Preparation of an Immobilized Lipase-Palladium Artificial Metalloenzyme as Catalyst in the Heck Reaction: Role of the Solid Phase

Article abstract of DOI:10.1002/adsc.201500014

A p-nitrophenylphosphonate palladium pincer was synthesized and selectively inserted by irreversible attachment on the catalytic serine of different commercial lipases with good to excellent yields in most cases. Among all, lipase from Candida antarctica B (CAL-B) was the best modified enzyme. The artificial metalloenzyme CAL-B-palladium (Pd) catalyst was subsequently immobilized on different supports and by different orienting strategies. The catalytic properties of the immobilized hybrid catalysts were then evaluated in two sets of Heck cross-coupling reactions under different conditions. In the first reaction between iodobenzene and ethyl acrylate, the covalent immobilized CAL-B-Pd catalyst resulted to be the best one exhibiting quantitative production of the Heck product at 70°C in dimethylformamide (DMF) with 25% water and particularly in pure DMF, where the soluble Pd pincer was completely inactive. A post-immobilization engineering of catalyst surface by its hydrophobization enhanced the activity. The selectivity properties of the best hybrid catalyst were then assessed in the asymmetric Heck cross-coupling reaction between iodobenzene and 2,3-dihydrofuran retrieving excellent results in terms of stereo- and enantioselectivity.

Full text of DOI:10.1002/adsc.201500014

FULL PAPERS
DOI: 10.1002/adsc.201500014
Preparation of an Immobilized Lipase-Palladium Artificial
Metalloenzyme as Catalyst in the Heck Reaction: Role of the
Solid Phase
Marco Filice,a,* Oscar Romero,a Antonio Aires,b Jose M. Guisan,a
Angel Rumbero,b and Jose M. Palomoa,*
a
Departamento de Biocatµlisis, Instituto de Catµlisis (CSIC), Marie Curie 2, Cantoblanco, Campus UAM, 28049 Madrid,
Spain
Fax: (+34)-91-585-4760; phone: (+34)-91-585-4768; e-mail: marcof@icp.csic.es or josempalomo@icp.csic.es
Departamento de Química Orgµnica, Facultad de Ciencias, Universidad Autónoma de Madrid, Cantoblanco, 28049
b
Madrid, Spain
Received: January 14, 2015; Revised: June 12, 2015; Published online: August 19, 2015
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201500014.
Abstract: A p-nitrophenylphosphonate palladium sulted to be the best one exhibiting quantitative pro-
pincer was synthesized and selectively inserted by ir- duction of the Heck product at 708C in
reversible attachment on the catalytic serine of dif- dimethylformamide (DMF) with 25% water and par-
ferent commercial lipases with good to excellent ticularly in pure DMF, where the soluble Pd pincer
yields in most cases. Among all, lipase from Candida was completely inactive. A post-immobilization engi-
antarctica B (CAL-B) was the best modified enzyme. neering of catalyst surface by its hydrophobization
The artificial metalloenzyme CAL-B-palladium (Pd) enhanced the activity. The selectivity properties of
catalyst was subsequently immobilized on different the best hybrid catalyst were then assessed in the
supports and by different orienting strategies. The asymmetric Heck cross-coupling reaction between
catalytic properties of the immobilized hybrid cata- iodobenzene and 2,3-dihydrofuran retrieving excel-
lysts were then evaluated in two sets of Heck cross- lent results in terms of stereo- and enantioselectivity.
coupling reactions under different conditions. In the
first reaction between iodobenzene and ethyl acry- Keywords: immobilization; lipases; metalloenzymes;
late, the covalent immobilized CAL-B-Pd catalyst re- palladium pincers
Introduction
Among the three strategies, the covalent protocol
presents several advantages, such as a more precise
Artificial metalloenzymes are formed by the combina- knowledge of the location of the metal complex by
tion of a biomolecule (peptide, DNA or protein)[1–3]
–
a site-selective coupling, no dependence on the bind-
as host – and an organometallic compound. Being ing affinity of the complex to the protein and the pos-
a second coordination sphere, the protein matrix con- sibility to promote the target modification in different
fers to the metal catalysts better performances espe- sites on the protein structure. Recently, elegant strat-
cially in terms of stereo- and regioselectivity.[4–8] egies based on the modification of the native protein
Sometimes, non-catalytic metals can even be trans- scaffold[13] or the combination of genetic modification
formed into very interesting catalysts by means of of proteins by insertion of unnatural amino acids, and
their insertion into a proteic core.[9]
orthogonal strategies promoting the selective covalent
The metal insertion on the protein scaffold has coupling of the organometallic complexes have been
been generally performed by three different strat- developed.[14] Nevertheless, despite their tremendous
egies:[1–3] (i) supramolecular anchoring (by a strong af- potential, most of the artificial enzymes described to
finity ligand, such as biotin[10]), (ii) dative anchoring date have been mainly developed under a more aca-
(by direct metal coordination with amino acid resi- demic point of view. Hence, with a more applied aim,
dues)[11] or (iii) covalent attachment (by selective an- the development of concepts such as versatility or
choring of a metal-binding ligand on an a-amino large availability of the used host protein must be
acid).[12]
strongly considered. For example, the combination of
Adv. Synth. Catal. 2015, 357, 2687 – 2696
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2687

FULL PAPERS
Marco Filice et al.
