Tài liệu Báo cáo khoa học: Active and regulatory sites of cytosolic 5¢-nucleotidase doc

Active and regulatory sites of cytosolic 5¢-nucleotidase
Rossana Pesi
1
, Simone Allegrini
2
, Maria Giovanna Careddu
1,2
, Daniela Nicole Filoni
1
,
Marcella Camici
1
and Maria Grazia Tozzi
1
1 Dipartimento di Biologia, Unita
`
di Biochimica, Universita
`
di Pisa, Pisa, Italy
2 Dipartimento di Scienze del Farmaco, Universita
`
di Sassari, Sassari, Italy
Introduction
Cytosolic 5’-nucleotidase (cN-II) is a ubiquitous enzyme
that catalyses either the hydrolysis or the transfer of
phosphate esterified in the 5¢ position of 6-hydroxypu-
rine monophosphate nucleosides [1]. The transfer of
phosphate can lead to phosphorylation of inosine,
guanosine and a number of their analogues [2]. There-
fore, in addition to being involved in regulation of
purine intracellular pool, the enzyme is also responsible

Keywords
cN-II active site; cN-II regulatory sites;
cN-II structure; cytosolic 5¢-nucleotidase II
Correspondence
M. G. Tozzi, Dipartimento di Biologia,
Via S. Zeno 51, Pisa, Italy
Fax: +39 502211450
Tel: +39 502211457
E-mail:
(Received 19 July 2010, revised
10 September 2010, accepted
21 September 2010)
doi:10.1111/j.1742-4658.2010.07891.x
Cytosolic 5¢-nucleotidase (cN-II), which acts preferentially on 6-hydroxypu-
rine nucleotides, is essential for the survival of several cell types. cN-II
catalyses both the hydrolysis of nucleotides and transfer of their phosphate
moiety to a nucleoside acceptor through formation of a covalent phospho-
intermediate. Both activities are regulated by a number of phosphorylated
compounds, such as diadenosine tetraphosphate (Ap
4
A), ADP, ATP,
2,3-bisphosphoglycerate (BPG) and phosphate. On the basis of a partial
crystal structure of cN-II, we mutated two residues located in the active
site, Y55 and T56. We ascertained that the ability to catalyse the transfer
of phosphate depends on the presence of a bulky residue in the active site
very close to the aspartate residue that forms the covalent phospho-
intermediate. The molecular model indicates two possible sites at which
adenylic compounds may interact. We mutated three residues that mediate
interaction in the first activation site (R144, N154, I152) and three in the
second (F127, M436 and H428), and found that Ap

4
A and ADP interact
with the same site, but the sites for ATP and BPG remain uncertain. The
structural model indicates that cN-II is a homotetrameric protein that
results from interaction through a specific interface B of two identical
dimers that have arisen from interaction of two identical subunits through
interface A. Point mutations in the two interfaces and gel-filtration experi-
ments indicated that the dimer is the smallest active oligomerization state.
Finally, gel-filtration and light-scattering experiments demonstrated that
the native enzyme exists as a tetramer, and no further oligomerization is
required for enzyme activation.
Structured digital abstract
l
MINT-8011572: cN-II (uniprotkb:O46411) and cN-II (uniprotkb:O46411) bind (MI:0407)by
dynamic light scattering (
MI:0038)
l
MINT-8011493, MINT-8011481: cN-II (uniprotkb:O46411) and cN-II (uniprotkb:O46411)
bind (
MI:0407)bymolecular sieving (MI:0071)
Abbreviations
cN-II, cytosolic 5¢-nucleotidase; cN-III, cytosolic 5¢-nucleotidase III; cN-IA, cytosolic 5¢-nucleotidase IA; cN-IB, cytosolic 5¢-nucleotidase IB;
cdN, cytosolic 5¢(3¢)-deoxyribonucleotidase; mdN, mitochondrial 5¢(3¢)-deoxyribonucleotidase.
FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS 4863
for pro-drug activation and inactivation [3,4]. It has
been demonstrated that the catalytic mechanism of
cN-II requires formation of a covalent phospho-inter-
mediate on an aspartate residue located in a conserved
motif (motif I) [5]. This motif, together with three
other conserved motifs, is shared among the members

of the haloacid dehalogenase (HAD) superfamily,
including the soluble 5¢-nucleotidase family: cytosolic
5¢-nucleotidases cN-II, cN-III, cN-IA and cN-IB, and
both cytosolic and mitochondrial 5¢(3¢)-deoxyribonu-
cleotidases [5]. Even though all soluble 5 ¢-nucleotidases
share the same reaction mechanism and possess con-
served structural motifs in the catalytic site, only two
members of this family possess phosphotransferase
activity, namely cN-II and cN-III.
cN-II has several unique aspects, such as its complex
regulation and the very high degree of primary
sequence conservation during evolution [5]. These
aspects indicate that this enzyme plays an important
role. cN-II knockdown through RNAi causes apopto-
sis in cultured cells [6]. Furthermore, over-expression
of cN-II by more than 10-fold in HEK293 cells has
proved impossible to achieve, probably because of an
adverse effect on cell viability [7]. The reaction rates
of both nucleotide hydrolysis and phosphate transfer
catalysed by cN-II appear to be regulated by a number
of phosphorylated compounds such as ADP, ATP,
BPG, Ap
4
A and polyphosphates. These compounds
act as allosteric activators. ATP causes an increase in
V
max
of approximately 10-fold, with little effect on K
m
for the substrate IMP [8,9].

