Letter Cite This: Org. Lett. 2018, 20, 3888−3891
pubs.acs.org/OrgLett
Enantioselective Total Synthesis of (+)-Anthecularin Yusuke Ogura,† Shoko Okada, Naoki Mori, Ken Ishigami,‡ and Hidenori Watanabe* Department of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
Downloaded via BOSTON UNIV on July 6, 2018 at 10:25:08 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
S Supporting Information *
ABSTRACT: An enantioselective total synthesis of (+)-anthecularin, an antiplasmodial and antitrypanosomal sesquiterpene lactone, has been achieved in 3.9% overall yield through 18 steps from a known dibromo alcohol. The key features of the synthesis include an intramolecular Claisen-type cyclization of a formyl-protected hydroxyl lactone to construct a bicyclic intermediate with a quaternary stereogenic center and a stereocontrolled 1,2-addition of vinyllithium to a methoxyethyl-protected spirocyclic hydroxyl enone to install a tetrasubstituted asymmetric center with excellent diastereoselection. This first enantioselective synthesis of anthecularin enabled the determination of its absolute configuration as 2R, 3R, 4S, 8R.
M
Structurally, anthecularin (1) features an unusual tetracyclic fused ring system composed of oxabicyclo[3.2.1]octene, cyclohexene, and γ-butyrolactone rings, as well as four contiguous asymmetric centers in the molecule, which makes it a synthetically challenging target. Although a total synthesis of the racemate of 1 via an intramolecular [5 + 2] cycloaddition reaction was previously achieved by Pattenden and co-workers,4 no report on its synthesis in an enantioselective manner has been disclosed so far. We describe herein the first enantioselective total synthesis of anthecularin (1), which enabled the determination of its absolute stereochemistry. Our retrosynthetic analysis of anthecularin (1) is shown in Scheme 1. For the total synthesis of anthecularin (1), the stereoselective construction of the doubly allylic tetrasubstituted C8 chiral center would be crucial. Keeping this in mind, we envisaged two synthetic approaches. The first one (route A) features a double ring-closing metathesis5 (RCM) of 2 to install the C8 asymmetric center, the stereochemistry of which would be controlled by the steric restriction offered by the tetrahydrofuran ring system in 2. The tetraene 2, in which the C8 carbon atom is a prochiral center, would be accessible by double vinyl addition to bicyclic lactone 5 in one pot and subsequent etherification of the resulting diol 3. In the second approach (route B), the vinyl addition/RCM sequence is conducted twice in a stepwise manner (5 → 4 → 7 → 6 → 1), although the diastereoselective addition of the second vinyl group to 7 to form 6 could be a challenging task. The optically active bicyclic compound 5, which serves as a common key intermediate for both strategies, would, in principle, be prepared by a sequence of reactions involving the Evans
alaria is one of the most prevalent parasitic diseases in the world, threatening the lives of millions of people each year.1 Although artemisinin-based combination therapies (ACTs) remain effective for the treatment of malaria, recent reports on the emergence of malaria parasites which resist existing antimalarial drugs, including artemisinin derivatives, clearly indicate the need for the development of new therapeutics.2 Anthecularin (1), a new skeletal type of sesquiterpenoid isolated from a lipophilic extract of the aerial parts of Greek Anthemis auriculata (Figure 1), exhibits
Figure 1. Structure and numbering of anthecularin (1).
