A general protein purification and immobilization method on controlled porosity glass: Biocatalytic applications

Article abstract of DOI:10.1039/c4cc02605e

A general combined purification and immobilization method to facilitate biocatalytic process development is presented. The support material, EziG, is based on controlled porosity glass (CPG) or polymer-coated versions thereof (HybCPG) and binds protein affinity tags. Biocatalytic reactions in aqueous and organic media with seven enzymes of biocatalytic interest are shown. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4cc02605e

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
A general protein purification and immobilization
method on controlled porosity glass: biocatalytic
applications†
Cite this: Chem. Commun., 2014,
50, 9134
Received 8th April 2014,
Accepted 20th June 2014
K. Engelmark Cassimjee,*a M. Kadow,a Y. Wikmark,a M. Svedendahl Humble,a
M. L. Rothstein,b D. M. Rothsteinb and J.-E. Backvall*a
¨
DOI: 10.1039/c4cc02605e
www.rsc.org/chemcomm
A
general combined purification and immobilization method environments at a pH below 10. The nature of the material on
to facilitate biocatalytic process development is presented. The support which an enzyme is immobilized is important for process
material, EziGt, is based on controlled porosity glass (CPG) or polymer- efficacy and reusability.9,10 CPG is a suitable material for this
coated versions thereof (HybCPG) and binds protein affinity tags. purpose.10,11 Its average pore diameter can be varied from 10 to
Biocatalytic reactions in aqueous and organic media with seven 300 nm with a maintained sharp pore size distribution. This
enzymes of biocatalytic interest are shown.
provides a favorable micro-environment for enzymes, where the
interconnecting porous structure facilitates the mass transfer of
The benefits of immobilizing an enzyme on a carrier before reactants and products throughout the material. Enzymes and
its use in biocatalytic reactions include increased reusability, other proteins have been immobilized into CPG10,12–15 using
stabilization and facile separation from the reaction mixture.1 glutaraldehyde as a linker to aminated CPG. This linker was also
The immobilization of an enzyme in a biocatalytic reaction is used to bind enzymes to aminopropyl-mesocellular foam.16,17
also a requirement for high process efficacy.2–4
The recently developed HybCPG8 may offer additional benefits
Various enzyme immobilization techniques have been published. as an immobilized enzyme carrier, but has not been explored
Porous materials are often used as enzyme carriers4 for biocatalytic earlier for this purpose. In HybCPG the otherwise hydrophilic
reactions in organic solvent, e.g. macroporous polypropylene porous surface is coated with an organic polymer, e.g. homo-
polymer beads (Accurels),5 and polymethacrylate.2 Furthermore, geneous polystyrene for a hydrophobic surface, homogeneous
enzymes can be coupled to one another to form cross-linked polyacrylonitrile for a hydrophilic surface, or co-polymeric
enzyme aggregates (CLEAs), consisting of almost only enzyme design for blended characteristics (see ESI† for more details).
molecules, that can be used as a heterogeneous catalyst.2,6 The
The aim of the present study was to construct a general
existing enzyme immobilization techniques are effective but not method for purification and immobilization of enzymes for
general. Testing and optimization of a number of methods are biotransformations based on modified CPGs and HybCPGs,
usually required before a suitable preparation method is found, which can bind protein affinity tags, named EziGt. The method
and pre-purification is necessary to achieve a pure enzyme catalyst. was explored using seven enzymes: one o-transaminase, three
A general method that allows purification and immobilization lipases, and a Baeyer–Villiger monooxygenase (BVMO) together
of enzymes in one step, providing biocatalysts for organic with two cofactor-regenerating enzymes.
transformations, is highly desirable.
Three types of EziGt were produced by the use of one type of
Glass, a material inert in nature, can be produced as long chain aminoalkyl CPG (LCAA CPG), a material commonly
particles of size-controlled macro- or meso-pores, CPG, and as used for solid phase oligonucleotide synthesis,18 and two types of
a hybrid inorganic–organic material thereof, HybCPG.7,8 These amino HybCPG. The HybCPG types employed were custom-made
materials are suited for various applications in fluid processes, products; one coated with a hydrophobic polymer including
since they are stable in most organic solvents and aqueous vinylbenzyl chloride units (HybCPG VBC) and another with a
hydrophilic coating consisting of a copolymer of styrene and
acrylonitrile (HybCPG copo). In these tests a pore diameter of
50 to 60 nm was employed.
