Total syntheses of (-)-hanishin, (-)-longmide B, and (-)-longmide B methyl ester via a novel preparation of N-substituted pyrrole-2-carboxylates

Article abstract of DOI:10.1021/ol203433c

A novel preparation of N-substituted pyrrole-2-carboxylates has been developed based upon 1,3-dipolar cycloaddition and a conventional hydrogenolysis. By using this method as the key step, total syntheses of natural alkaloids (-)-hanishin, (-)-longmide B,

Full text of DOI:10.1021/ol203433c

ORGANIC
LETTERS
2012
Vol. 14, No. 4
1062–1065
Total Syntheses of (ꢀ)-Hanishin,
(ꢀ)-Longmide B, and (ꢀ)-Longmide B
Methyl Ester via a Novel Preparation of
N-Substituted Pyrrole-2-Carboxylates
Guolin Cheng, Xinyan Wang,* Hailin Bao, Chuanjie Cheng, Nan Liu, and Yuefei Hu*
Department of Chemistry, Tsinghua University, Beijing 100084, P. R. China
wangxinyan@mail.tsinghua.edu.cn; yfh@mail.tsinghua.edu.cn
Received December 22, 2011
ABSTRACT
A novel preparation of N-substituted pyrrole-2-carboxylates has been developed based upon 1,3-dipolar cycloaddition and a conventional
hydrogenolysis. By using this method as the key step, total syntheses of natural alkaloids (ꢀ)-hanishin, (ꢀ)-longmide B, and (ꢀ)-longmide B
methyl ester were accomplished in the highest overall yields, respectively.
Pyrrole-2-carboxylic acid derivatives are a large family
of natural alkaloids.1 Most brominated pyrrole-2-carbox-
amides were isolated from marine sponges, such as
(ꢀ)-hanishin (1),2 (ꢀ)-longmide B (2),3 and (ꢀ)-longmide
B methyl ester (3)4 (Figure 1). These alkaloids not only have
a novel chiral dihydropyrrolo[1,2-a]pyrazin-1-one skeleton
but also showed biologically important properties. For
example, (ꢀ)-hanishin (1) was cytotoxic toward NSCLC-
N6 human nonsmall-cell-lung carcinoma (IC50 9.7 μg/mL);
(ꢀ)-longmide B (2) displayed activity against African try-
Figure 1. Three bioactive brominated pyrrole-2-carboxamides.
panosome (IC50 1.53 μ/mL); and (()-longmide B methyl
ester (3) exhibited cytotoxic activity against P-388 lympho-
cytic leukemia cells (ED50 30 μg/mL), respectively. There-
The first racemic total syntheses of 1ꢀ3 were accom-
plished a year after (ꢀ)-hanishin (1) was first isolated.5
fore, they are highly attractive targets for the total synthesis.
Thereafter a number of enantioselective total syntheses
were reported via chiral pool6 and asymmetric catalysis7
strategies. These reported routes clearly indicated that
(1) For selected references, see: (a) Bellina, F.; Rossi, R. Tetrahedron
2006, 62, 7213–7256. (b) Walsh, C. T.; Garneau-Tsodikova, S.; Howard-
Jones, A. R. Nat. Prod. Rep. 2006, 23, 517–531. (c) Handyab, S. T.;
Zhang, Y. Org. Prep. Proc. Int. 2005, 37, 411–445.
(2) Mancini, I.; Guella, G.; Amade, P.; Roussakis, C.; Pietra, F.
efficient total syntheses of 1ꢀ3 must resolve two major
Tetrahedron Lett. 1997, 38, 6271–6274.
(3) (a) Scala, F.; Fattorusso, E.; Menna, M.; Taglialatela-Scafati, O.;
(5) (a) Sun, X.-T.; Chen, A. Tetrahedron Lett. 2007, 48, 3459–3461.
(b) Banwell, M. G.; Bray, A. M.; Willis, A. C.; Wong, D. J. New J. Chem.
1999, 23, 687–690.
