Prebiotic carbohydrate synthesis: Zinc-proline catalyzes direct aqueous aldol reactions of α-hydroxy aldehydes and ketones

Article abstract of DOI:10.1039/b501512j

Zn-proline catalyzed aldolisation of glycoladehyde gave mainly tetroses whereas in the cross-aldolisation of glycoladehyde and rac-glyceraldehyde, pentoses accounted for 60% of the sugars formed with 20% of ribose. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b501512j

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Prebiotic carbohydrate synthesis: zinc–proline catalyzes direct
aqueous aldol reactions of a-hydroxy aldehydes and ketones†
Jacob Kofoed, Jean-Louis Reymond and Tamis Darbre*
Department of Chemistry and Biochemistry, University of Berne, Freiestrasse 3, 3012, Berne,
Switzerland. E-mail: tamis.darbre@ioc.unibe.ch; Fax: +41 (0)31 631 80 57;
Tel: +41(0) 31 631 43 70
Received 31st January 2005, Accepted 22nd March 2005
First published as an Advance Article on the web 14th April 2005
Zn–proline catalyzed aldolisation of glycoladehyde gave mainly tetroses whereas in the cross-aldolisation of
glycoladehyde and rac-glyceraldehyde, pentoses accounted for 60% of the sugars formed with 20% of ribose.
Introduction
The importance of carbohydrates for living organisms, and
especially the prominent role of ribose as building block of
ribonucleic acid (RNA), has prompted studies towards possible
routes for the synthesis of sugars under prebiotic conditions.1
The discovery that RNA acts as both the carrier of genetic
information and as an RNA autoreplicase indicates the existance
of an early “RNA world”,2 where ribozymes could perform
vital chemical functions such as phosphorylation,3 acyl transfer,4
peptide-bond formation5 and carbon–carbon bond formation.6
The early earth was probably rich in organic molecules7 making
it plausible to envision the formation of pentoses and hexoses by
condensation of smaller molecules. Recently, glycolaldehyde was
discovered in interstellar molecular clouds from which new stars
are forming providing circumstantial evidence for the existence
of glycolaldehyde on the early earth.8 Amino acids such as
glycine, alanine, glutamic acid, valine and proline have been
found in meteorites.9
Fig. 1 The two distinct aldolases.18
well as Zn(Arg)2 and Zn(Lys)2 are able to catalyze the direct
reaction between aldehydes and ketones in water with moderate
enantioselectivity (class II aldolase mimic).23 We have recently
reported that unprotected glycolaldehyde can be converted to
mainly tetroses and some hexoses in water using the Zn(Pro)2
complex as catalyst under conditions that are compatible with
a prebiotic environment.24 Zn is an abundant transition metal
and it is found in a large number of metalloenzymes including
aldolases involved in the metabolism of sugars.25 Furthermore,
chiral amino acids were possibly present in the prebiotic period,
either formed on earth or from an extraterrestrial origin.9,11
Herein, we report that in general unprotected a-hydroxy alde-
hydes and ketones can undergo aldol additions in the presence
of Zn(Pro)2 as catalyst in water. Depending on the starting
aldehyde, the products obtained included tetroses, pentonse,
hexoses, keto-pentoses, keto-hexoses and smaller amounts of
higher sugars. For ease of analysis the sugars were also reduced
to polyols using NaBH4.27 The reactions are summarized in
Scheme 1.
