Encapsulation of a catalytic imidazolium salt into avidin: Towards the development of a biohybrid catalyst active in ionic liquids

Article abstract of DOI:10.1002/chem.201303865

Herein, we report the development of biohybrid catalysts that are capable of catalyzing the aldol reaction. The use of biotinylated imidazolium salts in combination with racemic or enantiomerically pure catalytic anions allowed us to study the adaptive and cooperative positioning of the anionic catalyst inside the protein. Supramolecular encapsulation of the biotinylated catalyst into avidin resulted in good selectivity for the aldol reaction performed in ionic liquid/water mixtures. Biohybrid catalysts capable of catalyzing the aldol reaction are prepared from avidin and biotinylated imidazolium salts with either racemic or enantiomerically pure catalytic anions. Supramolecular encapsulation (see figure) of the biotinylated catalyst in avidin resulted in good selectivities for the aldol reaction when performed in ionic liquid/water mixtures and the adaptive and cooperative positioning of the anionic catalyst inside the avidin protein is discussed. Copyright

Full text of DOI:10.1002/chem.201303865

DOI: 10.1002/chem.201303865
Full Paper
&
Supramolecular Catalysis
Encapsulation of a Catalytic Imidazolium Salt into Avidin: Towards
the Development of a Biohybrid Catalyst Active in Ionic Liquids
Vincent Gauchot,[a] Mathieu Branca,[b] and Andreea Schmitzer*[a]
Chem. Eur. J. 2014, 20, 1530 – 1538
1530
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Abstract: Herein, we report the development of biohybrid
catalysts that are capable of catalyzing the aldol reaction.
The use of biotinylated imidazolium salts in combination
with racemic or enantiomerically pure catalytic anions al-
lowed us to study the adaptive and cooperative positioning
of the anionic catalyst inside the protein. Supramolecular en-
capsulation of the biotinylated catalyst into avidin resulted
in good selectivity for the aldol reaction performed in ionic
liquid/water mixtures.
Introduction
cally modified, even if this requires some time through long
methods in order to optimize the environment around the cat-
alyst. Imidazolium salts, best known as ionic liquids (ILs), have
gained major interest in the world of organic synthesis as
promising “green solvents” as they display many interesting
characteristics in terms of supramolecular architecture, non-
toxicity, atom economy, and protein stabilization.[11–13] More re-
cently, ILs have been used in asymmetric catalysis and biocatal-
ysis, either as solvents or as actual catalysts. Their potential as
solvents for the aldol reaction has been widely reported in the
literature, and their superior efficiency as catalysts was recently
highlighted.[14] Whereas catalysis using the cations of ILs has
been widely reported in the literature, use of the anions of or-
ganic salts as catalysts is still a developing topic.[15] We previ-
ously reported the use of an imidazolium salt bearing a chiral
catalytic anion that can be the source of stereoinduction in the
aldol and Michael reactions.[16] Moreover, having demonstrated
the beneficial effect of the second coordination sphere provid-
ed by the presence of a cyclodextrin unit in a supramolecular
complex,[17] we were interested in determining the influence of
a host protein on the activity and stereoselectivity of a hybrid
system composed of an imidazolium-based biotinylated
anchor and a nonchiral organocatalytic anion. The biotinylated
imidazolium cation therefore plays an important role, not only
in modulating the steric and electronic properties of the orga-
nocatalytic anion (first-coordination-sphere interactions), but
also in its position inside the protein and therefore in defining
the second coordination sphere.
Enzymes are evolved entities with supramolecular structures
possessing highly catalytic functions, which are usually accom-
panied by a variety of conformational states. It has been clearly
demonstrated that the motion of the enzyme structure confers
catalytic efficiency during catalysis.[1] Recent progress in host–
guest chemistry has allowed chemists to use noncovalent an-
choring strategies to build supramolecular complexes with
proteins that behave as biohybrid catalysts. These supramolec-
ular complexes show intricate and hierarchical architectures, as
well as dynamic features, all of which are required parameters
for the development of catalytic systems.
