Tài liệu Báo cáo khoa học: Insights into the interaction of human arginase II with substrate and manganese ions by site-directed mutagenesis and kinetic studies Alteration of substrate specificity by replacement of Asn149 with Asp docx

Insights into the interaction of human arginase II with
substrate and manganese ions by site-directed
mutagenesis and kinetic studies
Alteration of substrate specificity by replacement of Asn149
with Asp
Vasthi Lo
´
pez, Ricardo Alarco
´
n, Marı
´
a S. Orellana, Paula Enrı
´
quez, Elena Uribe, Jose
´
Martı
´
nez and
Nelson Carvajal
Departamento de Bioquı
´
mica y Biologı
´
a Molecular, Facultad de Ciencias Biolo
´
gicas, Universidad de Concepcio
´
n, Chile
Arginase (l-arginine urea amidino hydrolase,
EC 3.5.3.1) catalyzes the hydrolysis of l-arginine to
yield l-ornithine and urea, and exhibits an absolute

requirement for bivalent metal ions, especially Mn
2+
,
for catalytic activity. Metal ions are thought to acti-
vate a coordinated water molecule, by lowering the
pK
a
for proton ionization and generation of the
hydroxide that nucleophilically attacks the guanidino
carbon of the scissile bond of l-arginine [1–3].
The enzyme is widely distributed in living organisms,
where it serves several functions, including ureagenesis
and regulation of the cellular levels of l-arginine, a
precursor for the production of creatine, proline, poly-
amines and nitric oxide [4–7]. Mammalian tissues con-
tain two distinct isoenzymic forms: arginase I, which is
highly expressed in the liver and it has been tradition-
ally associated with ureagenesis, and the extrahepatic
arginase II, which is thought to provide a supply of
l-ornithine for proline and polyamine biosynthesis
[7–12]. Both arginase isoforms are also thought to par-
ticipate in the regulation of nitric oxide biosynthesis by
competing with nitric oxide synthases for the common
Keywords
manganese ions; histidine; agmatinase
activity; arginase II; human
Correspondence
N. Carvajal, Departamento de Bioquı
´
mica y

Biologı
´
a Molecular, Facultad de Ciencias
Biolo
´
gicas, Universidad de Concepcio
´
n,
Casilla 160-C, Concepcio
´
n, Chile
Fax: +56 41 239687
E-mail:
(Received 24 May 2005, revised 13 July
2005, accepted 19 July 2005)
doi:10.1111/j.1742-4658.2005.04874.x
To examine the interaction of human arginase II (EC 3.5.3.1) with sub-
strate and manganese ions, the His120Asn, His145Asn and Asn149Asp
mutations were introduced separately. About 53% and 95% of wild-type
arginase activity were expressed by fully manganese activated species of the
His120Asn and His145Asn variants, respectively. The K
m
for arginine (1.4–
1.6 mm) was not altered and the wild-type and mutant enzymes were essen-
tially inactive on agmatine. In contrast, the Asn149Asp mutant expressed
almost undetectable activity on arginine, but significant activity on agma-
tine. The agmatinase activity of Asn149Asp (K
m
¼ 2.5 ± 0.2 mm) was
markedly resistant to inhibition by arginine. After dialysis against EDTA,

the His120Asn variant was totally inactive in the absence of added Mn
2+
and contained < 0.1 Mn
2+
Æsubunit
)1
, whereas wild-type and His145Asn
enzymes were half active and contained 1.1 ± 0.1 Mn
2+
Æsubunit
)1
and
1.3 ± 0.1 Mn
2+
Æsubunit
)1
, respectively. Manganese reactivation of metal-
free to half active species followed hyperbolic kinetics with K
d
of
1.8 ± 0.2 · 10
)8
m for the wild-type and His145Asn enzymes and
16.2 ± 0.5 · 10
)8
m for the His120Asn variant. Upon mutation, the chro-
matographic behavior, tryptophan fluorescence properties (k
max
¼ 338–
339 nm) and sensitivity to thermal inactivation were not altered. The

Asn149fiAsp mutation is proposed to generate a conformational change
responsible for the altered substrate specificity of arginase II. We also con-
clude that, in contrast with arginase I, Mn
2+
A
is the more tightly bound
metal ion in arginase II.
4540 FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS
substrate l-arginine [11,13]. Particularly interesting has
been a possible role of arginase II in regulating the
availability of l-arginine for nitric oxide synthesis in
human penile and clitoral corpus cavernosum and
vagina, which converts this isoenzyme in a potential
target for the treatment of sexual arousal disorders
[9,14,15].
At present, there is considerable information about
the structural and functional properties of arginase I,
whose deficiency in humans results in hyperarginine-
mia, characterized by growth retardation and progres-
sive mental impairment [8]. Although significantly less
is known about arginase II, the enzyme has been
cloned [16–18], some of their kinetic properties have
been described [9–12] and the X-ray crystal structure
of a fully active, truncated form complexed with a
boronic acid transition state analog inhibitor was
determined at 2.7 A
˚
resolution [9].
Human arginases I and II are related by about 50%
amino acid sequence identity and, more importantly,