Free inorganic phosphate, on the other hand, acts as
an allosteric inhibitor, causing a 20-fold increase in K
m
for the substrate IMP with little effect on V
max
. Inter-
estingly, ATP partially counteracts the effect of phos-
phate, by increasing V
max
; however, it is unable to
reverse the increase in K
m
[9]. cN-II has been described
as a homotetramer with the ability to change its oligo-
merization state in response to the presence of activa-
tors or inhibitors. It has been suggested that the
change in the oligomerization state is accompanied by
a change in specific activity [10]. However, this simple
model does not explain the kinetic evidence described
above.
Despite its cytosolic location, cN-II has particularly
poor solubility. This is why it has been difficult to
obtain the crystal structure of the whole protein. A
truncated form of cN-II lacking the last 25 amino
acids is significantly more soluble than the wild-type
enzyme, and was recently crystallized [11]. The crystal-
lographic model, constructed by ordering 487 residues
(1-400 and 417-488) out of 561, indicates a homotetra-
meric protein resulting from interaction through a
specific interface (interface B) of two identical dimers

that arise from interaction of two identical subunits
through interface A. Because of the presence of adeno-
sine bound to the crystallized protein, Wallde
´
n et al.
[11] identified two putative effector sites, and suggested
that effector site 1, close to interface A, might interact
with BPG and Ap
4
A, while ATP and ADP might bind
to effector site 2.
In a previous paper, two active forms of cN-II were
identified in extracts from different organs [12]. The two
forms purified from calf thymus showed different behav-
iour with activators. The heavier form (form A) has a
high-affinity regulatory site for BPG, while ADP and
ATP share a different site. The lighter form (form B)
has three different sites for the three activators [5,12].
Although many papers have been published in the
last few years describing structural and functional fea-
tures of cN-II, fundamental aspects of how the enzyme
functions remain to be unravelled, in particular the
number and location of the interaction sites for the
activators and inhibitors, the relationship between
activity and enzyme oligomerization state, and, finally,
the amino acid residue(s) responsible for phospho-
transferase ability of cN-II. We have utilized a mecha-
nistic approach in order to increase knowledge on
these topics.
Results

Active site and phosphotransferase reaction
Figure 1 shows the aligned sequences of motif I for the
six intracellular 5¢-nucleotidases. Of the six enzymes,
only cN-II and cN-III possess a Thr instead of a Val
(position 56 of cN-II, boxed in Fig. 1) [11]. These are
the only two enzymes for which phosphotransferase
activity has been unquestionably ascertained, and
Wallde
´
n et al. [11] suggested that the presence of T56
Fig. 1. Aligned conserved motif I of the six known intracellular
human 5¢-nucleotidases: cN-II, cN-III, cN-IA, cN-IB, cytosolic 5¢(3¢)-
deoxyribonucleotidase (cdN) and mitochondrial 5¢(3¢)-deoxyribonu-
cleotidase (mdN). Residues 55 and 56 of cN-II are indicated in bold.
Regulation of cytosolic 5¢-nucleotidase R. Pesi et al.
4864 FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS
might be important for this activity. However, we
noted that there is another variable residue near T56.
In position 55 of cN-II, a Tyr is present that is substi-
tuted by Met in cN-III and by Ala or Gly in the other
enzymes. Therefore, the two phosphotransferases have
a bulkier amino acid in this position compared to the
other four enzymes, which have amino acids with very
small side chains. We decided to construct two
mutants of motif I at positions 55 (Y55G) and 56
(T56V) in order to ascertain whether one of them is
responsible for the phosphotransferase activity. Deter-
mination of kinetic parameters showed that substitu-
tion of Tyr55 by a smaller residue causes a dramatic
increase in the ratio of nucleotidase to phosphotrans-

ferase activity. The affinity for the substrates inosine
and IMP remains unaltered, and the effect of the
enzyme activators ATP and Mg
2+
is also unchanged
(Table 1). Furthermore, by simultaneously measuring
the rate of nucleoside and phosphate production from
IMP and monophosphate synthesis from inosine (see
Experimental procedures), we observed that, in the
wild-type, the transfer of phosphate to inosine
accounts for 53% of the phosphate produced from
IMP, but only for 11% with the mutated Y55G
enzyme under the same experimental conditions
(Fig. 2). However, substituting Thr56 by Val produced
an enzyme with a lower turnover but the same func-
tional characteristics as the wild-type enzyme, included
the nucleotidase ⁄ phosphotransferase ratio (Table 1).
Regulatory sites
Effector site 1 is located near subunit interface A, and
is believed to interact with the activator Ap
4
A, by
binding one adenosine moiety in each subunit (Fig. 3).
We mutated Arg144, which has been proposed to bind
the phosphate moiety of nucleotide activators, Ile152,
which is thought to interact with the adenylic moiety
of the activator, and Asn154, which probably interacts
with the purine ring through a hydrogen bond [11],
substituting these amino acids by negatively charged
residues. For putative effector site 2, we mutated

Phe127 and His428, as adenosine is presumably
stacked between these two residues, and Met436,
which forms a hydrogen bond with the purine ring
through its carbonylic group [11] (Fig. 3). We substi-
tuted the first two residues by a negatively charged res-
idue to discourage stacking, and replaced Met436 with
a bulkier amino acid (Trp).
Putative effector site 1
None of the mutations produced had a significant
effect on K
m
for the two principal substrates (Table 2).
Table 1. Effect of point mutations on various kinetic parameters of recombinant bovine cN-II. Nucleotidase and phosphotransferase activi-
ties were measured as described in Experimental procedures. Values are means ± SD of at least three independent assays. k
cat
refers to
nucleotidase activity and is measured at saturating concentrations of IMP and sub-saturating concentrations of inosine. The K
50
for ATP was
measured as phosphotransferase activity, while for Mg
2+
, it was measured as nucleotidase activity.
cN-II
Nucleotidase ⁄
phosphotransferase
K
m
(inosine)
(m
M)

