Redesign of the phosphate binding site of L -rhamnulose- 1-phosphate aldolase towards a dihydroxyacetone dependent aldolase

Article abstract of DOI:10.1002/adsc.201000719

The aldol addition of unphosphorylated dihydroxyacetone (DHA) to aldehydes catalyzed by L-rhamnulose-1-phosphate aldolase (RhuA), a dihydroxyacetone phosphate-dependent aldolase, is reported. Moreover, a single point mutation in the phosphate binding site

Full text of DOI:10.1002/adsc.201000719

FULL PAPERS
DOI: 10.1002/adsc.201000719
Redesign of the Phosphate Binding Site of l-Rhamnulose-
1-Phosphate Aldolase towards a Dihydroxyacetone Dependent
Aldolase
Xavier Garrabou,a Jesffls Joglar,a Teodor Parella,b Ramon Crehuet,a Jordi Bujons,a
and Pere ClapØsa,*
a
Department of Biological Chemistry and Molecular Modeling, Instituto de Química Avanzada de CataluÇa – CSIC, Jordi
Girona 18-26, 08034 Barcelona, Spain
Fax : (+34)-93-204-5904; phone: (+34)-93-400-6112; e-mail: pere.clapes@iqac.csic.es
Servei de Ressonància Magn›tica Nuclear, Department of Chemistry, Universitat Autònoma de Barcelona, Bellaterra,
b
Spain
Received: September 21, 2010
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.201000719.
Abstract: The aldol addition of unphosphorylated di- than the natural l-lactaldehyde. The RhuA N29D
hydroxyacetone (DHA) to aldehydes catalyzed by l- mutant modified the optimum enzyme design for the
rhamnulose-1-phosphate aldolase (RhuA), a dihy- natural substrate and changed its catalytic properties
droxyacetone phosphate-dependent aldolase, is re- making the aldolase more versatile to other aldol ad-
ported. Moreover, a single point mutation in the ditions of DHA.
phosphate binding site of the RhuA wild type, that
is, substitution of aspartate for asparagine at position Keywords: aldol reaction; amino aldehydes; enzyme
app
max
N29, increased by 3-fold the V
of aldol addition catalysis; l-rhamnulose-1-phosphate aldolase; muta-
reactions of DHA to other aldehyde acceptors rather genesis
Introduction
and protected hydroxy aldehydes, using catalytic
amounts of tertiary amines in solvent-free systems,
Asymmetric aldol additions of dihydroxyacetone furnishing aldol adducts with a high degree of syn rel-
(DHA) to aldehydes play a key role in the synthesis ative stereochemistry, but lacking 1,2-asymmetric in-
of carbohydrates and their analogues.[1–3] Hence, cata- duction.[1,19] Biocatalytically, class I d-fructose-6-phos-
lytic aldol addition methods that enable a precise con- phate aldolase (FSA) isoenzymes from E. coli[20,21]
ꢀ
trol over the stereochemistry of the newly formed C
and a mutant of transaldolase B from E. coli, TalB
C bond are of paramount importance.[4–8] Dihydroxy- F178Y,[22] are the only documented aldolases that
ACHTUNGTRENNUNGacetone phosphate (DHAP) and synthetic equivalents accept DHA, hydroxyacetone, hydroxybutanone and
of DHA, for example, 2,2-dimethyl-1,3-dioxan-5-one glycolaldehyde as donor substrates furnishing syn
(dioxanone), have been used in biocatalysis and orga- (3S,4R) configured aldol adducts.[23–27] Nevertheless,
nocatalysis, respectively.[2,3,9–14] Although the prepara- the access to the whole set of complementary stereo-
tion and synthetic applications of DHAP and DHA selective catalysts for DHA addition is currently a
equivalents have reached a high degree of sophistica- challenge pursued for both organocatalysis and bioca-
tion and efficiency[3,8,15,16] the preferred choice is by talysis.
far the inexpensive DHA, which reduces costs and
improves the atom-economy of the process.
In this work, we discovered that the DHAP-depen-
dent aldolase RhuA wild type and the RhuA N29D
In this regard, Barbas III and co-workers accom- mutant accept DHA as donor substrate, providing ste-
plished the organocatalytic syn-aldol addition of reocomplementary biocatalysts [i.e., furnishing aldol
DHA to aromatic and aliphatic aldehydes in dime- adducts with (3R,4S) stereochemistry] to the FSA.
thylformamide with moderate to good syn/anti dia- This eliminates the need of using arsenate salts
stereoselection.[17,18] Mahrwald and co-workers (>0.5M) or borate buffer (0.2M) in the medium to
achieved the addition of DHA to aliphatic, aromatic form DHA-arsenate or -borate esters, respectively, for
Adv. Synth. Catal. 2011, 353, 89 – 99
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
89

Xavier Garrabou et al.
FULL PAPERS
mimicking the DHAP.[28–30] Moreover, arsenate salts lose-1-phosphate (6) (Table 1), but with just 5.3% of
are not acceptable due to their high toxicity and the the wild type activity, as thus expected since the intro-
DHA-borate ester has shown limited activity for duction of aspartate residues may decrease the affini-
some of the tested aldehydes.[30]
ty for the phosphate anion.
RhuA wild type and the mutants were screened for
the aldol addition of DHA and DHAP to both 1, the
natural acceptor, and 2, (S)-N-Cbz-alaninal, well tol-
erated by RhuA wild type (Scheme 1, Table 2).[37]
The first interesting result was that RhuA wild type
Results and Discussion
To improve the reactivity of RhuA towards DHA a
structure-guided approach was envisaged since pre- accepts DHA as donor substrate[38] (Table 2) furnish-
liminary results obtained by directed evolution did ing adducts with the syn (3R,4S) configuration identi-
not present conclusive activities.[31] The main residues cal to those obtained with DHAP.[37] This represents,
interacting with the phosphate,[32] N29, N32, S75, to the best of our knowledge, the first example of an
T115 and S116 (Figure 1, panel A), were independ- aldol addition of DHA by a class II DHAP-aldo-
ently replaced by aspartate, intending to establish lase.[39] Most interestingly, RhuA N29D was a better
new polar contacts that may stabilize DHA.[33–35] biocatalyst (~2-fold) when a non-natural acceptor al-
RhuA N29D was the most active mutant for the ret- dehyde such as 2 was used (Table 2). The affinity of
roaldol reaction of the natural substrate, l-rhamnu- RhuA wild type for DHA could be envisaged since
crystallographic structures contained DHA in partial
occupation of the DHAP binding site (Figure 1, panel
A).[32] Hence, the results indicate that the enzyme is
Table 1. Retroaldol activities of RhuA wild type and RhuA
mutants[a] on l-rhamnulose-1-phosphate (6).
also able to stabilize the enediolate intermediate even
without the presence of the phosphate group.[40]
RhuA catalyst
Specific activity
[Umgꢀ1]
Relative specific
activity [%]
AHCTUNGTRENNUNG
wild type
N29D
S116D
N32D
3.8
0.2
0.006
0.005
0.002
nd[c]
100.0
5.3
0.2
0.1
0.1
–
S75D
T115D[b]
[a]
RhuA [Co(II)] was utilized through this work since its al-
dolase activity was higher than that of RhuA [Zn(II)].[36]
The substitution of glutamine for serine at the equivalent
position S71 of FucA, caused a large structural modifica-
tion that inactivated the protein, as it was thus observed
in the crystal structure.[33]
[b]
[c]
Scheme 1. RhuA wild type and the mutants catalyzed aldol
addition of DHA and DHAP to l-lactaldehyde (1) and (S)-
N-Cbz-alaninal (2).
nd: <0.001 Umgꢀ1 (1 unit catalyzes the cleavage of
1 mmol of 6 per minute at 258C).
