An anionic polycondensation strategy for the synthesis of dibenzoxanthenones: Progress toward the synthesis of hypoxyxylerone

Article abstract of DOI:10.1055/s-2004-835634

An anionic polycondensation has been used as the key step in a highly convergent strategy for the preparation of hypoxyxylerone derivatives.

Full text of DOI:10.1055/s-2004-835634

LETTER
2693
An Anionic Polycondensation Strategy for the Synthesis of Dibenzo-
xanthenones: Progress Toward the Synthesis of Hypoxyxylerone
Synthesis of
D
ib
m
enzoxanthenones manuel Chevenier, Christophe Lucatelli, Urvish Pandya, Wei Wang, Yves Gimbert,* Andrew E. Greene
Université Joseph Fourier, Chimie Recherche (LEDSS), 38041 Grenoble, France
Fax +33(4)76514494; E-mail: Yves.Gimbert@ujf-grenoble.fr
Received 16 August 2004
OH
HO
Abstract: An anionic polycondensation has been used as the key
step in a highly convergent strategy for the preparation of hypoxy-
xylerone derivatives.
O
HO
Key words: antitumor agents, condensation, heterocycles, natural
products, total synthesis
OH
O
OH
Hypoxyxylerone
1
Hypoxyxylerone (1), a dibenzoxanthenone isolated from
the fungus Hypoxylon fragiforme in 1991 by Edwards and
Figure 1
coworkers, has been shown to inhibit in vitro topo-
The anionic polycondensation reported over 20 years ago
by Kjaer and coworkers,3 although demonstrated only for
acetophenones, appeared to offer an attractive alternative.
This under-used methodology, if it could be successfully
extrapolated to acetonaphthones, seemed capable of
providing a short, highly convergent, and quite flexible
approach to hypoxyxylerone and diverse derivatives.
isomerase I.1
We have been interested in the synthesis of not only this
scarce metabolite, but also diverse analogs that might
provide insight into structure–activity relationships. We
have already disclosed a first-generation approach to the
structurally unique natural product, which culminated in
the preparation of its penta(O-methyl) derivative 5
(Scheme 1).2 This route, based on a novel anionic homo-
Fries rearrangement of ester 2 to secure the key xanthone
intermediate 4, via 3, was convergent, but the preparation
of each of the naphthalene moieties in 2 was long, which
made it ill-suited for the synthesis of analogs.
The key step in the envisaged approach, outlined in
Scheme 2, would be the formation of dibenzoxanthone I
through anionic polycondensation of acetonaphthone IV
with homophthalate III. Each of these components can be
readily prepared with considerable structural and func-
tional group variation.
OMe
MeO
OMe
MeO
BnO MeO
OBn
MeO
MeO
Br
O
OMe
OMe OMe O
OMe OMe O
OH
2
3
OMe
MeO
O
OMe
MeO
O
MeO
MeO
OMe OH
O
OMe O
OMe
OMe
4
5
Scheme 1
SYNLETT 2004, No. 15, pp 2693–2696
0
7
.1
2
.2
0
0
4
Advanced online publication: 08.11.2004
DOI: 10.1055/s-2004-835634; Art ID: G31904ST
© Georg Thieme Verlag Stuttgart · New York

