The first total synthesis of (±)-annosqualine by means of oxidative enamide-phenol coupling: pronounced effect of phenoxide formation on the phenol oxidation mechanism

Article abstract of DOI:10.1016/j.tetlet.2006.08.028

The first total synthesis of a spiro-isoquinoline alkaloid, (±)-annosqualine, was established by employing an enamide-phenol coupling of a 1-methylene-1,2,3,4-tetrahydroisoquinoline derivative with a hypervalent iodine reagent, where the formation of the

Full text of DOI:10.1016/j.tetlet.2006.08.028

Tetrahedron Letters 47 (2006) 7301–7306
The first total synthesis of ( )-annosqualine by means of oxidative
enamide–phenol coupling: pronounced effect of phenoxide
formation on the phenol oxidation mechanism
Hiroki Shigehisa, Jun Takayama and Toshio Honda*
Faculty of Pharmaceutical Sciences, Hoshi University, Ebara 2-4-41, Shinagawa-ku, Tokyo 142-8501, Japan
Received 10 July 2006; revised 3 August 2006; accepted 10 August 2006
Available online 30 August 2006
Abstract—The first total synthesis of a spiro-isoquinoline alkaloid, ( )-annosqualine, was established by employing an enamide–
phenol coupling of a 1-methylene-1,2,3,4-tetrahydroisoquinoline derivative with a hypervalent iodine reagent, where the formation
of the phenoxide was recognized to be an essential step for the reaction of the phenolic hydroxyl group with the hypervalent iodine
reagent leading to the formation of the desired product.
Ó 2006 Elsevier Ltd. All rights reserved.
Annosqualine 1, a novel isoquinoline alkaloid with an
unprecedented skeleton bearing a spirocyclohexadi-
enone function, was isolated from the stems of Annona
squamosa in 2004 as a minor component, and was sup-
posed to be a biogenetic precursor of protoberberine
and oxoprotoberberine alkaloids.1 Although the struc-
ture of 1 was elucidated spectroscopically, its synthesis
and biological activity have not been studied yet (Fig. 1).
to the formation of a spirocyclohexadienone moiety,
was involved as the key step2 (Fig. 2).
In relation to a project directed at the synthesis of bio-
logically active natural products by employing aromatic
oxidation with a hypervalent iodine reagent,3–5 we are
interested in establishing a concise synthesis of the
unique isoquinoline alkaloid, annosqualine 1. Prior to
the synthesis of the natural product, we decided to inves-
tigate efficient and mild reaction conditions for the
oxidation of a readily available enamide 5 as follows
(Scheme 1).
Recently, we have developed a facile synthetic procedure
for a proaporphine alkaloid, ( )-stepharine, where an
oxidative enamide–phenol coupling of an isoquinoline
derivative with a hypervalent iodine reagent, iodobenz-
ene diacetate (PIDA) in trifluoroethanol (TFE) leading
Condensation of the known 6,7-dimethoxy-1-methyl-
3,4-dihydroisoquinoline hydrochloride 3 with 4-(tert-
butyldimethylsilyloxy)benzoyl chloride 2, prepared from
4-(tert-butyldimethylsilyloxy)benzoic acid,6 afforded
enamide 4. Since an attempted isolation of the pheno-
lic enamide 5, derived from 4 by desilylation with
OMe
MeO
N
O
HO
MeO
MeO
PhI(OAc)2
NH
(PIDA)
NaBH4
MeO
O
N
CF3
MeO
-COCF3
CF3CH2OH
(TFE)
Figure 1. Structure of annosqualine 1.
O
O
Keywords: Annosqualine; Spiro-isoquinoline alkaloid; Iodobenzene
diacetate; Enamide–phenol coupling; Phenoxide formation.
OH
stepharine
*
Corresponding author. Tel.: +81 3 5498 5791; fax: +81 3 3787
0036; e-mail: honda@hoshi.ac.jp
Figure 2. Our previous synthesis of proaporphine alkaloid.
