Synthesis of Cyclohexane-Angularly-Fused Triquinanes

Abstract

Cyclohexane-angularly-fused triquinanes, 6-5-5-5 tetracycles, have attracted the attention of synthetic chemists due to their highly congested core structures and multiple quaternary carbon centers. This review focuses on the six completed total synthesis of naturally occurring cyclohexane-angularly-fused triquinanes in addition to seven notable methodologies that have been developed for the synthesis of these structures.

Graphical Abstract

1. Introduction

Polyquinanes are polycyclic frameworks consisting of five-membered carbocycles that are widely found in natural products.¹ Since the first polyquinane natural product, hirsutic acid C (1, Figure 1), was identified in 1966,² new polyquinanes of diverse skeletal types have been reported, stimulating the development of methodology for the efficient synthesis of these novel structural types.³ In particular, triquinane-bearing natural products have been extensively studied and have served as important platforms for the development of new synthetic methodology as numerous total syntheses of triquinane natural products have been reported to date.^1,4,5 Triquinanes are a class of the polyquinanes comprised of three five-membered rings and can be further categorized into three classes based on the nature of the ring connectivities: linear triquinanes (A),⁴ angular triquinanes (B), and [3.3.3] propellane triquinanes (C; Figure 1).⁵ Recently, cyclohexane-angularly-fused triquinane natural products, 6-5-5-5 tetracycles (D and E), have attracted widespread attention due to their structural novelty, complexity and intriguing biological activities (Figure 2). These structures are predominantly C₂₀-diterpenes bearing five methyl (or four methyl and one methylene) substituents (five carbons) on the core tetracyclic unit (fifteen carbons). To date, three natural products in this class have been reported. Cyclopiane natural products, represented by conidiogenone (2), contain linear triquinane subunits. A linear triquinane structure is also embedded in aberrarone (3). In contrast, waihoensene (4) contains an angularly-fused triquinane structure that is in turn fused to a cyclohexane ring. The cyclohexane-angularly-fused triquinanes are noteworthy for both their highly congested core structures, including multiple quaternary carbon centers, as well as up to eight contiguous stereogenic centers.

Figure 1.

Structures of hirsutic acid C (1), classical triquinanes (A–C) and cyclohexane-angularly-fused triquinanes (D and E)

Figure 2.

Representative naturally occurring 6-5-5-5 tetracycles (isolation year)

In this review, the total syntheses of naturally occurring 6-5-5-5 tetracycles, as well as notable methodologies that have been developed for the synthesis of these structures, will be described. Also included are the syntheses of 6-5-5-5 tetracarbocyclic compounds that have been employed as pivotal intermediates in the total synthesis of other structural types, although the focus is limited to the preparation of carbocyclic systems so that heteroatom-containing 6-5-5-5 tetracyclic compounds will not be discussed herein.

2. 6-5-5-5 Tetracycles Containing a Linear Triquinane

2-1. Total Synthesis of Cyclopiane Diterpenes

The cyclopiane diterpenes are characterized by a rigid 6-5-5-5 tetracyclic ring system that embodies a linear triquinane unit (Figure 3), and can further classified into two groups, conidiogenones and conidiogenols, based on oxidation state. They share identical carbon frameworks except for the oxidation state of C-1 oxygen. Since the isolation of conidiogenone (2) and conidiogenol (8) from extracts of the fermentation broth of Penicillium cyclopium in 2002,⁶ eighteen additional naturally occurring derivatives of conidiogenone/conidiogenol have been reported, most of which contain additional hydroxy functionalities at diverse positions on the carbon framework.⁷ The cyclopianes exhibit diverse biological activities. Conidiogenone (2) and conidiogenol (8) demonstrate potent conidiation-inducing activity with only trace amounts (20 ng) of these natural products required for full induction of conidiogenesis.^6b Conidiogenone (2) was also reported to show potent activity against Methicillin-resistant Staphylococcus aureus (MRSA),^7b whereas conidiogenone C (6) exhibited potent cytotoxic activity against both HL-60 and BEL-7402 cancer cell lines.^7a In addition, the biosynthetic pathway of conidiogenone (2) was recently fully elucidated.⁸

Figure 3.

Representative structures of conidiogenones and conidiogenols

To date, three total syntheses of cyclopiane natural products have been reported, the first of which was reported by Tu in 2016,⁹ fourteen years after the structures of these compounds were first disclosed, a testament to the significant synthetic challenges involved in the total synthesis of this sterically congested compound class. The second and third syntheses were achieved by Snyder (2019) and Zhai (2020), respectively.^10,11 Whereas the Tu group prepared the ABCD tetracyclic skeleton of the cyclopianes via introduction of the B ring onto an ACD spiro-tricycle, the Snyder and Zhai approaches are based on introduction of the A ring onto a pre-exisiting BCD linear triquinane (Scheme 1). Each of these approaches are outlined in detail below.

