Influence of the C(4)-C(3)-C=O dihedral angle of chiral NADH mimics on the stereoselectivity of reductions

Article abstract of DOI:10.1016/S0957-4166(02)00101-5

Herein we report the stereoselective synthesis of new chiral NADH mimics 2 and 13 of the benzo [b]-1,6-naphthryridine series. The synthesis of 2 and 13 relies upon a Friedlander-type condensation between the amino imine 3 and the piperidine-2,4-dione 4 bearing a stereogenic center at C(6). The resulting NADH models were involved in the reduction of methyl benzoylformate. A comparison of their performance with that of previously reported NADH mimics such as model 1, throws new light on the role played by the C(4)-C(3)-C=O dihedral angle (α) on the stereoselectivity of the hydride transfer.

Full text of DOI:10.1016/S0957-4166(02)00101-5

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 13 (2002) 227–232
Influence of the C(4)ꢀC(3)ꢀCꢁO dihedral angle of chiral NADH
mimics on the stereoselectivity of reductions
Jean-Luc Vasse, Vincent Levacher,* Jean Bourguignon and Georges Dupas
Laboratoire de Chimie Organique, Fine et He´te´rocyclique associe´ au CNRS, IRCOF-INSA. BP 08 F-76131,
Mont Saint Aignan Ce´dex, France
Received 25 January 2002; accepted 22 February 2002
Abstract—Herein we report the stereoselective synthesis of new chiral NADH mimics 2 and 13 of the benzo [b]-1,6-naphthryridine
series. The synthesis of 2 and 13 relies upon a Friedlander-type condensation between the amino imine 3 and the piperidine-2,4-
dione 4 bearing a stereogenic center at C(6). The resulting NADH models were involved in the reduction of methyl
benzoylformate. A comparison of their performance with that of previously reported NADH mimics such as model 1, throws new
light on the role played by the C(4)ꢀC(3)ꢀCꢁO dihedral angle (h) on the stereoselectivity of the hydride transfer. © 2002 Elsevier
Science Ltd. All rights reserved.
1. Introduction
show that the carboxamide group adopts a trans out-
of-plane orientation with a dihedral angle of about 30°
(h).1 It has been suggested that the stereospecificity of
the coenzyme originates from the trans out-of-plane
orientation adopted by the carboxamide group leading
to transfer of the Hsyn hydrogen with respect to the
carbonyl oxygen. An out-of-plane orientation of the
amide carbonyl may be regarded as a stereogenic unit,
i.e. a non-permanent chiral axis (C3ꢀCꢁO) which can
be partly responsible for the stereospecificity of reduc-
tions (Fig. 1).
In the area of enzymatic reduction, NADH coenzyme
plays an important role and has stimulated a great deal
of interest in the field of biomimetic chemistry. Numer-
ous chiral NADH models have been synthesized with a
view to developing new efficient chemo- and enantiose-
lective reducing agents and to elucidate the mechanism
of the stereospecific hydrogen transfer. In the enzymatic
process, it is known that depending on the nature of the
enzyme, either HR or HS at the C(4) position of the
dihydronicotinamide ring is stereospecifically trans-
ferred to a prochiral substrate. The most stable confor-
mation of free NADH exhibits a cis-conformation of
the carboxamide group (i.e. the CꢁO dipole is pointed
towards the N1 atom as depicted in Fig. 1). In contrast,
X-ray structures of the coenzyme bound to enzymes
To gain insight on how this conformational feature
may contribute to the stereochemical outcome of enzy-
matic reductions, numerous chiral NADH mimics
based on the configurational control of the C(3)ꢀCꢁO
chiral axis have been investigated. Among the most
Figure 1. Conformational changes of the nicotinamide moiety in NADH coenzyme when bound to enzymes.
