Volume 65, Issue 33, 15 August 2009, Pages 6571-6575
2008 Tetrahedron Prize for Creativity in Organic Chemistry. The Art of Synthesis: Methods, Strategies and Applications
doi:10.1016/j.tet.2009.05.023 | How to Cite or Link Using DOI
Copyright © 2009 Elsevier Ltd All rights reserved.
|Cited By in Scopus (1)|
|Permissions & Reprints|
A concise strategy for the synthesis of the carbocyclic portion of zoanthenol is disclosed. The key step involves a 6-endo radical-mediated conjugate addition that constructs the quaternary stereocenter at C(12) and closes the B ring in a stereoselective manner. The synthesis of the C-ring fragment uses an enantioselective desymmetrization to simultaneously establish the absolute stereochemistry of two vicinal quaternary stereocenters. In only 17 steps from known compounds, the route affords an ABC ring system containing all three quaternary stereocenters and appropriate functionality to complete the synthesis of zoanthenol.
1. IntroductionThe zoanthamine family of natural products has provided a challenging platform for synthetic innovation since the first member of the family was isolated 25 years ago (Fig. 1)., [1a], [1b], [1c], [1d], [1e], [1f] and  This structurally complex and stereochemically dense family of natural products exhibits a range of important biological activities including anti-osteoporotic, anti-inflammatory, antibacterial, and cytotoxic properties.2 A number of synthetic groups have made important contributions2 in efforts to access these complex alkaloids, including the total syntheses of norzoanthamine by Miyashita et al. in 2004 and by Kobayashi et al. in 2008., [3a], [3b] and [3c] To date no groups have reported the synthesis of zoanthenol (1), but Hirama et al. have disclosed an advanced strategy for its completion.4
|Full-size image (51K)|
Our synthetic efforts toward the zoanthamine alkaloids initially focused on the assembly of the challenging carbocyclic core (ABC rings) of zoanthenol (1). We were drawn to zoanthenol as our initial target due to its oxidized A ring, which allowed for a greater diversity of synthetic approaches because of the range of chemistry available to aromatic systems. The carbocyclic core was of special concern due to the density of stereochemical elements arrayed in the B and C rings. Our retrosynthetic analysis began by unraveling the heterocyclic DEFG rings of zoanthenol, revealing a tricyclic core with a functionalized side-chain (2, Scheme 1). The cyclization events2 required to convert compound 2 to the natural product were well-established as thermodynamically favorable by Kobayashi et al.,, [5a], [5b] and [5c] Williams and Cortez,6 and Miyashita et al.3a Disconnection of the side-chain from intermediate 2 at the C(8)–C(9) bond revealed tricyclic alkyne 3 and lactam 4. Tricycle 3 was envisioned to be accessible from tetracycle 5, which in turn could be formed from aryl bromide 6 by radical-induced intramolecular conjugate addition. This tethered A–C ring system would arise from the addition of benzylic Grignard reagent 7 to enal 8. Enal 8 would be derived from methyl ketone 9, which could be obtained from the desymmetrization of meso anhydride 10.
|Full-size image (49K)|
2. Results and discussion
Zoanthenol's C ring is arguably the most complex region of the molecule because it is fused to three other rings and boasts five consecutive stereocenters, of which three are all-carbon quaternary centers. It was our expectation that the C ring would pose the greatest challenges in stereochemical control, as well as in carbon–carbon bond-construction. The vicinal all-carbon quaternary centers (at C(9) and C(22)) posed a particular challenge, thus our strategy was to establish this functionality at an early stage. We targeted the known Diels–Alder cycloadduct 117 (Scheme 2) as a starting material that could be converted to a meso anhydride, which could in turn be desymmetrized with a chiral reagent to allow the enantioselective synthesis of zoanthenol. In fact, several reports indicated that the selective methanolysis of a meso anhydride was a viable approach., [8a] and [8b] We anticipated that employing a strong acid would induce desilylation of anhydride 11 followed by in situ dehydration. Gratifyingly, the treatment of Diels–Alder adduct 11 with 0.5 equivalents of sulfuric acid in 1,2-dichloroethane produced anhydride 10 in 94% yield. At this point, we were poised to attempt the key desymmetrization. To our delight, treatment of anhydride 10 with quinine and methanol in toluene at 22 °C resulted in the formation of half-ester 12 in >99% yield and 50% ee. Performing the reaction at −50 °C increased the enantiomeric excess to 77% while maintaining a good yield.9 The enantiomer of half-ester 12 could be obtained in similar yield and enantiopurity by substituting quinine for quinidine. Finally, the enantioselectivity could be further increased to 85% by employing O-(−)-(menthyl acetate)quinidine in the reaction at −50 °C for 10 days.10 To the best of our knowledge, this work represents the first example of the desymmetrization of a meso anhydride that simultaneously establishes the absolute stereochemistry of two vicinal all-carbon quaternary stereocenters.11
|Full-size image (17K)|
With our desymmetrized C ring in hand, we sought to oxygenate C(10) and C(21) using the carboxylic acid as a relay for the newly established stereochemical information. We were pleased to find that iodolactonization of half-ester 12 could be affected with good positional selectivity and yield (87%) to provide 13 (Scheme 3). Treatment of the iodolactone with silver acetate led to a syn-periplanar vinylogous attack by the acetate nucleophile to afford allylic acetate 14. Recrystallization enhanced the enantiomeric excess of this acetate to 98%. Methanolysis of iodolactone 14 produced allylic alcohol 15, the connectivity and relative stereochemistry of which were determined unambiguously by X-ray analysis. We sought to convert the ester and lactone functionalities at C(8) and C(23) to the alcohol oxidation state. Thus, allylic alcohol 15 was smoothly silylated upon treatment with TBSOTf and pyridine to provide 16, which was reduced with lithium aluminum hydride to form triol 17 in 83% yield over the two steps. Triol 17 is a versatile intermediate that has facilitated the exploration of a number of synthetic approaches to zoanthenol.12
|Full-size image (36K)|
We turned our attention to the conversion of triol 17 to the proposed C-ring synthon 8. Thus, allylic alcohol 17 was selectively oxidized,13 the primary alcohols were constrained with an acetal functionality, and the resulting enone was hydrogenated to afford ketone 18 in excellent yield over three steps (Scheme 4). The ketone was then α-methylated under standard conditions to provide methyl ketone 9 as a mixture of diastereomers. The mixture was enolized with KHMDS and trapped by N-phenyl bis(trifluoromethanesulfonimide) to afford enol triflate 19 in 92% yield. Treatment of enol triflate 19 under the reductive carbonylation conditions developed during our earlier studies14 led to the formation of enal 8 in 65% yield with complete recovery of the unreacted enol triflate 19.