2694
E. Chevenier et al.
LETTER
This encouraging result prompted efforts to prepare more
complex derivatives of hypoxyxylerone through the use
of this approach. For comparison with the previous
approach, the penta(O-methyl) derivative 4 was first
targeted. The requisite, known3 homophthalate diester 11
could be easily prepared on a large scale as shown in
Scheme 4. Methyl acetoacetate was autocondensed5 to
give methyl orsellinate which was converted to diether
10a with dimethyl sulfate and potassium carbonate in re-
fluxing acetone (56%, 2 steps). Carbomethoxylation at the
benzylic position of 10a was best accomplished indirectly
by carboxylation of the corresponding acid 10b to yield
the diacid derivative,6 followed by methylation to afford
diester 11 (65%, 2 steps).
R2
R2
R2
R2
O
R1
R1
O
O
O
O
a
b
R1
OH
R1
O
R3
R3
I
II
R2
R2
R1
CO2Me
CO2Me
HO
O
+
R1
R3
The 4-step preparation of the second moiety, acetonaph-
thone 15b, was based on chemistry that had previously
been applied for the synthesis of molecules similar to
naphthoate 15a, but required considerable modification to
obtain serviceable results (Scheme 5).7 The anion formed
from the methyl orsellinate derivative 10a was treated
with Weinreb amide 128 to provide in 68% yield a-meth-
oxy ketone 133,9,10 which was smoothly converted into
isocoumarin 1411 with sodium hydride in THF.9
IV
III
R
1–3 = H, OR
Scheme 2
For the purpose of rapidly ascertaining whether the
acetonaphthone condensation would in fact be feasible
and, if so, whether it would produce primarily the desired
L-shaped product I through selective cyclization at
carbonyl a in the likely intermediate II (as opposed to the
U-shaped product through cyclization at b), the reaction
of commercially available 1¢-hydroxy-2¢-acetonaphthone
with dimethyl homophthalate was first examined
(Scheme 3).
OMe
OMe
OMe
N
OMe
12
CO2Me
O
CO2Me
O
NaH, THF
99%
LDA, THF
68%
MeO
MeO
OMe
10a
13
O
NaH
THF
CO2Me
HO
O
+
OMe O
OH
OMe
CO2Me
COR
OMe
MeOAc, LDA;
OH
O
O
HOAc
75%
OMe
MeO
MeO
6
7
8
MeLi;
TMSCl
59%
14
15a, R = OMe
b, R = Me
Scheme 3
Under the conditions described by Kjaer and coworkers,3
a unique xanthone 8 was formed in 46% yield, which was
shown to have the desired L-shape geometry, as from
acetophenones, by X-ray analysis.4 Extensive optimiza-
tion of the reactions conditions (separate dianion forma-
tion over 1 h, 1.7 equiv of diester, 22 h reaction time) led
to the formation of the pentacyclic xanthone in a much
improved 69% yield.
Scheme 5
This isocoumarin was next transformed in 75% yield into
methyl naphthoate 15a by using a procedure similar to
that developed by Roush and Murphy for their synthesis
of olivine.12 Treatment of this ester with methyllithium
and trimethylsilyl chloride13 then provided acetonaph-
thone 15b (59%; 30% overall from 10a).
OMe
OMe
O
1. LDA, (MeO)2CO
CO2R
1. NaH, n-BuLi
CO2Me
CO2Me
CO2Me
2. (MeO)2SO2,
2. (MeO)2SO2,
K2CO3
MeO
MeO
K2CO3
56%
65%
9
11
10a, R = Me
b, R = H
KOH,
EtOH
99%
Scheme 4
Synlett 2004, No. 15, 2693–2696 © Thieme Stuttgart · New York

LETTER
Synthesis of Dibenzoxanthenones
2695
OH
OMe
OMe
CO2Me
COMe
OMe
NaH, THF
+
∆
MeO
MeO
CO2Me
MeO
11
15b
OMe
MeO
O
OMe
HO HO
MeO
MeO
+
MeO
OH
O
MeO
OH
O
OMe
OMe
16
4
H2O-PhMe
180 °C
37% overall
Scheme 6
The polycondensation of the 2 fragments, diester 11 and The o-quinone methide substructure found in hypoxyxy-
ketone 15b, was effected under the optimized model-sys- lerone (1) could be generated in 18a–d through lithium
tem conditions to give a mixture of naphthonaphthone 16 aluminum hydride reduction, followed by in situ hydro-
and the desired xanthone 4 (Scheme 6); the mixture, by lytic rearrangement of the intermediate xanthols, which
heating in water–toluene at 180 °C in a closed vessel, gave derivatives 19a–d.18,19 The hypoxyxylerone ana-
could be transformed completely into xanthone 4 (37% logues 19b,d were found to inhibit topoisomerase I in
overall yield).14,15
vitro, albeit less so than the natural product.
Although the present convergent route to 4 is only slightly In summary, the anionic polycondensation developed by
higher yielding overall than the previous (11% vs 10%), it Kjaer and coworkers for acetophenones has been success-
is considerably more reproducible and rapid (7 steps vs 13 fully extended to acetonaphthones for a new, flexible
steps, longest linear sequence) and, furthermore, lends it- approach to hypoxyxylerone derivatives. Current efforts
self more readily to analog preparation.
are focused on the use of this convergent strategy to ac-
cess other analogs as well as the natural product itself.
To demonstrate the flexibility inherent in this approach, a
family of hypoxyxylerone analogs with OMe’s in lieu of
OH’s and a Me in place of the CH2OH group has been pre-
pared. The polycondensations were carried out with the
homophthalates 6 and 11 and the acetonaphthones 17a,b
(Scheme 7).16 The resulting pentacyclic xanthones (18a–
d), obtained in quite acceptable yields in view of the con-
ciseness of the approach, were all assigned the L-shape ge-
ometry based on the structures of derivatives 4 and 8 and
X-ray analysis of 18d.17
Acknowledgment
We thank Prof. P. Dumy for his interest in our work, Dr. C. Philouze
for the X-ray structure determinations, and Prof. J.-F. Riou for the
biological tests. We are also grateful to the Research Ministry and
Aventis for fellowship awards (to C. L. and E. C., respectively) and
the Université Joseph Fourier and the CNRS (UMR 5616, FR 2607)
for financial support.
R2
R2
R2
R2
OH R2
O
R1
O
O
R1
O
6 or 11
LiAlH4, THF;
aq HCl
R2
R1 OH
R1
O
17a, R2 = H
b, R2 = OMe
18a, R1 = H, R2 = H (40%)
18b, R1 = OMe, R2 = H (33%)
18c, R1 = H, R2 = OMe (20%)
19a, R1 = H, R2 = H (43%)
6 + 17a
19b, R1 = OMe, R2 = H (41%)
19c, R1 = H, R2 =OMe (40%)
19d, R1 = OMe, R2 =OMe (40%)
11 + 17a
6 + 17b
18d, R1 = OMe, R2 = OMe (42%)
11 + 17b
Scheme 7
Synlett 2004, No. 15, 2693–2696 © Thieme Stuttgart · New York