Tierney, M.; Kaiser, M.; Tasdemir, D. Mar. Drugs 2010, 8, 2162–2174.
(b) Cafieri, F.; Fattorusso, E.; TaglialatelaEScafati, O. J. Nat. Prod.
1998, 61, 122–125.
(6) Patel, J.; Pelloux-Le0on, N.; Minassian, F.; Valle0e, Y. J. Org.
Chem. 2005, 70, 9081–9084.
ꢀ
(4) (a) Reddy, N. S.; Venkateswarlu, Y. Biochem. Syst. Ecol. 2000, 28,
1035–1037. (b) Umeyama, A.; Ito, S.; Yuasa, E.; Arihara, S.; Yamada,
T. J. Nat. Prod. 1998, 61, 1433–1434.
(7) (a) Trost, B. M.; Osipov, M.; Dong, G. J. Am. Chem. Soc. 2010,
132, 15800–15807. (b) Trost, B. M.; Dong, G. Org. Lett. 2007, 9, 2357–
2359.
r
10.1021/ol203433c
Published on Web 02/08/2012
2012 American Chemical Society

issues: (a) efficient construction of the pyrrole-2-carbox-
ylate skeleton; (b) efficient introduction of the chiral CꢀN
bond. Herein, we would like to report a highly practical
route for the total synthesis of 1ꢀ3, in which a novel
method for pyrrole synthesis has been established and
the enantiopure N-substituted pyrrole-2-carboxylate is
constructed conveniently.
In our recent work, the catalytic hydrogenolysis of
3-substituted 2-isoxazolines was employed as the key step
in the total synthesis of various natural alkaloids.15 When
3-alkyl-2-isoxazolines are hydrogenolyzed in acidic aque-
ous MeOH, a β-hydroxy ketone is obtained16 and Raney-
Ni or Pd/C are used as catalysts. Based on these experi-
ments, we proposed a novel method for the preparation of
N-substituted pyrrole-2-carboxylates shown in Scheme 1.
Initially, ethyl 2-chloro-2-(hydroxyimino)acetate (HONd
C(Cl)CO2Et) and N-Cbz-allylamine (5) would react in a
1,3-dipolar cycloaddition to yield 5-aminomethyl-2-isoxa-
zoline-3-carboxylate (6). Then, the hydrogenolysis of 6
would give a cyclization product 8 by an intramolecular
attack of the amine on the ketone in 7, due to the ketone
being activated by its R-carboxylate group. Finally, the
intermediate 8 would be aromatized to yield the expected
product 9.
In the total syntheses of natural alkaloids containing a
structure of dihydro-pyrrolo[1,2-a]pyrazin-1-one, chiral
N-substituted pyrrole-2-carboxylates serve as important
precursors. Therefore, many imaginative methods have
been developed for their preparation, such as through the use
of intramolecular SN2 reactions of chiral secondary
chlorides,8 the Mitsunobu reaction of chiral secondary
alcohols,9 asymmetric intramolecular N-Michael addition,10
or asymmetric allylic alkylation.7,11 However, these methods
usually afford enantioenriched or diastereoenriched products
and satisfactory results are only obtained in a few cases.
Therefore, to obtain enantiopure products, traditional
PaalꢀKnorr pyrrole synthesis has been used for this purpose.
As shown in Figure 2, PaalꢀKnorr pyrrole synthesis uses 4a
or 4b as 1,4-diketone precursors and can directly yield
enantiopure products in one step,12 but the methods are
limited by the exceedingly difficult preparation of 4a or 4b.13
The commercially available 2,5-dimethoxytetrahydrofuran
(4c) could give enantiopure N-substituted pyrroles smoothly,
but many steps are required for introduction of the
2-carboxylate groups (with low efficiency).6,14 Thus, there is a
great need to develop a more efficient and general protocol
for the synthesis of enantiopure N-substituted pyrrole-2-
carboxylates on a laboratory scale.