* Corresponding author. Fax: +33 (0) 2 35 52 29 62; e-mail: vincent.levacher@insa-rouen.fr
0957-4166/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(02)00101-5

228
J.-L. Vasse et al. / Tetrahedron: Asymmetry 13 (2002) 227–232
by the C(3)ꢀCꢁO bond. As can be seen from the
literature, numerous axially chiral NADH mimics
based on the stereocontrol of the C(3)ꢀCꢁO chiral axis
have been reported by us and others. However, few
investigations have been devoted to the influence of the
C(4)ꢀC(3)ꢀCꢁO dihedral angle (h) on the enantioselec-
tive reduction of this class of NADH mimics.6
representative examples of this class of models are
Ohno’s2 and Vekemans’3 reagents which were shown to
be highly stereoselective in the reduction of methyl
benzoylformate [(S) 95% e.e.]. In the course of this
asymmetric reduction, a second chirality transfer occurs
whereby the axial chirality of the pyridinium salt is
restored with a high level of stereoselectivity [(aS) 95%
e.e.]. This intramolecular chirality transfer is consistent
with the ternary complex proposed by Vekemans in
which the amide carbonyl group and the transferring
hydrogen are syn-oriented (Fig. 2). However, the pres-
ence of an additional stereogenic unit at C(4) prevents
us from assessing the role played by the C(3)ꢀCꢁO
chiral axis in the good performances of these NADH
mimics (Fig. 2).
2. Results and discussion
In an attempt to fill in this gap, we undertook the
synthesis of model 2, which involves the carbonyl group
in a six-membered lactam ring. According to molecular
modeling, the most stable conformer would set both
substituents R1 and R2 in a staggered conformation as
observed in model 1. Although both models 1 and 2
have in common many conformational and configura-
tional features, model 2 exhibits a much smaller
C(4)ꢀC(3)ꢀCꢁO dihedral angle (h) close to 10°. Models
1 and 2 offer the opportunity to study, independent of
other factors, the influence of this dihedral angle on the
stereoselective properties of this class of models (Fig.
3).
We recently reported5 a new class of models in which
the carbonyl amide is incorporated into a lactam struc-
ture, thereby mimicking the trans out-of-plane confor-
mation adopted by the coenzyme when bound to an
enzyme (Fig. 3). Molecular modeling4 of model 1
revealed a dihedral angle of about 45°. The presence of
a stereogenic center on the lactam moiety ensures the
configurational control of the resulting C(3)ꢀCꢁO chiral
axis.5 The stereoselective synthesis of (4R)-deuterated
model (aS,S)-1 gave experimental proof of Dsyn transfer
to methyl benzoylformate. Model (aS,S)-1 afforded
deuterated (R)-methyl mandelate with up to 84% e.e
(R). The high level of asymmetric induction observed
was ascribed to the presence of this chiral axis defined
We intended to synthesize model 2 of the benzo[b]-1,6-
naphthryridine series to protect the 5,6-double bond
from unwanted side reactions and according to previ-
ous work in our group.7 The target model 2 was
prepared following a Friedlander quinoline synthesis.
O
Me
Hsyn
(S)
N
S
O
-
O
ClO4
PhCOCOOMe
H OH
Mg2+
N
Me
O
-
ClO4
Hsyn
O
Me
O
Me
Ph
COOMe
N
(aS) N
N
(S) 95% ee
Me
N+ Me
-
ClO4
Proposed ternary complex
(aS) 95% ee
Figure 2. NADH mimics with axial and central chirality reported by Ohno2 and Vekemans.3
HO
Ph
(R) 84% ee
R1
D
PhCOCOOMe
Mg(ClO4)2
O
PhCOCOOMe
Mg(ClO4)2
?
COOMe
O
Hanti
Dsyn
Hanti
Hsyn
R1
MeO
N
*
MeO
MeO
aS
N
aS
R2
(S)
R2
MeO
N
N
(S)
Hsyn
α
Hanti
R1
Six-membered ring
model 2
Seven-membered ring
model 1
O
R1= Bn; R2= Me
R1= i-Pr; R2= Me
N
R2
α=10−15°
α=45−50°
*
Figure 3. Design of new NADH mimics to study the relationship between the C4ꢀC3ꢀCꢁO dihedral angle (h) and the
stereochemical outcome of the reduction.

J.-L. Vasse et al. / Tetrahedron: Asymmetry 13 (2002) 227–232
229
The poor stability of ortho-aminobenzaldehydes
prompted us to test the Borsche modification8 using
imine 3.9 The piperidine-2,4-dione 4 was obtained via
Dieckmann cyclization. The rest of the synthesis fol-
lows the classical steps of quaternization and regioselec-
tive reduction of the corresponding quinolinium salt
(Scheme 1).
