L -Rhamnulose-1-phosphate Aldolase from Thermotoga maritima in Organic Synthesis: One-Pot Multistep Reactions for the Preparation of Imino- and Nitrocyclitols

Article abstract of DOI:10.1002/adsc.201500187

Rhamnulose-1-phosphate aldolase from Thermotoga maritima (Rhu1PATm) has been recently cloned and characterised. This hyperthermophilic enzyme offers intriguing possibilities for practical catalysis. This is due to its high stability under extreme reaction

Full text of DOI:10.1002/adsc.201500187

FULL PAPERS
DOI: 10.1002/adsc.201500187
l-Rhamnulose-1-phosphate Aldolase from Thermotoga maritima
in Organic Synthesis: One-Pot Multistep Reactions for the
Preparation of Imino- and Nitrocyclitols
a
b
c,d
Isabel Oroz-Guinea, Karel Hernµndez, Flora Camps Bres,
c,d
c,d,
b,
Christine GuØrard-HØlaine, Marielle Lemaire, * Pere ClapØs, *
and Eduardo García-Junceda *
a,
a
Departmento de Química Bioorgµnica, Instituto de Química Orgµnica General (IQOG–CSIC), Juan de la Cierva 3, 28006
Madrid, Spain
Fax: (+34)-91-564-4853; e-mail: eduardo.junceda@csic.es
Departmento de Química Biológica y Modelado Molecular, Instituto de Química Avanzada de CataluÇa (IQAC–CSIC),
b
Jordi Girona 18-26, 08034 Barcelona, Spain
Fax: (+34)-93-204-5904; e-mail: pere.clapes@iqac.csic.es
Clermont UniversitØ, UniversitØ Blaise Pascal, ICCF, BP 10448, F-63000 Clermont-Ferrand, France
CNRS, UMR 6296, ICCF, BP 80026, F-63171 Aubi›re, France
c
d
Fax: (+33)-4-7340-7717; e-mail: marielle.lemaire@univ-bpclermont.fr
Received: February 13, 2015; Revised: April 14, 2015; Published online: May 18, 2015
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201500187.
Abstract: Rhamnulose-1-phosphate aldolase from Indeed, the diastereoselectivity control was not com-
Thermotoga maritima (Rhu1PATm) has been recent- plete, since mixtures of l-threo (2R,3S; natural ster-
ly cloned and characterised. This hyperthermophilic eopreference) and d-erythro (2R,3R) diastereoiso-
enzyme offers intriguing possibilities for practical mers were obtained as described for Rhu1PAEc.
catalysis. This is due to its high stability under ex- Rhu1PATm was able to catalyse aldol reactions
treme reaction conditions, such as high temperature using nitroaldehydes and N-Cbz-aminoaldehydes.
or the presence of organic co-solvents. The Rhu1- Conversion of the selected nitroaldehydes was com-
PATm potentiality in organic synthesis has been ex- plete, leading to mixtures of l-threo and d-erythro
plored focusing on (i) the reaction stereocontrol, (ii) diastereoisomers, in 95:5 and 75:25 ratios. When N-
the possibility of combining it with other mesophilic Cbz-aminoaldehydes were used, aldol reaction yields
enzymes in multienzyme systems and (iii) its applica- were lower and the stereoselectivity l-threo:d-eryth-
tion to the synthesis of imino- and nitrocyclitols. In ro ranged between 32:68 to>95:5.
our study, the diastereoselectivity of Rhu1PATm was
similar to the one reported for Rhu1PA from Escher-
ichia coli (Rhu1PAEc). However, we observed signif- Keywords: aldol reaction; azasugars; biocatalysis;
icant differences for some aldehyde acceptors. cascade reactions; lyases; thermophilic enzymes
Introduction
such as their ability to carry out the reactions under
moderate conditions of temperature, pH and pressure,
Carbon-carbon (CÀC) bond formation is one of the high efficiency and regio-, chemo- and stereoselectiv-
[
1]
[3]
cornerstone reactions in synthetic organic chemistry. ity. Enzymes also represent an environmentally at-
In this sense, the catalytic asymmetric CÀC bond for- tractive option because they can catalyse reactions in
[
4]
mation by the aldol addition reaction has long been water and are biodegradable. The high stereoselec-
recognized as one of the most useful tools for the con- tivity of aldolases in CÀC bond formation gives them
struction of new CÀC bonds, since the formation of a wide range of synthetic applications. Among the dif-
one or two new stereocentres is concomitant with the ferent families of aldolases, dihydroxyacetone phos-
[2]
bond forming process. The use of enzymes as cata- phate (DHAP)-dependent aldolases have particularly
lysts is a good alternative to non-enzymatic synthesis. attracted the attention of chemists. Furthermore, they
This is because they provide a number of advantages, have been extensively used in chemoenzymatic syn-
Adv. Synth. Catal. 2015, 357, 1951 – 1960
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1951

Isabel Oroz-Guinea et al.
FULL PAPERS
[5]
theses. DHAP-dependent aldolases lead to the for- Results and Discussion
mation of two new stereocentres with a high degree
of stereochemical control. Despite this well proven Study of the Stereochemical Outcome of Rhu1PATm-
synthetic utility of aldolases, there are still few pro- Catalysed Reactions
cesses that have been developed at an industrial scale
and none of them – to the best of our knowledge – We first performed a study on the stereochemical out-
[
6]
used DHAP-dependent aldolases. This is unfortu- come of the reaction catalysed by Rhu1PATm as
nate, considering the recent advances in our under- a function of the temperature. To this end, the per-
standing of the mechanism of aldolase catalysis. This centage of each diastereoisomer formed in the aldol
has allowed the successful application of many differ- addition reaction of DHAP to benzyloxyacetaldehyde
ent engineering approaches to alter the catalytic prop- 2 was measured at different times (Scheme 1). Reac-
erties of aldolases such as stability, substrate specifici- tions were conducted at 48C, 258C, 458C and 808C.
[
7]
ty and stereoselectivity.
As expected, the diasteroisomer l-threo (3R,4S),
One of the main difficulties in applying these pro- which is the one of the natural preference of the
cesses at an industrial scale is the lack of stability enzyme, was synthesised faster (i.e., kinetic product)
under the reaction conditions. An approach to over- than the d-erythro counterpart. In particular, this dif-
come this issue is the use of enzymes from extremo- ference was highlighted at 48C due to the slowing
[
8]
phile microorganisms. From a biocatalytical point of down of the reaction rate (Figure 1A). In all tempera-
view, thermophilic and hyperthermophilic microor- tures assayed, the ratio threo:erythro decreased with
ganisms have attracted particular attention, since they the progress of the reaction, until reaching a plateau
are sources of thermostable enzymes. Thermozymes at a 2:1 ratio (Figure 1).
display outstanding stability against high temperature,
However, thermodynamic equilibrium was reached
as well as against other denaturalisation conditions, faster with the increase of temperature. A similar
such as the presence of organic solvents, detergents or result was obtained when Rhu1PAEc was used with
[
9]
extreme pH values. In this sense, we have recently (R)-N-Cbz-2-aminohexanal as substrate. In this case,
described the heterologous expression of the rhamnu- the diastereoisomer ratio obtained did not vary at dif-
lose-1-phosphate aldolase from Thermotoga maritima, ferent reaction temperatures. However, the l-threo:d-
its functional characterisation and preliminary syn- erythro ratio changed with the temperature when (R)-
[
10]
thetic properties.
This enzyme presented an opti- N-Cbz-2-aminopentanal or N-Cbz-2-amino-4-methyl-
[
12]
mum temperature at 958C and displayed a great sta- pentanal were used as acceptor substrates. Thus, it
bility to extreme reaction conditions, such as high cannot be excluded that the Rhu1PATm stereoselec-
temperatures (i.e., half-life of more than 3 h at tivity could change with the temperature when using
1158C) and in the presence of different organic sol- other aldehydes as substrates.
vents. Furthermore, Rhu1PATm showed activity in
a wide range of pH, with an optimum activity at pH
[
11b]
at 8.0. A multienzyme system
previously devel- Synthesis of Nitrocyclitols Mediated by Rhu1PATm
oped by the García-Junceda group, coupling the in
situ phosphorylation of DHA with the aldolase cata- Nitrocyclitols belong to the large family of cyclitols
lysed aldol reaction, was successfully assayed with and are essentially found as synthetic precursors of
[11]
[13]
Rhu1PATm.