Table 2. Aldol adduct yield and initial reaction rate (vo) of the aldol addition of DHA, DHAP and DHA-borate to 1 and 2.
RhuA
DHA[a]
DHAP[b]
DHA-borate[c]
DHA[d]
DHAP[e]
DHA-borate[f]
[h]
[h]
[h]
[h]
[h]
[h]
5[g]
vo
6[g]
vo
5[g]
vo
7[g]
vo
8[g]
vo
7[g]
vo
wild type
N29D
N32D
S75D
S116D
63
43
nd
26
24
2.4 (100)
1.8 (78)
nd
1.2 (38)
0.6 (24)
100
100
100
100
100
960 (100)
84 (9)
102 (11)
36 (4)
100
75
nd
100
100
240 (100)
2.4 (1)
nd
18 (8)
12 (5)
40
90
nd
18
35
2.4 (100)
5.4 (225)
nd
1.8 (75)
1.8 (75)
60
60
40
15
55
240 (100)
234 (98)
36 (15)
6 (3)
100
84
nd
18
84 (100)
4.8 (6)
nd
3.6 (4)
2.4 (3)
102 (11)
72 (30)
20
[a]
[3]=100 mM, [1]=40 mM; triethanolamine hydrochloride (TEA) buffer 50 mM pH 7.0, 258C, 24 h reaction time.
[4]=50 mM, [1]=40 mM, medium adjusted to pH 6.9, 258C, 24 h reaction time.
[b]
[c]
[d]
[e]
[f]
[3]=100 mM, [1]=40 mM, borate buffer 200 mM pH 7.0, 258C, 24 h reaction time.
[3]=100 mM, [2]=65 mM; TEA buffer 50 mM pH 7.0/dimethylformamide 4:1, 258C, 24 h reaction time.
[4]=50 mM, [2]=65 mM, medium adjusted to pH 6.9, 258C, maximum productivity achieved after 4 h.
[3]=100 mM, [2]=65 mM, borate buffer 200 mM pH 7.0, 258C, 24 h reaction time.
Percentage of aldol adduct formed relative to the limiting substrate.
[g]
[h]
mmol of aldol adduct formed per h and per mg of protein (mmol hꢀ1 mgꢀ1), in parenthesis percentage of vo respect to the
wild type; nd: not detected.
90
asc.wiley-vch.de
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 89 – 99

Redesign of the Phosphate Binding Site of l-Rhamnulose-1-Phosphate Aldolase
Figure 1. Panel A: crystal structure of the RhuA wild type active center (PDB code 1OJR) showing in cyan the five residues
on the phosphate binding site (N29, N32, S75, T115 and S116) subjected to mutation. A bound phosphate (red) and a DHA
molecule (green) coordinated to the essential Zn(II)(light blue) are also shown. Panels B and C: minimized and superposed
modeled complexes of 6 (B) and 7 (C) into the active center of both RhuA wild type (green) and the N29D mutant (pink).
Residue N29 is shown in cyan and the D29 mutation in yellow. The mechanistically important residues R28, E117 and E171’
are labeled; a prime denotes that the residue belongs to the neighboring subunit. The conformation shown for residues
E171’ and R28 differs from that observed in the 1OJR crystal structure to reflect the putative arrangement of these residues
at physiological pH.[32] The new arrangement for these residues was obtained by manual modification of the crystal structure
followed by extensive minimization. Molecular dynamics simulations confirmed the stability of these models (see Supporting
Information). Panels D and E: minimized modeled complexes of 6 (D) and 5 (E) bound into the active center of RhuA
N32D. The mutated D32 (yellow) is shown forming a hydrogen bond with the phosphate group of 6 (D), or the catalytic car-
boxylate group of residue E117 (E). The mutated D32 (yellow) is shown forming a hydrogen bond with (D) the phosphate
group of 6, or (E) the catalytic carboxylate group of residue E117. The stability of these modeled structures was further as-
sessed by running molecular dynamics simulations with explicit solvent, which showed that the above hydrogen bonds are
ꢀ
ꢀ
maintained during the whole simulation (average distance O H···O: (D) 1.59ꢁ0.11 Š, (E) 1.60ꢁ0.09 Š; average O H···O
angle: (D) 162ꢁ8 degrees, (E) 167ꢁ8 degrees; see Supporting Information).
Adv. Synth. Catal. 2011, 353, 89 – 99
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
91

Xavier Garrabou et al.
FULL PAPERS
Table 3. Steady-state kinetic parameters[a] for the aldol addition reaction of DHA and DHAP to 2.
DHAP
DHA
Biocatalyst
V
/Kampp [d]
V
/Kampp [d]
app
max
app
max
app ½bꢂ
app ½cꢂ
app ½bꢂ
app ½cꢂ
V
K
V
K
max
m
max
m
RhuA wild type
RhuA N29D
606ꢁ36
702ꢁ12
0.6ꢁ0.2
7ꢁ1
1.0
0.1
30ꢁ2
1125ꢁ142
1339ꢁ236
3 10ꢀ5
8 10ꢀ5
108ꢁ6
[a]
The parameters were determined by a non-linear regression analysis of untransformed data direct to the Michaelis–
Menten kinetic model. The steady state kinetic parameters of 2 for the aldol addition using DHAP as donor were:
Vampapx =312ꢁ30 mmol hꢀ1 mgꢀ1 and Kmapp =15ꢁ4 mM, therefore a fixed [2]=65 mM was used in the assays.[41]
Vmax (mmol of aldol adduct formed hꢀ1 mgꢀ1 of protein).
[b]
[c]
[d]
Kampp (mM).
app
V
/Kampp (hꢀ1 mgꢀ1).
max
To gain insight into these results, the apparent the 1-OH can act as a hydrogen bond donor. This
steady-state kinetic parameters of DHA and DHAP new hydrogen bond could modify the energetics of
ꢀ
donors for RhuA wild type and RhuA N29D were de- the C3 C4 bond breaking/formation. To prove this,
termined for the aldol addition to 2 (Table 3).
the simplest possible model of the active site was gen-
As judged by the Kampp values, the affinity of DHAP erated (see the Supporting Information) and DFT cal-
for the RhuA N29D is ~12-fold lower than that for culations to determine the effects of this hydrogen
the wild type, however the DHAP is still a much bond on the energy barrier of the transformation
better donor than DHA for the RhuA N29D. Both were conducted. The calculations revealed that the
RhuA wild type and the N29D mutant had compara- hydrogen bond interaction between the 1-OH of the
app
max
ble V
values (Table 3) for the aldol addition of Co(II)-bound DHA and the carboxylate group of the
DHAP to 2. Hence, it appears that, for this reaction, acetate, mimicking the D29 residue, reduced the
the mutation N29D affects mainly the Kampp of DHAP. energy barrier for the formation of the C C bond
ꢀ
On the other hand, both RhuA wild type and the with the aldehyde by 3.1 kcalmolꢀ1, which is the rate-
N29D mutant exhibited a similar Kampp for DHA. This limiting step when other acceptors rather than the
indicates that the improved activity of the N29D natural one (1) are used.[42–46] This energy reduction
mutant towards DHA for this particular reaction was cannot be directly compared to change in activation
app
max
due to an increase of its V
(i.e. ~3.6-fold with re- barrier of the enzyme, since electrostatic and dynamic
spect to RhuA wild type), and thus is consistent with effects tend to reduce the net effect of a mutation.[47]
the observed vo values (Table 2) at 100 mM [DHA].