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.08.028

7302
H. Shigehisa et al. / Tetrahedron Letters 47 (2006) 7301–7306
MeO
MeO
MeO
OR
HCl
N
COCl
N
3
MeO
O
TBSO
1.0 equiv. of 3
2
4 R = TBS
5 R = H
TBAF, THF
rt, 20 min
2.0 equiv. of 2, NEt3
CH2Cl2, rt, 30 min
66% from 3
MeO
MeO
MeO
MeO
PIDA, TFE
rt, 10 min
+
N
O
N
O
O
7
O
6
NaBH4
TFE
rt, 1 h
MeO
MeO
MeO
MeO
OH
N
O
HN
O
O
8
9
O
Scheme 1. Preparation of enamide 4 and its conversion to 7.
tetrabutylammonium fluoride (TBAF) in tetrahydrofu-
ran, resulted in the easy formation of the hydrolysis
product 9, the crude enamide 5, obtained from the
above reaction mixture by evaporation of the solvent,
was subjected to oxidation without further purification.
the oxidation gave the desired product in 90% yield. As
for this reason, we thought that there are two reactive
sites against the oxidant (PIDA), in the starting com-
pound 5, a phenolic oxygen and an enamide carbon,
which might make the oxidation troublesome. Although
two reactive sites were also present in the starting mate-
rial of our proaporphine synthesis, the enamide nitrogen
of the starting isoquinoline was protected with a trifluo-
roacetyl group, a strong electron-withdrawing group,
which might diminish the reactivity of the enamide car-
bon with the oxidant to afford the desired product in
high yield. To prove this hypothesis, we decided to
add a base to the starting enamide prior to its further
oxidation, since generation of the phenoxide by addition
of a base would be expected to increase its reactivity
against an oxidant. Thus, the starting enamide 5 was
treated with 1.0 equiv of n-butyllithium in HFIP at
0 °C,8 and the resulting mixture was reacted with
1.0 equiv of PIDA at room temperature to give the de-
sired spiro-enamide 7 in 38% yield. When this reaction
was carried out in the presence of 2.0 equiv of n-butyll-
ithium, the yield was improved to 78%. The results
obtained are summarized in Table 1.
First, we investigated the oxidation of 5 with the use of
PIDA as the oxidant, in TFE at room temperature, and
subsequent reduction of the presumed intermediate 6
with sodium borohydride in a one-pot procedure
according to our previous procedure;2 however, none
of the desired product 8 could be isolated under these
reaction conditions producing only decomposed prod-
ucts. However, we were pleased to be able to isolate ena-
mide 7 by careful examination of the reaction mixture
when this oxidation was conducted under the same reac-
tion conditions, without further treatment of an inter-
mediate with sodium borohydride, although the yield
was lower than 10%. Encouraged by this result, we next
focused our attention on searching for optimal condi-
tions for the oxidation, in which we decided to isolate
spiro-enamide 7 instead of its one-pot conversion to 8
by subsequent reduction.
It is well recognized that the use of a solvent with less
nucleophilicity gives better result in an oxidative pheno-
lic coupling. Thus, a similar oxidation of 5 with PIDA
was carried out in hexafluoroisopropanol (HFIP) as
the solvent,7 instead of TFE; however, the desired
spiro-enamide 7 was again obtained in a trace amount.
At this point, we had to figure out the reason why the
oxidation of 5 did not proceed smoothly to give the
desired product, compared to our previous work,2 where
It is noteworthy that the reaction temperature for the
preparation of the phenoxide would be required to be be-
low 4 °C due to the instability of the starting enamide.9
Moreover, HFIP was obviously better than TFE as the
solvent in this reaction. The necessity of 2 equiv of n-
butyllithium would be attributed to trapping of acetic
acid generated from the reagent during the reaction
process, in addition to the formation of the phenoxide
to increase its reactivity against the oxidant.

H. Shigehisa et al. / Tetrahedron Letters 47 (2006) 7301–7306
Table 1. Oxidation of enamide 4 to the spiro-enamide 7
7303
MeO
MeO
1. TBAF, THF, rt, 20 min.
2. PIDA (1.0 equiv), n-BuLi (x equiv)
solvent, rt, 10 min.
