Organic Letters
Letter
Our retrosynthetic analysis is shown in Scheme 1. The
synthesis of 1 features the strategic use of two key reactions: an
intramolecular azomethine ylide cycloaddition12 that may also
be described as a tethered asymmetric [C+NC+CC] cyclo-
addition paired with the ring expansion of a functionalized
pyrroline to construct the target’s C- and D-rings. The
asymmetric [C+NC+CC] cycloaddition is a robust multi-
component reaction that enables the synthesis of highly
functionalized pyrrolidines in enantiomerically pure form via
the staged one-pot combination of an aldehyde (C-
component), a chiral glycylsultam (NC-component), and a
dipolarophile (CC-component) (a brief overview of this
multicomponent reaction is provided in the Supporting
+CC] cycloaddition using an alkyne dipolarophile not only
provides an enantiomerically pure pyrroline (the D-ring
precursor) but also installs a synthetic handle for the
subsequent ring-expansion reaction.
TBS protecting group followed by oxidation of primary alcohol
10 using 2-iodoxybenzoic acid (IBX) gave aldehyde 4, which
served as the CC−tether−C construct. The overall yield of 4
from 6 was 60%.
Assembly of the ergoline structure (Scheme 3) commenced
with the Ag(I)-catalyzed reaction of 4 and 5 to give
Scheme 3. Asymmetric Synthesis of Lysergic Acid
Significantly, the plan allows for potential structural
modification of the ergoline core that would expand ergot
alkaloid structural space. For example, the C-ring may be
enlarged by simply modifying the tether connecting the “C”
and “CC” components. Our strategy may also be applied to the
synthesis of the antipode of 1 and related analogues. Such
deep-seated variation in molecular structure is not found in the
natural ergot alkaloids. The resulting diversification of three-
dimensional structure space could prove useful for targeting
specific receptor subtypes employing fragment-based drug
design strategies. The use of an intramolecular azomethine
ylide cycloaddition to construct the C-ring and (pre)D-ring of
1 was inspired by Brewer’s synthesis of ( )-cycloclavine.13 The
ring-expansion reaction finds precedent in the work of Cossy14
and Charette.15
The synthesis of 1 began with the known16 trisubstituted
indole 6 (Scheme 2), itself prepared in five steps from
cycloadduct 3 in good yield on a multigram scale. On the
basis of our prior experience, this reaction was expected to
proceed through concerted pre-TS 11. It is unlikely that a
stepwise Michael addition/Mannich ring closure is involved
because the observed regioselectivity would be disfavored in
that case. The structure of 3 was confirmed by extensive NMR
as well as the eventual correlation with lysergic acid (1). It is
worth noting that attempted cycloaddition using an alkynyl
aldehyde structurally related to 4 but lacking the activating
sulfonyl group failed. Treatment of 3 with lithium borohydride
resulted in reductive removal of the auxiliary as well as
conjugate reduction of the α,β-unsaturated sulfone to give a
mixture of saturated sulfone 12 along with elimination product
13. Intermediate 12 could be converted to 13 by treating the
crude product with sodium hydride. The sulfonyl group thus
not only activates the dipolarophile but also enables
installation of the target’s double bond. Presumably, both the
conjugate reduction and sulfinate elimination reaction proceed
with the help of the hydroxyl moiety. The removal of the
sulfone group under nonreducing conditions was crucial
because it allowed the retention of the toluenesulfonyl indole
protecting group. At this point, the pyrroline was N-
methylated via reductive amination to give β-aminoalcohol 2
in 27% overall yield from cycloadduct 3 (an average of 65%
yield for each operation). Because the protons α to the
acylsultam and γ to the sulfone in 3 are potentially
epimerizable, we decided to assess the enantiomeric purity of
2. The enantiomer of 2 (ent-2) was prepared by the same
sequence of reactions but using ent-5 as the NC component.
Scheme 2. Synthesis of Disubstituted Indole 9
tion). Sonogashira coupling of trimethylsilyl acetylene and aryl
bromide 6 gave alkyne-substituted indole 7 in excellent yield.
Removal of the trimethylsilyl group produced terminal
acetylene 8, which reacted with the in situ-generated
phenylsulfonyl radical to produce alkynyl sulfone 9.17 The
sulfone moiety serves to activate the alkyne for the dipolar
cycloaddition but is readily removed afterward. This three-step
route from 6 to 9 was necessitated by the reluctance of 6 to
couple directly with ethynyl phenyl sulfone. Removal of the
B
Org. Lett. XXXX, XXX, XXX−XXX