Therefore, the value of 3.1 kcalmolꢀ1 can be consid-
Contrary to what was observed with the acceptor 2, ered as an upper limit of the H-bond effect, but it
the vo of RhuA wild type was much higher than that substantiates that the presence of D29 could reduce
of the N29D mutant for the aldol addition of DHAP the activation energy of the RhuA-promoted aldol
to 1, under the screening conditions (i.e., 50 mM addition reaction of substrates like 2.
[DHAP]). It is assumed that the residues N29 and
No improvement on the activity towards DHA and
E171ꢁ can interact and fix the position of the 5-OH a similar or stronger decrease of the activity (vo,
group of 6 (Figure 1, panel B), or, in the synthetic di- Table 1) towards DHAP, relative to RhuA N29D, was
rection, that of the 2-OH of 1.[32] The N29D mutation observed for the rest of the mutants studied. This
precludes one of the two putative hydrogen bonds might correlate with their inability to establish similar
and, consequently, it probably affects the binding of hydrogen bonding interactions with DHA to those
both the DHAP and 1.
found in RhuA N29D and with the closer proximity
value in the of the mutations to the putative location of the phos-
app
max
We suggest that the increased V
N29D mutant arises mainly from the direct interac- phate group of DHAP. Strikingly, mutant N32D
tion of DHA with D29. Hence, molecular models of showed no activity towards DHA but it did maintain
the complexes of adduct 7 with RhuA wild type and a significant residual activity towards DHAP. We rea-
N29D show that the terminal hydroxymethylene soned that the location of the N32D mutation deep
group of the adduct might adopt different conforma- inside the catalytic cavity, in close proximity to the es-
tions when binding to each protein (Figure 1, panel sential E117 residue and to the phosphate binding site
C). In the complex with the wild type protein, the 1- (Figure 1, panel A) probably shifts the pKa of the car-
OH group might be accepting hydrogen bonds from boxylate group of D32 towards higher values, favoring
the amide from N32 as well as, intramolecularly, from its protonated neutral form. Under these conditions,
the amide of the Cbz-moiety. With the N29D mutant D32 could act as a hydrogen bond donor with the
92
asc.wiley-vch.de
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 89 – 99

Redesign of the Phosphate Binding Site of l-Rhamnulose-1-Phosphate Aldolase
phosphate group of DHAP not interfering with the much lower extent by sulphate ions (Figure 2). Inter-
function of E117 (Figure 1, panel D). However, in the estingly, all the mutants with negative charges at the
absence of this phosphate group, that is, with DHA, DHAP binding site are much less inhibited by Pi,
the protonated D32 could be hydrogen bonded to which it is also consistent with the Kampp of DHAP ob-
E117 (Figure 1, panel E), forcing it into an unreactive served for the RhuA N29D mutant (Figure 2).
conformation or changing its function essential acid-
base properties.
RhuA N29D was successfully used as catalyst for
the aldol addition of DHA to other selected alde-
The presence of borate buffer (200 mM) improved hydes (Scheme 2, Table 4). The observed unbiased
the vo of the aldol addition of DHA to 1 and 2 cata- stereochemical outcome of the enzymatic aldol reac-
lyzed by RhuA wild type by ~100- and ~35-fold, re- tions indicates full equivalence with RhuA wild type,
spectively (Table 2). However, the effect on the reac- and that both the mutation and the unphosphorylated
tions catalyzed by the mutants was lower or practical- DHA did not perturb the correct substrate recogni-
ly null. Alteration of the phosphate-binding site with tion. It is noteworthy that N-formylglycinal (9d) fur-
aspartate residues appears to be dramatic for the nished, within the limits of high field NMR detection,
DHA-borate complex probably because of electro- only the syn product (see the Supporting Information)
static unfavorable interactions. Electrostatic repul- suggesting a better stereochemical control when small
sions of anions in the active site were substantiated protecting groups are used as compared with bulky
by inhibition studies of inorganic phosphate and sul- masking moieties.[48]
fate ions. The crystallographic structure of RhuA wild
type revealed one phosphate molecule coordinating
the essential Zn(II),[32] which is likely responsible of Conclusions
its inhibition (e.g., <10% RhuA wild type activity at
50 mM Pi) when DHAP is used as donor.[12] The aldol In summary, it was observed that RhuA wild type ac-
addition of DHA catalyzed by RhuA wild type was cepts DHA as donor although with high Kampp values.
strongly inhibited by Pi (IC50 =0.5 mM) and to a Mutations at the phosphate-binding site of the protein
provided a new RhuA catalyst, the RhuA N29D
mutant, with enhanced reactivity towards aldol addi-
tions of DHA to non-natural aldehydes. The installa-
tion of an aspartate residue, D29, near the hydroxy-
methyl group of DHA appears to reduce the activa-
tion energy of the rate-limiting aldol addition step.
Biocatalytic aldol additions of DHA catalyzed by
RhuA, previously restricted to the use of borate
buffer (200 mM), have now been expanded to low-
salt reaction media in practical yields. This new RhuA
N29D mutant will find broad applications in stereose-
lective aldol additions of DHA to aldehydes, since the
number of reported examples with RhuA wild type
can take benefit from using DHA instead of
Figure 2. Relative vo of the aldol addition reaction of DHA
DHAP.[44,49–53] Progress in understanding the binding
to 2 catalyzed by RhuA wild type and mutants in the pres-
mode and reactivity of DHA in RhuA and establish-
ing patterns that help to rationally design new and
more effective mutants is an ongoing work in our
group.
ence of Pi and SO42ꢀ: 10 mM Pi (white bars), 50 mM Pi
2ꢀ
(pale grey bars), 25 mM SO4 (grey bars) and 100 mM
2ꢀ
SO4 (black bars). No product formation was detected in
the presence of 50 mM Pi in reactions catalyzed by RhuA
wild type.
Scheme 2. Aldol addition reactions of DHA to 9a–9d catalyzed by RhuA wild type and RhuA N29D.
Adv. Synth. Catal. 2011, 353, 89 – 99
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
93

Xavier Garrabou et al.
FULL PAPERS
Table 4. Initial velocities (vo), yields and stereochemical outcome of the aldol addition reactions of DHA to 9a–9d catalyzed
by RhuA wild type and RhuA N29D mutant.