Scheme 1.

Elaboration of ABCD tetracyclic frameworks of cyclopiane natural products

2-1-1. Tu’s Asymmetric Total Synthesis

Tu and co-workers reported the first total syntheses of ent-conidiogenone (2), ent-conidiogenone B (5), and ent-conidiogenol (8), in which ent-conidiogenol (8) was synthesized from ent-conidiogenone B (5) via ent-conidiogenone (2).⁹ These pioneering efforts led to the correction of the assignments of the absolute configurations of 2, 5 and 8. Retrosynthetically, the ABCD tetracyclic core of common intermediate 9, including the C-9 quaternary carbon center, results from aldol cyclization of the ketoaldehyde derived from spirocyclic compound 10, in which the acetaldehyde moiety was introduced via selective addition of an acetaldehyde equivalent to diketone 11 (Scheme 2). The spiro carbon of 11 is selectively generated from the key vinyl cyclobutanol silyl ether precursor 12 via diastereoselective semipinacol-type ring expansion. Key intermediate 12 was envisioned as the result of addition of the anion derived from 13 to (–)-14, in which the cyclopropane is employed to reduce steric congestion imparted by the corresponding gem-dimethyl group that is present in 5.

Scheme 2.

Tu’s retrosynthetic analysis

In the synthetic direction, racemic bicyclo[3.2.0]heptanone 14 was prepared via Michael addition of unsaturated ester 15¹² and Grignard reagent 16,¹³ followed by intramolecular [2+2] cyclization of the derived ketene under conditions reported by Ghosez and co-workers (Scheme 3).¹⁴ To achieve high enantiopurity of (–)-14, with the absolute configuration to that originally assigned for the natural product, chiral resolution of rac-14 was performed by a sequence of stereoselective reduction to the corresponding alcohol, coupling with a chiral auxiliary,¹⁵ chromatographic separation, detachment of a chiral auxiliary, and re-oxidation to afford (–)-14 with 99% ee. Allylic ether 12 was then generated by 1,2-addition of the vinyl lithium derived from 13 to (–)-14, followed by silylation of the resulting tertiary alcohol. Subjection of 12 to semipinacol rearrangement conditions under the agency of Lewis acid led to migration of the more electron-rich tertiary methine of 12 (red bond), affording 17 with the incorrect relative stereochemical relationship of the newly generated spiro carbon to the existing bicyclo[3.3.0]octane for the synthesis of the conidiogenones.

Scheme 3.

Tu’s first attempt to prepare 11 via semipinacol rearrangement-mediated ring expansion of 12

The selectivity of the key bond migration semipinacol step was successfully reversed by introduction of a phenylthiol moiety to enhance the electron density of secondary carbon of methylene ketone 14 (Scheme 4).¹⁶ The C-10 phenylthiol substituent in 19 promoted the desired bond migration to furnish spirocycle 21 via the conformation shown in 20. The authors proposed that putative steric repulsion between TMS and Me group, which are highlighted in red in 20, led to the adoption of the indicated conformation which promoted Re face attack to establish the critical spiro carbon relative stereochemistry indicated in 21. Construction of the 6-5-5-5 tetracyclic ring system was completed by formation of the B ring of 23. Allylation of 21, followed by ozonolysis of 22 and intramolecular aldol reaction of the derived ketoaldehyde, obtained on ketal hydrolysis, provided tetracyclic compound 23, which was next transformed to 9 by desulfurization and deoxygenation. The completion of the synthesis was achieved by introduction of the remaining methyl group (C-16 of 5) via carbonyl transposition of 9 and β-face methylation of the derived vinylogous ester 24. After stereoselective methylation of 24, reductive carbonyl transposition and enone reduction was effected to afford ent-conidiogenone B (5), in 22 steps and in 0.3% overall yield from 15. The optical rotation and CD spectrum of synthetic 5 led to the correction of the absolute configuration of conidiogenone B to that shown in Figure 3. Epoxidation and reduction of 3 led to the synthesis of ent-conidiogenone (2), which upon α-face reduction was converted to ent-conidiogenol (8).

Scheme 4.

Completion of total syntheses of ent-conidiogenone B (5), ent-conidiogenone (2), and ent-conidiogenol (8) (Tu, 2016)

2-1-2. Snyder’s Asymmetric Total Synthesis

In 2019, Snyder and co-workers completed the second total synthesis of conidiogenones and conidiogenols, which featured their concept of quaternary-center-guided synthesis for complex polycyclic skeletons, which involved sequential annelation of each ring, starting from the C ring of the cyclopianes.¹⁰ As outlined retrosynthetically in Scheme 5, conidiogenone B (5), which was an intermediate in the syntheses of both 2 and 8 by Tu and co-workers (Scheme 4), could be accessed by intramolecular reductive Heck coupling of 26, the side chain of which could be prepared by homoallylation of aldehyde 27 (Scheme 5). The triquinane skeleton of 27 could be established by a series of C-C bond formation steps from cyclopentanone 29 via triflated diquinane 28, in which the stereoselectivity and regioselectivity in the formation of 28 resulted from the quaternary center of 29.

