Direct aldol reaction of pyruvic derivatives: Catalytic attempt to synthesize ulosonic acids

Article abstract of DOI:10.1002/ejoc.201200325

The catalytic asymmetric aldol reaction of pyruvic aldehyde dimethyl acetal and 2-acetylthiazole with sugar aldehydes is demonstrated to be the key step in the synthesis of 3-deoxy-2-ulosonic acids. Efficient and stereoselective pyruvate aldol reactions are catalyzed by metal-based chiral Trost and Shibasaki catalysts. The presented synthetic methodology mimics aldolase-catalyzed reactions by direct activation of C₃ pyruvate equivalents with the use of metal-based chiral catalysts en route to the synthesis of 2-keto-3-deoxy-D-glucosonic acid and 3-deoxy-D-manno-2-octulosonic acid derivatives. Copyright

Full text of DOI:10.1002/ejoc.201200325

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201200325
Direct Aldol Reaction of Pyruvic Derivatives: Catalytic Attempt To Synthesize
Ulosonic Acids
[a]
[a]
[a]
[a,b]
Osama El-Sepelgy, Darius Schwarzer, Piotr Oskwarek, and Jacek Mlynarski*
Keywords: Aldol reactions / Asymmetric catalysis / Carbohydrates / Enzymes
The catalytic asymmetric aldol reaction of pyruvic aldehyde
dimethyl acetal and 2-acetylthiazole with sugar aldehydes is
demonstrated to be the key step in the synthesis of 3-deoxy-
Shibasaki catalysts. The presented synthetic methodology
mimics aldolase-catalyzed reactions by direct activation of C
pyruvate equivalents with the use of metal-based chiral cata-
lysts en route to the synthesis of 2-keto-3-deoxy- -glucosonic
acid and 3-deoxy- -manno-2-octulosonic acid derivatives.
3
2-ulosonic acids. Efficient and stereoselective pyruvate aldol
D
reactions are catalyzed by metal-based chiral Trost and
D
Introduction
The direct asymmetric aldol reaction of unmodified
carbonyl compounds represents one of the most powerful
[
1]
carbon–carbon bond-forming reactions in both nature
and synthetic organic chemistry.[ Although enzymes are
increasingly recognized as useful catalysts for asymmetric
organic preparations, their application is mostly limited to
a narrow scope of substrates similar to that used by nature.
Nevertheless, observation of enzymatic modes of substrate
activation provides insight for achieving similar transforma-
tions by chemical synthesis. As a result, the efficient and
direct aldol reactions of important biological donors such
as hydroxyacetone and dihydroxyacetone have been success-
fully realized by using biomimetic pathways.[
2]
3]
Scheme 1. Biosynthesis of 3-deoxy-2-ketoacids by phosphoenolpyr-
Among the variants of asymmetric aldol reactions, pyr- uvate-dependent aldolases.
uvate-dependent aldol reactions are of particular interest,
as these processes are vital in the formation of 3-deoxy-2-
In recent years, a number of chemical and enzymatic
methodologies have been reported for the synthesis of
ulosonic acids (3-deoxy-2-keto acids), which are essential
sugar units for many biological processes and transforma-
[
6,7]
[
4]
higher sialic and ulosonic acids.
However, in spite of the
tions. The phosphorylated form of 2-keto-3-deoxy-d-
glucosonic acid (KDG), that is, KDPG (1), is part of the
Entner–Doudoroff pathway, whereas 3-deoxy-d-manno-2-
many published attempts, non-enzymatic syntheses based
on chain extension with the use of an appropriate sugar
unit as the electrophile and a C pyruvate as the nucleophile
3
octulosonic acid [KDO (2)] has been found in Gram-nega-
tive bacteria lipopolysaccharides (LPS).[5] Pyruvate-depend-
under asymmetric control (see Scheme 1 for the biosyn-
thetic pathway) is still troublesome, if simply not possible.[
In a broader context, the direct catalytic aldol reaction of
pyruvate and pyruvate derivatives still remains an elusive
goal despite the breakthroughs in many other direct asym-
8]
ent aldolase enzymes have catabolic function in vivo in the
degradation of 3-deoxy-2-ulosonic acids, whereas their
counterparts employ phosphoenolpyruvate as a donor in
the synthesis of target biomolecules (Scheme 1).[6]
[
2]
metric aldol methodologies. In spite of the tremendous
effort, the scope of the known methodologies are narrow
and are strictly limited to highly active and non-enolizable
[
[
a] Faculty of Chemistry, Jagiellonian University,
Ingardena 3, 30-060 Krakow, Poland
[9]
aldehydes or the self-condensation of pyruvate donors
E-mail: jacek.mlynarski@gmail.com
[
10]
www.jacekmlynarski.pl
with pyruvate-type acceptors. Even so, we assumed that
the direct activation of the donor could be possible in an
asymmetric manner by using biomimetic metal-based cata-
lysts.
b] Institute of Organic Chemistry, Polish Academy of Sciences
Kasprzaka 44/52, Warsaw, Poland
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200325.
2724
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2724–2727

