|Home | About | Journals | Submit | Contact Us | Français|
This manuscript describes a convergent synthesis and the revision of the relative stereochemistry of nakiterpiosin, a marine C-nor-D-homosteroid. Our synthesis features a late-stage carbonylative Stille cross-coupling reaction and a photo-Nazarov cyclization reaction that deliver the complete nakiterpiosin skeleton efficiently.
Nakiterpiosin (1) is a marine sponge metabolite that exhibits potent cytotoxicity against the P388 murine leukemia cell line (GI50 10 ng/mL) (Figure 1).1 It was the first C-nor-D-homosteroid isolated from a marine source. Its unique chemical structure and strong P388 growth inhibition property prompted us to initiate a research program to explore its laboratory synthesis2 and biological function. We report herein the synthesis and structure revision3 of nakiterpiosin.
The C-nor-D-homosteroids are skeletally rearranged steroids with their C-ring contracted and D-ring expanded by one carbon. The veratrum alkaloid cyclopamine (3) and veratramine (4) are arguably the best known members.4,5 The teratogenic alkaloid 3 inhibits the Hedgehog (Hh) signaling by binding to Smoothened (Smo),6,7 and the chronotropic alkaloid 4 induces serotonin (5-HT) syndrome.8 While 1 possesses potent cytotoxicity against P388, its molecular target is not known. Furthermore, the complete biological profile of 1 could not be obtained due to the scarcity of the material. From 30 kg of sponge T. hoshinota, only 0.4 mg of nakiterpiosin was obtained. Its chemical structure was assigned as 2 in the original reports.1
We were puzzled by the inconsistency between the C-20 stereochemistry reported for 2 and 3/4 and therefore set out to probe the relative stereochemistry of nakiterpiosin. Our model studies indicated potential misassignment of the C-6, C-20 and C-25 stereogenic centers.9 We next considered the biogenesis of the halogen atoms10 of nakiterpiosin to rationalize the C-6 and C-20 stereochemistry. We envisioned that the C-21 chlorine atoms of nakiterpiosin might be introduced by radical chlorination and the C-6 bromine atom by bromoetherification (as shown in 5) to result in retention of the C-20 configuration and the anti C-5,6 bromohydrin stereochemistry. Taken together, these considerations led us to propose 1 to be the correct structure of nakiterpiosin. Indeed, we found that the 1H and 13C NMR spectra of our synthetic sample of 1 agreed with those of the natural product.11 In contrast, those of synthetic 29,11 and the natural product are significantly different. We thus revised the relative stereochemistry of nakiterpiosin to be that indicated in 1, which shares the same configuration at the C-20 and C-25 positions with 3 and 4.
Our synthetic strategy is outlined in Figure 2. We dissected 1 into fragments 6 and 7 and constructed the central cyclopentanone ring with a carbonylative cross-coupling reaction12 and a photo-Nazarov cyclization reaction.13 The electrophilic coupling component 6 was synthesized by an intramolecular Diels–Alder reaction,14 and the nucleophilic coupling component 7 by vinylogous Mukaiyama aldol reaction.15
To synthesize 6, we first converted acid 816 to the corresponding Weinreb amide and then set the C-6 stereocenter by Noyori reduction17 (Scheme 1). The in-water hydrogenation protocol18 provided significant enhancement of the reaction rate, allowed low catalyst loading and suppressed the formation of the undesired lactone. Subsequently, isopropenyl Grignard addition to 9 followed by Me2AlCl-promoted intramolecular Diels–Alder reaction gave the exo cycloaddition product. The C-6 hydroxyl group controlled the stereoselectivity19 of this Diels-Alder reaction to afford single diastereomer. The C-6 hydroxyl group was then activated with an electron-deficient aryl sulfonate group for the introduction of the C-6 bromine atom. We found that 10 underwent retro-Diels–Alder reaction readily at elevated temperature, particularly in the presence of Lewis acids. This characteristic imposed significant challenges for the introduction of the C-6 bromine atom. We therefore dihydroxylated the olefin group to prevent the retro-Diels–Alder reaction even through this functionalization created serious steric congestion around the C-6 position. The C-6 bromine atom could then be introduced with inversion of configuration and concomitant acetonide protection. The absolute and relative configurations of 11 were confirmed by X-ray analysis.11 Finally, enol triflate installation completed the synthesis of 6.
