Synthetic Approaches and Total Syntheses of Vinigrol, a Unique Diterpenoid

Abstract

This review summarizes all published total and formal syntheses as well as synthetic approaches towards vinigrol. The content is divided into sections, which are focused on each research groups contributions and how far each approach was advanced towards vinigrol. Graphical summaries of all the published vinigrol structural perspectives, starting materials used for each routes and a discussion of preferred or privileged reactions employed is also presented.

Graphical Abstract

Introduction

The aim of this review is to present a detailed account of all published total syntheses and synthetic approaches toward the unique diterpenoid natural product vinigrol. We have chosen to summarize the efforts of each paper by graphically detailing the synthesis, and associated reaction details, of only the most advanced intermediate each research team made. This review¹ omits contributions only reported in graduate student theses. The remarkable structure of vinigrol, which is most simply described as a densely decorated cis-decalin core with a diaxial four carbon bridge connecting the two cyclohexyl rings, was first reported in 1987.² Vinigrol contains seven challenging stereocenters and is decorated with only three heteroatoms in the form of hydroxyl groups (primary, secondary and tertiary) all of which are located on one of the two cyclohexyl rings. Vinigrol has been reported to exhibit a promising biological profile, most notably in the areas of cancer and HIV.³ Shown in Scheme 1 is a compilation of all the vinigrol structural perspectives reported to date by the research teams whose efforts are summarized in the following sections. Each one of these perspectives will appear in later schemes as we utilize each authors chosen vinigrol perspective to highlight (with color) how far their synthetic efforts advanced.

Scheme 1.

Reported Vinigrol Structural Perspectives.

Summerized in Scheme 2 is a structural taste of the synthetic journeys you the reader are about to embark on. Shown are the structures of all the starting materials used by these research teams in their investigations towards vinigrol. It is evident that there are preferred structural patterns as all these approaches can represented by only five structural types: a) cyclohexanones (Mehta, Baran, Matsuda, Crowe, Hanna and Barriault), b) phenols (Njardarson and Liao), c) cyclic 1,3-dienes (Sun and Barriault), d) acyclic diols (Fallis and Barriault) and e) Evans auxiliary (Paquette and Kaliappan). What does this mean? It is not an easy question to answer, but as we summarize the key reaction types used in each one of these routes later in the review it becomes clear that certain key reactions are preferred as well and they might be one of the key reasons for the starting material choices. There are of course much simpler explanations such as the fact that vinigrol has two fused cyclohexyl rings, which quite reasonably calls for a six membered ring starting materials, which is indeed the preferred choice.

Scheme 2.

Starting Materials Used for Published Vinigrol Approaches and Total Syntheses.

Although the contents of this review can be organized in several ways we decided to focus each section of the review on authors and to present within these sections all of the author’s contributions to date. Furthermore, we selected to arrange these sections in order of the timeline of the first disclosure by each author. We have tried our best to faithfully recreate each author’s chosen structural perspective. Key transformative reactions are highlighted with instructive names in orange boxes. The final structure of each sequence is then compared with vinigrol and atom, bond forming and stereochemical accomplishments highlighted in blue.

1. Hanna’s Approaches to Vinigrol

The first vinigrol approach (1993)⁴ was reported by Professor Hanna and coworkers (Scheme 3). A Diels-Alder union between the silyl enol ether (2) of cyclohexenone (1) and para-quinone (3) affords a cyclohexyl fused [2.2.2] adduct 4. Following a sequence of Luche reduction, protection, deprotection and elimination the stage is set for selective reduction of a 1,3-cyclohexadiene moiety (7). This reduction challenge was accomplished with hydrogen in the presence of Wilkinson’s catalyst. Deprotection of the MOM-protecting group is followed by selective addition of a vinyl magnesium chloride reagent to the less hindered face of the ketone sets the stage for the key step in Hanna’s proposed route, a Cope-rearrangement to form vinigrol’s carbocyclic core. Hanna and his team demonstrated that by employing oxy-anionic Cope rearrangement condtions in refluxing tetrahydrofuran in the presence of the appropriate crown ether, they could indeed access the vinigrol core (11) in excellent yield from 10. Thus, in only ten steps from 1, vinigrol core (11) is rapidly accessed.

Scheme 3.

I. Hanna’s 1993 Approach Toward the Vinigrol Core.

Few years later, Hanna reported a short report which highlighted some of the challenges they would be facing in the coming years in trying to realize the success of their model system towards the actual target (vinigrol).⁵ Protection of 11 (Scheme 4), reduction of the ketone and protection as a methyl ether (12) set the stage for evaluating double hydration approaches. Interestingly, they learned that structures like 12 underwent a very facile acid catalyzed cascade to form fused tricyclic products like 13 instead of any desired hydrated products. Undeterred, Hanna’s team marched forward towards a synthetic solution.

Scheme 4.