Rhu1PATm was assayed using the aminocyclitols.
These amines are present in many
one-pot/two-steps protocol with four different alde- pharmacologically significant compounds exhibiting
hydes: l-lactaldehyde, benzyloxyacetaldehyde, phe- antitumour, antibacterial, antifungal, antimalarial, and
nylacetaldehyde and ethyl 3-methyl-4-oxocrotonate.
Rhu1PATm was able to react with these four alde-
hydes at 458C and 808C in combination with other
mesophilic enzymes, in a one-pot/two-steps approach.
The reactions were significantly faster at the highest
[10]
temperature.
Herein, we describe a study of the synthetic utility
of the Rhu1PATm, focusing on the stereocontrol of
the aldol reaction, its ability to integrate with other
mesophilic enzymes into multienzyme systems and, fi-
nally, its application to the synthesis of imino- and ni-
trocyclitols.
Scheme 1. Aldol addition reaction of DHAP (1) – in situ ob-
tained by enzymatic phosphorylation of DHA – to benzyl-
oxyacetaldehyde 2 catalysed by Rhu1PATm.
1952
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1951 – 1960

FULL PAPERS
l-Rhamnulose-1-phosphate Aldolase from Thermotoga maritima
[
14]
antiviral activities. Consequently, in the last decade
there has been an increasing demand for the develop-
ment of appropriate synthetic procedures to prepare
cyclitols as well as analogues with enhanced or more
[
15]
selective biological profiles.
Lemaireꢁs group reported a straightforward chemo-
enzymatic cascade to access nitro- and aminocyclitols.
This process involved two stereoselective CÀC bond
formations. One was catalysed by an aldolase and the
other was a spontaneous intramolecular Henry reac-
[
16]
tion. Herein, we have studied the application of the
Rhu1PATm in the context of one-pot/two-steps che-
moenzymatic cascade to access to nitrocyclitols
(
Scheme 2).
Aldehydes 4-nitrobutanal (4) and (Æ)-3-hydroxy-4-
nitrobutanal [(Æ)-6] were used as acceptor substrates
in the aldol addition to DHAP catalysed by Rhu1-
PATm. When the DHAP concentration reached its
wm ea rx ei ma du dm e,d.the aldolase and aldehydes 4 or (Æ)-6
In the first attempt, the aldol addition of DHAP to
4
at a 0.1-mmol scale, with a DHAP/nitroaldehyde
ratio of 2:1 at 458C failed. Unfortunately, nitroalde-
hyde 4 was labile at this temperature and induced
DHAP degradation. Thus, reactions were carried out
Figure 1. Diastereoisomer l-threo:d-erythro ratio obtained
in the aldol reaction of DHAP to 2 catalysed by Rhu1PATm
at 48C (A; ~), 258C (A; ~), 458C (B; *) and 808C (B; *).
Scheme 2. Cascade reactions for the synthesis of nitrocyclitols. First step: phosphorylation of DHA catalysed by recombinant
[18]
ATP-dependent DHAK from Citrobacter freundii CECT 4626 (AK=acetate kinase); second step: aldol reaction catalysed
by Rhu1PA and in situ intramolecular Henry reaction; third step: hydrolysis of the phosphate group using acid phosphatase.
[17]
Data for Rhu1PAEc, taken from ref.
Adv. Synth. Catal. 2015, 357, 1951 – 1960
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1953

Isabel Oroz-Guinea et al.
FULL PAPERS
Table 1. Percentage of aldol adducts (11, 14a–d, and 23a, b) explained because an all-equatorial arrangement of
formed in analytical scale reactions catalysed by Rhu1-
PATm.
substituents (e.g., 7a and almost all equatorial in 7b)
corresponds to the more stable product. The remain-
ing mixture of diastereoisomers can then be re-sub-
Aldehyde
Aldol product
Optimum reaction
time [min]
[a]
[b]
jected to further retroaldol-retro Henry equilibration
[%]
[
19]
and products 7a and 7b were accumulated.
1
1
1
1
1
2
2
0
11 (51)
15
20
30
60
10
10
30
This study was continued using the 3,3-diethoxypro-
panal (8) as an acceptor (Scheme 3). The aldol reac-
tion led to the linear compounds 9a and 9b with an
isolated yield of 67%. Remarkably, both diastereoiso-
mers l-threo and d-erythro were obtained in a 70:30
ratio for the first time in this work using a rhamnu-
lose-1-phosphate aldolase.
3a
3b
3c
3d
2a
2b
14a (61)
14b (74)
14c (39)
14d (49)
23a (37)
23b (38)
[
a]
Percentage of aldol adduct formed determined by HPLC
using an external standard method.
Time at which maximum accumulation of aldol adduct
was reached.
Products 9a and 9b were previously synthesized
using the l-fuculose-1-phosphate aldolase from E. coli
[
b]
[20]
(
Fuc1PAEc). As expected, the percentage obtained
from d-erythro diastereoisomer was higher in the re-
actions catalysed by Fuc1PAEc (l-threo:d-erythro
at 258C. In spite of its hyperthermophile nature, ratio 3:97) than when the catalyst was the Rhu1-
Rhu1PATm retains between 5 and 10% of its maxi- PATm. These compounds could be considered as ni-
[
10]
mum activity at 258C. In these conditions, aldehyde trocyclitol precursors. Indeed, deprotection of the al-
4
led to nitrocyclitols 5a and 5b and aldehyde (Æ)-6 dehyde followed by a Henry reaction with nitrome-
led to nitrocyclitols 7a and 7b (Scheme 2). These com- thane would lead to nitrocyclitols 7a and 7b.
pounds were isolated after purification in 70% and
71% yields for each couple of isomers, respectively.
These compounds are identical to those previously Synthesis of Iminocyclitols Mediated by Rhu1PATm
[17]
described. Their configurations were established on
the basis of the 3R stereoselectivity of the enzyme The second cascade reaction of the Rhu1PATm, with
1
and H NMR analysis.
the DHAK-mediating the DHAP formation was the
The stereoselectivity shown by the aldolase depend- one developed by ClapØsꢁs group for obtaining imino-
ed on the aldehyde acceptor (Table 1). Using the al- cyclitols (Scheme 4). The strategy consists of the aldol
dehyde 4 the major diastereoisomer obtained was the addition of DHAP to either (R)- or (S)-N-Cbz-amino-
l-threo, which corresponds to the expected stereo- aldehydes. This is followed by an acid phosphatase-
chemical preference for the Rhu1PA (Table 1).
catalysed hydrolysis of the phosphate group and sub-
The data presented in this work were consistent sequent intramolecular reductive amination by treat-
with the ones previously described using Rhu1PAEc ment with H in the presence of Pd/C. This strategy
2
[17]
as catalyst. Using (Æ)-6 the actual stereoselectivity allows the formation of three new stereogenic centres
[
12a,21]
of Rhu1PATm could not be disclosed. The reaction for each aldehyde tested (Scheme 4):
two of
produced 7a and 7b and the enzymatic aldol addition them controlled by the aldolase and one generated
appeared to virtually produce exclusively the l-threo during the reductive amination.
diastereoisomer from (R)-6 and the d-erythro from
First, the capability of the Rhu1PATm to catalyse
S)-6 instead of the four possible diastereoisomers the aldol addition reaction of DHAP to N-Cbz-ami-
Supporting Information, Scheme S3). This may be noaldehydes was tested on an analytical scale using
(
(
Scheme 3. Rhu1PATm-catalysed synthesis of aldol products 9a and 9b, which are putative precursors of nitrocyclitols 7a and
b.
7
1
954
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1951 – 1960

FULL PAPERS
l-Rhamnulose-1-phosphate Aldolase from Thermotoga maritima
Scheme 4. Cascade reaction for the synthesis of iminocyclitols employing Rhu1PATm.
