Organocatalytic asymmetric assembly reactions for the syntheses of carbohydrate derivatives by intermolecular Michael-Henry reactions

Article abstract of DOI:10.1073/pnas.1003350107

Given the significance of carbohydrates in life, medicine, and industry, the development of simple and efficient de novo methods to synthesize carbohydrates are highly desirable. Organocatalytic asymmetric assembly reactions are powerful tools to rapidly

Full text of DOI:10.1073/pnas.1003350107

Organocatalytic asymmetric assembly reactions for
the syntheses of carbohydrate derivatives by
intermolecular Michael-Henry reactions
Hisatoshi Uehara, Ritsuo Imashiro, Gloria Hernández-Torres, and Carlos F. Barbas III1
The Skaggs Institute for Chemical Biology and the Departments of Chemistry and Molecular Biology, The Scripps Research Institute, 10550 North Torrey
Pines Road, La Jolla, CA 92037
Edited by David W. C. MacMillan, Princeton University, Princeton, NJ, and accepted by the Editorial Board May 29, 2010 (received for review March 17, 2010)
Given the significance of carbohydrates in life, medicine, and indus-
try, the development of simple and efficient de novo methods to
synthesize carbohydrates are highly desirable. Organocatalytic
asymmetric assembly reactions are powerful tools to rapidly
construct molecules with stereochemical complexity from simple
precursors. Here, we present a simple and robust methodology
for the asymmetric synthesis of pyranose derivatives with talo-
and manno- configurations from simple achiral precursors through
organocatalytic asymmetric intermolecular Michael–Henry reaction
sequences. In this process, (tert-butyldimethylsilyloxy)acetalde-
hyde 1 was successfully utilized in two ways: as a donor in a highly
selective anti-Michael reaction and as an acceptor in a consecutive
Henry reaction. Varied nitroolefins served as Michael acceptors and
varied aldehydes substituted for 1 as Henry acceptors providing for
the construction of a wide range of carbohydrates with up to 5
stereocenters. In these reactions, a catalyst-controlled Michael re-
action followed by a substrate-controlled Henry reaction provided
3,4-dideoxytalose derivatives 6 in a highly stereoselective manner.
The Henry reaction was affected by addition of a simple base such
as triethylamine: A complex chiral base was not necessary. 3,4-Di-
deoxymannose derivatives 7 were produced by simply changing
the base from triethylamine to 1,8-diazabicyclo[5.4.0]undec-7-
ene. Extension of this methodology to a syn-Michael initiated
sequence was also successful. A mechanistic discussion is provided
to explain the unusual substrate-induced stereoselectivity of the
Henry reaction.
organocatalytic approach has been a driving force in the field
(14, 17–23).
Recently, we reported highly selective anti-Michael reactions
of (tert-butyldimethylsilyloxy)acetaldehyde 1 to form γ-nitroalde-
hydes 4 catalyzed by primary amine-thiourea 3 (24). This type of
catalyst provides for enamine-based activation of the aldehyde
while enforcing configurational control of enamine geometry. To-
gether with hydrogen bonding activation of β-nitroalkenes pro-
vided by the thiourea (25–29) functionality of the catalyst, this
catalyst effectively merges two key functionalities in organocata-
lysis. The α-oxyaldehyde structure in the Michael product 4 sug-
gested that successive Henry reactions of 4 to the parent aldehyde
1 could produce highly functionalized nitroalcohol 5, which might
exist as its cyclized 3,4-dideoxypyranose form 6 as shown in
Scheme 1. An asymmetric assembly reaction of this type would
link three substrates through the formation of two new C-C bonds
while installing four contiguous asymmetric centers. We were en-
couraged to explore this idea by development of several asym-
metric Henry reactions (30), including intermolecular Michael-
intramolecular Henry tandem reactions (31–33), an iminium
mediated intermolecular Michael–Henry sequence (34) and
Michael-aza-Henry reactions (35, 36). Here we demonstrate an
organocatalytic intermolecular one-pot Michael–Henry reaction
through enamine catalysis. As shown in Scheme 1, two stereocen-
ters at C2 and C3 position in 6 were controlled with near perfec-
tion by the anti-Michael aldehyde reaction [up to 98∶2
diastereomeric ratio (dr) and 99% enantiomeric excess (ee)].
The challenge was to link this reaction with an intramolecular
Henry reaction to produce a single product with defined C4
and C5 stereocenters. Herein we present our solution to this chal-
lenge and present a simple and robust methodology for syn-
thesis of pyranose derivatives with talo- and manno- configura-
tions through organocatalytic intermolecular Michael–Henry
reaction sequences.
amine-thiourea catalyst ∣ asymmetric reaction ∣ carbohydrates ∣
Michael reaction ∣ organocatalysis
arbohydrates are one of the most important classes of organic
C
molecules and play diverse and essential roles in life, medi-
cine, and industry. Since Emil Fischer’s structural elucidation and
synthesis of carbohydrates more than a century ago (1), carbohy-
drate synthesis has continued to challenge synthetic chemists.
Robust, simple, direct, and highly stereoselective methods to car-
bohydrate synthesis remain largely elusive and the development
of such methodologies is a driving force in synthetic chemistry.
Indeed, our discovery of the proline catalyzed intermolecular
aldol reaction (2, 3) and other related reactions (4, 5) were made
possible by our development of antibody aldolases as synthetic
tools for carbohydrate synthesis (6, 7). In the decade since this
discovery, organocatalysis has emerged as a promising route to
a wide range of chiral molecules (4, 5, 8–11). Our studies in
organocatalysis prompted us (5, 12–14) and later others (15, 16)
to develop cascade reactions and one-pot synthetic approaches
toward the synthesis of complex molecules containing multiple
stereocenters with the aim of producing robust and operationally
simple approaches to the synthesis complex asymmetric mole-
cules like carbohydrates. We have classified reactions of this type
broadly as organocatalytic asymmetric assembly reactions
because they provide for the asymmetric assembly of multiple
substrates into higher order products with stereochemical com-
plexity. The preparation of carbohydrates based on this type of
Results and Discussion
We envisioned that if the Michael reaction of aldehyde 1 with β-
nitrostyrene 2a was carried out in the presence of additional base,
the Michael product 4a would react with remaining aldehyde 1.