I. Hanna’s 1996 Vinigrol Core Rearrangement Observations.

Only a year later, an olefin hydration solution was identified by Hanna’s team (Scheme 5).⁶ They demonstrated that alcohol 14 could be hydrated in such a way to afford triol 15 in modest yield. This was a significant achievement as it completed the critical cis-decalin ring fusion of vinigrol. With enough quantities of 15 in hand they turned their attention toward installing some additional functional groups. The unprotected secondary alcohol was removed using a xanthate radical reduction approach (17) and the remaining secondary alcohol was oxidized using Dess-Martin periodinane to ketone 19. Installation of the methylene hydroxyl group (20) of vinigrol was accomplished in moderate yield by first forming a silylenol ether and then trapping it with formaldehyde with the aid of a lanthanide catalysts. With the decalin challenges mostly solved, Professor Hanna turned his attention to the four carbon bridge and its two chiral centers.

Scheme 5.

I. Hanna’s 1997 Decalin Functionalization Explorations.

Hanna next turned his attention to employing a more complex nucleophile in the addition step to the ketone to enable introduction of the two four carbon bridge substituents (Scheme 6).⁷ Towards that end, by starting with a slightly more complex cyclohexadiene (22) in the Diels-Alder reaction he was able to bring in the methyl group needed in the bridge. Following Luche reduction, a Mitsunobo inversion reaction was needed to set the correct vinigrol secondary alcohol stereocenter (25). With the secondary alcohol stereochemistry set a new approach was needed to install the methylene alcohol functionality. A creative solution was identified in the form of a silyl tethered radical cyclization (27) followed by peroxide mediated oxidation of the C-Si bond (28). Following standard protection and dehydration steps the stage was set for introduction of a more complex Grignard nucleophile (31) to afford propargylic alcohol 32. Syn reduction of the alkyne group with Rieke zinc afforded Z-olefin 33, which was essential for setting the appropriate stereochemistry of the isopropyl bridge stereocenter in the key oxy-anionic Cope rearrangement step (34). Unfortunately the methyl group stereocenter on the bridge was incorrect, which is why a three step solution involving reduction, dehydration (35 to 36) and substrate controlled hydrogenation (37 to 38) was needed. Advanced intermediate 38 has all the necessary vinigrol carbon atoms but it lacks a double bond and the ability to selectively reduce one of the oxirane C-O bonds to reach vinigrol.

Scheme 6.

I. Hanna’s 2003 Approach Toward Vinigrol.

In Hanna’s final attempt to reach vinigrol he needed to find a way to hydrate the tetrasubstituted olefin achieving a net anti addition (Scheme 7).⁸ Dihydroxylation with osmium tetraoxide achieved the first part of his task, the introduction of a hydroxyl group (39) at the decalin ring fusion. Before turning his attention toward how to reduce the other C-O bond, he first needed to flip the methyl stereochemistry using the substrate controlled olefin reduction logic described in Scheme 6 (41). This was accomplished by first protecting the 1,2-diol as a carbonate, followed by hydrogenation of the trisubstituted olefin and reprotection of the 1,3-diol group. Following carbonate group removal, diol 43 was functionalized as the mono-mesylate with the aim of eliminating it to form exocyclic olefin 44 exclusively. This non-trivial activation-elimination task was only achieved in dismal 11% yield, which is perhaps not surprising as both alcohols can react, or be mesylated and then eliminated in a variety of ways. All hopes were squashed when they realized that reduction of this exocyclic olefin only afforded the incorrect stereoisomer (45). Their most advanced intermediate (46) is only one double bond away from vinigrol in addition to being epimeric at one of the stereocenters.

Scheme 7.

I. Hanna’s 2009 Approach Toward Vinigrol – Almost there!

2. Mehta’s Vinigrol Approach

Professor Mehta⁹ reported a model study toward vinigrol soon after that of Professor Hanna’s (Scheme 8). The study commences with a Robinson annulation to build 50, which was then bis alkylated (50) and subjected to an addition of a vinyl magnesium bromide to afford allylic alcohol 52. As vinigrol contains a [5.3.1] bridged bicyclic core, similar to that of Taxol, within its structure, Professor Mehta postulated that an oxy-Cope rearrangement could be used to convert 52 into 53 and rapidly reavel this bicyclic core. Interestingly, oxy-anion accelerated Cope rearrangement conditions failed for 52 and the best results were achived by heating 52 in the absence of solvent to very high temperatures. These harsh thermal oxy-Cope rearrangement conditions afforded the expected product 53 as well as substantial amount of 54, which is the product of an alternate intramolecular ene-reaction.

Scheme 8.

G. Mehta’s 1996 oxy-Cope Rearrangement Approach Toward Vinigrol.

3. Matsuda’s Vinigrol Approaches

Professor Matsuda’s group was next in reporting their synthetic strategy towards vinigrol (Scheme 9).¹⁰ His approach commenced with an aldol union between dihydrocarvone (55) and aldehyde 56. The resulting aldol product was then dehydrated (58) and the enone subjected to a Luche reduction. Vanadium directed epoxidation of allylic alcohol 59 afforded oxirane 60. Following a Swern oxidation, samarium diiodide was employed as a reducing agent to break one of the oxirane C-O bonds via ketyl radical intermediate and protonation of the resulting enolate (62). Acylation of the secondary alcohol then set the stage for establishing the feasibility of the proposed key 6-exo-trig ketyl radical cyclization. Samarium diiodide turned out to to be the optimal reagent for forming cis-decalin 64, which contains all but one of the necessary cis-decalin substituents found in vinigrol.

Scheme 9.

F. Matsuda’s 1996 Ketyl Radical Cyclization Approach Toward Vinigrol.

In his second approach (Scheme 10). Matsuda focused his attention on accessing the [5.3.1] bicyclic core Professor Mehta had shown could be build using an oxy-Cope rearrangement strategy.¹¹ Again, starting from dihydrocarvone an aldol reaction was followed by a dehydration step to form 68. This time around, an allylmagnesium Grignard reagent was added to the ketone and the resulting allylic alcohol (69) protected as a MOM-ether (70). Hydroboration of the mono-substituted olefin and oxidation of the primary alcohol product with Dess-Martin periodinane afforded aldehyde 72. In this proposal, Professor Matsuda hoped to bring together the aldehyde and an allyl chloride in a samarium mediated reductive coupling reaction to form [5.3.1] bridged bicyclic core 73. This was indeed shown to be feasible using samarium(II) diiodide in the presence of HMPA as co-solvent.

Scheme 10.

F. Matsuda’s 1997 Approach Toward Vinigrol.

4. Paquette’s Vinigrol Approaches

With only several approaches towards vinigrol synthesis reported in the 1990’s, contributions from numerous laboratories starting emerging in the 21^st century. The first such contribution came in 2003 from the laboratory of Professor Paquette.¹² All his routes (Schemes 11–13) focus on building a fully functionalized cis-decalin precursor and then evaluate cyclization approaches to form the four carbon bridge. Synthesis of the common cis-decalin precursor and the first cyclization substrate are shown in Scheme 11. Evan’s asymmetric alkylation was used to set the first chiral center (the isopropyl stereocenter), and after standard redox manipulations, aldehyde 77 was then subjected to a Robinson annulation sequence to afford enone 78 (DB-18-Cr-6 = Dibenzo-18-Crown-6). An anionic double Michael addition cascade between enone 79 and a vinyl sulfoxide provided access to [2.2.2] bicyclic product 80 in a suboptimal 27% yield. Grignard reagent derived from vinyl iodide 81 was used to from allylic alcohol 82, which then underwent the key oxy-anionic Cope rearrangement to yield densely substituted cis-decalin product 83. Protection of the newly formed ketone as a cyclic ketal allowed selective manipulations of the two protected primary alcohol groups to take place to eventually afford pre-cyclization precursor 86. Unfortunately, all attempts to displace the sulfonate leaving group intramolecularly with the anion derived from the sulfone failed.

Scheme 11.

L. A. Paquette’s 2003 Vinigrol Approach.

In their second cyclization attempt, their goal was to use a carbanion or a ketyl radical to attack a more proximal and rigid strained lactone electrophile (Scheme 12).¹³ Their routes commenced with intermediate 83, which was advanced through a series of standard redox and protection-deprotectoin steps to lactones 89 and 92. The low yielding lactone steps are a testament to how difficult these bridge bicyclic lactones are to make. Sadly again, no productive cyclization was observed from either substrate.

Scheme 12.

L. A. Paquette’s 2005 Vinigrol Eight Membered Ring Cyclization Studies I.

Still undeterred, the Paquette group decided to use their advanced intermediate to evaluate a host of other cyclization approaches, including a ring closing metathesis, Ramberg-Bäcklund and Lactam-Sulfoxide ring contraction reactions (Scheme 13).¹⁴ All cyclization substrates were made using standard manipulations from common intermediate 83. Sadly, ring closing metathesis substrates 93–95 failed, as did lactam-sulfoxide ring contraction substrates 96–98, Ramberg-Bäcklund substrate 99 and reductive cyclization substrate 100. With these cis-decalin cyclization failures, no more advances towards vinigrol were made.

Scheme 13.