[
21a]
an N-Cbz-aminoaldehyde:DHAP ratio of 1:1.7. DMF but required longer reaction times (4–5 h).
For 13c,
(
30% v/v) was added to the reaction medium to im- 13d, 22a and 22b Rhu1PAEc gave superior conver-
[12a,21c]
prove the solubility of the aldehydes. Reaction tem- sions (58–69% conversion).
perature was set up at 458C to retain a good activity
of the aldolase, while preventing a high degradation stituents on the Ca is shared with its mesophilic coun-
While this ability to accept a wide variety of sub-
[
10]
rate of the DHAP.
terpart from E. coli, studies with other DHAP-depen-
Reactions were monitored by HPLC and stopped dent aldolases demonstrated that not all of them ex-
when the formation of aldol adduct was maximal hibit the same degree of tolerance towards structural-
[
12a,22,23]
(
Table 1), thus preventing the spontaneous retroaldol ly diverse N-Cbz-aminoaldehydes.
[
21a]
reaction as has been reported for similar reactions.
After these preliminary studies, reactions for ob-
In these conditions, all the N-Cbz-aminoaldehydes taining the corresponding iminocyclitols were scaled
tested were accepted as substrates by Rhu1PATm, up to the mmol scale (i.e., 2 mmol DHAP; 1.2 mmol
which was able to use aldehydes with both linear and aldehyde). Once the reaction reached the maximum
branched substituent at Ca (Table 1).
aldol adduct formation, the unreacted aldehyde was
Good conversions (61–74%) were obtained with N- removed and aldol adducts 11, 14a–d, 23a and 23b
Cbz-aminoaldehydes with small Ca substituents (e.g., were treated with acid phosphatase. Afterwards,
1
0, 13a and 13b) (Table 2) and moderate (37–49%) aldols were purified by reverse-phase HPLC, paying
with Ca branched substituents including prolinal de- special attention to prevent the losses of possible dia-
rivatives. Using the same aldehyde acceptors, Rhu1- stereoisomers formed during the reaction. Then, aldol
PAEc at 258C gave similar yields for 10 and 13a, 13b adducts were submitted to Cbz deprotection/intramo-
Adv. Synth. Catal. 2015, 357, 1951 – 1960
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1955

Isabel Oroz-Guinea et al.
FULL PAPERS
Table 2. Overall yield and stereochemical outcome obtained for the d-erythro stereochemistry, which is typically
in the synthesis of iminocyclitols catalysed by Rhu1PATm.
obtained with Fuc1PAEc catalyst.
In another example, the aldol addition of DHAP to
R)-N-Cbz-prolinal (22b), Rhu1PATm gave a mixture
of l-threo:d-erythro 65:35, whereas using Rhu1PAEc
Aldehyde
Iminocyclitol
Iminocyclitol
overall yield
l-threo:d-
(
erythro
[a]
[b]
[
%]
[%]
the reaction was reported to be fully stereoselec-
[
23]
1
1
1
1
1
2
2
0
12a+12b
15a
32
19
47
18
12
9
53:47
>95:5
67:33
32:68
>95:5
>95:5
65:35
tive. Given the high sequence identity and structur-
al homology between the Rhu1PA from T. maritima
and E. coli, it could be argued that the lack of selec-
tivity in the Rhu1PATm enzyme is determined for
similar reasons to those already determined for Rhu1-
3a
3b
3c
3d
2a
2b
16b+17b
18c+19c +20c
[c]
21
24a+24b
24c+24d
[
24]
PAEc. Thus, according to Kroemer et al., l-lactal-
dehyde binds to the catalytic centre of the Rhu1PAEc
through two hydrogen bonds, directing the methyl
group to the hydrophobic surroundings. As a result,
the substrate is anchored to the active site of the
enzyme favouring the nucleophilic attack by just one
enantiotopic face. Restrictions imposed by the struc-
ture of the different N-Cbz-aminoaldehydes, as well
as the lack of the hydroxy group, may induce an in-
12
[
[
a]
Purification procedures were not optimized.
b]
Ratio corresponding to the aldol adducts inferred from
the iminocyclitol compounds and determined within the
limits of NMR detection.
[
c]
18c and 19c are epimers at C-2 formed during reductive
amination.
lecular reductive amination to afford the correspond- correct orientation of the aldehyde in the catalytic
[
21a]
ing iminocyclitols, following the method previously pocket.
described.
Moreover, some important differences
[21b]
This procedure provided overall yields may occur in the active site of both aldolases since
ranging from 9 to 47% (Table 2). The structure and Rhu1PATm was not able to use DHA in presence of
relative stereochemistry of the iminocyclitols were de- borate buffer, an ability so far restricted to Rhu1-
[
22]
termined by NMR and in all cases were consistent PAEc.
with those previously described in the literatur-
As previous studies have reported, the stereochem-
[12a,21a,c,23]
e.
As in previous examples of aldol additions istry of the iminocyclitols at C-2 [C-3 in pyrrolizidines
catalysed by DHAP-dependent aldolases, it can be as- 24–27] revealed that the reductive amination with H2
sumed from mechanistic considerations that the abso- in the presence of Pd/C depended on the structure of
lute configuration at C-3 (Scheme 4) (i.e., the stereo- the aldol adduct precursor. In most cases, for exam-
genic centre generated from the DHAP donor) will ple, iminocyclitols 12a, 12b, 15a, 16b, 17b, 21d, 26b
be conserved throughout with any DHAP-aldolases and 27b reductive amination was highly diastereose-
[
5,21,22]
and with no exceptions known so far.
Moreover, lective. Here, the hydrogenation took place from the
the specific absolute configuration introduced by face opposite to the C-4 hydroxy group (C-1 hydroxy
some N-Cbz-aminoaldehyde chiral centres (e.g., 13a– group for the pyrrolizidine derivatives), regardless of
d, 22a and 22b, Scheme 4) allowed the unequivocal the relative stereochemistry of the other substituents.
assessment of the absolute stereochemistry of the imi- Exceptions were observed in the iminocyclitols 18c,
nocyclitols. Thus, from the stereochemical configura- 19c, 24a and 25a.
tion of the iminocyclitols, it was possible to infer the
stereochemical outcome of both the enzymatic aldol
addition and the reductive amination.
Conclusions
Mixtures of l-threo and d-erythro diastereoisomers
were obtained, which depended upon the structure of The work presented herein describes the synthetic
the acceptor aldehyde (Table 2). Rhu1PATm provided scope of the Rhu1PATm. Since this aldolase main-
a good stereoselectivity for the (S)-configured N-Cbz- tains a significant percentage of activity at tempera-
aminoaldehydes furnishing always the l-threo config- tures much lower than its optimum (958C), it is com-
ured aldol adducts while l-threo:d-erythro mixtures patible with multienzyme systems for CÀC bond for-
were formed with the (R)-configured ones (Table 2). mation involving mesophilic enzymes. This enzyme
This behaviour was similar to that observed for Rhu1- was successfully tested in two different chemo-enzy-
PAEc using both DHAP or DHA in the presence of matic cascades for the synthesis of nitrocyclitols and
[
12a,21a,22,23]
borate buffer as donors.
The aldol addition iminocyclitols. In both cases, Rhu1PATm was able to
of DHAP to (R)-N-Cbz-valinal (13c) gave a mixture catalyse the aldol addition using the corresponding al-
of l-threo:d-erythro (32:68) being the d-erythro the dehydes to provide the desired products with similar
major one (Table 2). Interestingly, the same reaction or higher yields than Rhu1PAEc. Additionally, the
catalysed by Rhu1PAEc was fully d-erythro stereose- stereoselectivity shown by the aldolase from T. mariti-
[12a]
lective.
For 13c both enzymes have a preference ma was, in general, similar to the one of the Rhu1-
1956
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1951 – 1960

FULL PAPERS
l-Rhamnulose-1-phosphate Aldolase from Thermotoga maritima
PAEc, even though there were significant differences Stereochemistry Study of the Rhu1PATm-Catalysed
on specific substrates. Our data indicated that the for- Reaction
mation of the l-threo isomer was faster than that of
the d-erythro isomer. No effect of the temperature
was observed in this process.
