Activated αβ-unsaturated aldehydes as substrate of dihydroxyacetone phosphate (DHAP)-dependent aldolases in the context of a multienzyme system

Article abstract of DOI:10.1002/adsc.200900603

The utility for carbon-carbon bond formation of a multienzyme system composed of recombinant dihydroxyacetone kinase (DHAK) from Citrobacter freundii, the fructose bisphosphate aldolase from rabbit muscle (RAMA) and acetate kinase (AK) for adenosine triphosphate (ATP) regeneration has been studied. Several aldehydes with great structural diversity, including three α,β-unsaturated aldehydes, have been analysed as acceptor substrates. It was found that α,β-unsaturated aldehydes bearing an electron-withdrawing group in the β position to the double bond with a trans configuration are good acceptors for RAMA in this multienzyme system. The aldol reaction proceeds with excellent D-threo enantioselectivity and the aldol adduct is obtained in good overall yield. The L-threo and D-erythro enantiomers are also accessible from rhamnulose 1-phosphate aldolase (Rha-1PA) and fuculose 1-phosphate aldolase (Fuc-1PA) catalysed reactions, respectively.

Full text of DOI:10.1002/adsc.200900603

FULL PAPERS
DOI: 10.1002/adsc.200900603
Activated a,b-Unsaturated Aldehydes as Substrate of
Dihydroxyacetone Phosphate (DHAP)-Dependent Aldolases
in the Context of a Multienzyme System
Israel Sꢀnchez-Moreno,a Laura Iturrate,a Elisa G. Doyagꢁez,a
Juan Antonio Martꢂnez,a Alfonso Fernꢀndez-Mayoralas,a
and Eduardo Garcꢂa-Juncedaa,*
a
Instituto de Quꢂmica Orgꢀnica General, CSIC, Juan de la Cierva 3. 28006 Madrid, Spain
Fax : (+34)-915-644-853; e-mail: eduardo.junceda@iqog.csic.es
Received: September 2, 2009; Published online: November 10, 2009
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200900603.
Abstract: The utility for carbon-carbon bond forma- double bond with a trans configuration are good ac-
tion of a multienzyme system composed of recombi- ceptors for RAMA in this multienzyme system. The
nant dihydroxyacetone kinase (DHAK) from Citro- aldol reaction proceeds with excellent d-threo enan-
bacter freundii, the fructose bisphosphate aldolase tioselectivity and the aldol adduct is obtained in
from rabbit muscle (RAMA) and acetate kinase good overall yield. The l-threo and d-erythro enan-
(AK) for adenosine triphosphate (ATP) regeneration tiomers are also accessible from rhamnulose 1-phos-
has been studied. Several aldehydes with great struc- phate aldolase (Rha-1PA) and fuculose 1-phosphate
tural diversity, including three a,b-unsaturated alde- aldolase (Fuc-1PA) catalysed reactions, respectively.
hydes, have been analysed as acceptor substrates. It
was found that a,b-unsaturated aldehydes bearing an Keywords: aldol reaction; aldolases; biotransforma-
À
electron-withdrawing group in the b position to the tions; C C bond formation; multienzyme processes
Introduction
A straightforward strategy for DHAP preparation
is the kinase-catalysed DHA phosphorylation, using
Aldolases have been extensively used in chemoenzy- ATP as phosphoryl donor. This approach was first de-
matic syntheses because of their ability to catalyse the scribed in 1983 by Wong and Whitesides using the
formation of C C bonds with a high degree of stereo- enzyme glycerol kinase.[6] ATP-dependent DHA kin-
À
chemical control. The main group of aldolases is the ases have also been used for the simple and efficient
one that uses dihydroxyacetone phosphate (DHAP) preparation of DHAP. Itoh et al.,[7] have shown that
as donor.[1] Their major drawback is their strict specif- DHAK isoenzyme I, from Schizosaccharomyces
icity for the donor substrate. Besides the limitations pombe strain IFO 0354, is a useful biocatalyst for the
imposed by this fact, DHAP is expensive to be used production of DHAP.
in large-scale synthesis and labile at neutral and basic
pH. Efforts to overcome the DHAP-dependence of system for one-pot C C bond formation, based on
Our research group has described a multienzyme
[8]
À
aldolases have involved the in situ formation of arsen- the use of a recombinant ATP-dependent DHAK
ACHTUNGTRENNUNG
ate or borate esters of dihydroxyacetone (DHA) from Citrobacter freundii CECT 4626[9] for in situ
which act as phosphate ester mimics,[2] the modifica- DHAP formation and fuculose 1-phosphate aldolase
tion of the substrate specificity of aldolases by in vitro (Fuc-1PA) for the catalysis of the aldol reaction. The
enzyme directed evolution strategies[3] and the use of multienzyme system was completed with the in situ
newly discovered enzymes.[4] In addition to these ef- regeneration of ATP from acetyl phosphate catalysed
forts, an efficient method of DHAP preparation is by acetate kinase (AK). This regeneration system
still essential and several chemical and enzymatic allows the use of ATP in catalytic amounts and avoids
routes for DHAP synthesis have been described.[5]
the accumulation of ADP, which is a strong inhibitor
of the DHAK activity.
Adv. Synth. Catal. 2009, 351, 2967 – 2975
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2967

Israel Sꢀnchez-Moreno et al.