Scheme 5.

Snyder’s retrosynthetic analysis

The total synthesis of cyclopiane diterpenes by Snyder commenced with enantioenriched cyclopentanone 29,¹⁷ which was oxidized to enone 30 via regioselective enolization by hindered tertiary amine (PMP, 1,2,2,6,6-pentamethylpiperidine), followed by IBX oxidation of the enol silane intermediate.¹⁸ Subjection of 30 to the reductive coupling conditions developed by Baran¹⁹ furnished diquinane 31, which on hydrazone-mediated alkylation as described by Corey²⁰ and regiospecific enolate triflate formation provided vinyl triflate 28. B ring formation was then achieved by exposure of 28 to catalytic Pd(II) catalyst in the presence of an oxygen source (n-Bu₄NOAc),²¹ to afford triquinane methyl alcohol 34, after basic hydrolysis of the intermediate acetate 33. This represents the first example of direct C–O bond formation from an unactivated C(sp₃)−Pd(II) intermediate in a Pd(0)/Pd(II) catalytic system. DMP oxidation of 34 gave aldehyde 27, which underwent radical-mediated, chemoselective NHK-type reaction²² with iodide 35 to produce 37 as a 3:1 mixture of diastereomers. The construction of the tetracyclic core was completed by reductive Heck cyclization²³ of 37 to furnish methylene cyclohexane 38. Finally, stereoselective α-face reduction of the exo-methylene, followed by IBX oxidation of the resultant alcohol, afforded conidiogenone B (5), which was converted to 2 and 8, as described by Tu. The Snyder approach led to the synthesis of conidiogenone B (5) in 13 steps with 6.4% overall yield from commercially available starting materials.

Additionally, the Snyder group achieved the syntheses of conidiogenone C (6) and D (7), more highly oxidized members of conidiogenone family (Scheme 7). Regioselective epoxidation of the terminal olefin of 29 and reductive cyclization afforded, after MOM protection of alcohol, bicyclic compounds 39 and 40. Subjection of 39 and 40 to the same series of steps described above for the preparation of conidiogenone B (5) afforded conidiogenone C (6) and D (7), respectively, which represent the first syntheses of these more highly oxidized congeners of the conidiogenone family.

Scheme 7.

Synthesis of conidiogenone C (6) and D (7)

2-1-3. Zhai’s Asymmetric Total Synthesis

In 2020, Zhai and co-workers reported the synthesis of conidiogenone B (5), which has been converted to conidiogenone (2) and conidiogenol (8) in the syntheses of Tu and Snyder, respectively. This approach was notable for the use of the Danheiser annulation to annelate the A ring of the cyclopiane diterpenes (Scheme 8).¹¹ The Zhai group envisioned that the cyclohexanone moiety (A ring) of 5 could be accessed from 41 via intramolecular aldol reaction with concomitant epimerization of C-5, in which the dicarbonyl functionality of 41 could be introduced via ozonolysis of alkene 42. The tetraquinane skeleton of 42 could be established from triquinane 44 via Danheiser annulation^24,25 with silyl allene 45. This reaction would be initiated by the conjugate addition of 45 to 44 and the resulting intermediate 43 would cyclize to produce tetraquinane 42. Cyclopentenone 44 would result from one-pot Nicholas/Pauson-Khand reaction of functionalized cyclopentane 46, which constitutes the D ring of the cyclopianes.

Scheme 8.

Zhai’s retrosynthetic analysis

Zhai’s asymmetric synthesis of conidiogenone B (5) began from the enantioenriched alcohol 48 (92% ee), which was generated in high optical purity by CBS reduction of readily available trimethylcyclopentenone 47 (Scheme 9).²⁶ Enol ether formation was achieved by reaction of the alcohol with n-butylvinyl ether 49. Subsequent Claisen rearrangement gave the corresponding aldehyde, which was subjected to van Leusen reaction²⁷ for one-carbon homologation to afford nitrile 50. Hydrogen atom transfer (HAT)^28,29 of 50 in the presence of an iron catalyst yielded diquinane 51, the C-2 position of which was oxygenated after stereoselective alkyne addition to the carbonyl group, leading to the formation of compound 52. A sequence of oxidative 1,2-diol cleavage and direct reduction gave diol 53, the primary alcohol of which was selectively eliminated via Grieco’s protocol³⁰ to afford enyne 46. The linear triquinane framework was completed via one-pot Nicholas/Pauson-Khand reaction sequence³¹ to give compound 44, which on Danheiser annulation with silyl allene (±)-45 in the presence of Lewis acid was converted to tetraquinane 42. After deoxygenation by a three step-sequence, 54 was subjected to ozonolysis followed by the treatment of hydrochloric acid to furnish, via intermediate 55, conidiogenone B (5), along with trisubstituted alkene 56, that could be subsequently isomerized to conidiogenone B (5), albeit in moderate yield. The Zhai approach led to the synthesis of conidiogenone B (5) in 14 steps and in 4.7% overall yield from 47.