Stereochemistry of the Rhu1PATm was evaluated by meas-
uring the percentage of each diasteroisomer formed depend-
ing on the reaction time and temperature. Benzyloxyacetal-
dehyde 2 (100 mmol) was dissolved into 600 mL of DMSO
Aldol adduct precursors of iminocyclitols 14a, 14d
and 23a were obtained with full l-threo (3R,4S) ste- (10% final volume) and, then, added to 200 mmol of DHAP.
reoselectivity. The rest aldol adducts from nitroalde- Afterwards, the reaction pH was adjusted to 6.9. Finally, al-
hydes and N-Cbz-aminoaldehydes were obtained as dolase (0.4 U) was added to start the reaction. Four differ-
ent temperatures were chosen to carry out the study: 48C,
room temperature, 458C and 808C. Aliquots of each reac-
mixtures of l-threo (3R,4S) and d-erythro (3R,4R)
diastereoisomers. The major diasteromer formed was
tion were taken at different times and formation of the dia-
the one with the natural stereopreference of the
steroisomer was monitored by HPLC, using an RP-HPLC
enzyme (l-threo), with the exception of aldehyde 13c.
XBridgeꢂ C18, 5 mm, 4.6”250 mm column (Waters). The
elution system used was: solvent (A); 0.1% trifluoroacetic
acid (TFA) in H O (v/v) and solvent (B); 0.095% (v/v) TFA
2
Experimental Section
in CH CN/H O 4:1; gradient elution from 10% to 70% B
3
2
À1
over 30 min. Flow rate was 1 mLmin , detection was at
215 nm. The amount of aldol adduct produced was quanti-
fied from the peak areas using an external standard.
Materials
All solvents and reagents were purchased from Sigma–Al-
drich or Acros. Triosephosphate isomerase (TIM), a-glycer-
ophosphate dehydrogenase (aGDH), phosphatase acid from
potato and acetate kinase (AK) was purchased from Sigma–
Aldrich. Rhu1PA from T. maritima and DHAK from C.
freundii were expressed and purified as previously de-
Synthesis of Nitrocyclitols 5a, 5b, 7a and 7b and
Nitrocyclitol Precursors 9a and 9b
Nitroaldehyde 4 was prepared by mixing acrolein with
a excess of nitromethane with KF as a mild base at À358C,
[10,18]
scribed.
as previously described by Oroz-Guinea et al.
modifications. Acetyl phosphate (7.15 mmol), MgSO4
1.25 mmol) and DHA (5 mmol) were added to phosphate
DHAP was enzymatically prepared from DHA
[16a]
as described by El Blidi et al.
Nitroaldehyde (Æ)-6 was
[11c]
with slight
obtained starting from aldehyde 8, as reported by El Blidi
[16b]
et al.
Finally, aldehyde 8 was prepared by ozonolysis of
(
[25]
the corresponding alkene,
procedure of Hoaglin et al.
previously obtained using the
[26]
buffer (30 mL; 60 mM; pH 7.5) and water to a final reaction
volume of 90 mL. DHAK (431 U) and AK (225 U) were
suspended in ultra-pure water (pHꢀ7.0) and added to the
reaction. To start the reaction ATP (163 mmol) was added.
The phosphorylation reaction was completed after 30 min.
For the storage of the mixture the pH was adjusted to 5.0 to
avoid the degradation of DHAP and was subsequently
frozen. Nickel/iminodiacetic acid (Ni /IDA) agarose was
supplied by Agarose Bead Technologies (Spain). Dialysis
was performed in SPECTRA/Porꢂ membranes (Spectrum)
with 12,000–14,000 Da cut-off.
0
.5 mmol of the corresponding aldehyde – 4 (58.5 mg),
(
Æ)-6 (66.5 mg) or 8 (73.1 mg) – and 3 U of Rhu1PATm
were added to 1 mmol of DHAP, previously adjusted to
pH 6.9. Distilled water was added to a final volume of
3
0 mL. DMSO (10% in the final volume) was added to the
reaction media to dissolve aldehyde 4. The reaction was
kept at room temperature for 24 h with stirring. The pH was
then adjusted to 4.7 with HCl (1M) and Rhu1PATm was
eliminated by centrifugation. Subsequently, acid phospha-
tase from potato (200 U) was added and the reaction was
carried out during 24 h. After this time, the phosphatase was
removed and the resulting solution was concentrated under
vacuum. Reactions were monitored spectrophotometrically
2+
Rhu1PATm Activity Assay
[11b]
The retro-aldol activity of Rhu1PATm was spectrophoto-
measuring the disappearance of DHAP
and by TLC
metrically measured using rhamnulose-1-phosphate (Rhu-
(Merck 60 F254 silica gel TLC plates). The residue was puri-
fied by column chromatography with Merck 60/230–400 and
60/40–63 mesh silica gel, using a solvent gradient from 9:1
to 8:2 of CH Cl /MeOH to afford the nitrocyclitols 5a, b,
1
P) as substrate. Rhu-1P was synthesised using the multien-
[11a]
zyme system described by Sµnchez-Moreno et al.
at
1
3
mmol of DHA scale. The reaction was carried out in
0 mL of HEPES buffer (20 mM; pH 7.5) containing DHA
2
2
and 7a, b and 95:5 of CH Cl /MeOH for the ketone diacetal
2
2
(
(
(
1 mmol), l-lactaldehyde (1.5 mmol), acetyl phosphate
2 mmol), ATP (68 pmol), MgSO4 (250 pmol), ZnCl2
0.3 pmol), DHAK (15 U), AK (30 U) and Rhu1PA from E.
9a, b. Overall yields of the desired products were: 5a+5b
(72 mg, 70%), 7a+7b (79.2 mg, 71%) and 9a+9b (77.1 mg,
1
13
67%). H and C NMR spectra of these products were re-
corded on a Bruker Avance 400 spectrometer in CDCl3,
D O, and CD OD. In all cases, resonance spectra were con-
coli. Activity assays were run at 458C following the decrease
NADH
À1
À1
in absorbance at l=340 nm (e
0 min in a reaction mixture (1 mL) containing Tris-HCl
buffer (40 mM; pH 8.0), NADH (0.2 mmol), aGDH/TIM
2.4 U of aGDH and 24.8 U of TIM), Rhu-1P (10 mmol),
=6220M cm ) for
2
3
1
sistent with those previously described for these com-
[17,19]
pounds.
(
1-(Hydroxymethyl)-6-nitrocyclohexane-1,2,3-triol: 2 iso-
mers. (2R,3S):(2R,3R) ratio 5a:5b, 95:5. Isomer 5a
and Rhu1PATm. 1 unit aldolase activity was defined as the
amount of enzyme that converts 1 mmol of Rhu-1P in
DHAP and l-lactaldehyde per minute in the above de-
scribed conditions.
1
(1R,2R,3S,6S): H NMR (400 MHz, MeOD, 298 K): d=4.79
(dd, J=13.0 Hz, J=4.0 Hz, 1H, H-6), 3.82 (d, J=11 Hz, 1H,
H-7 ), 3.73 (ddd, J=9 Hz, J=11.5 Hz, J=4.6 Hz, 1H, H-3),
B
Adv. Synth. Catal. 2015, 357, 1951 – 1960
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1957

Isabel Oroz-Guinea et al.
FULL PAPERS
3
.37 (d, J=9.3 Hz, 1H, H-2), 3.33 (d, J=11.0 Hz, 1H, H-
2 mmol of DHAP (previously synthesized and adjusted to
pH 6.9) and 10 U of Rhu1PATm. The reaction progress was
followed spectrophotometrically, measuring the aldol
formed by retro-aldol reaction catalysed by Rhu1PA from
E. coli. When a maximum of aldol conversion was achieved,
2 volumes of MeOH were added to stop the reaction and
precipitate the enzyme (10: 105 min; 13a: 90 min; 13b:
75 min; 13c: 150 min; 13d: 150 min; 22a: 60 min; 22b:
90 min). Then, the mixture was filtered through a 0.45 mm
cellulose membrane filter and MeOH was removed under
vacuum. The filtrate was washed with AcOEt (3”100 mL)
to remove the aldehyde that did not react.