FULL PAPERS
Scheme 1. Multienzyme system for the facile aldol addition reaction between DHAP – in situ obtained by enzymatic phos-
phorylation of DHA – and different aldehydes, catalysed by RAMA.
In our aim to broaden the scope of the multien- DHAP. In spite of the advantages of using acetyl
zyme system, we describe here its use with fructose phosphate as phosphoryl donor, the activity of the
1,6-bisphosphate aldolase from rabbit muscle AK is sensitive to changes in pH. Thus, continuous
(RAMA) for the catalysis of the aldol reaction of al- maintenance of the pH at 7.5 was needed to ensure
dehydes with diverse chemical structures as acceptor the optimal functioning of the system. Since DHAP is
substrates (Scheme 1). Among the aldehydes assayed unstable at pH 7.5, delicate adjustments of the activi-
three a,b-unsaturated aldehydes (compounds 3i, 3j ties of aldolase and/or DHAK were necessary to pre-
and 3l) were included although it has been described vent the accumulation of DHAP and to minimise its
that these aldehydes are not substrates of RAMA.[10] non-enzymatic degradation. The optimum RAMA/
Interestingly, ethyl 3-methyl-4-oxocrotonate (3l) DHAK activity ratio was calculated using acetalde-
was efficiently accepted as substrate by RAMA in the hyde as acceptor. In our first experiment with this
context of this multienzyme system. Therefore, we fo- enzyme using a RAMA/DHAK ratio of 1/1.5, there
cused on the study of the reaction with this aldehyde was considerable accumulation of DHAP, but almost
and explored if it is also a substrate of l-fuculose 1- no formation of the aldol product after 1 hour. Thus,
phosphate aldolase (Fuc-1PA) and l-rhamnulose 1- we decided to increase the RAMA/DHAK activity
phosphate aldolase (Rha-1PA). Structure identifica- ratio to 4/1. Under these conditions, 23 mmol DHAP
tion of the three aldol products obtained and the (23%) accumulated and only 9% of the aldol product
study of the acceptor influence on the steric outcome was formed after 1 hour. At longer reaction times, the
of the aldolase-mediated reactions are reported.
accumulation of DHAP increased but the amount of
the aldol product remained constant. These results in-
dicated that it was necessary to have an initially high
RAMA/DHAK activity ratio and also that the stabili-
ty of RAMA may be low under the reaction condi-
tions. After several attempts, we established that the
initial optimum RAMA/DHAK activity ratio was
11.5/1. Since, the half-life of RAMA was approxi-
Results and Discussion
Optimization of the Multienzyme System with
RAMA
In our previous paper,[8] we described the perfor- mately 5 h under the reaction conditions, it was neces-
mance of our multienzyme system employing a one- sary to supply the reaction with an additional aliquot
pot/one-step strategy. In this approach, all the neces- of 12 U of aldolase after 3 h of reaction (see the Sup-
sary components for the enzymatic phosphorylation porting Information).
of DHA and the aldol addition are present at the be-
ginning of the reaction. The main feature of this pro-
cess is that the DHAP must be formed at the same Screening of the Synthetic Applicability of the
rate as it is consumed by the aldolase.[8] Two condi- Multienzyme System. One-Pot/One-Step Approach
tions are necessary to ensure that this multienzyme
system functions optimally: (i) the pH must be main- Under these conditions, reactions with the selection
tained at 7.5 and (ii) the aldolase/kinase activity ratio of aldehydes (3a–3l, Scheme 1) were carried out in a
must be adjusted to avoid the accumulation of 0.1 mmol of DHA scale in a one-pot/one-step proto-
2968
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2967 – 2975

Activated a,b-Unsaturated Aldehydes as Substrate
FULL PAPERS
Table 1. Formation of aldols 4a–4l catalysed by the DHAK/RAMA multienzyme system.
One-Pot/One-Step
One-Pot/Two-Steps
Aldehyde
Aldol formation [%]
DHA consumption [%]
Aldol formation [%]
DHAP accumulation [%]
3a
3b
3c
3d
3e
53.2
7.8
63.3
52.2
10.8
0.0
70.6
32.1
1.5
98.0
97.4
95.0
96.0
98.2
0.0
97.1
98.0
99.4
22.8
95.1
61.8
4.5
61.5
95.0
>95.0
9.6
58.9
95.0
>95.0
3f[a]
3g[b]
3h
3i
4.1
3.0
81.0
48.7
95.0
3j[c]
3k
3l
1.7
70.5
7.4
>95.0
>95.0
>95.0
[a]
Aldehyde is provided as a 50% dilution in benzyl alcohol (reaction mixture contains 3.3% of benzyl alcohol).
Reaction mixture contains 10% of DMSO.
Reaction mixture contains 5% of DMSO.