K
m
(IMP)
(m
M) k
cat
(s
)1
)
K
50
(Mg
2+
)
(m
M)
K
50
(ATP)
(m
M)
Wild-type 1.8 ± 0.1 1.0 ± 0.2 0.1 ± 0.02 59.0 ± 6 2.0 ± 0.5 1.0 ± 0.3
Y55G 7.6 ± 2.1 1.3 ± 0.1 0.1 ± 0.03 323.0 ± 15 5.0 ± 2 0.40 ± 0.4
T56V 1.9 ± 0.1 1.0 ± 0.2 0.1 ± 0.02 6.0 ± 1 4.0 ± 1 0.80 ± 0.2
50
Rate of product formation (%)
100
0
Ino
IMP

P
i
Ino H
2
O
E + IMP
E + IMP E-P
E + P
i
Ino
Fig. 2. Rate of inosine, IMP and P
i
production catalysed by wild-
type cN-II (black bars) or mutant Y55G (white bars) in the presence
of 2 m
M IMP and 1.4 mM inosine (Ino) as substrates. The assays
were performed as described in Experimental procedures. For wild-
type and Y55G, 100% activity corresponds to 32 and 18 UÆmL
)1
,
respectively. The upper scheme indicates the catalytic mechanism
of cN-II. E, enzyme; P, phosphate. The products measured are
boxed.
R. Pesi et al. Regulation of cytosolic 5¢-nucleotidase
FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS 4865
Mutant R144E showed an altered affinity for all the
activators tested, N154D was normally activated by
ATP and BPG but activation by ADP and Ap
4
A was

completely prevented, as is the case for I152D, which
showed a higher K
50
for ATP and BPG.
Putative effector site 2
Study of cN-II crystallized in the presence of adeno-
sine indicates that this nucleoside binds to this site,
even though only the amino acid residues involved in
the binding of the purine base were identified, while
the ribose moiety was completely disordered [11].
Mutation of two residues involved in the binding of
adenine (F127E and M436W) had no effect on the
functional characteristics of the enzyme, except for an
increase in the K
50
for Ap
4
A for mutant M436W. The
mutations did not alter the activatory capacity of all
the compounds tested. The mutant H428D was almost
insensitive to all known activators of cN-II (Table 2).
CN-II subunit oligomerization
Using ‘FirstGlance in Jmol’ ( />fgij/), we analysed the cN-II crystal structure described
by Wallde
´
n et al. [11]. We mutated three amino acids
at interface A (Phe36, Tyr115 and Asp396) and two at
interface B (Lys311 and Gly319).
Interface A
In the model proposed on the basis of the crystal struc-

ture, interface A is involved in formation of a dimeric
structure (Fig. 3). Fifty-three amino acids contribute to
interaction of the two monomers by forming both
hydrogen bonds and salt bridges. Some of these residues
are near to effector site 1. We designed our mutants in
Fig. 3. Model of the homotetrameric quaternary structure of cN-II
showing interfaces A and B and the Mg
2+
site. The inset shows
the tertiary structure of each subunit. Effector sites 1 and 2 and
the active site are shown.
Table 2. Effect of point mutations on various kinetic parameters of recombinant bovine cN-II. Nucleotidase and phosphotransferase activi-
ties were measured as described in Experimental procedures. Values are means ± SD of at least three independent assays. K
50
values for
P
i
, ATP, ADP and Ap
4
A were measured as phosphotransferase activity, while those for BPG and Mg
2+
were measured as nucleotidase activ-
ity. The extent of activation, when present, was between 5- and 10-fold. (1) Mutation in putative effector site 1; (2) mutation in putative
effector site 2. NA, no activation.
cN-II
K
m
(IMP)
(m
M)

K
m
(inosine)
(m
M)
K
50
(Mg
2+
)
(m
M)
K
50
(P
i
)
(m
M)
K
50
(BPG)
(m
M)
K
50
(ATP)
(m
M)
K

50
(ADP)
(m
M)
K
50
(Ap
4
A)
(m
M)
Wild-type 0.10 ± 0.02 1.0 ± 0.2 2.0 ± 0.5 2.0 ± 0.3 0.3 ± 0.06 1.0 ± 0.3 2.2 ± 0.5 0.1 ± 0.05
R144E (1) 0.10 ± 0.05 1.0 ± 0.2 2.0 ± 1.0 6.5 ± 1.0 5.0 ± 1.50 30.0 ± 5.0 33.0 ± 6.0 1.6 ± 1.00
N154D (1) 0.1 ± 0.04 2.5 ± 0.5 0.3 ± 0.3 4.0 ± 1.0 0.3 ± 0.07 0.5 ± 0.2 NA NA
I152D (1) 0.2 ± 0.05 0.7 ± 0.2 0.9 ± 0.8 3.5 ± 1.2 5.0 ± 2.00 20.0 ± 4.0 NA NA
F127E (2) 0.1 ± 0.03 1.0 ± 0.3 0.9 ± 0.7 1.5 ± 0.5 0.7 ± 0.10 1.5 ± 1.0 2.5 ± 0.7 0.2 ± 0.03
M436W (2) 0.1 ± 0.02 0.9 ± 0.3 1.5 ± 1.0 1.5 ± 0.5 0.5 ± 0.06 1.5 ± 1.0 1.5 ± 1.0 1.0 ± 0.50
H428D (2) 0.3 ± 0.05 0.6 ± 0.5 1.5 ± 1.0 0.4 ± 0.3 NA NA NA NA
Regulation of cytosolic 5¢-nucleotidase R. Pesi et al.
4866 FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS
an attempt to interfere with monomer aggregation via
interface A by introducing or deleting charged residues
or altering molecular hindrance. One of the mutated
amino acids (Tyr115) is located very close to Arg144,
which is part of effector site 1 (Fig. S1). The mutant
F36R was inactive, while Y115A behaved normally and
D396A was only activated by BPG (Table 3).
Interface B
The structure described by Wallde
´