proteins with high catalytic versatility could generate
higher possibilities depending on the transition metal
or metal binding ligands inserted. Moreover, the ac-
cessibility to a high amount of the protein is mandato-
ry in order to obtain a high amount of the final artifi-
cial metalloenzymes.In this context, lipases – acylgly-
cerol hydrolases – fulfill both requirements. These are
very versatile enzymes with successful results in dif-
ferent processes using non-natural substrates[15] and
they are commercially available in large amounts. Re-
cently, these enzymes have been combined with
metals in the preparation of several hybrids cata-
Scheme 1. Different phosphonate-based structures. Organo-
palladium pincer (1) and lipase irreversible inhibitor (2).
lysts.[16–18] Considering the catalytic mechanism of metric Heck cross-coupling reaction with excellent re-
these enzymes – with a catalytic serine in the active sults.
site – p-nitrophenol (pnP) phosphonate esters (typi-
cally used as serine-protease irreversible inhibitors[19]
)
can be applied as practice anchoring units on the or- Results and Discussion
ganometallic complex for its specific incorporation
into the catalytic pocket.[19]
Site-Directed Incorporation of the Pd Complex 1 to
In general, these kinds of artificial metalloenzymes Different Lipases
have been created in solution and the host protein
showed a critical influence on the catalytic properties The anchoring of the p-nitrophenylphosphonate palla-
of the created artificial metalloenzymes. Indeed, espe- dium pincer to the lipase catalytic serine (Scheme 2)
cially the stability of the enzyme clearly results as the was followed by an enzymatic hydrolytic activity
bottleneck of the entire strategy mainly considering drop, determined by monitoring the difference in ab-
the drastic reaction conditions expected by the metal sorbance increase of the released p-nitrophenolate
transition application. Hence, this kind of catalyst anion by UV spectrophotometry.
cannot be reused for several cycles.
The reaction of 1 with different lipases was carried
Therefore in this paper we describe the preparation out using five equivalents of 1 with respect to the
of an artificial metalloenzyme on a solid support ful- lipase amount and the lipases were previously cova-
filling all these criteria. Thanks to the use of the solid- lently immobilized on CNBr-activated Sepharose
phase chemistry, many advantages such as the possi- (Sepha-CNBr) (Scheme 3).
bility to use excess of metal complexes, quantitative
transformations or easy purification have been gener- ganometallic insertion depending on the protein.
ated.[21]
Lipase from Candida antarctica B (CAL-B) was rap-
Herein, the design of immobilized Pd-lipase artifi- idly inactivated, and after 5 min the artificial metal-
Figure 1 shows the large differences in yield of or-
cial metalloenzymes is described. The p-nitrophenyl- loenzyme
phosphonate palladium pincer 1 (Scheme 1) was syn- (Figure 1).
Sepha-CNBr-CAL-B-1
was
formed
thesized and used as an organometallic complex.
Mass spectroscopic analysis (MALDI-TOF) corro-
For that purpose, lipases from different sources, dif- borated the anchoring of palladium pincer achieving
ferent enzyme immobilization protocols (covalent at- the desired organoPd-CAL-B hybrid (Figure S1, Sup-
tachment and physical adsorption) and different sup- porting Infomation).
port materials (Sepharose and Sepabeadsꢀ) were
Lipases from Rhizomucor miehei (RML) or Candi-
combined in order to create different immobilized ar- da rugosa (CRL) also achieved an almost complete
tificial enzymes.
Finally, a post-immobilization catalyst engineering
(based on the specific chemical modification of the
supporting matrix surface) was evaluated in order to
obtain the best artificial metalloenzyme. All the new
artificial metalloenzymes were evaluated as catalysts
À
in C C coupling Heck reactions under different con-
ditions. The best lipase-Pd artificial metalloenzyme
showed higher activity than the soluble Pd catalyst,
also maintaining a complete trans selectivity and it
was recycled several times. Beside the activity, even
the stereo- and enantioselectivity of the best hybrid
catalyst were evaluated by its application in an asym- teins.
Scheme 2. General strategy to prepare lipase metallopro-
2688
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2687 – 2696

FULL PAPERS
Preparation of an Immobilized Lipase-Palladium Artificial Metalloenzyme
exist in their closed conformation (active site blocking
by the lid). Thus, in these two unsuccessful cases the
metal incorporation was tested on the lipase immobi-
lized by interfacial activation on Sepharose function-
alized by octyl groups (Sepha-Octyl) – a hydrophobic
support – (Scheme 3b).
In this immobilization strategy the lipase is ad-
sorbed on the support fixing its open conformation[22]
and giving to the substrates the highest accessibility
to the active site. In the case of ROL immobilized on
Sepha-octyl, most of the lipase activity (~80%) was
lost by the insertion of 1 in the active site (Figure 2a).
However, no modification was observed for Sepha-
Octyl immobilized PFL (Figure 2b). This phenomen-
on could be explained because PFL shows the longer
peptide lid among all these enzymes with a huge hy-
drophobic oxyanion.[23] When the usual lipase inhibi-
tor p-diethyl-p-nitrophenyl phosphate (2) was used,
the catalytic serine of PFL immobilized on Sepha-
CNBr was modified at 80% in 125 min.
Therefore, these results demonstrated that 1 can be
incorporated in most of the immobilized lipases
tested, although the most effective result was ach-
ieved using CAL-B as protein scaffold.
Scheme 3. Different orienting strategies for lipase immobili-
zation.
Figure 1. Coupling reaction of organometallic pincer 1 and
different lipases immobilized on CNBr-activated Sepharose.
active site modification with 1 but with longer reac-
tion times (65 and 125 min, respectively). Using lipase
from Thermomyces lanuginosus (TLL), the highest
achieved conversion was around 80%. However, li-
pases from Pseudomonas fluorescens and Rhizopus
oryzae were not modified at all, maintaining the ini-
tial activity (Figure 1). The lipase modification has
been performed in aqueous media on enzymes immo-
bilized by covalent attachment at pH 7 (through the
Figure 2. Effect of different orienting immobilization strat-
egies during the coupling reaction between organometallic
pincer 1 and a) Rhizopus oryzae lipase (ROL) and b) Pseu-
N-terminus), where generally these enzymes mainly domonas fluorescens lipase (PFL).