[b]
[c]
col. In this initial screening, yields were not opti- One-Pot/Two-Step Approach
mized. For some acceptors the RAMA/DHAK activi-
ty ratio had to be experimentally optimised. The In order to overcome the failures found in the reac-
degree of substrate conversion to product was mea- tions of the third category, we decided to explore a
sured using the retro-aldol activity of the aldolase.[11] one-pot/two-steps strategy and apply it also to the
The results could be grouped into three categories second category reactions and to those aldehydes that
(Table 1): (i) those where the phosphorylation of were in the limit of the first category (aldehydes 3c
DHA worked properly, namely, the percentage of and 3k). With this protocol, only the components for
DHA consumption ranged between 95 and 100% and the enzymatic phosphorylation of DHA are added at
percentage of aldol product formation was over 25% the beginning of the reaction, allowing DHAP to ac-
(aldehydes 3a, 3c, 3d, 3g, 3h and 3k); (ii) those where cumulate. When at least 95% of DHAP was formed,
the extent of DHA phosphorylation was almost com- we reduced the pH to 6.8 – at this pH the degradation
plete (95–100%), but the aldol formation was lower rate of DHAP decreases significantly (see the Sup-
than 25% (aldehydes 3b, 3e and 3i); and (iii) those porting Information) – and then, both aldolase and al-
where the extent of DHA phosphorylation was lower dehyde were added to initiate the aldol addition as a
than 90% or even null reaction (aldehydes 3f, 3j and second step.
3l).
When using this strategy, we were able to restore
As expected, none of the three a,b-unsaturated al- the functionality of the multienzyme system and, in
dehydes was substrate of the RAMA-catalysed reac- most instances of the third category reactions, the per-
tion. In the case of aldehyde 3i, the result obtained – centage of the aldol formation increased significantly
as for other aldehydes belonging to the second cate- (Table 1). Thus, aldol 4f that was unaccessible with
gory – can be explained because this compound is a the one-pot/one-step approach, can be obtained in
poor substrate for the aldolase, however, in the case good yield using this protocol. However, the most in-
of aldehyde 3j and even 3l, the lack of reaction with teresting result was the obtained with the activated
the aldolase could be due to the low amount of a,b-unsaturated aldehyde 3l that in this approach per-
DHAP formed. In fact, the results belonging to the formed as a good substrate for RAMA. On the other
third category show clearly that the phosphorylation hand, the use of the one-pot/two-steps approach pro-
system had not worked properly either because the duced no significant increase in the percentage of
aldehydes themselves or a component in the reaction aldol formation in the reactions belonging to the
medium interfered or inhibited with the DHAK or second category, confirming that these aldehydes are
the AK. In this sense, the presence of benzyl alcohol poor substrates for RAMA. However, when the alde-
in the reaction mixture with aldehyde 3f could ex- hyde is a good substrate for RAMA applying the
plain the results obtained with this substrate.
one-pot/two-step approach the yield of aldol can be
improved, like in the case of 4k which yield increased
from 70% to 81%.
Adv. Synth. Catal. 2009, 351, 2967 – 2975
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2969

Israel Sꢀnchez-Moreno et al.
FULL PAPERS
Figure 1. 1H (A) and 13C (B) NMR spectra of aldol 4l.
Those reactions that gave a percentage of aldol for- sponding to the syn configuration of the hydroxy
mation exceeding 25%, were repeated in a 0.3 mmol groups in the new stereogenic centres (Figure 1).
of DHA scale, the aldol products were purified by
To distinguish between the two syn enantiomers,
precipitation with barium salt[12] and their identity es- we applied the enzymatic assay described by Sheldon
tablished by NMR analysis. Thus, we could validate and co-workers based on the reversibility of the aldol
that the a,b-unsaturated aldehyde 3l was accepted as reaction.[13] Aldol 4l was submitted to retro-aldol re-
substrate by RAMA in the context of this multi- action catalysed by RAMA, Rha-1PA and Fuc-1PA
A
(Figure 2).
a,b-Unsaturated Aldehyde 3l as Substrate of RAMA
In order to prove the synthetic utility of this ap-
proach, we repeated the reaction with aldehyde 3l in
a 3 mmol of DHA scale. In a first attempt, we used
proportionally, a lower amount of RAMA than in the
reaction at the 0.3 mmol scale. In these conditions we
obtained 51.4% of aldol 4l after 26 h of reaction.
Since some degree of RAMA precipitation was ob-
served, it was necessary to add more enzyme along
the reaction time. This precipitation was due to the
presence of 3l, since in its absence protein precipita-
tion was not observed. In subsequent reactions, we in-
creased the amount of RAMA and the reaction was
completed in 8 h with 60% of aldol product forma-
tion. The phosphorylated aldol adduct 4l was purified
1
by HPLC with 54% overall yield. H and 13C NMR of
Figure 2. Enzymatic determination of the stereoisomers
presents in aldol 4l. Retro-aldol reaction catalysed by
the purified aldol showed signals almost exclusively
for one diastereoisomer, presumably the one corre- RAMA (—); Rha-1PA (---) and Fuc-1PA (····).
2970
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2967 – 2975

Activated a,b-Unsaturated Aldehydes as Substrate
FULL PAPERS
Scheme 2.
Scheme 3. Resonance structures of the aldehyde ethyl 3-methyl-4-oxocrotonate (3l).