L. A. Paquette’s 2005 Vinigrol Eight Membered Ring Cyclization Studies II.

5. Barriaults’ Approaches and Total Synthesis of Vinigrol

Professor Barriault has pursued several creative approaches towards vinigrol, which finally culminated in a total synthesis in 2012. His first approach, reported in 2004, was focused on employing an ingenious pericyclic cascade (Scheme 14).¹⁵ Cyclohexadiene 101 served as his starting material for his model studies. Following standard oxidative manipulations, oxirane 103 was desymmetrized using isopropenyl magnesium bromide to afford cyclohexyl triol 104. Protection of the vicinal alcohols and oxidation of the remaining secondary alcohol furnished ketone 105, which was then converted into propargylic alcohol 106. Allylation of the tertiary alcohol provided requisite substrate (107) for the pericyclic reaction studies. Impressively, when ene-yne 107 was heated to high temperatures in toluene it underwent the proposed oxy-Cope/Claisen/Ene-rearrangement cascade to afford cis-decalin 108 in excellent yields. Unfortunately, this remarkable pericyclic cascade did not proceed when the olefin terminus was substituted, which ended efforts to further advance this route.

Scheme 14.

L. Barriault’s 2004 Approach Toward Vinigrol.

Professor Barriault’s second vinigrol attempt also aimed at employing a key pericyclic reaction (Claisen rearrangement) to form the four carbon decalin-bridge (Scheme 15).¹⁶ The substrate (116) needed for this key step was assembled from cyclohexenone 1. After vinylation of 1, enone 109 was reduced and the resulting diene subjected to a magnesium-directed Diels-Alder cycloaddition reaction to afford 110. Standard redox and protection/deprotection manipulations then advanced 110 to alcohol 113. Oxidation of the free alcohol, addition of a vinyl cerium nucleophile and deprotection furnished diol 114. Treatment of 114 with Ley oxidation conditions yielded lactone 115, which was readily converted into Claisen rearrangement substrate 116 using Tebbe’s olefination. Regrettably, enol ether 116 did not undergo the expected Claisen rearrangement.

Scheme 15.

L. Barriault’s 2005 Claisen Rearrangement Approach Toward Vinigrol.

Barriault and co-workers used the same decalin intermediates to also evaluate additional cyclization approaches. Shown in Scheme 16 is their approach towards ring closing metathesis substrate 122, whose precursor was also a candidate for a McMurry cyclization. Grignard addition of 118 to ketone 117 was followed by protection and elimination to afford trisubstituted olefin 120. High pressure ruthenium catalyzed reduction of olefin 120 afforded 121. Removal of both silyl groups with tetrabutylammonium fluoride followed by Ley oxidation furnished a bis-aldehyde, which did not cyclize when subjected to the requisite titanium reagents and conditions most commonly employed in McMurry type cyclizations. Olefination of the bis-aldehyde was easily accomplished under standard Wittig olefination conditions. Sadly, the resulting diene (122) did not undergo the expected ring closing metathesis reaction to form the vinigrol carbocyclic core.

Scheme 16.

L. Barriault’s 2005 Ring Closing Metathesis Approach Toward Vinigrol.

The Barriault group evaluated an alternative Claisen rearrangement strategy as well to form the vingirol core (Scheme 17). Towards that end, 1,4-butanediol was converted into enal 125 and then subjected to Weiler addition reaction conditions, which yielded alcohol 127. Directed reduction was then followed by acid mediated lactonization conditions (128) and the ensuing lactone carbonyl group was next converted into a diene (129) via a four step protocol. The diene was then reacted with benzyl protected maleimide in a Lewis acid directed Diels-Alder cycloaddition reaction to afford decalin 130. Following silyl protection of the alcohol, Barriault and his team were rewarded with a successful Claisen rearrangement under microwave heating conditions to afford the much desired vinigrol carbocyclic core 131.

Scheme 17.

L. Barriault’s 2005 Second Claisen Rearrangement Approach Toward Vinigrol.

Despite the success of the Claisen rearrangement in their 2005 approach, the Barriault group decided to go back to the drawing board and pursue a new synthetic approach (Scheme 18).¹⁷ Starting from 1,4-butanediol they rapidly form vinyl iodide 133, which is subjected to copper catalyzed etherification with allylic alcohol 134 to afford Claisen rearrangement precursor 135. Claisen rearrangement proceeded smoothly in the presence of a Lewis acid, which also served the critical role of rapidly reducing the sensitive aldehyde product. Standard redox and homologation procedures then advanced 136 to aldehyde 138, which was subjected to the Bestman-Ohira alkynlation conditions to afford alkyne 140. Deprotection of the primary alcohol was then followed by Ley oxidation and a Wittig olefination to yield 142, which underwent a high yielding intramolecular ene-yne metathesis cyclization reaction in the presence of Grubb’s second generation ruthenium catalyst. The nitrile of the ene-yne product (143) was converted into a vinyl ketone (144) using standard procedures. This triene underwent a remarkably mild and near quantitative intramolecular Diels-Alder cycloaddition reaction to form the vinigrol carbocyclic core (145). Following this excellent success, the Barriault group turned their attention toward completing the total synthesis.

Scheme 18.

