slow. Fortunately, substituting Cs2CO3 for K3PO4 gave a
faster reaction and a 64% yield of 31 after 15 h at 65 °C.
When the modified procedure was applied to 18 and 30, a
single coupling product (9) was isolated in 57-65% yield,
depending on the scale.
closely to those reported for N-alkoxycarbonyl indole rota-
mers.23 Two sets of N-BOC singlets were also observed at
-45 °C, δ 1.78 ppm (minor) and 1.68 ppm (major).
At rt, the C(27) oxazole proton was a sharp singlet at δ
6.45 ppm and the C(29) methylene protons appeared as an
AB quartet (δ 4.15 ppm). A positive NOE interaction
confirmed that one of these methylene protons is oriented
as shown in macrocycle 8, Scheme 4, proximal to the acetal
proton. At -45 °C, the methylene signals were broadened
and the oxazole C(27) proton was barely resolved into two
partially overlapping singlets, suggesting small long-range
effects on the chemical shifts due to N-BOC rotamers. Larger
differences in these signals would be expected if the oxazole
atropisomers had been frozen out. The NMR evidence is
consistent with the presence 8 as the dominant atropisomer
in solution, or with rapid interconversion of atropisomeric
oxazoles.
In conclusion, the macrocyclic core (8) of diazonamide
A was assembled in 11 linear steps from benzofuran 19 with
an overall yield of 11%.24 The C(16)-C(18) biaryl bond was
made by Suzuki coupling following the analogy from the
synthesis of 7.3c The key macrocycle was formed by an
imino-Dieckmann cyclization to close the C(29)-C(30)
bond.25 According to the X-ray and solution NMR evidence,
the cyclization produced the correct oxazole atropisomer.
High dilution conditions were not required for the cyclization,
presumably due to the favorable conformational constraints
present in the precursor 9. The C(30) ketone in the resulting
macrocycle 8 offers a number of options for the eventual
elaboration to a fused oxazole as in the diazonamides 1-4.
Further studies toward the synthesis of relevant targets will
be reported in due course.
The NMR spectrum of 9 in CDCl3 contained broad peaks
indicative of atropisomers interconverting at room tempera-
ture.3c The broad C(27) oxazole proton signals at δ 6.20 and
6.37 ppm suggested two species in solution with a ratio of
ca. 3:2. This ratio was also reflected in the C(29) methyl
signals at δ 1.41 and 2.09 ppm, but a single N-Boc signal
was present as a sharp singlet at δ 1.69 ppm. An NMR study
was carried out in toluene-d8 over a broad temperature range
to estimate the energy barrier for atropisomer interconversion.
This solvent altered the ratio of atropisomers to ca. 3:1 on
the basis of the methyl ester and the C(29) methyl signals.
A low-temperature spectrum at -20 °C clearly indicated the
presence of two species with interconversion frozen out on
the NMR time scale (methyl ester, δ 3.08 and 3.53 ppm;
C(29) methyl, δ 1.30 and 1.96 ppm). Upon raising the
temperature to 30 °C, the methyl ester and the C(29) methyl
signals became broad, while heating the sample to >70 °C
gave coalescence and one peak for both the C(29) methyl
group and the methyl ester (δ 1.76 and 3.47 ppm, respec-
tively). This behavior is characteristic of indoles containing
the C(24) oxazole and a hindered aryl substituent at C(18)
as described in an earlier publication from our laboratory.3c
In the current study, a value for ∆G* ) ca. 15.5 kcal/mol
was determined, corresponding to the barrier for rotation
about the C(16)-C(18) bond. Thus, 9 exists as a mixture of
atropisomers that interconvert on the laboratory time scale
at temperatures well below 0 °C.21
With subunits 18 and 30 coupled, macrocycle precursor
9 contains the C(28)-methyl and C(10)-ester functionalities
necessary to test the Dieckmann-type cyclization.22 The first
attempt by treatment of 9 with 3 equiv of lithium bis-
(trimethylsilyl)amide at room temperature (8 min) gave
macrocycle 8 in ca. 30% yield. Optimum conditions were
reached by using 3 equiv of lithium diisopropylamide (LDA)
at -23 °C (5 min), a procedure that afforded 8 in 57% yield.
The structure and oxazole orientation corresponding to the
macrocyclic biaryl subunit of naturally derived 3 were
verified in detail by X-ray crystallography as shown in
Scheme 4.
NMR analysis of 8 in CDCl3 revealed a single set of well-
resolved signals at room temperature, in contrast to 9.
However, cooling to -20 °C caused broadening of signals
for the protons closest to indole nitrogen. At -45 °C, two
sets of signals were frozen out in a 3:2 ratio: C(21)-H
doublets at δ 8.42 ppm (major) and 8.07 ppm (minor);
C(25)-H singlets at 7.97 ppm (minor) and 7.81 (major). The
chemical shifts and coalescence temperatures correspond
Acknowledgment. This work was supported by NIH
(CA17918). The authors thank Dr. D. A. Barda for working
out the conversion from 19 to 20 and Dr. J. W. Kampf for
the X-ray structure.
Supporting Information Available: Experimental de-
tails, spectroscopic characterization, and X-ray data tables
for 8. This material is available free of charge via the Internet
OL016097R
(23) Oldroyd, D. L.; Weedon, A. C. J. Org. Chem. 1994, 59, 1333.
Morales-Rios, M. S.; Joseph-Nathan, P. Magn. Reson. Chem. 1987, 25,
911.
(24) In a preliminary study, intermediate 21 has also been made from
benzofuran as shown below:
(21) For macrocyclization at -23 °C, ∆G* ) 15.5 kcal/mol corresponds
to a half-life of ca. 5 s for atropisomer interconversion.
(25) For leading references to oxazole side chain lithiation, see: Lipshutz,
B. H.; Hungate, R. W. J. Org. Chem. 1981, 46, 1410. Evans, D. A.; Cee,
V. J.; Smith, T. E.; Santiago, K. J. Org. Lett. 1999, 1, 87.
(22) A similar cyclization has been considered by Moody et al.4a
2454
Org. Lett., Vol. 3, No. 16, 2001
Vedejs