residues which are known to be involved in metal
coordination, substrate binding and catalysis are
strictly conserved between the two isoenzymes [9].
Moreover, a binuclear metal cluster (Mn
2+
A
–Mn
2+
B
)
is accepted to be required for maximal catalysis by
both enzyme forms [9]. Mammalian and all other
known arginases also shares a significant sequence
homology with all sequenced agmatinases (agmatine
ureo hydrolase, EC 3.5.3.11), which catalyzes the
hydrolytic production of urea from agmatine, a decar-
boxylated derivative of arginine [19–21]. In view of
this, a common evolutionary origin and subsequent
divergence, resulting in totally different substrate spe-
cificities, is considered for arginase and agmatinase
[20]. Such substrate discrimination is particularly
important for mammalian arginase II and agmatinase,
as both enzymes are mitochondrial and functionally
different [22]. A key factor is the a-carboxyl group of
the substrate, which makes the difference between
arginine and agmatine. Agmatine, which results from
decarboxylation of arginine by arginine decarboxylase,
is a metabolic intermediate in the biosynthesis of
putrescine and higher polyamines and may have
important regulatory roles in mammals [23].

The metal cluster of human arginase II was found
to be nearly identical to that of rat liver arginase I in
its complex with the transition state analog S-(2-boro-
noethyl)-l-cysteine (BEC). His120 and His145, and the
corresponding His101 and His126 in arginase I, were
described among the ligands for coordination of
Mn
2+
A
and Mn
2+
B
, respectively [9,24]. However, the
volume of the active site cleft was found to be larger
for arginase II. Moreover, the D232 (Od1)-Mn
2+
B
separation of 2.6 A
˚
was considered to be somewhat
long for an inner-sphere coordination interaction, as
that observed in arginase I [9]. Differences in the bind-
ing of the a-carboxylate and a-amino groups of BEC
were also ascribed to the larger volume of the active
site cleft of arginase II, which allows more water-medi-
ated enzyme–inhibitor interactions in this enzyme. For
example, Asn130 was identified as a ligand for the
a-carboxylate group of BEC in arginase I, but a water-
mediated hydrogen bond was proposed in place of a
direct hydrogen bond to the equivalent Asn149 in the

arginase II–BEC complex [9]. The isoenzymic forms
also differ in subcellular localization [8], immunologi-
cal properties [8] and sensitivity to inhibition by
ornithine [10], branched chain amino acids [25], fluor-
ide [26] and the transition state analog S-(2-borono-
ethyl)-l-cysteine [9].
In this study, the interaction of human arginase II
with substrate and metal ions was examined by site-
directed mutagenesis and kinetic studies. Selection of
target residues (Asp149, His120 and His145) was based
on the roles assigned to the equivalent residues in argi-
nase I [1,24] and the crystal structure of the arginase
II–BEC complex [9]. The Asn149fiAsp mutation
altered the substrate specificity of arginase II, yielding
enzyme species with almost undetectable activity on
arginine but significant activity on agmatine. From the
effects of replacement of His120 and His145 with aspa-
ragine we conclude that Mn
2+
A
, and not Mn
2+
B
,as
occurs in arginase I [1], is the more tightly bound ion
in arginase II.
Results and Discussion
General properties of the wild-type, His120Asn,
His145Asn and Asn149Asp variants of arginase II
Purified wild-type, His120Asn and His145Asn variants

of arginase II were active even in the absence of added
Mn
2+
, although preincubation with 5 mm Mn
2+
for
10 min at 60 °C was required to convert the enzymes
to their fully active state. In contrast, the arginase
activity of the Asn149Asp variant was practically
undetectable, both before and after the incubation with
the manganese ions.
Fully active His120Asn and His145Asn variants
exhibited about 53% and 95% of the corresponding
wild-type activity, with the K
m
value for l-arginine
remaining essentially unaltered (Table 1). Considering
His120 and His145 as metal ligands in arginase II [9],
the essentially invariant K
m
value upon mutation of
these residues agree with the currently accepted mech-
anism for the arginase reaction, which considers the
V. Lo
´
pez et al. Interaction of arginase II with substrate and manganese ions
FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS 4541
metal ion as being involved in the stabilization of the
transition state [27], but not in the stabilization of the
substrate in the Michaelis–Menten complex [24]. Also

in agreement with this, the K
m
value was not altered
by the full activation step. Although an effect of the
mutations on substrate binding can be excluded, the
significantly lower k
cat
value for His120Asn indicates
that the scissible guanidino group of the substrate is
not optimally oriented with respect to the metal-bound
hydroxide in this enzyme variant. The k
cat
and K
m
values for the wild-type enzyme are comparable with
previously reported values [9].
The tryptophan fluorescence properties (k
max
¼ 338–
339 nm) and sensitivity of arginase II to thermal inac-
tivation (Fig. 1) were not significantly altered by the
introduced mutations, and no differences between
the wild-type and mutant enzymes were detected by
the chromatographic procedures used for their purifi-
cation. In view of these results, at least gross structural
changes can be discarded as a consequence of the
mutagenic replacements.
Effects of the His120Asn and His145Asn
mutations on the affinity of metal binding
to arginase II

