Yihx-encoded haloacid dehalogenase-like phosphatase HAD4 from Escherichia coli is a specific α-d-glucose 1-phosphate hydrolase useful for substrate-selective sugar phosphate transformations

Article abstract of DOI:10.1016/j.molcatb.2014.09.004

Phosphomonoester hydrolases (phosphatases; EC 3.1.3.) often exhibit extremely relaxed substratespecificity which limits their application to substrate-selective biotransformations. In search of a phos-phatase catalyst specific for hydrolyzing α-d-glucose 1-phosphate (αGlc 1-P), we selected haloaciddehalogenase-like phosphatase 4 (HAD4) from Escherichia coli and obtained highly active recombinantenzyme through a fusion protein (Zbasic₂HAD4) that contained Zbasic₂, a strongly positively chargedthree α-helical bundle module, at its N-terminus. Highly pure Zbasic₂HAD4 was prepared directly fromE. coli cell extract using capture and polishing combined in a single step of cation exchange chro-matography. Kinetic studies showed Zbasic₂HAD4 to exhibit 565-fold preference for hydrolyzing αGlc 1-P (k_cat/K_M= 1.87 ± 0.03 mM^?1s^?1; 37°C, pH 7.0) as compared to d-glucose 6-phosphate (Glc 6-P). Alsoamong other sugar phosphates, αGlc 1-P was clearly preferred. Using different mixtures of αGlc 1-P and Glc 6-P (e.g. 180 mM each) as the substrate, Zbasic₂HAD4 could be used to selectively convert the αGlc1-P present, leaving back all of the Glc 6-P for recovery. Zbasic₂HAD4 was immobilized conveniently usingdirect loading of E. coli cell extract on sulfonic acid group-containing porous carriers, yielding a recyclableheterogeneous biocatalyst that was nearly as effective as the soluble enzyme, probably because proteinattachment to the anionic surface occurred in a preferred orientation via the cationic Zbasic₂module.Selective removal of αGlc 1-P from sugar phosphate preparations could be an interesting application ofZbasic₂HAD4 for which readily available broad-spectrum phosphatases are unsuitable.

Full text of DOI:10.1016/j.molcatb.2014.09.004

Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Yihx-encoded haloacid dehalogenase-like phosphatase HAD4 from
Escherichia coli is a specific ␣-d-glucose 1-phosphate hydrolase useful
for substrate-selective sugar phosphate transformations
Martin Pfeiffera, Patricia Wildbergera, Bernd Nidetzkya,b,∗
a Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, Petersgasse 12/1, A-8010 Graz, Austria
b ACIB – Austrian Centre of Industrial Biotechnology, Petersgasse 14, A-8010 Graz, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 April 2014
Received in revised form 9 September 2014
Accepted 11 September 2014
Available online 20 September 2014
Phosphomonoester hydrolases (phosphatases; EC 3.1.3.) often exhibit extremely relaxed substrate
specificity which limits their application to substrate-selective biotransformations. In search of a phos-
phatase catalyst specific for hydrolyzing ␣-d-glucose 1-phosphate (␣Glc 1-P), we selected haloacid
dehalogenase-like phosphatase 4 (HAD4) from Escherichia coli and obtained highly active recombinant
enzyme through a fusion protein (Zbasic2 HAD4) that contained Zbasic2, a strongly positively charged
three ␣-helical bundle module, at its N-terminus. Highly pure Zbasic2 HAD4 was prepared directly from
E. coli cell extract using capture and polishing combined in a single step of cation exchange chro-
matography. Kinetic studies showed Zbasic2 HAD4 to exhibit 565-fold preference for hydrolyzing ␣Glc
1-P (kcat/KM = 1.87 0.03 mM−1 s−1; 37 ◦C, pH 7.0) as compared to d-glucose 6-phosphate (Glc 6-P). Also
among other sugar phosphates, ␣Glc 1-P was clearly preferred. Using different mixtures of ␣Glc 1-P and
Glc 6-P (e.g. 180 mM each) as the substrate, Zbasic2 HAD4 could be used to selectively convert the ␣Glc
1-P present, leaving back all of the Glc 6-P for recovery. Zbasic2 HAD4 was immobilized conveniently using
direct loading of E. coli cell extract on sulfonic acid group-containing porous carriers, yielding a recyclable
heterogeneous biocatalyst that was nearly as effective as the soluble enzyme, probably because protein
attachment to the anionic surface occurred in a preferred orientation via the cationic Zbasic2 module.
Selective removal of ␣Glc 1-P from sugar phosphate preparations could be an interesting application of
Zbasic2 HAD4 for which readily available broad-spectrum phosphatases are unsuitable.
Keywords:
Phosphatase
HAD superfamily
Sugar phosphate
Substrate selectivity
Oriented immobilization
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
after the activity of its first structurally characterized member (l-2-
HAD from Pseudomonas sp. YL), the HAD-like protein superfamily
Approximately half of the intracellular carbohydrates in the
Escherichia coli metabolome contain a phosphate moiety [1]. A
plethora of enzymes is involved in the dynamic regulation and
homeostasis of sugar phosphate metabolite levels. One key enzyme
subset is the phosphatases [2]. They catalyze phosphomonoester
group hydrolysis to release inorganic phosphate from phosphory-
lated substrate.
The known sugar phosphate phosphatases represent a large
diversity in protein structures, catalytic mechanisms, and substrate
specificity. However, many of them are evolutionary related to
haloacid dehalogenase (HAD)-like proteins [3–5]. Originally named
currently comprises mainly phosphatases, with about 79% of the
classified and biochemically studied proteins exhibiting this activ-
ity [6].
Common structural element of HAD-like phosphatases is a
Rossmann fold-like ␣/␤ core domain that contains the active
site. According to the presence and structure of an additional
domain, the so-called CAP, HAD-like phosphatases are subdi-
vided into subfamilies CI (␣-helical CAP), CII (mixed ␣/␤ CAP),
and CIII (no CAP) [7]. The catalytic centre is highly conserved.
It features a key Asp that is proposed to function as catalytic
nucleophile in a double displacement-like enzyme reaction mech-
anism. Phospho-monoester hydrolysis would accordingly proceed
in two catalytic steps via a covalent aspartyl-phosphate enzyme
intermediate [6–8]. HAD-like phosphatases require divalent metal
ion (typically Mg2+) bound in the active site for their full activ-
ity [9,10]. Phosphatase substrate specificity is dictated by flexible
∗
Corresponding author at: Institute of Biotechnology and Biochemical Engineer-
ing, Graz University of Technology, Petersgasse 12/1, A-8010 Graz, Austria.
