S. Canesi et al.
the analogous allylic interaction is considerably more severe
in that it involves compression of a more sterically demand-
ing Me group against the same vinylic hydrogen. In all like-
lihood, this retards the Prins reaction step and diverts the
reactive electrophile created upon umpolung activation of
the phenol towards other reaction pathways. In either case,
the product of the reaction is cis-ketone 18.[14]
Consistent with the aforementioned rationale, tetrahydro-
furan 20 reacted with DIB, via intermediates 21 and 22, to
furnish compound 23 in considerably higher yield (70%,
Scheme 7). Compound 20 emerged in 88% yield as a single
Scheme 8. The structure of (À)-platensimycin and retrosynthetic logic.
PCC=pyridinium chlorochromate.
droperoxyketal 25 in 64% overall yield. This material was
immediately treated with FeSO4 and CuACTHUNTRGNEU(GN OAc)2 in MeOH to
give 26.
Scheme 9.
Scheme 7. TFA=trifluoroacetic acid; DMP=Dess–Martin periodinane.
Addition of K2CO3 to the reaction mixture induced con-
version into 27 in 44% overall yield from 25. Compound 27
was obtained as a 3:1 mixture of exocyclic (major) and en-
docyclic alkenes. As first demonstrated by Nicolaou and co-
workers,[8] both isomers converge to the tetracyclic core of
platensimycin (32). Accordingly, no separation was required.
The mixture of alcohols was thus advanced to a mixture of
the corresponding aldehydes (28), in 74% yield, by PCC ox-
idation. Aldehydes 28 are known synthetic intermediates for
(À)-platensimycin.[8] Hence, they were subjected to the
action of Kaganꢄs reagent (SmI2),[19] whereupon stereoselec-
tive cyclization to alcohols 29 occurred (Scheme 10). Finally,
the olefin regioisomers of 29 converged to the same final
product 31 upon treatment with TFA. The elaboration of 31
into (À)-platensimycin is well documented in the litera-
ture.[8] Therefore, the synthesis of 31 represents a formal
total synthesis of (À)-platensimycin.
diastereomer (presumably the thermodynamically more fa-
vorable one) upon acid treatment of triols 19, which, in
turn, were assembled by using Evans asymmetric alkylation
technology.[10] We found it convenient to subject crude 23 to
Dess–Martinꢄs oxidation[16] and isolate and characterize ke-
toaldehyde 24, which emerged in 60% overall yield as a
single diastereomer. As anticipated, based upon our mecha-
nistic hypothesis, compound 24 displayed a cis relative con-
figuration of the carbonyl branches. The reaction had creat-
ed contiguous tertiary and quaternary carbon centers with
complete stereocontrol.
As an initial application of the oxidative Prins–pinacol
tandem sequence, we now describe a formal total synthesis
of (À)-platensimycin, 32 (Scheme 8). This substance is an
exciting experimental antibiotic that is believed to act as a
FabF inhibitor.[17a] Its unusual structure and potent bioactiv-
ity have elicited enormous interest within the synthetic com-
munity.[17] Our approach to 32 relies upon the assumption
that one may reach hydroperoxide 25 by intercepting oxoni-
um ion 22 with H2O2. Subsequent Schreiber–Fenton frag-
mentation[18] would provide alkenes 27, which may be elabo-
rated into advanced intermediate 31 as described by Nico-
laou and co-workers.[8]
In conclusion, our continuing investigations centering
upon the concept of aromatic ring umpolung have now pro-
duced an unprecedented oxidative Prins–Pinacol tandem
process. This method allows rapid, stereocontrolled elabora-
The hypothesis was put into practice as shown in
Scheme 9. Oxidation of compound 20 with DIB, as detailed
earlier, followed by the addition of 30% hydrogen peroxide,
afforded a 3:2 mixture of unassigned diastereomers of hy-
Scheme 10. HMPA=hexamethylphosphoramide.
11226
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11224 – 11228