Scheme 9.

Total synthesis of conidiogenone B (Zhai, 2020)

2-2. Synthetic Approach toward Aberrarane Diterpenes

Aberrarone (3) is the unique member of the aberrarane class of diterpenes. As outlined in Figure 4, the orientation of the methyl groups of the AB ring system in 3 is different from that of the cyclopianes, i.e., 2. In addition, the CD moiety of 3, which contains three additional carbonyl functionalities, is more highly oxidized than that of 2.³² It was subsequently discovered that 3, which was isolated from the Caribbean Sea whip Pseudopterogorgia elisabethae, exhibits in vitro antimalarial activity against a chloroquine-resistant strain of Plasmodium falciparum.³³

Figure 4.

The structures of aberrarone (3) and synthetic target 57 of Ito’s work

Although the total synthesis for aberrarone (3) has not yet been achieved, synthetic studies directed toward the assembly of the core tetracyclic framework 57, lacking both the D ring carbonyl substitutents and the AB methyl groups that are present in 3, were reported by Ito and co-workers in 2015.³⁴ As shown in Scheme 10, they envisioned that the A ring of 57 could be prepared by intramolecular aldol reaction of 58, which would result from conjugate addition of a four-carbon appendage to the triquinane core 59. Enone 59 was envisioned as the product of intramolecular aldol reaction and dehydration of ketoaldehyde 60, which would ultimately result from 62 via 61.

Scheme 10.

Synthetic strategy toward tetracyclic compound 57

The conversion of 62 to the core tetracyclic framework 57 is outlined in Scheme 11. The preparation of 61 proceeded from 62 by a standard eight-step sequence. Stereoselective 1,4-addition of Grignard reagent 63 from convex face of 61, and treatment of the resulting Michael adduct 64 with 10% H₂SO₄, yielded the corresponding aldol product, the hydroxy group of which was mesylated and eliminated to give tricycle 59. Introduction of the cyclohexanone ring was achieved in an analogous manner to that described for 64 with the homologous Grignard reagent 65. However, instead of dehydration of the initially formed aldol product, oxidation of the ketol intermediate led to the formation of 57.

Scheme 11.

Synthesis of aberrarone tetracyclic core 57 (Ito, 2015)

2-3. Intermediates in the Total Synthesis of Magellanine-Type Alkaloids

Magellanine (68, Scheme 12), a natural product of the Lycopodium alkaloid class,³⁵ was first isolated from the club moss Lycopodium magellanicum in 1976 and contains a 6-5-5-6 tetracyclic core, including a N-methyl piperidine ring and six stereogenic centers.³⁶ Eight total syntheses of magellanine have been reported,³⁷ along with the synthesis of other naturally occurring analogs such as magellaninone (53)^37d–h,38 and paniculatine (54),^{37d–f,39,40} the oxidation state of which differ in either the A or B ring, respectively. Three of these approaches have utilized the 6-5-5-5 tetracyclic ring system of 71 as a key intermediate. D ring double bond oxidative cleavage would give either diol 72 or dialdehyde 73, which on either amine double displacement or double reductive amination, respectively, would generate the piperidine core of the magellanine natural products. Approaches to the synthesis of the key intermediate 71 as well as the magellanine core structure are outlined below.

Scheme 12.

Transformation 6-5-5-5 tetracyclic skeleton 71 to magellanine-type natural products (68–70)

2-3-1. Mehta’s Synthetic Approach

In 1990, Mehta and co-workers described the synthesis of the complete ring system of magellanines (Scheme 13).⁴¹ The synthesis began with the manipulation of Cookson’s diketone (74),⁴² which was selectively deoxygenated to give monoketone 75 in six steps. Thermally-induced [2+2]-cycloreversion of compound 75 under flash vacuum pyrolysis conditions afforded linear triquinane 76, albeit in low yield (18%).⁴³ Isomerization of 76 to 77, followed by conjugate addition of a C₄-unit afforded cis-anti-cis compound 78 as the major product (in a 2:1 ratio with the corresponding cis-syn-cis diastereomer). After oxidation of 78 to the corresponding cyclopentenone, Wacker-type oxidation gave ketone 79, which on intramolecular Michael addition gave 80, containing the requisite 6-5-5-5 tetracyclic core.⁴⁴ The complete ring system 81 of the magellenines was obtained by oxidative cleavage of 80 to generate the corresponding diol, followed by activation and double displacement by methylamine.