Adv. Synth. Catal. 2015, 357, 2687 – 2696
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2689

FULL PAPERS
Marco Filice et al.
Scheme 4. Heck cross-coupling reaction between iodoben-
zene and ethyl acrylate.
more sustainable conditions, the reaction was tested
at 708C in DMF but unfortunately 1 did not catalyze
the reaction. However, the organometallic complex
was active in the presence of 25% of water (v/v) in
DMF at 708C – greener conditions – achieving com-
plete conversion at 29 h with a TOF value of 142Æ
7 hÀ1. The presence of more water content in the reac-
tion [50% (v/v)] reduced the catalyst activity to
a 47% yield and a TOF value of 27Æ2 hÀ1 (Table 1).
The stability of the free and supported pincers has
been well characterized demonstrating no leaching of
the Pd molecules under the given reaction condi-
tions.[24–25]
Figure 3. Manual docking of the conjugation of 1 to catalytic
serine of Candida antarctica lipase (fraction B) (CAL-B)
based on the X-ray structure (pdb: 1TCA). Figure was gen-
erated using Pymol software.
Thus the activity of the immobilized Sepha-CNBr-
This result could be explained analyzing the 3D CAL-B-1 metalloenzyme was tested applying the con-
structure of the CAL-B and the size of 1. In fact, ditions where Pd pincer 1 was active in the Heck re-
among the tested lipases, this enzyme presents the action (Table 1, entries 1 and 3). We preliminarily
shorter lid with a small oxyanion around 11 Š width evaluated the immobilized enzyme without the Pd
and 17 Š long, similar to the size of 1 (15 Š long), pincer 1 but incorporating the inhibitor 2 on the
permitting to the phosphonate ester much more ac- active site (Table 2, entries 1 and 2).
cessibility for reacting with the catalytic serine
(Figure 3).
This Sepha-CNBr-CAL-B-2 was completely inac-
tive under both reaction conditions. Unfortunately,
Moreover, also the presence of several hydrophobic the immobilized Sepha-CNBr-CAL-B-1 did not cata-
groups around the oxyanion could stabilize the pres- lyze the reaction under any of the tested conditions
ence of the pincer located on the active site.
(Table 2, entries 3 and 4).
Pursuing our search for optimal immobilized bio-
catalysts, Novozym435 – the commercial immobilized
CAL-B catalyst where the lipase is immobilized
mainly by hydrophobic interactions on Lewatit resin –
was successfully modified with 1 and the new hybrid
Assessment of Immobilized CAL-B-Pd-1 Artificial
Metalloenzymes in Heck Reactions
First, we evaluated the catalytic capacity of the sole Novozym435-1 was used to catalyze the Heck reac-
soluble Pd pincer 1 in the Heck reaction (Scheme 4). tion. This hybrid slightly catalyzed the reaction under
Table 1 shows the results of the reaction between the applied conditions, achieving negligible yield
iodobenzene and ethyl acrylate in the presence of (ꢀ5%, Table 2, entries 5 and 6) of the desired prod-
1 (Scheme 4) under different experimental conditions. uct. In this case, the reason for this very low conver-
Under the standard conditions, 1208C in DMF, sion was attributed to the immobilization methodolo-
1 catalyzed the Heck reaction with >99% yield in gy. In fact, in this case, the immobilization was rever-
18 h with a TOF value of 230Æ11 hÀ1. In order to use sible and the enzyme was leaching from the support
Table 1. Optimization of reaction conditions for Heck cross-coupling.[a]
Entry
Catalyst
Water [%, v/v]
Temp. [8C]
Time [h]
Yield [%]
TOF [hÀ1]
1
2
3
4
1
1
1
1
0
0
25
50
120
70
70
18
72
29
72
>99
0
>99
47
230Æ11
nd
142Æ7
70
27Æ2
[a]
Reaction conditions: 0.247 mmol of iodobenzene, 0.55 mmol of ethyl acrylate, 0.54 mmol (0.024 mol% of Pd relative to io-
dobenzene), 0.412 mmol of triethylamine, 1 mL of DMF.
2690
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2687 – 2696

FULL PAPERS
Preparation of an Immobilized Lipase-Palladium Artificial Metalloenzyme
Table 2. Heck reaction catalyzed by different artificial metalloenzymes.[a]
Entry
Catalyst
Water [%, v/v]
Temp. [8C]
Time [h]
Yield [%]
TOF [hÀ1]
1
2
3
4
5
6
8
9
Sepha-CNBr-CAL-B-2
Sepha-CNBr-CAL-B-2
Sepha-CNBr-CAL-B-1
Sepha-CNBr-CAL-B-1
Novozym435-1
0
25
0
25
0
25
0
120
70
120
70
120
70
120
70
72
72
72
72
72
72
72
55
35.5
0
0
0
0
n.d
n.d
n.d
n.d
n.d
n.d
16Æ1
75Æ3
116Æ5
2
Novozym435-1
<5
28
>99
>99
SP-CHO-CAL-B-1
SP-CHO-CAL-B-1
SP-CHO-CAL-B-1
0
25
10
70
[a]
Reaction conditions: 0.247 mmol of iodobenzene, 0.55 mmol of ethyl acrylate, 200 mg of each catalyst (0.024 mol% of Pd
relative to iodobenzene), 0.412 mmol of triethylamine, 1 mL of DMF.