The amino CPGs were treated with 2,4-dihydroxyaceto-
phenone, and the imines formed were reduced followed by
a Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University,
SE-106 91 Stockholm, Sweden. E-mail: karim@organ.su.se, jeb@organ.su.se
b Prime Synthesis, Inc., 2 New Road Suite 126, Aston, PA 19014, USA
† Electronic supplementary information (ESI) available: Additional information
Co(II) complexation (see Scheme S1, ESI†). The EziGt could
thereafter directly be used as a binding matrix for His6-tagged
regarding the enzymes used in this study, the carrier material and the biocatalytic
example reactions, and materials and methods. See DOI: 10.1039/c4cc02605e
9134 | Chem. Commun., 2014, 50, 9134--9137
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
Table 1 Immobilization yield, enzyme loading, enantiomeric excess of
the product (eep) and the initial rate for o-TA-EziGt prepared from
different types of CPG, for the reaction depicted in Scheme 1
Yield of
immobilization
(% active enzyme) (% w/w)
Active enzyme
loading
Initial rate
eep
(mmol minÀ1
CPG type
(%, GC) gÀ1 EziGt)
LCAA CPG
HybCPG VBC 499
HybCPG copo 499
499
24
29
21
499 (S) 0.38
499 (S) 0.70
499 (S) 0.06
o-TA-EziGt(HybCPG VBC) was the most active preparation
under the conditions tested, plausibly due to the hydrophobic
nature of the surface of the carrier, which may prevent
denaturation. However, the reason for the observed differences
in the reaction rate with the tested carriers remains to be fully
investigated. Our results demonstrate that the enzyme func-
tions in MTBE in a pure form on a carrier with a measureable
active enzyme amount.
Fig. 1 EziGt with chelated Co(II) and bound His6-enzyme. The structure
of the R-group is not revealed, and varies between the CPG products.
Candida antarctica lipase B (CalB)24,25 was immobilized on
EziGt from the cell culture supernatant after expression and
secretion by Pichia pastoris. For comparison immobilization was
also done on Accurels, where 8 h of incubation was required. EziGt
immobilization was achieved within 30 min. The immobilized CalB
preparations CalB-EziGt and CalB-Accurels were used in the kinetic
resolution of 1-phenylethanol (5) with vinylbutyrate (6) as the acyl
donor (Scheme 2). The enantiomeric excess (ee) of (R)-1-phenylethyl
butyrate ((R)-7) was determined by chiral GC. The ee, enantio-
selectivity (E), and the activities are given in Table 2. An excellent
enantioselectivity (E 4 300) was obtained for both CalB-EziGt
and CalB-Accurels.
enzymes (Fig. 1), the HybCPG-based preparations showed higher
catalytic activity. Immobilization was performed using the cell
lysate in the case of intracellular expression or the cell culture
supernatant for extracellular expression, yielding a pure hetero-
geneous enzyme catalyst. Extraction of purified enzyme could be
performed by the standard immobilized metal affinity chroma-
tography (IMAC)-technique.
Chromobacterium violaceum o-transaminase (o-TA)19–21 was
immobilized on EziGt by immersing the carrier material in the cell
lysate with the overexpressed enzyme; binding was completed
within 20 min on an orbital shaker. The yield of immobilization
was measured by active site quantification22 to ensure that the
deactivated enzyme was not included in the measurement, by
measuring the active sites in the cell lysate before and after
immobilization and comparing that difference with the active sites
measured on the obtained o-TA-EziGt. More than 99% of the
extracted enzyme was active; the enzyme mass corresponding to the
found active sites on the carriers yielded an enzyme mass loading of
more than 20%. o-TA-EziGt was successfully used as a transamina-
tion catalyst in a model reaction in methyl tert-butyl ether (MTBE)
(see Scheme 1) in which the lyophilized cell lysate has previously
been reported to be active;23 see Table 1 and Table S1 (ESI†). In
this model reaction transamination of 1-phenoxypropan-2-one
(2) with 1-phenylethylamine (1) afforded enantiomerically pure
(S)-1-phenoxypropan-2-amine ((S)-4, 499% ee) and acetophenone
(3). The lyophilized cell lysate from our preparation did not display
any catalytic activity under the tested reaction conditions.