Afterwards, the aqueous phase was acidified (pH 4.7) and
phosphatase acid from potato was added (5.3 U per mmol
of phosphorylated product). Dephosphorylation reaction
was followed by HPLC using the same methodology as de-
scribed above. When the product was completely dephos-
phorylated the reaction was stopped by adding two volumes
of MeOH. After removal of the enzyme by filtration, the fil-
trate was adjusted to pH 3.0 and loaded onto a semi-prepa-
rative HPLC column X-Terraꢂ (19/250 mm). Products were
7A), 2.43 (m, 1H, H-5), 2.01 (m, 1H, H-5), 1.97 (m, 1H, H-
4B), 1.32 (m, 1H, H-4 ). Isomer 5b (1R,2R,3R,6S): H NMR
1
A
(
4
400 MHz, MeOD, 298 K): d=4.75 (dd, J=12.4 Hz, J=
.0 Hz, 1H, H-6), 4.00 (m, 1H, H-3), 3.88 (d, J=2.6 Hz, 1H,
H-2), 3.83 (d, J=11.5 Hz, 1H, H-7 ), 3.55 (d, J=11.5 Hz,
B
1
H, H-7 ), 2.41 (m, 1H, H-5), 1.93 (m, 2H, H-5, H-4 ), 1.76
A
B
(
m, 1H, H-4 ).
A
1-(Hydroxymethyl)-6-nitrocyclohexane-1,2,3,5-tetraol:
2
isomers. (2R,3S):(2R,3R) ratio 7a:7b, 75:25. Isomer 7a
1
(
1R,2R,3S,5R,6S): H NMR (400 MHz, MeOD, 298 K): d=
4
1
1
.67 (d, J=10.6 Hz, 1H, H-6), 4.50 (ddd, J=11.0 Hz, J=
1.0 Hz, J=4.9 Hz, 1H, H-5), 3.86 (m, 1H, H-3), 3.77 (d, J=
1.2 Hz, 1H, H-7 ), 3.41 (d, J=9.5 Hz, 1H, H-2), 3.24 (d,
B
J=10.8 Hz, 1H, H-7 ), 2.26 (m, 1H, H-4 ), 1.41 (pq, J=
1
A
B
2.1 Hz, J=12.1 Hz, J=12.1 Hz, 1H, H-4 ). Isomer 7b
A
1
(
1R,2R,3R,5R,6S): H NMR (400 MHz, MeOD, 298 K): d=
4
1
2
.55 (d, J=10.6 Hz, 1H, H-6), 4.43 (ddd, J=11.0 Hz, J=
1.0 Hz, J=4.9 Hz, 1H, H-5), 4.06 (m, 1H, H-3), 3.82 (d, J=
.4 Hz, 1H, H-2), 3.79 (d, J=11.3 Hz, 1H, H-7 ), 3.50 (d,
B
J=11.3 Hz, 1H, H-7 ), 2.02 (m, 1H, H-4 ), 1.82 (ddd, J=
1
A
B
1.9 Hz, J=11.9 Hz, J=10.8 Hz, 1H, H-4 ).
A
eluted using CH CN gradient. The flow rate was
3
6,6-Diethoxy-1,3,4-trihydroxyhexan-2-one:
2
isomers.
À1
1
0 mLmin and the products were detected at 215 nm.
(
3R,4S):(3R,4R) ratio 9a:9b, 70:30. Isomer 9a (3R,4S):
Fractions containing the pure product were pooled and
lyophilized. Overall yields of the desired products were:
12a+12b (112.8 mg, 32%), 15a (68.9 mg, 19%), 16b+17b
(168.2 mg, 47%), 18c+19c+20c (70.0 mg, 18%), 21d
(48.1 mg, 12%), 24a+25a (36.5 mg, 9%), 26b+27b
1
H NMR (400 MHz, CD OD, 298 K): d=4.66 (dd, 1H, J=
3
4
.5 Hz, J=6.8 Hz, H-6), 4.50 (d, 1H, J=19.4 Hz, H-1 ) 4.40
A
(
d, 1H, J=19.4 Hz, H-1 ), 3.88 (m, 1H, H-3), 3.59 (m, 4H,
B
OCH -7, OCH -9), 3.27 (m, 1H, H-4), 1.82 (m, 2H, CH -H-
5
CD OD, 298 K): d=213.47 (C-2), 102.38 (C-6), 79.67 (C-3),
7
1
CD OD, 298 K): d=4.69 (dd, 1H, J=4.1 Hz, J=7.5 Hz, H-
6
H-1 ), 4.05 (m, 6H, H-3, H-4, OCH -7, OCH -9), 1.77 (m,
2
2
2
2
1
3
), 1.15 (m, 6H, CH -8, CH -10); C NMR (100 MHz,
3 3
(
48.0 mg, 12%).
3
The product obtained was dissolved in H O/MeOH (1:1 v/
2
0.29, 67.95, 63.29, 62.72 (C-1, C-4, C-7, C-9), 38.57 (C-5),
v) and treated with H2 (50 psi) in the presence of Pd/C
2 equiv. w/w) at room temperature during 24 h. After that,
1
5.66 (C-8, C-10). Isomer 9b (3R,4R): H NMR (400 MHz,
(
3
the catalyst was removed and the filtrate was adjusted to
pH 6.4 with formic acid and lyophilized obtaining a solid
material.
), 4.48 (d, 1H, J=19.4 Hz, H-1 ) 4.37 (d, 1H, J=19.4 Hz,
A
B
2
2
13
H, CH -H-5), 1.15 (m, 6H, CH -8, CH -10); C NMR
2
3
3
Separation of the isomers of iminocyclitols 16b+17b,
(100 MHz, CD OD, 298 K): d=212.94 (C-2), 102.25 (C-6),
3
1
8c+19c+20c and 26b+27b was performed in a FPLC
7
3
9.77 (C-3), 70.89, 68.12, 63.02, 62.80 (C-1, C-4, C-7, C-9),
7.79 (C-5), 15.66 (C-8, C-10).
system using CM-Sepharose CL-6B (Amersham Pharmacia)
packed in a glass column (45”25 cm). The column was
washed with four volumes (1 vol ~220 mL) of a 100 mM so-
Analytical Scale Synthesis of Iminocyclitols
lution of NH and then equilibrated with four additional vol-
3
umes of distilled H O (pH ~7.0). Iminocyclitols were dis-
2
Analytical scale reactions (1 mL total volume) were con-
ducted in 2-mL test tubes with screw cap incubated at 458C.
solved in 10 mL of H O and the pH of the solution adjusted
2
to 6.0 with formic acid to be loaded in the column. The
15 mmol of aldehyde – 10, 13a and 13b (3.1 mg each); 13c
column was washed with three volumes of distilled H O
2
(
3.5 mg); 13d (3.74 mg); 22a and 22b (3.49 mg each) – were
(
pH ~7.0) and the products were eluted with a 10 mM solu-
dissolved in DMF (300 mL) and then mixed with freshly pre-
pared DHAP neutralised at pH 6.9. Reaction was started by
adding the aldolase (between 0.75 and 1.6 U measured at
À1
tion of NH at a flow of 1 mLmin .
3
Fractions containing the compounds were analysed by
1
13
NMR. H and C NMR spectra of iminocyclitols were re-
corded on a Varian System 400 spectrometer. The com-
pounds described were fully characterised using typical gra-
dient-enhanced 2D experiments: COSY, HSQC, recorded
under routine conditions (see the Supporting Information
for spectra).
4
58C). Reactions were monitored by HPLC, as described
above. To monitor those reactions with aldehydes 13c and
3d, 22a and 22b the gradient elution was changed to 30%
1
to 100% B over 30 min.
Preparative Scale Synthesis of Iminocyclitols
2
-(Hydroxymethyl)piperidine-3,4-diol:
2
isomers.