More than 95% of the aldol adduct was cleaved by Synthesis of the l-threo and d-erythro Enantiomers of
RAMA (d-threo configuration) and only about 4% 4l via Rha-1PA- and Fuc-1PA-Catalysed Reactions
was cleaved by Rha-1PA (l-threo configuration). No
reaction was observed with Fuc-1PA (d-erythro con- Since the aldol adduct 4l is a highly functionalised
figuration) and, althougth we could not perform the molecule and, therefore, of interest in organic synthe-
reaction with tagatose bisphosphate aldolase (l-eryth- sis, we decided to obtain the enantiomers l-threo and
ro configuration), the presence of a stereoisomer with d-erythro accessible via Rha-1PA- and Fuc-1PA-cata-
the corresponding configuration was discarded since lysed reactions, respectively. Reaction conditions were
the the sum of the products with d- and l-threo con- similar to those described for 4l. The highest yield
figurations represented the 100% of aldol 4l. These achieved for aldol adduct l-threo-4l, was 60% after
results clearly showed that the main compound ob- 6 h of reaction. Aldehyde 3l was a poor substrate for
tained was a d-threo configured aldol and that the re- Fuc-1PA. In this case the maximum yield achieved
1
action proceeded with high enantioselectivity.
was only 33% after 20 h of reaction. H and 13C NMR
a,b-Unsaturated aldehydes are not substrates of of the isolated aldols l-threo-4l and d-erythro-4l clear-
RAMA because the electron delocalisation through ly showed that the products were mixtures of two dia-
the double bond reduces the positive charge in the stereoisomers (Figure 3).
carbonyl carbon precluding the nucleophilic attack of
the enamine intermediate formed from the ketone ously assigned to diastereoisomer syn, therefore the
substrate and the enzyme (Scheme 2). other group of signals must correspond to the anti
One group of the signals was identical to that previ-
In the case of the aldehyde 3l, the presence of the diastereoisomer. As expected, in the reaction cata-
ester group introduces an additional resonance struc- lysed by Rha-1PA the major compound is the syn dia-
ture with a negative charge in the carbonyl oxygen of stereoisomer with a syn:anti ratio of 76:24. In the
the ester and a positive charge in alpha position of Fuc-1PA-catalysed reaction the syn:anti ratio is re-
carbonyl of the aldehyde (Scheme 3).
versed (32:68) and the anti diastereoisomer is the
Thus, the presence of an electron-withdrawing sub- major compound. Retro-aldol analysis of compounds
stituent must activate the aldehyde carbonyl group in- l-threo-4l and d-erythro-4l, showed that only the en-
creasing its reactivity. This effect is in agreement with antiomers l-threo (major product in Figure 2A) and
the general observation by Whiteside and co-work- d-erythro (major product in Figure 2B) were present
ers[10] in which electron-withdrawing groups activated in the samples. Although these results are consistent
the carbonyl group against the nucleophilic attack in- with the natural stereochemistry of these enzymes,
creasing, therefore, its reactivity.
they also show that in these cases the aldol reaction
Adv. Synth. Catal. 2009, 351, 2967 – 2975
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2971

Israel Sꢀnchez-Moreno et al.
FULL PAPERS
Figure 3. 1H (A, B) and 13C (C, D) NMR spectra of isolated products from Rha-1PA (A, C) and Fuc-1PA (B, D) catalysed
reactions.
2972
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2967 – 2975

Activated a,b-Unsaturated Aldehydes as Substrate
FULL PAPERS
out by charring with Ce2MoO4. HPLC analyses were carried
out on a chromatograph JASCO, Dual Gradient Pump, with
UV/VIS detector. Optical rotations were determined on a
Perkin–Elmer 241 MC polarimeter for D2O solutions and
are given in 10À1 deg·cm2 gÀ1. Aldehydes 3a–3l were pur-
chased from Sigma–Aldrich or Fluka. Fructose 1,6-bisphos-
phate aldolase from rabbit muscle (RAMA), acetate kinase
(AK) and a-glycerophosphate dehydrogenase/triose phos-
phate isomerase (a-GDH/TIM) were purchased from
Sigma–Aldrich. Rha-1PA, Fuc-1PA and DHAK were cloned
in our laboratory.[9,14] All other chemicals were purchased
from commercial sources as reagent grade.
proceeds with lower enantioselectivity control that in
the RAMA-catalysed reaction.
Conclusions
In conclusion, the described multienzyme system is
robust enough to be used with the three synthetically
useful DHAP-dependent aldolases and with a great
variety of aldehydes. The possibility to apply one-step
or two-steps strategies, makes the system sufficiently
flexible to be able to work under different reaction
conditions such as the presence of co-solvents. Inter-
estingly, within the frame of our studies of this multi-
enzyme system we have found that an a,b-unsaturat-
ed aldehyde bearing an electron-withdrawing group in
the b position to the double bond with a trans config-
uration is a good acceptor for RAMA. Aldol addition
catalysed by this enzyme proceeds with excellent d-
threo enantioselectivity and good conversion. The l-
threo and d-erythro enantiomers are also accessible in
good yield from Rha-1PA- and in moderate yield
from Fuc-1PA-catalysed, reactions respectively. In
both cases, the reaction proceeds with lower enantio-
selectivity than that of the RAMA-catalysed reaction
and a mixture of l-threo and d-erythro enantiomers
was obtained, in each case the major one being that
corresponding to the natural stereochemical prefer-
ence of each enzyme. The aldol products are function-
alised compounds that can be useful intermediates in
organic synthesis.