n et al. [11] indicates
that the active tetramer is formed by interaction between
two identical dimers via interface B: this interface con-
tains 28 aminoacid residues, of which 8 make hydrogen
bonds (Fig. 3). We constructed two mutants by introdu-
cing a negative charged residue or substituting a residue
with a non polar group (K311A), thus forming both
hydrogen and van der Waals interactions. The two
mutants showed kinetic characteristics similar to those
of the wild-type (Table 3). FPLC gel-filtration chroma-
tography of purified recombinant wild-type cN-II indi-
cated that this enzyme exists in solution as a tetramer
(molecular mass 260 kDa) (Fig. 4). Actually, the chro-
matogram showed two peaks, but the first was inactive.
Immunoblotting of both peaks indicated cross-reactivity
with specific antibodies against cN-II, but the first inac-
tive peak disappeared when Escherichia coli cells
expressing cN-II were grown at 20 °C instead of 37 °C
(data not shown). A dimeric active form (130 kDa) was
present in addition to the tetramer for mutants K311A
and G319D (mutations at interface B) (Fig. 4).
Finally, we investigated the chromatographic behav-
iour of recombinant wild-type enzyme in the presence of
ATP and Mg
2+
as an activator and P
i
as an inhibitor.
Figure 5 shows that the enzyme exists and functions as a
tetramer irrespectively of the presence of activator or

inhibitor. We also performed light-scattering measure-
ments of the enzyme alone or in the presence of its effec-
tors. There was no change in molecular mass in the
presence of enzyme activators or inhibitors (Table 4).
Discussion
Active site
The soluble 5¢-nucleotidase family shares four con-
served motifs with the HAD superfamily that are
involved in the reaction mechanism [5]. Resolution of
the crystal structure of some family members indicated
that the active site contains all the conserved motifs,
whose role in catalysis was determined through a
mechanistic approach [5,11,13]. Other than these con-
served motifs, cN-IA, cN-IB, cN-II, cN-III and the
cytosolic and mitochondrial 5¢(3¢)-deoxyribonucleotid-
ases differ considerably in their primary structure.
Despite this poor similarity, there is much evidence to
indicate that all members of soluble 5¢-nucleotidase
family share the same reaction mechanism, proceeding
through a covalent phospho-enzyme intermediate
[13,14]. Phospho-cN-II has been isolated, and the
phosphate has been found to be localized to a con-
served aspartate residue (D52) in the first of the four
conserved motifs [13]. Formation of a phospho-inter-
mediate suggests the possibility that the enzyme cataly-
ses transfer of phosphate to a suitable acceptor [15].
A number of HAD superfamily members are able to
catalyse a phosphotransferase reaction, including at
least two soluble nucleotidases (cN-II and cN-III). The
aligned sequence of motif I of soluble nucleotidases

indicates that both nucleotidases with phosphotransfer-
ase activity had a Thr instead of Val in the fifth posi-
tion after the phosphorylated Asp, and a relatively
bulky residue in the fourth position instead of Gly,
which is present in all the other nucleotidases. On this
basis, we mutated two residues in motif I, Thr56 and
Tyr55. Thr56 was substituted by Val, and Tyr55 was
substituted by Gly. Our results indicate that the pres-
ence of Thr or Val in position 56 is not relevant for
the nucleotidase ⁄ phosphotransferase ratio. However,
the increase in flexibility very close to Asp54, which
Table 3. Effect of point mutations on various kinetic parameters of recombinant bovine cN-II. Nucleotidase and phosphotransferase activi-
ties were measured as described in Experimental procedures. K
50
values for P
i
, ATP, ADP and Ap
4
A were measured as phosphotransferase
activity, while those for BPG and Mg
2+
were measured as nucleotidase activity. The extent of activation, when present, was between
5- and 10-fold. Values are means ± SD of at least three independent assays. (A) Mutation in interface A; (B) mutation in interface B. NM,
not measurable. NA, no activation.
cN-II
K
m
(IMP)
(m
M)

K
m
(inosine)
(m
M)
K
50
(Mg
2+
)
(m
M)
K
50
(P
i
)
(m
M)
K
50
(BPG)
(m
M)
K
50
(ATP)
(m
M)
K