The general procedure for the production of the aldols 11,
(2S,3S,4S):(2S,3S,4R) ratio 12a:12b 53:47. Isomer 12a
1
1
4 and, 23 was as described above, with only slight varia-
(3S,4S): H NMR (400 MHz, D
3.2, J=12.6 Hz, 1H, H-7 ), 3.73 (dd, J=5.7 Hz, J=12.7 Hz,
1H, H-7 ), 3.6 (dt, J=4.9 Hz, J=9.5 Hz, J=11.6 Hz, 1H, H-
4), 3.13 (t, 1H, H-3), 3.0 (br, 1H, H-6 ), 3.0 (m, 1H, H-2),
2.95 (dt, J=3.2 Hz, J=13.1 Hz, 1H, H-6 ), 1.91 (br, 1H, H-
2
O, 298 K): d=3.87 (dd, J=
tions of the stoichiometry: 1.2 mmol of the corresponding al-
A
dehyde – 10, 13a and 13b (249 mg each); 13c (282 mg), 13d
B
(
2
299 mg); 22a and 22b (278 mg each) – was dissolved in
4 mL of DMF (35% of final volume), followed by the addi-
tion of distilled water (until a final volume of 70 mL),
A
B
3
1
5 ), 1.38 (dq, J=4.3 Hz, 1H, H-5 ); C NMR (125 MHz,
A
B
1958
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1951 – 1960

FULL PAPERS
l-Rhamnulose-1-phosphate Aldolase from Thermotoga maritima
D O, 298 K): d=72.38 (C-4), 72.11 (C-3), 60.57 (C-7), 60.45
2
6 ), 3.83 (m, J=12.7, 6.0 Hz, 1H, H-6 ), 3.25 (m, 1H, H-2),
A B
(
C-2), 44.38 (C-6), 31.44 (C-5). Isomer 12b (3S,4R):
3.20 (m, 1H, H-5), 1.71 (m, 1H, H-8), 1.59 (m, 1H, H-7 ),
A
1
13
H NMR (400 MHz, D O, 298 K): d=4.0 (m, 1H, H-3), 3.80
1.56 (m, 1H, H-7 ), 0.94 (m, 6H, H-9, H-10); C NMR
2
B
(
m, 1H, H-4), 3.2 (m, 1H, H-2), 1.85 (m, 1H, H-5 ), 1.75
(100 MHz, D O): d=80.4 (C-4), 76.3 (C-3), 61.5 (C-2), 60.5
2
A
(
3
m, 1H, H-5 ), 3.76 (dd, 1H, H-7 ), 3.72 (dd, 1H, H-7 ),
(C-5), 58.5 (C-6), 41.4 (C-7), 24.4 (C-8), 22.4 (C-9), 20.9 (C-
10).
B
A
B
1
3
.18 (m, 1H, H-6 ), 2.95 (m, 1H, H-6 ); C NMR
A
B
(
125 MHz, D O, 298 K): d=68.26 (C-4), 66.45 (C-3), 60.17
(7aS)-3-(Hydroxymethyl)hexahydro-1H-pyrrolizine-1,2-
2
(
C-7), 59.65 (C-2), 42.40 (C-6), 25.13 (C-5).
diol: 2 isomers. (1S,2S,3S,7aS):(1R,2S,3S,7aS) ratio 24a:25a,
1
(
5S)-2-(Hydroxymethyl)-5-methylpyrrolidine-3,4-diol
60:40. Isomer 24a (1S,2S,3S,7aS): H NMR (400 MHz, D
2
O,
), 3.78
(bs, 1H, H-2), 3.74 (t, J=8.0 Hz, 1H, H1), 3.65 (m, J=11.8,
1
(
7
15a): H NMR (400 MHz, D O, 298 K): d=3.77 (t, J=
298 K): d=3.79 (m, J=12.2 Hz, J=8.2 Hz, 1H, H-8
A
2
.0 Hz, 1H, H-3), 3.66 (dd, J=4.2 Hz, J=12.0 Hz, 1H, H-
7A), 3.60 (t, J=7.5 Hz, 1H, H-4), 3.58 (dd, J=6.4 Hz, J=
6.5 Hz, 1H, H-8
2.80 (m, 1H, H-5
B
B
), 3.20 (m, 1H, H-7a), 2.94 (bs, 1H, H-5
), 2.75 (m, 1H, H-3), 1.96 (m, 1H, H-7
A
),
),
1
7
7
8
5
2.0 Hz, 1H, H-7 ), 3.13 (dt, J=4.2 Hz, J=6.4 Hz, J=
A
B
1
.88 (m, 1H, H-6 ), 1.79 (m, 1H, H-6 ), 1.76 (m, 1H, H-7 );
.0 Hz, 1H, H-2), 3.06 (dq, J=7.0 Hz, 1H, H-5), 1.16 (d, J=
A B B
1
3
1
3
C NMR (100 MHz, D O): d=79.5 (C-1), 77.5 (C-2), 69.2
.0 Hz, 3H, H-6); C NMR (125 MHz, D O, 298 K): d=
2
2
(
6
2
C-3), 66.6 (C-7a), 55.8 (C-8), 54.8 (C-5), 29.3 (C-7), 24.3 (C-
1.15 (C-4), 76.53 (C-3), 61.67 (C-2), 60.65 (C-7), 56.28 (C-
), 15.95 (C-6).
1
). Isomer 25a (1R,2S,3S,7aS): H NMR (400 MHz, D O,
2
98 K): d=4.33 (dd, J=3.8 Hz, J=1.5 Hz, 1H, H-2), 4.26
(
5R)-2-(Hydroxymethyl)-5-methylpyrrolidine-3,4-diol:
2
(
bt, J=1.9 Hz, 1H, H-1), 4.08 (m, 2H, H-8 , H-8 ), 4.02 (m,
isomers. (2S,3S,4S,5R):(2S,3S,4R,5R) ratio 16b:17b 67:33.
Isomer 16b (2S,3S,4S,5R): H NMR (400 MHz, D O, 298 K):
d=4.01 (dd, J=1.4 Hz, J=3.2 Hz, 1H, H-4), 3.95 (dd, J=
1
1
5
3
3
7
A B
1
1
2
H, H-3), 3.93 (m, 1H, H-7a). 3.50 (m, 2H, H-5 , H-5 ),
A B
2
.41 (m, 1H, H-7 ), 2.23 (m, 1H, H-6 ), 1.87 (m, 2H, H-7 ,
A
A
B
1
3
H-6 ); C NMR (100 MHz, D O): d=78.5 (C-1), 77.5 (C-2),
.8 Hz, J=3.2 Hz, 1H, H-3), 3.84 (dd, J=4.9 Hz, J=
B
2
7
2
4.8 (C-3), 65.8 (C-7a), 55.8 (C-8), 50.1 (C-5), 28.9 (C-7),
5.8 (C-6).
1.9 Hz, 1H, H-7 ), 3.48 (dq, J=3.5 Hz, J=6.5 Hz, 1H, H-
A
), 3.71 (dd, J=8.8 Hz, J=12.3 Hz, 1H, H-7 ), 3.16 (dt, J=
B
(7aR)-3-(Hydroxymethyl)hexahydro-1H-pyrrolizine-1,2-
.3 Hz, J=5.1 Hz, J=8.2 Hz, 1H, H-2), 1.24 (d, J=6.7 Hz,
13
diol: 2 isomers. (1S,2S,3S,7aR):(1R,2S,3R,7aR) ratio 26b:27b,
H, H-6); C NMR (125 MHz, D O, 298 K): d=76.3 (C-4),
2
1
6
5:35. Isomer 26b (1S,2S,3S,7aR): H NMR (400 MHz, D O,
2
6.3 (C-3), 67.5 (C-2), 59.6 (C-7), 56.51 (C-5), 11.8 (C-6).
