|Home | About | Journals | Submit | Contact Us | Français|
Recently Uemura and co-workers described the structure of aburatubolactam A (1, Figure 1), a macrolactam isolated from the culture broth of a Streptomyces sp. bacterium, SCRC-A20, that was separated from a marine mollusk collected near Aburatubo, Kanagawa Prefecture, Japan.[1a] Aburatubolactam A is a member of a growing class of tetramic acid-containing macrolactams including cylindramide A,[1b] geodin A,[1c] xanthobaccin A,[1d] ikarugamycin,[1e] discodermide,[1f] and the alteramides.[1g] These mixed polyketide-amino acid metabolites have been isolated from a number of sources including sponges, marine bacteria and terrestrial bacteria, and display a diverse range of biological activities including cytotoxicity, anti-microbial activity, and inhibition of superoxide generation. In this Communication we describe a synthesis of aburatubolactam A.
Our strategy for the synthesis of aburatubolactam A is based on the coupling of two domains: a subunit containing the bicyclo[3.3.0]octane (2), and a 3-hydroxyornithine-derived subunit (3, Figure 1). The bicyclo[3.3.0]octane ring system was ultimately envisioned to arise from a ring-opening—ring-closing metathesis of functionalized bicyclo[2.2.1]heptene 4. [3,4]
The synthesis commenced with a Diels-Alder reaction of commercially available ketone 6 with cyclopentadiene in the presence of 20 mol % of MacMillan's catalyst 8 to give ketone 7 in 65% yield (endo:exo >98:2, 93% ee, Scheme 1). Conversion to the enone 4 was readily achieved in 80% yield by Saegusa oxidation of the trimethylsilyl enol ether derived from 7. When enone 4 was subjected to 2.5 mol% of Grubbs’ catalyst 9 under an ethylene atmosphere, rapid and smooth reorganization to the desired bicyclo[3.3.0]octene 5 occurred in 90% yield.
Further elaboration of 5 was accomplished by reduction of both alkenes (Pd/C, H2) to give fused bicyclic ketone 10 in 94% yield (Scheme 2). Introduction of the C6 and C13 side chains was achieved by a sequence beginning with enolate acylation with Mander's reagent, followed by reduction of the ketone with NaBH4. Elimination of the resultant alcohol by mesylation and treatment with sodium hydride in MeOH-THF (5:1) provided 11 in 64% overall yield from 10. The C13 side chain was introduced by employing Majetich's fluoride-mediated Sakurai allylation9 in DMF-DMPU to give 12 in 78% yield as a 4:1 mixture of inseparable C6 diastereomers favoring the undesired stereochemistry (viz 12a). This ratio could be improved to 2:1 in favor of the desired stereochemistry (viz 12b) by protonation of the silylketene acetal derived from 12a with HCl. Subsequent iodolactonization facilitated separation of the diastereoisomers, and gave 13 in 58% yield for the two steps, along with recovered 12a. This sequence also provided a mechanism for recycling material. Treatment of iodolactone 13 with Zn dust in AcOH/EtOH, followed by reduction of the acid with LAH provided alcohol 14 in 86% yield (2 steps).
Advancement of alcohol 14 to 2 involved cross metathesis of butene-1,4-diol derivative 16 catalyzed by 15 to give 13 in 95% yield (Scheme 3). Oxidation with Dess-Martin periodinane, followed by olefination with (iodomethylene)triphenylphosphorane under Stork-Zhao conditions provided vinyl iodide 19 in 82% yield. Conversion of the iodide to the stannane by treatment with tert-BuLi in the presence of tributyltin chloride also resulted in removal of the pivalate to give alcohol 20 in 85% yield. Treatment of this alcohol with Dess-Martin periodinane and subsequent Horner-Wadsworth-Emmons reaction yielded stannyl dioxenone 2 in 60% overall yield (2 steps).
The synthesis of the β-hydroxyornithine subunit 3 began with Sharpless asymmetric dihydroxylation of α,β-unsaturated ester 21 to give diol 22 in 90% yield and >98% ee (Scheme 4). Introduction of the nitrogen was achieved by cyclic sulfite formation and opening with sodium azide. Subsequent silylation provided ether 23 in 80% yield over the three steps. Reduction of the azide and nosylation led to 24 in 94% yield, and was followed by introduction of the methyl group by Mitsunobu reaction to give 25, and removal of the nosyl group with thiophenoxide provided amine 3 (82% from 24).
After exploring a number of unsuccessful end game strategies that paralleled those employed for cylindramide, the synthesis was completed as shown in Scheme 5. Coupling of the two halves of the molecule was achieved by heating dioxenone 2 with amine 3 in toluene under reflux for 6 hours (Scheme 5). Subjection of the sensitive β-ketoamide product to Stille coupling with tert-butyl-β-iodoacrylate, followed by Lacey-Dieckmann cyclization, led to tetramic acid 26 in 50% yield (over three steps from 2). Macrocyclization was achieved by simultaneous removal of the Boc carbamate and tert-butyl ester with TFA and treatment of the resulting compound with DEPC[ and Et3N in DMF for 12 hours. Removal of the TBS group with HF provided aburatubolactam A in 46% yield for the 3 steps. Data for an analytical sample of synthetic aburatubolactam A obtained by semi-preparative HPLC matched that obtained for an authentic sample provided by Professor Daisuke Uemura.
In conclusion, we have described a 23 step route that leads to aburatubolactam A and that further highlights the utility of tandem metathesis reactions in a target oriented setting.
**This research was supported by the National Cancer Institute (NCI CA110246). We thank Professor Daisuke Uemura (Nagoya University) for providing a sample of aburatubolactam A.
James A. Henderson, Department of Chemistry and Biochemistry University of Colorado Boulder, CO 80309-0215, USA.
Andrew J. Phillips, Department of Chemistry and Biochemistry University of Colorado Boulder, CO 80309-0215, USA Fax: (+)1 303 4920439 ; Email: ude.odaroloc@spillihP.werdnA.