The activity of CalB-Accurels is approximately twice that of
CalB-EziGt per bulk volumetric unit (volume occupied by the
catalyst including the pore volume) when calculated on the dry
preparations, as shown in Table 2, and approximately six times
higher per mass unit since the EziGt is a more dense material,
see Table S2 (ESI†). In the reaction mixture the Accurels swells
to approximately double the volume while the EziGt is virtually
unchanged; therefore the catalysts have approximately equal
catalytic activity per bulk volume in situ (data not shown).
The CalB variant Trp104Ala has previously been shown to have
higher activity and reversed enantiopreference for bulky secondary
alcohols compared to the wild-type enzyme.26–28 This enzyme was
readily bound to EziGt within 30 min of incubation while ten
days of incubation was required for immobilization on Accurels.
For this variant, EziGt (HybCPG VBC) is a more suitable carrier
than Accurels based on the bulk volumetric activity and the time
needed for immobilization (see Table S2, ESI†).
Scheme 2 Model reaction for immobilized lipases; the kinetic resolution
of 1-phenylethylamine 5 (10 mM) by the transesterification with vinyl-
butyrate 6 (100 mM).
Scheme 1 Chosen model reaction for o-TA-EziGt, where 1-phenylethyl-
amine 1 (150 mM) was used as the amino donor in the transamination of
1-phenoxypropan-2-one 2 (50 mM) in methyl tert-butyl ether (MTBE).
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 9134--9137 | 9135

View Article Online
ChemComm
Communication
Table 2 Incubation time required for immobilization, E values, ee, and
bulk volumetric activity for a chosen model reaction (Scheme 2, 22 1C), for
four preparations of immobilized CalB
re-addition of O2 and reaction for another 3 h increased the
yield of 10 to 88%. An analogous experiment with dissolved
enzymes showed conversion but could not be restarted by
addition of O2, presumably due to deactivation of the unstable
2,5-DKCMO. These results show that EziGt immobilization
had a stabilizing effect on one or several of the enzymes, and
is effective for the co-immobilization of several His6-tagged
enzymes from the same mixture for cascade reactions.
In a preparative experiment a multi-phase reaction system was
used for the oxidation of (+)-camphor (9) to lactone 10 using 9 in
cyclohexane (100 mM) together with the aqueous reaction mixture
described above (Fig. S1, ESI†). The EziGt-preparation remained
in the aqueous phase while mild orbital shaking was applied due
to its density. The immobilized enzymes were thereby protected
from deactivation through contact with the solvent interface.
The three-phase (cyclohexane, aqueous phosphate buffer, and
solid EziGt) reaction system showed 56% conversion to lactone
10 after 72 h. The analogous experiment with dissolved enzymes
in a two-phase environment gave virtually no conversion within
the same time period.
Schulz et al. have reported a successful enzymatic cascade
BVMO-reaction with a single-component BVMO,36 which does
not require a distinct flavin reductase for its supply of FADH2.
The method described herein for co-immobilization of a three-
enzyme-based cascade reaction system opens up new possibi-
lities for applying the more demanding two-component BVMOs
in biocatalytic transformations.
EziGt was used for combined protein purification and
immobilization of seven enzymes from the cell lysate or the cell
culture supernatant, which were then employed in biocatalytic
reactions. o-TA-EziGt was active in MTBE. CalA- and CalB-EziGt
were active transesterification catalysts and could be obtained in a
significantly shorter time compared to the corresponding Accurels-
based preparations. Co-immobilization of 2,5-DKCMO, FRE
and AlaDH on the same EziGt-carrier led to increased stability
in a two-phase system resulting in a functioning enzymatic
cascade reaction. EziGt based on HybCPG was shown to be the
most suitable carrier for all enzymes tested; HybCPG enables
tailoring of the surface on which the enzyme is bound.
Financial support from the Swedish Research Council and
the Knut and Alice Wallenberg Foundation is gratefully
acknowledged. We thank Prof. Wolgang Kroutil (Univ. of Graz)
for the gene of AlaDH.