antiplasmodial activity against the drug-resistant Plasmodium falciparum K1 strain (IC50 23.3 μg/mL).3 This biological property is considered to be ascribable to the specific inhibition of Pf FabI (IG50 14 μg/mL) and Pf FabG (IG50 28.3 μg/mL), key enzymes of the plasmodial fatty acid synthase (Pf FAS) which is a type II multiple enzyme complex and differs radically from human type I fatty acid synthase. Besides the antiplasmodial action, 1 also shows trypanocidal activity against Trypanosoma brucei rhodesiense, a protozoa response for African trypanosomiasis, with an IC50 value of 10.1 μg/mL without any cytotoxicity on mammalian L6 cells. From these desirable pharmacological profiles, anthecularin (1) deserves attention as a potential lead for new antimalarial and antitrypanosomal agents. © 2018 American Chemical Society
Received: May 9, 2018 Published: June 4, 2018 3888
DOI: 10.1021/acs.orglett.8b01467 Org. Lett. 2018, 20, 3888−3891
Letter
Organic Letters Scheme 1. Retrosynthetic Analysis of (+)-Anthecularin (1)
asymmetric aldol reaction between 106 and 8, installation of the C2 side chain by use of 9, and an intramolecular Claisen-type cyclization to construct the protected cyclic hemiacetal moiety. The preparation of the common key intermediate 15a (i.e., 5: P = Me) is shown in Scheme 2. Our synthetic efforts
of which could be readily removed by SiO 2 column chromatography. It was also possible to epimerize the undesired C3 epimer to 12 by heating its benzene solution in the presence of DBU (12/3-epi-12 = 10:1). Hydrolytic removal of the chiral auxiliary of 12 and subsequent reduction of a mixed anhydride intermediate derived from the resulting carboxylic acid gave an alcohol, which was then protected as its TBS ether 13. Alkylation of 13 with homoallylic iodide 9 afforded an alkylation product as a single diastereomer, although its stereochemistry was not determined. The TBS group of the alkylated lactone was then removed with TBAF, and the alcohol produced was converted into the corresponding formate 14. One-carbon installation to the C2 position accompanied by concomitant formation of the C2 quaternary stereocenter was realized efficiently by an intramolecular Claisen-type cyclization of 14 using KHMDS in THF at −78 °C, giving rise to a hemiacetal alkoxide intermediate, which upon treatment with methyl triflate provided the key intermediate 15a as a single diastereomer. The stereochemistry of 15a was determined by observing diagnostic NOE correlations depicted in Scheme 2. With the bicyclic lactone 15a in hand, we attempted double vinyl addition to 15a to directly obtain tetraene intermediate 16 (i.e., 3: P = Me) (Scheme 3). Treatment of 15a with 2.3 equivalent of vinyllithium in THF at room temperature, however, did not afford the desired product 16 but, instead, lactol 17 as a diastereomeric mixture in 90% yield. The unexpected formation of 17 would be rationalized by 1,4addition of a second equivalent of vinyllithium to a temporarily formed alkoxy enone intermediate 18. An attempt to add 1propenyllithium to 15a resulted in an analogous result, giving 19 as a mixture of four diastereomers, contrary to our expectation that intermediary alkoxy enone 20 would be less susceptible to 1,4-addition due to the presence of the β-methyl substituent. Upon exposure to 10 equivalent of lithium (trimethylsilyl)acetylide, 15a was converted quantitatively into monoaddition product 21, but its transformation to 22 as a precursor of 16 was unsuccessful. These fruitless attempts to obtain 16 or its equivalents from 15a in one pot obliged us to focus on the alternative stepwise approach (route B, Scheme 1). Our first task in route B was the monoaddition of vinyllithium to the bicyclic lactone 15a to obtain 23a (Scheme
Scheme 2. Preparation of Bicyclic Lactone 15a
commenced with the preparation of the β,γ-unsaturated aldehyde 8 (see Scheme 1) needed for the Evans asymmetric aldol reaction with 10. All attempts to obtain 8 by oxidation of 3-methyl-3-buten-1-ol were, however, unsuccessful, resulting in the formation of the conjugated aldehyde 3-methyl-2-butenal, which made us utilize dibromo alcohol 117 as the oxidation substrate. The TEMPO-mediated oxidation of 11 gave an unstable dibromo aldehyde, which was immediately subjected to the Evans aldol reaction with the chiral oxazolidinone 10. Exposure of the resulting aldol adduct to acidic conditions (pTsOH, benzene) effected lactone ring formation, and zinc reduction of the vic-dibromo moiety of the product furnished 12 in an acceptable yield of 51% from 11. The ratio of 12 and its C3 epimer in the crude product mixture was 11:1, the latter 3889