MeO
MeO
N
O
N
O
4
7
O
OTBS
Entry
n-BuLi (equiv)
Solvent
Yield (%)
1
2
3
4
5
6
7
8
9
0
TFE
TFE
TFE
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
ꢀ10
12
ꢀ10
ꢀ10
38
60
78
72
69
1.0
2.0
0
1.0
1.5
2.0
2.5
3.0
the intermediate, which, without isolation, was further
treated with sodium borohydride to provide the spiro-
cyclohexadienone 8 in 52% yield. Thus, we were able
to establish a one-pot synthetic procedure for 8 in
reasonable yield.
MeO
MeO
MeO
MeO
NaBH4, HFIP
rt, 1 h
N
O
N
O
91%
8
O
7
By establishing a synthetic route to the basic skeleton of
the natural product, we started the synthesis of anno-
squaline as follows. Our synthesis was launched with
the preparation of the known aldehyde 1411 by an alter-
native route in improved yield (Scheme 3).
O
Scheme 2. Reduction of enamide 7.
With the desired spiro-enamide 7 in hand, we focused
our attention on its reduction with sodium borohydride
to obtain the basic carbon framework of the natural
product. Again, it should be noted that the reduction
of 7 with sodium borohydride in methanol, acetic acid,
or TFE gave a complex mixture of products, whereas
the use of HFIP as the solvent gave the desired product
810 in 91% yield (Scheme 2). Based on consideration of
the above results, we attempted a one-pot preparation
of 8 from 5 again, as follows.
Bromination of monobenzylpyrogallol 1012 with 1,3-
dibromo-5,5-dimethylhydantoin (DBDMH) gave the
bromide 11 together with its regioisomer 12, in a ratio
of 8:1.13 The structure of 11 was unambiguously deter-
mined based on its NMR analysis including NOE exper-
iment. After methylation of phenolic oxygen with
dimethyl sulfate, the resulting bromide 13 was treated
with n-butyllithium, subsequently with N-formylmor-
pholine14 to provide aldehyde 14 in 90% yield. Since
we were able to achieve the synthesis of the desired
aldehyde 14 in good overall yield, preparation of 1-
methyl-3,4-dihydroisoquinoline derivative 20 was then
investigated (Scheme 4).
Treatment of 5, derived from 4 as above, with 2 equiv of
n-butyllithium in HFIP at below 4 °C, followed by oxi-
dation of the resulting phenoxide with PIDA afforded
DBDMH, CHCl3
0 ºC to rt, 3 h
(11:12 = 8:1)
OH
OH
OH
HO
Br
HO
HO
+
BnO
BnO
BnO
Br
10
11
12
n-BuLi
Me2SO4, K2CO3
acetone, reflux, 1 h
(77% from 10)
N
-formylmorpholine
OMe
THF, -78 ºC to 0 ºC
4 h (90%).
OMe
CHO
MeO
BnO
Br
MeO
BnO
14
13
Scheme 3. Preparation of aldehyde 14.

7304
H. Shigehisa et al. / Tetrahedron Letters 47 (2006) 7301–7306
NH4OAc
CH3NO2
80 ºC, 3 h
LiAlH4
THF, reflux
12 h
OMe
OMe
MeO
BnO
NO2
CHO
MeO
BnO
15
14
AcCl, NEt3
CH2Cl2, 0 ºC
1 h
OMe
OMe
Pd(OH)2/C, H2
EtOH, rt, 2 h
MeO
BnO
MeO
BnO
NHAc
NH2
100%
67% from 14
17
16
allyl bromide
K2CO3, acetone
reflux, 14 h
OMe
OMe
POCl3, benzene
reflux, 2 h
MeO
HO
MeO
O
NHAc
NHAc
86%
19
18
OMe
OMe
1.5 equiv. of 2, NEt3
CH2Cl2, reflux, 30 min
MeO
O
MeO
N
O
NHCl
O
56% from 19
20
21
OTBS
Scheme 4. Preparation of enamide 21.