The condensation of formaldehyde (the formose reaction)
initited by alkaline earth catalysts leads to a complex mixture
of straight- and branched-chain sugars, including tetroses,
pentoses and hexoses.10 Sugars and especially ribose are however
unstable under such basic reaction conditions undergoing
rapid decomposition.11,12 The yield of pentoses in the formose
reaction has been improved by the addition of lead(II) salts
and dihydroxyacetone.13 Weber showed that rac-glyceraldehyde
could undergo aldolisation catalyzed by iron(III) hydroxide
oxide yielding keto-hexoses that were stable under the reaction
conditions.14 Recently, it has been shown that pentoses, formed
by the aldol reaction of glycolaldehyde and formaldehyde,
can be stabilized by complexation with borate.15 Ribose 2,4-
diphosphate can be obtained as a major product from the reac-
tion of glycoaldehyde phosphate and formaldehyde.16 Tetroses
are also formed by condensation of glycolaldehyde catalyzed by
amino acids.17
In Nature such aldol condensations are promoted by two
distinct classes of aldolases: Class I aldolases contains an active
site lysine involved in the formation of nucleophilic enamine
intermediate,18 or Class II aldolases that contains an active site
Zn(II) cofactor facilitatingthe enolate formation bycoordinating
to the carbonyl oxygen of the ketone donor (Fig. 1).18
Proline and short peptides incorporating N-terminal proline
residues are capable of catalyzing direct aldol reactions in
organic solvents operating via an enamine mechanism19,20,22
(class I aldolase mimics). A few class I mimics operating
in water has also been reported.21 The Zn(Pro)2 complex as
Results and discussion
Synthesis of Zn–proline
Zn(Pro)2 (Fig. 2) is easily synthesized by stirring Zn(OAc)2
(1 eq.), L-proline (2 eq.) and triethylamine (2 eq.) in methanol.
The complex precipitates from the methanolic solution and can
be easily isolated.
Zn–proline catalyzed aldolisation of glycoladehyde
Aqueous glycolaldehyde was stirred for 7 d at room temperature
in the presence of catalytic Zn(Pro)2 (10 mol%) and the product,
obtained after lyophilization, was treated with acetic anhydride.
† Electronic supplementary information (ESI) available: Selected NMR
spectra and GC chromatograms. See http://www.rsc.org/suppdata/
ob/b5/b501512j/
1 8 5 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 5 0 – 1 8 5 5
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
Scheme 1 Zn–proline catalyzed aldol reaction to give sugars and their reduction with NaBH4. The products were identified by GC analysis of the
peracetates. Further identification was obtained from the alditols by GC analysis of the peracetates.
were identified by co-injection with reference peracetylated sug-
ars obtained from commercially available hexoses and treoses.
Sodium borohydride reduction of the crude aldolization
product followed by peracetylation gave a less complex sugar
mixture and allowed the unambiguous identification of the
sugars as their alditol peracetylates.27 Especially, the rare idose
could be identified as iditol with the reference compound
Fig. 2 X-ray structure of zinc–proline complex.26
obtained from NaBH4 reduction of sorbose. Furthermore, the
threitol was formed with 10% enantiomeric excess of the D-
The crude peracetyled sugar mixture was subjected to GC
isomer.24 The tetroses consisted mainly of threose whereas the
analysis, reveling two discrete areas corresponding to tetroses
minor hexose fraction contained mostly mannose and glucose
and hexoses (Fig. 3). The individual sugars present in the mixture
(Table 1).
When the sugar distribution was analyzed regarding the ratio
of the erythrose and threose and the ratio of the combined
hexoses formed from erythrose and the combined hexoses
formed from threoses, it was found that although threose
was the major tetrose present in the mixture after 7 d, the
hexoses formed from erytrose (allose, altrose, glucose and
mannose) predominated over the threo-hexoses (gulose, idose,
galactose and talose) (Fig. 4).28 We attribute the origin of this
stereoselectivity to a faster aldol reaction of the less stable
erythrose with glycolaldehyde (Fig. 4).
In order to determine if the mixture obtained was derived
from equilibration, commercial hexoses and tetroses were stirred
under the reaction conditions. After 7 d, only the initial sugars
were obtained and epimerization was not observed indicating
that the kinetic products formed from glycolaldehyde did not
undergo isomerization.
In an attempt to increase the ratio of hexoses by preventing
cyclization of treoses, we synthesized glycolaldehyde phosphate
by periodate cleavage of 1-phosphate glycerol (Scheme 2).29 This
substrate was however inert to the reaction conditions.
Fig. 3 Crude GC analysis of peracetylation mixture from the aldolisa-
tion of glycolaldehyde.