Over the last decade, the development of artificial hybrid
biocatalysts inspired by natural ones has been of great interest
in the field of stereoselective synthesis. Ranging from synthetic
macrocyclic compounds to self-assembled nanometer-sized
objects, such complexes have been exploited as scaffolds to
design supramolecular biohybrid systems.[2] The supramolec-
ular anchoring strategy relies on noncovalent interactions be-
tween small molecules and the biomolecular scaffold. The cru-
cial point of this strategy is the affinity of the guest molecule
for the host biomolecular scaffold. For example, Harada et al.
used the high affinity of antibodies for the creation of an artifi-
cial hydrogenase.[3] In the same spirit, Keinan et al. presented
an antibody–metalloporphyrin assembly that catalyzed enan-
tioselective oxidations.[4] In an early report, Whitesides et al. de-
scribed the creation of an artificial metalloenzyme based on
the very high affinity of biotin for avidin and streptavidin.[5]
Since 2003, the Ward research group has intensively explored
the biotin–(strept)avidin technology for the creation of artificial
metalloenzymes, ranging from those mimicking natural en-
zymes to the design of efficient unnatural metalloenzymes.[6–10]
The presence of the biomolecular scaffold offers an additional
advantage for the optimization of the artificial metalloenzyme.
Whereas chemical optimization can be achieved by modifying
the ligand or by introducing a spacer between the biotin
anchor and the metal, the biomolecular scaffold can be geneti-
The use of biotinylated imidazolium salts in combination
with racemic or enantiomerically pure anions allowed us to
study the adaptive and cooperative positioning of the anionic
catalyst inside the protein (Scheme 1). We report the prepara-
tion of a new type of biohybrid catalysts active in IL/H2O mix-
tures and mechanistic insights into its organocatalysis by using
the aldol reaction as an example. To the best to our knowl-
edge, this is the first example of a biohybrid catalyst that is
able to function in an ionic liquid medium.
Results and Discussion
[a] V. Gauchot, Prof. Dr. A. Schmitzer
Avidin is a glycosylated protein that is naturally present in egg
white.[18] Avidin is a tetrameric eight stranded b-barrel protein
that binds up to four biotins with high affinity. The interaction
between biotin and avidin is extremely tight with Ka =1.7ꢁ
1015 mÀ1. This high value ensures a quasi-irreversible anchoring
of biotinylated compounds in the protein pocket as a result of
numerous interactions between the biotin and the protein,
such as hydrophobic interactions,[19] Van der Waals interactions,
Departement de Chimie, Universitꢀ de Montrꢀal
C. P. 6128 Succursale Centre-Ville, Montrꢀal, Quꢀbec H3C 3 J7 (Canada)
E-mail: ar.schmitzer@umontreal.ca
[b] Dr. M. Branca
Present address: Laboratoire d’Electrochimie Molꢀculaire
UMR 7591 CNRS, Universitꢀ Paris Diderot, Sorbonne Paris Citꢀ
15 rue Jean-Antoine de Ba¨ıf, 75205 Paris Cedex 13 (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201303865.
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1531
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The formation of a racemate
in this case also demonstrated
the lack of assistance of I in the
stereoselective control of the
aldol reaction (Scheme 3). To
obtain homogeneous conditions
for the aldol reaction, with
avidin being extremely hydrosol-
uble, the reaction medium was
partitioned between water and
an organic co-solvent, where all
the reactants were soluble.
No denaturation of avidin was
expected under these reaction
conditions, as the stability of its
complex with (+)-biotin was pre-
viously described.[25] The forma-
tion of the biohybrid catalyst
I·Av was performed in situ by
mixing I and avidin for 30 min at
Scheme 1. Supramolecular encapsulation of the biotinylated catalyst in avidin.
room temperature. The chiral in-
and hydrogen bonding.[20] Avidin is extremely robust and
stable at high temperatures,[21] at extreme pH,[22] and at high
concentrations of denaturating agents.[23] The robustness of
this protein allows its use as a scaffold for a wide range of ap-
plications, which also rely on the fact that the valeric acid side
chain of biotin can be derivatized with little effect on the re-
markable affinity of the biotin–avidin complex.[24]
The synthesis of the cationic anchor began with the forma-
tion of the imidazole-amine 3
duction brought about by the avidin was first studied under
un-optimized catalytic conditions (Scheme 4). Biohybrid cata-
lyst I·Av was active in the aldol reaction performed in H2O/
MeOH 4:1 mixture, but only a small enantiomeric excess (ee)
was observed for the formation of the major anti diastereoiso-
mer (Scheme 4).
For all the catalytic tests, avidin was used in a small excess
(1.15 equiv for 1 equiv of I) to ensure the complete binding of
through a typical Gabriel synthe-
sis. The coupling of 3 with the
activated (+)-biotin afforded
precursor 4, which was then al-
kylated with 1-bromobutane
under standard conditions to
obtain the air-stable biotinylated
imidazolium salt 5 (Scheme 2).