1
298 K): d=4.04 (dd, J=6.3, 1.4 Hz, 1H, H-1), 3.83–3.73 (m,
Isomer 17b (2S,3S,4R,5R): H NMR (400 MHz, D O, 298 K):
d=4.28 (t, J=3.7 Hz, 1H, H-3), 3.95 (dd, J=3.7 Hz, J=
2
4
1
H, H-2, H-7a, H-8 , H-8 ), 3.13 (m, 1H, H-3), 3.04 (m,
A B
H, H-5 ), 2.83 (dt, J=6.2 Hz, J=4. 3 Hz, 1H, H-5 ), 1.89
A
B
9
3
3
1
2
.0 Hz, 1H, H-4), 3.9 (dd, J=4.7 Hz, J=11.8 Hz, 1H, H-7 ),
.8 (dd, J=8.4 Hz, J=11.8 Hz, 1H, H-7 ), 3.58 (m, J=
A
(
bs, 1H, H-6 ), 1.79 (m, 2H, H-7 , H-7 ), 1.65 (m, 1H, H-
A
A
B
B
1
3
6B); C NMR (100 MHz, D O): d=75.9 (C-2), 72.9 (C-1),
2
.7 Hz, J=4.7 Hz, J=8.4 Hz, 1H, H-2), 3.31 (m, 1H, H-5),
1
3
65.8 (C-7a), 64.6 (C-3), 58.8 (C-8), 54.6 (C-5), 26.9 (C-6),
.32 (d, J=6.7 Hz, 3H, H-6); C NMR (125 MHz, D O,
2
1
2
2.8 (C-7). Isomer 27b (1R,2S,3R,7aR): H NMR (400 MHz,
98 K): d 76.7 (C-4), 70.4 (C-3), 61.5 (C-2), 58.4 (C-7), 55.5
D O, 298 K): d=4.27 (dd, J=3.6, 1.6 Hz, 1H, H-1), 3.94
2
(
C-5), 16.06 (C-6).
-(Hydroxymethyl)-5-isopropylpyrrolidine-3,4-diol: 3 iso-
(
dd, J=4.4 Hz, J=1.5 Hz, 1H, H-2), 3.87 (m, 1H, H-7a),
2
3
8
.70 (m, J=11.2 Hz, J=6.5 Hz, 1H, H-8 ), 3.45 (m, 1H, H-
A
mers.
1
(2R,3R,4S,5R):(2R,3R,4S,5S):(2R,3S,4S,5R)1
ratio
), 3.00 (m, 1H, H-5 ), 2.91 (m, 1H, H-3), 2.78 (m, 1H, H-
A
B
8c:19c:20c, 42:25:32. Isomer 18c (2R,3R,4S,5R): H NMR
5B), 1.97 (m, 1H, H-6 ), 1.87 (m, 1H, H-7 ), 1.82 (m, 2H,
A
A
(400 MHz, D O): d=4.22 (t, J=3.9 Hz, 1H), 4.13 (dd, J=
2
13
H-6 , H-7 ); C NMR (100 MHz, D O): d=77.9 (C-2), 74.2
B
B
2
8
1
1
1
.5, 4.3 Hz, 1H), 3.88 (dd, J=11.5, 6.3 Hz, 1H), 3.75 (dd, J=
1.4, 7.1 Hz, 1H), 3.47 (td, J=6.7, 3.6 Hz, 1H), 3.01 (m,
H), 1.89 (dq, J=13.7, 6.8 Hz, 1H), 1.03 (d, J=6.8 Hz, 3H),
(C-1), 70.3 (C-7a), 69.4 (C-3), 58.8 (C-8), 56.6 (C-5), 29.5 (C-
6), 25.4 (C-7).
1
3
.01 (d, J=6.7 Hz, 3H); C NMR (100 MHz, D O): d=71.4
2
(C-3), 74.2 (C-4), 65.8 (C-5), 60.4 (C-2), 59.9 (C-6), 29.9 (C-
7
), 19.6 (C-8), 18.3 (C-9). Isomer 19c (2R,3S,4S,5R):
1
H NMR (400 MHz, D O): d=4.13 (d, J=4.7 Hz, 1H), 4.00
2
Acknowledgements
(
(
dd, J=2.8, 0.9 Hz, 1H), 3.82 (dd, J=11.6, 5.6 Hz, 1H), 3.74
m, 1H), 3.23 (ddd, J=7.0, 5.6, 2.8 Hz, 1H), 3.01 (m, 1H),
Work at CSIC has been supported by the Spanish Ministerio
de Economia y Competitividad (Grant CTQ2010-19073/
BQU, CTQ2012-31605), Generalitat de Catalunya (2009
SGR 00281), ERA-IB MICINN, PIM2010EEI-00607
(EIB.10.012. MicroTechEnz-EIB., www.fkit.unizg.hr/miten).
COST Action: CM1303 “Systems Biocatalysis” is acknowl-
edged. K. Hernandez and I. Oroz acknowledge the CSIC for
a JAE Predoctoral contract program.
1
3
6
.97 (m, 1H), 1.01 (d, J=3.1 Hz, 3H), 1.00 (d, J=3.1 Hz,
13
H); C NMR (101 MHz, D O): d=74.7 (C-3), 74.6 (C-4),
2
8.2 (C-5), 62.3 (C-2), 58.0 (C-6), 25.7 (C-7), 18.7 (C-8), 18.4
1
(C-8). Isomer 20c: (2R,3S,4S,5R): H NMR (400 MHz, D O,
2
2
3
1
3
6
98 K): d=4.21 (t, J=4.0 Hz, 1H, H-3), 4.12 (m, 1H, H-4),
.95 (dd, J=11.5 Hz, J=5.0 Hz, 1H, H-6 ), 3.74 (dd, J=
1.2 Hz, J=7.4 Hz, 1H, H-6 ), 3.46 (dt, J=3.6 Hz, J=
.1 Hz, 1H, H-2), 3.00 (t, J=8.0 Hz, 1H, H-5), 1.88 (dt, J=
A
B
1
3
.8, 1H, H-7), 1.00 (m, 6H, H-8, H-9); C NMR (100 MHz,
D O): d=76.2 (C-4), 73.4 (C-3), 68.7 (C-5), 64.1 (C-2), 60.2
2
(
C-6), 32.0 (C-7), 21.4 (C-8), 20.5 (C-9).
References
2-(Hydroxymethyl)-5-isobutylpyrrolidine-3,4-diol
(21d):
1
H NMR (400 MHz, D O, 298 K): d=3.93 (t, J=6.9 Hz, 1H,
[1] E. J. Corey, X. M. Cheng, The Logic of Chemical Syn-
thesis; John Wiley & Sons, New York 1989.
2
H-3), 3.83 (m, 1H, H-4), 3.78 (m, J=12.6, 3.9 Hz, 1H, H-
Adv. Synth. Catal. 2015, 357, 1951 – 1960
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1959

Isabel Oroz-Guinea et al.
FULL PAPERS
[
[
2] a) C. J. Li, Chem. Rev. 2005, 105, 3095–3166; b) C.
Palomo, M. Oiarbide, J. M. García, Chem. Soc. Rev.
mations 2, (Eds.: J. Whittall, P. Sutton), John Wiley &
Sons, Ltd, 2012, pp 323–331.
[12] a) J. Calveras, M. Egido-Gabµs, L. Gómez, J. Casas, T.
Parella, J. Joglar, J. Bujons, P. ClapØs, Chem. Eur. J.
2009, 15, 7310–7328; b) A. T. Carmona, J. Fuentes, I.
Robina, E. R. Garcia, R. Demange, P. Vogel, A. L.
Winters, J. Org. Chem. 2003, 68, 3874–3883.
2
004, 33, 65–75.
3] E. García-Junceda, J. F. García-García, A. Bastida, A.
Fernµndez-Mayoralas, Bioorg. Med. Chem. 2004, 12,
1
817–1834.
4] P. Anastas, N. Eghbali, Chem. Soc. Rev. 2010, 39, 301–
12.
[
[
3
[13] X. Chen, Y. Fan, Y. Zheng, Y. Shen, Chem. Rev. 2003,
103, 1955–1977.
5] a) P. ClapØs, J. Joglar, in: Modern Methods in Stereose-
lective Aldol Reactions, (Ed.: R. Mahrwald), Wiley-
VCH Weinheim, 2013, pp 475–527; b) K. Fesko, M.