Table 1 Tetroses and hexoses obtained by aldolization of glycolaldehyde. The products were identified by GC using two different ways; peracetylation
and after reduction and acetylation
Peracetylated aldosesf
Yield from GC (%)
Peracetylated alditols
Yield from GC (%)
Erythrose
Threose
Allose
13
38
5
Erythritola
Threitolb
Allitol
18
33
6
Talose
3
Talitole
2
Altrose
2
4
6
5
N/a
7
19
100
Galactose
Glucose
Gulose
Galactitol
Sorbitolc
4
8
Idosed
Iditol
2
8
19
Mannose
Unidentified
Total
Mannitol
Unidentified
Total
100
a Erythritol is a meso-compound. b Threitol was formed with 10% enantiomeric excess of the D-isomer. c Corresponds to the reduced glucose and
gulose. d Reference sugar was not commercially available. e Corresponds to reduced talose and altrose. f Aldoses correspond to a a/b anomeric mixture.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 5 0 – 1 8 5 5
1 8 5 1

View Article Online
Fig. 4 The acylic forms of D-aldotetroses and -hexoses as their Fischer projections showing the preferential formation of threose (blue) and
erythro-hexoses (black and red).
preferentially. Thus, condensation between glycolaldehyde and
rac-glyceraldehyde is the most favorable reaction, yielding
pentoses as the major components (Fig. 5).
Scheme 2 Synthesis and aldolisation attempt of phosphate glycol-
aldehyde.
Zn–proline catalyzed cross-aldolisation of glycolaldehyde and
rac-glyceraldehyde
The possible formation of pentoses under our reaction condi-
tions was investigated with the cross-aldolization of glycolalde-
hyde and rac-glyceraldehyde. This reaction is of particular in-
terest because, under conditions described as possibly prebiotic,
pentoses proved to be unstable.13 The latter results raised some
questions about the choice of ribose by Nature as the building
block of RNA.
Thus, equimolar amounts of glycolaldehyde and rac-
Fig. 5 The acylic forms of D-aldo-tetroses (red and blue) and pentoses
glyceraldehyde were stirred using the same conditions as above.
as their Fischer projections showing the preferential formation of
pentoses (black and green).
After 7 d, tetroses (20%), pentoses (60%) and hexoses (20%,
complex mixture of aldo- and ketohexoses) were formed. The
reaction mixture was also reduced with NaBH4 to give tetrols,
pentitols and hexitols. The sugars identified by GC analysis
of peracetylated aldols and peracetylated alditols are shown in
Table 2. Interestingly, the percentage of tetroses decreased in this
experiment, whereas the amount of hexose remained unchanged.
rac-Glyceraldehyde, therefore, can act as the electrophile for the
aldol reaction, whereas the glycolaldehyde enolate is formed
The selectivity towards pentoses is remarkable in a reac-
tion run with equimolar amounts of glycolaldehyde and rac-
glyceraldehyde. Ribose accounts for 20% of the total product
and 30% of the pentoses. Ribose alone was stirred in the presence
of Zn–proline for two weeks without epimerization to more
stable pentoses.
Table 2 Tetroses, pentoses and hexoses obtained by cross-aldolization of glycolaldehyde and rac-glyceraldehyde. The products were identified by
GCc after acetylation and after reduction and acetylation and coinjection with prepared references
Peracetylated aldoses
Yield from GC (%)
Peracetylated alditol
Yield from GC (%)
Erythrose
Threose
Arabinose
Lyxose
7
13
14
21
19
9
14
3
100
Erythritol
Threitol
9
11
34
Arabitola
Ribose
Ribitol
Xylitol
Hexoses
Unidentified
Total
13
15
12
6
Xylose
Hexosesb
Unidentified
Total
100
a Corresponds to reduced arabinose and lyxose. b Complex mixture of aldo- and keto-hexoses. Peracetylated aldoses and peracetylated alditols were
analyzed by GC. c In ref. 24 we reported the following distribution of pentoses based on 1H NMR chemical shift of the anomeric protons: arabinose
(21%) ribose (34%), lyxose (32%,) and xylose (13%).
