7102
H. Holla et al. / Tetrahedron Letters 53 (2012) 7101–7103
Figure 2.
ether and the quinolinone ring. This would explain why the aryl
group in the cinnamyl ether and the quinolinone of the N-methylfl-
indersine are on the same face of the cyclobutane ring. The reaction
was also chemoselective, with only [2+2] cycloaddition between
the cinnamyl ether and the disubstituted double bond of N-meth-
ylflindersine being observed.
Acknowledgements
We thank Dr. Hoan T. Vu for HRMS measurements and Assoc.
Professor Craig Williams for use of the Rayonet photochemical reac-
tor. We thank the Australian Research Council (ARC) for support to-
wards NMR and MS equipment (LE0668477 and LE0237908).
Scheme 1. Synthesis of melicodenine C (1, relative stereochemistry shown).
Supplementary data
work in the synthesis of related compounds.3 Irradiation for 24 h
led to the desired intermolecular cycloaddition product 1 in modest
yield (22%) and with excellent chemo and regioselectivities. Irradi-
ation for longer periods did not result in any significant improve-
ment in the yield. Subsequent reactions were carried out in
acetonitrile solution using a conventional Rayonet reactor with
either a quartz or glass tube at 300 nm. The reaction was carried
out with 1 equiv of 5 and 3 equiv of the cinnamyl ether in acetoni-
trile for 72 h. The yield of 1 obtained using the Rayonet reactor was
15% (34% after allowing for recovered 5). Some isomerisation of the
cinnamyl ether (15% isomerised to the Z-isomer) was observed, as
well as the formation of small amounts of the [2+2] cycloaddition
dimer of N-methylflindersine. Changing the solvent to acetone, or
addition of a sensitiser (benzophenone) did not appear (by LCMS)
to have any significant effect on the conversion of 5 into 1. Use of
toluene as the solvent however, did appear (by LCMS) to give a
slight improvement in the conversion, but this was not further
investigated. Attempts at thermal [2+2] cycloaddition employing
AlCl3 or transition metal mediated reactions4 (such as [Ni(PPh3)2
Cl2] + Zn) were unsuccessful.
Supplementary data associated with this article can be found, in
References and notes
1. Nakashima, K.; Oyama, M.; Ito, T.; Akao, Y.; Witono, J. R.; Darnaedi, D.; Tanaka,
T.; Murata, J.; Iinuma, M. Tetrahedron 2012, 68, 2421–2428.
2. Wang, X.; Lee, Y. R. Synthesis 2007, 3044–3050.
3. Neve, J. E.; Wijesekera, H. P.; Duffy, S.; Avery, V.; Carroll, A. R.; Garavelas, A.;
Jenkins, I. D.; Le, P. V.; Leone, P. de A.; Nikolapoulos, G.; Pham, N.; Ripper, J. A.;
Teague, S. J.; Williams, N.; Quinn, R. J. unpublished data.
4. Huang, D.-J.; Rayabarapu, D. K.; Li, L.-P.; Sambaiah, T.; Cheng, C.-H. Chem. Eur. J.
2000, 6, 3706–3713.
5. Dictionary of Natural Products on DVD, CRC Press, Taylor and Francis, 2012.
6. General procedure for the [2+2] cycloaddition reaction employed to synthesise
melicodenines C (1), D (2) and E (3): A solution of N-methylflindersine (5)
(0.5 mmol) and one of the respective cinnamyl ethers (6, 7 or 8) (3 equiv) in
MeCN (2 mL) was purged with a gentle stream of argon in an ultrasonic bath for
15 min and transferred to either a glass or quartz test tube under an atmosphere
of argon. The mixture was irradiated using a Rayonet photochemical reactor at
k = 300 nm for 72 h, at room temperature. The solvent was evaporated under
reduced pressure and the remaining residue was purified either by flash column
chromatography or by reverse phase HPLC (C-18 betasil column), to give the
respective melicodenines C (1), D (2) or E (3).
Melicodenines D and E were synthesised from N-methylflinder-
sine (5) and the cinnamyl derivatives 7 and 8 (both of which also
occur naturally5), respectively, using the same general procedure
as that outlined in Scheme 1. The yields for the final [2+2] cycload-
ditions were 11% (2) and 10% (3) (22% after allowing for recovered
5), respectively. The cinnamyl ethers 7 and 8 (Fig. 2) were prepared
in 41% and 49% overall yields, respectively, via a three-step se-
quence analogous to that outlined for 6 in Scheme 1.