Figure 4. SDS-PAGE analysis of Novozym 435 after incuba-
tion in Heck reaction medium at different conditions: 1)
H2O (100%); 2) DMF 75%/H2O 25% (v/v) at 708C for 2 h;
3) 6 h; 4) 24 h. DMF 100% (v/v) at 1208C; for 5) 2 h, 6) 6 h,
and 7) 24 h.
Figure 5. SEM micrographs of the macroporous matrices
under the reaction conditions tested. This hypothesis
used as solid support. a) Sepharose. b) Sepabeads.
was confirmed by SDS-PAGE of the Novozym435
under the different conditions used (Figure 4).
No protein was observed on the support after incu- tered their structure under different reaction condi-
bation at 1208C in DMF, and the same occurred in tions (Figure 5b).
the presence of 25% water and 708C. This result also
Thus, considering the previously described draw-
demonstrated that the soluble CAL-B-1 metalloen- backs, CAL-B was immobilized by a multi-covalent
zyme did not work under these reaction conditions attachment on a Sepabeads resin activated with alde-
where the homogeneous pincer was active, the use of hyde groups (SP-CHO) (Scheme 3, bottom). This
the solid-phase thus being mandatory . This phospho- strategy has been demonstrated to confer a high sta-
nate system (irreversibly attached to the catalytic bility to the enzyme.[27] Furthermore, this immobiliza-
serine in the protein) seems to stabilize the palladium tion is performed through the richest area of lysine
complex into the protein avoiding the leaching of the residues of the enzyme. In the case of CAL-B this
metal in comparison with other chemistries.[26]
area is located on the opposite side respect to the lid
Therefore, considering these negative results (the and the oxyanion location (Scheme 3, bottom). CAL-
use of a very hydrophilic matrix such as Sepharose B-SP-CHO preparation was fully modified with 1 in
and the selection of a reversible immobilization strat- 5 min achieving SP-CHO-CAL-B-1. This new immo-
egy), we decided to use a much more hydrophobic bilized hybrid metalloenzyme catalyzed the Heck re-
resin such as Sepabeadsꢁ, with no swelling effect re- action, yielding 28% of product at 1208C in pure
lated to solvent changes. Figure 5 shows the extreme DMF and >99% yield after 35.5 h at 708C in DMF
differences between the Sepharose (Figure 5a) and with 25% (v/v) water, with a TOF value of 116Æ5 hÀ1
Sepabeadsꢁ (Figure 5a) in a low concentration of (Table 2, entries 8 and 10). Surprisingly, the immobi-
water.
lized metalloenzyme catalyzed the Heck reaction with
The Sepharose structure is collapsed under these >99% yield at 708C in pure DMF, conditions where
conditions resulting in a critical impediment for the the pincer 1 was not active (Table 2, entry 9). There-
entrance of the substrates and therefore being a possi- fore, this successful result demonstrates the critical
ble additional explanation for the inactivity of Sepha- role of the enzyme environment together with the
CNBr-CAL-B-1 besides the CAL-B conformational support matrix hydrophobicity.
changes previously described. On the other hand, Se-
In this direction, a redesign of the support surface
pabeads are homogeneous and rigid macroporous focused to increasing its hydrophobicity was per-
beads with high superficial surface that maintain unal- formed (Scheme 5).
Adv. Synth. Catal. 2015, 357, 2687 – 2696
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2691

FULL PAPERS
Marco Filice et al.
SP-CAL-B-C1-1 showed a 35% yield in 72 h, five
times slower than SP-CAL-B-1 (Table 3).
On the other hand, SP-CAL-B-C8-1 (hydrophobiza-
tion effect introduced, Scheme 5) catalyzed the Heck
reaction achieving >99% yield of product in 27 h,
with the highest TOF value achieved under these con-
ditions with this Pd complex, 153Æ9 hÀ1 (Table 3,
entry 4). When a more hydrophobic chain was intro-
duced, in SP-CAL-B-C12-1, no improvement com-
pared with C8 was observed (Table 3, entry 4). The
new SP-CAL-B-C8-1 artificial metalloenzyme was
reused for two cycles maintaining 70% activity.
In order to demonstrate if this decrease in activity
value is due to the loss of protein stability, IR experi-
ments on the catalyst before and after the Heck reac-
tion were performed (Supporting Information, Fig-
ure S2). The results show that no significant structural
changes on the protein structure occur after the reac-
tion. Therefore, considering that no Pd leaching was
observed, the activity decrease maybe due to a possi-
ble Pd catalyst deactivation.[24]
Scheme 5. Scheme of hydrophobicity modulation of immobi-
lized artificial metalloenzyme by means of chemical engi-
neering of support matrix.
Finally, to expand the general scope of the opti-
mized hybrid catalyst and better assess its stereo- and
The immobilization of the protein on the aldehyde enantioselectivity, we selected an asymmetric Heck
activated support involves a first step of incubation at cross-coupling reaction between halobenzene and 2,3-
pH 10 to generate the imine groups (between alde- dihydrofuran as benchmark reaction (Scheme 6).[28]
hydes of support and amines from the protein) which
must be reduced to irreversible amine groups in the soluble phosphine-based ligands in order to impart
second step (Scheme 3, bottom).
chirality to a catalytic Pd2+ precursor.[28] Hence, as
This reaction has been recently optimized using
Therefore, in order to generate a more hydrophobic first choice, we used the SP-CAL-B-C8-1 catalyst ap-
surface surrounding the enzyme environment, before plying the conditions reported in the literature and
the reducing step, the unreacted aldehyde groups of testing iodo- and bromobenzene. In both cases, no
the support were modified with different alkylamines conversion was observed (Table 4).