50
(ADP)
(m
M)
K
50
(Ap
4
A)
(m
M)
Wild-type 0.10 ± 0.02 1.0 ± 0.2 2.0 ± 0.5 2.0 ± 0.3 0.3 ± 0.06 1.0 ± 0.3 2.2 ± 0.5 0.10 ± 0.05
F36R (A) NM NM NM NM NM NM NM NM
Y115A (A) 0.06 ± 0.05 2.0 ± 0.5 1.0 ± 0.7 1.1 ± 0.5 0.3 ± 0.05 2.0 ± 0.4 NA NA
D396A (A) 0.10 ± 0.03 0.5 ± 0.5 0.2 ± 0.06 2.6 ± 0.6 0.5 ± 0.05 NA NA NA
K311A (B) 0.04 ± 0.03 0.2 ± 0.5 0.7 ± 0.5 2.0 ± 0.4 0.5 ± 0.04 7.3 ± 1.5 1.00 ± 0.6 0.25 ± 0.07
G319D (B) 0.05 ± 0.04 0.5 ± 0.5 2.5 ± 0.6 1.6 ± 0.7 0.5 ± 0.03 3.0 ± 1.0 2.00 ± 0.5 0.25 ± 0.05
R. Pesi et al. Regulation of cytosolic 5¢-nucleotidase
FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS 4867
participates in binding of the phosphate [13], obtained
with the Y55G mutant, causes an increase in turnover
for the nucleotidase reaction, and, as a consequence, a
large increase in the nucleotidase ⁄ phosphotransferase
ratio.
Regulatory sites
cN-II is activated by BPG, a number of triphosphate
and diphosphate nucleosides (ATP and ADP being the
best activators), Ap
4
A and polyphosphates. Con-

versely, orthophosphate has an inhibitory effect.
Kinetic studies have indicated that ATP (and presum-
ably other phosphorylated compounds) causes stabil-
ization of an enzyme form with a high k
cat
, without
substantial alteration of the K
m
for the substrates,
while orthophosphate stabilizes a form at high K
m
with
no effect on k
cat
. If ATP and phosphate are present at
the same time, an enzyme form with high K
m
and high
k
cat
is observed [9]. Therefore, depending on the effec-
tor, cN-II may be present as one of two structures
.
.
.
.

.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
670 kDa
210 kDa
150 kDa
443 kDa
Activity (Arbitrary units) (–)
840
72
Retention time (min)
A
B
C
Abs
254 nm
(Arbitrary units) (–)
.

.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
•
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
Fig. 5. FPLC profiles of wild-type cN-II alone (A), or in the presence
of 5 m
M ATP and 10 mM MgCl
2
(B), or in the presence of 5 mM P
i
and 10 mM MgCl
2
(C). The column, prepared and eluted as
described in Experimental procedures, was loaded with approxi-
mately 150 lg of purified protein, and the activity was measured as
the rate of IMP production in the presence of inosine (phospho-
transferase activity) as described in Experimental procedures.
Arrows indicate the molecular mass of the marker proteins thyro-
globulin (670 kDa), apoferritin (443 kDa), b-amylase (210 kDa) and
alcohol dehydrogenase (150 kDa).

.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
670 kDa
210 kDa
150 kDa
443 kDa
Activity (Arbitrary units) (–)
840
72

Retention time (min)
Abs
254 nm
(Arbitrary units) (–)
WT
G319D
K311A

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
•
Fig. 4. FPLC profiles of wild-type cN-II, and two mutants in inter-
face B: G319D and K311A. The column, prepared and eluted as
described in Experimental procedures, was loaded with approxi-
mately 150 lg of purified protein, and the activity was measured as
the rate of phosphate production in the presence of IMP (phospha-
tase) as described in Experimental procedures. Arrows indicate the
molecular mass of the marker proteins thyroglobulin (670 kDa),

apoferritin (443 kDa), b-amylase (210 kDa) and alcohol dehydroge-
nase (150 kDa).
Regulation of cytosolic 5¢-nucleotidase R. Pesi et al.
4868 FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS
with high and low K
m
. In addition, a low and high k
cat
may be associated with each structure. It has also been
suggested, on the basis of kinetic characterization, that
the enzyme has at least three effector sites, one for
ATP, one for ADP and one for BPG [5,12]. On the
basis of the results obtained from crystallization of a
truncated form of cN-II [11], we constructed point
mutants for a number of amino acid residues located
in putative effector sites 1 and 2 and involved in bind-
ing of adenylic nucleotides and BPG.
Our results partially confirm the molecular model-
ling, suggesting that effector site 1 is the binding site
for Ap
4
A and that ADP binds this site as well. Ap
4
A
binds between two subunits with one adenosine moiety
in each subunit [11], but ADP may possibly fill the
whole site in each subunit and attain the same result.
Mutant N154D shows a decrease in the value of K
50
for Mg

2+
. Moreover, the mutated enzyme showed
some activity even in the absence of added Mg
2+
,
whereas all the other active mutants, as well as the
wild-type enzyme, are completely dependent on addi-
tion of the metal. This behaviour may be due to the
proximity of Asn154 to the active site (Fig. S2), and, in
particular, a modification in the interaction of this resi-
due with Asp351 (directly involved in Mg
2+
binding)
and His352 both belonging to motif III. Remarkably,
mutant D396A, in which the mutation is located in
interface A, very close to effector site 1 and the active
site, shows a similar increase in affinity for Mg
2+
. Our
results also indicate that ATP and BPG probably bind
to a different site and that this is effector site 2. Point
mutations at this site result in either generalized impair-
ment of enzyme regulation or characteristics very simi-
lar to those of the wild-type. As virtually all purine and
pyrimidine triphosphates have some activatory effect
on cN-II, the interaction site is likely to be more selec-
tive for the phosphorylated sugar moiety than for the
base. Therefore, binding of adenosine to effector site 2,
as demonstrated by the cN-II crystal structure [11],
may not be indicative of the location of the nucleoside

triphosphate effector site.
CN-II subunit oligomerization
CN-II has been purified from various sources and has
always been described as a homotetramer [16,17]. The
crystal structure suggests that the tetramer arises from
interaction of two dimers through interface B. Muta-
tions in interface A, through which two monomers
interact, and which is very close to the effector site 1,
either resulted in a completely inactive enzyme or
strongly interfered with activation by ADP and Ap
4
A.
This indirectly confirms that effector site 1 is specific
for these compounds. Mutation of amino acid residues
located in interface B generated proteins for which an
active dimeric form was detected in addition to the tet-
ramer. However, stabilization of the dimeric form had
no effect on the catalytic capacity of the enzyme. Our
results show that the monomer is probably inactive,
and the dimer is the smallest active cN-II quaternary
structure. FPLC analysis of purified recombinant
enzymes shows a heavier protein (720 kDa) in addition
to the proteins at the expected molecular mass. This
protein was an unusual oligomerization state of a pro-
tein that was identified by immunoblotting as cN-II
but completely inactive. E. coli produces a small
amount of cN-II that is correctly folded and a large
amount of incorrectly folded and insoluble protein. It
is conceivable that a cN-II protein that is incorrectly
folded but still soluble is also produced. A decrease in

growth temperature resulted in disappearance of the
heavier inactive peak. Furthermore, FPLC analysis of
the recombinant wild-type cN-II showed the presence
of the tetrameric active form irrespective of the pres-
ence of effectors. This result disagrees with other
results obtained on recombinant human cN-II [10],
which is almost identical to our bovine enzyme. To
support our finding, we used light scattering, a tech-
nique that, unlike gel-filtration chromatography, does
not require protein dilution. However, light-scattering
experiments confirmed that enzyme activation or inhi-
bition is not followed by a change in enzyme subunit
oligomerization.
In conclusion, our results confirm that there is indeed
an activatory site specific for Ap
4
A located at the inter-
face between two subunits (interface A). This site may
also accommodate ADP at lower affinity, but resulting
in the same level of activation as Ap
4
A. ATP binds to
a different site; however, it was not possible to confirm
this as effector site 2 indicated by the crystal structure
obtained in the presence of adenosine. In contrast to
Table 4. Variation of the protein molecular mass in the presence
of specific effectors as estimated by the ratio of the scattered light
intensity. Values are means ± SD of at least three measurements.
Sample Ratio
a

cN-II in 20 mM Tris ⁄ HCl, pH 8, +0.2 M NaCl (control)
(control), +5 m
M ATP 1.08 ± 0.08
(control), +5 m
M ⁄ 10 mM ATP ⁄ Mg
2+
1.16 ± 0.06
(control), +10 m
M ATP 1.13 ± 0.09
(control), +10 m
M ⁄ 10 mM ATP ⁄ Mg
2+
1.15 ± 0.08
(control), +5 m
M P
i
1.08 ± 0.07
(control), +5 m
M ⁄ 10 mM P
i
⁄ Mg
2+
0.95 ± 0.07
(control), +100 l
M AP
4
A 1.11 ± 0.10
a
Ratio of molecular mass in the presence of specific effectors to
that of the free protein.

R. Pesi et al. Regulation of cytosolic 5¢-nucleotidase
FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS 4869
the suggestion by Wallde
´
n [11], we demonstrated that
BPG does not bind effector site 1. Finally, mutations
at the interfaces and gel-filtration experiments indicated
that the tetramer is the major quaternary structure of
cN-II, irrespectively of the presence of effectors, but
that the dimer may also be stable and active.
Experimental procedures
Materials
DpnI was provided by New England Biolabs (Ipswich, MA,
USA) and Pfu DNA polymerase by Promega (Madison,
WI, USA). Ni-NTA agars was purchased from Qiagen
(Valencia, CA, USA). [8-
14
C]-inosine and [8-
14
C]-IMP were
obtained from Sigma-Aldrich (St Louis, MO, USA). Opti-
phase ‘HiSafe’ 3 scintillation liquid was obtained from Per-
kin-Elmer (Waltham, MA, USA). Superdex-200 was
purchased from GE Healthcare (Piscataway, NJ, USA). All
other chemicals were of reagent grade.
Site-directed mutagenesis
The point mutants F36R, Y115A, F127E, I152D, R144E,
N154D, K311A, G319D, D396A and M436W were
obtained using the PCR-based site-directed mutagenesis
method described by Fisher and Pei [18], and the point

mutants Y55G, T56V and H428D were produced as
described in the QuikChange
Ò
site-directed mutagenesis kit
manual (Stratagene, La Jolla, CA, USA). The primers used
are listed in Table 5.
Expression and purification of the recombinant
proteins
Expression of the recombinant mutants was performed as
previously described [19]. The 6· His-tagged proteins were
purified using the Ni-NTA agar method as described in the
QIAexpressionistÔ handbook (Qiagen). The protein con-
centration was determined using the Bradford method [20],
with BSA as the standard.
Enzyme assays
The nucleotidase activity of cN-II and its mutants was mea-
sured as the rate of [8-
14
C]-inosine formation from 2 mm
[8-
14
C]-IMP in the presence of 1.4 mm inosine, 20 mm
MgCl
2
, 4.5 mm ATP and 5 mm dithiothreitol, as previously
described [9]. Phosphotransferase activity was measured as
the rate of [8-
14
C]-IMP formation from 1.4 mm [8-
14