As we have previously reported,
anion metathesis of zwitterion 6
and bromide 5 required the for-
mation of a non-isolable hydrox-
ide intermediate by using an
IRA-400 ion exchange resin.[16]
The biotinylated imidazolium
salt I containing the pyrrolidine
anion was obtained by simple
anion exchange between the hy-
droxide anion and zwitterion 6.
A control experiment with I as
the catalyst was performed to
assess its activity in the absence
of avidin. The desired aldol prod-
ucts were obtained with more
than 98% conversion under typi-
cal aldol conditions by using cy-
clohexanone as the solvent.[16]
Scheme 2. Synthesis of the biotinylated catalyst. Overnight=16 h.
www.chemeurj.org ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2014, 20, 1530 – 1538
1532

Full Paper
tion medium was incrementally varied. As shown in
Figure 1, a linear increase in the selectivity for the
major anti diastereomer was obtained by increasing
the amount of IL up to 80% v/v, where 70% ee was
obtained. Such linearity in the variation of enantiose-
lectivity suggests a possible rearrangement of the
avidin structure with increasing IL content. When
going beyond a 4:1 IL/water ratio, the enantioselec-
tivity decreased to 24% ee and no reaction occurred
when the reaction was performed in pure [Bmim]Br,
proving that a small amount of water was necessary
to completely solubilize the avidin and maybe to pre-
serve its 3D structure. To the best of our knowledge,
there are no reports in the literature on the confor-
Scheme 3. Initial reaction conditions for the aldol reaction.
Scheme 4. Reaction conditions using I·Av as the catalyst.
I. Based on these preliminary results, we evaluated the solvent
effect on the reaction by screening several organic solvents
and ILs (Table 1). When using water-miscible solvents such as
MeOH, DMF, and DMSO (entries 1–3), excellent conversions
were obtained after 48 h of reaction.
Hydrophobic solvents like AcOEt or CH2Cl2 (entries 4–5), pre-
dictably, led to lower conversion rates, the reactants not being
soluble in water. Carrying out the reaction in H2O/cyclohexa-
none led to complete conversion after two days. Despite an
excellent conversion rate, the enantioselectivity was still low
(under 20%). These results were improved only when we used
ionic liquids as co-solvents. Different 1-butyl-3-methylimidazoli-
um salts ([Bmim]X) were tested, and whereas the conversion
rates were lower than those obtained in common organic sol-
vents, the selectivity was improved to up to 28% ee when
using the water-soluble [Bmim]Br. Replacement of [Bmim]Br
with the highly hydrophobic [Bmim]NTf2 resulted in a decrease
of enantioselectivity (down to 4%, entry 9). The decrease in
conversion when using ILs could be associated with the high
viscosity of the IL at low temperatures. Moreover, the nature of
the solvent did not seem to affect the diastereomeric ratio
(d.r.). In an attempt to understand the effect of ILs on the se-
lectivity of the aldol reaction, the ratio of [Bmim]Br in the reac-
Figure 1. Influence of the ionic-liquid content on the enantioselectivity of
the reaction.
mational changes of avidin in an ionic liquid medium. We per-
formed circular dichroism (CD) measurements of avidin
(Figure 2, solid line) and I·Av (dotted line) in water and a 4:1
IL/water mixture. As shown in Figure 2, substantial changes in
avidin’s conformation can be observed in the IL/water mixture.
A significant difference between the spectra of both free
avidin and I·Av revealed a conformation change in both secon-
dary (far-UV spectrum) and tertiary (near-UV spec-
trum) structures of avidin. The distinctive peaks for b-
sheets at 197 nm (positive band) and 213 nm (nega-
tive band) in water were replaced by a negative ab-
sorption band of high magnitude at 203 nm in the
IL/water mixture, suggesting a possible appearance
of random coil structures. The broad peak around
260 nm in the near-UV IL/water spectrum suggests
Table 1. Solvent screening.
Entry Solvent (4:1)
Conversion [%][a] d.r. (syn/anti)[a,b] ee (syn [%]/anti [%])[b]
that the intra-monomeric disulfide bond between
Cys-4 and Cys-83 was reduced.[26] Further studies
must be performed to elucidate if specific structure
changes occur and to see if a new inter-monomeric
disulfide bond is formed under these conditions.
These conformational changes in avidin structure
may be responsible for the variation in the selectivity
of the reaction arising from the differing IL content.
