Biogenetically patterned synthesis of camptothecin and 20- deoxycamptothecin

Article abstract of DOI:10.1016/S0040-4039(99)02210-8

A biogenetically patterned synthetic route to the monoterpenoid quinoline alkaloids 20-deoxycamptothecin and (±)-camptothecin from secologanin and tryptamine via vincoside/strictosidine lactams has now been realized, and hence afforded likely biosynthetic intermediates for testing in vivo. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(99)02210-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 859–862
Biogenetically patterned synthesis of camptothecin and
20-deoxycamptothecin
Richard T. Brown,∗ Liu Jianli and Cid A. M. Santos †
Department of Chemistry, The University of Manchester, Manchester M13 9PL, UK
Received 14 September 1999; accepted 22 November 1999
Abstract
A biogenetically patterned synthetic route to the monoterpenoid quinoline alkaloids 20-deoxycamptothecin and
(±)-camptothecin from secologanin and tryptamine via vincoside/strictosidine lactams has now been realized,
and hence afforded likely biosynthetic intermediates for testing in vivo. © 2000 Elsevier Science Ltd. All rights
reserved.
Keywords: alkaloids; biogenesis; biomimetic; indole; quinoline; terpene.
The alkaloid (+)-camptothecin 1 was isolated in 1966 by Wall and co-workers from Camptotheca
acuminata, a tree native to China, and shown to possess unique anticancer properties.1 Hence, over the
subsequent three decades its relatively simple structure has been the objective of many total syntheses
using a variety of approaches.1,2 Early biogenetic speculation focused on the similarity of the skeleta
of camptothecin and vincoside/strictosidine lactams 6/7, given that conversion of indole to quinoline
heterocycles was well known, and this was confirmed by in vivo incorporation of labelled strictosidine
lactam 6 into camptothecin.2 It thus belongs to the monoterpenoid indole alkaloid family, derived
from tryptamine 2 and secologanin 3, but little is known about the details of its biosynthesis beyond
what is outlined in Scheme 1. Conversion of 6 to 1 is a net oxidative process requiring: (i) oxidative
rearrangement of rings B and C; (ii) reduction of C-17, 18 and 19; (iii) aromatisation of ring D; (iv)
hydrolysis of the glucoside; and (v) oxidation of C-21 and C-20. At least seven distinct steps are
involved, which could occur in various orders. Through the maze of possible pathways, a biogenetically
patterned synthetic route to camptothecin from tryptamine and secologanin has now been found, and
likely biosynthetic intermediates produced for eventual testing in vivo.
Condensation of tryptamine and secologanin in pH 4 buffer afforded a ca. 3:2 mixture of the C-3
epimers vincoside 5 and strictosidine 4, which was converted to a mixture of the corresponding lactams
6/7 by heating with aq. Na2CO3, or vincoside could be selectively lactamised with Et3N/MeOH to 7
∗
†
Corresponding author. Tel: +44 161 275 4632; fax: +44 161 275 4939; e-mail: r.t.brown@man.ac.uk (R. T. Brown)
On sabbatical leave from Universidade Federal do Paraná, Curitiba, PR, Brazil.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02210-8
tetl 16138

