Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
2
N.P.S. Hassan et al. / Tetrahedron xxx (2014) 1e7
annulation between a pyranone and a phenylsulfonyl phthalide,
with the lactone formed in the final steps of the synthesis.10b,c We
have previously found the HK annulation to be a powerful method
for the synthesis of a number of pyranonaphthoquinone and pyr-
anonaphthalene frameworks.2a,b,11 In a new approach to nanao-
mycin D 3, we envisaged the use of an HK annulation employing an
enone containing a pre-formed chiral lactone as a key step. Pyr-
anonaphthalene 11 has been previously reported by Fernandes and
resulted in undesired isomerisation to the a,b-unsaturated ester. It
was also necessary to activate the magnesium powder by washing
with hydrochloric acid.
Alcohol 21 was next protected as a TBDPS ether 17 and was
subsequently subjected to Sharpless asymmetric dihydroxylation16
with the (DHQ)2PHAL ligand. The resulting diol underwent con-
comitant lactonisation to give a 76% yield of lactone 16 in 79% en-
antiomeric excess, as determined by Mosher ester analysis.
Protection of the alcohol in lactone 16 as an ethoxymethyl ether
was unsuccessful using ethoxymethyl chloride with either diiso-
propylethylamine or sodium hydride as a base. The alcohol in lac-
tone 16 was therefore protected as a methoxymethyl ether using
dimethoxymethane and phosphorous pentoxide, a procedure that
Bruckner in a total synthesis of nanaomycin D 3,10f hence synthesis
€
of this advanced intermediate 11 comprises a formal synthesis of
nanaomycin D 3 (Scheme 1).
avoided the use of base given the presence of a base-sensitive g-
lactone. The TBDPS ether was then deprotected using TBAF buffered
with acetic acid to give alcohol 22 in 92% yield. Oxidation of alcohol
22 to an aldehyde was unsuccessful with TPAP, IBX, PCC, TEMPO,
Swern conditions and DesseMartin periodinane (DMP) using pyr-
idine as a base. Pleasingly, use of DMP buffered with sodium bi-
carbonate gave aldehyde 15, which was immediately subjected to
an HWE olefination with dimethyl 2-oxopropylphosphonate to
furnish the desired enone-lactone 14 for the key HK annulation.
Enone-lactone 14 represents an attractive chiral synthon for
the construction of other
g-lactone containing pyranonaph-
thoquinones such as griseusin A17 and crisamicin A.18
With an expedient route to enantiopure enone-lactone 14
established, we next investigated its use as a Michael acceptor in an
HK annulation with cyanophthalide 13, itself readily prepared from
o-anisic acid as described previously.2b The mechanism of the
HausereKraus annulation is well-established, whereby the stabi-
lised anion formed from the cyanophthalide undergoes a Michael
addition to the enone and a subsequent Dieckmann-like conden-
sation then delivers the cyclic product.19 The reaction is known to
produce a mixture of hydroquinone and quinone products; there-
fore it is common to carry out a subsequent reductive methylation
step in order to trap the hydroquinone product as the dimethyl
ether derivative. Hence, HK annulation between cyanophthalide 13
and enone-lactone 14 was attempted using t-BuOK and the crude
material immediately subjected to reductive methylation condi-
tions (Table 1, entry 1).20 Disappointingly, only a complex mixture
was recovered.
The HK annulation was attempted again without carrying out
the methylation step. As found in previous HK annulation studies,2a
the product obtained was not the desired naphthoquinone in-
termediate 24, but the Michael addition product 23 (entry 2). Al-
tering the base to t-BuOLi produced the same outcome (entry 3).
Use of sodium hydride as a base gave naphthoquinone 24 (tenta-
tively assigned based on the 1H NMR spectrum of the crude ma-
terial) in 16% yield, together with 56% of the Michael addition
product 23 (entry 4). The naphthoquinone product 24 was found to
degrade quickly, so a second attempt at the reaction was un-
dertaken using sodium hydride as base for the annulation step
immediately followed by reductive methylation using caesium
carbonate and methyl iodide. These conditions furnished the de-
sired trimethoxynaphthalene 12, albeit in a low 13% yield (entry 5).
The annulation was next attempted using potassium hydride as
base followed by reductive methylation, affording a slightly higher
15% yield of methylated product 12 (entry 6). Next, the HK annu-
lation step was conducted using potassium hydride, omitting the
subsequent methylation step, affording naphthoquinone 24 and
intermediate 23 that were separated by chromatography. Pure in-
termediate 23, which was obtained in 64% yield, was then treated
with sodium hydride followed by reductive methylation in com-
bination with previously isolated naphthoquinone 24, giving tri-
methoxynaphthalene 12 in a much improved 30% yield (entry 7).
However, it was later discovered that this stepwise process was
Scheme 1. Retrosynthesis from known nanaomycin intermediate 11.
The formal synthesis of nanaomycin D 3 presented herein
focusses on assembly of pyranonaphthalene-lactone 11. Discon-
nection of the dihydropyran ring in pyranonaphthalene 11 gives
trimethoxynaphthalene 12, that in turn is assembled by an HK
annulation of known cyanophthalide 132b with enone-lactone 14.
Enone-lactone 14 could be accessed from lactone aldehyde 15 using
a HornereWadswortheEmmons (HWE) reaction. The stereogenic
centres in lactone aldehyde 15 can be readily constructed via
a Sharpless asymmetric dihydroxylation of b,g-unsaturated ester
17, followed by lactonisation.
2. Results and discussion
Initial attention focussed on the synthesis of chiral enone-
lactone 14 from unsaturated ester 17. Accordingly, synthesis of
unsaturated ester 17 began with readily available -mannitol 18
D
(Scheme 2). Acetonide protection12 and periodate cleavage13 were
carried out using literature procedures to give aldehyde 19. HWE
coupling with commercially available triethyl phosphonoacetate
under mild aqueous conditions14 cleanly afforded the desired E-
alkene 20 in good yield. Alkene 20 was then converted via reductive
cleavage to a b,g
-unsaturated ester 21 using Mg/MeOH,15 in which
transesterification with the solvent results in formation of the
methyl ester. It was found that the outcome of the reductive
cleavage step was very dependent on the particle size of the
magnesium powder used and the reaction time. Use of coarser
particle sizes afforded low yields whereas long reaction times