Scheme 13.

Synthesis of the ring system of magellanines (Mehta, 1990)

2-3-2. Overman’s Total Synthesis

The first total syntheses of (−)-magellanine (68) and (+)-magellaninone (69), described by Overman and co-workers in 1993, each incorporated the Prins-pinacol rearrangement as a key step in the assembly of the 6-5-5-5 tetracycle (Scheme 14).⁴⁵ Ring expansion of 82 led to the formation of diquinane 83 by reaction with [bis(methylthio)methyl]lithium and copper(II) triflate as reported by Cohen.⁴⁶ Conversion of the α-methylthionyl ketone 83 to vinyl iodide 84, in five steps, was followed by elaboration to dienyl acetal 85, the substrate for the key Prins-pinacol rearrangement step. Exposure of 85 to tin(IV) chloride yielded the tetracyclic ketones 88 and 89 in a 2:1 ratio in 57% combined yield. The observed stereoselectivity was rationalized by selective Prins cyclization from the convex face of the diquinane fragment of 86 to give 87, which on pinacol rearrangement afforded 88 and 89,⁴⁷ which were ultimately converted to magellanine (68) and magellaninone (69), as described in Scheme 12.

Scheme 14.

Total synthesis of magellanine-type natural products (Overman, 1993)

2-3-3. Liao’s Total Synthesis

The total synthesis of magellanine (68) by Liao and co-workers, described in 2002,⁴⁸ featured the use of an oxa-di-π-methane (ODPM) rearrangement^49,50 to generate the linear triquinane skeleton (Scheme 15). Reaction of masked o-benzoquinone 91,⁵¹ which was generated from acetovanillone (90), in the presence of diacetoxyiodobenzene (DAIB) and methanol, with cyclopentadiene afforded a single Diels-Alder cycloadduct 92, the relative stereochemistry of which was established by endo cycloaddition. Irradiation of 92 afforded triquinane 93 through ODPM rearrangement. After cyclopropane ring-opening of 93, vinyl triflate 94 was obtained and then elaborated to alkenyl methyl ketone 95. Silyl enol ether formation was followed by intramolecular cyclization via the generation of oxa-π-allylpalladium(II) intermediates⁵² to afford tetracyclic enone 96, which was finally converted into magellanine (68) as described above (14 steps in 9% overall yield from commercially available starting materials).

Scheme 15.

Total synthesis of magellanine (Liao, 2002)

2-3-4. Mukai’s Total Synthesis

The total syntheses of three three magellanine-type alkaloids 68–70, reported by Mukai and co-workers in 2007,⁵³ featured two Pauson-Khand reactions⁵⁴ as key ring-forming steps. Enantiopure enyne 97,⁵⁵ derived from diethyl l-tartrate, underwent the first Pauson-Khand reaction to generate bicyclic enone 98 with higher stereoselectivity than previously described.⁵⁶ Subjection of silyl-protected 99 to the Ueno-Stork reaction⁵⁷ provided, via intermediate 100, cyclic acetal 101, which was subsequently oxidized to lactone 102. The observed C-1/C-9 relative stereochemistry, which is critical for the selective construction of the magellanines, could be attributed to the controlling effect of the C-2 stereochemistry, as well as addition of the allyl moiety from the convex face of the bicyclo[4.3.1]nonane moiety. An eleven-step reaction sequence was then required to affect the conversion of 102, via 103, to 104, the substrate for the second Pauson-Khand cyclization. Exposure of 104 to the same reaction conditions that were employed with 97 successfully yielded the 6-5-5-5 tetracyclic compound 105, which, after several additional steps, led to the formation of three magellanine-type alkaloids (68, 43 total steps, 1.7% overall yield; 69, 43 total steps, 1.9% overall yield; 70, 45 total steps, 2.8% overall yield).

3. 6-5-5-5 Tetracycles Containing an Angular Triquinane

3-1. Total Synthesis of Waihoensene

Waihoensene (4, Figure 5) is a tetracyclic diterpene which was isolated from the New Zealand podocarp, Podocarpus totara var waihoensis, by Weavers and co-workers in 1997.⁵⁸ Its structure features an angularly fused triquinane unit (indicated with blue bonds in 4) that is in turn angularly fused to a six membered ring, the total structure of which contains six contiguous stereogenic centers including four contiguous quaternary carbon centers (red dots).⁵⁹ The structure is further notable for the absence of any functional groups except for the exo-methylene. The twenty-year lapse between the isolation of 4 in 1997 and the first total synthesis in 2017⁶⁰ attest to the unique and complex structure of 4. Two additional total syntheses of waihoensene (4) have been reported as well as two model studies for the tetracyclic core skeleton of 4.