1 8 5 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 5 0 – 1 8 5 5

View Article Online
Table 4 Products obtained from aldol reaction between glycolaldehyde
The aldolization of glycoaldeahyde seems to be slower than
the crossaldolization with only 20% of tetroses formed, and
aldolization of rac-glyceraldehyde does not take place at a
significant rate. Although it is not possible to draw conclusions
about Nature’s choice of ribose over other sugars, it is a
remarkable result indicating that ribose is a viable and stable
sugar under aqueous conditions without additional stabilizing
procedures. These results shows the same trend, however not
as pronounced, as reported by Krishnamurthy et al. who found
ribose (48%), arabinose (16%), lyxose (25%) and xylose (11%)
as their 2,4-biphosphates starting from glycolaldehyde phos-
phate and glyceraldehyde-2-phosphate.16b The isomerisation of
rac-glyceraldehyde into dihydroxyacetone was not significant,
and dihydroxyacetone
Peracetylated alditol (corresponding ketose)
Yield from GC (%)a
Glycerine (dihydroxyacetone)
Erythritol (erythrose)
Threitol (threose)
Xylitol (xylulose)
Ribitol (ribulose)
Unidentified
39
8
16
11
7
20
100
1
however we did note a singlet at ca. 4.3 ppm in the H NMR
spectrum of the crude besides a very complex mixture in the
hexose region of the GC spectra. This prompted us to investigate
the self-condensation of rac-glyceraldehyde which should lead
to ketohexoses.
Total
a Obtained from reduction and peracetylation of the crude sugar mixture
and analysis by GC of the resulting alditols.
Zn–proline catalyzed self-condensation of rac-glyceraldehyde
in the case of aldopentoses (Table 2). Consequently, the cross-
aldolisation between glycolaldehyde and rac-glyceraldehyde
(Table 2) predominates over homo-aldolisation to give a remark-
ably clean reaction considering the many possible products (62%
of pentoses).
When rac-glyceraldehyde alone was stirred in the presence
of Zn(Pro)2,dihydroxyacetone (42%) and keto-hexoses (42%)
(Table 3) were observed after one week. Fructose (26%) ac-
counted for the major part of the keto-hexose mixture. Anomeric
1
protons of aldoses were absent in the H NMR spectrum of
the crude mixture (see electronic supplementary information).
These results indicate that the rac-glyceraldehyde is isomerized
into the more stable dihydroxyacetone via a zinc-chelated
glyceraldehyde enediolate. The isomerisation is apparently much
slower than the addition of glycolaldehyde-enolate to rac-
glyceraldehyde by comparison of the results in Tables 2 and
3. The present results are also in agreement with earlier studies13
of the iron(III) hydroxide oxide catalyzed condensation of
rac-glyceraldehyde (sorbose (15.2%), fructose (12.9%), psicose
(6.1%) and tagatose (5.6%) were reported).
Mechanistic considerations
We propose a Lewis acid catalyzed mechanism for the aldol
reaction mediated by Zn–proline complex. The steps involved in
the proposed mechanism are shown in Fig. 6. Coordination of
hydroxyaldehyde to the metal induces formation of the chelating
enolate. This is reminiscent of the class II aldolases in Nature
that contain a Zn2+ cofactor in the active site facilitating the
enolate formation. The electrophilic aldehyde reacts with the
enolate with or without coordination with zinc.
Zn–proline catalyzed condensation of glycolaldehyde and
dihydroxyacetone
Keto-pentoses were formed from glycolaldehyde and dihydrox-
yacetone as seen in Table 4. Although the yields are given for the
peracetylated alditols obtained by sugar reduction with NaBH4,
1
the crude product of the aldol reaction was analyzed by H
NMR before reduction. The absence of peaks corresponding
to the anomeric protons of aldoses indicated the formation of
ketoses. Thus, no isomerization of dihydroxyacetone to the more
electrophilic rac-glyceraldehyde was seen and it should also be
noted that the conversion into ketopentoses is much lower than
Table 3 Aldol condensation of rac-glyceraldehydes. The products were
identified by GC after acetylation with coinjection of prepared references
Peracetylated ketose
Yield from GC (%)a
Dihydroxyacetone
Fructose
Sorbose
42
26
13
3
16
100
Fig. 6 Proposed Lewis acid catalyzed mechanism for the aldolisation
of glycolaldehyde mediated by Zn–proline complex.