L. Barriault’s 2007 Intramolecular Diels-Alder Approach Toward Vinigrol.

With an exciting synthetic blueprint established, Professor Barriault and his team set out to complete the total synthesis of vinigrol (Scheme 19).¹⁸ Following basic diol (123) manipulations, allylic alcohol 146 was reacted with cyclohexyl ketal 147 under Johnson-Claisen rearrangement conditions to afford ketone 148. Reduction of the propene moiety was followed by triflate formation (149) and a Stille coupling of the resulting enol triflate with vinyl tributyltin (150). The undesired of the two pivaloate isomers was subjected to a four step sequence to funnel all of the material to pivaloate 151. After silyl deprotection the primary alcohol was converted into enone 152, which was subjected to the facile intramolecular Diels-Alder cycloaddtion to deliver the vinigrol carbocyclic core (153). Wittig olefination and platinum catalyzed hydrogenation then ensured installation of a key stereocenter (154). At this point, the alcohol protecting group was swaped and the trisubstituted olefin converted in a five step sequence to a methyl group and a tertriary alcohol. This is a slight variation of a cycloaddition solution designed by Professor Baran in his vinigrol synthesis (Scheme 22). Following reductive removal of the benzyl group and oxidation with TEMPO, ketone 157 was obtained. The enolate of ketone 157 was treated with an electrophilic oxygen source in the form of Davis oxaziridine to install the final vinigrol stereocenter (158). Shapiro reaction conditions, as described by Baran,²¹ where then used to convert 158 into vinigrol.

Scheme 19.

L. Barriault’s 2012 Formal Total Synthesis of Vinigrol.

Scheme 22.

P. S. Baran’s 2009 Vinigrol Total Synthesis.

6. Fallis’s Vinigrol Approach

Professor Fallis from the University of Ottawa reported a Diels-Alder focused approach towards vinigrol (Scheme 20).¹⁹ His route commenced with conversion of 1,3-propanediol to aldehyde 160, which was treated with diene bromide 161 in the presence of indium metal. The resulting alcohol (162) was then dehydrated in two steps to yield triene alcohol 164, which underwent a magnesium directed Diels-Alder cycloaddition reaction with N-methyl maleimide. Cycloaddition product 165 was then oxidized (166) and homologated to alkyne 167. Deprotonation of the alkyne and trapping with dimethylformamide (DMF) afforded ynal 168, which following a Lindlar reduction (producing 169) was treated with vinyl magnesium bromide (170). Upon mild manganese dioxide oxidation of divinyl alcohol 170, the resulting ketone (171) underwent an efficient intramolecular Diels-Alder reaction with the aid of gentle heating to form cycloadduct 172, which contains the carbocyclic vinigrol core.

Scheme 20.

A. G. Fallis’s 2007 Vinigrol Approach.

7. Baran’s Approach and Total Synthesis of Vinigrol

Professor Phil Baran model approach to the vinigrol carbocyclic core is summarized in Scheme 21.²⁰ Bis-silylated diene 176, obtained from 1,3-cyclohexadione (173), was reacted with enoate 175 in a Diels-Alder reaction. Cycloadduct 176, was then homologated via vinyl triflate 177 to 1,3-diene 178. Reduction of the ester afforded a primary alcohol (179), which was oxidized with Dess-Martin periodinane to aldehyde 180. Remarkably, addition of allyl magnesium bromide in toluene afforded intramolecular Diels-Alder cycloadduct 181 in one pot and excellent yield. In collaboration with Professor Houk further experimental and computational follow up studies have been performed in an attempt to explain the remarkably facile nature of this Diels-Alder reaction.²¹ Deprotection of the silyl group was followed by oxidation and reduction of the resulting ketone to set up the necessary orbital alignment needed for the key ring expansion reaction. Mono-mesylation of diol 184 afforded 185, which gratifyingly underwent a mild and high yielding Grob fragmentation reaction to reveal the vinigrol carbocyclic core (186). Substrate controlled epoxidation of the trisubstituted olefin demonstrated that the tertiary C-O bond could be installed, which is where model efforts ended and total synthesis plans began (Scheme 22).

Scheme 21.

P. S. Baran’s 2008 Vinigrol Approach.

To complete the first total synthesis of vinigrol, Baran and his team were able to take advantage of advanced intermediate 188 from their model studies (Scheme 22).²² Ketone 188 was used to install the four carbon bridge methyl group stereocenter in a substrate controlled manner (189) after which the Grob fragmentation was successfully executed to unravel the vinigrol core (190). At this point, Professor Baran and his coworkers were faced with solving a non-trivial challenge, the controlled conversion of the tri-substituted olefin into a tertiary alcohol and a secondary methyl group stereocenter with a syn-relationship. They solved this problem with a remarkable cycloaddition reaction, which afforded heterocycle 191 containing both groups with the necessary stereochemical relationships albeit with some additional redox manipulations needed to unravel the desired native vinigrol functionality. Before tackling the heteocycle, the remaining double bond and ketone were reduced (192) and the resulting secondary alcohol subjected to a Chugaev elimination reaction (193). Reduction and activation of 193 afforded formamide 194, which could be reduced in two steps to a methyl group (195). Substrate controlled dihydroxylation of olefin 195 afforded a diol whose less hindered secondary alcohol could be selectively oxidized to α-hydroxy ketone 196. Ketone 196 was then condensed to hydrazone 197, which was converted in one step to vinigrol using the Shapiro reaction conditions and formaldehyde as the vinyl anion trapping agent. This remarkable final step proceeds without any hydroxyl protecting groups.

