The AB-Spiroacetal Subunit

Previous work by our group had demonstrated that high levels of 1,4-syn diastereosclection were possible in boron-mediated aldol reactions of 

At the outset of our work on this section of the spongistatin structure.15 it was not known whether an internal acetalisation procedure, analogous to that employed for the AR subunit, would lead to the desired spiroacctal 22a rather than the potentially more stable 22b. However, the axial Cm hydroxyl may hydrogen bond to an acetal oxygen, providing structure 22a additional stability not available to 22b, and this might be exploited in our synthetic endeavours. 

A SURPRISING RESULT REMOTE I.5-AN77 STEREO]NDUCTtON 

Treatment of 34 with aqueous HF in MeCN solution led to desilylation with concomitant acetal formation favouring the undesired spiroacetal 35 over 36 (ca. 5 1). This was not a welcome result At this juncture, we surveyed some equilibration conditions to yield more favourable quantities of the desired spiroacetal isomer 36. Bearing in mind the possibility of an intramolecular hydrogen bond between the axial C25 hydroxyl and an acetal oxygen in 36, which is not possible in 35, we rapidly established that acid catalysed (HC1) equilibration in an aprotic solvent (CDCI3, CH2CI2, Et20) provided ca. 1 1 mixtures of 35 and 36. 

Separation of these spiroacetal isomers was initially achieved using normal phase semi-preparative HPLC, which allowed re-equilibration of the undesired spiroacetal 35 and thereby a method to convert essentially all of the material to the desired isomer 36 after several cycles. The scale-up of this procedure was greatly facilitated by the determination of a simple method for separating the spiroacetal isomers by flash column 

---

Paterson, I., Oballa, R.M., and Norcross, R.D. (1996) Studies in marine macrolide synthesis Stereocontrolled synthesis of the AB-spiroacetal subunit of spongistatin 1 (altohyrtin A). Tetrahedron Lett., 37, 8581-8584. 

---