Scheme 1. A Proposed Novel Method for the Preparation of N-
Substituted Pyrrole-2-carboxylates
Thus, N-Cbz-allylamine 5a (R = H) was employed as a
model substrate and the Cbz-group would play two roles.
First, it could activate the amine group to allow efficient
and chemoselective N-allylation for easy preparation of
the analogues (R ¼ H) of 5a. Second, it could be removed
by catalytic hydrogenolysis to unmask an NH2 group. As
shown in Scheme 2, 5a smoothly underwent a cycloaddi-
tion with HONdC(Cl)CO2Et to give desired 6a in 97%
yield. To our delight, instead of the expected intermediate
8a, the hydrogenolysis of 6a directly gave ethyl pyrrole-2-
carboxylate (9a) as the final product in 20% yield.
Figure 2. Three precursors in PaalꢀKnorr pyrrole synthesis.
(8) Mukherjee, S.; Sivappa, R.; Yousufuddin, M.; Lovely, C. J. Org.
Lett. 2010, 12, 4940–4943.
(9) Laha, J. K.; Cuny, G. D. J. Org. Chem. 2011, 76, 8477–8482.
(10) (a) Kwon, S.-H.; Lee, H.-J.; Cho, C.-W. Bull. Korean Chem. Soc.
2011, 32, 315–318. (b) Bandini, M.; Bottoni, A.; Eichholzer, A.; Miscione,
G. P.; Stenta, M. Chem.;Eur. J. 2010, 16, 12462–12473. (c) Dickson,
D. P.; Wardrop, D. J. Org. Lett. 2009, 11, 1341–1344. (d) Feldman, K. S.;
Saunders, J. C. J. Am. Chem. Soc. 2002, 124, 9060–9061. (e) Feldman,
K. S.; Saunders, J. C.; Wrobleski, M. L. J. Org. Chem. 2002, 67, 7096–
7109.
Scheme 2. A Two-Step Preparation of Pyrrole-2-carboxylate
(9a)
(11) Trost, B. M.; Dong, G. Chem.;Eur. J. 2009, 15, 6910–6919.
(12) (a) Wehn, P. M.; Du Bois, J. Angew. Chem., Int. Ed. 2009, 48,
3802–3805. (b) Sircar, I.; Winters, J R. T.; Quin, J., III; Lu, G. H.; Major,
T. C.; Panekt, R. L. J. Med. Chem. 1993, 36, 1735–1745.
(13) For the preparation of 4, see: (a) Crestia, D.; Guerard, C.; Bolte,
J.; Demuynck, C. J. Mol. Catal. B: Enzym. 2001, 11, 207–212. For the
preparation of 5, see: (b) Clauson-Kaas, N.; Limborg, F. Acta Chem.
Scand. 1952, 6, 551–555.
(14) (a) Yoshimitsu, T.; Ino, T.; Tanaka, T. Org. Lett. 2008, 10, 5457–
5460. (b) Mashiko, T.; Hara, K.; Tanaka, D.; Fujiwara, Y.; Kumagai,
N.; Shibasaki, M. J. Am. Chem. Soc. 2007, 129, 11342–11343. (c)
Nenajdenko, V. G.; Reznichenko, A. L.; Balenkova, E. S. Tetrahedron
2007, 63, 3031–3041. (d) Demir, A. S.; Subasia, N. T.; Sahin, E.
Tetrahedron: Asymmetry 2006, 17, 2625–2631.