Melicodenine C (1): Colourless oil; 1H NMR (500 MHz, CDCl3): d 8.02 (dd, J = 8.04,
1.3 Hz, 1H, H-10) 7.50 (td, J = 7.7. 1.2 Hz, 1H, H-8), 7.22 (m, 2H, H-7,9), 6.48 (d,
J = 8.2 Hz, 1H, H-50), 6.48 (d, J = 1.5 Hz, 1H, H-20), 6.38 (dd, J = 7.9, 1.5 Hz, 1H, H-
60), 5.80 (d, J = 1.4 Hz, 1H, OCH2O), 5.77 (d, J = 1.5 Hz, 1H, OCH2O), 3.85 (dd,
J = 8.5, 8.3 Hz, 1H, H-4), 3.69 (dd, J = 9.5, 8.8 Hz, 1H, H-70), 3.43 (dd, J = 10.0,
4.1 Hz, 1H, H-90a), 3.41 (dd, J = 10.0, 4.1 Hz, 1H, H-90b), 3.36 (s, 3H, 6-Me), 3.31
(s, 3H, 90-OMe), 2.64 (m, 1H, H-80), 2.60 (m, 1H, H-3), 1.52 (s, 3H, 2-Me), 1.18 (s,
3H, 2-Me); 13C NMR (125 MHz, CDCl3): d 162.6 (C-5), 155.5 (C-10b), 146.7 (C-30),
145.8 (C-40), 138.8 (C-6a), 133.9 (C-10), 130.1 (C-8), 123.0 (C-10), 121.3 (C-60),
121.2 (C-9), 116.6 (C-10a), 113.6 (C-7), 109.1 (C-20), 107.6 (C-4a), 107.2 (C-50),
100.5 (OCH2O), 76.5 (C-2), 74.3 (C-90), 59.0 (90-OMe), 44.0 (C-70), 41.6 (C-3), 40.1
(C-80), 33.1 (C-4), 28.9 (6-Me), 25.2 (2-Me), 23.8 (2-Me). LRMS (ESI): m/z 434
[MH]+. HRMS (ESI): Calcd for C26H28NO5 [MH]+: 434.1961. Found: 434.1981.
Melicodenine D (2): Colourless oil; 1H NMR (500 MHz, CDCl3): d 8.03 (dd, J = 8.1,
1.5 Hz, 1H, H-10), 7.50 (td, J = 8.5, 1.5 Hz, 1H, H-8) 7.21 (m, 2H, H-7,9), 6.63 (dd,
J = 8.1, 1.6 Hz, 1H, H-60), 6.62 (d, J = 8.1 Hz, 1H, H-50), 6.28 (d, J = 1.5 Hz, 1H, H-
20), 3.89 (m, 1H, H-4), 3.76 (s, 3H, 30OMe), 3.70 (m, 1H, H-70), 3.44 (m, 2H, H-90),
3.32 (s, 3H, 90-OMe), 3.31 (s, 3H, 6-Me), 3.23 (s, 3H, 40-OMe), 2.68 (m, 1H, H-80),
2.62 (dd, J = 8.8, 8.5 Hz, 1H, H-3), 1.53 (s, 3H, 2-Me), 1.19 (s, 3H, 2-Me); 13C NMR
(125 MHz, CDCl3): d 162.5 (C-5), 155.3 (C-10b), 147.8 (C-40), 147.3 (C-30), 138.7
(C-6a), 132.5 (C-10), 130.0 (C-8), 122.8 (C-10), 121.3 (C-9), 121.2 (C-60), 116.5 (C-
10a), 113.5 (C-7), 110.7 (C-20), 110.2 (C-50), 108.0 (4a), 76.6 (C-2), 74.5 (C-90),
59.0 (90-OMe), 55.7 (30-OMe), 55.0 (40-OMe), 43.7 (C-70), 41.7 (C-3), 40.0 (C-80),
32.8 (C-4), 28.8 (6-Me), 25.1 (2-Me), 23.6 (2-Me). LRMS (ESI): m/z 450 [MH]+.
HRMS (ESI): Calcd for C27H32NO5 [MH]+: 450.2275. Found: 450. 2276.
The 1H and 13C NMR spectra of 1–3 were virtually identical to
those reported by Oyama et al.1 for the natural products. An excep-
tion was the resonance reported for C-7 in melicodenine E, for which
Oyama et al.1 report a chemical shift of 107.1 ppm. This appears to
be a typographical error as C-7 in melicodenines C and D occurs at
113.7 and 113.6 ppm, respectively. The chemical shifts for C-7 in
(synthetic) 1–3 were 113.6, 113.5 and 113.4 ppm, respectively.
In conclusion, melicodenines C–E have been synthesised in
modest yields by an intermolecular [2+2] cycloaddition strategy.6
This provides confirmation of the structures and the relative ste-
reochemistry assigned by Oyama et al. to the natural products.
The [2+2] cycloaddition was regioselective, possibly as a result of
a
p-stacking interaction between the aryl group of the cinnamyl