Condensation of 14 with nitromethane in the usual man-
ner afforded nitrostyrene 15, which on reduction with
lithium aluminum hydride in refluxing THF furnished
phenethylamine 16. Acetylation of 16 with acetyl chlo-
ride gave amide 17 in 67% yield from 14. Benzyl ether
17 was transformed to its allyl ether 19, by two steps
including a catalytic hydrogenation and allylation of
the resulting phenol derivative 18 with allyl bromide in
the presence of potassium carbonate, on consideration
of the feasibility for its removal at the later stage of
the synthesis. Bischler–Napieralski cyclization of 19
with phosphoryl chloride in benzene gave the 3,4-di-
hydroisoquinoline hydrochloride 20, which, on treat-
ment with 4-(tert-butyldimethylsilyloxy)benzoyl chloride,
provided enamide 21 in 56% yield from 19.
Enamide–phenol coupling of 21, the key step in this
synthesis, was carried out as follows (Scheme 5).
Desilylation of 21 with tetrabutylammonium fluoride in
THF afforded the phenolic compound, which, without
isolation, was treated with 2 equiv of n-butyllithium in
OMe
1) TBAF, THF, rt, 20 min
2) 2 equiv.of n-BuLi, HFIP
then PIDA, 0 to 4 ºC, 10 min
MeO
O
NaBH3CN, HFIP
N
O
rt, 150 min
21
73%
86%
22
O
OMe
OMe
Pd(PPh3)2Cl2 (cat.)
Bu3SnH, AcOH, CH2Cl2
reflux, 22 h
MeO
O
MeO
HO
N
O
N
O
98%
1
O
23
O
Scheme 5. Synthesis of annosqualine 1.

H. Shigehisa et al. / Tetrahedron Letters 47 (2006) 7301–7306
7305
Iwata, M.; Kiyono, Y.; Egi, M.; Kita, Y. J. Am. Chem.
Soc. 2003, 125, 11235–11240.
6. Stem, A. J.; Swenton, J. S. J. Org. Chem. 1989, 54, 2953–
2958.
7. Begue, L. P.; Bonnet-Delpon, D.; Crousse, B. Synlett
2004, 1–29.
8. Potassium carbonate could also be used for the oxidation;
however, the yield (up to 20%) was found to be much
lower than that of the case of n-butyllithium. The reaction
could be carried out in a homogeneous phase by the use of
n-butyllithium. Similar aromatic oxidation of phenolic
compounds with a hypervalent iodine reagent in the
presence of sodium hydride was reported, although the
reason for the formation of the corresponding phenoxides
was not mentioned in the literature. See: (a) de Sousa, J.
D. F.; Rodrigues, J. A. R.; Abramovitch, R. A. J. Am.
Chem. Soc. 1994, 116, 9745–9746; (b) Rodrigues, J. A. R.;
Abramovitch, R. A.; de Sousa, J. D. F.; Leiva, G. C.
J. Org. Chem. 2004, 69, 2920–2928.
hexafluoroisopropanol (HFIP). To this mixture was
added iodobenzene diacetate (PIDA) at below 4 °C for
10 min to give spiro-enamide 22 successfully, in 73%
yield from 21. Sodium borohydride reduction of 22 in
HFIP gave the reduction product 2315 in 60% yield,
whereas the use of sodium cyanoborohydride as the
reducing agent could improve the formation of 23 to
86% yield.
Finally, deprotection of the allyl group of 23 with a cat-
alytic amount of bis(triphenylphosphine)palladium
dichloride and tributyltin hydride16 afforded ( )-anno-
squaline 1,17 mp 222–223 °C [(+)-natural annosqualine
was isolated as a syrup], in 98% yield. The spectroscopic
data (1H and 13C NMR) of the synthesized compound
were identical with those provided by Professor Wu
and Yang.
9. Treatment of n-BuLi with HFIP probably generated
lithium hexafluoroisopropoxide, which on further treat-
ment with 5 gave the corresponding phenoxide. These
processes should be carried out at below 4 °C, due to the
instability of the phenoxide. When these processes were
conducted at room temperature, the starting 3,4-dihydro-
isoquinoline 3 was recovered as the degradation product.
10. Selected data for 8: Mp 223–224 °C. FT-IR (film) mmax
In summary, we are able to demonstrate the versatility
of enamide–phenol coupling by its application to the
first total synthesis of a naturally occurring spiro-iso-
quinoline alkaloid, annosqualine 1. The strategy devel-
oped here would be applicable to the synthesis of
various types of alkaloids, and further extension of this
strategy is under investigation in this laboratory.