Figure 5.

Structures of (+)-waihoensene (4)

3-1-1. Lee’s Racemic Total Synthesis

The key feature of the Lee approach to the synthesis of waihoensene⁶⁰ is the use of trimethylenemethane (TMM) cycloaddition,⁶¹ a similar strategy to that which was disclosed by Lee and co-workers in 2004 for the synthesis of crinipellin (Scheme 17).⁶² Retrosynthetically, waihoensene (4) could be accessed from 106 via introduction of two methyl groups onto the D ring and ketone methylenation. The carbonyl group of 106 could be introduced by selective allylic oxidation of compound 107, the tetracyclic framework which is selectively generated from TMM diyl intermediate 108, via the more stable reacting conformation 108′′, which avoids the 1,3-diaxial interaction of two methyl groups as shown in the diastereomeric orientation illustrated in 108′. The TMM diyl intermediate 108 could be generated in situ via intramolecular tandem cycloaddition reaction of the biradical derived from 109, which is generated from allenyl diazo substrate 110, by loss of dinitrogen. The diazo allene 110 is prepared from allenyl aldehyde 111.

Scheme 17.

Lee’s retrosynthetic analysis toward waihoensene (4)

In the synthetic direction, racemic β-keto ester 112⁶³ was transformed to tosylate 113 in nine steps (Scheme 16), which was then treated with the Grignard reagent derived from bromide 114 for copper(I)-catalyzed S_N2′ reaction to give allenyl aldehyde 111 upon deprotection of TBS group, and Swern oxidation. Conversion of 111 to the corresponding allenyl diazo intermediate 110 was achieved via reaction of 111 with tosylhydrazone (Scheme 17).⁶⁴ Intramolecular [2+3] cycloaddition reaction of the diazo intermediate with the allene provided methylenepyrazole intermediate 109, which on elimination of dinitrogen led to the formation of the key TMM diyl intermediate 108.⁶⁵ Through the favored conformation 108′′, [2+3] cycloaddition of the reactive diyl moiety of 108 with the exo-methylene group furnished 6-5-5-5 tetracyclic 107 as the major product, the relative stereochemistry of which was consistent with that of waihoensene (Scheme 18).

Scheme 16.

Total synthesis of magellanine-type alkaloids based on two Pauson-Khand reactions (Mukai, 2007)

Scheme 18.

Racemic total synthesis of waihoensene (Lee, 2017)

The second part of the synthesis commenced with allylic oxidation of 107 to provide enone 106, which served as a precursor for the exo-methylene group that is present in 4 (Scheme 18). While attempted allylic oxidation of 107 using pyridinium chlorochromate (PCC) and tert-butyl hydroperoxide (TBHP) led to modest yields of 106, the authors developed a more efficient, albeit four-step approach to the synthesis of 106. Dihydroxylation of 107, followed by selective tosylation of the resulting secondary alcohol yielded the corresponding tosylate, which was subjected to elimination via treatment with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) to provide 115. Oxidative rearrangement of 115 using pyridinium dichromate (PDC) afforded enone 106, which on exposure to higher-order methyl cuprate in presence of boron trifluoride etherate⁶⁶ led to the diastereoselective formation of 116. Intriguingly, treatment of 116 with LiHMDS and iodomethane provided 117 as a single product, which possessed all methyl groups with the correct relative stereochemistry for the synthesis of waihoensene. Based on the X-ray crystal structure of 116, Lee and co-workers suggested that methylation of 116 from the α-face should be disfavored due to steric hindrance that could be attributed to the substituted C-15 carbon and the concave nature of the enolate derived from 116. Regioselective C-3 enolate formation, leading to 117 after methylation, was rationalized by the difference between the dihedral angles of the pseudo-axial C-3 H (H-3) and pseudoequatorial C-1 H (H-1). Methylenation of 117 using the Petasis reagent⁶⁷ then delivered waihoensene (4) (total 20 steps and overall yield 0.44% from commercially available starting material).

3-1-2. Huang-Yang’s Asymmetric Total Synthesis

In 2020, Huang, Yang, and co-workers reported an elegant asymmetric total synthesis of waihoensene (4) that featured both a Conia-ene reaction and Pauson-Khand reaction to construct the carbon framework of 4. In this approach, the relative stereochemical relationships in 4 are all controlled by the C-5a quaternary stereogenic center in diyne 122 (Scheme 19).⁶⁸ The formation of 4 would result from 118 via stereoselective reduction of the exo-methylene from the sterically more hindered concave face and subsequent ketone olefination. The two methyl groups of the cyclopentanone ring (D ring) of 118 would be stereo- and regioselectively introduced from 119, as previously reported by Lee (Scheme 18).⁶⁰ The CD ring system of 119 would be prepared by Pauson-Khand cyclization of enyne 120. The authors anticipated that Conia-ene-type reaction of 121, the enolate obtained on deprotonation of 122 (which would in turn be prepared from 123), would generate the cis-bicyclo[4.3.0]nonane skeleton 120.

