Angewandte
Communications
Chemie
16 to be obtained in well reproducible 82% on a 1.4 g scale
(single largest batch).[35]
With appreciable amounts of 16 in hand, the stage was set
for the key ring-opening/cross-coupling reaction. Because of
À
the presence of the free OH group, an extra equivalent of
MeMgBr had to be used; in practice, however, excess
Grignard reagent was necessary for good conversion. Largely
for the low solubility of the resulting magnesium alkoxide, the
original solvent system (Et2O/toluene)[20] was replaced by
Et2O/cyclopentyl methyl ether. Under these conditions, the
iron-catalyzed reaction proceeded cleanly at À308C to give
the desired Z,E-configured dienoic acid 17 in good yield and
favorable isomeric purity (75% (84% brsm), ꢁ 18:1) when
the reaction was stopped after 3.5 h; no competing attack of
the Grignard reagent onto the ester or lactone groups in 16
was noticed, and the compatibility of the iron catalyst with the
sulfur and nitrogen donor sites is rewarding. A practical
complication, however, arose from the fact that small
amounts of unreacted starting material remained at this
point which proved difficult to separate; actually, analytically
pure samples of 17 could only be obtained by preparative
TLC, which is obviously impractical on scale.
Because longer reaction times necessary for full conver-
sion resulted in stereochemical erosion, the crude material of
the iron-catalyzed pyrone ring opening reaction was directly
subjected to macrolactonization under modified Mukaiyama
conditions.[36] Specifically, the use of 18 as the activating
agent, which is escorted by a non-nucleophilic counterion,
furnished the macrocycle in isomerically pure form, whereas
more conventional macrolactonization protocols resulted in
massive isomerization.[37] At this stage, product 19 could be
purified by flash chromatography and unreacted pyrone 16 be
recovered. Although performed only with 250 mg batches,
this protocol proved practical and well reproducible, and
nothing augurs against applications on significantly larger
scale.
From this point onward, the completion of the synthesis of
DMDA-Pat A (2) followed the known route,[8,13] but not
without notable improvements. Specifically, the oxidation of
the primary alcohol 20 formed on deprotection of 19 with
buffered TBAF proved delicate. In our hands, reproducible
results were obtained when the reaction was performed under
Parikh–Doering conditions[38] and telescoped with the Wittig
olefination, as the intermediate aldehyde is rather sensitive
and epimerization-prone. Single crystals of the resulting
product 21 suitable for X-ray diffraction could be grown,
which revealed, for the first time, the conformation of the
pateamine macrocycle that is supposedly critical for eliciting
the biological response (Figure 1). The large ring adopts an
unusual triangular shape; the thiazole ring as well as the Z,E-
configured dienoate lie perpenticular to its plane, thus
forming a conspicuous “wall” each,[39] which is heightened
by the methyl residue at C22 introduced via the iron
chemistry. The two carbonyl groups are axially disposed but
oriented to the opposite sides of the macrocycle as to
minimize the overall dipole. As expected, the side chain is
attached in an equatorial orientation.
Scheme 2. a) Allylmagnesium bromide, Et2O, À308C, 50–68%; b) Zn,
LiCl, THF; c) allyl bromide, CuCN (5 mol%), THF, 08C, 89%; d) (i)
B2(pin)2, Pt(dba)3 (2.4 mol%), 8 (2.8 mol%), THF, 608C; (ii) aq. H2O2,
NaOH, 91% (91% ee); e) TBDPSCl, DMAP, Et3N, CH2Cl2, 08C!RT,
quant.; dba=dibenzylidene-acetone; DMAP=4-dimethylaminopyri-
dine; TBDPS=tert-butyldiphenylsilyl.
pectedly, attempted asymmetric dihydroxylation of the ter-
minal olefin in 5 proved unsatisfactory, even though several
ligands were surveyed (ee ꢀ 65%).[27] This problem was
conveniently solved by a two-step/one-pot procedure com-
prising
a platinum-catalyzed enantioselective diboration
controlled by the TADDOL-derived phosphonite ligand 8,
followed by oxidation of the vicinal boronate species primar-
ily formed.[28] This method furnished excellent results in terms
of selectivity and yield, and also scaled well; to the best of our
knowledge, it is the first application to a heterocyclic substrate
featuring possible coordination sites for the platinum catalyst.
Selective protection of the primary alcohol in 6 thus formed
gave building block 7 as necessary for the synthesis of
DMDA-Pat A (2).
The crucial pyrone fragment was also readily prepared
(Scheme 3). To this end, commercial (S)-9 was opened with
lithium acetylide and the resulting product 10 immediately
engaged in a Sonogashira coupling with iodide 12 derived
from propiolate 11 in one step. A loading of 1 mol% of
[(XPhos)AuNTf2] sufficed to convert the resulting product 13
into the corresponding pyrone 14.[29,30] The reaction was
exquisitely selective for the desired 6-endo-dig cyclization and
proceeded without any noticeable interference of the unpro-
À
tected OH group; alternative protocols from the litera-
ture[31] were not nearly as effective and selective as this
powerful p-acid catalyzed methodology.[32]
Esterification of 14 with 4-pentenoic acid followed by
hydroboration of 15 gave an adequate nucleophilic partner
for an alkyl-Suzuki coupling[33] with the bromothiazole frag-
ment 7. The use of Pd(OAc)2 in combination with RuPhos[34]
furnished a remarkably active catalyst that allowed product
Next, aldehyde 21 was subjected to a Takai olefination
which gave alkenyl iodide 22 in high yield and appreciable
Angew. Chem. Int. Ed. 2016, 55, 6051 –6056
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6053