Acceptor
vo (mmolhꢀ1 mgꢀ1 protein)
Aldol adduct formed [%] (time, [h])
syn:anti (10:11)[a]
wild type
N29D
wild type
N29D
9a
9b
9c
9d
0.6
0.6
4.2
2.4
1.2
2.4
6.6
2.4
23 (72)
42 (72)
53 (48)
88 (72)
45 (72)
68 (72)
63 (48)
92 (72)
70:30
90:10
90:10
>98:2
[a]
The syn:anti ratios were identical for the wild type and N29D mutant.
Experimental Section
Analytical Methods
Protein concentrations were calculated with the method of
Bradford.[56] Activity assays with the natural substrate l-
rhamnulose-1-phosphate (6) were carried out as described
by Vidal et al.[57]: 1 mL total volume, NADH (0.15 mM),
bis(cyclohexylamine) l-rhamnulose 1-phosphate (2.0 mM),
KCl (100 mM), Tris-HCl, pH 7.5 (50.0 mM), and glycerol
phosphate dehydrogenase (2.5 activity units mLꢀ1). The re-
action was incubated at 258C and the absorbance variation
with time monitored at 340 nm
Materials
Synthetic oligonucleotides were purchased from MWG-Bio-
tech. Acid phosphatase (PA, EC 3.1.3.2, 5.3 Umgꢀ1) was
from Sigma–Aldrich. Glycerol 3-phosphate dehydrogenase
from rabbit muscle (GDH) and NADH were from Sigma–
Aldrich. The plasmid pQE-RhuA containing the gene for
expression of His-Tagged l-rhamnulose-1-phosphate aldo-
lase was a generous gift from Departament d’Enginyeria
Química of the Universitat Autònoma de Barcelona. High
density IDA-Agarose 6BCL Nickel Charged was from His-
panagar. The precursor of dihydroxyacetone phosphate
(DHAP), dihydroxyacetone phosphate dimer bis(ethyl
ketal), was synthesized in our laboratory using a procedure
described by Jung et al.[54] with slight modifications. Deion-
ized water was used for preparative HPLC and Milli-Q-
grade water for analytical HPLC. All other solvents used
were of analytical grade. N-Cbz-amino aldehydes used in
these studies were synthesized in our laboratory using pro-
cedures published in previous works.[55] The dicyclohexyla-
mine salt of l-rhamnulose-1-phosphate (6) was synthesized
using RhuA aldolase as described in reported procedures.[49]
HPLC Analyses
HPLC analyses were performed on an RP-HPLC cartridge,
250”4 mm filled with Lichrosphereꢂ 100, RP-18, 5 mm
(Merck) or on a XBridgeꢂ C18, 5 mm, 4.6”250 mm column
(Waters). The solvent system used was (A): 0.1% (v/v) tri-
fluoroacetic acid (TFA) in H2O and (B): 0.095% (v/v) TFA
in acetonitrile/H2O 4/1, gradient elution from 10% to 70%
B over 30 min, flow rate 1 mLminꢀ1, detection 215 nm,
column temperature 308C. The amount of aldol adduct pro-
duced was quantified from the peak areas using an external
standard methodology.
Mutagenesis
General Methods
The pQE-RhuA plasmid was utilized as template for muta-
genesis. Plasmids containing the RhuA mutants were ob-
tained utilizing the “megaprimer” method.[58] Standard PCR
conditions were used. The external primers utilized were
pQE53 and pQE35. Internal primers are listed in the Sup-
porting Information. The resulting 0.9-kilobase pair PCR
fragments were purified, cleaved with BamHI plus HindIII,
and ligated with pQE-40 plasmid, which had been opened
likewise and purified. Strain JM109 was used for transfor-
mations; resulting clones were checked for their integrity by
restriction analyses and DNA sequencing.
Preparative column chromatography: Merck silica gel 60,
particle size 0.040–0.063 mm (230–240 mesh, flash). Analyti-
cal specific rotation values were measured with a Perkin–
Elmer Model 341 (Überlingen, Germany). NMR analysis:
1
High-field H and 13C nuclear magnetic resonance (NMR)
analyses were carried out using a Bruker AVANCE 500
spectrometer equipped with a high-sensitive CryoProbe for
[D2]H2O and MeOH-d4 solutions. Full characterization of
the described compounds was performed using typical gradi-
ent-enhanced 2D experiments: COSY, NOESY, HSQC and
HMBC, recorded under routine conditions. When possible,
NOE data were obtained from selective 1D NOESY ver-
sions using a single pulsed-field-gradient echo as a selective
excitation method and a mixing time of 500 ms. When nec-
essary, proton and NOESY experiments were recorded at
different temperatures in order to study the different behav-
ior of the exchange phenomena to avoid the presence of
false NOE cross-peaks that complicates both structural and
dynamic studies. Routine, 1H (400–500 MHz) and 13C
(101 MHz) NMR spectra of compounds were recorded with
Varian Mercury-400 and Varian Anova-500 spectrometers,
respectively.
Protein Expression and Purification
The plasmids were transformed into E. coli strain M-15
[pREP-4] (QIAGEN). Cells were grown at 378C in 5 L of
an LB medium containing ampicillin (100 mgLꢀ1) and kana-
mycin (25 mgLꢀ1) up to an optical density of 0.6 at
600 nm.[57] For protein expression, the temperature was low-
ered to 308C to avoid inclusion bodies formation and iso-
propyl-b-d-thiogalactoside (IPTG) was added to a final con-
centration of 50 mM. After additional 4 h cells were harvest-
ed, suspended in starting buffer (50 mM disodium hydrogen
94
asc.wiley-vch.de
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 89 – 99

Redesign of the Phosphate Binding Site of l-Rhamnulose-1-Phosphate Aldolase
phosphate, 300 mM NaCl, 20 mM imidazole, pH 8.0) and
lysed using a Cell Disrupter (Constant Systems). Cellular
debris was removed by centrifugation at 12,000 g for 10 min.
The clear supernatant was collected and purified by affinity
chromatography in an FPLC system (Amersham biosci-
ences). The crude supernatant was applied to a cooled HR
16/40 column (GE Healthcare) containing affinity bead
(50 mL) and was washed with start buffer (150 mL). The
protein was eluted with an aqueous buffered solution pH 8.0
containing disodium hydrogen phosphate (50 mM), NaCl
(300 mM) and imidazole (300 mM), at a flow rate of
3 mLminꢀ1. CoCl2 (up to 1 mM) was added to the eluted
protein and incubated for 15 min. Addition of (NH4)2SO4
(0.4 g per mL of liquid) caused protein precipitation. The re-
sulting pellet was centrifuged at 12,000 g for 10 min, sus-
pended in (NH4)2SO4 (50 mL, 0.4 gmLꢀ1) and centrifuged
again. The pellet was finally suspended in (NH4)2SO4
(50 mL, 0.4 gmLꢀ1) and stored at 48C. Electrospray ioniza-
tion mass spectrometry of proteins is described in the Sup-
porting Information.
catalyst was adjusted (0.032–0.19 mg of protein) to ensure
linear dependence of product concentration vs. time at reac-
tion conversions <20%. At 5, 10, 20, 30 and 45 min samples
were withdrawn and analyzed by HPLC as described above.
The amount of aldol adduct produced was quantified by an
external standard method. Linear correlations were found
for conversion lower than 15%. The estimated standard
error for three determinations was between 10–12%.
