Degradation of 1 to 7. The degradation pathway from 1 to 7 via 2-6 outlined in Scheme I6 is patterned after that developed in our laboratory in conjunction with the synthesis of amphoterolide B. 7· 8 Two comments are appropriate.(1) The normally problematic step of cleaving the mycosamyl moiety (step b) proceeded wellthrough oxidative deglycosidation of 2 with DDQ7a to provide tetraenone 3, andthis method appears generally ap-plicable to manyother polyenemacrolides.(2) Reduction of 3 with NaBH4 (step c) 8a led to the exclusive formation of a single tetraenol (4). The 15/? configuration was assigned to 4 through the observation of a negative Cotton effect in the CD spectrum of the p-nitrobenzoate derivative 4a. 9· 10 That boththe MOM ether of 4 (Jl415 2.8,/14. il5 8.0,/15> 16 7.0 Hz) and 2 (/1415 2.5, J 14^ 15 8.0,/l516 7.5 Hz) show very similar coupling patterns for

H14, H14/, H15, and Hl6 confirms the 15R assignment of l. 10 Synthesis of 7. NMR spectral comparison between the pimarolide derivative 4 and the amphoterolide B methyl ester as well as between a pair of their respective degradation products strongly suggests that the pyranmoieties [C (9)-C (13) in 1] of both antibiotics possess the same stereochemistry. Thus, a synthetic intermediate (8) representing the C (9)-C (15) fragment and used in our amphoterolide B synthesis76· 11 served as starting material and was converted into aldehyde 12 via 9-11 as shown in Scheme II. 12" 14 Two different reagent-controlled reactions were used to preparethe two possible configurations at C (7) in 7.(1) The asymmetric aldol reaction of aldehyde 12 with the enolate derived from 3-ethylpentyl ethanethioate and chiral (/?,/?)-di-methylborolanyl trifiate15 provided a 1: 8 mixture of 13a and 13b, while the use of (5', S)-dimethylborolanyl trifiate reversed the product ratio (10: 1 of 13a and 13b). Silylation of 13a and 13b followed by NaBH4 reduction16 and oxidation afforded the al-