(methyl, octyl and myristoyl) or ethylenediamine
Subsequently, we carried out the reactions by ap-
(EDA) as negative control (generation of positive plying our previously optimized conditions (25%
charge). After that, the incorporation of 1 into the
different newly engineered CAL-B immobilized prep-
arations was carried out similar to the previous modi-
fication on the other CAL-B preparations.
This chemical modification of the support caused
an important effect on the catalytic properties of the
final artificial metalloenzyme (Table 3).
The SP-CAL-B-EDA-1 hardly catalyzed the reac-
Scheme 6. Asymmetric Heck cross-coupling reaction be-
tion with <5% of product (Table 3, entry 2), and the tween halobenzene and 2,3-dihydrofuran.
Table 3. Effect of the chemical engineering of catalyst surface on the catalytic properties of SP-CAL-B-1 in the Heck cross-
coupling reaction.[a]
Entry
Catalyst
Surface Modification (R)
Yield [%]
Time [h]
TOF [hÀ1]
1
2
3
4
5
SP-CAL-B-1
–
>99
<5
35
>99
>99
37.5
72
72
27
31
110Æ5
n.d
20Æ1
153Æ9
133Æ7
SP-CAL-B-EDA-1
SP-CAL-B-C-1
SP-CAL-B-C8-1
SP-CAL-B-C12-1
CH2CH2NH2
CH3
(CH2)7CH3
(CH2)11CH3
[a]
Reaction conditions: 0.247 mmol of iodobenzene, 0.55 mmol of ethyl acrylate, 200 mg of each catalyst (0.024 mol% of Pd
relative to iodobenzene), 0.412 mmol of triethylamine, 0.75 mL of DMF and 0.25 mL water and 708C.
2692
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2687 – 2696

FULL PAPERS
Preparation of an Immobilized Lipase-Palladium Artificial Metalloenzyme
Table 4. Asymmetric Heck reaction of different aryl halides catalyzed by SP-CAL-B-C8-1 derivative.[a]
Entry
X
Solvent
Water [%, v/v]
Temp. [8C]
Yield [%]
s ratio[b]
ee [%]
1[c]
2[c]
3
4
5
I
Br
I
Br
I
Br
ethylene glycol
ethylene glycol
DMF
DMF
DMF
0
0
25
25
25
25
80
80
70
70
120
120
0
0
0
0
95
0
nd
nd
nd
nd
18
nd
nd
nd
nd
nd
96.6
nd
6
DMF
[a]
Reaction conditions: 0.247 mmol of halobenzene, 0.55 mmol of 2,3-dihydrofuran, 200 mg of SP-CAL-B-C8-1 catalyst
(0.024 mol% of Pd relative to halobenzene), 0.412 mmol of triethylamine, 1 mL of solvent mixture, 24 h.
Olefinic selectivity.
[b]
[c]
Standard conditions reported in ref.[28]
nas fluorescens lipase (PFL) was from Amano Pharmaceuti-
water in DMF at different temperatures). At 708C, no
conversion was observed in all the cases (Table 4, en-
tries 3 and 4). When the reaction temperature was in-
creased up to 1208C, the reaction in the presence of
iodobenzene showed an almost quantitative conver-
sion and especially a good stereo- (s ratio 18) and
enantioselectivity (>96%) (Table 4, entry 5). Under
these conditions, the bromobenzene was completely
inactive (Table 4, entry 6). These results definitively
confirmed the large and powerful applicability of our
hybrid catalyst.
cal (Japan). Octyl-Sepharose 4BCL and CNBr-activated Se-
pharose 4BCL were from GE Healthcare (Sweden). Sepa-
beadsꢁ epoxide (SP-EC) was kindly gifted by Resindion-
Mitsubishi (Italia). Sepabeads aldehyde support (SP-CHO)
was prepared as previously described with minor modifica-
tions.[28] Lewatit VO OC 1600 was from Lanxess (Germany).
Candida rugosa lipase (CRL), Ryzopus oryzae lipase
(ROL), Rhizomucor miehei lipase (RML), p-nitrophenol
butyrate (pNPB), iodobenzene, ethyl acrylate, ethylenedi-
amine, methylamine, octylamine, dodecylamine, Triton X-
100, cetyltrimethylammonium bromide, sodium borohydride
and sodium periodate were from Sigma-Aldrich. The com-
mercial analytical standards of isomers of ethyl cinnamate
were from Toronto Research Chemicals (Canada) (the Z-cis
isomer) and Sigma Aldrich (the E-trans isomer). Other used
reagents were of analytical grade. The scanning electron mi-
croscopy (SEM) imaging was performed on a TM-1000 (Hi-
tachi) microscope. The spectrophotometric analyses were
run on a V-630 spectrophotometer (JASCO, Japan). HPLC
spectrum P100 (Thermo Separation products) was used.
Analyses were run at 258C using an L-7300 column oven
and a UV6000LP detector. The commercial analytical stan-
dard of 2-phenyl-2,5-dihydrofuran was from Hong Kong
Chemhere Co Limited (China). Column chromatography
was carried out on silica gel (silica gel 60, from Merck, Ger-
many). TLC analysis was performed on Merck silica gel 60
F254. The CAL-B structure analysis was performed using the
PyMOL (DeLano Scientific) software.
