Portal de Periódicos da CAPES

Stereocontrolled Total Synthesis of (−)‐Aurisides A and B

2005; Wiley; Volume: 44; Issue: 7 Linguagem: Inglês

10.1002/anie.200462267

1521-3773

Ian Paterson, Gordon J. Florence, Annekatrin C. Heimann, Angela Mackay,

Marine Sponges and Natural Products

An expedient total synthesis of aurisides A and B (1), unusual cytotoxic macrolide glycosides isolated from the Japanese sea hare D. auricularia, takes advantage of a highly convergent aldol-based route for the stereocontrolled construction of the common macrolide core. This is followed by α-selective glycosylation to introduce the sugar moieties. Aurisides A (1) and B (2) are unique marine polyketides isolated in 1996 by Yamada and co-workers from the Japanese sea hare Dolabella auricularia,1a an organism that has proved to be a rich source of bioactive secondary metabolites.2 Initial biological screening of 1 and 2 highlighted significant cytotoxicity, with IC50 values against HeLa S3 cervical cancer cell lines of 0.17 and 1.2 μg mL−1, respectively. The aurisides are 14-membered glycosylated macrolides that contain a six-membered hemiacetal ring, an E-trisubstituted enone with an E,E bromodiene side chain appended at C13, and different sugar moieties attached at C5 (Scheme 1). The unusual structure of the aurisides, combined with their biological activity and low natural abundance (0.8 mg of 1 was obtained from 278 kg of D. auricularia), has generated growing interest in these compounds as synthetic targets.1b, 3 In 1998, the Yamada group reported a synthesis of the aglycon unit, confirming their initial stereochemical assignment. The low yield of their route was the result of several problematic steps in its late stages.1b Herein, we report the first total synthesis of aurisides A (1) and B (2) by appropriate attachment of the required sugar residue, which involves a highly convergent and expedient aldol-based route for the stereocontrolled construction of the common macrolide core. As outlined in Scheme 1, our synthetic strategy relied on a late-stage, α-selective glycosylation of the equatorial C5 alcohol in lactone 5 with the fluorosugar 3 or 4, each derived from L-rhamnose. The macrocyclic lactone, in turn, was envisaged to arise from a stereocontrolled Mukaiyama aldol coupling between aldehyde 6 (C1–C7) and silyl enol ether 7 (C8–C17) containing a bromodiene terminus, followed by a suitable macrolactonization step. Introduction of the remote C13 stereocenter in 7 was planned to rely on the application of an asymmetric vinylogous Mukaiyama (AVM) aldol reaction,4 while the C5 center in 6 would also be installed by a suitable aldol reaction.5 Retrosynthetic analysis for the aurisides. As shown in Scheme 2, the synthesis of the C1–C7 subunit 6 began with a highly stereoselective boron-mediated aldol reaction of the readily available methyl ketone 86 with 3-butenal (9), derived from the oxidative cleavage of 1,2-glycol 10.7 Enolization of 8 with (+)-Ipc2BCl/Et3N,6, 8 followed by the addition of a freshly prepared anhydrous solution of 9 at −78 °C, provided the corresponding 1,4-syn aldol adduct 11 (94 %, >97:3 d.r.). Treatment of 11 with PMBTCA in the presence of catalytic TfOH in Et2O at room temperature afforded the PMB ether 12 in 85 % yield.9 Alternatively, use of Sc(OTf)3 in toluene at 0 °C provided 12 in 76 % yield on a multigram scale,10 with decreased by-product formation. Subsequent ozonolysis of 12 with reductive PPh3 workup gave the 1,5-ketoaldehyde 6 in 97 % yield. Synthesis of the C1–C7 subunit 6. a) NaIO4, CH2Cl2, pH 4 buffer, 0 °C, 3 h; b) 1. (+)-Ipc2BCl, Et3N, Et2O, 0 °C, 1 h; 2. 