General Protocol for One-Pot/One-Step Reactions
The one-pot/one-step reactions were carried out in HEPES
buffer (3 mL, 20 mM pH 7.5) containing DHA (0.1 mmol),
aldehyde (0.15 mmol), acetyl phosphate (0.2 mmol) and
MgSO4 (25 mmol), DHAK (1 U), AK (6 U) and RAMA
(11.5 U). After 3 h of reaction an additional amount of
RAMA (12 U) was added due to the instability of the aldo-
lase. To reactions with aldehydes 3g and 3j 10% and 5% of
DMSO was added, respectively, to improve their solubility.
Aldehyde 3f was supplied in a 50% solution in benzyl alco-
hol that afforded 3.3% of benzyl alcohol in the reaction
mixture. Reactions were initiated upon addition of ATP
(6.8 mmol) and their progress was followed spectrophoto-
metrically. A continuous pH adjustment to 7.5 during the re-
action course was needed to keep the ATP regeneration
system working.
General Protocol for One-Pot/Two-Step Reactions
The one-pot/two-steps reaction were carried out in HEPES
buffer (2 mL, 20 mM, pH 7.5) containing DHA (0.1 mmol),
acetyl phosphate (0.2 mmol), MgSO4 (25 mmol), DHAK
(6.5 U) and AK (6 U). The reactions were initiated upon ad-
dition of ATP (6.8 mmol). When DHAP accumulation was
higher than 95%, the pH was adjusted to 6.8 and water
(1 mL) containing the corresponding aldehyde (0.15 mmol)
and RAMA (11.5 U) was added. As before, after 3 h of re-
action more RAMA (12 U) was added. For aldehydes 3g
and 3j 10% and 5% of DMSO, respectively, was added. The
reaction progress was followed spectrophotometrically. For
reactions at the 0.3 mmol DHA scale, the concentrations of
reactives were kept constant. To initiate the aldol reaction
RAMA (51.6 U) was added. Additional aliquots of RAMA
(12.3 U) were added after 1 h and 3 h of reaction.
Experimental Section
General Remarks
1H and 13C NMR spectra, using D2O as solvent, were record-
ed on a Varian System 500 spectrometer equipped with a
5 mm HCN cold probe with field z-gradient, operating at
500.13 and 125.76 MHz for 1H and 13C, respectively. The
sample temperature was maintained constant at 298 K. One-
dimensional NMR experiments were performed using stan-
dard Varian pulse sequences. Two-dimensional [1H, 1H]
NMR experiments (gCOSY and TOCSY) were carried out
with the following parameters: a delay time of 1 s, a spectral
width of 3000 Hz in both dimensions, 4096 complex points
in t2 and 4 transients for each of 256 time increments, and
linear prediction to 512. The data were zero-filled to 4096ꢄ
4096 real points. Two-dimensional [1H-13C] NMR experi-
ments (gHSQC and gHMBC) used the same 1H spectral
window, a 13C spectral windows of 15,000 Hz, 1 s of relaxa-
tion delay, 1024 data points, and 256 time increments, with a
linear prediction to 512. The data were zero-filled to 4096ꢄ
4096 real points. Typical numbers of transients per incre-
ment were 4 and 16, respectively. UV/Visible spectra were
recorded on a SpectraMax Plus 384 spectrophotometer at
258C. TLC was performed on silica-gel plates (GF254
Merck) with fluorescent indicator and detection was carry
Tracking of Reaction Progress
The amount of DHAP formed was quantified by an enzy-
matic assay based in the reduction of DHAP, catalysed by
the a-GDH with concomitant oxidation of NADH to NAD+
340
NADH
monitoring the decrease of absorbance at 340 nm (e
=
6220 cmÀ1 MÀ1).[11] The assays were run at room temperature,
measuring the absorbance at 340 nm for 15 min in a reaction
mixture (1 mL) containing the reaction aliquot, Tris-HCl
(40 mM, pH 8.0), NADH (0.2 mmol) and a-GDH/TIM
(2 U). For the measurement of the DHA consumption,
DHAK (0.375 U), ATP (5 mmol) and MgSO4 (3.75 mmol)
were added to the reaction mixture. The accumulation of
Adv. Synth. Catal. 2009, 351, 2967 – 2975
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2973

Israel Sꢀnchez-Moreno et al.
FULL PAPERS
the aldol product was measured using the retro-aldol activity
of the aldolase (0.05 U).
scribed before but in a 3 mmol of DHA scale. Thus, the re-
action mixture (60 mL) contained: HEPES pH 7.5 (20 mM),
DHA (3 mmol), MgSO4 (0.75 mmol), acetyl-P (6 mmol),
DHAK (66 U) and AK (180 U). The reaction was initiated
upon addition of ATP (0.2 mmol). When DHAP accumula-
tion was higher than 95%, the pH was adjusted to 6.8, the
aldehyde 3l (4.5 mmol) and RAMA (246 U) were added.
After 6 h of reaction another 123 U of RAMA were added.
Compound 4l was purified by HPLC (see the Supporting In-
5-Deoxy-d-threo-pent-2-ulose 1-phosphate (4a): 2 isomers
1
(3:1) – Major isomer: H NMR (500 MHz, D2O, 298 K): d=
4.54 (d, 1H, J=5.9 Hz, H-1A), 4.46 (d, 1H, J=J=6.2 Hz, H-
1B); 4.18 (d, 1H, J=2.3 Hz, H-3); 4.10 (dq, 1H, J=6.6,
2.3 Hz, H-4), 1.07 (d, 3H, J=6.6 Hz, CH3-5); 13C NMR
(125 MHz, D2O, 298 K): d=211.6 (C-2), 78.5 (C-3), 67.9 (C-
4), 67.7 (C-1), 18.4 (C-5). Minor isomer: 1H NMR
(500 MHz, D2O, 298 K): d=4.50 (d, 1H, J=6.2 Hz, H-1A),
4.42(d, 1H, J=J=6.2 Hz, H-1B); 4.24 (d, 1H, J=4.3 Hz, H-
3); 4.0–3.9 (m, 1H, 2.3 Hz, H-4), 0.96 (d, 3H, J=6.6 Hz,
CH3-5).
