Organic Letters
Letter
At the outset of our studies, the work of Stark had not been
reported and the only stereochemical information about the
structure of heronapyrrole C was the C8−C11 cis-geometry
(ROESY correlation) across one of the THF rings.1 Hence,
instead of using a cascade cyclization to form the two THF
rings simultaneously, we decided to construct heronapyrrole C
3, using a convergent synthesis and form the two THF rings at
different stages. The synthesis reported herein is therefore more
flexible and enabled access to all possible stereoisomers for
structure elucidation purposes.
Scheme 3. Synthesis of Aldehydes 14a and 14b
Nitration of 3-farnesyl pyrrole 4 was reported to be largely
nonselective by Fenical and co-workers, giving undesired 2-
nitropyrrole 5a as the major product.5 Stark and co-workers
used the same strategy in their approach, achieving slightly
better selectivity for 5b after optimization.
In light of this poor regioselectivity, we aimed to introduce
the 2-nitro group at the start of the synthesis, anticipating that
full control of regiochemistry would be achievable. To date, the
chemistry of 2-nitropyrroles has remained relatively unexplored,
but it was hoped that development of novel synthetic methods
would permit investigation of its potential utility in medicinal
chemistry.
Retrosynthetically, epoxide 6, accessible from Shi epoxida-
tion7 of olefin 7, constitutes a suitable cyclization precursor for
heronapyrrole C 3 (Scheme 2). Intermediate alkene 7,
TIPS) using chlorides or triflates proved unsuccessful. Finally,
N-benzyloxymethyl (BOM) 11a and N-benzoyloxymethyl
(Boz) 11b derivatives of 2-nitropyrrole were successfully
prepared and readily deprotected in excellent yield.10
With a range of N-protected 2-nitropyrroles in hand, the
pivotal regioselective halogenation step was examined (Scheme
3). Attempted iodination of either the N-Boc or N-Ts
derivatives returned only starting material or decomposition
products. Direct iodination of unprotected 2-nitropyrrole led to
mixtures of mono- and di-iodinated products, which were
inseparable by flash chromatography. Pleasingly, however,
iodination of the N-BOM 11a and N-Boz 11b 2-nitropyrroles
afforded the corresponding 4-iodo-2-nitropyrroles, 12a and
12b, respectively, in excellent yield, as the only observable
regioisomers. Subsequent Stille coupling with vinyltributyltin
proceeded uneventfully to give 13a and 13b, which were then
oxidized by exposure to iodobenzene diacetate11 to give the
requisite aldehydes 14a and 14b for subsequent Julia-Kocienski
olefination.
Scheme 2. Retrosynthesis of (+)-Heronapyrrole C
Synthesis of the key sulfone coupling partner 9 was initiated
from geraniol 15 (Scheme 4). Sharpless asymmetric epox-
idation with (−)-DIPT12 and subsequent mesylation afforded
epoxy mesylate 16 as a single stereoisomer in excellent yield for
the two steps.13 Sharpless asymmetric dihydroxylation of 16 (dr
>9:1) followed by acid-catalyzed ring opening of the epoxide,
with clean inversion of the quaternary center, gave mesylate 17
containing the required trans-tetrahydrofuran ring. Treatment
of 17 with potassium carbonate in methanol then led to
formation of terminal epoxide 18.
A direct dihydroxylation/epoxide transposition sequence
beginning from 16 provided 18 in a one-pot operation in
80% yield, however the product was contaminated with up to
10% of the undesired stereoisomer.14,15 As these dister-
eoisomers were inseparable at this or later stages, the two-
step sequence was adopted, such that 18 could be obtained
reliably in stereoisomerically pure form after chromatography.
Triethylsilyl protection of the tertiary hydroxy group was
effected under standard conditions, followed by ring opening of
the epoxide with allyl magnesium bromide in the presence of
copper(I) chloride. Silylation of the resultant secondary
alcohol, again with TESCl, then afforded alkene 19 in 80%
yield. In order to introduce the necessary sulfone unit, the
terminal alkene was subjected to Markovnikov oxymercuration-
containing the preformed trans-THF ring, was envisioned to
result from Julia-Kocienski olefination8 of 2-nitropyrrole
aldehyde 8 with advanced sulfone 9, derived from geraniol.
Our synthetic investigations toward heronapyrrole 3 began
with preliminary studies toward the required 2-nitropyrrole
building blocks, starting from 2-nitropyrrole 109 itself (Scheme
3). As the feasibility and regioselectivity of subsequent
tranformations of the pyrrole nucleus were expected to be
dictated by the nature of the pyrrole N-substituent, a
preliminary screen of suitable protecting groups was under-
taken.
Electron-withdrawing groups (Boc, Ts) were readily installed
and could also be removed in high yields under standard
conditions. A variety of benzyl subsituents could be appended,
but these could not be cleanly deprotected. N-Silylation (TES,
379
dx.doi.org/10.1021/ol403246j | Org. Lett. 2014, 16, 378−381