E-mail address: bernd.nidetzky@tugraz.at (B. Nidetzky).
http://dx.doi.org/10.1016/j.molcatb.2014.09.004
1381-1177/© 2014 Elsevier B.V. All rights reserved.

40
M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
loop structures on the ␣/␤ core domain and on the CAP domain
[11]. The spectrum of phosphoryl substrates hydrolyzed is usually
very broad, and it includes both sugar and non-sugar phosphates
[12]. Relationships between structure and specificity are only
weakly defined for HAD-like phosphatases. Substrate binding and
product release involve domain-closing and opening movements
of the CAP [9,13–15]. Protein conformational flexibility further
complicates efforts to infer substrate specificity from sequence
and three-dimensional structure information alone. Purposeful
selection or molecular design of a HAD-like phosphatase for biocat-
alytic conversion of specific target substrates is therefore difficult.
Evidence from biochemical characterization is vital for efficient
development.
2.2. Expression and purification of His6 HAD4
Expression vector pCA24N-yihx encoding HAD4 equipped with
N-terminal hexahistidine tag was kindly provided by Dr. Alexan-
der Yakunin. The vector was transformed into electro-competent
cells of E. coli BL21 (DE3), and single-colony transformants were
selected on agar plates containing 0.025 mg/mL chloramphenicol.
Recombinant protein was produced in 1-L baffled shaking flasks
at 37 ◦C using an agitation rate of 110 rpm (Certomat BS-1 shaking
incubator, Sartorius, Göttingen, Germany). Flasks contained 250 mL
Lennox-medium supplemented with 0.025 mg/mL chlorampheni-
col. At OD600 of 0.8, temperature was decreased to 18 ◦C, 0.01 mM
isopropyl-␤-d-thiogalactopyranoside (IPTG) was added, and culti-
vation was continued for 20 h. Centrifuged (4 ◦C, 30 min; 4,420 g)
and washed cells were suspended in 10 mL of 50 mM Mes pH
7.0 buffer and disrupted by triple passage through French pres-
sure cell at 150 bar (American Instruments, Silver Springs, USA).
After removal of cell debris (4 ◦C, 30 min; 20,000 g) and filtra-
tion through a 1.2 ␮m cellulose-acetate syringe filter (Sartorius,
Göttingen, Germany). The pre-treated cell lysate (350 mg; 8 mL)
was loaded on a 14 mL copper-loaded affinity column (Chelating
Sepharose Fast Flow; XK 16/20 column, GE Healthcare, Little Chal-
font, U.K.), beforehand equilibrated with buffer A (50 mM Tris
HCl, pH 7.0 containing 300 mM NaCl) and mounted on an Bio-
Logic DuoFlowTM system (Biorad, Hercules, CA, USA). Differential
elution was performed with buffer B (50 mM Tris HCl, pH 7.0
containing 300 mM NaCl and 400 mM imidazole) at 10 ◦C and a
flow rate of 3 mL/min. The elution protocol comprised 5 steps,
whereas concentration of buffer B was stepwise increased to 0,
10, 30, 60 and 100% in a volume of 100, 90, 60, 30 and 75 mL,
respectively. All buffers were degassed and filtered using 0.45 ␮m
cellulose-acetate or 0.2 ␮m polyamide filters. Protein elution was
monitored spectrophotometrically at 280 nm, and collected frac-
tions were assayed for protein concentration and phosphatase
activity against p-nitrophenyl phosphate (pNPP; see Assays). Active
fractions were pooled and buffer was exchanged to buffer A2
(50 mM KH2PO4-buffer, pH 7.0) with Amicon Ultra-15 Centrifugal
Filter Units (Millipore, Billerica, USA). Pooled fractions were fur-
ther purified using a 50 mL Fractogel EMD-DEAE column (XK 26/20,
Merck, Darmstadt, Germany) mounted on a BioLogic DuoFlowTM
system. Protein elution was performed with buffer B2 (50 mM
KH2PO4, 1 M NaCl, pH 7.0) applying a five-step purification pro-
tocol at a constant flow rate of 5 mL/min: step 1, isocratic flow of
100% buffer A2 for 50 mL; step 2, linear increase to 15% buffer B2
in 40 mL; step 3, isocratic flow of 85% buffer A2 and 15% buffer
B2 for 80 mL; step 4, linear increase to 100% buffer B2 in 180 mL;
step 5, isocratic flow of 100% buffer B2 for 70 mL. Active fractions
were pooled and buffer was exchanged to 50 mM Hepes pH 7.0
with Amicon Ultra-15 Centrifugal Filter Units. Protein purification
was monitored by SDS PAGE and phosphatase activity measure-
ments.
Selectivity is a prime feature of many enzymatic transfor-
mations [16]. A growing number of examples show that in a
comparison of conventional-chemical and biocatalytic process
options in organic synthesis [17,18], it is usually the bio-based
selectivity that provides a decisive advantage for process develop-
ment. In this paper, we address the problem of substrate-selective
hydrolysis of ␣-d-glucose 1-phosphate (␣Glc 1-P) by phosphatases
and report identification of HAD4 (yihx gene product) from E. coli
for that purpose. Major demand for the reaction is in reactive
processing of mixtures of sugar phosphates to eliminate all of
the ␣Glc 1-P present in the starting material. Sample work-up
for analytics, in the field of sugar phosphate metabolomics for
example, and facilitated sugar phosphate product recovery con-
stitute interesting applications of selective ␣Glc 1-P converting
phosphatases. The readily available broad-spectrum phosphatases
are however not suitable for these applications, and discovery
of new phosphatase catalysts is therefore required. HAD4 was
obtained as highly active recombinant enzyme through a specially
designed fusion protein (Zbasic2 HAD4) that contained Zbasic2, a
strongly positively charged three ␣-helical bundle module, at its
N-terminus [19]. Zbasic2 facilitated functional expression of solu-
ble protein in E. coli and enabled protein capture and polishing
efficiently combined in a single step of cation exchange chromatog-
raphy. Furthermore, the Zbasic2 module had an instrumental role
in the development of an immobilized HAD4 biocatalyst where
Zbasic2 HAD4 was bound directly from crude bacterial cell extract
on sulfonic acid group-containing porous carriers. Attachment of
fusion protein to the negatively charged carrier surface was not only
highly selective in that accompanying E. coli protein only showed
marginal binding under the conditions used, but it also appeared
to have occurred in a strongly preferred orientation via the cationic
binding module: immobilized Zbasic2 HAD4 was nearly as effec-
tive as the free enzyme. Kinetic characterization of Zbasic2 HAD4
with ␣Glc 1-P and the competing sugar phosphate substrate d-
glucose 6-phosphate (Glc 6-P) is reported. Utilization of different
sugar phosphates as substrates for hydrolysis by Zbasic2 HAD4 is
described, and some basic properties of the enzyme fused to Zbasic2
are reported. Clear-cut separation of Glc 6-P from ␣Glc 1-P through
selective hydrolysis of the sugar 1-phosphate at high concentra-
tion (up to 180 mM) and in the presence of up to 9-fold excess of
competing substrate is shown.