DOI: 10.1021/acs.orglett.8b01467 Org. Lett. 2018, 20, 3888−3891
Letter
Organic Letters
diastereomeric mixture was then subjected to ring-closing metathesis8 conditions followed by treatment with silica gel to afford a 5:1 mixture of 24a and its ring-opened form 25a. Reaction of the mixture with an excess amount of vinyllithium proceeded with a high degree of diastereoselection (dr 17:1), furnishing desired diastereomer 26a preferentially. The addition reaction, however, required a considerably longer reaction time (6 h), and disappointingly, the overall yield (17%) of 26a from 15a via the three steps was far from satisfactory, which prompted us to seek a better way to diastereoselectively install the C8 stereocenter. We envisaged the utilization of a protecting group (R) in 25 that is capable of forming a chelate with the lithium metal of vinyllithium, thereby directing the attack of the reagent to the bottom face of the cyclohexenone ring. As a promising protecting group for that purpose, we chose 2-methoxyethyl and set about the preparation of new intermediate 25b, which began with the intramolecular Claisentype cyclization of 14 followed by trapping of the resulting hemiacetalic alkoxide intermediate with 2-methoxyethyl triflate to afford 15b. This bicyclic lactone was converted into a 3:1 mixture of 24b and 25b by the same two-step sequence as used for the transformation of 15a to the mixture of 24a and 25a.9 To our delight, exposure of the 3:1 mixture to an excess amount of vinyllithium (5 equiv) in THF furnished 26b with excellent stereoselection (dr 27:1) in an acceptable overall yield of 54% from 15b within 30 min; the undesired C8 epimer was readily removed by silica gel column chromatography. The relative configuration of 26b was determined by NOE experiments. The significant enhancement in diastereoselectivity and reaction rate observed in the reaction of 24b/25b might support our aforementioned presumption that the 2-methoxyethoxy moiety of 25b could serve as a handle to direct the addition of vinyllithium. The final stage of our total synthesis of anthecularin (1) in an enantioselective manner is shown in Scheme 5. Treatment of
Scheme 3. Examination of Double Vinyl Addition to 15
4). After considerable experimentation, we found that this transformation could be effected cleanly by exposing 15a to 1 equivalent of the lithium reagent in ether (instead of THF) at 0 °C. The cyclic hemiacetal 23a produced as a ca. 3:1 Scheme 4. Diastereoselective Installation of the C8 Stereocenter
Scheme 5. Endgame of the Synthesis of (+)-Anthecularin (1)
the diol 26b with MsCl in pyridine in the presence of a catalytic amount of DMAP at 60 °C brought about selective mesylation of the hydroxyl group at the C4 position and subsequent intramolecular etherification to provide 27 in 94% yield. Lewis acid-mediated removal of the 2-methoxyethyl protecting group of 27 was followed by TPAP−NMO oxidation of the resulting cyclic hemiacetal intermediate, affording lactone 28. Finally, subjection of 28 to RCM conditions furnished 1, the 1H and 13 C NMR spectra of which were in good accordance with those of the natural product.3 The specific rotation of the synthetic material [[α]26D +24.0 (c 0.28, CHCl3)] was virtually identical 3890
DOI: 10.1021/acs.orglett.8b01467 Org. Lett. 2018, 20, 3888−3891
Letter
Organic Letters to that reported for the natural product [[α]26D +23.9 (c 0.15, CHCl3)], which enabled us to determine the absolute configuration of anthecularin as that depicted in Figure 1. This absolute configuration of (+)-anthecularin (1) was consistent with that suggested by Hodgson and co-workers based on their synthetic studies on putative biosynthetic precursors of 1.10 In conclusion, we have accomplished the first enantioselective total synthesis of (+)-anthecularin (1) in 3.9% overall yield from 3,4-dibromo-3-methyl-1-butanol (11) by an 18-step sequence that involves (1) the use of 3,4-dibromo-3methylbutanal as a substitute for unstable 3-methyl-3-butenal in the Evans asymmetric aldol reaction leading to the γmethallyl-γ-butyrolactone 12; (2) intramolecular Claisen-type cyclization that enabled cyclic hemiacetal formation, as well as concomitant establishment of the C2 quaternary asymmetric center (14 → 15b); and (3) highly stereocontrolled addition of vinyllithium to the lactone 25b to install the C8 tetrasubstituted stereocenter of 26b with excellent diastereoselection. The absolute configuration of anthecularin was determined to be 2R,3R,4S,8R by comparing the specific rotation of the synthetic material with that of the natural product.