Since7a(R = H) and 8a(R = H) werethe intermediates
in the conversion of 6a to 9a and their formation and
Org. Lett., Vol. 14, No. 4, 2012
1063

reaction would be influenced by acid,16 we evaluated dif-
ferent acid additives in the reaction. As shown in Table 1,
aq. HCl and aq. H2SO4 were better acid reagents than
B(OH)3 (entries 1ꢀ3) and HOAc gave the best results
(entry 4). When 6a was hydrogenolyzed at 40 °C, 9a was
obtained in 95% yield after 3 h (entry 5) and higher
temperatures were not necessary (entry 6). The results in
entries 7ꢀ9 proved that both MeOH and H2O were
essential and a 5:1 ratio was the best combination.
amines, the conversions of 6hꢀ6n to 9hꢀ9n needed longer
reaction times (24 h) and gave moderate to good yields. As
was expected, the enantiopure N-substituted pyrrole-2-
carboxylate 9o was obtained in 93% yield from the corre-
sponding chiral 2-isoxazoline 6o,17 which was not the case
by any N-alkylation methods. It was interesting to observe
that one isoxazoline ring in N,N-bis(2-isoxazoline-3-
carboxylate) 6p was cleaved to give 9p in 70% yield and
the other one stayed intact under these conditions.
Scheme 3. N-Substituted Pyrrole-2-carboxylates 9aꢀ9p Pre-
Table 1. Effects of Solvents, Acids, and Temperatures on the
Hydrogenolysis of 6a
pared
temp time yield of
entry
solvent
acid
(°C)
(h)
9a (%)a
1
2
3
4
5
6
7
8
9
MeOH/H2O (5:1)
MeOH/H2O (5:1)
MeOH/H2O (5:1)
MeOH/H2O (5:1)
B(OH)3
aq. HCl
aq. H2SO4
HOAc
25
25
25
25
40
60
40
40
40
12
12
12
12
3
20
38
43
67
MeOH/H2O (5:1) HOAc
95
MeOH/H2O (5:1)
MeOH/H2O (1:1)
MeOH
HOAc
HOAc
HOAc
HOAc
3
72
12
12
12
90
trace
trace
H2O
a The isolated yields were obtained.
To test the scope of this method, different N-substituted
N-Cbz-allylamines (6bꢀ6p) were prepared in 90ꢀ97%
yield. As shown in Scheme 3, under the standard hydro-
genolytic conditions, all the substrates 6aꢀ6g gave the
corresponding pyrroles 9aꢀ9g in excellent yield. Among
them, the conversion of 6e into 9e was especially important
because any 2-aminoethanol or 3-aminopropan-1-ol with-
out protection of the hydroxyl group could not be con-
verted into the corresponding N-(2-hydroxyethyl)pyrrole
or N-(3-hydroxypropyl)pyrrole by PaalꢀKnorr pyrrole
synthesis. Due to the weak nucleophilic ability of aromatic
a Reaction time was 3 h. bReaction time was 24 h.
(15) (a) Cheng, G.; Wang, X.; Zhu, R.; Shao, C.; Xu, J.; Hu, Y.
J. Org. Chem. 2011, 76, 2694–2700. (b) Su, D.; Wang, X.; Shao, C.; Xu,
J.; Hu, Y. J. Org. Chem. 2011, 76, 188–194.
(16) (a) Curran, D. P. J. Am. Chem. Soc. 1982, 104, 4024–4026. (b)
Kozikowski, A. P.; Stein, P. D. J. Am. Chem. Soc. 1982, 104, 4023–4024.
(17) After the TBS protective group in 9o was removed, the produced
alcohol compound 18 was acylated by (S)-l-[(4-methylphenyl)-
sulphonyl]-2-pyrrolidinecarbonyl chloride to give the corresponding
diastereomeric ester 19 in 94% yield as a single product (determined
by the spectra of 1H and 13C NMR) (see Supporting Information).
Thus, a novel preparation of N-substituted pyrrole-2-
carboxylates (9) was established by the conventional
hydrogenolysis of 5-aminomethyl-2-isoxazoline-3-carbo-
xylates (6). This transformation in fact was a one-pot five-
step process that included deprotection of N-Cbz, cleavage
of the NꢀO bond, hydrolysis of imine to the ketone,
intramolecular attack of the amino group on a ketone,
and aromatization by loss of two H2O molecules. By using
enantiopure amines as substrates, the enantiopure N-sub-
stituted pyrrole-2-carboxylates can be prepared con-
veniently.