Scheme 19.

Huang-Yang’s retrosynthetic analysis for the synthesis of waihoensene (4)

As shown in Scheme 20, reaction of 124 and Grignard reagent 125 gave enone 126, from which the requisite C-5a stereogenic center in 127 was installed (91% enantiomeric excess) using the method of Alexakis.⁶⁹ Sakurai allylation⁷⁰ of 127 gave 128, which on reaction with ozone and the Bestmann reagent afforded diyne 122 in modest overall yield. The preparation of the 6-5-5-5 tetracyclic ring system via Conia-type reaction was achieved by reaction of 122 with a catalytic amount of potassium tert-butoxide⁷¹ to deliver cis-fused bicyclic system of 120. Subjection of this product to Pauson-Khand cyclization produced tetracycle 119 with the desired relative stereochemistry for the synthesis of waihoensene. Elaboration of 119 to waihoensene required regio- and stereoselective introduction of the three remaining methyl groups and ketone methylenation to introduce the exocyclic olefin. Ni-catalyzed methylation of 119, followed by C-9 ketone olefination gave 129 through selective ketal protection/deprotection of the C-2 ketone. Attempted hydrogenation of the exocyclic Δ^9,15 double bond led exclusively to reaction from the more sterically accessible convex face. However, the desired product 116 was ultimately obtained via intramolecular hydrogen atom transfer (HAT) reaction²⁹ using conditions described by Baran.⁷² Experimental evidence was provided to support the intermediacy of 130 in the formation of 116. Methylation and olefination, based on the Lee synthesis (Scheme 18), gave (+)-waihoensene (4) (total 15 steps and 3.8% overall yield).

Scheme 20.

Asymmetric total synthesis of waihoensene (Huang and Yang, 2020)

3-1-3. Snyder’s Asymmetric Total Synthesis

Almost at the same time, Snyder and co-workers reported the asymmetric total synthesis of waihoensene (4)⁷³ using a closely related strategy to that disclosed by Huang-Yang.⁶⁸ They also employed a sequence of Conia-type ene reaction/Pauson-Khand cyclization to generate the requisite carbocyclic ring system. As shown in Scheme 21, the main difference between these approaches is the methodology employed for the synthesis of the cis-hydrindane moiety. In Huang-Yang’s work, the cis-hydrindane of 120 (blue bonds) was prepared from cyclohexanone 122 through Conia-ene reaction under basic conditions, whereas in the Snyder approach, the cis-hydrindane 131 was prepared via gold-catalyzed Conia-ene reaction of cyclopentanone 132. Both approaches involved the use of the C-5a stereocenter (red dots) to control the establishment of all other stereochemical relationships in the synthesis of 4.

Scheme 21.

Formation of cis-hydrindane of Huang-Yang’s and Snyder’s works

In the Snyder approach, the establishment of the C-5a absolute stereochemistry in 137 was achieved via asymmetric conjugate addition chemistry through the agency of N-heterocyclic carbene complex 136 as described by Hoveyda (Scheme 22).⁷⁴ Cyclopentenone 135, prepared from 133 and Grignard reagent 134, was enantioselectively converted to 137. Carboxylation of 137 using LiHMDS and Mander’s reagent gave a mixture of 138 and 139. The undesired regioisomer 139 could be recycled to 137 via Krapcho decarboxylation⁷⁵ and re-carboxylation. Desilylation of 138, followed by Conia-ene reaction, afforded 131, with establishment of the requisite cis-hydrindane stereochemistry. Interestingly, hydrogenation of 131 with PtO₂ led to predominant formation of the desired isomer 140 (3.2:1.0 dr), the result of the addition of hydrogen from the concave face, i.e., the same selectivity as observed in a closely related system by Shenvi.^29,76 After methylenation of 140, the 3-butyne moiety of 141, which is required for the Pauson-Khand reaction, was installed in a six-step sequence that involved ester reduction, HWE olefination, and alkyne formation using the Ohira-Bestmann reagent.⁷⁷ Finally, Pauson-Khand reaction of 141 at elevated temperature (160 °C) afforded tetracycle 106, which was spectroscopically identical to 106 prepared by Lee (Scheme 18; total 17 steps and 0.74% overall yield).

Scheme 22.