Direct Aldol Reaction of Pyruvic Derivatives
Results and Discussion
Table 1. Direct aldol reaction of pyruvic aldehyde dimethyl acetal
with glyceraldehyde.
In this paper, we report that chiral metal complexes can
act as a pyruvate-dependent aldolase, that is, a type II ald-
olase, to catalyze the direct asymmetric aldol reaction of
pyruvate derivatives and their equivalents with aliphatic al-
dehyde electrophiles. The products obtained from this reac-
tion can be further successfully transformed into 3-deoxy-
Entry
Catalyst (mol-%)
Time
Yield
anti/syn[a]
[b]
[%]
1
2
3
4
5
6
7
8
9
(R)-Pro (30)[
(S)-Pro (30)
c]
5 d
5 d
5 d
20 min
36 h
36 h
8 h
8
4
8
35
56
49
45
45
40
4:1
1:4
4:1
2.2:1
3:1
2.5:1
2:1
[c]
[c]
d]
(S)-Tet (30)
LDA (110)[
[
e]
(R)-ProPh (10)[
2-ulosonic acids by using previously elaborated and reliable
e]
(S)-ProPh (10)
methodologies.
(R)-LLB (10)[
f]
First, we focused on the troublesome aldol reaction be-
tween (R)-glyceraldehyde and dimethyl acetal 4 leading to
anti-configured 3-deoxy-2-keto-d-glucosonic acid precursor
[f]
(S)-LLB (10)
8 h
24 h
7:1
13:1
(S)-LLB (10)[
g]
a] Determined by 1H NMR spectroscopy. [b] Yield of isolated
(5 mmol),
overview, we simultaneously used organocatalysts mimick- (25 mmol), catalyst (1.5 mmol, 30 mol-%) in DMSO at 4 °C for
[
5
(Scheme 2). For our initial trials and to have a general product. [c] Reactions were performed with
3
4
5
d. [d] Reaction was performed with 3 (1 mmol), 4 (1 mmol), LDA
ing type I aldolases and metal complexes acting as type II
aldolases.
(
1.1 mmol, 110 mol-%) in THF at –78 °C for 20 min. [e] Reactions
were performed with 3 (1 mmol), 4 (1.5 mmol), catalyst (10 mol-%)
in DMF at r.t. for 36 h. [f] Reactions were performed with 3
(
(
1 mmol), 4 (1 mmol), catalyst (10 mol-%) and KHMDS additive
10 mol-%) in THF at r.t. for 8 h. [g] Reactions were performed
with 3 (1 mmol), 4 (1 mmol), catalyst (10 mol-%) and KHMDS
additive (10 mol-%) in THF at 0 °C for 24 h.
all reactions performed under organocatalytic control, we
observed predominant formation of the corresponding de-
hydration product, which greatly affected the yield of the
desired aldol. We believe that the elimination step is not an
aldol reaction step but rather results from the competitive
Mannich reaction/elimination pathway, so further optimiza-
tion of the proline-catalyzed reaction may not be pos-
[
12]
sible.
Our efforts in the application of various organocatalysts
to the reaction were also unsuccessful. After this initial
trial, we decided to focus on metal-assisted activation of
pyruvic aldehyde dimethyl acetal, which constitutes a white
area in the map of the asymmetric aldol reaction. This can
be compared mechanistically to type II aldolases, which
employ a zinc ion to acidify the α proton of the donor com-
Scheme 2. Direct aldol reaction of pyruvic aldehyde dimethyl acetal
with glyceraldehyde.
Initial investigations of the direct aldol reaction between ponent to form an enolate. As a reference example, we per-
selected acceptor 3 and donor 4 were conducted by em- formed the aldol reaction of the stoichiometrically gener-
[
13]
ploying (R)-proline as the enamine-formed organocatalyst ated lithium enolate of 4 with glyceraldehyde.
Table 1, Entry 1). tion was quite inspiring, and desired aldol 5 was obtained in
According to previous reports, “natural” (S)-proline 35% yield with a moderate anti/syn (ca. 2:1) ratio (Table 1,
The reac-
(
proved to be a better catalyst when the (S)-configured pro- Entry 4). Encouraged by these results, we decided to test an
tected glyceraldehyde was used, whereas with (R)-config- asymmetric dinuclear zinc catalyst with the ProPh ligand
ured 2,3-O-isopropylidene-d-glyceraldehyde (3), (R)-proline presented in 2000 by Trost for the reaction of aryl methyl
afforded the best results in terms of yield.[ Keeping this in ketones. Pleasingly, both enantiomeric catalysts exhibited
mind, we tested both enantiomeric proline organocatalysts, good catalytic activity, and more importantly, good stereo-
but the observed catalytic activity of both mirror molecules selectivity was observed for a catalyst loading of only
was simply disappointing. The reactions were unsuccessful, 10 mol-% (Table 1, Entries 5 and 6). For the reaction per-
which was quite surprising, as a similar reaction performed formed at room temperature, (R)-ProPh proved to be a
under asymmetric control was described previously by En- slightly better catalyst, as reflected in both the yield and the
3]
[14]
[
11]
ders. The best results were observed by us in the reaction diastereomeric ratio (3:1). Further progress in the reaction
with (R)-proline (30 mol-%) to afford aldol 5 in 8% yield was observed for the Shibasaki lanthanium-based catalyst
[
15]
when a fivefold excess amount of the ketone was used with (S)-BINOL ligands. The (S)-LLB catalyst proved to
(Table 1, Entry 1). Thus, the reaction cannot be seen as a be the most suitable catalyst when (R)-configured aldehyde
catalytic process when 30 mol-% of the catalyst resulted in 3 was used, as the best results in terms of the anti/syn ratio
the formation of 8% of the desired product. Moreover, for (7:1) were obtained, even at room temperature (Table 1, En-
Eur. J. Org. Chem. 2012, 2724–2727
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2725