Tagatose
Unidentifiedc
Total
The product distribution obtained by aldolization of glyco-
ladehyde (Table 1) shows a preference for the hexoses formed
by reaction of erythrose with glycoladehyde. The yields of
individual hexoses are too low to rationalize further for a
preferential stereochemistry around the new C2–C3 bond.29
a Obtained from peracetyletion of the crude sugar mixture and anal-
ysis by GC. b Without addition of dihydroxyacetone. c Consists most
probably of psciose and dendro-ketoses. Psicose was not commercially
available at a reasonable price.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 5 0 – 1 8 5 5
1 8 5 3

View Article Online
In the pentose series, however, ribose and lyxose, the two
pentoses with erythro-configuration in the new C2–C3 bond,
are formed in higher ratio (40%) compared to the arabinose
and xylose, the two sugars with threo configuration in the C2–
C3 bond (23%). These results are similar to the ones obtained
in the studies with glycoladehyde phosphate17 and can be
interpreted by analysis of the possible conformers resulting from
the interaction of aldehyde and enolate as shown in Fig. 7. The
conformation A (and the corresponding enantiomer) leads to
ribose and lyxose whereas the less favored conformation B (and
its enantiomer) gives arabinose and xylose.
N2 and H2-flame. The temperature programs were as follows:
(A) starting at 180 ◦C fo◦r 5 min (isotermic) then heating at
5 ◦C ◦min−1 to 200 ◦C; 200 C for 5 min (isotermic) then heating
at 5 C min◦−1 to 220 ◦C; 220 ◦C for 20 min (isotermic) then
heating at 5 C min−1 to 350 ◦C; and (B) starting at 180 ◦C and
heating at 5 ◦C min−1 to 220 ◦C; 220 ◦C for 20 min (isotermic)
then heating at 10 ◦C min−1 to 350 ◦C. All sugar mixtures, except
for reduced pentoses which were analyzed using method (B),
were analyzed using method (A). Data recording was done with
the ChromCard Software and processing done with the free
CHROMuLAN package (http://chromulan.org).
Aldolization of glycolaldehyde
A solution of glycolaldehyde (60 mg, 1 mmol) and Zn(Pro)2–
complex (29 mg, 0.10 mmol) in H2O (5 mL) was stirred for 7 d
at room temperature. The solvent was removed by lyophilization
and the 1H NMR spectrum in D2O was recorded. A portion of
the sugar mixture was peracetylated using the general method
below and injected on GC. Another portion was reduced using
the general method below, and then peracetylated and injected
on GC.
Fig.
7
Possible interactions of glycoaldehyde enolate and rac-
glyceraldehyde with A giving ribose and (in the enantiomeric form)
lyxose and B giving arabinose and xylose.
Cross-aldolisation between glycolaldehyde and
rac-glyceraldehyde
The rate of decomposition of pentoses have been measured
at neutral pH and 100 ◦C. Ribose and lyxose are less stable
than xylose and arabinose14 indicating that the mixture obtained
in the aldol reaction does not represent a thermodynamic
equilibrium.
A
solution of glycolaldehyde (60 mg, 1 mmol), rac-
glyceraldehyde (90 mg, 1 mmol) and Zn(Pro)2–complex (29 mg,
0.10 mmol) in H2O (5 mL) was stirred for 7 d at room
temperature. The solvent was removed by lyophilization and
the 1H NMR spectrum was recorded in D2O. A portion of the
sugar mixture was peracetylated using the general method below.
Another portion was reduced using the general method below,
both samples were analyzed by GC.