Enzymatic Aldol Reactions with DHA or DHA in
Borate Buffer: Determination of Reaction
Conversion to Aldol Adduct
Reactions were conducted on an analytical scale (see
above). Aldehydes 2, 9a, 9b or 9c (19.5 mmol) were dissolved
in dimethylformamide (60 mL) and aldehydes 1 or 12
(12.0 mmol) in water (60 mL). These solutions were mixed
with a DHA solution (150 mL, 30.0 mmol) and buffer (60 mL,
triethanolamine-HCl 250 mM for DHA as donor or sodium
borate 1000 mM for DHA-borate ester as donor; pH 7.0).
The aldolase (30 mL, 0.19 mg) was then added to start the
aldol reaction. Productivities were measured at 24 h reaction
time. Reaction monitoring with aldehydes 2, 9a, 9b and 9c
was as follows: samples (12–25 mL) were withdrawn, diluted
with methanol (250–1000 mL) and analyzed by HPLC under
the conditions described above. Reaction monitoring with
aldehydes 1 and 9d was as follows: samples (10 mL) were
Enzymatic Aldol Reactions with DHAP as Donor:
Determination of Reaction Conversion to Aldol
Adduct
Analytical scale reactions (300 mL total volume) were con-
ducted in test tubes (2 mL) stirred with a vortex mixer
(VIBRAX VXR basic, Ika) at 1000 rpm and 258C. Alde-
hydes 2, 9a, 9b or 9c (19.5 mmol) were dissolved in dimethyl-
formamide (60 mL) and aldehydes 1 or 9d (12.0 mmol) in
water (60 mL). Each of the aldehydes was mixed with freshly
neutralized (pH 6.9) DHAP solution (150 mL, 15.0 mmol),
and the aldolase (90 mL, 0.19 mg protein, amount corre-
sponding to 2 U·mLꢀ1 reaction for RhuA wild type) was
added to start the aldol reaction. Productivities were mea-
sured at 4 h of reaction time. Reaction monitoring with alde-
hydes 2, 9a, 9b or 9c was as follows: samples (12–25 mL)
were withdrawn, diluted with methanol (250–1000 mL) and
analyzed by HPLC under the conditions described above.
Reaction monitoring with aldehydes 1 and 9d was as fol-
lows: samples (10 mL) were mixed with a solution (10 mL) of
acid phosphatase (5.3 U·mLꢀ1 in sodium citrate buffer
400 mM pH 4.5) and incubated for 24 h. The resulting crude
was mixed with a solution of O-benzylhydroxylamine hydro-
chloride (30 mL, 21.2 mgmLꢀ1; 0.14 mmolmLꢀ1) in pyridine:
methanol:water 33:15:2. After incubation at 508C for
60 min, samples were diluted in methanol (950 mL) and di-
rectly analyzed and quantified by HPLC using an external
standard method. Conversions were calculated with respect
to the initial concentration of DHAP, which was the limiting
reactant.
mixed with a solution of O-benzylhydroxylACTHNUGRTENUNGamine hydrochlo-
ride (30 mL, 21.2 mgmLꢀ1; 0.14 mmolmLꢀ1) in pyridine:me-
thanol:water 33:15:2. After incubation at 508C for 60 min,
samples were diluted in methanol (960 mL) and directly ana-
lyzed by HPLC. Samples were analyzed, and quantified by
HPLC as described above. Conversions were calculated re-
spect to the initial concentration of aldehyde acceptor,
which was the limiting reactant.
Enzymatic Aldol Reactions with DHA or DHA in
Borate Buffer as Donor: Initial Velocity
The initial aldol velocities (vo) were determined by measur-
ing the amount of aldol adduct produced by HPLC at the
initial reaction times in a similar manner as used for reac-
tions using DHAP. Reactions were conducted on an analyti-
cal scale (see above). The amounts of aldehydes, dimethyl-
formamide and DHA used for these reactions were identical
to those for the analysis of aldol adduct formed. The aldo-
lase (30 mL, 0.09–0.19 mg) was then added to start the aldol
reaction, being the amount of catalyst adjusted to ensure
linear dependence of product concentration vs. time at reac-
tion conversions <20%. At different times, 20, 40, 80 and
120 min and 10, 20, 40, 60 and 120 min for reactions with
borate buffer, samples were withdrawn, analyzed, and quan-
tified by HPLC as described above.
Enzymatic Aldol Reactions with DHAP as Donor:
Initial Aldol Velocities
Steady-State Kinetic Parameters for DHAP and
DHA
The initial aldol velocities (vo) were determined by measur-
ing the amount of aldol adduct produced by HPLC at the
initial reaction times. Reactions were conducted as de-
scribed above for the analytical scale reactions. The amounts
of aldehydes, dimethylformamide and freshly neutralized
DHAP solution used for these reactions were identical to
those for the analytical reactions. In this case, the amount of
app
max
To determine the Kampp and V
values of DHAP and DHA,
the aldol additions of DHAP and DHA to 2 catalyzed by
both RhuA wild type and RhuA N29D were run on an ana-
lytical scale (see above).
Steady-state kinetic parameters for DHAP: Aldehyde 2
(19.5 mmol) was dissolved in dimethylformamide (60 mL)
Adv. Synth. Catal. 2011, 353, 89 – 99
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
95

Xavier Garrabou et al.
FULL PAPERS
and mixed with a solution (120 mL) containing different
amounts of DHAP (0.03, 0.06, 0.10, 0.15, 0.30, 0.60, 0.90,
1.5, 3.0, 6.0 mmol) freshly neutralized at pH 6.9 and buffer
(30 mL, triethanolamine-HCl 500 mM, NaCl 500 mM,
pH 7.0). To this mixture, the aldolase, RhuA wild type or
RhuA N29D, (90 mL, 0.2–1.60 mg of protein) was added to
start the reaction. Samples (10 mL) were withdrawn at differ-
ent times (120, 165, 210, 255 and 300 seconds) and the reac-
tion stopped by addition of methanol (90 mL). The amount
of free DHAP in the samples was immediately measured
spectrophotometrically utilizing a coupled assay with glycer-
ol-3-phosphate dehydrogenase.
Steady-state kinetic parameters for DHA: Aldehyde 2
(19.5 mmol) was dissolved in dimethylformamide (60 mL)
and mixed with a solution (120 mL) containing different
amounts of DHA (30, 60, 120, 180, 240, 300, 450, 630,
720 mmol) and buffer (30 mL, triethanolamine-HCl 500 mM,
NaCl 500 mM, pH 7.0). To this mixture, the aldolase, RhuA
wild type or RhuA N29D, (90 mL, 0.19 mg protein) was
added to start the reaction (total final volume, 300 mL). For
each DHA concentration, samples (25 mL) were withdrawn
at different times (5, 10, 20 and 30 min), diluted with metha-
nol (250 mL), centrifuged and analyzed by HPLC as de-
scribed for the determination of initial velocities. The exper-
imental and fitted data of the activity vs. DHAP and DHA
concentration, shown in Figure S1 and Figure S2, respective-
ly, in the Supporting Information, was adjusted by comput-
er-based nonlinear regression of untransformed data direct
to the kinetic Michaelis–Menten equation.
a rotary evaporator and the suspension was filtered through
a 0.45 mm cellulose membrane filter. The filtrate was adjust-
ed to pH 3.0 and loaded onto a semi-preparative column X-
Terraꢂ (19”250 mm). Products from aldehydes 2, 9a–9c
were eluted using a CH3CN gradient from 0% to 52%
CH3CN in 30 min. The flow rate was 10 mLminꢀ1 and the
products were detected at 215 nm. Analysis of the fractions
was accomplished under gradient conditions (10 to 70% of
solvent B in 30 min) on the analytical HPLC. Pure fractions
were pooled and lyophilized obtaining the corresponding
products 7 (226 mg, 0.76 mmol, 59% isolated yield), 10a+
11a (101 mg, 0.34 mmol, 26% isolated yield), 10b+11b
(159 mg, 0.57 mmol, 44% isolated yield), and 10c+11c
(147 mg, 0.61 mmol, 47% isolated yield). The aldol adducts
7, 10a+11a and 10b+11b were converted into the corre-
sponding iminocyclitols. To this end, aldol adducts were dis-
solved in H2O/MeOH 1:1 and treated with H2 (50 psi) in the
presence of Pd/C (50 mg) at room temperature during 24 h.