Alternatively, the pre-treated cell lysate (280 mg; 8 mL) was
loaded on 1.6 cm × 2.5 cm; 5 mL HisTrap FF column (GE Health-
care), equilibrated with buffer C (50 mM MES, pH 7.4 containing
125 mM NaCl, 20 mM imidazole) and mounted on an ÄKTA prime
plus (GE Healthcare) system. Protein was eluted using an imid-
azole gradient from 0 to 80% buffer D (50 mM MES, pH 7.4
containing 125 mM NaCl and 500 mM imidazole) at 10 ◦C and
at a flow rate of 4 mL/min. At 20% buffer D, the linear gradient
was halted for 2 column volumes to facilitate differential elu-
tion of bound protein. Active fractions were pooled and buffer
was exchanged to 50 mM MES, pH 7.0. Protein purification was
monitored by SDS PAGE and phosphatase activity measurements.
Densitometry evaluation of the SDS PAGE was performed with
the GelAnalyzer software (http://www.gelanalyzer.com/index.
html).
2. Materials and methods
2.1. Chemicals and reagents
Unless stated otherwise, all chemicals were of highest purity
available from Sigma–Aldrich (Vienna, Austria) or Roth (Karlsruhe,
Germany). Sugar phosphates were obtained as sodium or ammo-
nium salts. Oligonucleotide primers were from Sigma–Aldrich
(Vienna, Austria). DNA sequencing was done at LGC Genomics
(Berlin, Germany).

M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
41
reduced by l-ascorbic acid in the presence of Zn2+, which produces
a chromophore that can be measured at 850 nm.
2.3. Molecular cloning, expression and purification of
Zbasic2 HAD4
Reported coupled enzyme assays were used to measure ␣Glc 1-
P and Glc 6-P [23]. In contrast to Eis and Nidetzky [23], however, no
trehalose phosphorylase was present in our reaction mixture. Glu-
cose 6-P dehydrogenase from Leuconostoc mesenteroides (G8404;
Sigma Aldrich) and ␣-phosphoglucomutase from rabbit muscle
(P3397; Sigma Aldrich) were added sequentially. After addition of
glucose 6-P dehydrogenase or phosphoglucomutase, formation of
NADH was monitored at 340 nm and the NADH end concentra-
tion was used to calculate the corresponding Glc 6-P or ␣Glc 1-P
concentration.
The HAD4 gene was amplified from pCA24N-yihx by PCR using
Phusion DNA polymerase (Thermo Scientific, Waltham, USA) and
the following pair of oligonucleotide primers:
5ꢀ-GAAGCTCTGTTCCAGGGTCCGCTCTATATCTTTGATTTAG-3ꢀ,
and 5ꢀ-CTTTGTTAGCAGCCGGATCTCTTAGCATAACACCTTCG-3ꢀ.
PCR consisted of a pre-heating step at 98 ◦C for 5 min and was
followed by 30 reaction cycles of denaturation at 98 ◦C for 30 s,
annealing at 70 ◦C for 30 s, and elongation at 72 ◦C for 1.5 min. The
final extension step was carried out at 72 ◦C for 5 min. The result-
ing PCR product comprised 5ꢀ- and 3ꢀ-overhangs complementary
to the target sequence of the pT7ZbQGKlenow destination vector
[19]. Vector pT7ZbQGKlenow was amplified by PCR using Phusion
DNA polymerase and the following pair of oligonucleotide primers:
5ꢀ-CGGACCCTGGAACAGAGC-3ꢀ, and 5ꢀ-GAGATCCGGCTGCTAAC
AAAG-3ꢀ.
2.5. Kinetic characterization
All reactions were performed in 50 mM MES buffer, pH 7.0, con-
taining 25 mM MgCl2 and 100 mM NaCl. Incubations were done at
37 ◦C using agitation at 650 rpm on a Thermomixer comfort (Eppen-
dorf, Hamburg, Germany). ␣Glc 1-P or Glc 6-P was used as substrate
in the concentration range 0.1–30 mM and 0.1–700 mM, respec-
tively. To account for the lower activity of Zbasic2 HAD4 with Glc 6-P
as compared to ␣Glc 1-P, the corresponding enzyme concentration
was 0.39 (␣Glc 1-P) and 16 ␮M (Glc 6-P), respectively. Reactions
were started by addition of enzyme and allowed to run for up to
120 min while taking samples every 15 min. Thermal inactivation
(99 ◦C, 5 min) was used to stop reactions. Free phosphate concen-
tration was measured in supernatants of centrifuged samples.
It was confirmed that phosphate release was linear with incu-
bation time. Phosphate concentration after 15 min, or 120 min of
reaction was used to determine the initial rate for ␣Glc 1-P and
Glc 6-P conversions, respectively. Phosphate blanks from reactions
lacking enzyme were recorded.
Kinetic parameters were obtained from non-linear least squares
fits (SigmaPlot, Systat, Erkrath, Germany) of initial rates to
Michaelis–Menten equation expanded to include substrate inhibi-
tion (1). V is the initial rate, Vmax is the maximum initial rate, KM is
the Michaelis constant, KiS is the substrate inhibition constant, and
[S] is the initial substrate concentration. Under conditions where
KiS < KM (see later under Results) fitting of Eq. (1) did not succeed
and individual determination of all three kinetic parameters was
not possible. However, Vmax/Km could be determined from the ini-
tial part of the V–[S] curve where V is linearly dependent on [S].
PCR protocol was the same as above, except that elongation at
72 ◦C was extended to 3 min. Amplification products of the two
PCRs were individually subjected to parental template digest by
DpnI and then recombined via Gibson assembly [20]. The resulting
construct encodes a fusion protein where the small module Zbasic2
is joined by a flexible linker (11 amino acids) to the N-terminus of
HAD4. Sequenced plasmid vector encoding the Zbasic2 HAD4 fusion
protein was transformed into electrocompetent cells of E. coli BL21-
Gold (DE3), and single-colony transformants were selected on agar
plates containing 0.05 mg/mL kanamycin.
Protein production was done as described above, except for the
use of 0.05 mg/mL kanamycin. Harvested cells were resuspended in
10 mL of buffer A3 (50 mM MES buffer, pH 7.0, containing 100 mM
NaCl). Preparation of cell extract was done exactly as already
described.
Protein was purified at 10 ◦C using pre-packed 1.6 cm ×
2.5 cm; 5 mL HiTrap SP FF column (Ni Speharose; 16/25 mm, GE
Healthcare) on ÄKTA prime plus system. The column was equili-
brated with buffer A3 and 8 mL of the pre-treated cell extract were
loaded onto the column. Protein elution was performed using a con-
tinuous salt gradient ranging from 0 to 100% of buffer B3 (50 mM
HEPES buffer supplemented with 2 M NaCl, pH 7.0) in 150 mL at
a flow rate of 4 mL/min. The target protein eluted at 50% buffer
B3. Pooled fractions were collected and buffer was exchanged to
50 mM MES, pH 7.0, containing 250 mM NaCl and 50 mM MgCl2, as
described before. The concentrated preparation was aliquoted and
stored at −20 ◦C. Protein purification was monitored by SDS PAGE
and phosphatase activity measurements.
Vmax
S
V =
(1)
KM + S(1 + (S/KiS))
Besides ␣Glc 1-P and Glc 6-P various other sugar phosphates
were examined as substrates of Zbasic2 HAD4: ˛-d-galactose 1-P
(␣Gal 1-P), ˛-d-mannose 1-P (␣Man 1-P), ˛-d-galactose 6-P (␣Gal
6-P) and ˛-d-mannose 6-P (␣Man 6-P). They were used at constant
concentration of 20 mM, and the initial rate of phosphate release
was measured.
2.4. Assays
Protein was measured with Roti-Quant reagent (Roth, Karl-
sruhe, Germany) referenced against bovine serum albumin (BSA).
Purified Zbasic2 HAD4 was measured by UV absorbance at 280 nm
using a molar extinction coefficient of 25440 M−1 cm−1, calculated
from the sequence with ProtParam analysis tool from the expasy
server [21]. Standard phosphatase activity was measured with
pNPP (4.0 mM), the release of p-nitrophenol (pNP) was recorded
continuously at 405 nm with a Beckman Coulter DU 800 UV/VIS
spectrophotometer (Krefeld, Germany).
2.6. pH and temperature profiles of activity and stability
pH effects were studied at 37 ◦C (activity) or 25 ◦C (stability) in
the range 4.0–9.0 using multicomponent buffer of 50 mM sodium
acetate, 50 mM TES, 50 mM MES supplemented with 100 mM NaCl
and 25 mM MgCl2. Temperature effects (20–60 ◦C) were examined
at pH 7.0 using the salt-supplemented 50 mM MES buffer. pH or
temperature effects on activity were determined using ␣Glc 1-P
(20 mM) as substrate and measuring the phosphate release within
10 min of incubation. Effects on stability were determined from
non-agitated incubations of enzyme solution (200 ␮L; 23 ␮M).
Samples were withdrawn at certain times and residual activity was
measured under ␣Glc 1-P assay conditions (pH) as described above.
Reaction was performed at 30 ◦C using 50 mM MES, pH 7.0, con-
taining 25 mM MgCl2 and 100 mM NaCl. One unit of activity (U) is
the amount of enzyme releasing 1 ␮mol of pNP under the condi-
tions used. Specific activity is expressed in U/mg protein.
Inorganic phosphate was measured with a spectrophotometric
assay described by Saheki et al. [22]. Shortly, free phosphate forms
a complex with molybdate and the phosphomolybdate complex is

42
M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
Fig. 1. (A) Purification of Zbasic2 HAD4 monitored by SDS-PAGE. 1: PAGE-ruler protein ladder, 2: cell extract (1.5 ␮g), 3: insoluble protein fraction from E. coli expression
culture (1.5 ␮g), 4: flow-through (proteins that did not bind to the column) (1.5 ␮g), 5: fraction with minor phosphatase activity (1.0 ␮g), 6: fraction with minor phosphatase
activity (0.5 ␮g), 7: purified Zbasic2 HAD4 (1.5 ␮g); (B) Chromatogram of Zbasic2 HAD4 purification with 1.6 cm × 2.5 cm; 5 mL HiTrap SP FF column. The absorbance trace at
280 nm is shown as solid line. The salt gradient used is shown as dashed line, and the measured conductivity is shown as dotted line.
2.7. Zbasic2 HAD4 immobilization
high phosphatase background in the E. coli cell extract restricted
ꢀ determination to conditions where purified Zbasic2 HAD4 was
immobilized. To measure the activity of the immobilized enzyme,
3 mg of immobilizate were incubated at 37 ◦C in 500 ␮L solution
containing 20 mM ␣Glc 1-P as substrate. Agitation was at 330 rpm
to keep particles in suspension. Release of phosphate was measured
in samples after 10 min of incubation.
Basic protocol from Wiesbauer et al. [24] was used. Briefly,
−
Fractogel EMD SO3 (Merck, Darmstadt, Germany) and Relisorb
SP400 (Resindion, Binasco Milano, Italy) were the carriers used,
and their technical specifications are provided in Supplementary
Table T1. Hundred mg of wet carrier were weighed into Eppen-
dorf tubes, washed twice with water, and equilibrated with 50 mM
MES (pH 7.0) containing 200 mM NaCl. Two mL of enzyme solution
(5 mg/mL E. coli cell extract; 0.5 mg/mL purified Zbasic2 HAD4) were
added and the tubes were incubated in an end-over-end rotator at
25 ◦C for 60 min. Supernatant was removed (25 ◦C, 5 min; 20,000 g)
and weakly bound protein was washed away twice with loading
buffer. Note that only at 2 M NaCl the specifically and strongly
bound protein was eluted from the carrier. Enzyme immobilizates
were stored in 50 mM MES (pH 7.0) containing 100 mM NaCl and
25 mM MgCl2 at 4 ◦C. Enzyme immobilization was monitored from
the decrease in protein concentration and phosphatase activity in
solution, observing that enzyme was stable during the immobiliza-
tion.
2.8. Substrate-selective hydrolysis of ˛Glc 1-P in the presence of
Glc 6-P using free or immobilized Zbasic2 HAD4
Reaction conditions were the same as above under 2.5 Kinetic
characterization except that temperature was 30 ◦C unless indi-
cated. The substrate and soluble enzyme concentrations used are
shown under 3. Results. Samples were taken at certain times
and assayed for ␣Glc 1-P and Glc 6-P. Immobilized enzyme pre-
pared from E. coli cell extract was used identically as the free
enzyme, applying 3 mg of carrier in 500 ␮L substrate solution.
Agitation at 330 rpm with a magnetic stirrer bar was sufficient
to have carrier particles well suspended. Once ␣Glc 1-P was
converted almost exhaustively (∼90 min), particles were spinned
down (25 ◦C, 5 min; 20,000 g), washed once with buffer, and then
suspended again with substrate solution. Restart of the reaction
was done twice.
Effectiveness factors (ꢀ) of the Zbasic2 HAD4 immobilizates
were determined by comparing the observable activity of the
carrier-bound enzyme to the theoretically bound enzyme activity
calculated from the balance in solution. Therefore, ꢀ is the ratio of
observable and theoretically bound enzyme activity. The relatively

M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
43
Fig. 3. Phosphatase activity of Zbasic2 HAD4 with different sugar phosphate sub-
strates and pNPP. Reactions were performed at 37 ◦C in 50 mM MES buffer (pH 7.0)
containing 25 mM MgCl2 and 100 mM NaCl. The substrate concentration was 20 mM,
except for pNPP (4.0 mM). kcat app is an apparent turnover number calculated from
the initial rate of phosphate release divided by the molar enzyme concentration
which in turn is obtained from the protein concentration and the enzyme molecular
mass of 31 kDa. For further details about methodology, see Section 2.5.
Fig. S2). According to densitometric analysis of the protein gel, the
sample was composed roughly of 30–40% His6 HAD4. Yield of tar-
get protein was therefore estimated to have been around 5 mg/L
culture.
To establish a more efficient procedure of isolation of HAD4
and also in view of convenient immobilization of the enzyme
of ion-exchange support, we chose not to optimize protocols for
His6 HAD4, but opted for an alternative fusion protein approach.
A chimeric enzyme termed Zbasic2 HAD4 was constructed where
the so-called Zbasic2 module was attached to the N-terminus of
HAD4. Zbasic2 is an engineered arginine-rich variant of the Z domain,
a 58 amino acid (7 kDa) three-helix bundle obtained from the B
domain of staphylococcal protein A [25]. Fusion to Zbasic2 was
known from previous studies to enhance solubility of proteins oth-
erwise poorly soluble under conditions of recombinant production
in E. coli [19,26]. Zbasic2 was furthermore used as affinity module
for protein purification by cation exchange chromatography [25].
Finally, it was also applied for non-covalent immobilization of dif-
ferent enzymes on negatively charged carrier surfaces [24,27].
Zbasic2 HAD4 was isolated by cation exchange chromatography
of the E. coli cell extract on HiTrap SP FF resin. Purified protein
exhibited the expected size of 31 kDa. Capture of target protein,
from the complex protein matrix and its purification to apparent
homogeneity (Fig. 1A) were combined efficiently in a single work-
up step. Zbasic2 HAD4 eluted from the column as discrete protein
peak at high salt concentration, as shown in Fig. 1B. Analysis by
SDS PAGE (Fig. 1A) revealed that the flow-through fraction did not
contain Zbasic2 HAD4, demonstrating that all of the applied target
protein had bound to the column. Fig. 1A also shows that insol-
uble protein recovered from the E. coli cells contained significant
amounts of Zbasic2 HAD4, indicating that fusion to Zbasic2 was not
fully sufficient to prevent aggregation of HAD4 during recombi-
nant production in E. coli. About 11 mg of pure Zbasic2 HAD4 were
recovered from each L of bacterial cell culture. Purification with
HiTrap SP FF resin was extremely effective, considering that the
total protein amount present in the sample was almost 270 mg/L.
It is interesting but currently not understood why the purifica-
tion of Zbasic2 HAD4 was much more efficient than purification of
His6 HAD4. The specific activity towards pNPP of Zbasic2 HAD4 was
determined as 0.06 U/mg.
Fig. 2. Initial-rate kinetic characterization of sugar phosphate conversion by
Zbasic2 HAD4. (A) Michaelis–Menten plot for the hydrolysis of ␣Glc 1-P. (B) Michaelis-
Menten plot for the hydrolysis of Glc 6-P. For further details about methodology, see
2.5. Kinetic characterization. Symbols show the data, and the solid lines are non-linear
fits of Eq. (1) to the data.
3. Results and discussion
3.1. Functional expression and purification of HAD4
Kuznetsova et al. reported isolation of ␮g amounts of His6 HAD4
using automated high-throughput work-up procedures performed
at ␮L scale [12]. We attempted to scale up the purification of
His6 HAD4 to a processing capacity of about 400 mg total protein
from the crude E. coli cell extract. Contrary to the earlier study
where Ni-NTA beads were used [12], Cu2+ loaded immobilized
metal affinity chromatography (IMAC) in 14 mL column format
(16 cm × 20 cm) was used herein. Even though about one-third of
the applied protein was bound to the column, differential elution
with an imidazole step gradient did not result in fractionation of
protein(s) exhibiting clear phosphatase activity towards pNPP sub-
strate. Protein pools potentially containing phosphatase activity
showed a prominent band of correct size (25 kDa) next to other
major contaminants in the SDS polyacrylamide gel (Supplemen-
tary Fig. S1). They were therefore further fractionated by anion
exchange chromatography. Collected fractions showed only trace
amounts of phosphatase activity, and their analysis by SDS PAGE
indicated that purification had not been successful (Supplementary
Fig. S1).