Conclusions
In conclusion, we have here presented an effective
strategy to produce a heterogeneous immobilized,
active and reusable artificial metalloenzyme in rela-
tive good amount using a commercial lipase as pro-
tein scaffold. A p-nitrophenyl phosphonate palladium
pincer was selectively inserted by irreversible attach-
ment on the catalytic serine of the lipase. A fine de-
signed solid-phase strategy based on the careful selec-
tion of the supporting matrix together with the proper
orienting immobilization strategy demonstrated that
not only the enzyme scaffold but also the solid-phase
chemistry present a crucial role in order to obtain
a Pd catalyst with excellent activity, stereo- and enan-
tioselectivity and reusability in Heck cross-coupling
reactions. The application of this technology has per-
Enzymatic Activity Assay
The activities of the soluble and immobilized enzyme deriv-
atives were analyzed spectrophotometrically by measuring
the increment in absorbance at 348 nm produced by the re-
lease of p-nitrophenol (pNP) (e=5,150M<M->1 cmÀ1) in the
hydrolysis of 0.4 mM p-nitrophenol butyrate (pNPB) in
25 mM sodium phosphate at pH 7 and 258C. To initialize
the reaction, 0.05–0.2 mL of lipase solution or suspension
was added to 2.5 mL of substrate solution. Enzymatic activi-
ty is given as micromole of hydrolyzed pNPB per minute
per milligram of enzyme (IU) under the conditions de-
scribed above.
À
mitted us to carry out this C C bond reaction under
milder conditions than usual which is of critical im-
portance to industrial applications.
Experimental Section
Materials
Candida antarctica B lipase (CAL-B), Thermomyces lanugi-
nosus lipase (TLL) and Novozym 435 solid catalyst were
generously donated by Novozymes (Denmark). Pseudomo-
Adv. Synth. Catal. 2015, 357, 2687 – 2696
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2693

FULL PAPERS
Marco Filice et al.
Purification of Commercial Lipase Crude Extracts
and Immobilization on Sepharose Supports
Irreversible Inhibition of Lipase Immobilized
Preparations by Diethyl p-Nitrophenylphosphate (D-
pnP)
The purification and immobilization on Sepharose supports
of each commercial lipase here used were carried out as re-
ported elsewhere.[22] Briefly, to 10 mL of 5 mM phosphate
buffer pH 7 containing 10 mg of each commercial lipase
(measured by Bradford assay of crude extract), 1 g of com-
mercial octyl-Sepharose was added. After 3 h, the purified
enzyme derivative was recovered by filtration and washed
with abundant distilled water. The immobilization yield was
almost quantitative in all cases. In order to obtain the solu-
tion of pure soluble lipase to be immobilized on the com-
mercial CNBr-activated Sepharose support, 1 g of hydropho-
bic derivative was added to 10 mL of 25 mM phosphate
buffer solution containing 0.5% (v/v) of Triton X100 deter-
gent [only in the case of TLL, 0.6% (v/v) cetyltrimethylam-
monium bromide (CTAB) was used as detergent]. The re-
sulting desorption mixtures were kept on mechanical stirring
for 1 h and subsequently the supernatant containing the
soluble enzyme solution was recovered by filtration. The
lipase immobilization on CNBr activated Sepharose support
– with an enzyme loading of about 10 mg of pure lipase per
gram of support – was carried out as reported in the litera-
ture.[22]
0.2 g of different immobilized preparations were suspended
in 4 mL of 25 mM sodium phosphate buffer solution at pH 7
and 258C with or without the presence of 0.5% (w/v) of
Triton X-100. Then, 1.5 mM of inhibitor (D-pnP) was added
to this solution. The reaction was maintained until the activ-
ity of the immobilized enzyme – measured using the assay
previously described – was zero or reached a plateau.
Synthesis of Organo-Palladium Complex 1
The synthesis of compound 1 was carried out as described in
the literature.[20]
Modification of Lipase Catalysts by the Organo-
Palladium Complex 1
0.2 g of each lipase catalyst were added to 4 mL of 25 mM
phosphate buffer pH 7. After that, 20 mL of 300 mM aceto-
nitrile solution of 1 (final concentration 1.5 mM) were
added and the resulting mixture was maintained under
gentle stirring, periodically checking the enzyme activity de-
crease by the lipase activity assay previously described. The
modification yields were calculated as follow: catalyst modi-
fication (%) = (Ar/Ai)”100 where Ar is the catalyst hydro-
lytic activity after the modification reaction and Ai the cata-
lyst hydrolytic activity before the modification reaction. The
insertion of the organometallic 1 into the lipase cavity was
analyzed by MADI-TOF mass spectrometry, using native
CAL-B as calibration material (Supporting Information,
Figure S1). A peak of 35418.53 was found corresponding to
a lower mass than the calculated value ([MCAL-B +1=
35616.32]. This difference corresponds to the release of the
bromine and palladium ions during the measurement in
MALDI mass spectra.
Immobilization on Sepabeadsꢀ Aldehyde Supports
One gram of the aldehyde support (SP-CHO)[29] was added
to 10 mL of an enzymatic solution of purified CAL-B
(1 mgmLÀ1) prepared in 100 mM sodium bicarbonate buffer
pH 10.1 and the resulting immobilization mixture was main-
tained under gently stirring for 24 h. In order to reduce the
imino and aldehyde groups and stop the immobilization re-
action, 10 mg sodium borohydride were added under gentle
stirring. After 30 min, the immobilized enzyme was recov-
ered by filtration and washed with abundant distilled water.
Catalyst Engineering by Alkylamine Modification
In the cases where the modification yields were not quan-
titative by enzymatic assay measurements, in order to avoid
any possible unwanted hydrolytic side reaction of ethyl
esters present in the Heck reaction, the residual enzymatic
activity was inhibited as previously described.