As a starting point, we used triethylamine as a second catalyst
(Method A in Table 1, entry 1). The Michael reaction followed
by the Henry reaction preceded stereoselectively to provide pro-
duct 6a with the D-talo-configuration as a major product in good
yield with only small amounts of the D-manno-isomer 7a. Only
Author contributions: C.F.B. designed research; H.U., R.I., and G.H.-T. performed research;
H.U., R.I., G.H.-T., and C.F.B. analyzed data; and H.U., R.I., G.H.-T., and C.F.B. wrote
the paper.
The authors declare no conflict of interest.
This article is
a PNAS Direct Submission. D.W.M. is a guest editor invited by the
Editorial Board.
1To whom correspondence should be addressed. E-mail: carlos@scripps.edu.
This article contains supporting information online at www.pnas.org/lookup/suppl/
doi:10.1073/pnas.1003350107/-/DCSupplemental.
20672–20677 ∣ PNAS ∣ November 30, 2010 ∣ vol. 107 ∣ no. 48
www.pnas.org/cgi/doi/10.1073/pnas.1003350107

H
R
OH
R
HO
O
Ph
1
OHC
3
OTBS
NO2
OHC
NO2
2
4
OHC
OTBS
OTBS
OHC
Et3N (50 mol%)
NO2
+
3
5
OHC
NO2
4
R
+
OTBS
TBSO
OTBS
4
up to 98:2 dr
up to 99% ee
TBSO NO2
OTBS
OTBS
4a
dr 97:3, 98% ee
CH2Cl2
30 °C, 1 h, 87%
2
1
Ph
6a
5
1
3 equiv.
dr > 10:1, 98% ee
5
HO
O
F3C
F3C
S
NH NH2
Scheme 2. Henry reaction of isolated Michael product 4a.
3
TBSO
NO2
4
2
NH
R
6
that of the original Michael reaction. The reaction at 30 °C was
more reproducible than the reaction at room temperature. We
found that a short reaction time suppressed formation of C4-epi-
mer 7a, preventing base-promoted epimerization at C4 position.
To identify the actual catalyst of the Henry reaction, the iso-
lated Michael product 4a was treated with triethylamine without
primary amine-thiourea catalyst 3 (Scheme 2). The product 6a
was obtained in excellent yield and high ee. Thus, the actual
catalyst is the tertiary amine (in this case, triethylamine). The
Henry reaction provided predominantly one of the possible four
isomers. This is a very rare example of stereoselective intermo-
lecular Henry reaction controlled by the configuration of the
nitroalkane (35, 36).
Next, the effect of the second catalyst was studied. Reaction
with sterically hindered diisopropylethylamine gave comparable
results to those with triethylamine (Table 1, entry 8). Less hin-
dered bases such as 1,4-diazabicyclo[2.2.2]octane (DABCO) and
4-dimethylaminopyridine (DMAP) were poorer catalysts (entries
9 and 10). When 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) was
used, complete epimerization at C4 position was observed, af-
fording 3,4-dideoxy-D-mannose derivative 7a with excellent ee.
Derivative 7a existed only in the cyclized form in CDCl3.
Dimeric catalyst 11 provided the product 6a in moderate yield
with 10 mol% catalyst loading. The optimized one-pot conditions
that provide for the synthesis of 3,4-dideoxy-D-talose derivative
6a with excellent ee and in good yield from two molecules of (tert-
butyldimethylsilyloxy)acetaldehyde 1 and one β-nitrostyrene 2a
via an anti-Michael-Henry reaction are listed in Table 1, entry 7.
3
Scheme 1. anti-Michael–Henry reaction sequence to construct carbo-
hydrate structures.
the α-anomers of 6a and 7a were observed in accord with the
preference typical of manno- and talo-type carbohydrates. Small
quantities of other diastereomers were removed by column chro-
matography. The closed form 6a and its open form 5a existed as
an equilibrium mixture in CDCl3 (ca. 3∶1). This process provided
stereoisomer 6a with five continuous stereocenters in good iso-
lated yield; however, the enantiomeric excess was only 88%,
which was lower than that of the original Michael adduct 4a
as catalyzed by 3 (98% ee). Other chiral catalysts such as Take-
moto catalyst 8, quinine 9, and quinidine 10 were tested as the
second catalyst, but enantiomeric excess was not improved
(entries 2–4). Because the enantioselectivity should be mainly in-
duced at the irreversible Michael reaction step, we anticipated
that these second catalysts promoted the Michael reaction non-
selectively and that the competitive reaction resulted in a de-
creased ee. In fact, triethylamine and Takemoto catalyst 8 pro-
duced product 6 even in the absence of primary amine-thiourea
3 (entries 5 and 6).
To overcome this problem, the second catalyst was added after
the Michael reaction was complete (Method B). After insuring
complete conversion of β-nitrostyrene 2a, 50 mol% of triethyla-
mine was added to the reaction mixture, which was kept at 30 °C
for 1 h (entry 7). An ee value of 98% was observed, comparable to
Table 1. Optimization of Michael–Henry reaction
†
‡
§
Entry
Catalyst A
Catalyst B (mol%)
Time (T1 þ T2)
Method *
Yield (%)
dr (5a þ 6a∶7a)
ee (%)
1
2
3
4
5
6
7
8
3
3
3
3
-
Et3N (20)
8 (20)
9 (20)
10 (20)
Et3N (20)
8 (20)
Et3N (50)
i Pr2 EtN (50)
DABCO (25)
DMAP (50)
DBU (50)
Et3N (50)
1 d
1 d
1 d
1 d
1 d
A
A
A
A
A
A
B
B
B
B
B
B
62
54
62
58
27
17
68
68
44
15
51
38
7∶1
4∶1
88
89
87
87
—
-35
98
98
98
98
98
96
6∶1
6∶1
3∶1
-
1 d
3∶1
3
3
3
3
3
4 + 1 h
4 + 1 h
4 + 1 h
4 + 1 h
4 + 1 h
8 + 2 h
>10∶1
>10∶1
4∶1
9
10
11
1∶1
0∶1
¶
∥
12
11
8∶1
*Method A: catalyst A (20 mol%), catalyst B and 2a (0.1 or 0.2 mmol) were reacted with 1 (4 equiv.) in CH2Cl2 at rt for T1. Method B: catalyst A (20 mol%) and
2a (0.1 or 0.2 mmol) were reacted with 1 (4 equiv.) in CH2Cl2 at 30 °C for T1, then catalyst B was added (see text).