1700, 1660, 1520, 1260, 1230, 1100 cmꢁ1 1H NMR
;
(500 MHz, CDCl3) d 2.27 (dd, 1H, J = 12.7, 9.3 Hz, 9a-
H), 2.76 (dd, 1H, J = 16.1, 2.6 Hz, 4-H), 2.81 (dd, 1H,
J = 12.7, 6.5 Hz, 9b-H), 2.97 (ddd, 1H, J = 16.1, 11.6,
6.1 Hz, 4-H), 3.16 (br dt, 1H, 3-H), 3.87 (s, 3H, OMe),
3.89 (s, 3H, OMe), 4.34 (ddd, 1H, J = 13.2, 6.1, 1.9 Hz, 3-
H), 5.00 (br t, 1H, 1-H), 6.42 (dd, 1H, J = 10.0, 2.9 Hz, 13
or 15-H), 6.45 (dd, 1H, J = 10.0, 2.9 Hz, 13 or 15 H), 6.56
(s, 1H, 8-H), 6.67 (s, 1H, 5-H), 6.70 (dd, 1H, J = 10.0,
2.9 Hz, 16-H), 6.99 (dd, 1H, J = 10.0, 2.9 Hz, 12-H); 13C
NMR (125 MHz, CDCl3) d 28.0 (4-C), 38.4 (3-C), 40.0 (9-
C), 53.4 (10-C), 53.6 (1-C), 56.0 (OMe), 56.1 (OMe), 107.3
(8-C), 111.8 (5-C), 125.3 (40-C), 127.5 (80-C), 130.5 (16 or
17-C), 131.8 (16 or 17-C), 144.4 (12-C), 148.0 (16-C), 148.4
(2C, 6 and 7-C), 168.3 (11-C), 185.3 (14-C); MS [EI+] m/z
325 (M+); HR-MS [EI(+)] calcd for C19H19NO4 (M+):
325.1314; found 325.1323.
Acknowledgments
We greatly acknowledge Professors Wu and Yang,
Kaohsiung Medical University, Taiwan, for their gener-
ous gift of H and 13C NMR spectra of annosqualine.
1
We also thank Central Glass Co., Ltd., for providing
HFIP. This work was supported by a Grant-in-Aid from
the Ministry of Education, Culture, Sports, Science and
Technology of Japan.
References and notes
1. Yang, Y.-L.; Chang, F.-R.; Wu, Y.-C. Helv. Chim. Acta
2004, 87, 1392–1399.
11. Chen, C.-M.; Fu, Y.-F.; Yang, T.-H. J. Nat. Prod. 1995,
58, 1767–1771.
2. Honda, T.; Shigehisa, H. Org. Lett. 2006, 8, 657–659.
3. (a) Mizutani, H.; Takayama, J.; Honda, T. Tetrahedron
Lett. 2002, 43, 2411–2414; (b) Mizutani, H.; Takayama, J.;
Honda, T. Synlett 2005, 328–330.
12. Dietl, F.; Gierer, G.; Merz, A. Synthesis 1985, 626–631.
13. Bromination of 1-benzyloxy-2,3-dimethoxybenzene under
the same reaction conditions as described in Scheme 1 gave
bromide 11 and its regioisomer 12 in a ratio of ca. 1:1.
14. Formylation of 13 with n-BuLi and DMF gave the desired
aldehyde in 40–60% yield, and the corresponding primary
alcohol was isolated in 30–40% yield, although the
mechanism for its formation still remains obscure.
15. Selected data for 23: Mp 171–172 °C; FT-IR (film) mmax
1700, 1660, 1520, 1100 cmꢁ1; 1H NMR (270 MHz, CDCl3)
d 2.25 (dd, 1H, J = 12.6, 9.4 Hz), 2.66–2.72 (m, 1H), 2.77
(dd, 1H, J = 12.6, 6.4 Hz), 2.91–2.98 (m, 1H), 3.10 (dt,
1H, J = 11.7, 4.6 Hz), 3.89 (s, 3H), 3.90 (s, 3H), 4.30–4.37
(m, 1H), 4.57 (br d, 2H), 4.95 (br dd, 1H), 5.30 (br dd,
1H), 5.42 (br dd, 1H), 5.99–6.13 (m, 1H), 6.39 (m, 3H),
6.70 (dd, 1H, J = 10.2, 3.0 Hz), 6.97 (dd, 1H, J = 10.2,
3.0 Hz); 13C NMR (67.8 MHz, CDCl3) d 22.2, 37.9, 39.7,
53.3, 53.5, 60.7, 60.8, 69.9, 105.0, 117.8, 120.0, 130.4,
131.0, 131.7, 132.9, 141.7, 144.3, 147.9, 151.4, 151.7, 168.1,
185.2; MS [EI+] m/z 382 (M+); HR-MS [EI(+)] calcd for
C22H24NO5 (M+): 382.1654; found 382.1672.