860
Scheme 1.
and separated from strictosidine.3,4 Since H-3 is lost in aromatisation of ring D, we have carried out the
complete reaction sequence below on both the lactam mixture and vincoside lactam alone, but for clarity
only the latter is described.
In early work it was established that oxidative rearrangement of the tetrahydro-β-carboline system
of the lactams 6/7 to a pyrroloquinolone could be achieved.5–7 Also, ring D in 18,19-dihydro-6/7 could
be aromatised with DDQ in methanol, if only with incorporation of a methoxy group at C-17, but the
product could not be converted further to a quinolone.5 It was concluded that prior reduction of the 16,17
double bond would facilitate ring D aromatisation. Accordingly, catalytic hydrogenation of vincoside
lactam tetra-acetate with Raney Ni afforded an excellent yield of the 16S,17,18,19-tetrahydro derivative,
mp 279°, but unfortunately conversion to a quinolone and subsequent steps went in unacceptably low
yields.8,9 Eventually the following successful route was devised, as summarised in Scheme 2.
With NaIO4 in refluxing aq. MeOH the indole 2,7-bond in vincoside lactam tetra-acetate was cleaved to
a ketolactam 8, as indicated inter alia by λmax(MeOH) 246 nm; M+ 698.232 (C34H38N2O14).7 Heating 8
with Et3N/MeOH led to aldol condensation between C-2 and C-6 and formation of the pyrroloquinolone
9 [λmax 234, 316, 328 nm; νmax(film) 1750, 1720, 1660, 1570 cm−1; M+ 680.225 (C34H36N2O13)] with a
characteristic 1H NMR (200 Mhz, CDCl3) signal for the C-5 methylene at δ 5.20. Significantly, the C-3
epimers of 8 and 9, which we made from strictosidine lactam some time ago,6,7 have now been isolated
as natural products.10 Treatment of the quinolone with SOCl2 in DMF gave the 7-chloroquinoline 10,
mp 203–205°, identified by a typical quinoline UV spectrum with λmax 218, 240, 372 nm, loss of the NH
signal at δ 9.35, νmax 1752, 1663 cm−1 and (M+H)+ 699.195 (C34H36N2O1235Cl).
In a crucial step, the 7-chloroquinoline in ethanol was treated with hydrogen (50 psi) over freshly
prepared Raney Ni catalyst for 24 h. In the event, not only were the 16,17-β-alkoxyacrylamide and 18,19-
vinyl groups hydrogenated, and the chlorine hydrogenolysed, but the quinoline was also partially reduced
to afford as a single stereoisomer the dihydroquinoline 11. Its structure was indicated by M+ 670.276
(C34H42N2O12), νmax 1662 cm−1 (δ-lactam), λmax 250 nm (similar to an aniline) and corroborated by
the 1H NMR spectrum11 with a new NH peak and signals for new CH bonds at C-7, 16, 17, 18 and 19.
Subsequent aromatisation of both rings B and D in 11 was achieved with DDQ in dioxane to give the
conjugated quinolinopyridone 12, possessing the same UV chromophore as camptothecin, as indicated
by a similar spectrum with λmax 252, 288, 360 nm. In the NMR spectrum there were new aromatic proton
singlets at δ 7.15 and 8.34 for H-14 and H-7, respectively, and the molecular ion at m/z 664.229 analysed
for the required C34H36N2O12.

861
Scheme 2. Reagents and conditions: (i) Ac2O/py 12 h; (ii) NaIO4/aq. MeOH/∆ 30 min; (iii) Et3N/MeOH/∆ 30 min;
(iv) SOCl2/DMF/0°C 10 min; (v) Raney Ni/H2/EtOH 24 h; (vi) DDQ/dioxane/∆ 10 min; (vii) NaOMe/MeOH 2 h; (viii)
β-glucosidase/pH 5 buffer 3 days; (ix) PCC/DCM 30 min; (x) O2/CuCl2/DMF 10 h
After Zemplen deacetylation, the glucose was removed with β-glucosidase in pH 5.0 buffer to yield
the lactol 13, characterised by a hydroxyl peak at 3385 cm−1 in the IR spectrum and a doublet at δ
4.81 assigned to H-21, together with loss of all sugar protons in the NMR spectrum. No molecular ion
could be detected in its MS but a peak at m/z 317.127 corresponded to ready loss of OH (C20H17N2O2).
Oxidation of the lactol with PCC/CH2Cl2 gave a lactone, as indicated by loss of the IR peak at 3385
cm−1 and the NMR signal for H-21, together with a new carbonyl band at 1738 cm−1. Corroboration
that the product corresponded to the known alkaloid 20-deoxycamptothecin12 14, was obtained from M+
332.118 (C20H16N2O3), and comparison of UV and 1H NMR spectra.13