8. Njardarson’s Approaches and Total Synthesis of Vinigrol

Professor Njardarson and coworkers have pursued vinigrol using several approaches all of which employ a strategic oxidative dearomatization/Diels-Alder reaction and a late stage fragmentation step. Their first published route, which focues on the Wessely oxidation, is shown in Scheme 23.²³ Alkylation of resorcinol aldehyde 198 with bromoenoate 199 afforded 200, which was cyclized with an organocatalyst (201) using the Stetter reaction. Sodium borohydride reduction of ketone 202 afforded benzylic alcohol 203. The Njardarson group devised a remarkably mild method for forming a quinone methide in situ and trapping it with ethyl vinyl ether to afford 204 as a single diastereomer. Selective reduction of the ester to an aldehyde and homologation with phophonate 205 afforded enoate 206 in a high overall yield. Ethyl acetal 206 was hydrolyzed in the presence of camphorsulfonic acid and the resulting lactol oxidized to lactone 207 using Jones oxidation conditions. Careful hydrolysis of the lactone afforded a sensitive phenol-acid product, which was subjected to an extensive range of Wessely oxidation conditions. This turned out to be most challenging and it was finally revealed that in the presence of lead(IV) acetate (LTA) in hexafluoroisopropanal (HFIP), small amounts of the desired intramolecular Wessely oxidation products could be obtained. This intermediate ortho-quinone spiro lactone underwent a near quantitive intramolecular Diels-Alder cycloadditon to afford 208. The incredibly low overall 8% yield for these three steps is exclusively the result of the Wessely oxidation step. Although gratifying as a proof of concept result, the miserably low yielding dearomatization step required a redesign.

Scheme 23.

J. T. Njardarson’s 2009 Vinigrol Wessely Oxidation Approach.

The Njardarson group’s second vinigrol approach was also focused on trapping a reactive dearomatized phenol intermediate with an adjacent nucleophile followed by an intramolecular Diels-Alder reaction (Scheme 24).²⁴ This time, the Adler-Becker reaction served as the oxidative dearomatization platform. Exhaustive alkylation of ketone 209 followed by palladium mediated deprotection of the unwanted allyl ether afforded bis-allylated ketone product 210. Reduction of the ketone yielded phenol alcohol 211, which was shown to be an excellent Adler-Becker oxidation substrate. Dearomatized quinone spiro-epoxide 212 underwent an intramolecular Diels-Alder cycloadditon to furnish 213 in excellent yield. Despite these promising results, more highly substituted substrates did not fair as well and a redesign was required.

Scheme 24.

J. T. Njardarson’s 2009 Vinigrol Adler-Becker Oxidation Approach.

The Njardarson groups third approach centered on realizing an oxidative dearomatization of a pyrogallol precursor to generate a reactive quinone intermediate capable of undergoing the intramolecular Diels-Alder reaction in the same pot (Scheme 25).²⁵ Horner-Wadsworth-Emmons condensation between phosphonate 214 and aldehyde 215 afforded enoate 216. The terminal alkyne group was then converted into a vinyl iodide and the ester group reduced with diisobutylaluminum hydride to allylic alcohol 217. Remarkably, the oxidative dearomatization union between 217 and commercially available pyrogallol 218 afforded the proposed [2.2.2] bicyclic product 219 in one pot at room temperature. Vinyl iodide 219 was then converted to 220 via a 6-exo-trig radical cyclization. The two ketones were then treated with a cerium nucleophile and the resulting adducts reacted with potassium hydride to facilitate a base mediated Peterson type elimination and formation of tetraene 221. Ring closing metathesis using the Grubbs-Hoveyda second generation catalyst (222) in the presence of benzoquinone to suppress unwanted olefin migration converted 221 to tetracyclic cage 223, which contains the pre-fragmented vinigrol carbocyclic core.

Scheme 25.