As previously reported, the binding of I has no signif-
icant effect on the structure of avidin in water.[27–28]
1
2
3
4
5
6
7
8
9
H2O/MeOH
H2O/DMF
H2O/ DMSO
H2O/AcOEt
H2O/CH2Cl2
93
94
95
31
21
33:66
41:59
41:59
34:66
49:61
39:61
41:59
37:63
41:59
rac.:18
rac.:12
rac.:11
4:16
10:14
rac.:20
6:28
H2O/cyclohexanone >99
H2O/[Bmim]Br
H2O/[Bmim]BF4
H2O/[Bmim]NTf2
47
43
46
14:26
18:4
[a] d.r.=diastereomeric ratio. [b] Determined by chiral HPLC.
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1533
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
described in the literature.
A
control experiment was carried
out under these dilution condi-
tions with I as the catalyst in the
absence of avidin. No traces of
the aldol products were ob-
served after 48 h, confirming
that the catalytic activity of I
outside avidin was negligible in
these high dilution conditions
(Table 2, entry 2). This experi-
ment also allowed us to dis-
count any possible interference
by competitive reactions cata-
lyzed by the free catalyst I out-
side the avidin active site. These
results were also proof of the
supramolecular
complexation
between I and avidin, showing
that encapsulation of I inside
avidin is required for the reac-
tion to occur in an acceptable
time frame. A catalytic run was
also performed with zwitterion 6
with avidin to test the possible
complexation of the anionic cat-
alyst with any charged residue
Figure 2. Circular dichroism spectra of avidin (solid line) and I·Av (dotted line) at 0.5 mgmLÀ1
.
However, in the IL/water mixture, an important structural
change of avidin was observed in both the far- and near-UV
spectra when I was encapsulated in the protein.
inside avidin (Table 2, entry 3). No conversion was observed,
implying that the absence of the imidazolium moiety prevents
the correct encapsulation of the catalytic anion inside the
avidin pocket. A final control experiment was performed to de-
termine whether avidin in the 4:1 IL/water mixture exhibits
any catalytic activity (Table 2, entry 4). Once more, no conver-
sion was observed after 48 h.
At this point, several control experiments were carried out
to assess the role of each component of the reaction (Table 2).
A first reaction was performed under the optimized conditions
with 5·Av instead of I·Av, but no traces of the aldol products
were observed, highlighting the importance of the counter
anion in the catalytic process (Table 2, entry 1). One important
observation is that avidin, exhibiting a molecular weight of ap-
proximately 88 kDa, working in high substrate dilution condi-
tions was necessary. Since avidin exhibits a molecular weight
of approximately 88 kDa, substrate in high dilution was neces-
sary to ensure full solubilization of the protein and to lower
the viscosity of the reaction media; these high-dilution condi-
tions were approximately 100 times lower than those usually
As pH has a major impact on the behavior of proteins in
aqueous media, several tests were performed using different
buffers in the IL mixtures, allowing the pH to be varied from
1 to 12. Carrying out the reaction at pH 1 led to the degrada-
tion of I·Av, preventing catalysis (Table 3, entry 1). Unsurprising-
ly, pH has a dramatic influence on the catalyst efficiency: at
higher pH, the conversion was higher, but the selectivity was
lower. By plotting the diastereomeric ratio versus pH, it can be
seen that the overall trend shows that a more basic medium
leads to a decrease of enantioselectivity, but also gives
rise to the proportion of the syn isomers (Figure 3). The
hypothesis that a possible base-catalyzed reaction was
being carried out outside avidin was confirmed when
a blank experiment without avidin was performed using
NaOH as the sole catalyst (Table 3, entry 13). A racemic
mixture of aldol products was obtained, indicating that
the drop in selectivity observed at high pH results from
competition between the reaction occurring inside the
protein and the base-catalyzed reaction happening out-
side avidin, where there is no stereocontrol. As low pH
is required to favor the “inside” reaction, we chose to
keep the pH of the reaction medium at 3. The problem
of low conversion at this pH was solved by increasing
Table 2. Control experiments.
Entry
Catalyst
Conversion [%]
[a]
1
2
3
4
5·Av
–
–
–
–
[a]
I, no avidin
6, with avidin
avidin only
[a]
[a]
[a] No conversion was observed by HPLC.
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1534
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
how these locations could influence the reactivity of the
catalyst.
Table 3. Influence of pH.
Ward et al. recently reported that the insertion of an
amino acid in biotinylated catalysts increased the perfor-
mance of their metalloenzymes.[29] Based on these obser-
vations, we speculated that: 1) Aromatic amino acids
with hydrophobic side chains (l-Phe and l-Trp) would
display p–p interactions with hydrophobic residues
inside the cavity; 2) Adding a longer alkyl chain between
the biotin moiety and the imidazolium cation would
allow more flexibility inside the protein; 3) l-Proline,
with its high conformational rigidity, would restrain the
liberty of movement of the anionic catalyst. All tert-bu-
toxycarbonyl (Boc)-protected amino acids were first cou-
pled with 3 through a simple and efficient procedure
Entry
pH[a]
Conversion [%][b]
ee (anti) [%][b]
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
9
10
11
12
–
–
43
43
38
41
69
81
89
86
89
69
85
>99
61
70
63
59
42
27
22
17
13
16
10
rac.
9
10
11
12
13[c]
using
N-hydroxysuccinimide/dicyclohexylcarbodiimide
(NHS/DCC) in acetonitrile. A simple deprotection step,
followed by the coupling of (+)-biotin N-hydroxysuccini-
mide ester, led to the imidazole precursors. Alkylation
using 1-bromobutane, followed by anion metathesis
with 6 afforded II, III, IV, and V as air-stable compounds.
Full experimental details for the synthesis of these com-
pounds can be found in the Supporting Information.
The formation of the biohybrid catalysts was performed in situ
and catalysis was carried out by using the optimized condi-
tions already described. The catalytic activity of II–V and their
contributions to the stereocontrol of the reaction were also as-
sessed with avidin-free catalytic runs. All results are shown in
Table 4.
NaOH 0.1 m
[a] pH 1–2 and 12, chloride buffer; pH 3–6, acetate buffer; pH 7–8, phosphate
buffer; pH 9, trishydroxymethylaminomethane (TRIS) buffer; pH 10–11, carbonate
buffer. [b] Determined by chiral HPLC. [c] No avidin was added.
As previously observed, no chiral induction and extremely
low conversions were obtained for all the tested catalysts in
the absence of avidin, owing to the extremely diluted condi-
tions. The lack of reactivity underlines the crucial role of avidin
in terms of acceleration of the reaction rate. The variation of
the structure of the cationic thread did not affect the complex-
ation efficiency between the catalyst and avidin. However, no
significant variation in the obtained ee was observed when
using different cationic threads. These results suggest that the
pyrrolidine moiety has enough freedom of motion inside the
Figure 3. Trends in selectivity with varying pH.
the substrate concentration to
5ꢁ10À2 m, leading to 94% con-
version after 48 h with the same
enantioselectivity.
To gain a better understand-
ing of the interactions between
the biotinylated imidazolium salt
and avidin, as well as the mobili-
ty of the catalytic anion in avi-
din’s binding site, we designed
and synthesized different cation-
ic threads (Scheme 5). The struc-
ture of the biotinylated thread
was modified in an attempt to
position the anionic catalyst at
different locations inside the
avidin binding site and to see Scheme 5. Variation of the structure of the cationic scaffold.
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1535
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
To get a better insight into the impact of the intrinsic
chirality of the anion, the chiral diastereomeric analogues
Table 4. Effect of the cationic scaffold.[a]
of
I bearing l- and d- pyrrolidinemethanesulfonate
anions were tested in the absence of avidin, using an ap-
propriate substrate concentration (1m). The results are
compiled in Scheme 6, where only the major diastereo-
mers of the obtained aldol products are shown. This
time, the syn diastereomers were the major product and
the selectivities depended on the absolute configuration
of the anion, each anion favoring one enantiomer of the
anti products. However, the enantiomeric excess for the
anti diastereomer was very low (20% ee max.). Two con-
clusions can be drawn from these results: 1) There is no
obvious match–mismatch evidence in the case where the
racemic anion of I is embedded in the protein, as both
versions of the catalyst led to similar results in the forma-
tion of the (S,R) anti enantiomer (Scheme 6); 2) Even if
the intrinsic chirality of the anion is directly responsible
for the stereoselective outcome of the reaction in the ab-
sence of avidin, it seems that the chirality inside the pro-
tein governs and improves the selectivity of the biohybrid cat-
alyst.
Entry
Catalyst
Conversion [%][b]
ee (anti) [%][b]
I·Av
I
II·Av
II
III·Av
III
IV·Av
IV
94
<1
88
<2
89
<1
86
<1
88
70
<3
68
<4
66
<4
66
<4
66
1
2
3
4
5
V·Av
V
<2
<5
[a] Substrate concentration: 5ꢁ10À2 m in each case. [b] Determined by chiral
HPLC.
avidin cavity to adopt the same optimized position in all these
cases, independent of the structure of its cationic counterpart.