†Yield of isolated product.
‡Determined by 1H NMR analysis of purified product, (5a þ 6a∶7a).
§Determined by chiral phase HPLC analysis of corresponding alcohol.
¶The reaction was carried out at rt.
∥10 mol% of 11 was used.
Uehara et al.
PNAS
∣
November 30, 2010
∣
vol. 107
∣
no. 48
∣
20673

Table 2. Michael–Henry reaction to 3,4-dideoxy-D-talose
Table 3. Michael–Henry reaction to dideoxy-D-mannopyranose
derivatives 6*
derivatives 7*
Time
(T1, h)
Time
(T2, h)
Yield
(%)
†
‡
Entry
1
Product
ee (%)
98
H
Time Time Yield
dr
ee
(%)
HO
O
OTBS
NO2
‡
§
Entry
Product
(T1, h) (T2, h) (%) † (6 þ 5∶7) ‡ 6∶5
4
4
1
1
51
65
TBSO
1
2
3
4
Ph−
4 − BrC6H4−
4
4
4
1
68
>10∶1 3∶1
>10∶1 4∶1
>10∶1 3∶1
>10∶1 6∶1
1∶0 0∶1
7∶1 13∶1 97
6∶1 1∶0
>10∶1 1∶0
98
98
97
97
99
Ph
0.5 62
1.5 76
0.5 68
7a
4 − MeOC6H4−
3 − BrC6H4−
2;6 − Cl2C6H3−
2 − Thiophenyl
ðEÞ − PhCH ¼ CH−
n − C7H15−
H
HO
O
5
OTBS
NO2
¶∥
5
16
7
6
4
37
TBSO
6**
0.5 63
0.3 43
2
3
96
7¶
8
93
96
¶
5
18
44
7b
Br
O
*3 (20 mol%) and 2 (0.2 mmol) were reacted with 1 (4 equiv.) in CH2Cl2 at
30 °C for T1, then Et3N (50 mol%) was added and reacted for T2.
†Yield of isolated product.
H
HO
OTBS
NO2
‡Determined by 1H NMR analysis of purified product in CDCl3.
§Determined by chiral phase HPLC analysis of corresponding alcohol.
¶3 (50 mol%) was used.
TBSO
23
1
1
48
57
95
98
∥Et3N (100 mol%) was used.
OMe 7c
**Et3N (30 mol%) was used.
H
O
HO
OTBS
The C4-epimer 7a in the manno-configuration was produced by
simply changing the second catalyst from triethylamine to DBU.
With these optimized conditions in hand, we surveyed scope of
the reaction using triethylamine as the second catalyst (Table 2).
Nitrostyrenes with both electron withdrawing groups and
donating groups on the aromatic ring were good substrates for
the reaction and the corresponding talose derivatives were ob-
tained in good yield and excellent ee (entries 2–4). To suppress
epimerization, substrates with electron withdrawing groups were
subjected to short second reaction times (T2), whereas longer re-
action times were necessary with electron rich substrates. Reac-
tion of 2,6-dichloronitrostyrene 2e required 50 mol% of catalyst 3
and equimolar amounts of triethylamine to provide the product,
5e, with high ee (99% ee) in moderate yield (entry 5); note that 5e
was present in the open form. Heteroaromatic substrates were
also good candidates for this reaction: 2-(2-Nitrovinyl)thiophene
2f was converted to the product 6f in good yield and selectivity
(entry 6). Enantiomeric excess of products 6 was comparable to
that of Michael products 4 we reported previously (24). When
nitrodiene 2g was used, alkenyl-substituted dideoxytalose deriva-
tive 6g formed with acceptable ee (entry 7). A second reaction
time of 18 h was used to convert alkyl substituted nitroolefin
2h to 6h with good dr and ee (entry 8). Both 6g and 6h with
smaller substituents existed in only in the cyclic form.
Next, we investigated synthesis of 3,4-dideoxy-D-mannose
derivatives 7 using DBU as the second catalyst (Table 3). The cy-
clized products were obtained in reasonably good yield via a
three-step sequence: anti-Michael reaction, syn-Henry reaction,
and C4-epimerization. We observed immediate consumption of
Michael product 4 upon addition of DBU, followed by relatively
slow but complete conversion of talo-type compound 6 to manno-
configured 7. All tested nitroolefins with substituted phenyl,
heteroaromatic, alkenyl, and alkyl groups were converted into
their corresponding derivatives with excellent ee under these con-
ditions, reflecting wide scope of this reaction. The exception was
the sterically demanding 2,6-dichloro-β-nitrostyrene 2e; we could
not isolate the corresponding manno-product in pure form
probably due to poor selectivity of the Henry reaction and decom-
position of product 5e during the prolonged reaction time.
TBSO
NO2
4
5
5
7d
Br
H
HO
O
O
OTBS
NO2
7
20
5
1
2
1
59
66
50
96
93
96
TBSO
S
H
7f
HO
OTBS
§
6
TBSO
NO2
7g
Ph
H
HO
O
OTBS
§
7
TBSO
NO2
n-C7H15
7h
*3 (20 mol%) and 2 (0.2 mmol) were reacted with 1 (4 equiv.) in CH2Cl2 at
30 °C for T1, then DBU (50 mol%) was added and reacted for T2.