Incubation
time for
immobili-
zation (h)
Initial rate
ee of
per bulk volumec
(mmol minÀ1
(R)-7b
Carriera
E
(%, GC) mlÀ1 carrier)
EziGt
(LCAA CPG)
EziGt
(HybCPG VBC)
EziGt
0.5
0.5
0.5
8
4300 (R) 499.5
4300 (R) 499.5
4300 (R) 499.5
0.23
0.21
0.26
(HybCPG copo)
Accurels
4300 (R) 499.5
o0.60d
a
b
See Table S1 (ESI) for EziGt data. Calculated at 49% conversion.
c
d
Initial rate of consumption of 5. Without inclusion of swelling
which occurs in contact with solvent.
Candida antarctica lipase A (CalA),29–31 a lipase with a wider
substrate range and lower enantiospecificity compared to CalB,
was immobilized on EziGt (LCAA CPG) and applied for the
chosen model reaction (Scheme 2). As expected the E value was
low (1.3), but with a higher activity than that of CalB (Table S2,
ESI†). The results show that EziGt is suitable for immobiliza-
tion of CalA and further optimizations with subsequent appli-
cations are ongoing.
The FMNH2-dependent BVMO 2,5-diketocamphane monooxy-
genase from Pseudomonas putida (2,5-DKCMO)32 was co-immobilized
on EziGt with two cofactor-reconverting enzymes (flavin reductase
(FRE) from E. coli33,34 for FMNH2 recycling and alanine dehydrogenase
(AlaDH) from B. subtilis35 for NADH recycling). This immobilization
was performed by exposing the EziGt carrier to a mixture of all three
cell lysates after overexpression of the three enzymes. The three-
enzyme EziGt was then used for an enzymatic cascade reaction for
Baeyer–Villiger oxygenation of (+)-camphor (9) (2.0 mM) in phos-
phate buffer, with FMNH2 and NADH regeneration by consumption
of L-alanine (Scheme 3). The reaction was started by addition of
molecular oxygen followed by sealing of the reaction vessel and
subsequent orbital shaking. Preparation with EziGt based on LCAA
CPG displayed virtually no conversion, whereas the HybCPG VBC
preparation gave 63% of lactone 10 within 3 h. The latter vessel was
left for 24 h without any improvement in conversion; however,
Notes and references
1 U. T. Bornscheuer, Angew. Chem., Int. Ed., 2003, 42, 3336–3337.
2 M. D. Truppo, H. Strotman and G. Hughes, ChemCatChem, 2012, 4,
1071–1074.
3 R. A. Sheldon, Adv. Synth. Catal., 2007, 349, 1289–1307.
4 M. Hartmann and D. Jung, J. Mater. Chem., 2010, 20, 844–857.
5 J. A. Bosley and A. D. Peilow, J. Am. Oil Chem. Soc., 1997, 74, 107–111.
6 R. A. Sheldon, R. Schoevaart and L. M. van Langen, Appl. Microbiol.
Biotechnol., 2011, 92, 467–477.
7 F. Hoffmann, M. Cornelius, J. Morell and M. Froba, Angew. Chem.,
Int. Ed., 2006, 45, 3216–3251.
8 M. L. Rothstein, D. M. Rothstein and D. P. Lee, US Pat., 12 145 051,
2008.
9 A. Liese and L. Hilterhaus, Chem. Soc. Rev., 2013, 42, 6236–6249.
10 M. Hartmann and X. Kostrov, Chem. Soc. Rev., 2013, 42, 6277–6289.
Scheme 3 Enzymatic cascade reaction with EziGt co-immobilized
single-component BVMO (2,5-DKCMO), FRE and AlaDH for oxidizing
(+)-camphor 9 to (1R)-1,8,8-trimethyl-2-oxabicyclo[3.2.1]octan-3-one 10,
L-alanine conversion to pyruvate is used for recycling the cofactors NAD+
and FMNH2.
9136 | Chem. Commun., 2014, 50, 9134--9137
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
11 D. Enke, F. Janowski and W. Schwieger, Microporous Mesoporous 25 E. M. Anderson, K. M. Larsson and O. Kirk, Biocatal. Biotransform.,
Mater., 2003, 60, 19–30.
