of the starting ratio of 6 and 10.10In order to determine the
absolute configuration of natural 1a, a highly enantioen-
riched sample was prepared from (ꢀ)-7.6a As the absolute
stereochemistry of (ꢀ)-7 is known to be 3S, 16S,11 our
sample of optically enriched 1a has the absolute config-
uration aR,3S,14R,16S. Comparison of the optical rota-
tion of our synthetic 1a with the literature value1a,12a
showed that synthetic 1a was in the opposite enantiomeric
series, and therefore the absolute configuration of naturally
occurring 1a can now be assigned as aS,3R,14S,16R.12b
A detailed assessment of NOEs associated with 3 and
X-ray crystallographic analysis confirmed the relative
stereochemistry of 3 to be as shown in Scheme 6 and
indicated that, unlike the reactions involving 10 and 12,
the oxidative cleavage of 14 had occurred with retention of
the C14 stereochemistry (Scheme 6). A highly enantioen-
riched sample of 315 was prepared from (ꢀ)-7 in three steps
and was found to be in the opposite enantiomeric series to the
isolated natural product. Natural occurring (þ)-3 can there-
fore be assigned the absolute configuration aS,3R,14R,16R.
Scheme 5. Synthesis of 1a and a Representation of the X-ray
Crystal Structure of 1a
Scheme 7. Proposed Mechanism of Photooxidative Cleavage
Our attention next turned to the synthesis of 3, as the
reported structure showed that 3 was epimeric to 1a at C14.
O-Ethyleburnamine (14) was prepared as a single diastereo-
mer by alkylation of a mixture of 6and 10 under phase transfer
conditions.13 Subsequent oxidative cleavage of 14 gave 3 as a
single diastereoisomer (Scheme 6), the 13CNMRofwhichwas
identical to that previously reported in the literature.2b,14
A possible rationalization of the stereochemical out-
come of the reaction involves addition of singlet oxygen
to the indole C7 position from the convex face of 6 and 10
to give diastereomers 15/16. Subsequent ring closure
would form dioxetanes 17/18 which could undergo C14
equilibration, via the aldehyde 19, to give only 17 with the
required stereochemistry to yield 1a after collapse of the
dioxetane ring.16 Alternatively collapse of the dioxetane in
17/18 could give 1a and its C14 epimer which could
undergo equilibration to only 1a via 21 (Scheme 7).17
When 1a was first isolated it was proposed that it could
serve as a biosynthetic precursor of quinolone alkaloids.1a
More recently a quinolone alkaloid was isolated alongside its
proposed biosynthetic keto-lactam and eburnane type alka-
loid precursors,2a adding weight to this proposal. With a
reliable route to keto-lactams of type 1a and 3 in hand, we
decided to investigate this intramolecular aldol reaction.
In our case, treatment of 1a with sodium methoxide brought
about the desired condensation, giving a 3:1 mixture of
C14 diastereomers 22 and 23 (Scheme 8).18 Similar reaction
Scheme 6. Preparation of 3 and a Representation of the X-ray
Crystal Structure of 3
(10) Performing the reaction starting with varying diastereomeric
ratios of 6/10 always exclusively returned 1a. A more detailed discussion
of the mechanism of this reaction is provided later in the text.
(11) Toh-Seok, K.; Poh-suan, T.; Chen, W. Phytochemistry 1993, 33,
921–924.
(12) (a) (ꢀ)-1a [R]D = ꢀ154 (c = 0.29, CHCl3) (lit. [R]D þ166
(c = 0.18, CHCl3).1a1a (b) Our assignment of the absolute configuration
of the isolated natural product raises interesting issues relating to the
structural assignments of the potential precursors of 1a in M. henryi
previously reported in the literature (see Supporting Information).
(13) (a) Arambewela, L. S. R.; Khoung-Huu, F. Phytochemistry 1981,
20, 349–350. (b) Khoung-Huu et al. reported the synthesis of O-iso-
methyleburnamine from epieburnamine 6 using this methodolgy; however
we were unable to replicate this when alkylating with either MeI or EtBr.
(14) For a detailed comparison of the 1H NMR spectrum of our
material with the reported spectrum for 3, see Supporting Information.
(15) [R]D = ꢀ92 (c = 0.27, CHCl3) (lit.2b2b þ126.66 (c = 0.65,
CHCl3)).
(16) An alternative C14 epimerization mechanism via a N1ꢀC14
iminium ion is unlikely, as these reactions were run in methanol and no
C14 OMe substituted products were observed.
(17) As the C14 O-alkyl substrate 14 did not undergo C14 epimeriza-
tion during the oxidative cleavage reaction, the presence of a C14
hydroxy group seems to be a requirement for this mechanism to operate.
Org. Lett., Vol. XX, No. XX, XXXX
C