Communication
isomers. Olefin metathesis with vinylic boronic ester 18 has
been reported to occur with much higher selectivity[30] and
was therefore explored. We were pleased to find that subject-
ing alkene 8 and vinylic boronic ester 18 to the identical cross
metathesis conditions, yielded 20 as a single geometric isomer
(0.4 mmol scale, 68% yield). However, upon scale up we en-
countered two major problems: i) a dramatic reduction in con-
version (10% after 14 h, 3.8 mmol scale), and; ii) the formation
of 1,2-disubstituted alkene 21 (15%). The latter observation
has been described previously,[30] possibly due to the transposi-
tion of boron from the internal to the terminal position of
alkene 18 and subsequent metathesis with 8. As this product
was only observed by GC-MS after extended reaction times
(over 10 h), it was attributed to a transmuted catalyst of 19
causing the isomerization of 18.[31] Therefore, it was imperative
to increase conversion over a short reaction time to avoid cata-
lyst degradation. Through running the reaction at higher con-
centration (1.0m) and adding the catalyst portion-wise (5+
5 mol%), we increased conversion to 45% and reduced the
amount of 21 formed (5%). Finally, periodic degassing of the
reaction every two hours removed the ethylene content of the
solution and further pushed the equilibrium towards vinyl bor-
onate 20, achieving a 60% yield on a 5.3 mmol scale over 10 h
with minimal formation of alkene 21 (<5%).
20 in 66% over three steps. Once again, these three steps
could also be carried out sequentially, without intermediate
purification (“one-pot”), in an increased 86% yield. As a result,
significant amounts (>900 mg) of 3 was obtained over eight
steps from (R)-3-hydroxybutyrate in 38% overall yield.
Completion of the synthesis followed literature precedent
(Scheme 6).[12,15] Selective deprotection of the primary silyl
ether with TBAF (85%), followed by a two-step TEMPO/Pinnick
oxidation, yielded acid 23 in 81%. Lactonisation of the 12-
membered core proceeded efficiently under Yamaguchi condi-
tions (81% yield), and subsequent global deprotection with
HF·pyridine gave the mycolactone core 2 in 80%. In forming
the lactone ring, minor diastereomers observed in the forma-
tion of 3 were separated completing the synthesis of lactone
2, which was identical in all respects to the literature, in a total
of 13 steps and 17% overall yield.
Scheme 6. Completion of synthesis. a) TBAF, THF, 85%; b) i. TEMPO
(15 mol%), BAIB, CH2Cl2/H2O (2:1); then b) ii. NaClO2, 2-methyl-2-butene,
Na2H2PO4 buffer/tBuOH (2:1), 81%; c) (C6H2Cl3)COCl, DMAP, PhH, 81%;
d) HF·pyridine, THF, 80%.
In conclusion, the shortest synthesis of the mycolactone
core to date has been completed both in terms of longest
linear sequence (13 vs 14[14] steps) and total step count (17 vs
28[14] steps). Moreover if the sequenced iterative homologation
is counted as one step, then the mycolactone core is achieved
in only 11 steps. Although a scalable route has already been
accomplished, our synthesis is able to rapidly deliver significant
amounts (>100 mg) of highly enantio- and diastereoenriched
mycolactone core through utilization of simple carbamate
building blocks. Both in terms of step count and scale, the syn-
thesis showcases the power of lithiation–borylation methodol-
ogy for the efficient and convergent synthesis of complex mol-
ecules.
Scheme 5. Olefin metathesis of 8.
With our two key building blocks in hand, we examined our
final iterative lithiation–borylation process. Matteson one-
carbon homologation of 20 proceeded in near quantitative
yield and homologation of 7 with lithiated carbamate 6
(1.5 equiv) gave our required fragment 4 in 81% isolated yield.
Oxidation and NMR analysis showed it to be 97:3 d.r., consis-
tent with the homologation of 6 with analogous allylic boro-
nate 10. For the final step, lithiation of 5 in the presence of
(À)-sparteine was required, but in explorative lithiation–deuter-
ation experiments we isolated diene 22[32] in 10% yield, in ad-
dition to the required deuterated product (90%). This showed
that 10% lithiation of 5 occurred at the allylic position fol-
lowed by E2 elimination of the carbamate. We therefore used
an excess of carbamate 5 (1.5 equiv) with respect to boronic
ester 4 (1.0 equiv), and the final homologation and subsequent
oxidation gave known intermediate 3 in 82% yield and high
d.r.[33] with 950 mg prepared. With isolation and chromato-
graphic purification of each intermediate, 3 was formed from
Acknowledgements
We thank EPSRC (EP/I038071/1) and the European Research
Council (FP7/2007-2013, ERC grant no. 246785) for financial
support. We thank Dr. Eddie Myers and Dr. Daniele Leonori for
helpful suggestions.
Keywords: asymmetric synthesis
·
boron
·
iterative
homologation · mycolactone · natural products
Chem. Eur. J. 2015, 21, 13900 – 13903
13902
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim