Stereospecific total synthesis of (-)-8-epi-hyperaspine

Article abstract of DOI:10.1016/j.tetlet.2003.09.149

Condensation of protected δ-hydroxy-β-amino ester 7 with a β-keto ester provides vinylogous urethane 8, which is cyclized under the action of t-BuOK followed by decarboxylation to afford enone 12. Hydrogenation of 12 or its N,O-diprotected derivative 13 g

Full text of DOI:10.1016/j.tetlet.2003.09.149

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 8609–8612
Stereospecific total synthesis of (−)-8-epi-hyperaspine
Dawei Ma* and Wei Zhu
State Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences, 354 Fenglin Lu, Shanghai 200032, China
Received 17 June 2003; revised 4 September 2003; accepted 10 September 2003
Abstract—Condensation of protected d-hydroxy-b-amino ester 7 with a b-keto ester provides vinylogous urethane 8, which is
cyclized under the action of t-BuOK followed by decarboxylation to afford enone 12. Hydrogenation of 12 or its N,O-diprotected
derivative 13 gives 2,6-cis-disubstituted piperdines. Using these intermediates, (−)-8-epi-hyperaspine is synthesized.
© 2003 Elsevier Ltd. All rights reserved.
Hyperaspine is a new member of ladybird alkaloids,
which was isolated from Hyperaspia campestris, a num-
ber of an as yet uninvestigated tribe.1 This compound
possesses a 3-oxoquinolizidine skeleton, which was not
discovered among over 50 known ladybird alkaloids.
By using 2D NMR and MS methods, Braekman and
his co-workers established its structure as shown in
Figure 1. However, its absolute configuration was not
determined. Obviously, this molecule can be con-
structed using piperidine A as the key intermediate.
During the investigation on the synthesis from enan-
tiopure b-amino acid derivatives,2 we have developed
an efficient method to assemble 4-hydroxy-2,6-disubsti-
tuted piperdines diastereoselectively.2i It was our hope
to apply this methodology to the synthesis of hyper-
aspine. The studies thus undertook are reported here.
98%. Next, hydrogenolysis with Pearlman’s catalyst
was employed to remove all the benzyl protecting
groups and the resultant hydroxyamine was treated
with t-butyl dimethylsilyl chloride to give 7 in 77%
yield for two steps.
With the b-amino ester 7 in hand, we planned to
condense it with a b-keto ester and then carry out an
intramolecular Dieckmann reaction to obtain the
cyclization product using the procedures as described
before.2i However, we found that the previous reaction
As outlined in Scheme 1, our synthesis started from
(S)-3-hydroxybutanoic acid ethyl ester, which was pre-
pared from ethyl acetate using baker’s yeast mediated
reduction in light petroleum.3 Treatment of this b-
hydroxyl ester with benzyl 2,2,2-trichloroacetimidate
under acid condition afforded the benzyl ether 4.4 After
the ester moiety of 4 was reduced with diisobutylalu-
minium hydride to give the crude aldehyde, Wittig
reaction was carried out to provide a,b-unsaturated
ester 5 in 90% yield. Subjected the ester 5 to a
diastereoselective Michael addition with lithium (S)-N-
benzyl-a-methylbenzylamide according to Davis’s pro-
cedure provided b-amino ester 6 in 85% yield.5 Only
Figure 1. Structure of hyperaspine and its retrosynthetic analy-
sis.
1
one isomer was determined by H NMR, which indi-
cated that the diastereoselectivity of this step was over
* Corresponding author. Tel.: 86-21-64163300; fax: 86-21-64166128;
e-mail: madw@pub.sioc.ac.cn
Scheme 1.
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.09.149

8610
D. Ma, W. Zhu / Tetrahedron Letters 44 (2003) 8609–8612
conditions2i for these two steps were not suitable for the
present substrate. Because the TBDMS protecting
group could not survive at acid-catalyzed azeotropic
removal of water condition,2i we decided to condense 7
Treatment of 9 with aqueous NaOH in refluxing etha-
nol provided the decarboxylation and desilylation
product 12. After the free hydroxyl group of 12 was
reprotected with TBDMSCl, its amine group was pro-
tected with (Boc)2O mediated with n-BuLi to afford 13
in 90% yield. We expected that the installation of this
Boc group would help us to get a trans-2,6-dialkylpipe-
ridine in hydrogenation step because direct hydrogena-
with 3-oxooctanoic acid ethyl ester under solvent-free
2d
,
condition with the assistance of MgSO4 and 4 A MS.
This mild condition allowed us to obtain the vinylogous
urethane 8 in 85% yield. The cyclization step of 8 was
found to be a difficult task and several methods were
thus investigated (Scheme 2). The results are summa-
rized in Table 1.
tion of 12 would give
a cis-2,6-dialkylpiperidine
through the conformation B (Fig. 2) as demonstrated
before.2i The similar selectivity change has been
observed by Comins and his co-workers during the
synthesis of indolizidine alkaloids.6 The strong A(1,3)
strain between the alkyl substituent at C-2 and N-pro-
tecting group was used to explain the selectivity as
indicated in conformation C of Figure 2, which forced
C-2 substituent into an axial orientation thereby result-
ing in attack from the C-2 substituent face.6 One could
logically think if the hydrogenation of 13 occurred in a
similar manner the desired trans-2,6-dialkylpiperidine
would be obtained. Thus, hydrogenation of 13 over
10% Pd/C in MeOH was carried out and a single
diastereoisomer 147 was obtained in 95% yield. After
cleavage of all the protecting groups of 14 with tri-
fluoroacetic acid in methylene chloride, reprotection
was carried out with 37% formaldehyde to provide a
bicyclic product 15.8 The stereochemistry of 14 was
easily determined by the NOE correlation as indicated
in Scheme 3. After careful analysis, we disappointingly
found that 14 was a cis but not trans-2,6-dialkylpipe-
ridine. This conclusion was further confirmed by fol-
lowing two evidences: (1) the presence of Bohlman
bands in the FTIR spectrum of 15 strongly supported
that 15 exists in a trans fused ring conformation as
indicated in Scheme 3;9 (2) Pd/C-catalyzed hydrogena-
tion of 12 followed by treatment with formaldehyde
and Dess–Martin oxidation to oxidate the resultant
hydroxyl group in hydrogenation step also produced
15.
Initially, sodium ethoxide mediated Dieckmann reac-
tion in refluxing ethanol was attempted.2i We found
that at this condition the reaction worked but gave the
desired product 9 in less than 20% yield (entry 1).
Considerable amount of enamine 10 was isolated,
which implied that the b-elimination of 8 to form 10
and 11 occurred at the present condition. The similar
result was observed when the reaction was taken place
in refluxing t-BuOH under the action of t-BuOK (entry
4). In order to depress the b-elimination we decided to
run this reaction at lower temperature. Accordingly,
LDA or NaH was employed as the base to carry out
the reaction at 0°C or room temperature. In these two
cases no reaction occurred (entries 2 and 3). However,
when the base was switched to t-BuOK the reaction
worked in THF at 0°C to give 9 in good yield (entry 5).
After optimization of experimental conditions, the yield
was further improved to 79% by adding 8 in one
portion to a refluxing solution of 2.5 equiv. of t-BuOK
in THF and quickly quenching the reaction (in 2 min)
after the addition (entry 6). By this mean, the b-elimi-
nation of 8 was minimized.
Although the reason for selectivity in hydrogenation of
13 was not clear, a possible explanation was proposed
based on a conformation of 13 as demonstrated in
Figure 3. Through this conformation stereoelectroni-
cally preferred axial attack took place to give a kineti-
cally favored chairlike transition state and finally
provide cis-2,6-dialkylpiperidine 14.10
In order to change the stereoselectivity, reduction of 13
with sodium borohydride or sodium cyanoborohydride,
and hydrogenation of 13 with other catalysts such as
Pt/C were attempted. However, in all cases only cis-2,6-
dialkylpiperidine was determined, which allowed us to
Scheme 2.
Table 1. Intramolecular Dieckmann reaction of 8
Entry
Reaction condition
Isolated yield of 8 (%)
1
2
3
4
5
6
NaOEt/EtOH, rt–reflux
LDA/THF, −78°C to 0°C
NaH/THF, 0°C to rt
KOBu-t/t-BuOH, reflux
KOBu-t/THF, 0°C
<20
a
–
–
a
<20
67
79
KOBu-t/THF, reflux
Figure 2. Conformation analysis of 2,3-dihydro-4-pyridones
during the addition reactions.
a The starting material was recovered.