The Friedlander reaction was achieved in refluxing
ethanol in the presence of piperidine providing quino-
line 514 in 74% yield. The presence of the PMB group
offers an additional chelation site for the magnesium
ion, thereby preventing us from establishing a correla-
tion between the dihedral angle (h) and the stereoselec-
tive properties of the models. The PMB protective
group was thus cleaved under oxidative conditions to
give 10 in 72% yield.12a Treatment of quinoline 10 with
benzyl bromide furnished the benzylated quinoline 11
in 75% yield. Quaternization of the former was accom-
plished in nearly quantitative yield affording 12,15
which was subsequently reduced regioselectively under
classical conditions leading to model 216 (Scheme 3).
Methods for the preparation of piperidine-2,4-diones
are essentially based on Dieckmann cyclization.10 Only
few papers deal with the preparation of enantiopure
piperidine-2,4-diones bearing a stereogenic center at
C(6).11 The configurational control of this stereogenic
center was achieved by the known diastereoselective
conjugate addition of chiral lithium amide 6 on trans-
methyl crotonate.12 The corresponding b-amino ester 7
was obtained in 67% yield and 95% d.e. The absolute
configuration of the newly created stereogenic center
was assigned as (S) according to literature reports.12a
The resultant amino ester 7 was chemoselectively
debenzylated to afford 8 in 95% yield. Treatment of 8
with the acid chloride derived from ethyl malonate then
furnished 9 in 91% yield and Dieckmann cyclization led
to the desired piperidine-2,4-dione 413 in 69% yield
(Scheme 2).
The staggered conformation of 5 placing the methyl
group in a pseudo-axial position was deduced from
measurement of the coupling constants between the
C(3%) proton and the two C(4%) protons. Further sup-
port for this assignment came from a NOESY experi-
ment (Fig. 4). Molecular modeling4 confirms this
finding and shows a dihedral angle (h) of about 15°.
The same set of coupling constants was observed in
model 2, indicating that the methyl group occupies a
pseudo-axial position as well.
Model 2 was involved in the reduction of methyl ben-
zoylformate in the presence of magnesium perchlorate.
As can be seen in Scheme 4, (R)-methyl mandelate was
obtained with e.e. of 4%, whereas 1 led to (R)-methyl
mandelate with up to 84% e.e. under the same condi-
O
O
Hanti
Hsyn
R
R
R
R
N
N
N
Me
Friedlander
N
Me
Tol
5
Quaternization
Reduction
N
O
O
R=OMe
R2
R1
R1
R1
R1
PMB
Me
(i)
N
Model 2
N
α=10−15°
+
74%
PMB: p-methoxybenzyl
NH2
N
Me
O
Tol
O
Dieckmann
3
N
5: R2=PNB
10: R2=H
4
PMB
Me
R
72%
(ii)
(iii)
N
+
R1=OMe
O
R
NH2
75%
3
4
11: R2=Bn
O
O
Hanti
Hsyn
Scheme 1. Retrosynthetic analysis of model 2.
R1
R1
Bn
R1
R1
Bn
(v)
90%
N
(iv)
N
100%
N
+
Me
N
Me
TfO-
R
PMB
12
(R)
(i)
N
PMB
(S)
α=10−15°
Ph
N
Model 2
MeOOC
67%
Me
Li
6
d.e.>95%
Scheme 3. Reagents and conditions: (i) piperidine few drops,
D, 12 h; (ii) CAN, CH3CN/H2O: 9/1, rt, 4 h; (iii) KOH,
benzyl bromide, DMSO, 1 h; (iv) MeOTf, CH2Cl2, rt; (v)
Na2S2O4, Na2CO3, H2O.
Ph
7: R=
PMB: p-methoxybenzyl
(ii)
95%
8: R= H
O
O
PMB
Me
EtOOC
PMB
(iii)
91% MeOOC
N
(iv)
69%
N
O
Me
9
4
Scheme 2. Reagents and conditions: (i) trans-methyl croto-
nate, 30 min, −78°C; (ii) H2, Pd(OH)2, MeOH, rt, overnight;
(iii) EtOOCCH2COCl, NEt3, CH2Cl2, rt, 6 h; (iv) MeONa,
THF, D 2 h; then HCl (6N), D, 2 h.
Figure 4. Conformational analysis of 5.

230
J.-L. Vasse et al. / Tetrahedron: Asymmetry 13 (2002) 227–232
The mixture of diastereomers syn-13a and anti-13b was
involved in the reduction of methyl benzoylformate
under the same conditions as those applied with model
2. The progress of the reaction was monitored, at
1
regular intervals, by H NMR. After two days, methyl
mandelate was obtained in 80% yield with low enan-
tioselectivity (20% e.e.) in favor of the (R) enantiomer
(Fig. 6).