These results also show that the cation does not actively par-
ticipate in the catalytic process and that the movement of the
anion inside the avidin pocket is the crucial parameter of this
biohybrid species. To support these observations, the two
enantiomers of the 2-pyrrolidinemethanesulfonate salt of 6
were prepared. Each catalyst bearing a chiral counter anion
was tested under the optimized conditions and this, surprising-
ly, led to similar results compared with their racemic counter-
parts in the presence of avidin (Scheme 6). Interestingly, the d-
proline derivative did not give the expected enantiomer of the
usual anti aldol product, but gave the (S,R) enantiomer with an
enantiomeric excess similar to the results obtained by using I.
As both enantiomers of the syn aldol product can be ob-
tained in similar excess regardless of whether the l- or d-pyrro-
lidinemethanesulfonate anion is used, it is clear that the sulfo-
nate moiety dictates the approach of the aldehyde towards
the enamine. Complementary experiments were carried out
with three non-biotinylated catalysts to understand the role of
both the anion and cation in the model aldol reaction
(Scheme 7). The presence of the imidazolium cation close to
the sulfonate group favors the approach of the aldehyde
through hydrogen bonding of H-2 of the imidazolium to the
oxygen atom of the aldehyde. Steric hindrance results in attack
of the enamine on its Si face, leading to the formation of the
syn diastereomer. When the imidazolium H-2 is replaced by
a methyl group, the possibility of hydrogen bonding is re-
moved and the selectivity of the reaction is ruled purely by
steric restrictions; thus, the aldehyde is forced to approach the
“opposite face” of the enamine (Scheme 8).
The importance of the imidazolium cation was also high-
lighted by the results obtained when using the deprotonated
version of 6 as the catalyst; in this instance, the selectivity de-
creased considerably.
While it is still unclear why the anti diastereomer was prefer-
entially formed when the reaction was carried out inside
avidin, one could suppose that the sulfonate group can inter-
act with a cationic residue inside avidin, resulting in a confor-
mational change to the sulfonated arm. In this case, the steric
hindrance initially caused by the catalyst would play less of
a role in the discrimination of the enamine faces and the ob-
served stereocontrol would be induced by the chiral pocket of
the avidin (for more details see the Supporting Information).
This hypothesis is also consistent with the results obtained
with the different enantiomerically pure anions, which gave
the same aldol product. These results also suggest the possibil-
ity of free motion of the anions inside avidin and show that
racemic anions can be used to form the diastereomeric cata-
Scheme 6. Influence of the chirality of the anion.
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1536
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
about by the protein around the catalyst was dis-
cussed. Further studies on the substrate tolerance of
our system are currently under investigation in our
group and efforts are currently being made towards
the synthesis of new catalytic anions with different
structures to improve the catalyst’s activity. Also, the
design of novel anionic compounds that can act as li-
gands for transition metals could lead to the assem-
bly of new biohybrid catalytic systems.
Experimental Section
All organic compounds were purchased in their highest
available purity and used without further purification.
Egg-white Avidin was purchased from Lee Biosolutions.
NMR experiments were recorded on an Avance 300
Brucker, at 300 and 75.5 MHz, and an Avance 400 Bruck-
er, at 400 and 100 MHz, with non-spinning samples. All
NMR experiments were obtained by the use of the com-
mercially available sequence on Brucker spectrometers.
Coupling constants are given in Hertz (Hz) and chemical
shifts are given in ppm (d) and measured relative to re-
sidual solvent. Mass spectral data were obtained by the
Universitꢂ de Montrꢂal Mass Spectrometry Facility and
were recorded on a Mass spectrometer TSQ Quantum Ultra
(Thermo Scientific) with accurate mass options instrument.
Scheme 7. The role of the imidazolium cation in the catalytic process.
Typical procedure for the Aldol reaction
Stock solutions of starting materials and catalysts were prepared:
S1 refers to a 0.2m solution of p-nitrobenzaldehyde in [Bmim]Br; S2
refers to a 2m solution of cyclohexanone in [Bmim]Br; S3 refers to
a 0.03m solution of catalyst in water/buffer. In a vial, avidin (12 mg,
13.8 U/mg, binds 0.678 mmol of biotinylated catalyst, 1.13 equiv
compared with catalyst) was suspended in [Bmim]Br (60 mL). S2
(10 mL, 20 mmol, 10 equiv) and S3 (20 mL, 0.6 mmol, 30 mol%) were
added. Once the avidin was completely dissolved, the reaction
mixture was stirred for 30 min at 238C to allow the formation of
the catalytic species. S1 (10 mL, 2 mmol, 1 equiv) was added and the
reaction mixture was stirred for 48 h at 48C. Diethyl ether (0.5 mL)
was then added and the vial was vortexed for 1 min. The ether
phase was analyzed by chiral HPLC (ChiralPak AD-H column, hex-
anes/isopropyl alcohol (IPA) 95/5, 0.5 mLminÀ1, l=254 nm) tR (anti
isomer)=40.84 (minor), 56.69 (major), tR (syn isomer)=27.96
(minor), 37.35 min (major).