Asymmetric total synthesis of waihoensene (Snyder, 2020)

3-2. Miscellaneous Approaches

In 1999, Moore and co-workers described the synthesis of the tetracyclic core of waihoensene (4) that featured a sequence of inter- and intramolecular Michael reactions.⁷⁸ As shown in Scheme 23, reaction of bicyclo[3.2.0]heptanone 142 with lithiated cyclohexene 143 afforded anionic intermediate 144, which underwent oxy-Cope rearrangement in situ to generate 147, upon elimination of methanol from 145 and acidic workup of intermediate 146. Transannular ring closure of 148 to give 150, containing the tetracyclic ring system of waioensene, was triggered by intermolecular conjugate addition of thiophenolate to 147. The authors proposed that the elimination reaction to give 150 is effected by the ketone enolate of 149 based on the proximity of the enolate to the β-proton (highlighted in red) that is oriented trans to the thiophenyl group.

Scheme 23.

Synthesis of the core structure of waihoensene (4) (Moore, 1999)

In 2020, Wang, Tu and co-workers reported a tandem reaction sequence involving Castro-Stephens coupling/acyloxy shift/cyclization/semipinacol rearrangement for the synthesis of highly functionalized spirocyclo[4.5]decane derivatives.⁷⁹ As illustrated in Scheme 24a, Castro-Stephens coupling⁸⁰ of propargyl ester 151 and allylic bromide 152 under basic condition give 153, on which 1,3-acyl shift generate cationic intermediate 154 via gold(I) catalysis. The cationic cyclization/semipinacol rearrangement of 154 leads to the synthesis of multisubstituted spirocyclo[4.5]decane 155 and related skeleta. Wang, Tu and colleagues subsequently applied this strategy to the synthesis of the 6-5-5-5 carbon framework of the waihoensene as outlined in Scheme 24b. Intramolecular [2+2] cycloaddition of the keteniminium intermediate derived from amide 156 gave bicyclo[3.2.0]heptane 157.^14c,81 The elaborated bicyclo[3.2.0]heptane 158 was reacted with the acetate of butynol to give spirocyclo[4.5]decane 160 via the intermediacy of 159 in moderate yield (3.2:1 mixture of diastereomers). Intramolecular aldol reaction of the ketoaldehyde derived from 160 furnished tetracycle 161, containing the tetracyclic ring system of waihoensene (4).

Scheme 24.

(a) Strategy design toward the spiro-cyclo[4.5]decane skeleton 155. (b) Synthetic route toward the ring system of waihoensene (Wang and Tu, 2020)

4. Conclusion

This review has focused on the preparation of tetracyclic ring systems in which a cyclohexane is angularly fused to a triquinane core. This tetracyclic moiety has become an attractive target for the synthetic community due to the challenges afforded by multiple stereocenters and quaternary carbon centers in these structurally complex ring systems. Diverse strategies have been reviewed, from semipinacol rearrangement to oxa-di-π-methane rearrangement, trimethylenemethane cycloaddition, and Pauson-Khand cyclization, which have led to efficient approaches to these structures.

Scheme 6.

Total synthesis of cyclopiane diterpenes (Snyder, 2019)

Biographies

Jeffrey D. Winkler matriculated at Harvard College where he participated in the research programs of Professor James Wuest, Dr. Larry Blaszczak, and Professor E. J. Corey. He graduated cum laude in Chemistry in 1977. He completed his Ph.D. degree in the laboratories of Professor Gilbert Stork in 1981. He next moved to the laboratories of Professor Ronald C. D. Breslow as an American Cancer Society Postdoctoral Fellow. In 1983, he joined the Chemistry Department at the University of Chicago as an Assistant Professor, and moved in 1990 to the University of Pennsylvania where he is currently Merriam Professor of Chemistry, Undergraduate Chair in Chemistry, and a member of both the Abramson Cancer Center and the Center for Cancer Pharmacology.

In his independent career first at Chicago and now at Penn, he has established an active research laboratory engaged in both synthetic organic and bioorganic chemistry. His early work featured the first total syntheses of manzamine A, saudin, and ingenol. His more recent work has focused on collaborative efforts at the interface of chemistry and medicine with researchers at the University of Pennsylvania School of Medicine, the Wistar Institute and the Children’s Hospital of Philadelphia.

Hongjun Jeon graduated from the College of Pharmacy at the Seoul National University (Republic of Korea) and received his Ph.D. in 2017 from the same college under the supervision of Professor Sanghee Kim. His Ph.D. research focused on the total synthesis of Cephalotaxus alkaloids and related synthetic methodologies. He is currently a postdoctoral fellow in the Winkler laboratories in the Department of Chemistry at the University of Pennsylvania, working on the total synthesis of naturally occurring diterpenes and the synthesis of BRAF-kinase inhibitors.