O. El-Sepelgy, D. Schwarzer, P. Oskwarek, J. Mlynarski
SHORT COMMUNICATION
try 8). The reaction stereoselectivity could be further im-
proved by performing the reaction for 24 h at 0 °C. anti-
Configured KDG precursor 5 was formed predominantly
with a syn/aldol ratio of 13:1 (Table 1, Entry 9). For all
tested catalysts, a slight excess of the ketone donor was suf-
ficient to achieve a good yield of the aldol product.
Both epimeric aldols of (S,R)- and (R,R)-configuration
were easily deprotected with an acidic ion-exchange resin
(Amberlyst 15) to give the cyclic forms of the desired
sugars. It is important to stress that we did not observe
racemization of glyceraldehyde under the elaborated reac-
tion conditions. Application of both types of chiral catalyst
did not affect the enantiomeric excess values of the product,
which depends only on the enantiomeric excess value of the
starting material. This was proved by using HPLC analysis Scheme 4. Synthesis of KDO precursor.
on a chiral stationary phase.
Subsequently, en route to the ulosonic acid derivatives,
we decided to test also the use of 2-acetylthiazole (6), suc-
Conclusions
cessfully introduced by Dondoni, to the chemistry of ulos-
In conclusion, we have successfully developed efficient
and stereoselective pyruvate aldol reactions catalyzed by
metal-based chiral catalysts. This methodology allowed the
first catalytic synthesis of 3-deoxy-2-keto acid precursors
through direct aldol reaction of sugar aldehydes with pyr-
uvic derivatives. An elaborated direct aldol reaction pro-
moted by chiral metal complexes acting as type II aldolases
can be seen as resembling the natural pathway of substrate
[
16]
onic acids.
This was also an interesting challenge in the
field of the catalytic asymmetric aldol reaction, as applica-
tion of heterocyclic methyl ketone could possibly expand
[
2]
the still-narrow scope of the Shibasaki and Trost catalysts.
Application of (R)-ProPh resulted in the formation of aldol
in 50% yield with ca. 5:1 anti/syn ratio. To our gratifica-
7
tion, only 5 mol-% of the chiral (S)-LLB catalyst was neces-
sary to afford a promising yield and stereoselectivity in the
aldol reaction of 6 with (R)-glyceraldehyde (3, Scheme 3).
Desired aldol 7 was formed in 53% yield with 75%de for
equimolar amounts of both substrates on a 2-mmol scale.
Efficient transformation of this KDG-precursor to the ulo-
sonic acid was described previously by Dondoni.[
activation. The presented protocol, which utilizes C pyr-
3
uvic aldehyde dimethyl acetal, provides an attractive and
biomimetic approach to ulosonic acids and constitutes an-
other interesting field of application for the powerful Trost
and Shibasaki catalysts. A variety of pyruvate derivatives
were efficiently activated under the elaborated protocol, in-
cluding 2-acetylthiazole, which is a fundamental achieve-
ment when compared to previously published methodolo-
gies for the synthesis of ulosonic acids. We believe that this
work can be seen as the intermingling of two fundamental
streams of chemical research: biomimetic direct aldol meth-
odology and the synthesis of defined sugar structures.
8b]
Experimental Section
Supporting Information (see footnote on the first page of this arti-
Scheme 3. Synthesis of the KDG precursor.
1
cle): Detailed experimental procedures and copies of the H NMR
1
3
and C NMR spectra of all key intermediates and products.
Next, we carried out a more demanding synthesis of a
higher 2-ulosonic acid – KDO precursor, starting from d-
arabinose diacetonide (8, Scheme 4). The synthesis of de-
sired anti-configured aldol 9 was performed by (S)-ProPh-
Acknowledgments
controlled addition of 2-acetylthiazole to protected sugar Project operated within the Foundation for the Polish Science
aldehyde 8. A low catalyst loading (5 mol-%) and the op- TEAM and MPD Programmes co-financed by the EU European
Regional Development Fund.
tion of performing the reaction at room temperature proved
to be additional practical advantages of the described meth-
odology. By using only a 1.3-fold excess of the aldehyde to [1] M. Brovetto, D. Gamenara, P. Saenz Méndez, G. A. Seoane,
ketone, the isolated yield increased to 77%. From this key
intermediate, the synthesis proceeded over three previously
described steps, namely, intramolecular hemiketalization,
thiazole-to-formyl conversion, and oxidation of the aldos-
ulose to the target KDO.[
Chem. Rev. 2011, 111, 4346–4403.
[
[
2] B. M. Trost, C. S. Brindle, Chem. Soc. Rev. 2010, 39, 1600–
1632.
3] a) D. Enders, A. A. Narine, J. Org. Chem. 2008, 73, 7857–7870;
b) M. Markert, R. Mahrwald, Chem. Eur. J. 2008, 14, 48–48;
c) N. Mase, C. F. Barbas III, Org. Biomol. Chem. 2010, 8, 4043–
8b]
2726
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2724–2727