C]-ino-
sine, in the presence of 2 mm IMP, 20 mm MgCl
2
, 4.5 mm
ATP and 5 mm dithiothreitol, as previously described [9].
For determination of kinetic parameters (K
m
and k
cat
), the
concentration of the labelled substrates ranged from 0.02 to
4mm. A plot of the dependence of the rate of phospho-
transferase activity on MgCl
2
, ATP, ADP and BPG con-
centration was used to determine the value of K
50
for these
compounds. Under these experimental conditions, the accu-
mulation of radiolabelled inosine (nucleotidase activity) rep-
resents the sum of the phosphatase and the
phosphotransferase activities. It has been previously
reported that, at a concentration close to the K
m
value
(1.4 mm), inosine reduces phosphatase activity to 50%
without affecting the V
max
for both reactions [9]. Thus, the
expected value of 2 was determined for the ratio between

nucleotidase and phosphotransferase activities under the
experimental conditions used for the wild-type recombinant
cN-II assay. Accordingly, an alteration of this ratio for a
mutant was considered as caused either by an alteration of
the K
m
value for one of the two substrates or by a variation
of the k
cat
value for one of the two activities. When
required, the rate of phosphate formation (phosphatase
activity) was measured as described by Chifflet et al. [21].
One unit of enzyme activity is the amount of enzyme
Table 5. Primers used for site-directed mutagenesis.
Mutant Forward primer (5¢ to 3¢) Reverse primer (5¢ to 3¢)
F36R GCGCGTGAACCGGAGTT ACCCGATGATAGGCTTC
Y115A CGCTGGAAACCTCTTGG GCATCAACTTTCAAAAGAT
F127E CGAGATAAGGGGACCAG TTAAATCCATGTGCACAG
R144E AGAAGATGACACTGAAAG TGAATAAATTTATTTGGATAC
I152D CGATCTGAACACACTATTC TAAAATCTTTCAGTGTCAT
N154D GGACACACTATTCAACCT AGAATGTAAAATCTTTCAGT
K311A CGCGCTGAAAATTGGTAC CCAGTTTTAGTATCCACC
G319D GGACCCCTTACAGCA GTGTAGGTACCAATTTTC
D396A GGCTATTTTCTTGGCTGA AAGCTCTGAAGCTCTTC
M436W GTGGATGGGGAGCCTG CCGTAGCACATGTCCA
H428D AAGAAAGTAACTGACGACATGGACATGTG CACATGTCCATGTCGTCAGTTACTTTCTT
Y55G AGTGTTTTGGGTTTGACATGGATGGCACACTTGCTG CAGCAAGTGTGCCATCCATGTCAAACCCAAAACACT
T56V AGTGTTTTGGGTTTGACATGGATTATGTGCTTGCTG CAGCAAGCACATAATCCATGTCAAACCCAAAACACT
Regulation of cytosolic 5¢-nucleotidase R. Pesi et al.
4870 FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS

required to convert 1 lmol of substrate to product per min-
ute under the assay conditions.
Gel-filtration chromatography
The gel-filtration chromatography was performed on a
FPLC system utilizing a Superdex-200 column
(1.2 · 32 cm). Purified wild-type or mutant cN-II (150 lg)
was loaded onto the column, and the chromatography was
performed at a flow rate of 0.3 mLÆmin
)1
using 50 mm
Tris ⁄ HCl, pH 7.4, with addition of 200 mm NaCl. Frac-
tions of 0.1 mL were collected.
Light scattering
The intensity of light scattered at 90° from the incident
beam was measured using a spectrofluorometer (Fluoro-
max-4; Horiba, Edison, NJ, USA). The wavelength of the
incident light was 350 nm, and the band-pass used in the
excitation and emission monochromators was 1.0 nm. The
sample cell (fluorescence cell 1 · 1cm
2
cross-section) was
cleaned using water filtered through 0.22 lm pore mem-
brane. Protein and additive solution were passed through
identical filters. All reagents were dissolved in 20 mm
Tris ⁄ HCl + 0.2 m NaCl, pH 8, and the enzyme concentra-
tion ranged between 0.09 and 0.14 mgÆmL
)1
. Scattering
intensities were compared between samples that had the
same protein concentration and refractive index. Under

these conditions, their ratio, according to Parr and Ham-
mes [22], is proportional to the ratio of molecular mass.
Acknowledgements
We would like to thank Dr Giovanni Strambini and
Dr Margherita Gonelli of the Institute of Biophysics
(National Research Centre, Pisa, Italy) for the light-
scattering analysis. We would also like to thank Dr
Adrian Wallwork for careful language revision of the
manuscript. This work was supported by a grant from
the Ministero dell’Istruzione, dell’Universita
`
e della
Ricerca and by local funds from the University of
Pisa.
References
1 Tozzi MG, Camici M, Pesi R, Allegrini S, Sgarrella F
& Ipata PL (1991) Nucleoside phosphotransferase activ-
ity of human colon carcinoma cytosolic 5¢-nucleotidase.
Arch Biochem Biophys 291, 212–217.
2 Banditelli S, Baiocchi C, Pesi R, Allegrini S, Turriani
M, Ipata PL, Camici M & Tozzi MG (1996) The
phosphotransferase activity of cytosolic 5¢-nucleotidase;
a purine analog phosphorylating enzyme. Int J Biochem
Cell Biol 2, 711–720.
3 Hunsucker SA, Mitchell BS & Spychala J (2005) The
5¢-nucleotidases as regulators of nucleotide and drug
metabolism. Pharmacol Ther 107, 1–30.
4 Galmarini CM, Jordheim L & Dumontet C (2003) Role
of IMP-selective 5¢-nucleotidase (cN-II) in haematologi-
cal malignancies. Leuk Lymphoma 44, 1105–1111.

