Toward the total synthesis of maoecrystal V: an intramolecular Diels–Alder route to the maoecrystal V pentacyclic core with the appropriate relative stereochemistry

Abstract

A diastereoselective route to the maoecrystal V core compound (6) has been achieved. Key transformations include an intramolecular Diels–Alder cyclization and an exo-glycal epoxide rearrangement sequence.

The unique and complex structure of maoecrystal V (1, Fig. 1), first isolated by Sun and co-workers in 2004, serves to render it a challenging target for total synthesis. Adding to the level of interest in maoecrystal, is its potent in vitro activity against HeLa cells (IC₅₀ = 20 ng/ml).¹ Particularly noteworthy in this regard is the apparent specificity of its cytotoxicity against gynecological cells.

Figure 1.

Among the defenses which maoecrystal V mounts against those who would undertake its total synthesis, are two contiguous quaternary carbon stereocenters and a tertiary alcohol, embedded within a rigid pentacyclic scaffold. Not withstanding the complexity of the problem, a number of groups, including our own,^2c have initiated investigations directed toward the total synthesis of maoecrystal V.²

In designing our program, we envisioned recourse to an intramolecular Diels–Alder (IMDA) cyclization, hoping thereby to gain access to the bicyclo [2.2.2]-octane core substructure of maoecrystal. In preliminary investigations, the highly functionalized precursor, 2, was indeed found to undergo IMDA cyclization, albeit with the undesired sense of facial selectivity (Fig. 1).^2c We sought to circumvent this setback through the design of a modified sequence, featuring IMDA of a less densely functionalized and symmetrical precursor (cf. 4→5). In devising this new route, we were not insensitive to the potential difficulties inherent in fashioning the requisite trans-fused furanoid ring junction³ (6, see ring B and its junction to ring A). Fortunately, these challenges have now been met, and a concise synthesis of the maoecrystal V core compound (6) in the required stereochemical sense has now been accomplished for the first time. The manner in which we reached this important milestone may well have broader implications in the design of total syntheses.

Our route to the IMDA precursor 4 commenced with a Birch-type vinylogous acylation between substrates 7 and 8, thus providing intermediate 9, possessing one of the requisite quaternary carbon centers (Scheme 1). The latter was subjected to global DIBAL-H reduction, followed by selective MnO₂-mediated reoxidation of the resulting allylic alcohol, to afford 10. A two-step sequence, involving esterification with acyl chloride 11,⁴ followed by formation of the TBS enol ether, provided the target IMDA substrate, 4. In the event, compound 4 readily underwent thermally-induced IMDA cyclization in reasonable yield to generate the desired cycloadduct. Upon exposure to TBAF, the TBS group was hydrolyzed and the phenyl sulfone moiety suffered spontaneous elimination to afford key intermediate 5 in 62% isolated yield.

Scheme 1.

Synthesis of intermediate 5. Reagents and conditions: (a) LDA, −78 °C, THF, 40%; (b) DIBAL-H, −78 °C, DCM, 2 h; (c) MnO₂, rt, DCM, 40 min, two steps, 68%; (d) 11, Py, 0 °C, DCM, 30 min, 86%; (e) TBSOTf, TEA, DCM, −78 °C, 12 h, 91%; (f) toluene, sealed tube, 166 °C, 1 h, then TBAF, THF, 62%.

We next turned to the stereoselective emplacement of the requisite C₁₀ tertiary alcohol. Initial attempts to accomplish direct installation of the C₁₀ alcohol through a cascade 1,4-reduction/oxidation sequence resulted in a 1.5:1 diastereomeric mixture of compounds 12a and 12b (Scheme 2).⁵ Alternatively, we were able to gain access to 12a as a single diastereomer through a three-step sequence. As shown, compound 5 was first treated with basic H₂O₂ to provide, with apparent stereospecificity, the desired epoxide 13. The observed outcome is believed to arise from the presence of the C₁₆ ketone, which creates a stereoelectronic environment that promotes epoxidation from the β-face, as shown.⁶ Upon exposure to sequential epoxide opening and reduction of the resultant iodohydrin, epoxide 13 was converted to 12a in good overall yield. With compound 12a in hand, attentions were directed to installation of the tetrahydrofuran ring. Efforts to directly achieve etherification of 12a through activation of the A-ring 5-exo alkene moiety were unsuccessful. However, a two-step solution was possible. Thus, upon exposure to m-CPBA, 12a was found to undergo regio- and stereoselective epoxidation to afford intermediate 14. Following treatment with p-TsOH · H₂O, a cyclized product was obtained in good yield, as a single diastereomer. However, X-ray analysis revealed the stereochemistry at C₅ to be opposite that required for maoecrystal V (see structure 15, where the A–B rings are cis-fused, Scheme 2).

Scheme 2.

Synthesis of Intermediate 15. Reagents and conditions: (a) PhSiH₃, Mn(dpm)₃, O₂, DCM/ⁱPrOH, 0 °C, 80%; (b) H₂O₂, NaOH, MeOH, 0 °C, 95%; (c) MgI₂, DCM, 45 °C, 40 min; (d) Bu₃SnH, AIBN, toluene, reflux, two steps, 50%; (e) m-CPBA, DCM, rt, 18 h, 72%; (f) p-TsOH · H₂O, DCM, rt, 12 h, 90%.

Our initial attempt to convert 15 to the requisite [6,5] trans-fused system through epimerization⁷ of the C₅ stereocenter is outlined in Scheme 3. Compound 15 was converted to 16 in a straightforward manner. However, exposure of 16 to the action of sodium methoxide in methanol afforded a new product, which was determined to be compound 17. Thus, under these conditions, compound 16 had undergone an intramolecular aldol reaction, while the non-natural cis-fusion of the A–B junction persisted.

Scheme 3.

Attempted Epimerization of C₅. Reagents and conditions: (a) Pd/C, H₂, EtOH, rt, 12 h; (b) DMP, DCM, rt, 12 h, two steps, 75%; (c) NaOCH₃, oxygen free, MeOH, 40 °C, 36 h, 80%.

In order to circumvent the undesired aldol pathway, we sought to remove the C₁₆ ketone functionality. As shown in Scheme 4, dithioketal formation, followed by reduction with Raney-Ni and subsequent oxidation with Dess-Martin reagent furnished the target monoketone 18 in 70% overall yield. This intermediate served as a useful model compound in the context of the actual maoecrystal V synthesis. With 18 in hand, we examined its susceptibility to thermal, base-induced epimerization, hoping to generate at least some traces of trans-fused epimer. Unfortunately, under standard conditions (NaOMe, MeOH, 50 °C), no epimerization was observed. Apparently, the β-face of the A-ring is too congested to allow for the intermolecular delivery of a proton to the ring junction position.

Scheme 4.

Epimerization of 15 and synthesis of the maoecrystal V core (6). Reagents and conditions: (a) ethanedithiol, BF₃ · OEt₂ DCM; (b) Raney-Ni, ethanol, 75 °C, 8 h; (c) DMP, DCM, rt, 12 h, three steps, 70%; (d) NaBH₄, DCM/MeOH, −78 °C to −50 °C, 96%; (e) MsCl, DMAP, DCM, 50 °C, 12 h, 58%, 95% brsm; (f) DBU, toluene, 128 °C, 4 h, 90%; (g) DMDO, DCM, 0 °C, then ether, BF₃, 75%.

There occurred to us the possibility of intramolecular transfer of a strategically placed hydrogen atom to allow for generation of the required A–B trans-fusion.⁸ We proceeded as follows (Scheme 4). Reduction of compound 18 with NaBH₄ afforded intermediate 19 as a single diastereomer. A two-step dehydration sequence provided exo-glycal 20 in good yield.⁹ In the key transformation, exo-glycal epoxide formation was accomplished through exposure to DMDO.¹⁰ Happily, in situ treatment with BF₃ · OEt₂ generated the rearrangement product, target compound 6. The structure of the reduced product 21 was verified by X-ray analysis (Scheme 5).¹¹

Scheme 5.

Synthesis and X-ray structure of compound 21. Reagents and conditions: (a) NaBH₄, −78 °C to −50 °C, 95%.

In summary, we have accomplished the stereoselective synthesis of the maoecrystal V pentacyclic core structure (6). Key transformations include an IMDA reaction (4→5) and apparently for the first time, an exo-glycal epoxide rearrangement sequence to install the rigid tetrahydrofuran moiety. Efforts are underway to apply these teachings to the total synthesis of maoecrystal V.¹²