The graph in Fig. 6 clearly shows that only the major
diastereomer syn-13a is consumed as methyl mandelate
is formed. Since anti-13b does not participate in the
reduction of the substrate, this result strongly suggests
that only Hsyn in model 2 is transferred to methyl
Scheme 4. Reduction of methyl benzoylformate.
tions (Scheme 4). Since both models 1 and 2 possess the
same configurational features, this result clearly estab-
lished that changes in the dihedral angle (h) lead to
important variations in the stereoselectivity of reduc-
tions. One can wonder if the dihedral angle has an
influence either on the stereodifferentiation of both
diastereotopic C(4) protons or on the recognition of the
two faces of the prochiral substrate.
Models syn-13a and anti-13b bearing a methyl group at
C(4) may help to tackle this question. If both
diastereotopic hydrogen atoms are effectively differenti-
ated in model 2, only one of the two epimers syn-13a
and anti-13b should react with the substrate or at least
much faster. According to the stereoselective behavior
of reagent 1,5 one can assume that epimer syn-13a, with
H-4 in a syn position with respect to the carbonyl
group, is the most reactive (Fig. 5)
O
O
Me
N
O
R1
R1
R2
R1
R1
PMB
Me
N
N
(i)
55%
N
+
NH2
O
14
R1=OMe
R1
4
15: R2=PMB
67%
16: R2=H
Me
O
(ii)
Bn
(iv)
(iii)
75%
Bn
17: R2=Bn
(v)
R1
N
+
95%
TfO-
O
18
Me
Hanti
R1
R1
N
(S)
H
O
O
N
(S)
O
S
The desired tricyclic precursor 1517 was prepared in a
similar manner with 5, in 55% yield, by using
dimethoxy aminobenzophenone 14 in a Friedla¨nder
condensation under Fehnel conditions.18 For the same
reasons as those previously mentioned with model 2,
the PMB protecting group was cleaved and the amine
product 16 was subsequently benzylated to furnish 17
in 56% overall yield. Quaternization and regioselective
reduction of quinolinium salt 18 by means of sodium
dithionite, afforded 13a19 and 13b in 91% overall yield
and 80% d.e. The diastereoselectivity observed in this
former step can be rationalized considering an
approach of sodium dithionite to the opened face of the
quinolinium salt, followed by the rearrangement of the
resulting sulfinate intermediate.20 Although the stereo-
chemical aspect of this rearrangement has not been
studied in the literature, it could be assumed that it
occurs with retention of configuration. According to
these considerations, the major epimer formed would
be syn-13a (Scheme 5).
Me
anti-13b
R1
R1
Bn
80% de
91%
+
O
N
Hsyn
Me
R1
R1
Bn
N
N
(R)
N
(S)
syn-13a
anti-13b/syn-13a : 10/90
Scheme 5. Access to both epimers syn-13a and anti-13b.
Reagents and conditions: (i) H2SO4 (few drops), AcOH, reflux,
2 h; (ii) CAN, CH3CN/H2O: 9/1, rt, 4 h; (iii) KOH, benzyl
bromide, DMSO, 2 h; (iv) MeOTf, CH2Cl2, rt; (v) Na2S2O4,
Na2CO3, H2O.
HO
Ph
H
(i)
syn-13a (90%)
anti-13b (10%)
+
COOMe
80%
+
+
+
(R) 20% ee
Yield
100%
90%
Methyl
mandelate
+
syn-13a
+
+
Substrate
+
50%
O
O
+
Hanti
Me
+
Me Hsyn
+
MeO
MeO
Bn
MeO
MeO
Bn
+
N
N
+
+
anti-13b
Hours
10%
+
N
N
anti-13b
24h
48h
syn-13a
Figure 5. Difference on reactivity of both epimers 13a and
13b depending on the syn and anti position of H-4 with
respect to the carbonyl group.
Figure 6. Reduction of methyl benzoylformate with a 90/10
mixture of epimer syn-13a and anti-13b: (i) PhCOCOOMe/
Mg(ClO4)2/CH3CN/rt/48 h.