9, CH2Cl2, −78 → −27 °C, 2.5 h; 3. H2O2 (30 % aq), pH 7 buffer, MeOH, 0 °C→RT, 1 h; c) PMBTCA, TfOH (0.3 mol %), Et2O, room temperature, 3 h; d) PMBTCA, Sc(OTf)3 (1 mol %), PhMe, 0 °C, 15 min; e) O3, NaHCO3, CH2Cl2, −78 °C, 10 min; then PPh3, −78 °C→RT, 3 h. Ipc=isopinocampheyl, PMBTCA=para-methoxybenzyltrichloroacetimidate, Tf=trifluoromethanesulfonyl, TIPS=triisopropylsilyl. Synthesis of the C8–C17 subunit 7 (Scheme 3) commenced with the bromination of potassium glutaconaldehyde (13),11 according to the method of Duhamel and co-workers.12 Treatment of 13 with Br2 and PPh3 provided the corresponding E,E bromodienal 14 (68 %). The stage was now set for the critical AVM reaction between 14 and silyl dienolate 1513 to introduce the C13 stereocenter, along with the 10E-trisubstituted alkene functionality.5, 14 Gratifyingly, treatment of aldehyde 14 with [(R)-binol-Ti(OiPr)2] (50 mol %) in THF at −78 °C, generated in situ from (R)-binol and Ti(OiPr)4, followed by addition of silyl dienolate 15 provided the vinylogous aldol adduct 16 exclusively in 89 % yield and 94 % ee. Synthesis of the C8–C17 subunit 7. a) Br2, PPh3, CH2Cl2, 0 °C→RT, 4 h; b) Ti(OiPr)4 (50 mol %), (R)-binol (50 mol %), CaH2, THF, −78 °C, 72 h; c) TBSCl, imidazole, CH2Cl2, room temperature, 4 h; d) DIBAL, CH2Cl2, −78 °C, 2 h; e) MnO2, Et2O, room temperature, 3 h; f) CH2CH(Me)MgBr, THF, −78→0 °C, 1.5 h; g) MnO2, Et2O, room temperature, 5 h; h) L-Selectride, CaH2, THF, −78 °C, 15 min; TMSCl⋅Et3N, −78→−20 °C, 45 min. Binol=1,1′-bi(2-naphthol), TBS=tert-butyldimethylsilyl, DIBAL=diisobutylaluminum hydride, L-Selectride=lithium tri-sec-butylborohydride, TMS=trimethylsilyl. TBS ether formation on the alcohol 16 was followed by conversion into aldehyde 17 by treatment with DIBAL. Subsequent reoxidation with MnO2 proceeded in 87 % yield. Addition of isopropenyl magnesium bromide to 17 and oxidation of the resulting alcohol with MnO2 provided enone 18 (86 %), anticipated as a direct precursor to the silyl enol ether 7. Initial attempts at this transformation employing Chan hydrosilylation,15 LiAlH4/CuI/TMSCl,16 or the Stryker reagent17 proved unsuccessful. However, when 18 was subjected to L-Selectride in THF at −78 °C, regioselective 1,4-reduction of the less sterically encumbered enone was observed. The resultant enolate was quenched with TMSCl to provide the C8–C17 subunit 7, which was used directly in the subsequent coupling step. With key subunits 6 and 7 in hand, attention was focused on their Mukaiyama-type aldol union to introduce the C7 stereocenter, relying on 1,3-anti induction from the C5 ether through an open transition state, following the Evans polar model,18 thus completing the carbon backbone of aglycon 5 (Scheme 4). In practice, exposure of 6 and 7 to BF3⋅Et2O in CH2Cl2 at −95 °C provided the adduct in 66 % yield (from 18) with 95:5 d.r. which was present in solution in the closed hemiacetal form 19. Treatment of 19 with PPTS, (MeO)3CH, and MeOH cleanly provided methyl acetal 20. To confirm the stereochemistry of 20, irradiation of 5-H provided a diagnostic NOE interaction with the C3-OMe, consistent with their 1,3-diaxial relationship, whereas irradiation of the C3-OMe in turn, provided a further NOE interaction with 7-H. Cleavage of the silyl ethers with TASF in wet DMF gave diol 21 (72 %).19 Selective oxidation of the C1 terminus to seco acid 22 was next attempted and required careful optimization. Under the conditions developed by Piancatelli and co-workers,20 treatment of 21 with TEMPO and PhI(OAc)2 in MeCN/pH7 buffer (5:1) provided the intermediate aldehyde, which was further oxidized with NaClO2 to give 22 in 60 % yield. This intermediate readily underwent Yamaguchi macrolactonization to provide the desired 14-membered macrocycle 23 cleanly (86 %).21 Cleavage of the PMB ether with DDQ and hydrolysis of the methyl acetal then afforded auriside aglycon 5 in 74 % yield. At this stage, the 1H and 13C NMR spectroscopic data and specific rotation agreed with those reported by the Yamada group.1b Synthesis of the auriside aglycon 5. a) BF3⋅OEt2, CaH2, CH2Cl2, −95 °C, 10 min; b) PPTS (20 mol %), CH(OMe)3, MeOH, room temperature, 4 h; c) TASF, H2O, DMF, 0→15 °C, 5 h; d) 1. TEMPO, PhI(OAc)2, MeCN, pH 7 buffer, 6 h; 2. NaClO2, NaH2PO4, tBuOH, H2O, 2-methyl-2-butene, room temperature, 30 min; e) 1. 