We therefore tested IMAC on Ni2+ Sepharose using a differ-
ent column format (1.6 cm × 2.5 cm; 5 mL; HisTrapFF) at smaller
scale. Pooled protein fractions exhibiting phosphatase activity were
shown by SDS PAGE to contain the expected band of 25 kDa but
also a major contaminant of around 37 kDa size (Supplementary
Stock solutions of purified Zbasic2 HAD4 were prepared by con-
centrating desalted protein in 50 mM MES buffer, pH 7.0, to about

44
M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
Fig. 4. Selective hydrolysis of ␣Glc 1-P by Zbasic2 HAD4. Reaction mixtures contained (A) 20 mM ␣Glc 1-P (•) and 20 mM Glc 6-P (ꢁ), (B) 180 mM ␣Glc 1-P (•) and 180 mM Glc
6-P (ꢁ), and (C) 20 mM ␣Glc 1-P (•) and 180 mM Glc 6-P (ꢁ). Reactions further contained purified Zbasic2 HAD4 at 3.2 ␮M (A), 16 ␮M (B) and 1.6 ␮M (C). They were performed
at 30 ◦C and pH 7.0, except in A where 37 ◦C was used. Note: the Glc 6-P concentration in panel C at 150 min was measured but includes effect of slight water evaporation.
10 mg/mL. After frozen storage at −20 ◦C, we observed significant
protein precipitation in thawed samples. Addition of 250 mM NaCl
resulted in partial re-solubilization of the precipitate concomitant
with recovery of some of the original phosphatase activity in solu-
tion, suggesting effect of ionic strength on solubility and stability of
Zbasic2 HAD4. Supplementing the storage buffer with a combination
of salts (250 mM NaCl, 50 mM MgCl2) prevented the precipita-
tion of Zbasic2 HAD4 completely. Storage in liquid state at 4 ◦C was
preferred where the enzyme was stable for several weeks. Ten-
tative explanation for tendency of Zbasic2 HAD4 to form insoluble
aggregates at low ionic strength is the large difference in the calcu-
lated isoelectric points for Zbasic2 (pI = 11.5) and HAD4 (pI = 5.18).
Intramolecular and intermolecular domain–domain interactions
driven by charge complementarity might promote a native-like
protein aggregation that is partly reversible at high salt concen-
tration. Mg2+ ions could have two effects in this context. First of
all, Mg2+ is an active-site ligand of HAD4 [12]. Thus the structure of
the catalytic center (and perhaps that of the whole protein) might
be stabilized when Mg2+ is bound. Secondly, adsorption of Zbasic2
fusions to negatively charged surfaces is strongly weakened in the
presence of Mg2+ ions [28]. Capacity of Mg2+ to perturb charged
interactions of the Zbasic2 module are much more pronounced than
that of Na+ (each as chloride salt) [28].
reaction occurring next to the hydrolysis. Both the phosphorylated
substrate and the glucose product could in principle function as
acceptor of the phosphoryl group and thereby compete with water
in the overall reaction. However, it was confirmed that initial rates
of substrate utilization and substrate hydrolysis were identical within
limits of experimental error.
Fig. 2 shows initial rates determined at different substrate con-
centrations for hydrolysis of ␣Glc 1-P (Fig. 2A) and Glc 6-P (Fig. 2
B). For each substrate, the V–[S] curve increased at low [S] to reach
a distinct maximum value of V at intermediate [S], only to decrease
again in the high-[S] range. The kinetic behavior indicates inhibi-
tion of Zbasic2 HAD4 by the phosphorylated substrate. Interestingly,
His6 HAD4 did not show substrate inhibition at high ␣Glc 1-P (data
not shown), suggesting substrate inhibition to be a feature caused
by introduction of the Zbasic2 module.
␣Glc 1-P differed from Glc 6-P in affinity and reactivity
for Zbasic2 HAD4-catalyzed hydrolysis. Maximum value of V was
observed at around 10 mM for ␣Glc 1-P while for Glc 6-P the cor-
responding substrate concentration was about 100 mM (Fig. 2).
Substrate inhibition also occurred at lower concentrations of ␣Glc
1-P as compared to Glc 6-P. An apparent kcat was calculated from the
maximum V value (kcat app = Vmax app/[E]), and ␣Glc 1-P (10.2 s−1
)
was about 110-fold more active in kcat app terms than Glc 6-P
(0.09 s−1). By way of comparison, kcat app for hydrolysis of pNPP
(4 mM) was 0.032 s−1. For comparison with Zbasic2 HAD4, we also
examined the preparation of His6 HAD4 under otherwise identical
conditions and measured its hydrolytic activity against ␣Glc 1-P
(10 mM). Using correction for partial purity of the enzyme prepa-
ration used where only 30–40% of total protein was His6 HAD4
(Fig. S2), we calculated a kcat app between 7.1 and 9.5 s−1 from
the experimentally determined initial rate. Therefore, this result
3.2. Kinetic analysis of hydrolysis of ˛Glc 1-P, Glc 6-P and other
sugar phosphates catalyzed by Zbasic2 HAD4
Initial reaction rates at steady state were obtained from time-
course measurements of substrate consumption and phosphate
release. Disparity of initial rates determined from the different
measurements would have indicated a phosphoryl transfer side

M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
45
be considered a relatively specific ␣Glc 1-P hydrolase, overall
consistent with findings of Kuznetsova et al. on the substrate
specificity of His6 HAD4 [21].
3.3. Effects of pH and temperature on activity and stability of
Zbasic2 HAD4
Results are shown in Figs. S3 (pH effects) and S4 (tempera-
ture effects). The pH-profile of Zbasic2 HAD4 activity at 37 ◦C was
bell-shaped with a broad plateau between pH 5.0 and pH 7.0
where enzyme showed at least 85% of its maximum activity at pH
6.0. pH dependence of the activity is similar for Zbasic2 HAD4 and
His6 HAD4 [21]. Recorded at 25 ◦C, the enzyme was fully stable for
up to 20 h at all pH values in the range 4.0–9.0. Temperature profile
at pH 7.0 showed optimum activity of Zbasic2 HAD4 at 40 ◦C, with
rapid decline of activity at higher temperatures. Activity was fully
stable for 20 h at or below 30 ◦C while at the higher temperatures
(≥ 40 ◦C) the activity was very rapidly lost.
3.4. Substrate-selective hydrolysis of ˛Glc 1-P in a mixture of
˛Glc 1-P and Glc 6-P
High selectivity of Zbasic2 HAD4 for hydrolyzing ␣Glc 1-P as com-
pared to Glc 6-P, as implied by the ratio of the corresponding kcat/KM
values, was put to critical test in a biotransformation experiment in
which both sugar phosphates were offered as substrate. The sub-
strate concentrations (20 mM; 180 mM) and also the molar ratio of
␣Glc 1-P/Glc 6-P were varied. Time courses of the enzymatic con-
versions are shown in Fig. 4 (panels A–C). In all cases, ␣Glc 1-P was
effectively depleted from the substrate mixture while Glc 6-P was
not utilized within error limit.
Zbasic2 HAD4-catalyzed selective hydrolysis of ␣Glc 1-P thus
provided effective mean for separation of the two sugar phos-
phates. Specialized separation procedures such as borate complex
ion exchange chromatography [29] would otherwise be needed to
isolate the Glc 6-P. In the case of the alternative separation problem
to remove Glc 6-P from ␣Glc 1-P, another HAD-like phosphatase,
namely the enzyme BT4131 from Bacteroides thetaiotaomicron,
could be of interest [9]. The phosphatase is highly active in hydroly-
sis of hexose 6-phosphates and pentose 5-phosphates, but exhibits
barely any activity with hexose 1-phosphates. Taken together,
these results suggest that HAD-like phosphatases present a poten-
tially valuable source of selective sugar phosphate hydrolases that
could be applied to resolution of mixtures of sugar phosphates
that are otherwise very difficult to separate. In addition to phos-
phatases that exhibit broad substrate spectrum and are already
available commercially, development of a toolbox of selective phos-
phomonoester hydrolases could provide interesting opportunities
for biocatalysis.
Fig. 5. Selective hydrolysis of ␣Glc 1-P by Zbasic2 HAD4 immobilized on (A) Relisorb
SP400 and (B) Fractogel EMD SO3−. Reaction mixtures contained 15 mM ␣Glc 1-P (•)
and 20 mM Glc 6-P (ꢁ). Reactions were performed at 30 ◦C and pH 7.0. Three mg of
immobilizate prepared from E. coli cell extract as described under 2.7. Zbasic2 HAD4
immobilization were suspended in 500 ␮L substrate solution. Agitation was applied
to keep particles in suspension.
is important in showing that fusion to Zbasic2 did not diminish
the intrinsic activity of HAD4 relative to the His6 HAD4 refer-
ence.
Eq. (1) could be used for data fitting by nonlinear least squares
regression, as shown in Fig. 2; however, the resulting kinetic param-
eters (kcat, KM, KiS) were not well defined statistically. Covariance
analysis revealed strong interdependence of all parameters, but
especially of KM and KiS, implying that independent determination
of each single parameter was not possible from the data avail-
able. An approximate ratio between KM/KiS can be given however,
and this was ∼21 (60/2.82) for ␣Glc 1-P and ∼0.15 (51/351) for
Glc 6-P. To nevertheless obtain parameter of enzyme specificity,
we determined kcat/KM from the part of the V–[S] curve where V
was linearly dependent on [S]. The relationship V = (kcat/KM) [S] [E]
holds under these conditions. kcat/KM for hydrolysis of ␣Glc 1-P
(1.87 0.03 mM−1 s−1) exceeded the kcat/KM for hydrolysis of Glc
6-P (3.2 0.1 × 10−3 mM−1 s−1) by 565-fold. Zbasic2 HAD4 strongly
discriminates between ␣Glc 1-P and Glc 6-P as substrate for hydrol-
ysis.
3.5. Immobilization of Zbasic2 HAD4 and repeated use of
immobilizates
Immobilization on Fractogel and Relisorb carriers was first done
with the purified enzyme that was offered in a protein/wet carrier
mass ratio of 0.01. Nearly all (≥95%) of the offered protein and phos-
phatase activity was bound to each carrier under the conditions
used. An effectiveness factor (ꢀ) of 0.9 ( 0.1) was determined for
both immobilizates, indicating that immobilized Zbasic2 HAD4 was
nearly as active as the corresponding soluble enzyme preparation. It
was pointed out in earlier studies of Zbasic2 enzyme fusions [20,23]
and is confirmed here using Zbasic2 HAD4, that highly directed
(“oriented”) immobilization via the Zbasic2 module offers the par-
ticular advantage of preparing carrier-bound insoluble biocatalyst
with optimized values of ꢀ. Note also that despite its noncova-
lent character, the immobilization is quite stable [24,27], as will be
Besides ␣Glc 1-P and Glc 6-P we examined other sugar phos-
phates as substrates of Zbasic2 HAD4, focusing in particular on
the enzyme’s ability to discriminate between the ␣-configured
1-phosphate and the corresponding 6-phosphate. Fig. 3 shows the
results. All sugar 6-phosphates (d-Glc, d-Man, d-Gal) were slow
enzyme substrates, exhibiting turnover rates of about 1 × 10−1 s−1
.
Activity with ␣Man 1-P and ␣Gal 1-P was enhanced by about
one magnitude order in comparison (Fig. 3), yet the activity with
␣Glc 1-P stood out among all substrates tested. Zbasic2 HAD4 can

46
M. Pfeiffer et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 39–46
shown later for Zbasic2 HAD4 immobilizates in biocatalyst recycling
experiment.
University of Toronto, Canada) for plasmid vector pCA24N-yihx;
and Sophia Hober (Department of Biotechnology, Alba Nova Uni-
versity Center, Stockholm, SE) for the plasmid vector containing the
Zbasic2 gene.
When Fractogel or Relisorb carrier was charged with E. coli cell
extract at protein/carrier mass ratio of 0.1, about 12–18% of total
protein was bound to the solid support. Due to high background
phosphatase activity in cell extract, balance for Zbasic2 HAD4 activ-
ity was not possible under these conditions. However, from the
phosphatase activity of the immobilizate and the protein eluted
from the carrier at high salt concentration, we determined a spe-
cific activity of immobilized Zbasic2 HAD4 of about 0.05 U/mg which
is close to the specific activity of the purified soluble enzyme.
Enzyme immobilizates thus prepared were applied to repeated
batchwise hydrolysis of ␣Glc 1-P from a substrate mixture that
contained 20 mM Glc 6-P and 15 mM of ␣Glc 1-P. Heterogeneous
biocatalyst was recycled two times in each experiment, as shown
in Fig. 5 where panels A and B display results for the Relisorb and
Fractogel immobilizate, respectively. The Relisorb immobilizate
was re-used without significant loss of activity over three reac-
tion cycles where ␣Glc 1-P was degraded almost completely and
the Glc 6-P present was not affected. Using the Fractogel immo-
bilizate, by contrast, the rate of conversion of ␣Glc 1-P decreased
about twofold from cycle 1 to 3 (Fig. 5, panel B). Assuming that
Zbasic2 HAD4 was stable under the conditions used (cf. Fig. S4), grad-
ual elution of enzyme from this carrier is a plausible reason for the
observed decay of the reaction rate and thus the space-time yield.
However, further examination of the operational stability of the
Fractogel immobilizate of Zbasic2 HAD4 was beyond the scope of
this study.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.2014.
09.004.
References
[1] I. Nobeli, H. Ponstingl, E.B. Krissinel, J.M. Thornton, J. Mol. Biol. 334 (2003)
697–719.
[2] K. Papenfort, Y. Sun, M. Miyakoshi, C.K. Vanderpool, J. Vogel, Cell 153 (2013)
426–437.
[3] T. Hisano, Y. Hata, T. Fujii, J.-Q. Liu, T. Kurihara, N. Esaki, K. Soda, J. Biol. Chem.
271 (1996) 20322–20330.
[4] I.S. Ridder, H.J. Rozeboom, K.H. Kalk, D.B. Janssen, B.W. Dijkstra, J. Biol. Chem.
272 (1997) 33015–33022.
[5] E.V. Koonin, R.L. Tatusove, J. Mol. Biol. 244 (1994) 125–132.
[6] K.N. Allen, D. Dunaway-Mariano, Curr. Opin. Struct. Biol. 19 (2009)
658–665.
[7] A.M. Burroughs, K.N. Allen, D. Dunaway-Mariano, L. Aravind, J. Mol. Biol. 361
(2006) 1003–1034.
[8] S.D. Lahiri, G. Zhang, D. Dunaway-Mariano, K.N. Allen, Biochemistry 41 (2002)
8351–8359.
[9] Z. Lu, D. Dunaway-Mariano, K.N. Allen, Biochemistry 44 (2005) 8684–8696.
[10] M.C. Morais, W. Zhang, A.S. Baker, G. Zhang, D. Dunaway-Mariano, K.N. Allen,
Biochemistry 39 (2000) 10385–10396.
[11] H.H. Nguyen, L. Wang, H. Huang, E. Peisach, D. Dunaway-Mariano, K.N. Allen,
Biochemistry 49 (2010) 1082–1092.
[12] E. Kuznetsova, M. Proudfoot, C.F. Gonzalez, G. Brown, M.V. Omelchenko,
I. Borozan, L. Carmel, Y.I. Wolf, H. Mori, A.V. Savchenko, C.H. Arrow-
smith, E.V. Koonin, A.M. Edwards, A.F. Yakunin, J. Biol. Chem. 281 (2006)
36149–36161.
[13] S.D. Lahiri, G. Zhang, J. Dai, D. Dunaway-Mariano, K.N. Allen, Biochemistry 43
(2004) 2812–2820.
[14] G. Zhang, J. Dai, L. Wang, D. Dunaway-Mariano, L.W. Tremblay, K.N. Allen,
Biochemistry 44 (2005) 9404–9416.
[15] S.D. Lahiri, G. Zhang, D. Dunaway-Mariano, K.N. Allen, Bioorg. Chem. 34 (2006)
394–409.
[16] T. van Herk, A.F. Hartog, A.M. van der Burg, R. Wever, Adv. Synth. Catal. 347
(2005) 1155–1162.
[17] K. Faber, Biotransformations in Organic Chemistry, Springer, New York, 2011.
[18] K. Buchholz, Biocatalysts and Enzyme Technology, 2nd ed., Wiley-Blackwell,
Weinheim, 2012.
[19] M. Hedhammar, S. Hober, J. Chromatogr. A 1161 (2007) 22–28.
[20] D.G. Gibson, L. Young, R.-Y. Chuang, J.C. Venter, C.A. Hutchison, H.O. Smith, Nat.
Methods 6 (2009) 343–345.
[21] E. Gasteiger, C. Hoogland, A. Gattiker, M.R. Wilkins, R.D. Appel, A. Bairoch, The
Proteomics Protocols Handbook, Springer, Totowa, 2005, pp. 571–607.
[22] S. Saheki, A. Takeda, T. Shimazu, Anal. Biochem. 148 (1985) 277–281.
[23] C. Eis, B. Nidetzky, Biochem. J. 341 (1999) 385–393.
[24] J. Wiesbauer, J.M. Bolivar, M. Mueller, M. Schiller, B. Nidetzky, ChemCatChem
3 (2011) 1299–1303.
4. Conclusions
First-time functional expression, isolation, and characterization
of E. coli HAD4 as fusion protein harboring the charged module
Zbasic2 at the N-terminus is reported. Single-step capture and pol-
ishing of Zbasic2 HAD4 by cation exchange chromatography was
notably efficient to obtain multi-mg amounts of highly purified
enzyme. Kinetic evidence and results from mixed sugar phosphate
conversion experiments reveal that Zbasic2 HAD4 is highly selec-
tive for hydrolysis of ␣Glc 1-P. Zbasic2 HAD4 is proposed to be a
useful catalyst for hydrolytic transformation (e.g. removal) of ␣Glc
1-P from complex substrate solutions containing multiple sugar
phosphates. Immobilization of Zbasic2 HAD4 on porous carriers con-
taining surface sulfonate groups provides a recyclable and highly
effective heterogeneous biocatalyst. Zbasic2 HAD4 complements the
current portfolio of phosphatases for use in biocatalysis by adding
new characteristics of specificity and selectivity.
[25] T. Graslund, G. Lundin, M. Uhlén, P.-A. Nygren, S. Hober, Protein Eng. 13 (2000)
703–709.
[26] E. Samuelsson, T. Moks, M. Uhlen, B. Nilsson, Biochemistry 33 (1994)
4207–4211.
[27] J.M. Bolivar, B. Nidetzky, Biotechnol. Bioeng. 109 (2012) 1490–1498.
[28] J.M. Bolivar, B. Nidetzky, Langmuir 28 (2012) 10040–10049.
[29] M.J. Lefebvre, N.S. Gonzalez, H.G. Pontis, J. Chromatogr. A 15 (1964) 495–500.
Acknowledgments
We would like to acknowledge financial support from the Aus-
trian Science Funds FWF (project L586-B03); Alexander Yakunin
(Department of Chemical Engineering and Applied Chemistry,