J.-L. Vasse et al. / Tetrahedron: Asymmetry 13 (2002) 227–232
231
benzoylformate, as was already demonstrated with
model 1. Consequently, decreasing the dihedral angle
(h) does not affect the stereodifferentiation of both
faces of the model but, has in return a dramatic effect
on the stereodifferentiation of both faces of the sub-
strate (Fig. 7a).
present study demonstrates that this chiral axis is effec-
tive in promoting a highly enantioselective Hsyn transfer
to methyl benzoylformate, only if it displays a fairly
large dihedral angle (h). Finally, this work has impor-
tant implications on the interpretation of the asymmet-
ric induction with previously reported chiral NADH
mimics bearing a carboxamide group at C(3.) In partic-
ular, these results raise the general question of the
participation of this carboxamide group, as a chiral
relay with this former class of models.21
This study also gives further understanding on the
mechanism of asymmetric induction with models previ-
ously reported by Ohno2 and Vekemans3 (Fig. 2, 7b).
Although this class of models possess the same configu-
rational features as those encountered in syn-13a, in
particular a stereogenic center at C(4), their perfor-
mances in terms of stereoselectivity are much more
superior (Figs. 2 and 7b). This observation demon-
strated that the stereogenic center at C(4) in Ohno’s
and Vekemans’ models is not responsible for the
stereodifferentiation of the substrate. Thus, in contrast
to models 2 and syn-13a, the dihedral angle h in Ohno’s
and Vekemans’ models close to 60° is sufficiently pro-
nounced so that the resultant chiral relay C(3)ꢀCꢁO
would exert efficient stereodifferentiation of both faces
of the prochiral substrate (Fig. 7b).
Acknowledgements
We thank the CNRS, the re´gion Haute Normandie and
the CRIHAN for technical and financial support.
References
1. (a) Ecklund, H.; Samama, J. P.; Jones, T. A. Biochem-
istry 1984, 23, 5982; (b) Skarzynski, T.; Moody, P. C. E.;
Wonacott, A. J. J. Mol. Biol. 1987, 193, 171; (c)
8
Almarsson, O.; Bruice, T. C. J. Am. Chem. Soc. 1993,
115, 2125–2138; (d) Wu, Y.-D.; Houk, K. N. J. Org.
Chem. 1993, 58, 2043–2045.
3. Conclusion
2. (a) Ohno, A.; Kashiwagi, M.; Ishihara, Y.; Ushida, S.;
Oka, S. Tetrahedron 1986, 42, 961–973; (b) Mikita, Y.;
Hayashi, K.; Mizukami, K.; Matsumoto, S.; Yano, S.;
Yamazaki, N.; Ohno, A. Tetrahedron Lett. 2000, 41,
1035–1038.
In summary, this work has allowed us to establish that
the efficiency of axially chiral NADH mimics based on
the out-of-plane orientation of the carbonyl amide is
highly dependent on the dihedral angle (h). We could
gain insight into the role of this resultant masked chiral
axis. It confers a different reactivity of the two
diastereotopic C(4) protons in favor of Hsyn with
respect to the amide carbonyl group. Moreover, the
3. (a) De Kok, P. M. T.; Bastiaansen, L. A. M.; van Lier, P.
M.; Vekemans, J. A. J. M.; Buck, H. M. J. Org. Chem.
1989, 54, 1313–1320; (b) Buck, H. M. Recl. Trav. Chim.
Pays-Bas 1996, 115, 329–332; (c) Vekemans, J. A. J. M.;
Boogers, J. A. F.; Buck, H. M. J. Org. Chem. 1991, 56,
10–16.
(a) Small dihedral angle (α) --> "off" position of the chiral relay
4. PCMODEL® 7.0 implementation of MM2 force-field.
COOMe
CERIUS 2®, MSI CVFF force-field.
O
R=H 2: (R) 4% ee
5. (a) Vasse, J. L.; Dupas, G.; Duflos, J.; Que´guiner, G.;
Bourguignon, J.; Levacher, V. Tetrahedron Lett. 2001, 42,
4613–4616; (b) Vasse, J. L.; Dupas, G.; Duflos, J.;
Que´guiner, G.; Bourguignon, J.; Levacher, V. Tetra-
hedron Lett. 2001, 42, 3713–3716.
6. (a) Be´dat, J.; Levacher, V.; Dupas, G.; Que´guiner, G.;
Bourguignon, J. Chem. Lett. 1996, 359–360; (b) Be´dat, J.;
Levacher, V.; Dupas, G.; Que´guiner, G.; Bourguignon, J.
Chem. Lett. 1995, 327–328.
R=Me syn-13a: (R) 20% ee
Hsyn
α
R
HO
Ph
H
α =15°
O
Bn
COOMe
N
Low e.e.