2,4,6-trichlorobenzoyl chloride, Et3N, PhMe, room temperature, 1 h; 2. DMAP, room temperature, 4 h; f) DDQ, CH2Cl2, pH 7 buffer, room temperature, 30 min; g) pTsOH⋅H2O, THF, H2O, room temperature, 16 h. PPTS=pyridinium p-toluenesulfonate, TASF=(diethylamino)sulfur trifluoride, DMF=N,N-dimethylformamide, TEMPO=4-amino-2,2,6,6-tetramethylpiperidine-1-oxyl, DMAP=4-N,N-dimethylaminopyridine, DDQ=2,3-dichloro-5,6-dicyano-1,4-benzoquinone, pTsOH=para-tolunenesulfonic acid. With aglycon 5 in hand, attention was now directed toward the preparation of the activated fluorosugar units 3 and 4 (Scheme 5). Following work reported by the Nicolaou group, a nine-step sequence starting from L-rhamnose was used to access common precursor 24.22, 23 Formation of the disaccharide 3 began with the activation of 24 with DAST/NBS to provide exclusively α-fluorosugar 25 in 60 % yield.24 Mukaiyama glycosylation between 25 and 26,25 obtained by deprotection of 24 with TBAF, provided 27 as the α-anomer exclusively (84 %). Activation of 27 with DAST/NBS then afforded 3 in 79 % yield. The fluorosugar 4 required for auriside B (2) was readily synthesized in 77% yield by treatment of 26 with trichloroacetyl isocyanate to afford carbamate 28 (99 %),26 followed by activation with DAST/NBS. Synthesis of fluorosugar units 3 and 4. a) DAST, NBS, −15 °C, 15 min; b) TBAF, THF, 0 °C→RT, 4 h; c) SnCl2, AgClO4, molecular sieves (4 Å), Et2O, 0 °C→RT, 4 h; d) Cl3CC(O)NCO, CH2Cl2, room temperature, 1 h; K2CO3, MeOH, room temperature, 2 h. DAST=(diethylamino)sulfur trifluoride, NBS=N-bromosuccinimide, TBAF=tetrabutylammonium fluoride. Completion of the total synthesis, as shown in Scheme 6, required the coupling of aglycon 5 with either activated sugar 3 or 4 to directly provide auriside A (following silyl deprotection) and auriside B, respectively. Under the Mukaiyama protocol, the reaction of 3 and 5 followed by desilylation with HF⋅pyr afforded (−)-auriside A (1) in 37 % yield ([α]=−16.3 (c=0.033 in MeOH) Ref. 1a −43.0 (c=0.050 in MeOH)). Similarly, the reaction of 4 and 5 proceeded smoothly to afford (−)-auriside B in 74 % yield ([α]=-21.6 (c=0.10 in MeOH), Ref. 1a −30.0 (c=0.090 in MeOH)). In each case, analytical data (1H, 13C NMR and IR spectroscopy, MS, and specific rotation) for the synthetic material were in excellent agreement with those reported for natural aurisides A and B, allowing confirmation of the relative and absolute configurations of these compounds.27 Completion of the total synthesis of aurisides A (1) and B (2). a) SnCl2, AgClO4, molecular sieves (4 Å), Et2O, 0 °C→RT, 16 h; b) HF⋅py, THF, 0 °C→RT, 16 h; c) SnCl2, AgClO4, molecular sieves (4 Å), Et2O, 0 °C→RT, 8 h. In conclusion, we have completed an expedient total synthesis of (−)-aurisides A and B that proceeds in 18 steps (1.7 % overall yield) and 17 steps (3.5 % overall yield), respectively. By building the E,E bromodiene of the side chain into the silyl enol ether 7, the key Mukaiyama aldol coupling with aldehyde 6 delivers the advanced intermediate 19 in a highly convergent manner. This can then be converted into the aurisides by α-selective glycosylation of the derived macrolide core 5 with the fluorosugars 3 and 4. This work also highlights an efficient enantioselective vinylogous Mukaiyama aldol reaction, which in tandem with our diastereoselective boron-mediated aldol methodology provides a rapid synthetic entry into this structurally unique class of bioactive marine macrolides. Supporting information for this article is available on the WWW under http://www.wiley-vch.de/contents/jc_2002/2005/z462267_s.pdf or from the author. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.