Deductively, no racemization occurred during the preparation and trans-
formations of the compounds 6o, 9o, or 18. In fact, no racemization of the
enantiopure N-substituted pyrrole-2-carboxylates was reported in published
articles to date.6,12a,14aꢀ14d
Then, the new method was employed in the total synthe-
sis of (ꢀ)-hanishin (1), (ꢀ)-longmide B (2), and (ꢀ)-long-
mide B methyl ester (3). According to the routine
1064
Org. Lett., Vol. 14, No. 4, 2012

procedure, the commercially available (S)-ethyl 3-(N-Cbz-
amino)-4-hydroxybutanate (10) was initially converted
into the corresponding azide 11 in one pot. However,
NaH-promoted N-allylation of 11 gave a N,N-diallyl
product 12 instead of the expected product 13, accompa-
nied by cleavage of the CꢀN bond. As shown in Scheme 4,
it may be caused by a NaH-promoted retro-Michael
addition through an intramolecular E1cB elimination
mechanism.
(3) were accomplished in 7, 8, and 9 steps (from the starting
material 10) in 44%, 43%, and 43% overall yield, respec-
tively.
Scheme 5
Scheme 4. NaH-Promoted retro-Michael Addition
Thus, the relatively weak base LiHMDS was employed
and the N-allylic product 13 was obtained in moderate
yield (Scheme 5). Then 13 was converted into the key
precursor 14 by reduction of its azide with PPh3 followed
by N-Boc protection in one pot. As was expected, the 1,3-
dipolar cycloaddition between 14 and HONdC(Cl)CO2Et
gave2-isoxazoline 15as a mixture ofdiastereoisomers (1:1)
in excellent yield under mild conditions (the diastereoi-
somers would produce the same product in the next step).
When the suspension of 15 and Pd/C (10 wt %) in aqueous
MeOH and HOAc was stirred for 3 h at 40 °C under a
hydrogen atmosphere (balloon), the N-substituted pyr-
role-2-carboxylate 16 was obtained in 92% yield. Accord-
ing to the literature method,6 16 was regioselectively
brominated using 2 equiv of NBS to yield the 4,5-dibro-
mopyrrole 17 in 85% yield.
Since intermediate 17 has two different carboxylates, its
amidization may generate a mixture of γ-lactam and δ-
lactam. Luckily, when the N-deprotected product of 17
was heated at reflux in toluene for 1 h in the presence of
Et3N, (ꢀ)-hanishin (1) was obtained as the sole product in
93% yield. Experiments proved that the weak base Et3N
and nonpolar toluene were essential for this high chemos-
electivity. By simple saponification of 1 in solution of
NaOHinaqueousEtOH, (ꢀ)-longmideB (2) was obtained
in 98% yield. After 2 was treated with the solution of HCl
in MeOH for 2 h, it took an esterification to give (ꢀ)-
longmide B methyl ester (3) in almost quantitative yield.
Thus, the total syntheses of natural alkaloids (ꢀ)-hanishin
(1), (ꢀ)-longmide B (2), and (ꢀ)-longmide B methyl ester
In conclusion, chiral N-substituted pyrrole-2-carboxy-
lates are important precursors in the synthesis of alkaloids
containing the structure of dihydropyrrolo[1,2-a]pyrazin-
1-one. However, PaalꢀKnorr pyrrole synthesis usually
shows low efficiency and chemoselectivity, while other
methods normally afford enantioenriched or diastereoen-
riched products. In this article, a novel preparation of N-
substituted pyrrole-2-carboxylates has been developed and
applied in the total synthesis of several natural products.
Further works will be reported in due course.
Acknowledgment. This work was supported by the Na-
tional Natural Scientific Foundation of China (21072112).
Supporting Information Available. Experiments, char-
acterization, 1H and 13C NMR spectra for products
intermediates 6aꢀp, 9aꢀp, 11, 13ꢀ17, 19, and the alka-
loids 1ꢀ3. This material is available free of charge via the
Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 4, 2012
1065