One gram of the aldehyde support (SP-CHO) was added to
10 mL of an enzymatic solution of purified CAL-B
(1 mgmLÀ1) prepared in 100 mM sodium bicarbonate buffer
pH 10.1 and the resulting immobilization mixture was main-
tained under gently stirring for 24 h. After that, the enzy-
matic derivative was recovered by filtration and, without
any washing, 10 mL of a 100 mM sodium bicarbonate buffer
pH 10.1 solution containing 500 mM of methylamine, ethyle-
nediamine, octylamine or dodecylamine was directly added.
In the last two cases, 20% and 40% (v/v) dioxane was
added, respectively, in order to ensure the full solubility of
the modifier compounds. Hence, the resulting mixture was
maintained under gentle mechanical stirring for 30 min.
After that, 20 mg of NaBH4 were added and the suspension
was maintained under gentle mechanical stirring for 30 min.
Subsequently, the engineered immobilized enzyme was re-
covered by filtration and washed with distilled water, 50%
(v/v) dioxane aqueous solution and finally abundant distilled
water.
Desorption Study of Novozym 435 Derivative by
SDS-PAGE Analysis
200 mg of Novozym 435-1 modified derivative were incubat-
ed in 1 mL of dimethylformamide (DMF) or an aqueous so-
lution thereof, in order to assess the stability of the catalyst
À
under the reaction conditions expected by the Heck C C
coupling. After different times, the enzyme derivatives were
recovered by filtration and analyzed by SDS-PAGE (gel
polymerization: 12%).
General Procedure for Heck Cross-Coupling
Reaction
In a 1.5-mL screw-cap sealed vessel, 200 mg of each catalyst
were added to
a
solution containing iodobenzene
(0.0306 mL, 0.274 mmol) and ethyl acrylate (0.059 mL,
0.55 mmol) in DMF or DMF/distilled water (final volume
1 mL). The mixture was preheated at 708C under magnetic
2694
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2687 – 2696

FULL PAPERS
Preparation of an Immobilized Lipase-Palladium Artificial Metalloenzyme
stirring for 5 min. After that, to initialize the reaction, trie-
thylamine (0.057 mL, 0.412 mmol) was added. The final sus-
pension was left under magnetic stirring at 708C for the in-
dicated times.
Acknowledgements
This work has been sponsored by CSIC. M.F. thanks CSIC
for a JAE-Doc contract (“Junta para la Ampliacion de estu-
dios”) cofounded by ESF (European Social Fund). O.
Romero is grateful to CONICYT and Programa Bicentenario
Becas-Chile for financial support The authors also thank Dr.
Ramiro Martinez from Novozymes for the generous gift of li-
pases and Dr. Daminatti from Resindion for the gift of the
Sepabeads resin.
HPLC Monitoring of Heck Cross-Coupling Reaction
The reaction outcome was monitored by HPLC analysis of
samples of the reaction withdrawn at different times. The
analysis condition determinations were performed with
a Kromasil-C8 (150”4.6 mm and 5 mm Ø), at a flow of
1 mLminÀ1; l=267 nm; and a mobile phase: 50% (v/v)
ACN in MilliQ water. The E configuration of achieved
product was confirmed by HPLC identity using the commer-
cial standards (provided as described in Materials) the Rts
being 8.4 min for Z-cis isomer and 9.4 min for E-trans
isomer. The yields were obtained by extrapolating the
values through a calibration curve of the E-products (R2 =
0.9975).
To confirm the suitability of the HPLC method, the final
reaction mixture was purified by silica gel chromatography.
In this case, as well as to perform the recyclability studies,
the reaction amounts were scaled up 5 times, maintaining
the reactant proportions described above. Hence, the reac-
tion performed under the best conditions [DMF in the pres-
ence of 25% (v/v) of distilled water], once it reached its
maximum conversion, was extracted with Et2O. The com-
bined filtrate was dried over anhydrous NaSO4, concentrat-
ed under reduced pressure and finally purified by silica-gel
chromatography (hexane). The purified yields were in
agreement with those obtained by HPLC (215.4 mg of pure
E-ethyl cinnamate, 99% yield).
References
[1] a) M. Durrenberger, T. R. Ward, Curr. Opin. Chem.
Biol. 2014, 19, 99–106; b) M. Genz, V. KÅhler, M.
Krauss, D. Singer, R. Hoffmann, T. R. Ward, N. Sträter,
ChemCatChem 2014, 6, 736–740.
[2] J. C. Lewis, ACS Catal. 2013, 3, 2954–2975.
[3] Y. Lu, N. Yeung, N. Sieracki, N. M. Marshall, Nature
2009, 460, 855–862.
[4] J. Bos, G. Roelfes, Curr. Opin. Chem. Biol. 2014, 19,
135–143.
[5] F. W. Monnard, E. S. Nogueira, T. Heinisch, T. Schirm-
er, T. R. Ward, Chem. Sci. 2013, 4, 3269–3274.
[6] M. R. Ringenberg, T. R. Ward, Chem. Commun. 2011,
47, 8470–8476.
[7] C. A. Denard, J. F. Hartwig, H. Zhao, ACS Catal. 2013,
3, 2856–2864.
[8] M. T. Reetz, Chem. Rec. 2012, 12, 391–406.
[9] T. Reiner, D. Jantke, A. N. Marziale, A. Raba, J. Ep-
pinger, Chem. Open 2013, 2, 50–54.
[10] C. M. Thomas, T. R. Ward, Chem. Soc. Rev. 2005, 34,
337–346, and references cited therein.
Asymmetric Heck Cross-Coupling Reaction
[11] a) T. Ueno, H. Tabe, Y. Tanaka, Chem. Asian J. 2013, 8,
1646–1660; b) H. Inaba, S. Kanamaru, F. Arisaka, S. Ki-
tagawaa, T. Ueno, Dalton Trans. 2012, 41, 11424–11427.