5-Deoxy-5-ethyl-d-threo-pent-2-ulose 1-phosphate (4c): 2
isomers (10:1) – Major isomer: 1H NMR (500 MHz, D2O,
298 K): d=4.55 (dd, 1H, J=18.7, 6.1 Hz, H-1A), 4.44 (dd,
1H, J=18.7, 6.1 Hz, H-1B); 4.22 (d, 1H, J=2.1 Hz, H-3);
3.87 (dt, 1H, J=7.0, 2.1 Hz, H-4), 1.4–1.3 (m, 2H, CH2-5),
1.3–1.1 (m, 2H, CH2-6), 0.71 (t, 3H, J=7.4 Hz, CH3-7);
13C NMR (125 MHz, D2O, 298 K): d=212.2 (C-2), 77.6 (C-
3), 71.5 (C-4), 68.1 (C-1), 34.5 (C-5), 18.5 (C-6), 13.4 (C-7).
5-Deoxy-5-propyl-d-threo-pent-2-ulose 1-phosphate (4d):
1H NMR (500 MHz, D2O, 298 K): d=4.60 (dd, 1H, J=18.8,
5.9 Hz, H-1A), 4.50 (dd, 1H, J=18.8, 5.9 Hz, H-1B); 4.27 (s,
1H, H-3); 3.90 (t, 1H, J=7.1 Hz, H-4), 1.5–1.4 (m, 2H,
CH2-5), 1.3–1.1 (m, 4H, CH2-6, CH2-7), 0.72 (t, 3H, J=
6.9 Hz, CH3-8); 13C NMR (125 MHz, D2O, 298 K): d=212.1
(C-2), 77.6 (C-3), 71.9 (C-4), 68.1 (C-1), 32.1 (C-5), 27.4 (C-
6), 22.1 (C-7), 13.6 (C-8).
1
formation); [a]2D5: À1.67 (c 0,9, D2O); H NMR (400 MHz,
D2O, 298 K): d=5.85 (s, 1H, H-7), 4.72 (dd, 1H, J=18.3,
7.5 Hz, H-1A), 4.63 (dd, 1H, J=18.3, 7.5 Hz, H-1B), 4.52 (d,
1H, J=2.2 Hz, H-3), 4.48 (s, 1H, H-4), 4.00 (c, 2H, J=
7.1 Hz, H-9), 1.91 (s, 3H, CH3-6), 1.08 (t, 3H, J=7.1 Hz,
CH3-10); 13C NMR (100 MHz, D2O, 298 K): d=208.7 (C-2),
168.9 (C-8), 157.3 (C-5), 116.0 (C-7), 75.5 (C-3), 75.1 (C-4),
69.1 (C-1), 61.3 (C-9), 15.5 (C-6), 13.5 (C-10); EM (IES-
MFE): m/z=311.0536 [M]+; anal. calcd. for C10H15O9P2À
(310.0465): C 38.47, H 5.49; found: C, 38.61; H, 5.75.
l-threo-4l: The procedure was similar to that described
for 4l. 42 U of Rha-1PA were added at the beginning of the
reaction. After 3 h of reaction other 30 U of Rha-1PA were
1
added; H NMR (400 MHz, D2O, 298 K): d=5.73 (s, 1H, H-
7), 4.71 (dd, 1H, J=18.3, 7.5 Hz, H-1A), 4.62 (dd, 1H, J=
18.3, 7.5 Hz, H-1B), 4.31 (d, 1H, J=5.9 Hz, H-3), 4.18 (d,
1H, J=6.1 Hz, H-4), 3.90 (c, 2H, J=7.1 Hz, H-9), 1.99 (s,
3H, CH3-6), 1.08 (t, 3H, J=7.1 Hz, CH3-10); 13C NMR
(100 MHz, D2O, 298 K): d=208.7 (C-2), 168.6 (C-8), 156.2
(C-5), 117.7 (C-7), 76.9 (C-4), 75.2 (C-3), 69.4 (C-1), 61.3 (C-
9), 15.7 (C-6), 14.5 (C-10); EM (IES-EM): m/z=311.0 [M]+,
623.2 (2ꢄ[M]+).
5-Deoxy-5-phenyl-d-threo-pent-2-ulose 1-phosphate (4f):
1
2 isomers (8:1) – Major isomer: H NMR (500 MHz, D2O,
d-erythro-4l: The procedure was similar to that described
for 4l. 110 U of Fuc-1PA were added at the beginning of the
reaction. After 6 h of reaction other 40 U of Fuc-1PA were
298 K): d=7.2–7.1 (m, 5H, Ar), 4.53 (d, 1H, J=6.3 Hz, H-
1A), 4.49 (d, 1H, J=6.3 Hz, H-1B), 4.2–4.1 (m, 2H, H-3, H-
4), 2.9–2.7 (m, 2H, CH2-5); 13C NMR (125 MHz, D2O,
298 K): d=211.9 (C-2), 138.2 (Cq-Ar), 129.3–128.4 (C-Ar),
77.1 (C-3), 73.1 (C-4), 70.8 (C-1), 39.8–38.2 (C-5).