After removal of the catalyst by filtration through a 0.45 mm
nylon membrane filter, the filtrates were adjusted to pH 6.4
with formic acid and lyophilized obtaining a solid material.
1
The H, 13C NMR and optical rotations and diastereomeric
distribution values matched those previously described.[37]
Aldol adducts 10c+11c were characterized directly as de-
scribed in a previous study.[59]
Scale-Up of the Addition of DHA to Aldehyde 9d
A similar standard procedure to that described above was
followed to synthesize the aldol adduct obtained by the
aldol addition of DHA to aldehyde 9d (1.0 mmol) catalyzed
by RhuA N29D. Purification on silica gel eluted with
CHCl3:AcOEt (1:1)-MeOH from 1:0 to 7:3 yielded 92 mg
(65% isolated yield) of 10d+11d, which were characterized
directly, without converting them into iminocyclitols. The
1H, 13C NMR spectra and assigments are depicted in Sup-
porting Information Figure S16.
Inhibition by Inorganic Phosphate and Sulphate Ions
Reactions were conducted on an analytical scale (see
above). Aldehyde 2 (19.5 mmol) was dissolved in dimethyl-
formamide (60 mL). This solution was mixed with a DHA
solution (100 mL, 30.0 mmol), buffer (60 mL, triethanolamine-
HCl 250 mM pH 7.0) and a variable amount of sodium
phosphate solution (50 mL, pH 7.0; 0, 7.5 or 30 mmol for de-
termination of initial velocities in Figure 2; 0.03, 0.15, 0.3,
0.75, 1.5 mmol for IC50 determination with RhuA wild type)
or sodium sulphate solution (50 mL, pH 7.0; 0, 3.0 or 15
mmol). The corresponding aldolase (30 mL, 0.09–0.19 mg)
was then added to start the aldol reaction. Samples (25 mL)
were withdrawn at different times (20, 40, 80 and 120 min
for determination of initial velocities (Figure 2); 15, 30, 45
and 60 min for IC50 determination with RhuA wild type) di-
luted with methanol (250–500 mL) and analyzed by HPLC
under the conditions described above. The graphical repre-
sentation of the activity vs. the logarithm of phosphate con-
centration gave the IC50 value of phosphate ion for RhuA
wild type. The experimental and fitted curve is shown in Fig-
ure S3 of the Supporting Information.
Computational Methods
Protein complexes were modeled with the package Schrç-
dinger Suite 2009 (Schrçdinger, LLC, New York) through
its graphical interphase Maestro (Maestro, version 9.0,
Schrçdinger, LLC, New York, NY, 2009). The program Mac-
roModel (MacroModel, version 9.7, Schrçdinger, LLC, New
York, NY, 2009) with its default force field OPLS 2005, a
modified version of the OPLS-AA force field,[60] and GB/
SA water solvation conditions[61] were used for all energy
calculations.
Coordinates of E. coli l-rhamnulose-1-phosphate aldo-
lase[32] were obtained from the Protein Data Bank[62] at
Brookhaven National Laboratory (PDB code 1OJR). By ap-
plying the necessary crystallographic symmetry operators to
this structure, the homotetramer that constitutes the biologi-
cal unit was built. Since the active center of RhuA is located
at the interface between each pair of contiguous monomers
each homotetramer contains four catalytic centers. Howev-
er, in order to reduce the computational time, all modeling
was performed on a RhuA dimer containing just one cata-
lytic site. Further modification of the RhuA wild type struc-
ture was required to adjust the conformation of residues
E27, R28 and E171ꢁ to their putative arrangement at physio-
logical pH.[32] The structures of RhuA N29D and N32D mu-
Scale-Up of the Addition of DHA to Aldehydes 2,
9a–9c; General Procedure
Reactions with RhuA N29D were scaled up to 20 mL total
volume following the composition of the corresponding ana-
lytical reactions (1.3 mmol aldehydes 2, 9a–9c). When the
product concentration was constant by HPLC (~72 h, con-
versions similar to those described for analytical reactions),
MeOH (20 mL) was added to stop the reaction and precipi-
tate the enzyme. The excess of MeOH was then removed in
96
asc.wiley-vch.de
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 89 – 99

Redesign of the Phosphate Binding Site of l-Rhamnulose-1-Phosphate Aldolase
tants were generated by mutating in silico the wild type pro-
tein. Protein structures were prepared using the Protein
Preparation Wizard included in Maestro to remove solvent
molecules and ions, adding hydrogens, setting protonation
states and minimizing the energy using the OPLS force
field. The structures of the different adducts bound into the
active centre of RhuA enzymes were modeled starting from
the coordinates of a dihydroxyacetone and a phosphate mol-
ecule, present in the 1OJR crystal structure and bound into
the active center of RhuA wild type, by manually adding the
moiety derived from the corresponding aldehyde substrate.
The structures were then minimized, first applying con-
straints to the protein (force constant=100 kcalŠꢀ2 molꢀ1)
to avoid large changes on its structure, and afterwards allow-
ing free movement of the whole system until reaching a gra-
dient <0.01 kcalmolꢀ1 Šꢀ1. Molecular dynamics simulations
were carried out for the complexes of 6 bound into the
active center of RhuA wild type and RhuA N32D, and for
l-rhamnulose (5) bound into the active center of RhuA
N32D. The program Desmond 2.2.9.1(Desmond Molecular
Dynamics System, version 2.2, D. E. Shaw Research, New
York, NY, 2009)[63] through the Maestro interphase (Maes-
tro-Desmond Interoperability Tools, version 2.2, Schrçding-
er, New York, NY, 2009) was used for that purpose. Parame-
ters from the OPLS 2005 force field were used for proteins
and ligands. Protein-adduct systems were prepared by im-
state. The H-bond remains strong in both structures, with no
significant changes between them: from 1.56 Š in reactants
to 1.59 Š in the transition state. The model without acetate
(see Supporting Information, ꢀF1igure S11), has a calculated
ꢀ
energy barrier of 8.2 kcalmol . In this case, the C C dis-
tance in the transition state is shorter, 2.05 Š. The shorter
distance is in agreement with a higher barrier, considering
Hammondꢁs postulate.