862
14
Finally, H-20 was oxidised to a hydroxyl group by O2/CuCl2 but with consequent loss of chirality
so that the product was racemic camptothecin, identical with an authentic sample of (+)-1 by TLC, MS,
1
IR, UV and H NMR spectra.15 However, routes to the (+)-isomer using enzymatic oxidation of 14 or
the intrinsic chirality of 11 are in progress. We have thus achieved the target of a biomimetic synthesis of
camptothecin, and also prepared the likely biosynthetic intermediates 12, 13 and 14. These and others,
such as the C-3 epimers of 8 and 9 from strictosidine lactam 4, can now be prepared with labels for in
vivo experiments.
Acknowledgements
We thank Drs. L. Akhter, D. Curless, S. B. Fraser, A. G. Lashford, and P. Richards for their pioneering
work in this area, and the British Council, the CVCP and the Chemistry Department, Manchester
University for financial support (JL).
References
1. Wall, M. E.; Wani, M. C. In The Alkaloids; Cordell, G. A., Ed.; Academic Press: San Diego, 1998; Vol. 50, Chapter 13, pp.
509–520 and references cited therein.
2. Cai, J.-C.; Hutchinson, C. R. In The Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1983; Vol. XXI, Chapter 4, pp.
101–137 and references cited therein.
3. Battersby, A. R.; Burnett, A. R.; Parsons, P. G. J. Chem. Soc. (C) 1969, 1193–1200.
4. Brown, R. T.; Leonard, J.; Sleigh, S. K. Phytochemistry 1978, 17, 899–900.
5. Fraser, S. B. Ph.D. Thesis, University of Manchester, 1975.
6. Lashford, A. G. Ph.D. Thesis, University of Manchester, 1978.
7. Hutchinson, C. R.; O’Loughlin, G. J.; Fraser, S. B.; Brown, R. T. Chem. Commun. 1974, 928.
8. Akhter, L. Ph.D. Thesis, University of Manchester, 1979.
9. Curless, D. Ph.D. Thesis, University of Manchester, 1985.
10. Carte, B. A.; DeBrosse, C.; Eggleston, D.; Hecht, S. M.; Hemling, M.; Mentzner, M.; Poeland, P.; Troupe, N.; Westley, J.
W. Tetrahedron 1990, 46, 2747–2760.
1
11. Compound 11: H NMR (200 MHz, CDCl3): δ 7.90 (bs, NH), 7.00–6.45 (m, 4 Ar-H), 5.37 (bs, H2-5), 5.20 (t, J=9 Hz,
H-30), 5.08 (dd, J=9, 8 Hz, H-40), 5.05–4.88 (m, H-20, H-3, H-10), 4.42 (d, J=9 Hz, H-21), 4.30 (m, J=6 Hz, H2-60), 4.12
(dd, J=12, 5.5 Hz, H-17β), 3.90 (dd, J=12, 9 Hz, H-17α), 3.75 (m, J=8, 6 Hz, H-50), 3.40 (d, J=13 Hz, H-7b), 3.28 (d, J=13
Hz, H-7a), 2.78 (m, J=9, 5.5, 5 Hz, H-16β), 2.57 (m, J=5, 6, 6, 12 Hz, H-15), 2.34 (m, J=13, 6, 2.5 Hz, H14β), 2.10–1.95 (4
s, 4 OAc), 2.00–1.80 (m, H-14α, H-19b), 1.65–1.38 (m, H-20β, H-19a), 0.98 (t, J=8 Hz, H3-18).
12. Adamovics, J. A.; Cina, J. A.; Hutchinson, C. R. Phytochemistry 1979, 18, 1085–1086.
13. Compound 14: λmax 254, 288, 360 nm; 1H NMR (500 MHz, CDCl3): δ 8.39 (s, H-7), 8.20 (d, J=8 Hz, H-12), 7.93 (d, J=8
Hz, H-9), 7.82 (t, J=8 Hz, H-10/11), 7.66 (t, J=8 Hz, H-11/10), 7.44 (s, H-14), 5.56 (d, J=17 Hz, H-17b), 5.38 (d, J=17 Hz,
H-17a), 5.29 (s, H2-5), 3.62 (t, J=6 Hz, H-20), 2.08 (m, H2-19), 1.07 (t, J=8 Hz, H3-18).
14. Winterfeldt, E.; Korth, T.; Pike, D.; Boch, M. Angew. Chem., Int. Ed. Engl. 1972, 11, 289–290.
15. Compound 1: λmax 252, 288, 367 nm; 1H NMR (500 MHz, CDCl3): δ 8.39 (s, H-7), 8.22 (d, J=8 Hz, H-12), 7.92 (d, J=8
Hz, H-9), 7.82 (t, J=8 Hz, H-10/11), 7.70 (s, H-14), 7.65 (t, J=8 Hz, H-11/10), 5.75 (d, J=17 Hz, H-17b), 5.33 (s, H2-5), 5.31
(d, J=17 Hz, H-17a), 1.87 (m, H2-19), 1.07 (t, J=8 Hz, H3-18).