To further examine the effects of the His120fiAsn
and His145fiAsn mutations on the interaction of the
enzyme with manganese ions, maximally activated spe-
cies of the wild-type and mutant enzymes were dia-
lysed for 2 h at 4 °C against 10 mm EDTA in 10 mm
Tris ⁄ HCl pH 7.5, followed by two changes of the same
buffer but without EDTA. The dialyzed enzymes were
then assayed for catalytic activity and metal content
by atomic absorption analysis. As shown in Fig. 2,
after incubation with 5 mm Mn
2+
for 10 min at 60 °C
and assay in the presence of added 2 mm Mn
2+
, all of
the enzymes were active and measured activities were
essentially equal to the initial activity of the corres-
ponding fully activated control. However, when the
preincubation step was omitted and the assays were
performed in the absence of added Mn
2+
, the
His120Asn variant was found to be totally inactive,
whereas half of full activity was expressed by the
His145Asn mutant and wild-type enzymes. In agree-
ment with the inactive state of dialyzed species of the
His120Asn variant, its manganese content was almost
undetectable (< 0.1 Mn
2+
Æsubunit

)1
). On the other
hand, the half active wild-type and His145Asn enzymes
Table 1. Kinetic properties of the arginase activities of wild-type
and mutant variants of human arginase II. Values, derived from two
separate experiments in duplicate, represent the means ± SD. Argi-
nase activities were determined in 50 mm glycine ⁄ NaOH pH 9.0.
ND, Not determined, because the N149D variant expressed almost
undetectable activity on arginine.
Enzyme k
cat
(s
-1
) K
m
Arg
(mM) k
cat
⁄ K
m
Arg
(M
-1
Æs
-1
)
Wild-type 249 ± 10 1.4 ± 0.1 177.9
His120Asn 131 ± 8 1.6 ± 0.1 81.9
His145Asn 238 ± 12 1.5 ± 0.1 158.6
Asn149Asp ND ND ND

Fig. 1. Fluorescence spectra (A) and sensitivity to thermal inactiva-
tion (B) of wild-type (s), H120N (d), H145N (h) and N149D (,)var-
iants of human arginase II. Fluorescence spectra were recorded at
25 °C; protein concentrations were 73, 82, 59 and 79 lgÆmL
)1
, for
the wild-type, His120Asn, His145Asn and Asn149Asp variants,
respectively. The line in (B) is for the average of experimental
values for all the enzyme variants.
Interaction of arginase II with substrate and manganese ions V. Lo
´
pez et al.
4542 FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS
contained 1.1 ± 0.1 Mn
2+
Æsubunit
)1
and 1.2 ± 0.1
Mn
2+
Æsubunit
)1
, respectively. Considerably more dras-
tic conditions were necessary to obtain metal-free,
inactive species of the wild-type and His145Asn
enzyme variants. Routinely, this was performed by
incubation for 1 h at 25 °C with 25 mm EDTA and
3 m guanidinium chloride in 10 mm Tris ⁄ HCl pH 7.5,
followed by overnight dialysis at 4 °C against 5 mm
Tris ⁄ HCl pH 7.5.

Clearly, the affinity of the arginase–manganese inter-
action was significantly altered by replacement of
His120 with asparagine. This aspect was quantitatively
evaluated by following the Mn
2+
-dependent reactiva-
tion of metal-free species of the wild-type, His120Asn
and His145Asn variants. Reactivation by free Mn
2+
concentrations in the nanomolar range followed hyper-
bolic kinetics, consistent with the absence of coopera-
tivity between metal binding sites. The estimated K
d
values were 1.8 ± 0.2 · 10
)8
m for the wild-type and
His145Asn enzymes and 16.2 ± 0.5 · 10
)8
m for the
His120Asn variant, whereas the V
max
values were
nearly equal to a half of those determined after incu-
bation of the respective enzyme variant with 5 mm
Mn
2+
for 10 min at 60 °C. These results indicate the
existence of high and low affinity bindings of mangan-
ese ions to arginase II and provide an explanation for
the manganese stoichiometries determined here. Con-

sidering the stoichiometry of 2 Mn
2+
Æsubunit
)1
derived
from EPR analysis of fully active arginase II [10], our
conclusion is that a weakly bound Mn
2+
is removed
by EDTA during the preparation of the samples for
atomic absorption analysis of the wild-type and
His145Asn variants. In addition to removal of the
more weakly bound Mn
2+
, the lower affinity for that
more tightly bound to the protein may explain the
absence of Mn
2+
from the EDTA-treated species of
the His145Asn variant. Even though under our condi-
tions the wild-type and His145Asn variants behaved
essentially in the same manner and expressed practi-
cally the same catalytic activity, an effect of the
His145Asn mutation on the affinity for the more
weakly bound metal ion cannot be discarded. The
presence of tightly and weakly bound manganese ions
was also demonstrated for fully active species of argi-
nase I [1,2]. Moreover, hyperbolic kinetics with dissoci-
ation constants for the more tightly bound Mn
2+

in
the range of those determined here, were also reported
for arginase I [28,29,30].
A binuclear motif (Mn
2+
A
–Mn
2+
B
) was derived
from the X-ray crystal structure of fully active arginase
II complexed with BEC, a boronic acid transition state
analog inhibitor of the enzyme [9]. As indicated by the
crystal structure, Mn
2+
A
and Mn
2+
B
are, respectively,
coordinated by His120 and His145, which are equival-
ent to the histidines at position 101 and 126 in the
sequence of arginase I [10]. However, they are clearly
differentiated by the consequences of their replace-
ments with asparagine. In fact, in contrast with argi-
nase II, dialysis against EDTA results in species of the
His101Asn variant of arginase I which are half active
and contain 1 Mn
2+
Æsubunit