Me
*
(b) large dihedral angle (α) --> "on"position of the chiral relay
7. For previous work related to stable annelated NADH
models, see: (a) Dupas, G.; Levacher, V.; Bourguignon,
J.; Que´guiner, G. Heterocycles 1994, 39, 405–429; (b)
Levacher, V.; Dupas, G.; Que´guiner, G.; Bourguignon, J.
Trend Heterocyclic Chem. 1995, 4, 293–302; (c) Vitry, C.;
Vasse, J.-L.; Dupas, G.; Levacher, V.; Que´guiner, G.;
Bourguignon, J. Tetrahedron 2001, 57, 3087–3098.
8. (a) Borsche, W.; Ried, W. Liebigs Ann. Chem. 1943, 554,
269; (b) Borsche, W.; Barthenheier, J. Liebigs Ann. Chem.
1941, 548, 50.
O
Hsyn
Me
N
(R)
Ar
COOMe
N
Me
O
Ohno's and Vekemans' models
Hsyn
α
Me
N
HO
Ph
H
α =60°
O
COOMe
High e.e.
9. Porter, H. K. The Zinin Reduction of Nitroarenes. Org.
React. 1973, 20, 455–481.
Figure 7. Relationship between the dihedral angle (h) and the
stereoselective course of reduction.
10. (a) Ashton, M. J.; Hills, S. J.; Newton, C. G.; Taylor, J.
B.; Tondu, S. C. D. Heterocycles 1989, 28, 1015; (b)

232
J.-L. Vasse et al. / Tetrahedron: Asymmetry 13 (2002) 227–232
Micovic, I. V.; Roglic, G. M.; Ivanovic, M. D.; Dosen-
Micovic, L.; Kiricojevic, V. D.; Popovic, J. B. J. Chem.
Soc., Perkin Trans. 1 1996, 2041.
107.75, 121.92, 124.36, 127.85, 127.94, 128.89, 136.14,
139.48, 141.98, 151.95, 153.37, 159.27, 159.52.
1
16. Selected data for 2 H NMR (CDCl3, 200 MHz) l 1.29
(3H, d, J=6.5 Hz), 2.13 (1H, d, J=16.0 Hz), 2.50 (1H,
dd, J=15.6 Hz, J=4.5 Hz), 3.14 (3H, s), 3.43 (1H, m),
3.60 (1H, d, J=18.5 Hz), 3.84 (1H, d, J=18.5 Hz), 3.86
(3H, s), 3.87 (3H, s), 4.11 (d, 1H, J=15 Hz), 5.44 (d, 1H,
J=15 Hz), 6.42 (1H, s), 6.68 (1H, s), 7.31 (5H, m).
11. (a) Waldmann, H.; Braun, M.; Dra¨ger, M. Tetrahedron:
Asymmetry 1991, 2, 1231; (b) Davies, F.; Chao, B.; Fang,
T.; Szewczyk, J. M. Org. Lett. 2000, 2, 1041–1043; (c)
Murray, P. J.; Starkey, I. D. Tetrahedron Lett. 1996, 37,
1875–1878; (d) Leflemme, N.; Dallemagne, P.; Rault, S.
Tetrahedron Lett. 2001, 42, 8997–8999.
1
17. Selected data for 15 H NMR (CDCl3, 200 MHz) l 1.12
(3H, d, J=6.6 Hz), 2.90 (1H, dd, J=15.6 Hz and 1.7
Hz), 3.11 (3H, s), 3.37 (1H, dd, J=15.6 Hz, J=5.8 Hz),
3.79 (1H, m), 3.80 (3H, s), 4.03 (3H, s), 4.05 (3H, s), 4.07
(1H, d, J=14.6 Hz), 5.38 (1H, d, J=14.6 Hz), 6.88 (2H,
d, J=8.6 Hz), 7.30–7.35 (4H, m). 13C NMR (CDCl3, 50
MHz) l 18.38, 30.09, 39.78, 48.53, 49.92, 55.67, 56.45,
56.62, 103.28, 107.90, 114.47, 120.29, 123.94, 129.79,
130.50, 144.15, 145.62 149.86, 153.74, 155.01, 159.44,
164.59.
12. (a) Davies, S. G.; Ichihara, O. Tetrahedron Lett. 1998, 39,
6045–6048; (b) Davies, S. G.; Ichihara, O. Tetrahedron:
Asymmetry. 1991, 2, 183–186.