Conclusion
Sugars were formed by aldol reaction of simpler aldehydes in
the presence of Zn–proline complex. By studying small sub-
systems of the formose reaction we were able to identify the
reaction of glycoaldehyde and rac-glyceraldehyde as the most
favorable, giving rise to pentoses (62%) with ribose accounting
for ca. 30% of the pentoses. The lack of stability of pentoses
under several conditions has been evoked as the major problem
when considering pentoses as possible prebiotic reagents. Under
basic conditions, condensation of formaldehyde yields less than
1% of ribose.14 Consequently, methods for increasing the yields
and stabilizing the formed pentoses were investigated. We have
shown that ribose and other pentoses can be formed by Lewis
acid catalyzed aldol reactions and that the products are stable
in aqueous media and at room temperature. The Zn–amino acid
complex required as catalyst could be present in the prebiotic
environment. Indeed, we formulated such complexes as class
II aldolase precursors able to catalyze carbon–carbon bond
formation and maybe promote chirality transfer from amino
acid to sugars. Our results leads to think that ribose could
have been present in an early stage of evolution, under possible
prebiotic conditions, and that further stabilization procedures
were not strictly necessary.
Aldol reaction between glycolaldehyde and dihydroxyacetone
A solution of glycolaldehyde (60 mg, 1 mmol), dihydroxyacetone
(90 mg, 1 mmol) and Zn(Pro)2–complex (29 mg, 0.10 mmol) in
H2O (5 mL) was stirred for 7 d at room temperature. The solvent
was removed by lyophilization and the 1H NMR spectrum was
recorded in D2O. The sugar mixture was peracetylated using the
general method below and injected on GC.
Aldol reaction between rac-glyceraldehyde and dihydroxyacetone
A solution of rac-glyceraldehyde (90 mg, 1 mmol), dihydroxyace-
tone (90 mg, 1 mmol) and Zn(Pro)2–complex (29 mg, 0.10 mmol)
in H2O (5 mL) was stirred for 7 d at room temperature.
1
The solvent was removed by lyophilization and the H NMR
spectrum was recorded in D2O. The sugar mixture was reduced
using the general method below, and then peracetylated and
injected on GC.
Sodium borohydride reduction of sugar mixtures
A solution of a sugar mixture in H2O (5 mL) was added NaBH4
(ca. 100 mg, ca. 2.6 mmol) in H2O (10 mL) and was stirred for
3 h, then, treated with acetic acid and lyophilized.
Experimental
NMR spectra and GC chromatograms are found in the supple-
mentary information. Chemicals and solvents were purchased
in the best possible quality from commercial suppliers. For thin-
layer chromatography (TLC), silica gel plates Merck 60 F254
were used and compounds were visualized by treatment with a
solution of phosphomolybdic acid (25 g), Ce(SO4)2·H2O (10 g),
conc. H2SO4 (60 mL), and H2O (940 mL) followed by heating.
Flash chromatography was performed using silicagel Merck 60
(particle size 0.040–0.063 mm), Solvents were distilled before
Peracetylation of the sugar mixtures
The dry residue was stirred for 24 h in a mixture of acetic acid
(2 mL) and pyridine (2 mL) with DMAP (10 mg), then quenched
with H2O (20 mL) and extracted with CH2Cl2 (3 × 20 mL). The
organic phase was extracted with, respectively, 1 N HCl (60 mL),
brine (60 mL) and H2O (60 mL), dried (Na2SO4) and evaporated
to dryness.
1
use. H NMR spectra were recorded on Bruker AVANCE300
(300 MHz) or Bruker DRX500 (500 MHz). Chemical shifts
are given in ppm referred to solvent residual peak. Coupling
constants (J) are reported in Hertz (Hz). GC experiments were
run on a MACHERY-NAGEL OPTIMA d 3 column (30 m ×
0.25 mm) with He (0.8 ml min−1) as carrier gas, makeup gas
Tetrol tetraacetates
The tetrol tetraacetates were separated from the hexitol
hexaacetates by flash chromatography (hexane–ethyl acetate
(2 : 1). 1H NMR (CDCl3, 300 MHz) d 2.04–2.08 (m, 12H,
1 8 5 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 5 0 – 1 8 5 5

View Article Online
4 × OAc), 4.03 (m, 1H, CH2), 4.16 (m, 1H, CH2), 4.30 (m,
12 G. Springsteen and G. F. Joyce, J. Am. Chem. Soc., 2004, 126, 9578–
9583.