)1
, and metal-free, inactive
species of the His126Asn variant [31]. Our conclusion
is that the more weakly bound metal ion, which is
preferentially removed by EDTA, is Mn
2+
A
in argi-
nase I and Mn
2+
B
in arginase II. Because ligands to
the metal ions are strictly conserved in these enzymes,
the difference would reside in the length of the ligand–
metal separations. In this connection, the Asp232
(Od1)-Mn
2+
B
separation of 2.6 A
˚
in the arginase II–
BEC complex was considered somewhat long for an
inner-sphere coordination interaction, as that observed
in arginase I [9]. A lengthened His124 (Nd1)–Mn
2+
B
bond was also associated to the preferential release of
Mn
2+
B

by EDTA during preparation of a substrate
complex of Bacillus caldovelox arginase for crystallo-
graphic analysis [32].
According to our present results, the catalytic activ-
ity of the partially active species of arginase II is
associated to the more tightly bound Mn
2+
A
. The
increased activity resulting from the addition of the
more weakly bound Mn
2+
B
may be explained by a
Fig. 2. Effect of dialysis of fully activated wild-type and mutant
species of human arginse II. Fully active species were dialyzed for
4 h at 4 °C against 10 m
M EDTA in 10 mM Tris ⁄ HCl pH 7.5, and
then assayed for arginase activity in the absence of added Mn
2+
(open bars) and after full activation with the metal ion and assay in
the presence of added 2 m
M Mn
2+
(filled bars). Arginase activities
are expressed as percentage of the initial activity of the corres-
ponding fully activated form.
V. Lo
´
pez et al. Interaction of arginase II with substrate and manganese ions

FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS 4543
lower pK
a
for a water molecule bound to a binuclear
metal cluster and, consequently, by a higher concentra-
tion of the nucleophilic metal-bound hydroxide [33–35]
and increased stabilization of the transition state affor-
ded by the more weakly bound metal ion [27].
Altered substrate specificity accompanying the
Asn149Asp mutation
Interestingly, while essentially inactive on l-arginine,
the An149Asp variant exhibited a significant activity on
its decarboxylated derivative, agmatine (Fig. 3). The
possibility of interference from the endogenous agma-
tinase of the bacterial vector was excluded by the
DEAE-cellulose chromatographic step of the purifica-
tion protocol. In fact, like for all the arginase variants
examined in this study, 0.10–0.15 m KCl was required
to elute the Asn149Asp variant from a DEAE-cellulose
column equilibrated with 5 mm Tris ⁄ HCl pH 7.5,
whereas about 0.45 m KCl was required for elution of
the endogenous bacterial agmatinase. Moreover, the
arginase variants, including Asn149Asp, were not detec-
ted by Western blot analysis using an anti-Escherichia
coli agmatinase polyclonal antibody. Finally, in contrast
with E. coli agmatinase, the Asn149Asp was markedly
resistant to inhibition by arginine (Fig. 4).
Clearly, arginine was very poorly recognized as a
substrate or inhibitor by the Asn149Asp variant. The
opposite occurred with the wild-type, His120Asn and

His145Asn variants, which were practically inactive on
agmatine and markedly resistant to inhibition by the
substrate analog. As an example, only about 25% inhi-
bition of the wild-type and mutant enzymes was pro-
duced by 20 mm agmatine and production of urea was
practically absent using this agmatine concentration
as a potential substrate. Although a more detailed ana-
lysis of the effect was not performed, the inhibition by
20 mm agmatine was eliminated by saturation with
l-arginine, indicating the competitive character of the
inhibition produced by the substrate analog. In general
agreement with our present results, agmatine was pre-
viously described as a very poor alternate substrate
and inhibitor for human arginase II [11].
Like the arginase activity of the wild-type,
His120Asn and His145Asn variants, agmatine hydro-
lysis by the Asn149Asp mutant enzyme was maximal
at pH 9–9.5 and strictly dependent on manganese
ions, because metal-free species were totally inactive in
the absence of added Mn
2+
. At the optimum pH, the
K
m
for agmatine (2.5 ± 0.2 mm) was very close to the
K
m
of the wild-type arginase for arginine. However,
the hydrolytic activity of Asn149Asp on agmatine was
only about 5% of the arginase activity of the wild-

type enzyme. As measured by k
cat
⁄ K
m
, the catalytic
efficiency of the Asn149Asp variant was found to be
about 36-fold lower than that of wild-type arginase II
acting on arginine. For comparison, the catalytic effi-
ciency of E. coli agmatinase [36] is only twofold lower
than that corresponding to wild-type arginase II. At
this connection, residues known to be involved in
binding and hydrolysis of the guanidino group of
l-arginine by arginase are strictly conserved in the act-
ive site of the agmatinases [19]. Moreover, modeling
studies have revealed that essentially the same position
with respect to the metal ions and conserved catalyti-
Fig. 3. Catalytic activity of the Asn149Asp variant of human argi-
nase II. Substrates were agmatine (s) and
L-arginine (d). The buf-
fer was 50 m
M glycine ⁄ NaOH pH 9.0.
Fig. 4. Effect of L-arginine on agmatine hydrolysis by the N149D
variant of arginase II (s)andE. coli agmatinase (d). The buffer was
50 m
M glycine ⁄ NaOH pH 9.0.
Interaction of arginase II with substrate and manganese ions V. Lo
´
pez et al.
4544 FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS
cally important residues may be adopted by agmatine

in E. coli agmatinase and l-arginine in B. caldovelox
arginase [37], indicating that the substrate specificity
of these enzymes rely mainly in substituents at C-a.
This has been, in fact, demonstrated for arginase
[38,39] and the same may be safely deduced for agma-
tinase. Therefore, as an explanation for the low cata-
lytic efficiency of Asn149Asp, we conclude that the
guanidino group of agmatine is not optimally posi-
tioned and oriented for a more efficient nucleophilic
attack by a metal-bound hydroxide, most probably
due to a nonoptimal positioning of the nonguanidino
portion of the substrate molecule.
Residue Asn149 is totally conserved among all the
arginases [19] and the equivalent Asn130 has been
considered as providing a hydrogen bond to the
a-carboxyl group of the substrate l-arginine in argi-
nase I [26]. However, against a functional equivalence
between these residues is the observation that Asn130,
but not Asn149, interacts with the a-carboxylate
group of the transition state analog BEC in the
corresponding binary enzyme–analog complex [9].
Certainly, if an interaction between Asn149 and the
a-carboxylate group of arginine were also operative
for arginase II, both the lack of arginase activity as
well as the resistance of the Asn149Asp mutant to
inhibition by l-arginine, would be explained by elec-
trostatic repulsion between the a-carboxyl group of
the amino acid and the introduced aspartic residue at
position 149. On the other hand, as agmatine lacks
the a-carboxyl group, there would be no impediment

for its binding and hydrolysis by the Asn149Asp vari-
ant. However, if the only change were in the charge at
position 149, it would hard to explain why agmatine
not only is practically not hydrolysed by wild-type
arginase II, but it is also very poorly inhibitory to this
enzyme form. Thus, the altered specificity most prob-
ably reflect an active site conformational change
resulting from the Asn149fiAsp substitution. As
deduced from the unaltered fluorescence properties,
thermal stability and chromatographic behavior, the
conformational change is not expected to be extensive
enough to cause gross alterations in the enzyme struc-
ture. Studies addressed to further define the expected
conformational change, using experimental and com-
putational methods, will be initiated soon in our
laboratory.
General conclusions
In addition to substantiate the participation of His120
and His145 as ligands for the manganese ions in
human arginase II, our results have provided addi-
tional evidence for the differences between the active
sites of this enzyme and arginase I. In spite of the rel-
atively low agmatinase activity of the Asn149Asp vari-
ant, it is clear that the interactions of arginase II with
l-arginine and agmatine are greatly altered by replace-
ment of this residue with aspartate. To the best of our
knowledge, this is the first report in which the sub-
strate specificity of arginase was altered by using site-
directed mutagenesis.
Experimental procedures

Materials
All reagents were of the highest quality commercially avail-
able (most from Sigma Chemical Co., St Louis, MO, USA)
and were used without further purification. Restriction
enzymes, and enzymes and reagents for PCR were from
Promega. The plasmid pBluescript II K(+), bearing the gene
of human arginase II, was kindly supplied by S. Cederbaum
(University of California, Los Angeles). Synthetic nucleotide
primers were obtained from Invitrogen and the QuickChange
site-directed mutagenesis kit was from Stratagene. Purified
E. coli agmatinase was obtained as described previously [36].
The rabbit anti-E. coli agmatinase polyclonal antibody was
supplied by M. Salas (Universidad de Concepcio
´
n, Chile).
Enzyme preparations
Bacteria were grown with shaking at 37 °C in Luria broth
in the presence of ampicillin (100 lgÆmL
)1
). The wild-type
and mutant arginase II cDNAs were directionally cloned
into the pBluescript II K(+) E. coli expression vector and
the enzymes were expressed in E. coli strain JM109, follow-
ing induction with 1 mm isopropyl thio-b-d-galactoside.
The bacterial cells were disrupted by sonication on ice
(5 · 30 s pulses) and the supernatant of a centrifugation for
20 min at 12 000 g was precipitated with ammonium sulfate
(60% saturation). The pellet, recollected by centrifugation
at 12 000 g for 10 min was resuspended in 5 mm Tris ⁄ HCl
pH 7.5 containing 2 mm MnCl

2
and dialyzed for 6 h at
4 °C against the same buffer. After incubation with 5 mm
MnCl
2
for 10 min at 60 °C, the enzyme solution was separ-
ated by chromatography on a CM-cellulose column equili-
brated with 5 mm Tris ⁄ HCl pH 7.5; active fractions, eluting
with the washings of the column, were then chromato-
graphed on a DEAE-cellulose column equilibrated with
5mm Tris ⁄ HCl pH 7.5. Active fractions, eluting at 0.10–
0.15 m KCl, were pooled and dialyzed against 5 mm
Tris ⁄ HCl pH 7.5 containing 2 mm MnCl
2
. A single protein
band was detected by SDS ⁄ PAGE and Coomassie blue
staining of purified enzymes.
Metal-free species of purified enzymes were obtained by
incubation for 1 h at 25 °C with 25 mm EDTA and 3 m
V. Lo
´
pez et al. Interaction of arginase II with substrate and manganese ions
FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS 4545
guanidinium chloride in 10 mm Tris ⁄ HCl pH 7.5, followed
by overnight dialysis at 4 °C against 5 mm Tris ⁄ HCl
pH 7.5.
Site-directed mutagenesis
The His120Asn, His145Asn and Asn149Asp mutant forms of
human arginase II were obtained by a two-step PCR [40],
using the QuickChange site-directed mutagenesis kit (Strata-

gene, La Jolla, CA, USA). The antisense mutagenic oligo-
nucleotide primers were: 5¢-gattgccaggctgttgtctcctcccag-3¢,
5¢-GTTGATGTCAGCATTGGCATCAACCCA-3¢ and
5¢-GGGGTGTGTCGATGTCA-3¢ for His120Asn, His145Asn
and Asn149Asp, respectively. The corresponding sense muta-
genic oligonucleotide primers were 5¢-CTGGGAGGAGA
CAACAGCCTGGCAATC-3¢ for His120Asn, 5¢-TGGGTT
GATGCCAATGCTGACATCAAC-3¢ for His145Asn and
5¢-TGACATCGACACACCCC-3¢ for Asn149Asp.
Fluorescence spectra and thermal inactivation
studies
Fluorescence measurements were made at 25 °C on a Shim-
adzu RF-5301 spectrofluorimeter (Columbia, MD). The
protein concentration was 40–50 lgÆmL
)1
and emission
spectra were measured with the excitation wavelength at
295 nm. The slit width for both excitation and emission
was 1.5 nm, and spectra were corrected by substracting the
spectrum of the buffer solution (5 mm Tris ⁄ HCl, pH 7.5) in
the absence of protein.
The stability to thermal inactivation was examined by
incubation of the enzymes at 75 °C in a solution containing
10 mm Tris ⁄ HCl pH 7.5 and 2 mm Mn
2+
. At several times
(up to 30 min), aliquots were removed and assayed for
residual enzymatic activity at pH 9.5, in the presence of
added 2 mm Mn
2+

.
Atomic absorption analysis
The manganese contents of arginase preparations were deter-
mined by atomic absorption on a Perkin Elmer 1100 atomic
absorption spectrometer (NY, USA) equipped with a
graphite furnace and a deuterium arc background corrector.
Recovery was nearly 100%. For analysis, the purified enzyme
was activated by incubation with 2 mm MnCl
2
in 10 mm
Tris ⁄ HCl pH 8.0 for 30 min at 37 °C, and then the free metal
ion was removed by dialysis against 10 mm Tris ⁄ HCl pH 7.5,
10 mm EDTA for 2 h at 4 °C, followed by two changes of
10 mm Tris ⁄ HCl pH 7.5 as the dialysis buffer.
Enzyme assays and kinetic studies
Routinely, enzyme activities were determined by measuring
the formation of urea from l-arginine or agmatine in
50 mm glycine ⁄ NaOH pH 9.0. In studying the effect of
pH on enzyme activities, buffers used were 50 mm
Tris ⁄ HCl pH 7–8.7 and 50 mm glycine ⁄ NaOH pH 8.7–10.
Urea was determined by a colorimetric method with a-iso-
nitrosopropiophenone [41]. As urea is also produced by
agmatine hydrolysis, in studying the inhibitory effect of
agmatine on arginine hydrolysis, reactions were followed
by measuring the formation of ornithine, determined by
the method of Chinard [42]. Protein concentrations were
determined by the method of Bradford [43], with BSA as
standard.
Steady-state initial velocity studies were performed at
37 °C and all assays were initiated by adding the enzyme

to a previously equilibrated buffer substrate solution.
Data from initial velocity studies, performed in duplicate
and repeated at least twice, were fitted to the Michaelis–
Menten equation, by using nonlinear regression with
prism 4.0 (GraphPad Software Inc., San Diego, CA,
USA).
To evaluate the affinity for the more tightly bound metal
ion, metal-free enzymes were incubated with varied concen-
trations of Mn
2+
in 10 mm Tris ⁄ HCl pH 8.5, 50 mm KCl
and 10 mm nitrilotriacetic acid as a metal ion buffer [28].
After equilibration for 15 min at 37 °C, arginase activities
were determined in 50 mm Tris ⁄ HCl pH 8.5. Free-Mn
2+
concentrations were calculated using a dissociation constant
of 3.98 · 10
)8
m and a pK
a3
value of 9.8 for nitrilotriacetic
acid [28]. Dissociation constants (K
d
) and V
max
values were
determined from double reciprocal plots of velocity vs. free
metal ion concentrations.
Acknowledgements
This research was supported by Grants 1030038 from

FONDECYT and Grant CONICYT to support the
PhD thesis of V. Lo
´
pez.
References
1 Kanyo ZF, Scolnick LR, Ash DE & Christianson DW
(1996) Structure of a unique binuclear manganese clus-
ter in arginase. Nature 382, 554–557.
2 Christianson DW & Cox JD (1999) Catalysis by metal-
activated hydroxide in zinc and manganese metallo-
enzymes. Annu Rev Biochem 68, 33–57.
3 Ash DE, Cox JD & Christianson DW (2000) Arginase:
a binuclear manganese metalloenzyme. Met Ions Biol
Syst 37, 407–428.
4 Morris SM (2002) Regulation of enzymes of the urea
cycle and arginine metabolism. Annu Rev Nutr 22, 87–
105.
5 Jenkinson CP, Grody WW & Cederbaum SD (1996)
Comparative properties of arginases. Comp Biochem
Physiol 114B, 107–132.
Interaction of arginase II with substrate and manganese ions V. Lo
´
pez et al.
4546 FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS
6 Iyer RK, Kim HK, Tsoa RW, Grody WW & Ceder-
baum SD (2002) Cloning and characterization of human
agmatinase. Genet Metab 75, 209–218.
7 Mistry SK, Burwell TJ, Chambers RM, Rudolph-Owen
L, Spaltmann F, Cook WJ & Morris SM (2002) Cloning
of human agmatinase An alternate path for polyamine

synthesis induced in liver by hepatitis B virus. Am J
Physiol Gastrointest Liver Physiol 282, G375–G381.
8 Cederbaum SD, Yu H, Grody WW, Kern RM, Yoo P
& Yyer RK (2004) Arginases I and II: do their func-
tions overlap? Mol Genet Metab 81, S38–S44.
9 Cama E, Colleluori DM, Emig FA, Shin H, Kim SW,
Kim NN, Traish AB, Ash DE & Christianson DW
(2003) Human arginase II: crystal structure and physio-
logical role in male and female sexual arousal. Biochem-
istry 42, 8445–8451.
10 Colleluori DM, Morris SM & Ash DE (2001) Expres-
sion, purification, and characterization of human type II
arginase. Arch Biochem Biophys 389, 135–243.
11 Colleluori DM & Ash DE (2001) Classical and slow-
binding inhibitors of human type II arginase. Biochemis-
try 40, 9356–9362.
12 Ash DE (2004) Structure and function of arginases.
J Nutr 134, 2760S–2764S.
13 Morris SM (2004) Recent advances in arginine meta-
bolism. Curr Opin Clin Nutr Metab Care 7, 45–51.
14 Kim NN, Cox JD, Baggio RF, Emig FA, Mistry SK,
Harper SL, Speicher DW, Morris SM, Ash DE, Traish
A & Christianson DW (2001) Probing erectile function.
S-(2-boronoethyl)-L-cysteine binds to arginase as a tran-
sition state analogue and enhances smooth muscle
relaxation in human penile corpus cavernosum, Bio-
chemistry 40, 2678–2688.
15 Cox JD, Kim NN, Traish AM & Christianson DW
(1999) Arginase-boronic acid complex highlights a phy-
siological role in erectile function. Nat Struct Biol 6,

1043–1047.
16 Vockley JG, Jenkinson CP, Shukla H, Kern RM, Grody
WW & Cederbaum SD (1996) Cloning and characteriza-
tion of the human type II arginase gene. Genomics 38,
118–123.
17 Gotoh T, Sonoki T, Nagasaki A, Terada K, Takiguchi
M & Mori M (1996) Molecular cloning of cDNA for
nonhepatic mitochondrial arginase (arginase II) and
comparison of its induction with nitric oxide synthase in
a murine macrophage-like cell line. FEBS Lett 395,
119–122.
18 Morris SM, Bhamidipati D & Kepka-Lenhart D (1997)
Human type II arginase: sequence analysis and tissue-
specific expression. Gene 193, 157–161.
19 Perozich J, Hempel J & Morris SM (1998) Roles of con-
served residues in the arginase family. Biochim Biophys
Acta 1382, 23–27.
20 Ouzounis CA & Kyrpides NC (1994) On the evolution of
arginases and related enzymes. J Mol Evol 39, 101–104.
21 Ahn HJ, Kim KH, Lee J, Ha JY, Lee HH, Kim D,
Yoon HJ, Kwon AR & Suh SW (2004) Crystal struc-
ture of agmatinase reveals structural conservation and
inhibition mechanism of the ureohydrolase superfamily.
J Biol Chem 279, 50505–50513.
22 Dallmann K, Junkder H, Balabanov S, Zimmermann
U, Giebel J & Walther R (2004) Human agmatinase is
diminished in the clear cell type of renal cell carcinoma.
Int J Cancer 108, 342–347.
23 Carvajal N, Orellana MS, Salas M, Enrı
´

quez P, Alarco
´
n
R, Uribe E & Lo
´
pez V (2004) Kinetic and site directed
mutagenesis studies of Escherichia coli agmatinase. A
role for Glu274 in binding and correct positioning of
the substrate guanidinium group. Arch Biochem Biophys
430, 185–190.
24 Cox JD, Cama E, Colleluori DM, Pethe S, Boucher J-L,
Mansuy D, Ash DE & Christianson DW (2001)
Mechanistic and metabolic inferences from the binding
of substrate analogues and products to arginase.
Biochemistry 40, 2689–2701.
25 Carvajal N & Cederbaum SD (1986) Kinetics of inhibi-
tion of rat liver and kidney arginases by proline and
branched-chain amino acids. Biochim Biophys Acta 870,
181–184.
26 Tormanen CD (2003) Substrate inhibition of rat liver
and kidney arginase with fluoride. J Inorg Biochem 93,
243–246.
27 Shin H, Cama E & Christianson DW (2004) Design of
amino acid aldehydes as transition-state analogue inhibi-
tors of arginase. J Am Chem Soc 126, 10278–10284.
28 Carvajal N, Torres C, Uribe E & Salas M (1995) Inter-
action of arginase with metal ions: studies of the enzyme
from human liver and comparison with other arginases.
Comp Biochem Physiol 112B, 53–159.
29 Hirsch-Kolb H, Kolb HJ & Greenberg DM (1971)

Nuclear magnetic resonance studies of manganese bind-
ing of rat liver arginase. J Biol Chem 246, 395–401.
30 Kuhn NJ, Ward S, Piponski M & Young TW (1995)
Purification of human hepatic arginase and its manga-
nese (II) -dependent and pH-dependent interconversion
between active and inactive forms: a possible pH-sen-
sing function of the enzyme on the ornithine cycle. Arch
Biochem Biophys 320, 24–34.
31 Orellana MS, Lo
´
pez V, Uribe E, Fuentes M, Salas M &
Carvajal N (2002) Insights into the interaction of
human liver arginase with tightly and weakly bound
manganese ions by chemical modification and site-direc-
ted mutagenesis studies. Arch Biochem Biophys 403 ,
155–159.
32 Bewley MC, Jeffrey PD, Patchett ML, Kanyo ZF &
Baker EN (1999) Crystal structures of Bacillus caldove-
lox arginase in complex with substrate and inhibitors
reveal new insights into activation, inhibition and cata-
lysis in the arginase superfamily. Structure Fold Des 7,
435–448.
V. Lo
´
pez et al. Interaction of arginase II with substrate and manganese ions
FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS 4547
33 Christianson DW (1997) Structural chemistry and bio-
logy of manganese metallenzymes. Prog Biophys Mol
Biol 67, 217–252.
34 Sossong TM, Khangulov SV, Cavalli RC, Soprano DR,

Dismukes GC & Ash DE (1997) Catalysis on dinuclear
Mn (II) centers: hydrolytic and redox activities of rat
liver arginase. J Bioinorg Chem 2, 433–443.
35 Ivanov I & Klein ML (2005) Dynamical flexibility and
proton transfer in the arginase active site probed by a
initio molecular dynamics. J Am Chem Soc 127, 4010–
4020.
36 Carvajal N, Lo
´
pez V, Salas M, Uribe E, Herrera P &
Cerpa J (1999) Manganese is essential for catalytic
activity of Escherichia coli agmatinase. Biochem Biophys
Res Commun 258, 808–811.
37 Salas M, Rodrı
´
guez R, Lo
´
pez N, Uribe E, Lo
´
pez V
& Carvajal N (2002) Insights into the reaction
mechanism of Escherichia coli agmatinase by site-
directed mutagenesis and molecular modelling A
critical role for aspartate 153. Eur J Biochem 269,
5522–5526.
38 Han S, Moore RA & Viola RE (2002) Synthesis and
evaluation of alternative substrates for arginase. Bioorg
Chem 30, 81–94.
39 Reczkowski RS & Ash DE (1994) Rat liver arginase:
kinetic mechanism, alternate substrates. Inhibitors, Arch

Biochem Biophys 312, 31–37.
40 Ho SN, Hunt HD, Horton RM, Pullen JK & Pease LR
(1989) Site-directed mutagenesis by overlap extension
using the polymerase chain reaction. Gene 77, 51–59.
41 Archibald RM (1945) Colorimetric determination of
urea. J Biol Chem 157, 507–518.
42 Chinard FP (1952) Photometric estimation of proline
and ornithine. J Biol Chem 199, 91–95.
43 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein util-
izing the principle of protein-dye binding. Anal Biochem
72, 248–254.
Interaction of arginase II with substrate and manganese ions V. Lo
´
pez et al.
4548 FEBS Journal 272 (2005) 4540–4548 ª 2005 FEBS