1
13. Selected data for 4: H NMR (CDCl3, 200 MHz) l 1.19
(3H, d, J=6.7 Hz), 2.46 (1H, dd, J=15.9 Hz and J=2.5
Hz), 2.65 (1H, dd, J=15.9 Hz and J=5.8 Hz), 3.38 (2H,
m), 3.74 (1H, m), 3.80 (3H, s), 4.04 (1H, d, J=14.7 Hz),
5.25 (1H, d, J=14.7 Hz), 6.88 (2H, d, J=8.7 Hz), 7.23
(2H, d, J=8.7 Hz). 13C NMR (CDCl3, 50 MHz) l 19.84,
46.50, 47.13, 48.04, 48.81, 55.63, 114.55, 129.11, 129.81,
159.60, 166.36, 204.21.
18. (a) Fehnel, E. A.; Deyrup, J. A.; Davidson, M. B. J. Org.
Chem. 1958, 23, 1996; (b) Fehnel, E. A. J. Org. Chem.
1966, 31, 2899.
1
14. Selected data for 5 H NMR (CDCl3, 200 MHz) l 1.18
1
19. Selected data for syn-13a H NMR (CDCl3, 200 MHz) l
(3H, d, J=6.6 Hz), 2.98 (1H, dd, J=16.1 Hz and J=1.9
Hz), 3.47 (1H, dd, J=16.1 Hz and J=5.9 Hz), 3.81 (3H,
s), 3.85 (m, 1H); 3.90 (3H, s), 4.02 (3H, s), 4.09 (1H, d,
J=14.6Hz), 5.48 (1H, d, J=14.6Hz), 6.89 (2H, d, J=8.5
Hz), 7.16 (1H, s), 7.31 (2H, d, J=8.5 Hz), 7.38 (1H, s),
8.74 (1H, s). 13C NMR (CDCl3, 50 MHz) l 18.80, 38.57,
48.08, 50.28, 55.61, 56.50, 56.57, 106.51, 107.53, 114.43,
121.32, 123.06, 129.73, 130.08, 135.45, 146.89, 150.11,
154.38, 154.38, 159.41, 163.56.
1.12 (3H, d, J=7.2 Hz), 1.28 (3H, d, J=6.6 Hz), 2.58
(1H, dd, J=16.6 Hz and 3.7 Hz), 2.68 (1H, dd, J=16.6
Hz and 5.9 Hz), 3.25 (3H, s), 3.55 (1H, m), 3.86 (3H, s),
3.88 (3H, s), 3.89 (1H, d, J=15.2 Hz), 4.23 (1H, q, J=6.5
Hz), 5.44 (1H, d, J=15.2 Hz), 6.47 (1H, s), 6.74 (1H, s),
7.18–7.34 (5H, m). 13C NMR (CDCl3, 50 MHz) l 18.24,
25.16, 30.49, 32.70, 33.41, 47.37, 48.69, 56.73 (2), 98.98,
104.56, 112.48, 122.04, 127.38, 128.13, 128.55, 134.09,
139.39, 143.37, 145.45, 147.79, 166.49.
1
15. Selected data for 12 H NMR (CDCl3, 200 MHz) l 1.13
20. For the rearrangement of the sulfinate intermediate, see:
(a) Biellman, J. F.; Callot, H. J.; Bull. Soc. Chem. Fr.
1968, 1154; (b) Caughey, W. S.; Schellenberg, K. A. J.
Org. Chem. 1966, 31, 1978.
21. Vitry, C.; Be´dat, J.; Prigent, Y.; Levacher, V.; Dupas, G.;
Salliot, I.; Que´guiner, G.; Bourguignon, J. Tetrahedron
2001, 57, 9101–9108.
(3H, d, J=6.6 Hz), 3.54 (1H, dd, J=18.0 Hz and 3.6
Hz), 3.65 (1H, dd, J=18.0 Hz and 6.3 Hz), 3.94 (1H, m),
4.00 (3H, s), 4.14 (3H, s), 4.16 (1H, d, J=15.0 Hz), 4.45
(3H, s), 5.36 (1H, d, J=15.0 Hz), 7.32 (5H, m), 7.42 (1H,
s), 7.69 (1H, s), 9.30 (1H, s). 13C NMR (CDCl3, 50 MHz)
l 19.31, 34.13, 40.41, 48.53, 49.10, 56.69, 57.99, 99.07,