2696
E. Chevenier et al.
LETTER
The cooled solution was worked up in the usual way with
CHCl3 to afford a solid product, which was triturated with
Et2O and filtered to give 4 (79 mg, 37%) as an orange solid.
Mp 268–269 °C (CHCl3). IR (neat): 3411, 1649, 1620, 1602
cm–1. 1H NMR (300 MHz, CDCl3): d = 3.62 (s, 3 H), 3.89 (s,
3 H), 3.90 (s, 3 H), 3.98 (s, 3 H), 4.04 (s, 3 H), 5.16 (s, 2 H),
6.32 (d, J = 2.2 Hz, 1 H), 6.52 (d, J = 2.2 Hz, 1 H), 6.57 (d,
J = 2.2 Hz, 1 H), 6.70 (d, J = 2.2 Hz, 1 H), 6.98 (s, 1 H), 7.65
(s, 1 H). 13C NMR (75.4 MHz, CDCl3): d = 55.5 (CH3), 55.6
(CH3), 56.3 (CH3), 56.4 (CH3), 59.1 (CH3), 73.6 (CH2), 96.9
(CH), 97.9 (CH), 99.6 (CH), 99.8 (CH), 100.8 (CH), 104.4
(C), 108.2 (C), 109.7 (C), 112.4 (C), 119.7 (CH), 137.5 (C),
140.4 (C), 141.8 (C), 152.0 (C), 157.2 (C), 159.8 (C), 161.0
(C), 161.7 (C), 161.8 (C), 163.9 (C), 183.8 (C). MS (DCI,
NH3 + isobutane): m/z (%) = 477 (100) [MH+]. HRMS: m/z
calcd for C27H24O8: 476.1471. Found: 476.1479 (M+).
Xanthone 4 was identical in all respects with the material
prepared earlier2 through the homo-Fries route.
References
(1) (a) Edwards, R. L.; Fawcett, V.; Maitland, D. J.; Nettleton,
R.; Shields, L.; Whalley, A. J. S. J. Chem. Soc., Chem.
Commun. 1991, 1009. (b) Gimbert, Y.; Chevenier, E.;
Greene, A. E.; Massardier, C.; Piettre, A. EP 014025514,
2001.
(2) Piettre, A.; Chevenier, E.; Massardier, C.; Gimbert, Y.;
Greene, A. E. Org. Lett. 2002, 4, 3139.
(3) (a) Kjaer, D.; Kjaer, A.; Risbjerg, E. J. Chem. Soc., Perkin
Trans. 1 1983, 2815. (b) Kjaer, A.; Kjaer, D. Acta. Chem.
Scand. B 1982, 36, 417. (c) To the best of our knowledge,
bikaverin is the only natural product prepared to date
through the use of this chemistry.
(4) Crystal data for 8: C21H12O3 monoclinic. P21/n; a = 7.729 (1)
Å, b = 8.127 (3) Å, c = 23.067 (3) Å; b = 95.95 (1)°;
V = 1441.1 (5) Å3; D = 1.44 gcm–3; m = 0.96 cm–1;
l = 0.71073 Å. 2qmax = 48°. 2501 measured reflections, 2440
independent reflections; 220 parameters. Reflections/
parameters ratio: 6.8; R [I>1.1s(I)] = 5.8%; wR [all
data] = 5.8%. G. O. F. (all data) = 1.54. Full lists of
fractional atomic co-ordinates, bond lengths and angles, and
thermal parameters have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication
no. CCDC-247512.
(5) (a) Chiarello, J.; Joullié, M. Tetrahedron 1988, 44, 41.
(b) Barrett, A. G. M.; Moris, T. M.; Barton, D. H. R. J.
Chem. Soc., Perkin Trans. 1 1980, 2272.
(6) (a) Hauser, F. M.; Rhee, R. Synthesis 1977, 245.
(b) Hauser, F. M.; Rhee, R. P. J. Org. Chem. 1977, 42, 4155.
(c) Hauser, F. M.; Rhee, R. J. Am. Chem. Soc. 1977, 99,
4533.
(16) (a) Although acetonaphthone 17a was known,16b the
reported yield was only 3.5%. Therefore, a new route was
developed (Scheme 8, R = H; 30% overall yield).
Acetonaphthone 17b16c could be prepared through a route
analogous to that used for the preparation of acetonaphthone
15b (Scheme 5, N-methoxy-N-methylacetamide replaces
12); however, a route analogous to that used for 17a proved
superior (Scheme 8, R = OMe; 52% overall yield).
(b) Yang, N. C.; Lin, L. C.; Shani, S.; Yang, S. S. J. Org.
Chem. 1969, 34, 1845. (c) Dodd, J. H.; Garigipati, R. S.;