|Full-size image (38K)|
The A and C rings were smoothly coupled by treatment of enal 8 with benzylic Grignard 7, and  affording alcohol 20 in 87% yield as a 10:1 mixture of diastereomers (Scheme 5). Subsequent oxidation of this alcohol with Dess–Martin periodinane15 provided the corresponding enone (21) in 89% yield. To explore the possibility of a radical-induced cyclization of the A ring into the C-ring enone, the C(13) aryl bromide derivative of enone 21 was synthesized. N-Bromosuccinimide is known to brominate positions para to electron releasing groups., [16a] and [16b] While there was little precedent to suggest the superior directing ability of silyl ethers relative to methyl ethers in this reaction, there was significant evidence that hydroxyls were superior to methyl ethers.17 As a result, we carried out a three-step protocol to regioselectively produce aryl bromide 6. The yield for the sequence was disappointingly low, owing to competitive desilylation and general decomposition during the silylation and bromination steps. We were pleased to find that direct bromination of enone 21 afforded bromoarene 6 in 80% yield as a 4:1 mixture of isomers resulting from bromination at C(13) and C(14), respectively. Gratifyingly, the desired isomer was the major product, and the mixture was adequate for the investigation of the radical cyclization.
|Full-size image (40K)|
With aryl bromide 6 in hand, we began to explore radical cyclization reactions to close the B ring of zoanthenol. Although endo radical conjugate addition cyclization reactions are much less common than exo reactions,18 good precedent for arene radical conjugate addition to make a quaternary center and a six-membered ring did exist.19 To our delight, we found that in the presence of azo radical initiators (e.g., AIBN) and a hydrogen atom donor (e.g., Bu3SnH), significant amounts of ketone 5 were formed from 6 (Scheme 6). In addition to the desired product, debrominated material (21) was obtained. Fortunately, the reduced material is readily separated from tetracycle 5 and can be rebrominated to allow for sufficient material throughput. The most effective conditions for the cyclization employed the azo initiator V-70 with slow addition of Ph3SnH at 32 °C. The use of V-70, which decomposes more readily (t1/2≈10 h at 30 °C) than AIBN (t1/2≈10 h at 80 °C), allows us to initiate the radical reaction at lower temperatures and reduces the amount of debrominated enone recovered, resulting in improved yields of 5. In the event, cyclization occurs to provide tetracycle 5 in 40% yield as a single diastereomer possessing the desired stereochemistry at both the newly formed all-carbon quaternary center and the tertiary center at the B–C ring junction. The stereochemical outcome of the key radical cyclization was unambiguously confirmed by X-ray structure analysis of alcohol 22, which was obtained by DIBAL reduction of ketone 5.20 Importantly, this result completes the synthesis of the carbocyclic core of zoanthenol, requiring just 17 steps from known Diels–Alder adduct 11.
|Full-size image (46K)|
In summary, we have described an efficient synthetic approach to the carbocyclic core of zoanthenol that addresses the demanding combination of a fused polycyclic framework and multiple contiguous stereocenters. The synthetic approach is highlighted by a versatile desymmetrization of a bis-quaternary meso anhydride that enables access to either enantiomer of zoanthenol by selecting the appropriate cinchona alkaloid. The resulting stereochemical information is then relayed around the C ring by a series of diastereoselective reactions. Additionally, good selectivity is observed during the Grignard addition to couple the A and C rings, and the superior directing ability of a silyl ether in the presence of a methyl ether was established in arene bromination reactions. Finally, the key radical cyclization occurs with excellent diastereoselectivity to construct the third all-carbon quaternary center within a single six-membered ring and establishes the correct relative stereochemistry for the tertiary center at the B–C ring junction. Overall, the approach allows access to the challenging carbocyclic core of zoanthenol in 17 steps.
The authors wish to thank Novartis (graduate fellowship to J.L.S.), the Philanthropic Education Organization (Scholar Award to J.L.S.), the Fannie and John Hertz Foundation (graduate fellowship to D.C.B.), Abbott, Amgen, Boehringer-Ingelheim, Bristol-Myers Squibb, Merck, and Caltech for their generous financial support. Additionally, we acknowledge Prof. Li Deng of Brandeis University for the kind donation of O-(−)-(menthyl acetate)quinidine and for helpful discussions.