1
added; H NMR (400 MHz, D2O, 298 K): d=5.84 (s, 1H, H-
7), 4.72 (dd, 1H, J=18.3, 7.5 Hz, H-1A), 4.63 (dd, 1H, J=
18.3, 7.5 Hz, H-1B), 4.51 (d, 1H, J=2.2 Hz, H-3), 4.48 (s,
1H, H-4), 4.00 (c, 2H, J=7.1 Hz, H-9), 1.91 (s, 3H, CH3-6),
1.08 (t, 3H, J=7.1 Hz, CH3-10); 13C NMR (100 MHz, D2O,
298 K): d=208.7 (C-2), 168.9 (C-8), 157.3 (C-5), 116.0 (C-7),
75.5 (C-3), 75.1 (C-4), 69.1 (C-1), 61.3 (C-9), 15.5 (C-6), 13.5
(C-10). EM (IES-EM): m/z=311.0 [M]+, 623.2 (2ꢄ[M]+).
5-O-Benzyl-d-threo-pent-2-ulose
1-phosphate
(4g):
1H NMR (500 MHz, DO, 298 K): d=7.2–7.1 (m, 5H, Ar),
4.52 (dd, 1H, J=18.5, 5.8 Hz, H-1A), 4.40 (dd, 1H, J=18.5,
5.8 Hz, H-1B); 4.4–4.3 (m, 2H, CH2Ph), 4.1–4.0 (m, 1H, H-
3), 3.5–3.4 (m, 1H, H-4), 3.5–3.4 (m, 1H, H-4),3.27 (d, 2H,
J=4.9 Hz); 13C NMR (125 MHz, D2O, 298 K): d=212.3 (C-
2), 129.5 (Ar), 129.3 (Ar), 128.9 (Ar), 128.6 (Ar), 75.5 (C-1),
73.1 (C-5), 72.4 (CH2Ph), 70.5 (C-4); 70.2 (C-3).
5-Deoxy-5-(methylthio)methyl-d-threo-pent-2-ulose
1-
1
Acknowledgements
phosphate (4h): H NMR (500 MHz, D2O, 298 K): d=4.55
(dd, 1H, J=18.8, 6.6 Hz, H-1A), 4.45 (dd, 1H, J= J=18.8,
6.6 Hz, H-1B); 4.28 (d, 1H, J=2.2 Hz, H-3); 4.05 (ddd, 1H,
J=14.4, 11.5, 9.2 Hz, H-4), 2.5–2.4 (m, 2H, H-6), 1.95 (s,
3H, Me), 1.8–1.7 (m, 2H, H-5); 13C NMR (125 MHz, D2O,
298 K): d=211.5 (C-2), 77.7 (C-3), 70.4 (C-4), 68.1 (C-1),
29.6 (C-6), 23.5 (C-5), 14.3 (Me).
We thank the Spanish Ministerio de Ciencia e Innovaciꢀn for
financial support (Grant CTQ2007-67403/BQU). I.S.-M and
L.I were supported by a predoctoral fellowship from Comu-
nidad de Madrid. E.G.D is a JAEPredoc fellow from CSIC.
5-O-Methyl-5-methoxy-d-threo-pent-2-ulose 1-phosphate
1
(4k): H NMR (500 MHz, D2O, 298 K): d=4.54 (dd, 1H, J=
References
18.8, 6.1 Hz, H-1A), 4.48 (dd, 1H, J=18.8, 6.1, H-1B), 4.42
(m, 1H, H-3), 4.34 (d, 1H, J=7.5 Hz, H-5), 3.9–3.8 (m, 1H,
H-4), 3.30 (s, 3H, OMe), 3.28 (s, 3H, OMe); 13C NMR
(125 MHz, D2O, 298 K): d=211.3 (C-2), 105.1 (C-5), 75.1
(C-3), 70.9 (C-4), 68.2 (C-1), 56.1 (OMe), 54.8 (OMe).
(E)-5,6-Dideoxy-6-(ethoxycarbonyl)-5-methyl-d-threo-
pent-5-en-2-ulose 1-phosphate (4l): Enzymatic synthesis of
4l was carried out by the one pot/two steps approach as de-
[1] a) L. Hecquet, V. Helaine, F. Charmantray .M. Lem-
aire, in. Modern Biocatalysis: Stereoselective and Envi-
ronmentally Friendly Reactions, (Eds.: W.-D. Fessner, T.
Anthonsen), Wiley-VCH, Weinheim, 2009, pp 287–
298; b) P. Clapes, G. A. Sprenger, J. Joglar, in: Modern
Biocatalysis: Stereoselective and Environmentally
Friendly Reactions, (Eds.: W.-D. Fessner, T. Anthon-
2974
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2967 – 2975

Activated a,b-Unsaturated Aldehydes as Substrate
FULL PAPERS
sen), Wiley-VCH, Weinheim, 2009, pp 299–306; c) L.