J. T. Njardarson’s 2009 Vinigrol Pyrogallol Oxidation Approach.

The clues from the Njardarson group pyrogallol model system motivated an adjustment, which culminated in the second total synthesis of vinigrol being accomplished (Schemes 26–27).²⁶ Etherification of phenol 224 with alcohol 225 afforded 226, whose lactone was reduced and the resulting free phenol strategically protected with a trifluoroethylether, whose role was also to deactivate and guide the pending dearomatization to the more electron rich ether. Dakin oxidation of aryl aldehyde 227 yielded phenol 228, which underwent the proposed oxidative dearomatization-Diels-Alder to form cycloadduct 229. A tandem 6-exo-trig/6-exo-dig palladium cyclization completed the pre-fragmentation vinigrol core (230) in only two steps from the flat aromatic precursor 228. Installation of the methyl group stereocenter on the four carbon bridge was easily accomplished using a substrate controlled hydrogenation. Olefination of ketone 231 proved tricky, but was eventually realized via an addition/Chugaev elimination sequence. Substrate controlled hydrogenation of 232 only afforded the incorrect methyl diastereomer, which is why an iridium catalyzed (233) directed homogeneous hydrogenation conditions were required to form 234. Mild acetal deprotection using lithium tetrafluoroborate and oxidation to aldehyde 235 set the stage for a remarkably selective and mild Baeyer-Villiger oxidation, which afforded diol 236 after reduction with diisobutylaluminum hydride. Diol 236 was converted to pre-fragmentation mesylate 238 using a directed reduction as the key step.

Mesylate 238 underwent a mild Grob fragmentation to form vinigrol core 239 in excellent yield (Scheme 27). Interestingly, the mesylate derived from diol 236 (i.e. the product of direct mesylation of the secondar alcohol of 236) also underwent the fragmentation despite having much poorer orbital overlap. Following hydrogenation of the newly formed double bond, ketone 240 was converted to α-hydroxy ketone 241. Dehydration of this alcohol turned out to be remarkably challenging and revealed an unforeseen and unwanted Grob fragmentation wherein the trifluoroethyl ether served the role of a leaving group. Finally, a solution was realized employing the Burgess dehydrating agent. Reduction of the enone under basic conditions set the stereochemistry required for the isopropyl stereocenter, which was then completed using a two step Wittig olefination/reduction sequence (243). Deprotection of the allyl methyl ether was accomplished in an unusal way with selenium dioxide and a reduction of the resulting enal. Directed epoxidation afforded oxirane 245, which was then strategically ring opened to allylic alcohol 246 with the help of a primary iodide and zinc. A remarkable interrupted selenium dioxide allylic oxidation not only oxidized olefin 246 but also transposed it to the thermodynamically more stable allylic alcohol 247. The final step required removal of a new and stable protecting group, trifluoroethyl ether, from the tertiary alcohol. By taking advantage of the high aciditity of the α-trifluoromethyl methylene protons, enol ether 248 could be formed in situ upon treatment with lithium diisopropyl amide (LDA). This surprisingly stable enol ether was then oxidatively cleaved with osmium tetraoxide to liberate vinigrol.

Scheme 27.

J. T. Njardarson’s 2013 Vinigrol Total Synthesis (Part II).

9. Wang’s Vinigrol Approach

Professor’s Wang and Crowe have evaluated an intramolecular alkylation approach to construct parts of the vinigrol skeleton (Scheme 28).²⁷ Enone 249 is converted to [3.3.1] bicyclic product 250 upon treatment with methyl acetoacetate and base via a Michael/Aldol addition cascade. Methylation of the β-keto ester followed by decarboxylation affords ketone 252, which is then temporarily protected as silylenol ether to allow a hydroboration-oxidation to be performed on the terminal olefin. Deprotection and iodination yielded iodide 256, whose desired epimer underwent the proposed (257) intramolecular enolate alkylation with the help of a lithium tetramethylpiperidide (LiTMP) base.

Scheme 28.

D. Wang’s 2014 Vinigrol Approach.

10. Sun’s Vinigrol Approach

Professor Sun evaluated an alternative intramolecular alkylation approach to form the [5.3.1] bridged bicyclic part of vinigrol (Scheme 29).²⁸ Cyclooctene monoxide 259 was ring opened with methyl cuprate and the resulting alcohol oxidized with IBX to dienone 261. Conjugate addition of isopropyl cuprate afforded primarily diastereomer 262, which then underwent a second substrate controlled cuprate addition in the presence of trimethyl silyl chloride (TMSCl) activator. Alkylation of silyl enol ether 264 then set the stage for the key intramolecular alkylation, which unfortunately did not afford any of 267, but instead provided fused oxirane 268. Interestingly, when ketone 266 was subjected to palladium mediated alkylation conditions inside-out ring system 269 was obtained in high yield.

Scheme 29.

B.-F. Sun’s 2014 Vinigrol Approach.