Acknowledgements
This work was supported by the Spanish MICINN
CTQ2009-07359 and CTQ2009-08328, Generalitat de Catalu-
nya (2009 SGR 00281), and ESF COST CM0701 projects. X.
Garrabou acknowledges the CSIC for a I3P predoctoral
scholarship.
References
[1] M. Markert, R. Mahrwald, Chem. Eur. J. 2008, 14, 40–
48.
[2] D. Enders, A. A. Narine, J. Org. Chem. 2008, 73, 7857–
7870.
[3] D. Enders, M. Voith, A. Lenzen, Angew. Chem. 2005,
117, 1330–1351; Angew. Chem. Int. Ed. 2005, 44, 1304–
1325.
[4] D. A. Evans, J. V. Nelson, T. R. Taber, Top. Stereochem.
1982, 13, 1–115.
3
mersing them in a ~110”90”90 Š box that contained
~29000 SPC water molecules and 10 (5 complexes) or 12 (6
complexes) sodium ions to achieve a zero net charge. The
full systems (~95,000 atoms each) were minimized, first with
the solute restrained and then without restrains, and equili-
brated stepwise at 300 K for 500 ps of MD. Simulations
(1.2 ns, 2 fs time step) were performed under periodic boun-
dary conditions in the NPT ensemble at 300 K and 1 bar,
using the Berendsen coupling scheme. Bond lengths to hy-
drogen atoms were constrained using the Shake algo-
rithm.[64] A cut-off of 9 Š was applied to van der Waals and
short-range electrostatic interactions, while long-range elec-
trostatic interactions were computed using the Particle
Mesh Ewald method (see Figures S6, S7 and S8 in the Sup-
porting Information).[65]
[5] R. Mahrwald, Modern Aldol Reactions, Vol. 1, Eno-
lates, Organocatalysis, Biocatalysis and Natural Product
Synthesis, Wiley-VCH, Weinheim, 2004.
[6] R. Mahrwald, Modern Aldol Reactions, Vol. 2, Metal
Catalysis, Wiley-VCH, Weinheim, 2004.
[7] B. Schetter, R. Mahrwald, Angew. Chem. 2006, 118,
7668–7687; Angew. Chem. Int. Ed. 2006, 45, 7506–
7525.
[8] S. Mukherjee, J. W. Yang, S. Hoffmann, B. List, Chem.
Rev. 2007, 107, 5471–5569.
[9] W.-D. Fessner, V. Helaine, Curr. Opin. Biotechnol.
2001, 12, 574–586.
[10] S. M. Dean, W. A. Greenberg, C.-H. Wong, Adv. Synth.
Density Functional Theory (DFT) Calculations
To gain insight on the effects that the N29D mutation can
exert on the enzyme kinetics, we generated the simplest pos-
sible model of the active site studying the aldol addition of
dihydroxyacetone (DHA) to formaldehyde. The model gen-
erated (see Supporting Information, Figures S9 and S10),
contains an acetate group to mimic the side chain of D29. A
second model was build, removing the acetate group to
assess the influence of the H-bond between the carboxylate
and the donor dihydroxyacetone (DHA). Both models were
optimized at the B3LYP/6–31+G(d) level. Calculations
were performed with the Gaussian 03 package.[66] For both
models, with and without the acetate, we have optimized the
reactants and the transition state geometries. The reactants
correspond to the aldehyde and DHA coordinated to
Catal. 2007, 349, 1308–1320.
[11] C.-H. Wong, R. L. Halcomb, Y. Ichikawa, T. Kajimoto,
Angew. Chem. 1995, 107, 453–474; Angew. Chem. Int.
Ed. Engl. 1995, 34, 412–432.
[12] W.-D. Fessner, C. Walter, Top. Curr. Chem. 1997, 184,
97–194.
[13] M. Schumperli, R. Pellaux, S. Panke, Appl. Microbiol.
Biotechnol. 2007, 75, 33–45.
[14] L. Iturrate, I. Sanchez-Moreno, E. G. Doyaguez, E.
Garcia-Junceda, Chem. Commun. 2009, 1721–1723.
[15] W.-D. Fessner, in: Asymmetric Organic Synthesis with
Enzymes, (Eds.: V. Gotor, I. Alfonso, E. Garcia-Ur-
diales), Wiley-VCH Verlag GmbH & Co. KGaA,
Weinheim, 2008, pp 275–318.
[16] P. ClapØs, W.-D. Fessner, G. A. Sprenger, A. K. Sam-
land, Curr. Opin. Chem. Biol. 2010, 14, 154–167.
[17] S. S. V. Ramasastry, K. Albertshofer, N. Utsumi, C. F.
Barbas, Org. Lett. 2008, 10, 1621–1624.
ꢀ
Co(II), and the transition state has the new C C bond being
formed. The model with the acetate (see Supporting Infor-
mation, Figure S10) has a calculated energy barrier of
ꢀ1
ꢀ
5.1 kcalmol . The C C distance is 2.19 Š at the transition
Adv. Synth. Catal. 2011, 353, 89 – 99
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
97

Xavier Garrabou et al.
FULL PAPERS
[18] S. S. V. Ramasastry, K. Albertshofer, N. Utsumi, F.
Tanaka, C. F. Barbas III, Angew. Chem. 2007, 119,
5668–5671; Angew. Chem. Int. Ed. 2007, 46, 5572–
5575.
[19] M. Markert, M. Mulzer, B. Schetter, R. Mahrwald, J.
Am. Chem. Soc. 2007, 129, 7258–7259.
[20] M. Schürmann, G. A. Sprenger, J. Biol. Chem. 2001,
276, 11055–11061.
[21] M. Schürmann, M. Schürmann, G. A. Sprenger, J. Mol.
Catal. B: Enzym. 2002, 19, 247–252.
[22] S. Schneider, T. Sandalova, G. Schneider, G. A. Spreng-
er, A. K. Samland, J. Biol. Chem. 2008, 283, 30064–
30072.
[23] J. A. Castillo, J. Calveras, J. Casas, M. Mitjans, M. P. Vi-
nardell, T. Parella, T. Inoue, G. A. Sprenger, J. Joglar,
P. ClapØs, Org. Lett. 2006, 8, 6067–6070.
[24] 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.
[25] A. L. Concia, C. Lozano, J. A. Castillo, T. Parella, J.
Joglar, P. ClapØs, Chem. Eur. J. 2009, 15, 3808–3816.
[26] X. Garrabou, J. A. Castillo, C. GuØrard-HØlaine, T. Par-
ella, J. Joglar, M. Lemaire, P. ClapØs, Angew. Chem.
2009, 121, 5629–5633; Angew. Chem. Int. Ed. 2009, 48,
5521–5525.
[40] Aldol addition of DHA to 2 catalyzed by l-fuculose-1-
phosphate aldolase failed, indicating that DHA accept-
ance appears to be an exclusive property of RhuA.
[41] An accurate treatment of the kinetic data would re-
quire the use of the Michaelis–Menten model for two-
substrate reactions. Under these conditions the deter-
app
max
mined V
and Kampp, depended on the concentration of
the acceptor aldehyde, which is fixed, (see H. Biss-
wanger, Enzyme Kinetics: Principles and Methods,
Wiley-VCH Verlag GmbH, Weinheim, 2002, pp 108–
119) thus allowing comparison of the kinetic properties
among the different enzymes, as stated in the text.