Iturrate, E. Garcꢂa-Junceda, in: Multi-Step Enzyme
Catalysis: Biotransformations and Chemoenzymatic
Synthesis, (Ed.: E. Garcꢂa-Junceda), Wiley-VCH, Wein-
heim, 2008, pp 61–81; d) W.-D. Fessner, in: Modern
Aldol Reactions, Vol. 1, (Ed.: R. Mahrwald), Wiley-
VCH, Weinheim, 2008, pp 201–272; e) W.-D. Fessner,
in: Asymmetric Organic Synthesis with Enzymes, (Eds.:
V. Gotor, I. Alfonso, E. Garcꢂa-Urdiales), Wiley-VCH,
Weinheim, 2008, pp 275–318; f) A. K. Samland, G. A.
Sprenger, Appl. Microbiol. Biotechnol. 2006, 71, 253–
264; g) L. J. Whalen, C.-H. Wong, Aldrichimica Acta
2006, 39, 63–71; h) J. Sukumaran, U. Hanefeld, Chem.
Soc. Rev. 2005, 34, 530–542; i) M. Breuer, B. Hauer,
Curr. Opin. Biotechnol. 2003, 14, 570–576; j) W.-D.
Fessner, V. Helaine, Curr. Opin. Biotechnol. 2001, 12,
574–586; k) T. D. Machajewski, C.-H. Wong, Angew.
Chem. 2000, 112, 1406–1430; Angew. Chem. Int. Ed.
2000, 39, 1352–1375; l) N. Wymer, E. J. Toone, Curr.
Opin. Chem. Biol. 2000, 4, 110–119.
Sprenger, J. Joglar, P. Clapes, Org. Lett. 2006, 8, 6067–
6070; d) M. Sugiyama, Z. Hong, P. H. Liang, S. M.
Dean, L. J. Whalen, W. A. Greenberg, C.-H. Wong, J.
Am. Chem. Soc. 2007, 129, 14811–14817; e) X. Garra-
bou, J. A. Castillo, C. Guꢅrard-Hꢅlaine, T. Parella, J.
Joglar, M. Lemaire, P. Clapꢅs, Angew. Chem. 2009, 121,
5629–5633; Angew. Chem. Int. Ed. 2009, 48, 5521–
5525.
[5] For an in-depth discussion of the chemical and enzy-
matic routes to DHAP see: M. Schꢁmperli, R. Pellaux,
S. Panke, Appl. Microbiol. Biotechnol. 2007, 75, 33–45,
and references therein.
[6] C.-H. Wong, G. M. Whitesides, J. Org. Chem. 1983, 48,
3199–3205.
[7] N. Itoh, Y. Tujibata, J. Q. Liu, Appl. Microbiol. Biotech-
nol. 1999, 51, 193–200.
[8] I. Sꢀnchez-Moreno, J. F. Garcꢂa-Garcꢂa, A. Bastida, E.
Garcꢂa-Junceda, Chem. Commun. 2004, 1634–1635.
[9] I. Sꢀnchez-Moreno, L. Iturrate, R. Martꢂn-Hoyos, M. L.
Jimeno, M. Mena, A. Bastida, E. Garcꢂa-Junceda,
ChemBioChem 2009, 10, 225–229.
[10] M. D. Bednarski, E. S. Simon, N. Bischofberger, W. D.
Fessner, M. J. Kim, W. Lees, T. Saito, H. Waldmann,
G. M. Whitesides, J. Am. Chem. Soc. 1989, 111, 627–
635.
[11] H. U. Bergmeyer, Methods of Enzymatic Analysis, 3rd
edn., Verlag Chemie, Deerfield, FL, 1984; Vol. 2,
pp 146–147.
[2] a) D. G. Drueckhammer, J. R. Durrwachter, R. L. Ped-
erson, D. C. Crans, L. Daniels, C.-H. Wong, J. Org.
Chem. 1989, 54, 70–77; b) R. Schoevaart, F. Van Rant-
wijk, R. A. Sheldon, J. Org. Chem. 2001, 66, 4559–
4562; c) M. Sugiyama, Z. Hong, L. J. Whalen, W. A.
Greenberg, C.-H. Wong, Adv. Synth. Catal. 2006, 348,
2555–2559.
[3] M. Sugiyama, Z. Hong, W. A. Greenberg, C.-H. Wong,
Bioorg. Med. Chem. 2007, 15, 5905–5911.
[12] K. K.-C. Liu, T. Kajimoto, L. Chen, Z. Zhong, Y. Ichi-
kawa, C.-H. Wong, J. Org. Chem. 1991, 56, 6280–6289.
[13] R. Schoevaart, F. v. Rantwijk, R. A. Sheldon, Biotech-
nol. Bioeng. 2000, 70, 349–352.
[14] L. Iturrate, PhD thesis, Autꢆnoma University, Madrid,
Spain, 2008.
[4] a) M. Schꢁrmann, G. A. Sprenger, J. Biol. Chem. 2001,
276, 11055–11061; b) M. Schꢁrmann, M. Schꢁrmann,
G. A. Sprenger, J. Mol. Catal. B: Enzymatic 2002, 19–
20, 247–252; c) J. A. Castillo, J. Calveras, J. Casas, M.
Mitjans, M. P. Vinardell, T. Parella, T. Inoue, G. A.
Adv. Synth. Catal. 2009, 351, 2967 – 2975
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2975