Weinreb, S. M. J. Org. Chem. 1982, 47, 4045.
R
R
MeO2C
CO2Me
(7) For an excellent review of modern methods for the synthesis
of naphthalenes, see: de Koning, C. B.; Rousseau, A. L.; van
Otterlo, W. A. L. Tetrahedron 2003, 59, 7.
(8) Graham, S. L.; Scholz, T. H. J. Org. Chem. 1991, 56, 4260.
(9) For similar transformations, see: Lewis, C. N.; Spargo, P. L.;
Staunton, J. Synthesis 1986, 944.
(10) Weinreb amides are superior to acid chlorides for this type
of reaction. See: Carpenter, T. A.; Evans, G. E.; Leeper, F.
J.; Staunton, J.; Wilkinson, M. R. J. Chem. Soc., Perkin
Trans. 1 1984, 1043; see also ref. 6c.
(11) Matsumoto, N.; Nakashima, T.; Isshiki, K.; Kuboki, H.;
Hirano, S.-I.; Kumagai, H.; Yoshioka, T.; Ishizuka, M.;
Takeuchi, T. J. Antibiotics 2001, 54, 285.
CH2(CO2Me)2
Pyridine, TiCl4
O
R
R
O
OH
R
1. MeSO3H
2. MeLi, TMSCl
R
17a, R = H
b, R = OMe
Scheme 8
(17) Crystal data for 18d: C26H22O7 triclinic. P-1; a = 10.404 (3)
Å, b = 11.124 (3) Å, c = 15.392 (6) Å; a = 67.91 (2)°, b =
77.29 (2)°, g = 66.97 (2)°. V = 1513.5 (9) Å3; D = 1.49 gcm–3;
m = 1.97 cm–1, l = 0.56083 Å; 2qmax = 35.7°. 3863 measured
reflections, 3863 independent reflections. 424 parameters.
Reflections/parameters ratio: 5; R [I>2s(I)] = 9.7%; wR [all
data] = 9.9%. G. O. F. = 2.22. The poor resolution is due to
problems associated with severely disordered CHCl3
molecules in the asymmetric cell and desolvation (capillary
tube). Full lists of fractional atomic co-ordinates, bond
lengths and angles, and thermal parameters have been
deposited with the Cambridge Crystallographic Data Centre
as supplementary publication no. CCDC-247513.
(12) Roush, W. R.; Murphy, M. J. Org. Chem. 1992, 57, 6622.
(13) (a) Rubottom, G. M.; Kim, C. J. Org. Chem. 1983, 48, 1550.
(b) Cooke, M. P. J. Org. Chem. 1986, 51, 951.
(14) Shi, J.; Zhang, X.; Neckers, D. C. J. Org. Chem. 1992, 57,
4418.
(15) Procedure for the Preparation of 4 – Polycondensation of
11 and 15b: To a stirred suspension of pentane-washed NaH
(79 mg, 3.29 mmol, from a 60% dispersion in mineral oil) in
anhyd THF (2.5 mL) under argon at 0 °C was added
dropwise a solution of 15b (130 mg, 0.45 mmol) in THF (5.0
mL). The mixture was stirred at 25 °C for 1 h before the
addition of a solution of homophthalate 11 (204 mg, 0.76
mmol) in THF (3.5 mL). The resultant mixture was stirred at
25 °C for 2 h and then heated at 80 °C for 20 h (pressure
tube). The solution was cooled and aq HCl (6 N) was added,
and the resultant mixture was extracted with CHCl3. After
the usual workup, the product was flash chromatographed
(SiO2, EtOAc in pentane, 30–50%) to give a mixture of 4 and
16 (12% and 25%, respectively, NMR), which was heated in
H2O–toluene 5:1 (12 mL) at 180 °C for 16 h (pressure tube).
(18) This procedure for accessing the o-quinone methides was
found to be more convenient than that previously used2
based on Saegusa dehydrosilylation.
(19) The corresponding unprotected derivatives 19b–d (R1,
R2 = H, OH) could also be secured in moderate yield through
treatment of the xanthones 18b–d with boron tribromide
prior to reduction–hydrolysis.
Synlett 2004, No. 15, 2693–2696 © Thieme Stuttgart · New York