D. Ma, W. Zhu / Tetrahedron Letters 44 (2003) 8609–8612
8611
Acknowledgements
The authors are grateful to the Chinese Academy of
Sciences, National Natural Science Foundation of
China (grant 29725205 and 29972049), and Qiu Shi
Science & Technologies Foundation for their financial
support.
References
1. Lebrun, B.; Braekman, J.-C.; Daloze, D.; Kalushkov, P.;
Pasteels, J. M. Tetrahedron Lett. 2001, 42, 4621–4623.
2. (a) Pu, X.; Ma, D. J. Org. Chem. 2003, 68, 4400–4405; (b)
Ma, D.; Xia, C.; Jiang, J.; Zhang, J.; Tang, W. J. Org.
Chem. 2003, 68, 442–451; (c) Ma, D.; Ma, N. Tetrahedron
Lett. 2003, 44, 3963–3965; (d) Ma, D.; Pu, X.; Wang, J.
Tetrahedron: Asymmetry 2002, 13, 2257–2260; (e) Ma, D.;
Zhu, W. Org. Lett. 2001, 3, 3927–3929; (f) Ma, D.; Xia,
C.; Jiang, J.; Zhang, J. Org. Lett. 2001, 3, 2189–2191; (g)
Wang, Y.; Ma, D. Tetrahedron: Asymmetry 2001, 12,
725–729; (h) Ma, D.; Sun, H. J. Org. Chem. 2000, 65,
6009–6016; (i) Ma, D.; Sun, H. Org. Lett. 2000, 2,
2503–2505; (j) Ma, D.; Sun, H. Tetrahedron Lett. 2000,
41, 1947–1950; (k) Ma, D.; Sun, H. Tetrahedron Lett.
1999, 40, 3609–3612; (l) Ma, D.; Zhang, J. Tetrahedron
Lett. 1998, 39, 9067–9068.
Scheme 3.
3. Medson, C.; Smallridge, A. T.; Trewhella, M. A. Tetra-
hedron: Asymmetry 1997, 7, 1049–1054.
4. Keck, G. E.; Murry, J. A. J. Org. Chem. 1991, 56,
6606–6611.
5. (a) Davies, S. G.; Ichihara, O. Tetrahedron: Asymmetry
1991, 2, 183–186; (b) Davies, S. G.; Ichihara, O.; Walters,
I. A. S. J. Chem. Soc., Perkin Trans. 1 1994, 1141–1147.
6. Comins, D. L.; LaMunyon, D. H.; Chen, X. J. Org.
Chem. 1997, 62, 8182–8187.
Figure 3. Stereochemical course during the hydrogenation of
13.
7. Selected data: [h]2D0 −2.1 (c 1.23, CHCl3); 1H NMR
(CDCl3, 400 MHz) l 4.76 (br s, 1H), 4.56 (br s, 1H),
3.90–3.84 (m, 1H), 2.68–2.62 (m, 2H), 2.43 (dt, J=15.3,
1.4 Hz, 1H), 2.33 (dd, J=15.3, 1.9 Hz, 1H), 1.86–1.78 (m,
1H), 1.61–1.58 (m, 3H), 1.49 (s, 9H), 1.43 (br s, 6H), 1.19
(d, J=6.0 Hz, 3H), 0.88–0.86 (m 12H), 0.05 (s, 6H); 13C
NMR (CDCl3, 75 MHz) l 209.2, 154.9, 80.4, 66.1, 52.9,
49.5, 45.7, 43.9, 43.7, 37.2, 31.7, 28.6 (3C), 26.8, 26.0