[42] Using DHAP as donor, it was suggested that its depro-
tonation was the rate-limiting step when 1 was the ac-
ceptor (see ref.[32]). When other acceptors are used, dif-
ferent reaction rates were obtained, suggesting that the
rate-limiting step would be the aldol addition.
[43] 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.
[44] W.-D. Fessner, G. Sinerius, A. Schneider, M. Dreyer,
G. E. Schulz, J. Badia, J. Aguilar, Angew. Chem. 1991,
103, 596–599; Angew. Chem. Int. Ed. Engl. 1991, 30,
555–558.
[27] S. Schneider, M. GutiØrrez, T. Sandalova, G. Schneider,
P. ClapØs, G. A. Sprenger, A. K. Samland, ChemBio-
Chem 2010, 11, 681–690.
[28] D. G. Drueckhammer, J. R. Durrwachter, R. L. Peder-
son, D. C. Crans, L. Daniels, C.-H. Wong, J. Org. Chem.
1989, 54, 70–77.
[45] W.-D. Fessner, J. Badia, O. Eyrisch, A. Schneider, G.
Sinerius, Tetrahedron Lett. 1992, 33, 5231–5234.
[46] X. Garrabou, L. Gomez, J. Joglar, S. Gil, T. Parella, J.
Bujons, P. ClapØs, Chem. Eur. J. 2010, 16, 10691–10706.
[47] T. Ishida, J. Am. Chem. Soc. 2010, 132, 7104–7118.
[48] J. Calveras, J. Bujons, T. Parella, R. Crehuet, L. Espelt,
J. Joglar, P. ClapØs, Tetrahedron 2006, 62, 2648–2656.
[49] W.-D. Fessner, A. Schneider, O. Eyrisch, G. Sinerius, J.
Badia, Tetrahedron: Asymmetry 1993, 4, 1183–1192.
[50] W.-C. Chou, L. Chen, J.-M. Fang, C.-H. Wong, J. Am.
Chem. Soc. 1994, 116, 6191–6194.
[29] R. Schoevaart, F. van Rantwijk, R. A. Sheldon, J. Org.
Chem. 2001, 66, 4559–4562.
[30] M. Sugiyama, Z. Y. Hong, L. J. Whalen, W. A. Green-
berg, C.-H. Wong, Adv. Synth. Catal. 2006, 348, 2555–
2559.
[31] M. Sugiyama, Z. Hong, W. A. Greenberg, C.-H. Wong,
Bioorg. Med. Chem. 2007, 15, 5905–5911.
[51] I. Henderson, K. B. Sharpless, C.-H. Wong, J. Am.
Chem. Soc. 1994, 116, 558–561.
[32] M. Kroemer, I. Merkel, G. E. Schulz, Biochemistry
[52] K. K. C. Liu, T. Kajimoto, L. Chen, Z. Zhong, Y. Ichi-
kawa, C.-H. Wong, J. Org. Chem. 1991, 56, 6280–6289.
[53] O. Eyrisch, W.-D. Fessner, Angew. Chem. 1995, 107,
1738–1740; Angew. Chem. Int. Ed. Engl. 1995, 34,
1639–1641.
2003, 42, 10560–10568.
[33] A. C. Joerger, C. Mueller-Dieckmann, G. E. Schulz, J.
Mol. Biol. 2000, 303, 531–543.
[34] G. DeSantis, J. J. Liu, D. P. Clark, A. Heine, I. A.
Wilson, C.-H. Wong, Bioorg. Med. Chem. 2003, 11, 43–
52.
[54] S.-H. Jung, J.-H. Jeong, P. Miller, C.-H. Wong, J. Org.
Chem. 1994, 59, 7182–7184.
[35] M. Cheriyan, E. J. Toone, C. A. Fierke, Protein Sci.
2007, 16, 2368–2377.
[55] J. Calveras, M. Egido-Gabµs, L. Gómez, J. Casas, T.
Parella, J. Joglar, J. Bujons, P. ClapØs, Chem. Eur. J.
2009, 15, 7310–7328.
[56] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[57] L. Vidal, O. Durany, T. Suau, P. Ferrer, M. D. Benaiges,
G. Caminal, J. Chem. Technol. Biotechnol. 2003, 78,
1171–1179.
[58] G. Sarkar, S. S. Sommer, BioTechniques 1990, 8, 404–
407.
[59] L. Espelt, J. Bujons, T. Parella, J. Calveras, J. Joglar, A.
Delgado, P. ClapØs, Chem. Eur. J. 2005, 11, 1392–1401.
[60] W. L. Jorgensen, D. S. Maxwell, J. Tirado-Rives, J. Am.
Chem. Soc. 1996, 118, 11225–11236.
[36] M. Hixon, G. Sinerius, A. Schneider, C. Walter, W.-D.
Fessner, J. V. Schloss, FEBS Lett. 1996, 392, 281–284.
[37] L. Espelt, T. Parella, J. Bujons, C. Solans, J. Joglar, A.
Delgado, P. ClapØs, Chem. Eur. J. 2003, 9, 4887–4899.
[38] This result was observed with either RhuA[Zn(II)] or
RhuA[Co(II)].
[39] l-Ribulose-5-phosphate 4-epimerase (AraD, EC
5.1.3.4) from E. coli, a class II enzyme closely related
to FucA, and variants of this enzyme displayed a back-
ground activity on the aldol addition reaction of DHA
to glycolaldehyde; see J. Samuel, Y. Luo, P. M. Morgan,
N. C. J. Strynadka, M. E. Tanner, Biochemistry 2001, 40,
14772–14780.
[61] W. C. Still, A. Tempczyk, R. C. Hawley, T. Hendrick-
son, J. Am. Chem. Soc. 1990, 112, 6127–6129.
98
asc.wiley-vch.de
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 89 – 99

Redesign of the Phosphate Binding Site of l-Rhamnulose-1-Phosphate Aldolase
[62] H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland,
T. N. Bhat, H. Weissig, I. N. Shindyalov, P. E. Bourne,
Nucleic Acids Res. 2000, 28, 235–242.
[63] K. J. Bowers, E. Chow, H. Xu, R. O. Dror, M. P. East-
wood, B. A. Gregersen, J. L. Klepeis, I. Kolossvµry,
M. A. Moraes, F. D. Sacerdoti, J. K. Salmon, Y. Shan,
D. E. Shaw, Proceedings of the ACM/IEEE Conference
on Supercomputing (SC06), Tampa, Florida, 2006. Free
pdf of the paper: http://sc06.supercomputing.org/sched-
ule/pdf/pap259.pdf.
Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Na-
katsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gom-
perts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Mo-
rokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Ra-
ghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G.
Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G.
Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A.
Pople, Gaussian 03, Revision E.01, Gaussian, Inc.,
Wallingford, CT, 2004.
[64] V. Krautler, W. F. Van Gunsteren, P. H. Hunenberger,
J. Comput. Chem. 2001, 22, 501–508.
[65] T. Darden, D. York, L. Pedersen, J. Chem. Phys. 1993,
98, 10089–10092.
[66] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuse-
ria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery
Jr, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam,
S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M.
Adv. Synth. Catal. 2011, 353, 89 – 99
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
99