Scheme 8. Possible transition states for the aldol reaction performed with a)
[Bmim]-l-6 and b) [Bdmim]-l-6.
lyst. Combined with its ionic liquid compatibility, this could be
an asset for the preparation of substrate-tolerant biocatalysts
without additional optimization of the avidin active site. More-
over, current work in our group aims at the synthesis of new
chiral pyrrolidine-based anions to achieve a fine tuning and
a better understanding of this system in catalysis, especially for
the development of new reactions.
Acknowledgements
This work was supported by the Natural Sciences and Engi-
neering Research Council of Canada, the Fonds Quꢂbꢂcois de
la Recherche sur la Nature et les Technologies, the Centre of
Green Chemistry and Catalysis, the Canada Foundation for In-
novation and the Universitꢂ de Montreal. We thank colleagues
for careful discussions of this manuscript.
Conclusion
We presented the assembly of a biohybrid catalyst obtained
by the complexation of biotinylated imidazolium salts with rac-
emic or enantiomerically pure catalytic anions and their activity
in the aldol reaction performed in IL/water mixtures. Circular
dichroism studies revealed the influence of the ionic liquid on
the conformation of avidin and the biohybrid catalyst. The
high degree of freedom of the anionic catalyst inside the pro-
tein was demonstrated for different biohydrid catalysts and
the influence of the second-sphere coordination brought
Keywords: aldol reaction · biotin–avidin · ionic liquids ·
supramolecular catalysis
[1] a) E. Z. Eisenmesser, O. Millet, W. Labeikovsky, D. M. Korzhnev, M. Wolf-
Watz, D. A. Bosco, J. J. Skalicky, L. E. Kay, D. Kern, Nature 2005, 438, 117–
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1537
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
121; b) Z. Kurkcuoglu, A. Bakan, D. Kocaman, I. Bahar, P. Doruker, PLoS
Comput. Biol. 2012, 8, e1002705.
[2] a) P. J. Deuss, R. Heeten, W. Laan, P. C. J. Kamer, Chem. Eur. J. 2011, 17,
4680–4698; b) Z. Dong, Q. Luo, J. Liu, Chem. Soc. Rev. 2012, 41, 7890–
7908.
[16] a) V. Gauchot, A. R. Schmitzer, J. Org. Chem. 2012, 77, 4917–4923; b) V.
Gauchot, J. Gravel, A. R. Schmitzer, Eur. J. Org. Chem. 2012, 6280–6284.
[17] L. Leclercq, A. R. Schmitzer, Organometallics 2010, 29, 3442–3449.
[18] a) M. D. Melamed, N. M. Green, Biochem. J. 1963, 89, 591–599; b) N. M.
Green, Adv. Protein Chem. 1975, 29, 85–133; c) Avidin-Biotin Technology
in Methods in Enzymology, Vol. 184 (Eds.: M. Wilchek, E. A. Bayer) Aca-
demic Press, Inc., San Diego, California, 1990.
[3] H. Yamaguchi, T. Hirano, H. Kiminami, D. Taura, A. Harada, Org. Biomol.
Chem. 2006, 4, 3571–3573.
[4] a) D. Shabat, H. Itzhaky, J. L. Reymond, E. Keinan, Nature 1995, 374,
143–145; b) S. Nimri, E. Keinan, J. Am. Chem. Soc. 1999, 121, 8978–
8982.
[5] M. E. Wilson, G. M. Whitesides, J. Am. Chem. Soc. 1978, 100, 306–307.
[6] J. Collot, J. Gradinaru, N. Humbert, M. Skander, A. Zocchi, T. R. Ward, J.
Am. Chem. Soc. 2003, 125, 9030–9031.
[7] C. Letondor, A. Pordea, N. Humbert, A. Ivanova, S. Mazurek, M. Novic,
T. R. Ward, J. Am. Chem. Soc. 2006, 128, 8320–8328.
[8] J. Pierron, C. Malan, M. Creus, J. Gradinaru, I. Hafner, A. Ivanova, A.