(3C), 24.0, 22.7, 18.1, 14.1, −4.0, −4.6; ESI-MS m/z 442
(M+H)+; ESI-HRMS calcd for C24H47NNaO4Si 464.3167
(M+Na)+, found 464.3174.
synthesize the 8-epi-hyperaspine at this moment. Thus,
reduction of 15 with LS-Selectride in THF at −78°C
produced an axial alcohol,11 which was reacted with
pyrrole 2-carboxylic acid chloride under the action of
triethylamine to give 8-epi-hyperaspine 1612 in 66%
yield (Scheme 4).
In conclusion, we have developed a stereocontrolled
route (12 linear steps and 14.8% overall yield) for
synthesizing (−)-8-epi-hyperaspine, which will be useful
for structure–activity relationship studies of hyper-
aspine. In addition, the chemistry presented here should
be of benefit for synthesizing other polysubstituted
piperidines.
8. Selected data: [h]2D0 −5.1 (c 1.1, CHCl3); 1H NMR
(CDCl3, 300 MHz) l 4.86 (d, J=7.8 Hz, 1H), 3.60 (d,
J=7.8 Hz, 1H), 3.60–3.55 (m, 1H), 2.47–2.31 (m, 5H),
1.68–1.47 (m 5H), 1.27–1.20 (m, 9H), 0.91–0.86 (m, 3H);
13C NMR (CDCl3, 75 MHz) l 207.8, 83.4, 73.0, 60.0,
58.7, 47.7, 46.3, 40.5, 32.5, 32.0, 23.6, 22.4, 21.0, 13.9; MS
m/z 239 (M+); HRMS calcd for C14H25NO2 239.1885
(M+), found 239.1915.
9. For a review, see: Crabb, T. A.; Newton, R. F.; Jackson,
D. Chem. Rev. 1971, 71, 109–126.
10. (a) Deslongchamps, P. Stereoelectronic Effects in Organic
Chemistry; Pergamon Press: Oxford, 1983; (b) Stevens,
R. V. Acc. Chem. Res. 1984, 17, 289–296.
11. Bardot, V.; Gardette, D.; Gelas-Mialhe, Y.; Gramain, J.;
Remuson, R. Heterocycles 1998, 48, 507–518.
Scheme 4.

8612
D. Ma, W. Zhu / Tetrahedron Letters 44 (2003) 8609–8612
0.90–0.85 (m, 3H); 13C NMR (CDCl3, 75 MHz) l 160.6,
123.1, 115.2, 115.0, 83.8, 73.7, 68.5, 55.6, 53.0, 40.5, 37.1,
35.7, 32.7, 32.7, 30.1, 24.3, 23.0, 21.6, 14.2; ESI-MS m/z
335 (M+H)+; HRMS calcd for C19H30N2O3 335.2329
(M+H)+, found 335.2332.
12. Selected data: [h]2D0 −8.0 (c 0.35, MeOH); 1H NMR
(CDCl3, 300 MHz) l 9.19 (br s, 1H), 7.00–6.98 (m, 1H),
6.96–6.94 (m, 1H), 6.31–6.30 (m, 1H), 5.28–5.27 (m, 1H),
4.83 (d, J=7.9 Hz, 1H), 3.60 (d, J=7.9 Hz, 1H), 3.57–
3.55 (m, 1H), 2.52–2.42 (m, 2H), 2.04–1.22 (m, 17H),