and CD studies.

Proof of the Structure of the Stemodia chilensis Tetracyclic Diterpenoid (+)-19-Acetoxy-stemodan-12-ol by Synthesis from (+)-Podocarpic Acid (2016)
The aim of this work was to confirm the structure of (+)-19-acetoxystemodan-12-ol (2f) isolated in 1991 from Stemodia chilensis [52]. The structure of this diterpenoid had been established only on the basis of 1 H-and 13 C-NMR spectroscopic data and on their similarity with those of co-occurring (+)-2-deoxystemodinone (2d). The starting material was compound (+)-48, obtained from (+)-podocarpic acid (4) as previously described (see Section 7). The strategy adopted is that described in Scheme 5. The HO-C(12) in (+)-48 is properly oriented for rearrangement to the stemodane system. Nevertheless, prior to rearrangement it was necessary to remove the C(15) carbonyl group, which would have prevented the rearrangement to the stemodane system by inhibiting the formation of a positive charge on the adjacent bridgehead carbon, and to convert the HO-C(12) into a better leaving group after selective protection of the HO-C (19).

Proof of the Structure of the Stemodia chilensis Tetracyclic Diterpenoid (+)-19-Acetoxy-stemodan-12-ol by Synthesis from (+)-Podocarpic Acid (2016)
The aim of this work was to confirm the structure of (+)-19-acetoxystemodan-12-ol (2f) isolated in 1991 from Stemodia chilensis [52]. The structure of this diterpenoid had been established only on the basis of 1 H-and 13 C-NMR spectroscopic data and on their similarity with those of co-occurring (+)-2-deoxystemodinone (2d). The starting material was compound (+)-48, obtained from (+)-podocarpic acid (4) as previously described (see Section 7). The strategy adopted is that described in Scheme 5. The HO-C(12) in (+)-48 is properly oriented for rearrangement to the stemodane system. Nevertheless, prior to rearrangement it was necessary to remove the C(15) carbonyl group, which would have prevented the rearrangement to the stemodane system by inhibiting the formation of a positive charge on the adjacent bridgehead carbon, and to convert the HO-C(12) into a better leaving group after selective protection of the HO-C (19).

Synthesis of (+)-Stemar-13-ene, (+)-18-Deoxystemarin from (+)-Podocarpic Acid via 9(11)-Podocarpen-12-one (1991-2012)
After the successful conclusion of the work on aphidicolane and stemodane diterpenoids, (±)-stemarin (3a) became the target of our studies. This tetracyclic diterpenoid had been synthesized in racemic form by Kelly and co-workers in 1980 [49]. Unfortunately this synthesis suffered a non-diastereoselective step, i.e., the completion of the synthesis with the minor epimer produced by the aldol condensation by which the bicyclo[2.2.2]octane system was formed. Developing a diastereoselective route to the key 6-hydroxy-1-methylbicyclo[2.2.2]octane intermediate appeared to us a worthwhile synthetic challenge. As it will be seen in the sequel this synthetic problem engaged us for twenty years until a simple solution was found. Being the synthetic problem confined in the construction of the C/D ring system, (+)-stemar-13-ene (57) and (+)-18-deoxystemarin (3b) the simplest compounds in the class were chosen as targets.

Retrosynthetic Analysis and Strategy
The first approach evaluated was based on the inversion of configuration at C(12) of suitable derivatives of the 13-methyl-12-endo-hydroxybicyclo[2.2.2]octan-15-one intermediate. Intermediates of type 55 were considered suitable to this task owing to the presence of the carbonyl group at C(15) which preventing the formation of a positive charge on the adjacent bridgehead carbon does not allow the rearrangement (Scheme 9).

Synthesis of (+)-Stemar-13-ene, (+)-18-Deoxystemarin from (+)-Podocarpic Acid via 9(11)-Podocarpen-12-one (1991-2012)
After the successful conclusion of the work on aphidicolane and stemodane diterpenoids, (±)-stemarin (3a) became the target of our studies. This tetracyclic diterpenoid had been synthesized in racemic form by Kelly and co-workers in 1980 [49]. Unfortunately this synthesis suffered a non-diastereoselective step, i.e., the completion of the synthesis with the minor epimer produced by the aldol condensation by which the bicyclo[2.2.2]octane system was formed. Developing a diastereoselective route to the key 6-hydroxy-1-methylbicyclo[2.2.2]octane intermediate appeared to us a worthwhile synthetic challenge. As it will be seen in the sequel this synthetic problem engaged us for twenty years until a simple solution was found. Being the synthetic problem confined in the construction of the C/D ring system, (+)-stemar-13-ene (57) and (+)-18-deoxystemarin (3b) the simplest compounds in the class were chosen as targets.

Diastereoselective Synthesis of (+)-Stemar-13-ene and (+)-18-Deoxystemarin by Inversion of Configuration at C(12) of Suitable Derivatives of the 13-Methyl-12-endo-hydroxybicyclo[2.2.2]octan-15-one Intermediate (1991)
The starting material was known 33a (vide supra) which was converted [54] into the corresponding tosylate 55 (Scheme 10). The latter was then treated with Et4N(PhCOO) in acetone at reflux affording the exo-benzoate 56. Later the same conversion was performed in benzene resulting in drastically reduced reaction times [55]. This methodology, described in the past for the inversion of configuration of acyclic secondary alcohols [56], was particularly suitable to our case in that it produces a locked exo-ketol which cannot therefore re-equilibrate to the more stable endo epimer. Deoxygenation of 56 via 58 by standard methods gave the benzoate 59. Debenzoylation of the latter produced 42b whose HO-C(12) is properly oriented for the rearrangement to the stemarane system. Thus refluxing the latter in benzene in the presence of TsOH gave (+)-stemar-13-ene (57) (see Section 8.5). Finally, epoxidation of the latter followed by LiAlH4 epoxide ring opening gave (+)-3b [54]. This synthesis was accomplished in 1991. In 2001 (+)-stemar-13-ene (57) was isolated from the fungus Phoma betae [57]. The two materials were identical in all respects. Deoxygenation of 56 via 58 by standard methods gave the benzoate 59. Debenzoylation of the latter produced 42b whose HO-C(12) is properly oriented for the rearrangement to the stemarane system. Thus refluxing the latter in benzene in the presence of TsOH gave (+)-stemar-13-ene (57) (see Section 8.5). Finally, epoxidation of the latter followed by LiAlH 4 epoxide ring opening gave (+)-3b [54]. This synthesis was accomplished in 1991. In 2001 (+)-stemar-13-ene (57) was isolated from the fungus Phoma betae [57]. The two materials were identical in all respects.
Later, in view of a methodology [58] by which esters are cleaved in the presence CH 3 ONa/La(OTf) 3 at neutral pH, thus ensuring no epimerization at HO-C(12), the desulphurization/debenzoylation steps could be reversed [59]. Later, in view of a methodology [58] by which esters are cleaved in the presence CH3ONa/La(OTf)3 at neutral pH, thus ensuring no epimerization at HO-C(12), the desulphurization/debenzoylation steps could be reversed [59].

Retrosynthetic Analysis and Strategy
Finally, a new, general and simple solution for the construction of the stemarane diterpene C/D ring system was found. This solution is based on the existing endo/exo equilibrium under acidic conditions between 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones (Scheme 12a) and on the acid catalyzed rearrangement of 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones to 4-methylbicyclo[3.2.1]oct-3-en-6-ones (Scheme 12b) described in 2005 within the framework of a study on the reactivity of isotwistanes [61].  Later, in view of a methodology [58] by which esters are cleaved in the presence CH3ONa/La(OTf)3 at neutral pH, thus ensuring no epimerization at HO-C(12), the desulphurization/debenzoylation steps could be reversed [59].  The equilibrium endo/exo ratio resulted 4:6. Hydroxydithioacetals 41a and 41b could be separated by HPLC. By repeating three times the equilibration/separation cycle, practically all the material (94%) could be converted into the desired exo epimer 41b [60].

Retrosynthetic Analysis and Strategy
Finally, a new, general and simple solution for the construction of the stemarane diterpene C/D ring system was found. This solution is based on the existing endo/exo equilibrium under acidic conditions between 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones (Scheme 12a) and on the acid catalyzed rearrangement of 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones to 4-methylbicyclo[3.2.1]oct-3-en-6-ones (Scheme 12b) described in 2005 within the framework of a study on the reactivity of isotwistanes [61]. The equilibrium endo/exo ratio resulted 4:6. Hydroxydithioacetals 41a and 41b could be separated by HPLC. By repeating three times the equilibration/separation cycle, practically all the material (94%) could be converted into the desired exo epimer 41b [60].