13 G. Zubay, Origins Life Evol. Biosphere, 1998, 28, 13–26.
14 A. L. Weber, J. Mol. Evol., 1992, 35, 1–6.
2H, CH2), 5.24 (m, 1H, CH), 5.30 (m, 1H, CH). Anal. chiral GC
◦
(240 C, 106 KPa, 1.35 mL min−1): tR = 90.7 (meso-erythritol
tetraacetate), tR = 109.6 (D-threitol tetraacetate), tR = 113.8
15 A. Ricardo, M. A. Carrigan, A. N. Olcott and S. A. Benner, Science,
2004, 303, 196.
16 (a) D. Mu¨ller, S. Pitsch, A. Kittaka, E. Wagner, C. E. Wintner
and A. Eschenmoser, Helv. Chim. Acta, 1990, 73, 1410; (b) R.
Krishnamurthy, S. Pitsch and G. Arrhenius, Origins Life Evol.
Biosphere, 1999, 29, 139–152.
17 (a) A. L. Weber, Origins Life Evol. Biosphere, 2001, 31, 71–86; (b) S.
Pizzarello and A. L. Weber, Science, 2004, 303, 1151.
18 For a review of chemoenzymatic synthesis of carbohydrates see:
H. J. M. Gijsen, L. Qiao, W. Fitz and C.-H. Wong, Chem.Rev., 1996,
443–473.
19 A proline catalyzed aldolization of propionaldehyde in DMF has
been reported to give carbohydrates with 47% ee: N. S. Chowdari,
D. B. Ramachary, A. Co´rdova and C. F. Barbas III, Tetrahedron
Lett., 2002, 43, 9591; a proline catalyzed asymmetric aldol reaction
to give sugars has also been reported in mixed solvents (DMSO–H2O
10 : 1, dioxane–H2O 10 : 1):; A. Co´rdova, W. Notz and C. F. Barbas
III, Chem. Commun., 2002, 24, 3024.
20 (a) For a recent proline catalyzed synthesis of protected sugars see:
A. B. Northrup, I. K. Mangion, F. Hettche and D. W. C. Macmillan,
Angew. Chem., Int. Ed., 2004, 43, 2152; (b) A. B. Northrup and
D. W. C. Macmillan, Science, 2004, 305, 1752–1755.
21 For amine-catalyzed aldol reactions in water see: T. J. Dickerson and
K. D. Janda, J. Am. Chem. Soc., 2002, 124, 3220; Y. Chen and J.-L.
Reymond, J. Org. Chem., 1995, 60, 6970.
22 (a) J. Kofoed, J. Nielsen and J.-L. Reymond, Bioorg. Med. Chem.
Lett., 2003, 15, 2445–2447; (b) Z. Tang, Z.-H. Yang, L.-F. Cun,
L.-Z. Gong, A.-Q. Mi and Y.-Z. Jiang, Org. Lett., 2004, 6, 2285–
2287; (c) H. J. Martin and B. M. List, SYNLETT, 2003, 12, 1901–
1902.
23 T. Darbre and M. Machuqueiro, Chem. Commun., 2003, 1090.
24 J. Kofoed, M. Machuqueiro, J.-L. Reymond and T. Darbre, Chem.
Commun., 2004, 1540–1541.
25 G. Schultz and M. Dreyer, J. Mol. Biol., 1996, 259, 458.
26 C.-H. Ng, H.-K. Fun, S.-B. Teo, S.-G. Teoh and K. Chinnakali, Acta
Crystallogr., Sect. C, 1995, 51, 244.