For the synthesis of 7, we utilized the Sharpless asymmetric epoxidation reaction20 of 1211 to set the C-20 stereochemistry (Scheme 2). The subsequent TBS protection and Yamamoto epoxide rearrangement reaction21 gave aldehyde 13. We note that aldehyde 13 was sensitive to both acidic and basic conditions and underwent elimination and racemization readily. The enantiomeric purity was eroded after the rearrangement reaction. We also found that Bi(OTf)3 promoted the vinylogous Mukaiyama aldol reaction of 13 with good levels of diastereoselectivity.22 However, further loss of enantiomeric purity could not be avoided. With the complete C-20–C-26 carbon framework of the side-chain in place, we then sought to set its anti–anti–trans configuration. The C-25 stereochemistry was set by a chelation-controlled hydrogenation of 14 using Crabtree's catalyst.23,24 We then inverted the C-22 stereocenter of 15 and protected the hydroxyl group as TBS ether to give 16 with the desired anti–anti–trans configuration. Selective deprotection of the primary TBS ether followed by Dess-Martin oxidation yielded aldehyde 17. The gem-dichloromethyl group was introduced by Cl2/P(OPh)3 to give 18.25 No epimerization of the C-20 stereocenter was observed. The X-ray analysis on des-TBS-18 confirmed the absolute and relative configurations of 18.11 Finally, the palladium-catalyzed stannylation afforded the nucleophilic coupling component 7.
The carbonylative coupling of 6 and 7 was achieved with a modified Stille's protocol26 using Pd(PPh3)4/CuCl in DMSO under 1 atm CO. The CuCl additive and DMSO solvent provided dramatic rate enhancement27 and were crucial to the success of this reaction. It is worth noting that 6 and 19 were highly sensitive to both acidic and basic reaction conditions. Addition of LiCl and prolonged heating led to the elimination of bromide. Enone 19 was obtained as a 4:1 mixture of inseparable diastereomers due to the diminished enantiomeric purity of 7. Nonetheless, the minor diastereomer could be removed at a later stage (20).
With 19 in hand, we next explored the key Nazarov cyclization reaction to complete the construction of the C-nor-D-homosteroid skeleton of nakiterpiosin. Remarkably, irradiation of a solution of 19 in acetonitrile at 350 nm smoothly delivered the desired annulation product as a 1:1 mixture of C-9 diastereomers, which converged to 20 upon treating with diisopropylamine in methanol. The photo-Nazarov cyclization reaction of aryl vinyl ketones was first reported by Smith and Agosta.28 Subsequent mechanistic studies by Leitich and Schaffner revealed the reaction mechanism to be thermal electrocyclization induced by photolytic enone isomerization.29 The mildness of the reaction conditions and the selective activation of the enone functional group allowed the facile transformation of the densely functionalized 19 to 20. It should be noted that the Lewis acid-promoted Nazarov cyclization of aryl vinyl ketones normally requires much harsher reaction conditions or activated substrates.30 Indeed, exposure of the model systems of 19 with simplified side-chains to various Lewis acids only resulted in substrate decomposition.
To complete the synthesis of 1, we first removed the acetonide protecting group of 20 and cleaved the diol to afford the corresponding bis-hemiacetal. Selective reduction of the less hindered hemiacetal followed by TBS deprotection31 furnished 1. The spectroscopic data of 1 is fully consistent with that of the natural sample.11 In contrast, synthetic 2, which was obtained by a similar approach,9,11 exhibits significantly different 1H and 13C NMR spectra. Notably, synthetic 2 existed as an equilibrium mixture of the C-4 hemiacetal and aldehyde forms.11 We have also compared the chemical shifts of the Mosher esters of synthetic 1 with those reported for the natural product and found that they were in agreement.11,32
In conclusion, the structure of nakiterpiosin is revised as 1. The newly assigned structure is supported by total synthesis and biogenesis rationale. Our synthetic approach is highly convergent and allows for the synthesis of the analogs and derivatives of 1 for further biological studies. We are investigating the biological function and molecular target of 1. These results will be reported in due course.
NMR spectroscopic data of synthetic 1 and 2, CD spectrum of synthetic 1, synthetic route for 2, experimental procedures, crystallographic data, and characterization data. This material is available free of charge via the Internet at http://pubs.acs.org.
Financial supports are provided by NIH (Grant NIGMS R01-GM079554), Welch Foundation (I-1596), and UT Southwestern. C.C. is a Southwestern Medical Foundation Scholar in Biomedical Research. We thank Prof. John MacMillan and Dr. Ana Paula Espindola for assistance in NMR experiments, and Dr. Vincent Lynch (UT Austin) and Dr. Radha Akella for X-ray analysis. We also thank Prof. Daisuke Uemura (Keio University) for providing a copy of the 1H NMR spectrum of the natural sample.