5 Allegrini S, Scaloni A, Careddu MG, Cuccu G, D’Am-
brosio C, Pesi R, Camici M, Ferrara L & Tozzi MG
(2004) Mechanistic studies on bovine cytosolic 5¢-nucle-
otidase II, an enzyme belonging to the HAD superfam-
ily. Eur J Biochem 271, 4881–4891.
6 Careddu MG, Allegrini S, Pesi R, Camici M, Garcia-
Gil M & Tozzi MG (2008) Knockdown of cytosolic
5¢-nucleotidase II (cN-II) reveals that its activity is
essential for survival in astrocytoma cells. Biochim
Biophys Acta 1783, 1529–1535.
7 Rampazzo C, Gazziola C, Ferraro P, Gallinaro L,
Johansson M, Reichard P & Bianchi V (1999) Human
high-K
m
5¢-nucleotidase: effects of overexpression of
the cloned cDNA in cultured human cells. Eur J
Biochem 261, 689–697.
8 Spychala J, Madrid-Marina V & Fox IH (1988) High
K
m
soluble 5¢-nucleotidase from human placenta. Prop-
erties and allosteric regulation by IMP and ATP. J Biol
Chem 263, 18759–18765.
9 Pesi R, Turriani M, Allegrini S, Scolozzi C, Camici M,
Ipata PL & Tozzi MG (1994) The bifunctional cytosolic
5¢-nucleotidase: regulation of the phosphotransferase
and nucleotidase activities. Arch Biochem Biophys 312,
75–80.
10 Spychala J, Chen V, Oka J & Mitchell BS (1999) ATP
and phosphate reciprocally affect subunit association of

human recombinant high Km 5¢-nucleotidase. Role for
the C-terminal polyglutamic acid tract in subunit associ-
ation and catalytic activity. Eur J Biochem 259, 851–
858.
11 Wallden K, Stenmark P, Nyman T, Flodin S, Graslund
S, Loppnau P, Bianchi V & Nordlund P (2007) Crystal
structure of human cytosolic 5¢-nucleotidase II: insights
into allosteric regulation and substrate recognition.
J Biol Chem 282, 17828–17836.
12 Pesi R, Baiocchi C, Allegrini S, Moretti E, Sgarrella F,
Camici M & Tozzi MG (1998) Identification, separation
and characterisation of two forms of cytosolic 5¢-nucle-
otidase ⁄ nucleoside phosphotransferase in calf thymus.
Biol Chem 379, 699–704.
13 Allegrini S, Scaloni A, Ferrara L, Pesi R, Pinna P,
Sgarrella F, Camici M, Eriksson S & Tozzi MG (2001)
Bovine cytosolic 5¢-nucleotidase acts through the forma-
tion of an aspartate 52-phosphoenzyme intermediate.
J Biol Chem 276, 33526–33532.
14 Baiocchi C, Pesi R, Camici M, Itoh R & Tozzi MG
(1996) Mechanism of the reaction catalysed by cytosolic
5¢-nucleotidase ⁄ phosphotransferase: formation of a
phosphorylated intermediate. Biochem J 317, 797–801.
R. Pesi et al. Regulation of cytosolic 5¢-nucleotidase
FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS 4871
15 Fersht A (1999) Structure and Mechanism in Protein
Science – A Guide to Enzyme Catalysis and Protein
Folding, 2nd edn. W.H. Freeman and Company, New
York, NY.
16 Zimmermann H (1992) 5¢-nucleotidase: molecular struc-

ture and functional aspects. Biochem J 285, 345–365.
17 Itoh R (1993) IMP–GMP 5¢-nucleotidase. Comp
Biochem Physiol B 105, 13–19.
18 Fisher CL & Pei GK (1997) Modification of a PCR-
based site-directed mutagenesis method. BioTechniques
23, 570–571.
19 Allegrini S, Pesi R, Tozzi MG, Fiol CJ, Johnson RB &
Eriksson S (1997) Bovine cytosolic IMP ⁄ GMP-specific
5¢-nucleotidase: cloning and expression of active enzyme
in Escherichia coli. Biochem J 328, 483–487.
20 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein uti-
lizing the principle of protein–dye binding. Anal Bio-
chem 72, 248–254.
21 Chifflet S, Torriglia A, Chiesa R & Tolosa S (1988)
A method for the determination of inorganic phosphate
in the presence of labile organic phosphate and high
concentrations of protein: application to lens ATPases.
Anal Biochem 168, 1–4.
22 Parr GR & Hammes GG (1975) Subunit dissociation
and unfolding of rabbit muscle phosphofructokinase by
guanidine by hydrochloride. Biochemistry 14, 1600–
1605.
Supporting information
The following supplementary material is available:
Fig. S1. Proximity between effector site 1 and inter-
face A.
Fig. S2. Proximity between effector site 1 and the
active site.
This supplementary material can be found in the

online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Regulation of cytosolic 5¢-nucleotidase R. Pesi et al.
4872 FEBS Journal 277 (2010) 4863–4872 ª 2010 The Authors Journal compilation ª 2010 FEBS