Retrosynthetic Analysis and Strategy
Finally, a new, general and simple solution for the construction of the stemarane diterpene C/D ring system was found. This solution is based on the existing endo/exo equilibrium under acidic conditions between 1-methyl- 6 The following retrosynthetic analysis therefore resulted (Scheme 13).

Synthesis of (+)-Stemar-13-ene
Thus the 85:15 endo/exo mixture 33, was dissolved in toluene and heated at 85 °C in the presence of TsOH (Scheme 14). After 24 h the rearrangement of 33 to (+)-63 was completed (see Section 8.6). Thioacetalization of (+)-63 by standard procedures afforded then (+)-64 which was desulphurized to (+)-57 [55,62]. The remarkable feature of this approach is that, owing to the stereospecificity of the rearrangement and to the 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones endo/exo equilibrium, the whole endo/exo mixture 33 is converted into the rearrangement product in which the C(13)-C(14) double bond is also present, a characteristic feature of some stemarane diterpenoids such as (+)-oryzalexin S (65) (Figure 4) [63,64] (induced in rice leaves in response to the invasion of the fungus Pyricularia oryzae or when exposed to UV radiation or heavy metals [65]) and a necessary tool for the introduction of the α-configured HO-C(13) if necessary. The following retrosynthetic analysis therefore resulted (Scheme 13). The following retrosynthetic analysis therefore resulted (Scheme 13).

Synthesis of (+)-Stemar-13-ene
Thus the 85:15 endo/exo mixture 33, was dissolved in toluene and heated at 85 °C in the presence of TsOH (Scheme 14). After 24 h the rearrangement of 33 to (+)-63 was completed (see Section 8.6). Thioacetalization of (+)-63 by standard procedures afforded then (+)-64 which was desulphurized to (+)-57 [55,62]. The remarkable feature of this approach is that, owing to the stereospecificity of the rearrangement and to the 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones endo/exo equilibrium, the whole endo/exo mixture 33 is converted into the rearrangement product in which the C(13)-C(14) double bond is also present, a characteristic feature of some stemarane diterpenoids such as (+)-oryzalexin S (65) (Figure 4) [63,64] (induced in rice leaves in response to the invasion of the fungus Pyricularia oryzae or when exposed to UV radiation or heavy metals [65]) and a necessary tool for the introduction of the α-configured HO-C(13) if necessary.

Synthesis of (+)-Stemar-13-ene
Thus the 85:15 endo/exo mixture 33, was dissolved in toluene and heated at 85 • C in the presence of TsOH (Scheme 14). After 24 h the rearrangement of 33 to (+)-63 was completed (see Section 8.6). Thioacetalization of (+)-63 by standard procedures afforded then (+)-64 which was desulphurized to (+)-57 [55,62]. The remarkable feature of this approach is that, owing to the stereospecificity of the rearrangement and to the 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones endo/exo equilibrium, the whole endo/exo mixture 33 is converted into the rearrangement product in which the C(13)-C(14) double bond is also present, a characteristic feature of some stemarane diterpenoids such as (+)-oryzalexin S (65) (Figure 4) [63,64] (induced in rice leaves in response to the invasion of the fungus Pyricularia oryzae or when exposed to UV radiation or heavy metals [65]) and a necessary tool for the introduction of the α-configured HO-C(13) if necessary. The remarkable feature of this approach is that, owing to the stereospecificity of the rearrangement and to the 1-methyl-6-hydroxybicyclo[2.2.2]octan-2-ones endo/exo equilibrium, the whole endo/exo mixture 33 is converted into the rearrangement product in which the C(13)-C (14) double bond is also present, a characteristic feature of some stemarane diterpenoids such as (+)-oryzalexin S (65) (Figure 4) [63,64] (induced in rice leaves in response to the invasion of the fungus Pyricularia oryzae or when exposed to UV radiation or heavy metals [65]) and a necessary tool for the introduction of the α-configured HO-C(13) if necessary.  The presence of a double bond endo to the leaving group ensures a quantitative rearrangement by stabilizing the positive charge on the former bridgehead carbon (Scheme 17).

Skeletal Rearrangements of 33 and 48 into 63 and 77, Respectively
The presence of a methyl group on the bridgehead carbon ensures a quantitative rearrangement by stabilizing the positive charge on the former bridgehead carbon (Scheme 21). The alternative rearrangement cannot occur owing to the presence of the carbonyl group at C(12) which prevents the formation of a positive charge on the adjacent bridgehead carbon (Scheme 22).