11. Kaliappan’s Formal Total Synthesis of Vinigrol

Professor Kaliappan most recently completed a nice formal total synthesis of vinigrol (Scheme 30).²⁹ Evans auxiliary based reagent 270 set the stage for the introduction of the first two stereocenters via Michael-alkylation sequence. Removal of the auxiliary and reduction then afforded aldehyde 275, which was homologated using the Corey-Fuchs reaction protocol to alkyne 277. Oxidative cleavage of the terminal olefin followed by addition of chiral Brown’s allylation reagent then yielded 280, which was then promptly deprotected (silyl group and MOM-ether) and strategically reprotected with a silyl and a pivaloate group to afford, after silyl deprotection, alcohol 285. Swern oxidation was followed by addition of vinyl magnesium bromide and another Swern oxidation to furnish enone 288. This enone then underwent a clever one pot intramolecular ene-yne ring closing metathesis followed by a facile intramolecular Diels-Alder cycloaddition with the aid of Grubb’s second generation catalyst and tin(IV) tetrachloride. Wittig olefination of vinigrol core 289 afforded an olefin (290), which could be selectively hydrogenated with a platinum catalyst to afford 291, which is an intermediate in Professor Barriault’s¹⁸ total synthesis of vinigrol.

Scheme 30.

K. P. Kaliappan’s 2014 Vinigrol Formal Total Synthesis.

12. Liao’s Vinigrol Approach

Professor Liao has recently demonstrated a nice oxidative dearomatization approach towards vinigrol (Scheme 31).³⁰ Dearomatization of phenol 292 in the presence of allylic alcohol 293 affords cycloadduct 294 in excellent yield. After a Stille coupling reaction a vinyl nucleophile was added to ketone 295 to form triene 296. Microwave mediate intermolecular Diels-Alder union between 296 and methyl vinyl ketone in dimethylformamide (DMF) then delivered 297. Following protection of ketone 297 as a cyclic ketal the stage was set to test the key oxy-anionic Cope rearrangement step. Gratifyingly, in the presence of potassium base and crown ether in refluxing tetrahydrofuran, the proposed rearrangement proceeded to form 300 in high yield. Interestingly, a thermal oxy-Cope rearrangemet of 299 affords none of the desired product.

Scheme 31.

C. C. Liao’s 2014 Vinigrol Approach.

It is instructive to compare and analyze the key reactions utilized by the twelve research groups approaches towards vinigrol featured in this review. For example, 50% (6/12) of these groups employ a Diels-Alder reaction as a key step while 33% (4/12) use an anionic-oxy Cope rearrangement and 25% (3/12) a metathesis reaction. Other classic reactions playing a key role in these approaches are the Grob fragmentation, Claisen rearrangement and Robinson type annulations. Clearly the “classic” reaction types are holding their own and still playing a vital role in supporting the assembly of complex diterpenoid natural products like vinigrol.

Scheme 26.

J. T. Njardarson’s 2013 Vinigrol Total Synthesis (Part I).

Biographies

Jón was born and raised in the small town of Akranes, Iceland. After graduation, he left his hometown and moved to Reykjavik to start his studies at the University of Iceland. Jon then followed in the footsteps of his Icelandic ancestors and moved west, to America. This journey brought him to New Haven Connecticut, where he chose to pursue a graduate career in Organic Chemistry at Yale University. While at Yale, he joined the research group of a newly hired assistant professor, John L. Wood, after becoming affected by his enthusiasm, energy and exciting new research program. During his doctoral studies Jon worked on the total synthesis of the nonadride natural products CP-225,917 CP-263,114. At the end of his graduate studies Jon was presented with the irresistible offer of moving to New York City to work in the laboratory of Professor Samuel J. Danishefsky the Memorial Sloan-Kettering Cancer Center (MSKCC). While in the Danishefsky group, as a General Motors Cancer Research Scholar, he worked on the total syntheses of the natural products epothilone 490 and migrastatin. Jon moved to Ithaca in 2004 to start his independent career at Cornell University, where he launched a research program focused on natural products and the development of new methods. In 2010 Jon and his group loaded the wagons, journeyed across the continent, and settled in Tucson where he is currently an associate professor of chemistry at the University of Arizona.

Cristian was born in the Moldovian hills of northeast Romania, and for 18 years after that never even contemplated the thought of becoming a chemist. After moving to the US everything changed. He first graduated from Western Connecticut State University with a B.A. in chemistry/biochemistry, and then began his doctoral studies at the University of Vermont, working for his first two years on radical additions to hydrazones in Gregory Friestad’s lab and the latter three under the supervision of Matthias Brewer, developing synthetic methodology for the fragmentation of cyclic C-C bonds. Cristian began his postdoctoral studies in Jon T. Njardarson’s labs at both Cornell University and The University of Arizona, working on the total synthesis of vinigrol. After the Sonoran desert adventure Cristian made the move back east and continued his postdoctoral studies at Yale in Professor Spiegel’s group. There he focused his efforts on the synthesis and evaluation of the most prevalent cross-linking type advanced glycation end product, glucosepane. He still sometimes wonders: why chemistry?… but not for too long.