Sardo, T. R. Ward, Angew. Chem. 2008, 120, 713–717; Angew. Chem. Int.
Ed. 2008, 47, 701–705.
[19] a) O. Livnah, E. A. Bayer, M. Wilchek, J. L. Sussman, Proc. Natl. Acad. Sci.
USA 1993, 90, 5076–5080; b) A. Feltus, S. Ramanathan, S. Daunert, Anal.
Biochem. 1997, 254, 62–68; c) C. Rosano, P. Arosio, M. Bolognesi,
Biomol. Eng. 1999, 16, 5–12.
[20] a) O. H. Laitinen, H. R. Nordlund, V. P. Hytonen, S. T. H. Uotila, A. T. Martti-
la, J. Savolainen, K. J. Airenne, O. Livnah, E. A. Bayer, M. Wilchek, M. S.
Kulomaa, J. Biol. Chem. 2003, 278, 4010–4014; b) D. E. Hyre, I. Le Trong,
E. A. Merritt, J. F. Eccleston, N. M. Green, R. E. Stenkamp, P. S. Stayton,
Protein Sci. 2006, 15, 459–467.
[21] N. M. Green, Methods Enzymol. 1990, 184, 51–67.
[22] H. R. Nordlund, V. P. Hytonen, O. H. Laitinen, S. T. H. Uotila, E. A. Niska-
nen, J. Savolainen, E. Porkka, M. S. Kulomaa, FEBS Lett. 2003, 555, 449–
454.
[9] T. R. Ward, Chem. Eur. J. 2005, 11, 3798–3804; T. R. Ward, Acc. Chem.
Res. 2011, 44, 47–57.
[10] T. K. Hyster, L. Knçrr, T. R. Ward, T. Rovis, Science 2012, 338, 500–503.
[11] N. Noujeim, L. Leclercq, A. R. Schmitzer, Curr. Org. Chem. 2010, 14,
1500–1516.
[12] a) T. Welton, Chem. Rev. 1999, 99, 2071–2084; b) V. I. Pꢃrvulescu, C. Har-
dacre, Chem. Rev. 2007, 107, 2615–2665; c) M. Deetlefs, K. R. Seddon,
Green Chem. 2010, 12, 17–30.
[13] a) K. Fujita, M. Forsyth, D. R. MacFarlane, R. W. Reid, G. D. Elliott, Biotech-
nol. Bioeng. 2006, 94, 1209–1213; b) D. Constatinescu, C. Herrmann, H.
Weingartner, Phys. Chem. Chem. Phys. 2010, 12, 1756–1763; c) J. V. Ro-
drigues, V. Prosinecki, I. Marrucho, L. P. N. Rebelo, C. M. Gomes, Phys.
Chem. Chem. Phys. 2011, 13, 13614–13616.
[23] T. Sano, M. W. Pandori, X. M. Chen, C. L. Smith, C. R. Cantor, J. Biol.
Chem. 1995, 270, 28204–28209.
[24] E. A. Bayer, E. Skutelsky, M. Wilchek, Methods Enzymol. 1979, 62, 308–
315.
[25] M. Gonzꢅlez, C.-E. ArgaraÇa, G. D. Fidelio, Biomol. Eng. 1999, 16, 67–72.
[26] L. Pugliese, A. Coda, M. Malcovati, M. Bolognesi, J. Mol. Biol. 1993, 231,
698–710.
[27] N. M. Green, M. D. Melamed, Biochem. J. 1966, 100, 614–621.
[28] F. Zsila, Anal. Biochem. 2009, 391, 154–156.
[29] U. E. Rusbandi, C. Lo, M. Skander, A. Ivanova, M. Creus, N. Humbert, T. R.
Ward, Adv. Synth. Catal. 2007, 349, 1923–1930.
[14] a) K. Bica, P. Gaertner, Eur. J. Org. Chem. 2008, 3235–3250; b) F. van
Rantwijk, Roger A. Sheldon, Chem. Rev. 2007, 107, 2757–2785; c) Y.
Kong, R. Tan, L. L. Zhao, D. H. Yin, Green Chem. 2013, 2422–2433; d) Y.
Qian, X. Zheng, Y. Wang, Eur. J. Org. Chem. 2010, 3672–3677.
[15] E. P. ꢄvila, G. W. Amarante, ChemCatChem 2012, 4, 1713–1721.
Received: October 2, 2013
Published online on January 2, 2014
Chem. Eur. J. 2014, 20, 1530 – 1538
www.chemeurj.org
1538
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim