Purification of an l-arabinose isomerase from Enterococcus faecium DBFIQ E36 employing a biospecific affinity strategy

Article abstract of DOI:10.1016/j.molcatb.2014.01.023

l-Arabinose isomerase is an intracellular enzyme that can convert l-arabinose to l-ribulose and d-galactose to d-tagatose, a promising but rare nutraceutical. Most of l-arabinose isomerases purified up to date employed the combination between DNA recombinant technology and affinity chromatography based on poly-histidine tail recognition, but few of the enzymes were obtained and purified in a non-recombinant way. For these reasons, a specific affinity bioadsorbent containing l-arabitol as ligand, a competitive inhibitor of the enzyme, was designed and synthesized for achieving pure preparations of the enzyme l-arabinose isomerase from wild-type Enterococcus faecium DBFIQ E36 strain, isolated from raw cow milk. The two-step purification procedure consisted in fractionation by ammonium sulphate precipitation followed by affinity chromatography with obtained bioadsorbent, allowing the purification, to electrophoretical homogeneity, of target enzyme. Characterization studies were performed with purified l-arabinose isomerase in order to increase knowledge of their physicochemical properties. In this sense, enzyme exhibited an optimum temperature of 50 C and optimum pH of 7.0, maintaining good stability in the ranges 20-45 C and pH 6.5-8. K_i were calculated, employing d-galactose as substrate, for l-arabitol and l-ribitol, achieving values of 7.9 mM and 183 mM, respectively. K_m and V_max values obtained were 35 mM and 81 U mg^-1 at 50 C, respectively. Mass spectrometry assay revealed a 48 kDa monomer whereas gel permeation chromatography achieved a 187 kDa molecular weight for native enzyme. Finally, 2D-electrophoresis and isoelectrofocusing analysis revealed an isoelectric point value of 3.80. Results have unveiled both an acidic nature and promising properties for l-arabinose isomerase isolated from E. faecium DBFIQ E36.

Full text of DOI:10.1016/j.molcatb.2014.01.023

Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Purification of an l-arabinose isomerase from Enterococcus faecium
DBFIQ E36 employing a biospecific affinity strategy
Pedro R. Torresa,1, Ricardo M. Manzob,2, Amelia C. Rubiolob,2
,
Francisco D. Batista-Vieraa,1, Enrique J. Mammarellab,∗
a Cátedra de Bioquímica, Departamento de Biociencias, Facultad de Química, Universidad de la República, Avenida Gral. Flores 2124, CP 11800 Montevideo,
Uruguay
b Grupo de Ingeniería de Alimentos y Biotecnología, Instituto de Desarrollo Tecnológico para la Industria Química, Universidad Nacional del Litoral (UNL),
Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Colectora RN 168 Km 472 “Paraje El Pozo” S/N, S3000GLN Santa Fe, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 October 2013
Received in revised form 3 January 2014
Accepted 25 January 2014
Available online 13 February 2014
l-Arabinose isomerase is an intracellular enzyme that can convert l-arabinose to l-ribulose and d-
galactose to d-tagatose, a promising but rare nutraceutical. Most of l-arabinose isomerases purified up
to date employed the combination between DNA recombinant technology and affinity chromatography
based on poly-histidine tail recognition, but few of the enzymes were obtained and purified in a non-
recombinant way. For these reasons, a specific affinity bioadsorbent containing l-arabitol as ligand, a
competitive inhibitor of the enzyme, was designed and synthesized for achieving pure preparations of
the enzyme l-arabinose isomerase from wild-type Enterococcus faecium DBFIQ E36 strain, isolated from
raw cow milk.
Keywords:
l-Arabinose isomerase
d-Tagatose
The two-step purification procedure consisted in fractionation by ammonium sulphate precipitation
followed by affinity chromatography with obtained bioadsorbent, allowing the purification, to elec-
trophoretical homogeneity, of target enzyme. Characterization studies were performed with purified
l-arabinose isomerase in order to increase knowledge of their physicochemical properties. In this sense,
enzyme exhibited an optimum temperature of 50 ◦C and optimum pH of 7.0, maintaining good sta-
bility in the ranges 20–45 ◦C and pH 6.5–8. Ki were calculated, employing d-galactose as substrate, for
l-arabitol and l-ribitol, achieving values of 7.9 mM and 183 mM, respectively. Km and Vmax values obtained
were 35 mM and 81 U mg−1 at 50 ◦C, respectively. Mass spectrometry assay revealed a 48 kDa monomer
whereasgel permeation chromatography achieveda187 kDamolecularweightfornativeenzyme. Finally,
2D-electrophoresis and isoelectrofocusing analysis revealed an isoelectric point value of 3.80. Results
have unveiled both an acidic nature and promising properties for l-arabinose isomerase isolated from E.
faecium DBFIQ E36.
d-Galactose
Affinity chromatography
Enterococcus faecium strain
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
is of great interest for both technological and nutritional
reasons, and the development of many alternative processes for
l-Arabinose isomerase (EC 5.3.1.4) is an intracellular enzyme
that catalyses, in vivo, the isomerization of l-arabinose to l-ribulose
[7,31,45]. Besides, in vitro experiments have shown that this bio-
catalyst can also convert d-galactose to d-tagatose, a rare sugar
with promising nutraceutical properties [3,14]. This ketohexose
obtaining it, apart from chemical synthesis [1], has been stud-
ied [10–13,28,29,38,40,46,47]. In this way, biological conversion
of d-galactose to d-tagatose employing the enzyme l-arabinose
isomerase is the most economically viable biological d-tagatose
manufacturing process because d-galactose can be readily obtained
from cheese whey, an industrial by-product produced during
cheese elaboration [9,16,24,41]. Biotechnological processes for the
production of d-tagatose requires the development of easy purifi-
cation methods with potential application to obtain l-arabinose
isomerases from safe sources.
∗
Corresponding author. Tel.: +54 342 4511546x1077; fax: +54 342 4511170.
E-mail addresses: fbatista@fq.edu.uy (F.D. Batista-Viera), ejoma@intec.unl.edu.ar
(E.J. Mammarella).
Many l-arabinose isomerases have been characterized nowa-
days, from different bacterial species, mainly thermophilic
1
Tel.: +598 2 9241806; fax: +598 2 9241906.
Tel.: +54 342 4511546x1077; fax: +54 342 4511170.
2
http://dx.doi.org/10.1016/j.molcatb.2014.01.023
1381-1177/© 2014 Elsevier B.V. All rights reserved.

100
P.R. Torres et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
or hyperthermophilic organisms [4,15,17,18,20,23–26,35]. l-
Arabinose isomerases from extremophiles, present better isomer-
ization rate and conversion yield but, on the other hand, are difficult
to obtain by using non-recombinant methods, mainly due to the
intrinsic difficulties in culture of these microorganisms. Enzyme
production by mesophilic bacteria was studied in lower exten-
sion [19,30,34,36,39,48,50], but recently the fact that most of them
are GRAS organisms, which possess better technological applica-
bility for manufacturing either food products or can be utilizing as
food ingredients, increased the interest in these l-arabinose sources
[21,37].
Certain conventional protocols for purification of l-arabinose
isomerases from non-recombinant sources have been studied in
several mesophilic bacteria such as Escherichia coli, Lactobacil-
lus plantarum, Enterobacter aerogenes, Lactobacillus gayonii and
Mycobacterium smegmatis. These purification procedures have
involved many sequential and time-consuming steps that include
salting out, heat treatments and many chromatographic tech-
niques such as ion-exchange, gel filtration, HPLC and hydrophobic
interaction chromatography. However, most of the technologi-
cally relevant l-arabinose isomerases studied up to date were
achieved as heterologous recombinant fusion proteins in which
a tag sequence is introduced and acts as target for affinity
support recognition and subsequent purification [15,24,37]. In
that way, these affinity techniques employed to quickly purify
l-arabinose isomerase did not use a specific property of the
mentioned enzyme. Biospecific affinity purification procedures
were not usually reported for l-arabinose isomerases, although
other keto-isomerases (e.g. glucose isomerase and fucose iso-
merase) were successfully purified by this strategy. Finally,
others l-arabinose isomerases were produced by recombinant
technology but not as fusion proteins, so they were conven-
tionally purified employing traditional purification techniques
[5].
2.2. Enzyme production
2.2.1. Bacterial strain and culture preservation
E. faecium was chosen as l-arabinose source according to Manzo
et al. report [27]. Long-term preservation was done by lyophiliza-
tion, in order to obtain safe and stable stock culture. Enterococcus
strain was frozen at −20 ◦C and −80 ◦C in MRS supplemented with
15% (v/v) glycerol [8].
2.2.2. Cell-free extract production
The initial culture was prepared by adding 0.2 mL of frozen stock
culture to 5 mL of MRS broth and incubating for 24 h at 37 ◦C. This
culture was transferred to 200 mL of modified MRS broth with
the following composition in % (w/v): 1% meat extract; 0.5% yeast
extract; 1% peptone; 0.2% K2HPO4; 0.02% MgSO4·7H2O; 0.005%
MnSO4·4H2O; 0.2% triammonium citrate; 0.5% sodium acetate
trihydrate; 0.108% polysorbate 80; 0.1% d-glucose and 0.5% l-
arabinose and incubated at 37 ◦C for 24 h. The final culture was done
by adding 180 mL of the last propagation culture in the late expo-
nential growth phase to 6000 mL of modified MRS broth [3% (v/v)
inocula] and allow growing at 37 ◦C for 24 h [27]. Cells obtained
were harvested by centrifugation at 4000 × g for 20 min at 4 ◦C,
resuspended in 500 mL of 50 mM phosphate buffer pH 7.0 (activity
buffer) and treated with 1 mg/mL lysozyme for 3 h at 37 ◦C. Then,
the cell suspension was disrupted with ultrasonic vibrations (60 W;
20 kHz; M.S.E.-Mullard Ultrasonic Disintegrator, Mullard Ltd., Lon-
don, UK) for a period of 40 min at 4 ◦C and, afterwards, remaining
cell debris was removed by centrifugation at 16,000 × g for 30 min
at 4 ◦C. Cell-free extract was sterilized by filtration using 0.22 ␮m
pore diameter membranes (Sartorius AG, Göttingen, Germany),
alicuoted and stored at −20 ◦C.
2.2.3. Enzyme assays
l-Arabinose isomerase activity was determined by measure-
ment of the amount of d-tagatose generated from d-galactose. The
reaction mixture contained 1 mM MnCl2, 500 mM of d-galactose,
200 ␮L of enzyme preparation properly diluted and 50 mM phos-
phate buffer pH 7.0 (activity buffer) to bring the final volume to
1 mL. The assay was done by incubating the reactive mixture and
enzyme at 50 ◦C for 1 h. Subsequently, enzymatic reaction was
stopped by cooling the samples on ice. The generated d-tagatose
was determined spectrophotometrically at 560 nm by the cysteine
carbazole sulfuric acid method [6]. One unit of l-arabinose iso-
merase activity was defined as the amount of enzyme catalyzing
the formation of 1 ␮mol of keto-sugar per min under the specified
conditions. Protein concentration was determined by the bicin-
choninic acid method employing BSA as standard [43]. Additionally,
the Ki for l-arabitol and l-ribitol (substrate and product analogues,
respectively) were determined in the range 0–250 mM performing
the enzymatic assay with d-galactose as substrate in the presence
of different concentrations of l-arabitol or l-ribitol in the reactive
mixture.
In a previous paper authors reported the screening of bacte-
rial strains isolated from raw cow milk in relation to its potential
to produce l-arabinose isomerase [27]. As a result, Enterococcus
faecium DBFIQ E36 was selected as one of the probe strains for
enzyme production attending to its characteristics. In this work,
a fast and reliable two-step purification procedure employing
a specially designed affinity chromatography adsorbent with l-
arabitolas ligand, acompetitiveinhibitor ofthe enzyme l-arabinose
isomerase, was developed for achieving pure preparations of
this enzyme from a cell-free extract obtained from a GRAS and
non-recombinant E. faecium DBFIQ E36 strain. Different physic-
ochemical properties of the enzyme l-arabinose isomerase such
as molecular weight, isoelectric point, effect of relevant putative
inhibitors and activators and quaternary structure have been deter-
mined.
2. Materials and methods
2.3. l-Arabinose isomerase purification
2.1. Materials, sugars and culture media
2.3.1. Fractionation with ammonium sulphate
MRS broth was provided by Difco Laboratories (Detroit, MI,
USA). Sepharose 4B, PD-10 (Sephadex G 25) columns, and empty
NAP-5 columns were purchased from GE Healthcare (Upp-
sala, Sweden). Epiclorhydrin, l-arabitol, d-galactose, d-tagatose
l-ribitol, d-arabinose, l-arabinose, cysteine hydrocloride, car-
bazole and bovine serum albumine (BSA) were purchased from
Sigma–Aldrich (St. Louis, MO, USA). 2-Mercaptoethanol was
purchased from Bio-Rad Laboratories (Hercules, CA, USA), bicin-
choninic acid (BCA) reagent from Pierce (Rockford, IL, USA), and all
other chemicals were of analytical grade.
The protein of cell-free extract was fractionated by adding solid
(NH4)2SO4 at 85% saturation, followed by overnight incubation at
4 ◦C. This fraction was centrifuged at 12,000 × g during 30 min at
4 ◦C and the pellet was dissolved in activity buffer and gel filtered
employing a Sephadex G-25 PD-10 column (GE Healthcare, Upp-
sala, Sweden).
2.3.2. Affinity purification
For the application of this technique, the affinity adsorbent was
developed. For activation of the support, 12 g of Sepharose 4B beads

P.R. Torres et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
101
were suspended in 2 M NaOH. Immediately, 0.2 g NaBH4 and 2 mL
epichlorohydrin were added to the mixture and incubated during
24 h at room temperature with continuous stirring. Afterwards,
suspension was washed sequentially with 2 M NaOH and water
until pH 7.0 was achieved. In order to determine the epoxy group
content of activated gel, 4 g of epoxy-activated Sepharose was sus-
pended in 10 mL of deionized water and pH was adjusted to 7.0.
Then, 15 mL of 1 M Na2S2O3 were added and allowed to react for 1 h
with continuous stirring. Finally, suspension was back-titrated with
60 mM HCl employing phenolphthalein as indicator. The amount
of epoxy groups present is calculated from the amount of HCl
required to reach pH 7.0 [42]. Coupling of ligands (l-arabitol, l-
ribitol and d-arabinose) was carried out by incubating overnight at
room temperature, with permanent stirring, 8 g of epoxy-activated
gel in 25 mL of 10% (w/v) l-arabitol dissolved in 0.5 M NaOH. The
gel derivative was filtered in sintered glass filter and washed with
0.5 M NaOH. Then, remaining epoxy groups were blocked by adding
200 ␮L of 2-mercaptoethanol in 25 mL of 0.5 M NaOH and incubated
for 1 h at room temperature. Affinity adsorbent was washed with
0.5 M NaOH and water until neutrality and stored at 4 ◦C in 100 mM,
pH 7.5 phosphate buffer.
Subsequently, l-arabitol-agarose affinity adsorbent was
employed for l-arabinose isomerase purification. For this purpose,
2 mL of affinity derivative was packed in a NAP-5 column and
equilibrated with 100 mM phosphate buffer pH 7.5. Gel filtered
dissolved pellet was applied to column either manually (protocol
1, discontinuous mode) or with a peristaltic pump (protocol
2, continuous mode) and incubated (protocol 1) or recycled at
0.5 mL/min (protocol 2) during 2 h at room temperature. Then, the
pass-through fraction was reserved and the column was washed
with 2 volumes of the same activity buffer. Elution was performed
with 30 mM of l-arabitol in activity buffer. Finally, samples were
gel filtered, in order to eliminate competitive inhibitor. All samples
achieved after each purification step were monitored by determin-
ing absorbance at 280 nm; activity and protein assays of relevant
fractions were performed.
to 150,000 m/z lineal range using a 200 Hz laser and a wave-
length of 355 nm. Peptide mass fingerprinting of protein selected
spots was carried out by in-gel sequencing-grade modified trypsin
treatment (Promega, Fitchburg, WI, USA) overnight at 37 ◦C. Pep-
tides were extracted from the gels using 60% (v/v) acetonitrile in
0.2% (v/v) trifluoroacetic acid (TFA), concentrated by vacuum dry-
ing and desalted using C18 reverse phase micro-columns (OMIX
Pippete tips, Varian, Palo Alto, CA, USA). Peptide elution from
micro-column was performed directly into the mass spectrometer
sample plate with 3 ␮L of matrix solution (5 mg/mL of ␣-cyano-4-
hydroxycinnamic acid in 60% (v/v) aqueous acetonitrile containing
0.2% (v/v) TFA.
Similarly, mass spectra of digestion mixtures were acquired in
a 4800 MALDI-TOF/TOF instrument in reflector mode and were
externally calibrated using a mixture of peptide standards (Applied
Biosystems, Foster City, CA, USA) including des-Arg1-Bradykinin
(m/z 904), Angiotensin I (m/z 1296), Glu1-fibrinopeptide B (m/z
1570), ACTH (1–17, m/z 2093), ACTH (18–39, m/z 2465) and ACTH
(7–38, m/z 3657). Collision-induced dissociation MS/MS experi-
ments of selected peptides were performed. Resulting MS and
MS/MS spectres were searched with MASCOT software and com-
pared against the available protein sequence database [33].
2.4.3. Gel-filtration chromatography for quaternary structure
determination
In order to estimate the molecular weight of the native enzyme,
6 mL of the purified preparation were applied to a preparative col-
umn containing the resin Superdex 200, previously equilibrated
with 50 mM phosphate buffer pH 7.0. Column was previously cal-
ibrated employing high-molecular-weight standard kit (HMW kit,
GE Healthcare), where elution volumes of each calibration protein
were determined. Elution of the enzyme was performed in the same
equilibration buffer and followed measuring the absorbance at
280 nm. Partition coefficient was calculated and molecular weight
of the sample was determined from calibration curve.
2.4.4. Properties of soluble enzyme
2.4. Primary characterization
The influence of pH and temperature on the activity and stability
of enzyme was studied. Determination of kinetic parameters (Km
and Vmax) in the range 30–50 ◦C was also performed.
2.4.1. Electrophoresis and isoelectric focusing
Native polyacrylamide gel electrophoresis (PAGE) was per-
formed on an homogeneous 12.5% Polyacrylamide PhastGelTM
gel. Besides, denaturing polyacrylamide gel electrophoresis (SDS-
PAGE) was done on a 8–25% T polyacrylamide PhastGelTM gradient
gel and on an homogeneous 12.5% T polyacrylamide PhastGelTM gel
employing the PhastSystem equipment (Pharmacia LKB, Uppsala,
Sweden). For molecular weight determination, low-molecular-
weight protein markers (14.4–94 kDa) for SDS-PAGE from GE
Healthcare was used. PAGE and SDS-PAGE were made following the
protocols described by [22]. Isoelectric focusing (IEF) was also per-
formed on Polyacrylamide PhastGelIEF media (gradient, pH 3–10)
employing the broad pI calibration kit (3.5–9.3) from GE Health-
care, according to the manufacturer’s instructions. For 2D-PAGE,
IEF was performed as previously described. Prior to the second
dimension, the IEF strips were equilibrated by incubation for 15 min
in a solution containing 2% (w/v) SDS, 50 mmol/L Tris–HCl (pH
8.8), 30% (v/v) glycerol, and 1% (w/v) 2-mercaptoethanol. Follow-
ing equilibration, the strips were placed on 12.5% T SDS-PAGE gels
and electrophoresis was made using the PhastSystem apparatus
according to manufacturer’s instructions. The gels were visualized
by silver staining.
2.4.5. Effect of temperature on activity
The influence of temperature on the activity of enzyme was
determined in the range of 20–70 ◦C in activity buffer using d-
galactose as substrate.
2.4.6. Thermal stability
The thermostability assays were performed at different tem-
peratures in the range 20–70 ◦C in activity buffer, by incubating
enzyme in a shaking bath. The inactivation kinetics of the enzyme
was monitored for 1–24 h by monitoring residual activity employ-
ing the enzyme assay described above. Half-lives were calculated
according to Arrhenius model.
2.4.7. Optimum pH and pH stability
The influence of pH on the activity of enzyme was studied in
the range of pH 4–9 at 25 ◦C, using d-galactose as substrate. The
effect of pH on the stability of the enzyme was determined in the
same range by measuring the residual activity after incubating the
biocatalysts for 1–10 h at 25 ◦C.
2.4.2. Mass spectrometry
2.4.8. Determination of kinetic parameters
Mass spectrum of whole protein was acquired in a 4800 MALDI
TOF/TOF (Ab Sciex, Framingham, Massachusetts, USA) in lineal
mode (lineal High) and using sinapinic acid as matrix from 10,000
The kinetic parameters of enzyme were determined with l-
arabinose (at 50 ◦C) and d-galactose (at 30–50 ◦C) 0–500 mM in
activity buffer (pH 7.0). The parameters were calculated according

102
P.R. Torres et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
90
80
70
60
50
40
30
20
10
0
900
0 mM
800
3.5 mM
7 mM
20 mM
700
600
500
400
300
200
100
35 40 45 50 55 60 65 70 75 80 85 90 95 100 105
0.0010
0.0015
0.0020
0.0025
0.0030
0.0035
0.0040
Ammonium sulphate saturation (%)
1/[D-galactose] (1/mM)
Fig. 1. Enzymatic activity recovery after the precipitation of the protein contained
in cell-free extract with different ammonium sulphate concentrations. The results
are the average of the assays performed in triplicate.
Fig. 2. Lineweaver–burk plots for l-arabitol Ki determination. l-Ribitol graphic is
not shown. The results are the average of the assays performed in triplicate.
activation degree of 43 ␮mol of epoxy groups per gram of filter-
dried gel. The support activation degree was selected in order to
achieve an l-arabitol concentration that has assured an adequate
amount of available ligand for l-arabinose isomerase interaction
with the bioadsorbent and, simultaneously reduced undesirable
steric restrictions when enzyme is being purified by affinity chro-
matography.
to Michaelis–Menten model, applying the Eadie–Hofstee lineariza-
tion procedure.
3. Results and discussion
3.1. Fractionation by ammonium precipitation
As seen in Fig. 1, salting out screening study was performed
employing ammonium sulphate saturation percentages from 40 to
100%. After salting out of crude enzyme extracts from E. faecium
DBFIQ E36 with ammonium sulphate at 60 and 80% saturation,
low l-arabinose isomerase recoveries were achieved (9 and 21%
respectively). A dramatic increase in enzyme recovery up to 78%
was reached at saturation percentages equal or above 85%. How-
ever, an increase of this percentage, even at 100% of salt saturation,
did not increase this recovery, probably because part of the enzyme
activity was present, as enzymatic aggregates, in the foam pro-
duced during protein precipitation so, unfortunately, this enzyme
could not be recovered completely. In addition, these results clearly
reveal the hydrophilic nature of target enzyme owe to the need of a
high ammonium sulphate concentration for l-arabinose isomerase
salting-out, in concordance with the acidic isoelectric point deter-
mined by means of isoelectric focusing. Furthermore, crude enzyme
extract had a poor enzyme amount so, salting out step, has allowed
concentrating it, which represents an additional advantage when
this methodology was applied.
3.3. l-Arabinose isomerase two-step purification
To study an enzyme, a highly purified native form of the protein
must be achieved. Although, a major protein is not so difficult to
be purified, a minor one may need several purification steps and
high skills on the manipulation of the techniques. In that sense,
the methods of purification vary from protein to protein, mak-
ing impossible to design a general purification strategy valid for
all cases. However, in this particular case, most of l-arabinose iso-
merases reported far are strongly inhibited by l-arabitol [21].
In the adopted strategy, l-arabinose isomerase was purified
to homogeneity from a crude extract of E. faecium DBFIQ E36
through a two-step procedure including ammonium sulphate pre-
cipitation followed by biospecific affinity chromatography using an
l-arabitol-agarose adsorbent and employing two different purifica-
tion protocols. As shown in Table 1, protocol included fractionation
by ammonium sulphate precipitation at 85% saturation followed
by the affinity chromatography step. An increase of only 5% in
(NH4)2SO4 saturation (85%) allowed the recovery of 78 of total
activity % (against the 21% recovery when precipitation was car-
ried out with 80% saturation) with a slight increase in protein
quantity. The selected protocol has made possible to achieve a 4.4
purification fold value, as seen in Table 1. Activity recovery results
from salting out step was one of the highest reported up to date,
being only surpassed by those reported by [7,48] which obtained
an 83% and a 79% yield, respectively, from conventional purification
of l-arabinose isomerase from L. plantarum. None of the scientific
reports which purified the enzyme from recombinant sources and
employed salting out as part of their purification protocols achieved
values as high as described above [5,17,35].
3.2. l-Arabinose isomerase bioadsorbents
Affinity chromatography uses specific binding interactions
between biomolecules and immobilized ligands. In that way, for
a given research purpose, each specific affinity system will require
their own set of conditions revealing its own peculiar challenges.
The inhibitory effects of l-arabitol (substrate analogue) and
l-ribitol (product analogue) were studied. The two polyols pro-
duced the competitive inhibition of the enzyme. However, l-ribitol
showed a very high value of Ki (183 mM) compared with l-arabitol
Ki (7.9 mM) when d-galactose was used as substrate, indicating a
high affinity that could be exploited for rational affinity bioadsor-
bent design (Fig. 2).
Besides, a 44% of total activity was achieved when apply-
ing a continuous elution mode which depicts more than twice
activity yield percentage in contrast with batch elution mode proto-
col. However, activity recovery after affinity chromatography step
decayed at 56% when comparing with ammonium sulphate step
which could be attributed to partial loss of activity by enzyme
In that way, a bioadsorbent based on covalent linkage of l-
arabitol to an epoxy-activated agarose matrix was developed
(Fig. 3). The agarose matrix (Sepharose 4B) had undergone an

P.R. Torres et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
103
Fig. 3. Graphic summary of the synthesis of l-arabitol-agarose bioadsorbent (specific position of linkage in l-arabitol-agarose is uncertain).
inactivation, probably due to the demanding conditions at which
the enzyme was exposed when interaction with the bioadsorbent
had occurred.
Besides, it can be seen that this band is not present in the pass-
through fraction.
Furthermore, Table 1 shows that l-arabinose isomerase was
purified between 609 and 720 folds in relation with cell-free
extract, respectively. A single active fraction was obtained after
elution with 35 mM l-arabitol which means that the enzyme was
found at low concentrations and was purified to homogeneity
from a complex cell-free extract containing many undesirable pro-
teins and other substances, which are present in bigger quantities.
These purification steps were only managed from cell-free extracts
obtained from wild-type bacteria and purified to homogeneity
with conventional techniques as described by [7] and [48]. Besides,
the purification factors reported were the highest obtained up to
date and this could be possible by means of employing a highly
specific interaction between the enzyme and its natural compet-
itive inhibitor, l-arabitol. This also reveals that cell-free extract
enzyme concentration is low being necessary to be increased either
by improving lytic methods or by culture medium optimization.
Moreover, the developed bioadsorbent was regenerated after each
purification procedure and reused efficiently for a long time.
Fractions containing l-arabinose isomerase activity obtained by
ammonium sulphate precipitation were also applied onto columns
containing d-arabinose-agarose and l-ribitol-agarose bioadsor-
bents based on the covalent immobilization of these ligands,
representing similar structures to l-arabinose and l-ribulose,
respectively (data not shown). Results have showed that only the
l-arabitol-agarose bioadsorbent allowed a successful purification
of the enzyme. Furthermore, preliminary kinetics studies indicated
that enzyme was not inhibited by its product, fact that could explain
the weak bioaffinity adsorption on l-ribitol-agarose, owe to the
similarities in their chemical structures.
3.5. Isoelectrofocusing and 2D-electrophoresis
The pI value turned out to be 3.8 (Fig. 4B), being the lowest of the
l-arabinose isomerases reported up to date [25,37]. The acidic pI,
generated by the presence of a great number of acidic residues, was
in accordance with the high ammonium sulphate concentration
needed for salting out of enzyme.
The 2D-electrophoresis has proved that two-step purification
method allowed obtaining a single band around 56 kDa with an iso-
electric point of 3.80. As previously mentioned, IEF also has shown
same isoelectric point as 2D.
3.6. Molecular weight determinations
MALDI-TOF pattern showed a putative monomer of about
48 kDa. The peaks visualized up to 150,000 Da in molecular weight
were assigned to monomer, dimer and trimer structure (Fig. 5). The
difference between mass values estimated by electrophoretic anal-
ysis and mass spectrometry (about 8 kDa) could be due to distortion
in electrophoretic behaviour of protein, as reported in literature
also for other enzymes and proteins [49]. The presence of a sin-
gle peak in gel permeation chromatography on Superdex 200 has
shown the homogeneity of the purified enzyme obtained by affin-
ity chromatography. The mass of the native enzyme determined by
this procedure was 187 kDa, in congruence to that calculated from
MALDI-TOF, supposing a homotetrameric structure (Fig. 5). Results
of MALDI-MS rendered no match in MASCOT database.
3.7. Enzyme properties
3.4. Native and SDS-PAGE analysis
3.7.1. Effects of pH on activity and stability
The effects of pH on activity and stability were determined. l-
Arabinose isomerase showed an optimum at pH 7.0, and activity at
acidic pH (70% of maximal activity at pH 4.5) was better than activ-
ity at alkaline pH (data not shown). The pH stability profile for the
enzyme was similar in the pH range 4.0–6.0, with maximum sta-
bility at pH 7.0–7.5. The stability decreased moderately at alkaline
pH (data not shown). Enzyme retained more than 60% of its activ-
ity when incubated 6 h at pH 6.5–8.5. Thus, enzyme presented in
this work showed good activity and stability at pH near neutrality
and slightly acidic pH, which is technologically important because
by-product formation was minimized at acidic pH [5].
Native PAGE electrophoresis was performed on purified l-
arabinose isomerase. A single band was seen and the presence of
higher molecular weight aggregates was not detected (data not
shown). In their native state, this enzyme has been characterized
as tetramer [32] and hexamer [44] also being found more rarely
as a dimer [37]. Furthermore, l-arabinose isomerase was purified
to electrophoretical homogeneity with a single band of 56 kDa on
12.5% T SDS-PAGE gels by affinity chromatography using the syn-
thesized adsorbent, as shown in Fig. 4A. All l-arabinose isomerases
from different organisms previously characterized as monomers,
were found in the molecular weight range from 52 to 60 kDa [21].
Table 1
Two-step purification of l-arabinose isomerase from E. faecium DBFIQ E36.
Fraction
Volume (mL)
Total protein (mg)
Total activity (U)a
Specific Activity (U mg−1
)
Purification fold
Yield (%)
Cell-free extract
85% ammonium sulphate precipitation
Affinity chromatography
73
16
1
56
10
0.034
14
11
6.2
0.25
1.1
180
1
4.4
720
100
78
44
a
Total activity was multiplied per 1 × 103. The results are the average of the assays performed in triplicate.

104
P.R. Torres et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
Fig. 4. (A) SDS-PAGE analysis on a 12.5% T polyacrylamide precast gel of different fractions obtained after enzyme purification. Lane 1: redissolved 85% ammonium sulphate
precipitate; lane 2: pass-through fraction from affinity chromatography; lane 3: eluted fraction with 30 mM l-arabitol solution; lane 4: molecular mass standards (LMW
Marker, GE Healthcare). (B) Isoelectric focusing of the purified E. faecium DBFIQ E36 l-arabinose isomerase. Lane 1, standard pI calibration kit; lane 2, affinity chromatography
purified enzyme. For assays, 4 ␮L of each sample were employed. Gels were silver stained according to Blum protocol [2].
Fig. 5. (A) Elution chromatogram obtained after application of two-step purified l-arabinose isomerase to a Superdex 200 exclusion chromatography column. (B) Gel
filtration selectivity curve (red circle: l-arabinose isomerase). Molecular weight standards employed were: ribonuclease (13.7 kDa), egg white albumine (43 kDa), bovine
serum albumine (67 kDa), aldolase (158 kDa), catalase (232 kDa), ferritin (440 kDa), thyroglobuline (669 kDa). (C) Mass spectra of pure l-arabinose isomerase.
3.7.2. Optimum temperature and thermal stability
the substrate, an advantage considering the applications of these
biocatalysts. The low value of Km for d-galactose reinforces the
technological interest of this enzyme. Significant differences were
observed also in Vmax (Vmax for d-galactose at 30 ◦C was 8 times
l-Arabinose isomerase activity showed an optimum at 50 ◦C,
and activity declined over 60 ◦C (data not shown). The thermal
stability profile exhibited that enzyme possesses good stability
at 45–50 ◦C, with half-life of 10 h at 50 ◦C. The stability strongly
decreased in the range 55–70 ◦C (data not shown).
Table 2
Kinetic parameters of l-arabinose isomerase from E. faecium DBFIQ E36 at different
temperatures and employing l-arabinose and d-galactose as substrate.
Substrate
Temperature (◦C)
Km (mM)
Vmax (U mg−1
)
3.7.3. Kinetic parameters
The Km for the enzyme was determined at 30 ◦C, 40 ◦C and
50 ◦C using d-galactose as substrate (Table 2). As expected, Km
decreased when temperature increased following the general ten-
dency. Besides, when temperature was increased from 30 ◦C to
50 ◦C, decreased almost 3-fold, thus improving the affinity for
l-Arabinose
d-Galactose
d-Galactose
d-Galactose
50
30
40
50
42
101
71
30
11
41
81
35
The results are the average of the assays performed in triplicate.

P.R. Torres et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 99–105
105
lower than Vmax at 50 ◦C). In a comparative way, kinetic parame-
ters for l-arabinose (the natural substrate of enzyme) were also
determined at 50 ◦C utilizing the same methodology (Table 2).
Results revealed that Km for d-galactose was approximately 3-fold
higher than the corresponding Km for l-arabinose. Nevertheless, the
enzyme from E. faecium exhibited kinetic advantages with respect
to other l-arabinose isomerases reported far in terms of d-galactose
biotransformation [4].
[2] H. Blum, H. Beier, H.J. Gross, Electrophoresis 8 (1987) 93–99.
[3] P.S.J. Cheetham, A.N. Wootton, Enzyme Microb Technol 15 (1993) 105–108.
[4] L. Cheng, W. Mu, T. Zhang, B. Jiang, Appl Microbiol Biotechnol 86 (2010)
1089–1097.
[5] H. Chouayekh, W. Bejar, M. Rhimi, K. Jelleli, M. Mseddi, S. Bejar, FEMS Microbiol
Lett 277 (2007) 260–267.
[6] Z. Dische, E. Borenfreund, J Biol Chem 192 (1951) 583–587.
[7] E.C. Heath, B.L. Horecker, P.Z. Smyrniotis, Y. Takagi, J Biol Chem 231 (1958)
1031–1037.
[8] R.J. Heckly, Adv Appl Microbiol 24 (1978) 1–53.
[9] O.O. Ibrahim, J.E. Spradlin, Process for manufacturing d-tagatose, US Patent
6,057,135, 2000.
[10] Y. Ishida, T. Kamiya, H. Itoh, Y. Kimura, K. Izumori, J Ferment Bioeng 83 (1997)
529–534.
4. Conclusion
[11] H. Itoh, T. Sato, T. Takeuchi, A.R. Khan, K. Izumori, J Ferment Bioeng 79 (1995)
184–185.
[12] K. Izumori, T. Miyoshi, S. Tokuda, K. Yamabe, Appl Environ Microbiol 46 (1984)
1055–1057.
[13] K. Izumori, K. Tsuzaki, J Ferment Bioeng 66 (1988) 225–227.
[14] K. Izumori, Y. Ueda, K. Yamanaka, J Bacteriol 133 (1978) 413–414.
[15] F. Jørgensen, O.C. Hansen, P. Stougaard, Appl Microbiol Biotechnol 64 (2004)
816–822.
[16] B.C. Kim, Y.H. Lee, H.S. Lee, D.W. Lee, E.A. Choe, Y.R. Pyun, FEMS Microbiol Lett
212 (2002) 121–126.
[17] H.J. Kim, J.H. Kim, H.J. Oh, D.K. Oh, J Appl Microbiol 101 (2006) 213–221.
[18] H.J. Kim, D.K. Oh, J Biotechnol 120 (2005) 162–173.
[19] J.H. Kim, P. Prabhu, M. Jeya, M.K. Tiwari, H.J. Moon, R.K. Singh, J.K. Lee, Appl
Microbiol Biotechnol 85 (2010) 1839–1847.
[20] J.W. Kim, Y.W. Kim, H.J. Roh, H.Y. Kim, J.H. Cha, K.H. Park, C.S. Park, Biotechnol
Lett 25 (2003) 963–967.
[21] P. Kim, Appl Microbiol Biotechnol 65 (2004) 243–249.
[22] U.K. Laemmli, Nature 227 (1970) 680–685.
[23] D.W. Lee, E.A. Choe, S.B. Kim, S.H. Eom, Y.H. Hong, S.J. Lee, H.S. Lee, D.Y. Lee, Y.R.
Pyun, Arch Biochem Biophys 434 (2005) 333–343.
[24] D.W. Lee, H.J. Jang, E.A. Choe, B.C. Kim, S.J. Lee, S.B. Kim, Y.H. Hong, Y.R. Pyun,
Appl Environ Microbiol 70 (2004) 1397–1404.
[25] S.J. Lee, D.W. Lee, E.A. Choe, Y.H. Hong, S.B. Kim, B.C. Kim, Y.R. Pyun, Appl Environ
Microbiol 71 (2005) 7888–7896.
[26] Y. Li, Y. Zhu, A. Liu, Y. Sun, Extremophiles 15 (2011) 441–450.
[27] R.M. Manzo, A.C. Simonetta, A.C. Rubiolo, E.J. Mammarella, Braz J Chem Eng 30
(4) (2013) 711–720.
[28] M. Manzoni, M. Rollini, Process Biochem 36 (2001) 971–977.
[29] S. Muniruzzanman, H. Tokunaga, K. Izumori, J Ferment Bioeng 78 (1994)
145–148.
[30] J. Patrick, N. Lee, Methods Enzymol 41 (1975) 453–458.
[31] J.W. Patrick, N. Lee, J Biol Chem 243 (1968) 4312–4318.
[32] J.W. Patrick, N. Lee, J Biol Chem 244 (1969) 4277–4283.
[33] D.N. Perkins, D.J.C. Pappin, D.M. Creasy, J.S. Cottrell, Electrophoresis 20 (1999)
3551–3567.
In this work, an l-arabinose isomerase from wild-type strain
E. faecium DBFIQ E36 was successfully purified by affinity
chromatography using a specific l-arabitol-agarose bioadsorbent
conveniently designed. The enzyme (present in low amount
in sample applied to affinity column) was purified 720-fold
in relation to cell-free extract. Purification factor reported was
the highest obtained up to date and this could be possible
by means of employing a highly specific interaction between
the enzyme and its immobilized natural competitive inhibitor,
l-arabitol. This methodology, greatly employed for other industri-
ally relevant keto-isomerases, has several advantages compared
with conventional purification techniques or recombinant pro-
duction of enzyme, regarding reusability of affinity bioadsorbent,
time/equipment requirements and safety, respectively.
The purified enzyme exhibited a more acidic pI compared
with other l-arabinose isomerases explaining the charged and
hydrophilic nature which, in turns resulted in a high salt con-
centration necessary for enzyme salting-out. Mass spectrometry
technique revealed a molecular weight of 48 kDa for monomer
whereas molecular mass of native enzyme (tetrameric) determined
by gel-filtration was 187 kDa. The enzyme exhibited an optimum
pH of 7.0 and optimum temperature of 50 ◦C, and showed good pH
and temperature stability. Preliminary kinetics studies indicated
that the enzyme was not inhibited by its product, a technologically
attractive characteristic, explaining the weak bioaffinity adsorption
on l-ribitol-agarose bioadsorbent.
[34] P. Prabhu, M.K. Tiwari, M. Jeya, P. Gunasekaran, I.W. Kim, J.K. Lee, Appl Microbiol
Biotechnol 81 (2008) 283–290.
[35] M. Rhimi, S. Bejar, Biochim Biophys Acta 1760 (2006) 191–199.
[36] M. Rhimi, R. Ilhammami, G. Bajic, S. Boudebbouze, E. Maguin, R. Haser, N. Agha-
jari, Bioresour Technol 101 (2010) 9171–9177.
[37] H.J. Roh, S.H. Yoon, P. Kim, Biotechnol Lett 22 (2000) 197–199.
[38] M. Rollini, M. Manzoni, Process Biochem 40 (2005) 437–444.
[39] N. Salonen, A. Nyyssölä, K. Salonen, O. Turunen, Appl Biochem Biotechnol 168
(2012) 392–405.
[40] T. Shimonishi, Y. Okumura, K. Izumori, J Ferment Bioeng 79 (1995) 620–622.
[41] M.I.G. Siso, Bioresour Technol 57 (1996) 1–11.
[42] L. Sundberg, J. Porath, J Chromatogr A 90 (1974) 87–98.
[43] J.M. Walker, in: J.M. Walker (Ed.), The Protein Protocols Handbook, 2nd ed.,
Humana Press Inc., Totowa, NJ, 2002, pp. 11–14.
[44] L.J. Wallace, F.A. Eiserling, G. Wilcox, J Biol Chem 253 (1978) 3717–3720.
[45] K. Yamanaka, W.A. Wood, Methods Enzymol 9 (1966) 596–602.
[46] H. Yoshida, M. Yamada, T. Nishitani, G. Takada, K. Izumori, S. Kamitori, Acta
Crystallogr Sect F Struct Biol Cryst Commun 63 (2007) 123–125.
[47] K. Yoshihara, Y. Shinohara, T. Hirotsu, K. Izumori, J Biosci Bioeng 100 (2006)
219–222.
Acknowledgments
This work was partially sponsored with funds of the projects
CAI + D 2009 Tipo II PI-64-325 (Universidad Nacional del Litoral,
Santa Fe, Argentina), PIP N◦ 112-200801-01331 (Consejo Nacional
de Investigaciones Científicas y Técnicas, Buenos Aires, Argentina),
PICT-2004 N◦ 20152 (Agencia Nacional de Promoción Científica
y
Tecnológica, Buenos Aires, Argentina) and FPR/F/BI/80/03-
UR/07/BVI/002 (MINCyT, Argentina and MEC, Uruguay), PEDECIBA
(Programa de Desarrollo de las Ciencias Básicas, Uruguay) and ANII
(Agencia Nacional de Investigación e Innovación, Uruguay). The
authors declare no competing financial interest.
References
[48] H. Zhang, B. Jiang, B. Pan, World J Microbiol Biotechnol 23 (2007) 641–646.
[49] W. Zhang, S. Niu, B.T. Chait, J Am Soc Mass Spectrom 9 (1998) 879–884.
[50] X. Zhou, J.C. Wu, World J Microbiol Biotechnol 28 (2012) 2205–2212.
[1] J.R. Beadle, J.P. Saunder, T.J. Wajada, Process for manufacturing tagatose, US
Patent 5,078,796, 1992.