4. For reviews, see: (a) Ochiai, M. Rev. Heteroat. Chem.
1989, 2, 92–111; (b) Moriarty, R. M. Synthesis 1990, 431–
447; (c) Moriarty, R. M.; Vaid, R. K.; Koser, G. F. Synlett
1990, 365–383; (d) Varvoglis, A. The Organic Chemistry of
Polycoordinated Iodine; VCH: New York, 1992; (e) Kita,
Y.; Tohma, H.; Yakura, T. Trends Org. Chem. 1992, 3,
113–128; (f) Stang, P. J.; Zhdankin, V. V. Chem. Rev.
1996, 96, 1123–1178; (g) Varvogolis, A. Hypervalent Iodine
in Organic Synthesis; Academic Press: San Diego, 1997; (h)
Kitamura, T.; Fujiwara, Y. Org. Prep. Proc. Int. 1997, 29,
409–458; (i) Zhdankin, V. V.; Stang, P. J. Chem. Rev.
´
2002, 102, 2523–2584; (j) Rodrıguez, S.; Wipf, P. Synthesis
2004, 2767–2783; (k) Wirth, T. Angew. Chem., Int. Ed.
2005, 44, 3656–3665.
5. For recent outstanding reports on natural product syn-
thesis by using hypervalent iodine reagents, see: (a)
Tohma, H.; Harayama, Y.; Hashizume, M.; Iwata, M.;
Egi, M.; Kita, Y. Angew. Chem., Int. Ed. 2002, 41, 348–
350; (b) Tohma, H.; Harayama, Y.; Hashizume, M.;
16. Corey, E. J.; Gin, D. Y.; Kania, R. S. J. Am. Chem. Soc.
1996, 118, 9202–9203.

7306
H. Shigehisa et al. / Tetrahedron Letters 47 (2006) 7301–7306
17. Selected data for the synthetic 1: Mp 222–223 °C. FT-IR
(film) mmax 1700, 1680 cmꢁ1 1H NMR (500
3.1 Hz), 6.39 (dd, 1H, J = 10.0, 3.1 H), 6.47 (d, 1H,
J = 0.6 Hz), 6.87 (dd, 1H, J = 10.0, 3.1 Hz), 7.22
(dd, J = 10.0, 3.1 Hz); 13C NMR (125 MHz, CD3OD)
d 23.2, 39.3, 40.4, 55.0, 55.4, 60.9, 61.0, 108.6, 119.3,
130.5, 132.0, 133.5, 141.3, 147.7, 151.2, 151.3,
152.4, 170.2, 187.7; MS [EI+] m/z 341 (M+); HR-MS
[EI(+)] calcd for C19H19NO5 (M+): 341.1263; found
341.1287.
;
MHz, CD3OD) d 2.30 (dd, 1H, J = 12.5, 9.3 Hz), 2.69
(dddd, 1H, J = 16.2, 11.3, 6.2, 1.1 Hz), 2.82 (dd, 1H,
J = 12.8, 6.8 Hz), 2.90 (ddd, 1H, J = 16.2, 4.3, 2.3 Hz),
3.15 (ddd, 1H, J = 13.2, 11.3, 4.6 Hz), 3.81 (s, 3H), 3.83
(s, 3H), 4.17 (ddd, 1H, J = 13.2, 6.2, 2.3 Hz), 4.61 (s, 1H),
5.06 (dd, 1H, J = 9.3, 6.8 Hz), 6.38 (dd, 1H, J = 10.0,