Gruber-Khadjawi, ChemCatChem 2013, 5, 1248–1272;
c) M. Müller, Adv. Synth. Catal. 2012, 354, 3161–3174;
d) M. Brovetto, D. Gamenara, P. Saenz MØndez, G. A.
Seoane, Chem. Rev. 2011, 111, 4346–4403; e) P. ClapØs,
W.-D. Fessner, Stereoselective Reactions of Carbonyl
and Imino Groups, in: Science of Synthesis. Stereoselec-
tive Synthesis 2, Vol. 2, (Ed.: G. A. Molander), Georg
Thieme Verlag, Suttgart, 2011, pp 677–734; f) P. ClapØs,
W.-D. Fessner, G. A. Sprenger, A. K. Samland, Curr.
Opin. Chem. Biol. 2010, 14, 154–167.
6] a) R. N. Patel, ACS Catal. 2011, 1, 1056–1074; b) S.-H.
Baik, H. Yoshioka, Biotechnol. Lett. 2009, 31, 443–448;
c) M. Wolberg, B. H. N. Dassen, M. Schürmann, S. Jen-
newein, M. G. Wubbolts, H. E. Schoemaker, D. Min,
Adv. Synth. Catal. 2008, 350, 1751–1759; d) J. A. Castil-
lo, J. Calveras, J. Casas, M. Mitjans, M. P. Vinardell, T.
Parella, T. Inoue, G. A. Sprenger, J. Joglar, P. Clapes,
Org. Lett. 2006, 8, 6067–6070; e) M. Müller, Angew.
Chem. 2005, 117, 366–369; Angew. Chem. Int. Ed. 2005,
[
14] a) J. Duchek, D. R. Adams, T. Hudlicky, Chem. Rev.
011, 111, 4223–4258; b) F. Cagide-Fagín, O. Nieto-
García, H. Lago-SantomØ, R. Alonso, J. Org. Chem.
012, 77, 11377–11382.
2
2
[
15] a) R. S. Babu, Q. Chen, S.-W. Kang, M. Zhou, G. A.
OꢁDoherty, J. Am. Chem. Soc. 2012, 134, 11952–11955;
b) Y. Leshch, A. Jacobsen, J. Thimm, J. Thiem, Org.
Lett. 2013, 15, 4948–4951; c) R. Mohanrao, A. Asokan,
K. M. Sureshan, Chem. Commun. 2014, 50, 6707–6710;
d) K. P. Stockton, B. W. Greatrex, D. K. Taylor, J. Org.
Chem. 2014, 79, 5088–5096; e) A. Vidyasagar, K. M.
Sureshan, Eur. J. Org. Chem. 2014, 2349–2356.
[
16] a) L. El Blidi, D. Crestia, E. Gallienne, C. Demuynck,
J. Bolte, M. Lemaire, Tetrahedron: Asymmetry 2004,
[
1
5, 2951–2954; b) L. El Blidi, M. Ahbala, J. Bolte, M.
Lemaire, Tetrahedron: Asymmetry 2006, 17, 2684–2688;
c) L. El Blidi, Z. Assaf, F. Camps Bres, H. Veschambre,
V. ThØry, J. Bolte, M. Lemaire, ChemCatChem 2009, 1,
4
63–471.
[
17] F. Camps Bres, C. GuØrard-HØlaine, V. HØlaine, C. Fer-
nandes, I. Sµnchez-Moreno, M. Traïkia, E. García-Jun-
ceda, M. Lemaire, J. Mol. Catal. B: Enzym. 2015, 114,
4
4, 362–365; f) W. A. Greenberg, A. Varvak, S. R.
5
0–57.
Hanson, K. Wong, H. J. Huang, P. Chen, M. J. Burk,
Proc. Natl. Acad. Sci. USA 2004, 101, 5788–5793.
[
18] I. Sµnchez-Moreno, L. Iturrate, R. Martín-Hoyos, M. L.
Jimeno, M. Mena, A. Bastida, E. García-Junceda,
ChemBioChem 2009, 10, 225–229.
19] a) R. Ballini, L. Barboni, F. Fringuelli, A. Palmieri, F.
Pizzo, L. Vaccaro, Green Chem. 2007, 9, 823–838; b) F.
Camps Bres, C. GuØrard-HØlaine, C. Fernandes, J. A.
Castillo, M. Lemaire, Tetrahedron: Asymmetry 2013,
[
7] For a recent review on engineering of aldolases for use
as industrial biocatalysts, see: C. L. Windle, M. Müller,
A. Nelson, A. Berry, Curr. Opin. Chem. Biol. 2014, 19,
[
2
5–33.
[
8] a) P. Falcicchio, S. Wolterink-Van Loo, M. C. R. Frans-
sen, J. van der Oost, Extremophiles 2014, 18, 1–13; b) F.
Canganella, J. Wiegel, Naturwissenschaften 2011, 98,
2
4, 1075–1081; c) W.-D. Fessner, C. Walter, Top. Curr.
Chem. 1996, 184, 97–194.
253–279; c) L. D. Unsworth, J. van der Oost, S. Koutso-
[
20] W.-D. Fessner, G. Sinerius, A. Schneider, M. Dreyer,
poulos, FEBS J. 2007, 274, 4044–4056.
G. E. Schulz, J. Badia, J. Aguilar, Angew. Chem. 1991,
[
9] a) S. Elleuche, C. Schrçder, K. Sahm, G. Antranikian,
Curr. Opin. Biotechnol. 2014, 29, 116–123; b) H. Atomi,
T. Sato, T. Kanai, Curr. Opin. Biotechnol. 2011, 22,
1
03, 596–599; Angew. Chem. Int. Ed. Engl. 1991, 30,
5
55–558.
[
21] a) L. Espelt, T. Parella, J. Bujons, C. Solans, J. Joglar,
618–626; c) C. Bertoldo, G. Antranikian, in: Enzyme
A. Delgado, P. ClapØs, Chem. Eur. J. 2003, 9, 4887–
Catalysis in Organic Synthesis: A Comprehensive Hand-
4
899; b) L. Espelt, J. Bujons, T. Parella, J. Calveras, J.
Joglar, A. Delgado, P. ClapØs, Chem. Eur. J. 2005, 11,
392–1401; c) J. Calveras, J. Casas, T. Parella, J. Joglar,
nd
book, 2 edn., (Eds.: K. Drauz, H. Waldmann), Wiley-
VCH Weinheim, 2008, pp 313–334; d) H. Atomi, Curr.
Opin. Chem. Biol. 2005, 9, 166–173.
1
P. ClapØs, Adv. Synth. Catal. 2007, 349, 1661–1666.
22] X. Garrabou, J. Calveras, J. Joglar, T. Parella, J. Bujons,
P. ClapØs, Org. Biomol. Chem. 2011, 9, 8430–8436.
[23] X. Garrabou, L. Gomez, J. Joglar, S. Gil, T. Parella, J.
Bujons, P. ClapØs, Chem. Eur. J. 2010, 16, 10691–10706.
[24] M. Kroemer, I. Merkel, G. E. Schulz, Biochemistry
2003, 42, 10560–10568.
[
[
10] I. Oroz-Guinea, I. Sµnchez-Moreno, M. Mena, E.
[
García-Junceda, Appl. Microbiol. Biotechnol. 2015, 99,
3
057–3068.
11] a) I. Sµnchez-Moreno, J. F. García-García, A. Bastida,
E. García-Junceda, Chem. Commun. 2004, 1634–1635;
b) I. Sanchez-Moreno, L. Iturrate, E. G. Doyagüez,
J. A. Martinez, A. Fernandez-Mayoralas, E. Garcia-
Junceda, Adv. Synth. Catal. 2009, 351, 2967–2975; c) I.
Oroz-Guinea, I. Sµnchez-Moreno, E. García-Junceda,
in: Practical Methods for Biocatalysis and Biotransfor-
[25] B. M. Trost, J. Dumas, M. Villa, J. Am. Chem. Soc.
1992, 114, 9836–9845.
[26] R. I. Hoaglin, D. G. Kubler, A. E. Montagna, J. Am.
Chem. Soc. 1958, 80, 5460–5463.
1960
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1951 – 1960