1998, 16, 181–204.
¨
´
¨
12 D. L. Eaton, Med. Res. Eng., 1976, 12, 17–22.
13 F. Janowski, G. Fischer, W. Urbaniak, Z. Foltynowicz and
B. Marciniec, J. Chem. Technol. Biotechnol., 1991, 51, 263–272.
14 J. M. Bolivar and B. Nidetzky, Biotechnol. Bioeng., 2012, 109,
1490–1498.
26 K. Engstrom, M. Vallin, P. O. Syren, K. Hult and J. E. Backvall, Org.
Biomol. Chem., 2011, 9, 81–82.
27 A. O. Magnusson, M. Takwa, A. Hamberg and K. Hult, Angew. Chem.,
Int. Ed., 2005, 44, 4582–4585.
28 A. O. Magnusson, J. C. Rotticci-Mulder, A. Santagostino and K. Hult,
ChemBioChem, 2005, 6, 1051–1056.
15 H. H. Weetall, Appl. Biochem. Biotechnol., 1993, 41, 157–188.
¨
16 K. Engstrom, E. V. Johnston, O. Verho, K. P. J. Gustafson, M. Shakeri, C.-W. 29 R. Kourist and U. T. Bornscheuer, Appl. Microbiol. Biotechnol., 2011,
¨
Tai and J.-E. Backvall, Angew. Chem., Int. Ed., 2013, 52, 14006–14010.
91, 505–517.
¨
¨
¨
17 M. Shakeri, K. Engstrom, A. G. Sandstrom and J.-E. Backvall, 30 S. H. Krishna, M. Persson and U. T. Bornscheuer, Tetrahedron:
ChemCatChem, 2010, 2, 534–538.
Asymmetry, 2002, 13, 2693–2696.
31 M. Martinelle, M. Holmquist and K. Hult, Biochim. Biophys. Acta,
1995, 1258, 272–276.
32 R. Gagnon, G. Grogan, E. Groussain, S. Pedragosa-Moreau, P. F.
Richardson, S. M. Roberts, A. J. Willetts, V. Alphand, J. Lebreton and
R. Furstoss, J. Chem. Soc., Perkin Trans. 1, 1995, 2527–2528.
18 R. T. Pon, Methods Mol. Biol., 1993, 20, 465–496.
19 U. Kaulmann, K. Smithies, M. E. B. Smith, H. C. Hailes and
J. M. Ward, Enzyme Microb. Technol., 2007, 41, 628–637.
20 C. Sayer, M. N. Isupov and J. A. Littlechild, Acta Crystallogr., Sect. F:
Struct. Biol. Cryst. Commun., 2007, 63, 117–119.
21 M. S. Humble, K. E. Cassimjee, M. Håkansson, Y. R. Kimbung, 33 M. Fontecave, R. Eliasson and P. Reichard, J. Biol. Chem., 1987, 262,
B. Walse, V. Abedi, H.-J. Federsel, P. Berglund and D. T. Logan, FEBS
J., 2012, 279, 779–792.
22 K. E. Cassimjee, M. S. Humble, V. Miceli, C. G. Colomina and
P. Berglund, ACS Catal., 2011, 1, 1051–1055.
23 F. G. Mutti and W. Kroutil, Adv. Synth. Catal., 2012, 354, 3409–3413.
24 J. Uppenberg, N. Oehrner, M. Norin, K. Hult, G. J. Kleywegt,
12325–12331.
¨
34 M. Kadow, K. Balke, A. Willetts, U. Bornscheuer and J. E. Backvall,
Appl. Microbiol. Biotechnol., 2014, 98, 3975–3986.
35 F. G. Mutti, C. S. Fuchs, D. Pressnitz, N. G. Turrini, J. H. Sattler,
A. Lerchner, A. Skerra and W. Kroutil, Eur. J. Org. Chem., 2012,
1003–1007.
S. Patkar, V. Waagen, T. Anthonsen and T. A. Jones, Biochemistry, 36 F. Schulz, F. Leca, F. Hollmann and M. T. Reetz, Beilstein J. Org.
1995, 34, 16838–16851.
Chem., 2005, 1, 10.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 9134--9137 | 9137