■
(3) Karioti, A.; Skaltsa, H.; Linden, A.; Perozzo, R.; Brun, R.; Tasdemir, D. J. Org. Chem. 2007, 72, 8103−8106. (4) Li, Y.; Nawrat, C. C.; Pattenden, G.; Winne, J. M. Org. Biomol. Chem. 2009, 7, 639−640. (5) (a) Scholl, M.; Trnka, T. M.; Morgan, J. P.; Grubbs, R. H. Tetrahedron Lett. 1999, 40, 2247−2250. (b) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953−956. (6) Hajra, S.; Giri, A. K.; Karmakar, A.; Khatua, S. Chem. Commun. 2007, 2408−2410. (7) Cane, D. E.; Weiner, S. W. Can. J. Chem. 1994, 72, 118−127. (8) Chatterjee, A. K.; Grubbs, R. H. Org. Lett. 1999, 1, 1751−1753. (9) A small amount of a mixture composed of a tricyclic hemiacetal and the corresponding bicyclic hydroxyl ketone produced via RCM between the C4 methallyl and the C8 vinyl groups of 23b was also obtained. (10) For details, see: (a) Hodgson, D. M.; Talbot, E. P. A.; Clark, B. P. Org. Lett. 2011, 13, 5751−5753. (b) Hodgson, D. M.; Talbot, E. P. A.; Clark, B. P. Chem. Commun. 2012, 48, 6349−6350.
■
NOTE ADDED AFTER ASAP PUBLICATION Reference 10 and the associated citation in the text were added June 8, 2018.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.8b01467. Experimental procedures, characterization data, and copies of NMR spectra for new compounds (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Hidenori Watanabe: 0000-0003-1031-7805 Present Addresses †
(Y.O.) Laboratory of Applied Bioorganic Chemistry, Graduate School of Agricultural Science, Tohoku University, 468-1 Aramaki-Aza-Aoba, Aoba-ku, Sendai 980-0845, Japan. ‡ (K.I.) Department of Chemistry for Life Sciences and Agriculture, Tokyo University of Agriculture, 1-1-1 Sakuragaoka, Setagaya-ku, Tokyo 156-8502, Japan. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are grateful to Dr. Kazuo Furihata (The University of Tokyo) for 2D-NMR analysis and to Dr. Baldip Kang and Dr. Anna Goldys (University of Oxford) for their help with manuscript preparation.
■
REFERENCES
(1) World Health Organization. World Report 2017; WHO: Geneva, 2017. http://www.who.int/malaria/publications/world-malaria-report2017/report/en/ (accessed May 4, 2018). (2) (a) Miotto, O.; Amato, R.; Ashley, E. A.; MacInnis, B.; AlmagroGarcia, J.; Amaratunga, C.; Lim, P.; Mead, D.; Oyola, S. O.; Dhorda, M.; et al. Nat. Genet. 2015, 47, 226−234. (b) Blasco, B.; Leroy, D.; Fidock, D. A. Nat. Med. 2017, 23, 917−928. 3891
DOI: 10.1021/acs.orglett.8b01467 Org. Lett. 2018, 20, 3888−3891