Direct Aldol Reaction of Pyruvic Derivatives
4
5
050; d) J. Mlynarski, B. Gut, Chem. Soc. Rev. 2012, 41, 587–
[9] a) For the reaction with chloral monohydrate, see: H. Torii, M.
Nakadai, K. Ishihara, S. Saito, H. Yamamoto, Angew. Chem.
2004, 116, 2017; Angew. Chem. Int. Ed. 2004, 43, 1983–1986;
b) for the reaction with activated aromatic aldehydes, see: S.
Luo, H. Xu, L. Chen, J.-P. Cheng, Org. Lett. 2008, 10, 1775–
1778.
96.
[
4] B. Lindberg, Adv. Carbohydr. Chem. Biochem. 1990, 48, 279–
3
18.
[
5] a) F. M. Unger, Adv. Carbohydr. Chem. Biochem. 1981, 38, 323–
3
1
4
88; b) R. Schauer, Adv. Carbohydr. Chem. Biochem. 1982, 40,
31–234; c) T. Angata, A. Varki, Chem. Rev. 2002, 102, 439–
70.
[10] N. Gathergood, K. Juhl, T. B. Poulsen, K. Thordrup, K. A.
Jørgensen, Org. Biomol. Chem. 2004, 2, 1077–1085.
[
6] A. Banaszek, J. Mlynarski, Stud. Nat. Prod. Chem. 2005, 30,
[
[
11] D. Enders, T. Gasperi, Chem. Commun. 2007, 88–90.
12] B. List, P. Pojarliev, C. Castello, Org. Lett. 2001, 3, 573–575.
4
19–482.
[
[
7] I. Hemeon, A. J. Benet, Synthesis 2007, 1899–1926.
8] a) For the diastereoselective base-catalyzed aldol condensation
[13] K. Narasaka, F.-C. Pai, Tetrahedron 1984, 40, 2233–2238.
[14] B. M. Trost, H. Ito, J. Am. Chem. Soc. 2000, 122, 12003–12004.
[15] N. Yoshikawa, Y. M. A. Yamada, J. Das, H. Sasai, M. Shiba-
saki, J. Am. Chem. Soc. 1999, 121, 4168–4178.
[16] A. Dondoni, Org. Biomol. Chem. 2010, 8, 3366–3385.
Received: March 15, 2012
between d-arabinose and C
carboxylation, see: R. Winzar, J. Philips, M. J. Kiefel, Synlett
4
oxalacetic acid, followed by de-
2
010, 583–586 and references cited therein; b) for the stereo-
selective aldol condensation of aldoses with lithiated 2-acetyl-
thiasole, see: A. Dondoni, P. Merino, J. Org. Chem. 1991, 56,
5294–5301.
Published Online: April 10, 2012
Eur. J. Org. Chem. 2012, 2724–2727
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2727