†Yield of isolated product.
‡Determined by chiral phase HPLC analysis.
§3 (50 mol%) was used.
Most typically, the Henry reaction proceeded in a stereo-spe-
cific manner with relatively slow C4-epimerization. Two pathways
are possible for isomerization from the talo-configuration in 6 to
the manno-type of 7: One is the direct epimerization of 6 by
deprotonation at C4 and the other is the retro-Henry–Henry
process (32). To clarify the mechanism, talo-product 6a was
converted to acetyl pyranose 12a (Scheme 3). Both open form 5a
and closed form 6a were converted to 12a and the minor manno-
type product 13a could be separated at this stage. DBU treatment
of 12a provided manno-product 13a, suggesting the epimeriza-
tion occurred via direct epimerization at the C4 position. More-
over, both 12a and 13a were successfully converted into 4-amino-
3,4-dideoxy-3-phenyl-D-talose derivative 14 and 4-amino-3,4-di-
deoxy-3-phenyl-D-mannose derivative 15, respectively. A 4-ami-
no-substituted mannose scaffold related to 13a is found in the
antitumor antibiotic spicamycin (37). The absolute and relative
20674
∣
www.pnas.org/cgi/doi/10.1073/pnas.1003350107
Uehara et al.

H
H
H
H
DBU (50 mol%)
CH2Cl2
Ac2O
pyridine
AcO
O
HO
O
AcO
O
16 (4 equiv.)
Et3N (50 mol%)
30 °C, 0.5 h, 55%
HO
O
CO2Et
NO2
A
OTBS
NO2
OTBS
OTBS
NO2
4
4
4
TBSO
TBSO
NO2
TBSO
30 °C, 0.5 h
77%
rt, 1h
TBSO
Ar
Ar
6a: Ar = Ph
6b: Ar = 4-BrC6H4
Ph
13a
Ph
98% ee
3
12a: Ar = Ph, 61%
12b: Ar = 4-BrC6H4, 56%
OHC
(20 mol%)
17
NO2
+
4a
Raney Ni, H2
EtOH, rt, 24 h
then Boc2O
Ph
1) Zn, AcOH
rt, 1 h
2) Boc2O, Et2O
rt, 13 h, 54%
OTBS
1
CH2Cl2
30 °C, 4 h
H
2a
HO
O
CO2Et
rt, 24 h, 79%
16 (4 equiv.)
DBU (50 mol%)
30 °C, 1h, 74%
4
H
TBSO
NO2
H
AcO
O
AcO
O
O
OTBS
OTBS
Ph
18 98% ee
TBSO
NHBoc
TBSO
NHBoc
H
CO2Et
Ph
Ph
16
14
15
B
HCHO (5 equiv.)
Et3N (50 mol%)
NO2
Scheme 3. Epimerization experiment of 12a and conversion to 4-amino-3,4-
dideoxy-3-phenyl-D-talose derivative 14 and 4-amino-3,4-dideoxy-3-phenyl-
D-mannose derivative 15.
HO
O
Ph
OHC
4
TBSO
NO2
OTBS
4a
30 °C, 1 h, 34%
Ph
19 97% ee
configurations of the 3,4-dideoxy-D-talose derivatives 6 were de-
termined by X-ray crystallography of 12b (Fig. 1). A 1,3-diaxial
interaction between the C2-alkoxy group and the C4-nitro group
strained the tetrahydropyran ring and this is presumably the rea-
son 6 exists as an equilibrium mixture between closed structure 6
and its open form 5 in most cases.
Scheme 4. Michael–Henry reaction with other Henry acceptors. (A)
Michael–Henry reaction with ethyl glyoxylate (16) as an acceptor. (B) Henry
reaction with formaldehyde as an acceptor.
reaction for syn-Michael adducts, we chose to study the Michael
reaction of isovaleraldehyde 20 with diphenylprolinol silyl ether
21 (40) under our one-pot Henry reaction conditions (Scheme 5).
The Michael–Henry product 24 formed in good yield with
excellent enantiomeric excess in the presence of p-nitrobenzalde-
hyde 23 and triethylamine. It should be noted that isovaleralde-
hyde 20 did not act as the Henry acceptor under these conditions.
The C4 and C5 stereocenters in the Henry product 24 possessed
the same configurations as those in 6. Hence, the stereoselectivity
of Henry reaction was controlled by the C3 stereocenter for
both anti-Michael adduct 4 and syn-Michael product 22 regard-
less configuration of C2 stereocenter.
In all of the reactions evaluated, only one isomer with four
successive stereocenters was formed as the major product. It is
clear that C2 and C3 stereochemistry was controlled in the
anti-Michael reaction by primary amine-thiourea catalyst 3. Sub-
sequent asymmetric induction at C4 and C5 positions occurred in
the consecutive Henry reaction catalyzed by an achiral amine
such as triethylamine. Substrate-controlled stereo induction in
the Henry reaction is known for chiral aldehyde acceptors. How-
ever, diastereoselective Henry-type reactions of chiral nitroalk-
anes are rare (35, 36) and a general kinetic mechanism to
explain their stereo induction is not reported.
To control C4 asymmetry in the Henry reaction, acceptor al-
dehyde 1 should approach from the si-face of the nitronate anion
generated from Michael product 4a and base. Because both
anti-Michael adducts 4 and syn-product 22 provide the same
stereo-induction in the successive Henry reaction, asymmetric in-
duction at this step is controlled by the chirality at C3 of the
Michael products. It is known that allylic 1,3-strain restricts rota-
tion of the σ-bond connected to enolates and nitronates (41).
Therefore, the nitronate anion generated from 4a should exist
predominantly in the conformation shown in Fig. 2A. The ap-
proach of aldehyde 1 should occur predominantly on the less
crowded face. Fleming and Lewis have discussed diastereoselec-
tivity of enolate alkylation, in which a Ph group was effectively
smaller than the iPr group (42). The “effective radius” of phenyl
group is smaller than that of a methyl group (43). Therefore, we
A method for synthesis of both talo- and manno-type sugars
has been established using (tert-butyldimethylsilyloxy)acetalde-
hyde 1 as both a donor in first Michael reaction and an acceptor
in second Henry reaction. If other aldehydes could be used as
acceptors in the second Henry reaction, the utility of the reaction
would increase dramatically. Therefore, we evaluated other ac-
ceptors (Scheme 4). When ethyl glyoxylate 16 was added as
the second acceptor in the presence of triethylamine, the talo-
configured product 17 was obtained. In this case, the contaminat-
ing C4-epimer 18 was readily removed by column chromato-
graphy. By changing the second catalyst from triethylamine to
DBU, the manno-type product 18 was obtained in good yield.
In both cases, aldehyde 1 served as a donor only in the Michael
reaction and products 17 and 18 were obtained in excellent ee.
When aqueous formaldehyde was used as an acceptor in the sec-
ond step (Scheme 4B), both the Henry reaction and epimeriza-
tion occurred to give pentose derivative 19, which could serve as a
precursor to human NK1 antagonists (38). Because α-oxy-
acetaldehyde 1 is a good acceptor in the Henry reaction, isolation
of Michael adduct 4a and removal of unreacted 1 followed by
addition and reaction with the second aldehyde provided better
yields of products in cases where remaining 1 resulted in a
competing Henry reaction in the one-pot format.
Although a large number of organocatalytic syn-Michael reac-
tions with aldehyde nucleophiles have been reported since 2001
(39), there is no precedence for intermolecular Henry reactions of
these Michael products. To evaluate this type of Michael–Henry
NO2
21
23 (3 equiv.)
Et3N (50 mol%)
Ph
3
H
OHC
(10 mol%)
HO
O
3
OHC
NO2
NO2
+
5
Ph
CH2Cl2
rt, 18h
rt, 4 h, 77%
4
NO2
2a
20
22
Ph
24
= ~4:1, 99% ee
OHC
Ph
Ph
OTMS
:
N
H
NO2
21
23
Scheme 5. syn-Michael–Henry reaction.
Fig. 1. X-ray crystal structure of 12b.
Uehara et al.
PNAS
∣
November 30, 2010
∣
vol. 107
∣
no. 48
∣
20675

O-
Ph
OH
vatives with talo- and manno- configurations. 3,4-Dideoxytalose
derivatives 6 were synthesized by combination of our anti-
Michael reactions of (tert-butyldimethylsilyloxy)acetaldehyde 1
to form γ-nitroaldehydes 4 and a subsequent syn-Henry reaction
with 1 as an acceptor. The extremely high selectivity of the anti-
Michael reactions (up to 98∶2 dr and 99% ee) established the C3
stereocenter that controlled the stereochemistry at successive C4
and C5 stereocenters in the syn-Henry reaction resulting in the
control of four contiguous asymmetric centers. These are the first
reported examples of enamine catalysis of asymmetric intermo-
lecular one-pot Michael–Henry reactions and are among rare ex-
amples of stereoselective intermolecular Henry reactions
controlled by the configuration of the nitroalkanes and catalyzed
by a simple base such as triethylamine. Use of DBU instead of
triethylamine provided 3,4-dideoxymannose derivatives 7 via a
three-step sequence: anti-Michael reaction, syn-Henry reaction,
and C4-epimerization. Various products with aromatic, vinyl,
and alkyl substituents at the C3 position were prepared demon-
strating the wide scope and efficiency of this strategy.
H
A
OHC
N+
OHC
2
4
OTBS
4
3
5
O-
si-face
3
TBSO
Ph
TBSO
NO2
H
OHC
OTBS
5a
4a-nitronate
1
si-face
B
C
Ph
Ph
H
H
NO2
Ph
H
H
NO2
-
-
-
H
NO2
OHC
H
TBS
TBS
O+
OHC
O
OHC
TBSO
O-
N+
D
OHC
H
H
H
5a
O
Base
O
Ph
TBSO
H
OTBS
In addition, the acceptor of the Henry reaction could be var-
ied. When ethyl glyoxylate 16 was added as the second acceptor,
both the talo-configured product 17 and the manno-type product
18 were obtained in the presence of triethylamine and DBU, res-
pectively. Moreover, substitution of catalyst 21 for 3 allowed us to
initiate the assembly reaction with a syn-Michael reaction to pro-
vide product 24 following the Henry reaction. Here, product 24
stereocenters at C4 and C5 possessed the same configurations as
those in 6. This result implies that this type of diastereoselective
Henry reaction is applicable to synthesis of a wide range of
Michael products because the syn-Michael reaction works with
a broad range of aldehydes and nitroolefins. As with the anti-
Michael product, there is no previous literature precedence for
the intermolecular Henry reaction with the syn-Michael products.
Stereoselectivity at C2 and C3 positions is explained by our
anti-Michael reaction design as reported previously (24). Mean-
while, asymmetric induction at C4 and C5 positions occurred in
the consecutive Henry reaction catalyzed by achiral bases. Stereo-
control at C4 position could be explained by si-face approach to
the nitronate anion generated from Michael product 4a and base.
Attack occurred from the side of the relatively small Ph group on
the nitronate anion; rotation here might be restricted by allylic
1,3-strain. On the other hand, the C5 stereocenter could be
induced by a syn-selective Henry reaction. In the proposed six-
membered ring transition state shown in Fig. 2D, side chains
of both nitronate and aldehyde occupy equatorial positions.
These two stereo-controlling factors may result in the unusually
high selectivity of the Henry reaction and consequent production
of D-talo-configured product 6 with four contiguous stereocen-
ters. In summary, we have demonstrated that catalytic asym-
metric assembly reactions based on sequential Michael–Henry
reactions allow rapid stereoselective assembly of varied aldehydes
and nitroolefins into carbohydrates with very high levels of enan-
tio- and diastereocontrol.
Fig. 2. Proposed transition state of the Henry reaction. (A) Configuration of
the nitronate is restricted by allylic 1,3-strain favoring si-face approach of the
nitronate to control the developing C4 stereocenter. (B) Less favorable nitro-
nate conformer. (C) Favored conformer that participates in the transition
state (Left) and hyperconjugated structure of the conformer (Right). (D)
Six-membered ring transition state structure to provide the syn-Henry
product with C5 stereo-induction.
rationalize that as Ph is smaller than the branched alkyl group
(-CH(OTBS)CHO) on the C3 position in Fig. 2A, the electrophi-
lic approach of the aldehyde should occur from the si-face of the
nitronate. It is worth mentioning that the stereoselectivity at the α
position to nitro group in other nitronate addition reactions, in-
cluding Enders’ cascade reaction (16, 44), may be explained in a
similar manner.
In addition, the stereoelectronic effect should enhance the
facial selectivity. A theoretical study indicated that allylic bonds
are staggered with respect to partially formed bonds in the transi-
tion states (45) and the two possible conformers are as shown in
Fig. 2 B and C. In electrophilic attack on a π system such as a
nitronate, a higher and more reactive highest occupied molecular
orbital is obtained by mixing the π orbital and the σ orbital when
the highest energy σ orbital is located perpendicular to the π
system in the nucleophile. It is suggested that a homoallylic
heteroatom (i.e., oxygen atom in Fig. 2C) raises the energy level
of the σ orbital via lone pair participation (46). This hyperconju-
gative interaction, shown in Fig. 2C, makes a transition state
through the conformer in Fig. 2C more likely.
On the other hand, C5 asymmetric induction could be ex-
plained by the syn-selective Henry reaction considering the six-
membered ring transition state. In the proposed cyclic six-
membered transition state shown in Fig. 2D, side chains of both
nitronate and aldehyde occupy equatorial positions. This cyclic
six-membered transition state was proposed for the highly syn-
selective asymmetric Henry reactions catalyzed by transition
metal complexes (47, 48). A chiral guanidine thiourea catalyst
is also reported to provide the syn-Henry product (49). Thus,
these two stereocontrolling factors, si-face approach to the nitro-
nate anion and syn-selective Henry reaction, may control both C4
and C5 asymmetric induction and may result in the observed high
selectivity of the Henry reaction described here.
Materials and Methods
Typical experimental procedure for synthesis of 3,4-dideoxy-D-talose deriva-
tive 6a: (tert-butyldimethylsilyloxy)acetaldehyde 1 (152 μL, 0.8 mmol) was
added to the solution of thiourea catalyst 3 (15.4 mg, 40 μmol) and β-nitros-
tyrene 2a (29.8 mg, 0.2 mmol) in CH2Cl2 (0.2 mL). The resulting solution was
stirred at 30 °C for 4 h and then triethylamine (13.9 μL, 0.1 mmol) was added.
After 1 h at 30 °C, Et2O and 1N HCl (0.2 mL) was added to the solution at rt.
The aqueous layer was separated and extracted three times with Et2O. The
combined organic layers were dried over MgSO4, concentrated, and purified
by flash column chromatography to afford 3,4-dideoxy-D-talose derivative 6a
(67.4 mg, 68%) as a colorless oil. 1H NMR (500 MHz, CDCl3) major (6a): δ
7.40 − 7.24 (m, 5H), 5.29 (brs, 1H), 4.79 (dd, J ¼ 4.6, 2.5 Hz, 1H), 4.47 (ddd, J ¼
8.5, 6.2, 2.3 Hz, 1H), 3.90 (dd, J ¼ 2.8, 1.4 Hz, 1H), 3.88 (dd, J ¼ 9.9, 6.2 Hz, 1H),
3.84 (dd, J ¼ 9.9, 8.6 Hz, 1H), 3.55 (dd, J ¼ 4.7, 2.7 Hz, 1H), 3.03 (brs, 1H), 0.93
(s, 9H), 0.84 (s, 9H), 0.03 (s, 3H), 0.01 (s, 6H), −0.22 (s, 3H); 13C NMR (100 MHz,
CDCl3) major (6a): δ 137.19, 129.95, 128.38, 128.11, 95.08, 81.65, 69.98, 68.85,
Conclusions
The development of simple, robust and efficient methods for the
de novo synthesis of carbohydrates is a significant challenge in
organic synthesis. Our approach to this problem was based on
the development of organocatalytic asymmetric assembly reac-
tions. Here we describe asymmetric syntheses of pyranose deri-
20676
∣
www.pnas.org/cgi/doi/10.1073/pnas.1003350107
Uehara et al.

62.44, 43.47, 26.00, 25.95, 18.41, 18.29, −4.43, −4.98, −5.46, −5.55; high res-
3.9 Hz, 1H), 0.93 (s, 9H), 0.83 (s, 9H), 0.08 (s, 3H), 0.08 (s, 3H), −0.21 (s, 3H),
−0.57 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 136.18, 129.08, 128.60, 127.92,
94.33, 82.45, 72.45, 71.32, 63.15, 45.66, 26.07, 25.95, 18.53, 18.13, −5.22,
þ
olution mass spectrometry (HRMS) (m∕z): ½M þ Hꢀþ calcd for C24H44NO6Si2
498.2702, found 498.2700. Enantiomeric excess: 98%, determined by HPLC
after reduction to corresponding alcohol (Chiralpak IC, hexane∕i-PrOH ¼
97∶3, flow rate 1.00 mL∕ min, λ ¼ 220 nm, rt): tRðmajorÞ ¼ 13.6 min,
tRðminorÞ ¼ 15.2 min.
−5.24, −5.47, −5.73; HRMS (m∕z): ½M þ Hꢀþ calcd for C24H44NO6Si2
þ
498.2702, found 498.2705. Enantiomeric excess: 98%, determined by
HPLC (Chiralpak AD-H, hexane∕i-PrOH ¼ 99∶1, flow rate 1.00 mL∕ min, λ ¼
220 nm, rt): tRðmajorÞ ¼ 11.6 min, tRðminorÞ ¼ 8.8 min.
The procedure for synthesis of 3,4-dideoxy-D-mannose derivative 7a was
similar to that for 6a, except for the use of DBU rather than triethylamine.
3,4-Dideoxy-D-mannose derivative 7a (50.6 mg, 51%) was obtained as a
colorless oil. 1H NMR (400 MHz, CDCl3) δ 7.33 − 7.21 (m, 5H), 5.47 (dd,
J ¼ 12.0, 9.8 Hz, 1H), 5.13 (d, J ¼ 1.3 Hz, 1H), 4.47 (dt, J ¼ 9.8, 3.5 Hz, 1H),
3.94 (dd, J ¼ 12.0, 2.5 Hz, 1H), 3.82 − 3.77 (m, 2H), 3.74 (dd, J ¼ 11.5,
ACKNOWLEDGMENTS. We thank the Skaggs Institute for Chemical Biology
for financial support. G.H.-T. thanks the Spanish Ministry of Education for
a fellowship.
1. Freudenberg K (1966) Emil Fischer and his contribution to carbohydrate chemistry. Adv
Carbohydr Chem Biochem 21:1–38.
2. List B, Lerner RA, Barbas CF, III (2000) Proline-catalyzed direct asymmetric aldol
reactions. J Am Chem Soc 122:2395–2396.
3. Sakthivel K, Notz W, Bui T, Barbas CF, III (2001) Amino acid catalyzed direct asymmetric
aldol reactions: A bioorganic approach to catalytic asymmetric carbon–carbon bond-
forming reactions. J Am Chem Soc 123:5260–5267.
26. Liu Y, Sun B, Wang B, Wakem M, Deng L (2009) Catalytic asymmetric conjugate
addition of simple alkyl thiols to α,β-unsaturated N-acylated oxazolidin-2-ones with
bifunctional catalysts. J Am Chem Soc 131:418–419.
27. Mei K, et al. (2009) Simple cyclohexanediamine-derived primary amine thiourea
catalyzed highly enantioselective conjugate addition of nitroalkanes to enones.
Org Lett 11:2864–2867.
28. Okino T, Hoashi Y, Takemoto Y (2003) Enantioselective michael reaction of malonates
4. Barbas CF, III (2008) Organocatalysis lost: Modern chemistry, ancient chemistry, and an
unseen biosynthetic apparatus. Angew Chem Int Ed 47:42–47.
to nitroolefins catalyzed by bifunctional organocatalysts.
125:12672–12673.
J Am Chem Soc
5. Notz W, Tanaka F, Barbas CF, III (2004) Enamine-based organocatalysis with proline and
diamines: The development of direct catalytic asymmetric aldol, Mannich, Michael,
and Diels–Alder reactions. Acc Chem Res 37:580–591.
29. Wang J, Li H, Yu X, Zu L, Wang W (2005) Chiral binaphthyl-derived amine-thiourea
organocatalyst-promoted asymmetric Morita–Baylis–Hillman Reaction. Org Lett
7:4293–4296.
6. Tanaka F, Barbas CF, III (2004) Antibody-catalyzed aldol reactions. Modern Aldol Reac-
tions, Volume 1: Enolates, Biocatalysis, and Natural Product Synthesis, ed R Mahrwald
(Wiley, New York), pp 273–310.
7. Wagner J, Lerner RA, Barbas CF, III (1995) Efficient aldolase catalytic antibodies that
use the enamine mechanism of natural enzymes. Science 270:1797–1800.
8. Bertelsen S, Jørgensen KA (2009) Organocatalysis—After the gold rush. Chem Soc Rev
38:2178–2189.
9. Doyle AG, Jacobsen EN (2007) Small-molecule H-bond donors in asymmetric catalysis.
Chem Rev 107:5713–5743.
10. Takemoto Y (2005) Recognition and activation by ureas and thioureas: stereoselective
reactions using ureas and thioureas as hydrogen-bonding donors. Org Biomol Chem
3:4299–4306.
11. Tian S, et al. (2004) Asymmetric organic catalysis with modified cinchona alkaloids. Acc
Chem Res 37:621–631.
12. Bui T, Barbas CF, III (2000) A proline-catalyzed asymmetric Robinson annulation reac-
tion. Tetrahedron Lett 41:6951–6954.
13. Chowdari NS, Ramachary DB, Barbas CF, III (2003) Organocatalytic asymmetric assem-
bly reactions: One-pot synthesis of functionalized β-amino alcohols from aldehydes,
ketones, and azodicarboxylates. Org Lett 5:1685–1688.
14. Chowdari NS, Ramachary DB, Córdova A, Barbas CF, III (2002) Proline-catalyzed asym-
metric assembly reactions: enzyme-like assembly of carbohydrates and polyketides
from three aldehyde substrates. Tetrahedron Lett 43:9591–9595.
15. Enders D, Grondal C, Hüttl MRM (2007) Asymmetric organocatalytic domino reactions.
Angew Chem Int Ed 46:1570–1581.
30. Palomo C, Oiarbide M, Laso A (2007) Recent advances in the catalytic asymmetric
nitroaldol (Henry) reaction. Eur J Org Chem 2561–2574.
31. Adibekian A, et al. (2008) Stereocontrolled synthesis of fully functionalized D-gluco-
samine monosaccharides via a domino nitro-Michael/Henry reaction. Chem Commun
3549–3551.
32. Hayashi Y, Okano T, Aratake S, Hazelard D (2007) Diphenylprolinol silyl ether as a
catalyst in an enantioselective, catalytic, tandem Michael/Henry reaction for the
control of four stereocenters. Angew Chem Int Ed 46:4922–4925.
33. Xu D, et al. (2008) A novel enantioselective catalytic tandem oxa-Michael–Henry
reaction: One-pot organocatalytic asymmetric synthesis of 3-Nitro-2H-chromenes.
Adv Synth Catal 350:2610–2616.
34. García Ruano JL, Marcos V, Suanzes JA, Marzo L, Alemán J (2009) One-pot synthesis of
pentasubstituted cyclohexanes by a Michael addition followed by a tandem inter–in-
tra double Henry reaction. Chem Eur J 15:6576–6580.
35. Han B, Xiao Y, He Z, Chen Y (2009) Asymmetric Michael addition of γ,γ-disubstituted
α,β-unsaturated aldehydes to nitroolefins via dienamine catalysis. Org Lett
11:4660–4663.
36. Jakubec P, Helliwell M, Dixon DJ (2008) Cyclic imine nitro-Mannich/lactamization
cascades: a direct stereoselective synthesis of multicyclic piperidinone derivatives.
Org Lett 10:4267–4270.
37. Sakai T, et al. (1995) Absolute configuration of spicamycin, an antitumor antibiotic
produced by Streptomyces alanosinicus. J Antibiot 48:899–900.
38. Morriello GJ, et al. (2008) Fused bicyclic pyrrolizinones as new scaffolds for human NK1
antagonists. Bioorg Med Chem 16:2156–2170.
39. Betancort JM, Barbas CF, III (2001) Catalytic direct asymmetric Michael reactions:
Taming naked aldehyde donors. Org Lett 3:3737–3740.
40. Hayashi Y, Gotoh H, Hayashi T, Shoji M (2005) Diphenylprolinol silyl ethers as efficient
organocatalysts for the asymmetric Michael reaction of aldehydes and nitroalkenes.
Angew Chem Int Ed 44:4212–4215.
41. Hoffmann RW (1989) Allylic 1,3-strain as a controlling factor in stereoselective trans-
formations. Chem Rev 89:1841–1860.
42. Fleming I, Lewis JJ (1985) A paradigm for diastereoselectivity in electrophilic attack on
trigonal carbon adjacent to a chiral centre: The methylation and protonation of some
open-chain enolates. J Chem Soc Chem Commun 149–151.
43. Bott G, Field LD, Sternhell S (1980) Steric effects. A study of a rationally designed
system. J Am Chem Soc 102:5618–5626.
16. Enders D, Hüttl MRM, Grondal C, Raabe G (2006) Control of four stereocentres in a
triple cascade organocatalytic reaction. Nature 441:861–863.
17. Enders D, Grondal C (2005) Direct organocatalytic de novo synthesis of carbohydrates.
Angew Chem Int Ed 44:1210–1212.
18. Jiang H, et al. (2009) Achieving molecular complexity by organocatalytic one-pot stra-
tegies—A fast entry for synthesis of sphingoids, amino sugars, and polyhydroxylated
α-amino acids. Angew Chem Int Ed 48:6844–6848.
19. Northrup AB, MacMillan DWC (2004) Two-step synthesis of carbohydrates by selective
aldol reactions. Science 305:1752–1755.
20. Ramasastry SSV, Albertshofer K, Utsumi N, Tanaka F, Barbas CF, III (2007) Mimicking
fructose and rhamnulose aldolases: Organocatalytic syn-aldol reactions with unpro-
tected dihydroxyacetone. Angew Chem Int Ed 46:5572–5575.
21. Ramasastry SSV, Zhang H, Tanaka F, Barbas CF, III (2007) Direct catalytic asymmetric
synthesis of anti-1,2-amino alcohols and syn-1,2-diols through organocatalytic anti-
Mannich and syn-aldol reactions. J Am Chem Soc 129:288–289.
44. Shinisha CB, Sunoj RB (2008) Unraveling high precision stereocontrol in a triple cascade
organocatalytic reaction. Org Biomol Chem 6:3921–3929.
45. Paddon-Row MN, Rondan NG, Houk KN (1982) Staggered models for asymmetric
induction: attack trajectories and conformations of allylic bonds from ab initio
transition structures of addition reactions. J Am Chem Soc 104:7162–7166.
46. McGarvey GJ, Williams JM (1985) Stereoelectronic controlling features of allylic
asymmetry. Application to ester enolate alkylations. J Am Chem Soc 107:1435–1437.
47. Arai T, Yokoyama N (2008) Tandem catalytic asymmetric Friedel–Crafts/Henry reaction:
Control of three contiguous acyclic stereocenters. Angew Chem Int Ed 47:4989–4992.
48. Sasai H, et al. (1995) Efficient diastereoselective and enantioselective nitroaldol
reactions from prochiral starting materials: Utilization of La–Li–6,6′-disubstituted
BINOL complexes as asymmetric catalysts. J Org Chem 60:7388–7389.
49. Sohtome Y, et al. (2007) Organocatalytic asymmetric nitroaldol reaction: Cooperative
effects of guanidine and thiourea functional groups. Chem Asian J 2:1150–1160.
22. Suri JT, Ramachary DB, Barbas CF, III (2005) Mimicking dihydroxy acetone phosphate-
utilizing aldolases through organocatalysis: A facile route to carbohydrates and ami-
nosugars. Org Lett 7:1383–1385.
23. Zhang H, Ramasastry SSV, Tanaka F, Barbas CF, III (2008) Organocatalytic anti-Mannich
reactions with dihydroxyacetone and acyclic dihydroxyacetone derivatives: A facile
route to amino sugars. Adv Synth Catal 350:791–796.
24. Uehara H, Barbas CF, III (2009) anti-Selective asymmetric Michael reactions of
aldehydes and nitroolefins catalyzed by a primary amine/thiourea. Angew Chem
Int Ed 48:9848–9852.
25. Lalonde MP, Chen Y, Jacobsen EN (2006) A chiral primary amine thiourea catalyst for
the highly enantioselective direct conjugate addition of α,α-disubstituted aldehydes to
nitroalkenes. Angew Chem Int Ed 45:6366–6370.
Uehara et al.
PNAS
∣
November 30, 2010
∣
vol. 107
∣
no. 48
∣
20677