(L-threitol tetraacetate), ee ca. 10%. L-Threitol tetraacetate was
1
synthesized as reference for chiral GC and NMR: H NMR
(CDCl3, 300 MHz) d 2.09–2.13 (m, 12H, 4 × OAc), 4.08 (m, 2H,
CH2), 4.36 (m, 2H, CH2), 5.35 (m, 2H, CH). Anal. chiral GC
(240 ◦C, 106 KPa, 1,35 mL min−1): tR = 114.7.
Acknowledgements
This work was supported by the University of Berne, the Otto
Mønsted Foundation, the COST D25 program and the Swiss
National Science Foundation, the COST program, the Otto
Mønsted Fond and the Office Federal Suisse de la Recherche
Scientifique.
References
1 For reviews of prebiotic chemistry see: (a) J. D. Sutherland and J. N.
Whitfield, Tetrahedron, 1997, 53, 11493–11527; (b) R. A. Hughes,
M. P. Robertson, A. D. Ellington and M. Levy, Curr. Opin. Chem.
Biol., 2004, 8, 629–633; (c) S. A. Benner, A. Ricardo and M. A.
Carrigan, Curr. Opin. Chem. Biol., 2004, 8, 672–689.
2 (a) W. Gilbert, Nature, 1986, 319, 618; (b) P. A. Sharp, Cell, 1985, 42,
397–400; (c) for a review on the construction of an “RNA-world” see:
D. P. Bartel and P. J. Unrau, Trends. Cell. Biol., 1999, 12, M9–M13.
3 J. R. Lorsch and J. W. Szostak, Nature, 1994, 371, 31–36.
4 M. Illangasekare, Science, 1995, 267, 643–647.
5 B. L. Zhang and T. R. Cech, Nature, 1997, 390, 96–100.
6 T. M. Tarasow, Nature, 1997, 389, 54–57.
7 S. L. Miller and H. C. Urey, J. Am. Chem. Soc., 2000, 77, 2351.
8 J. M. Hollis, F. J. Lovas and P. R. Jewell, Astrophys. J. Lett., 2000,
540, L107–L110.
9 Comets and the Origin and Evolution of Life, ed. P. J. Thomas,
C. P. McKay and C. F. Chyba, Springer-Verlag, New York, 1997,
pp. 3–19, J. Oro´ and A. Lazcano; pp. 29–62, A. Delsemme; for a
recent update see:; S. Pizzrello, Origins Life Evol. Biosphere, 2004,
34, 25–34.
10 (a) A. Butlerow, Liebigs Ann. Chem., 1861, 120, 295; (b) E. H.
Ruckert, E. Pfeil and G. Scharf, Chem. Ber., 1965, 98, 2558; (c) G.
Harsch, H. Bauer and W. Voelter, Liebigs Ann. Chem., 1984, 623;
(d) N. W. Gabel and C. Ponnamperuma, Nature, 1967, 216, 453;
(e) C. Reid and L. E. Orgel, Nature, 1967, 216, 455; (f) R. F. Socha
and A. H. Weiss, J. Catal., 1981, 67, 207.
27 J. S. Sawardeker, J. H. Sloneker and A. Jeanes, Anal. Chem., 1965,
12, 1602–1604.
28 When erythrose was stirred with glycolaldehyde and Zn(Pro)2 under
the conditions described, a predominant formation of glucose over
the other hexoses is observed by GC; when threose was stirred with
glycolaldehyde and Zn(Pro)2, talose, and galactose were formed in
larger amount.
29 (a) W. R. Jones and P. C. Dedon, J. Am. Chem. Soc., 1999, 121,
9231–9232; (b) a potential prebiotic pathway from glycolaldehyde
to glycolaldehyde phosphate using amidotriphosphate has been
reported by R. Krishnamurthy, G. Arrhenius and A. Eschenmoser,
Origins Life Evol. Biosphere, 1999, 29, 333–354.
11 (a) R. Larralde, M. P. Robertson and S. L. Miller, Proc. Natl. Acad.
Sci. USA, 1995, 92, 8158–8160; (b) R. Shapiro, Origins Life Evol.
Biosphere, 1988, 18, 71–85.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 5 0 – 1 8 5 5
1 8 5 5