[12] D. K. Garner, L. Liang, D. A. Barrios, J.-L. Zhang, Y.
Lu, ACS Catal. 2011, 1, 1083–1089.
In a 1.5 mL screw-cap sealed vessel, 200 mg of SP-CAL-B-
C8-1 catalyst were added to a solution containing iodoben-
zene (0.0306 mL, 0.274 mmol) and 2,3-dihydrofuran
(0.041 mL, 0.55 mmol) in 1 mL of DMF/distilled water 75/25
solution. The mixture was preheated at 1208C under mag-
netic stirring for 5 min. After that, to initialize the reaction,
triethylamine (0.057 mL, 0.412 mmol) was added. The final
suspension was left under magnetic stirring at 1208C for the
indicated times.
[13] I. D. Petrik, J. Liu, Y. Lu, Curr. Opin. Chem. Biol.
2014, 19, 67–75.
[14] H. Yang, P. Srivastava, C. Zhang, J. C. Lewis, ChemBio-
Chem 2014, 15, 223–227.
[15] a) P. Adlercreutz, Chem. Soc. Rev. 2013, 42, 6406–6436;
b) J. Brabcova, M. Filice, M. Gutarra, J. M. Palomo,
Curr. Biol. Compd. 2013, 9, 113–136.
HPLC Monitoring of Asymmetric Heck Cross-
Coupling Reaction
[16] a) M. Filice, M. Marciello, M. d. P. Morales, J. M.
Palomo, Chem. Commun. 2013, 49, 6876–6878; b) G.
Van Koten, J. Organomet. Chem. 2013, 730, 156–164.
[17] a) M. Filice, J. M. Palomo, ACS Catal. 2014, 4, 1588–
1598; b) K. Engstrom, E. V. Johnston, O. Verho, K. P.
Gustafson, M. Shakeri, C. W. Tai, J. E. Bäckvall,
Angew. Chem. Int. Ed. 2013, 52, 14006–14010.
[18] M. Filice, O. Romero, J. Gutierrez-Fernandez, B. de
Las Rivas, J. A. Hermoso, J. M. Palomo, Chem.
Commun. 2015, 51, 9324–9327.
The reaction outcome was monitored by HPLC analysis of
samples of the reaction withdrawn after 24 h. The analyses
were performed with a Kromasil-C4 (150”4.6 mm and 5 mm
Ø), at a flow of 1 mLminÀ1; l=254 nm; and a mobile
phase: 50% (v/v) ACN in MilliQ water. The formation of
the product was confirmed by HPLC identity using the com-
mercial standards provided as described in Materials. The
yields were obtained by extrapolating the values through
a calibration curve (R2 =0.9912). To analyze the enantiose-
lectivity, the final reaction mixture was purified by silica gel
chromatography with pentane/Et2O (10:1 to 5:1) as a color-
less oil. The analysis conditions were: a Daicel Chiralcel OJ-
H, 98:2 n-hexane/isopropyl alcohol, flow rate=0.5 mLminÀ1
in agreement with the conditions reported in the litera-
ture.[28]
[19] J. C. Powers, J. L. Asgian, W. D. Ekici, K. E. James,
Chem. Rev. 2002, 102, 4639–4750.
[20] C. A. Kruithof, M. A. Casado, G. Guillena, M. R.
Egmond, A. van der Kerk-van Hoof, A. J. R. Heck,
R. J. M. Klein Gebbink, G. van Koten, Chem. Eur. J.
2005, 11, 6869–6877.
Adv. Synth. Catal. 2015, 357, 2687 – 2696
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2695

FULL PAPERS
Marco Filice et al.
[21] P. J. H. Scott, in: Linker Strategies in Solid-Phase Or-
ganic Synthesis, John Wiley & Sons, Ltd., Chichester.
2009.
[25] D. E. Bergbreiter, P. L. Osburn, J. D. Frels, Adv. Synth.
Catal. 2005, 347, 172–184.
[26] K. Yu, W. Sommer, J. M. Richardson, M. Weck, C. W.
[22] a) A. Bastida, P. Sabuquillo, P. Armisen, R. Fernandez-
Lafuente, J. Huguet, J. M. Guisan, Biotechnol. Bioeng.
1998, 58, 486–493; b) M. Filice, JM. Guisan, M. Terreni,
J. M. Palomo, Nat. Protoc. 2012, 7, 1783–1796.
[23] J. D. Schrag, Y. Li, M. Cygler, D. Lang, T. Burgdorf,
H. J. Hecht, R. Schmid, D. Schomburg, T. J. Rydel, J. D.
Oliver, L. C. Strickland, C. M. Dunaway, S. B. Larson,
J. Day, A. McPherson, Structure 1997, 5, 187–202.
[24] N. C. Mehendale, J. R. A. Sietsma, K. P. de Jong, C. A.
van Walree, R. J. M. Klein Gebbink, G. van Koten,
Adv. Synth. Catal. 2007, 349, 2619–2630.
Jones, Adv. Synth. Catal. 2005, 347, 161–171.
[27] C. Mateo, J. M. Palomo, M. Fuentes, L. Betancor, V.
Grazu, F. Lopez-Gallego, B. C. C. Pessela, A. Hidalgo,
G. Fernandez-Lorente, R. Fernandez-Lafuente, J. M.
Guisan, Enzyme Microb. Technol. 2006, 39, 274–280.
[28] C. Wu, J. Zhou, J. Am. Chem. Soc. 2014, 136, 650–652.
[29] J. M. Guisan, Enzyme Microb. Technol. 1988, 10, 375–
382.
2696
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2687 – 2696