Structural and functional studies of the human
phosphoribosyltransferase domain containing protein 1
Martin Welin
1,
*, Louise Egeblad
2,
*, Andreas Johansson
1
,Pa
˚
l Stenmark
1,
, Liya Wang
2
,
Susanne Flodin
1
, Tomas Nyman
1
, Lionel Tre
´
saugues
1
, Tetyana Kotenyova
1
, Ida Johansson
1
,
Staffan Eriksson
2
, Hans Eklund
3
and Pa
¨
r Nordlund
1
1 Structural Genomics Consortium, Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Sweden
2 Department of Anatomy, Physiology and Biochemistry, Swedish University of Agricultural Sciences, Uppsala, Sweden
3 Department of Molecular Biology, Biomedical Center, Swedish University of Agricultural Sciences, Uppsala, Sweden
Keywords
characterization; crystal structure; homolog;
HPRT; phosphoribosyltransferase; PRTFDC1
Correspondence
P. Nordlund or S. Eriksson, Structural
Genomics Consortium, Department of
Medical Biochemistry and Biophysics,
Karolinska Institutet, S-17177 Stockholm,
Sweden; Department of Anatomy,
Physiology and Biochemistry, Swedish
University of Agricultural Sciences, Box 575,
SE-75123 Uppsala, Sweden
Fax: +46 8 524 868 50; +46 18 55 0762
Tel: +46 8 524 868 60; +46 18 471 4187
E-mail: par.nordlund@ki.se;
staffan.eriksson@afb.slu.se
*These authors contributed equally to this
work
Present address
Center for Biomembrane Research,
Department of Biochemistry and Biophysics,
Stockholm University, Sweden
Database
Structural data are available in the Protein
Data Bank under the accession number
2JBH
(Received 15 November 2009, revised 18
August 2010, accepted 30 September 2010)
doi:10.1111/j.1742-4658.2010.07897.x
Human hypoxanthine-guanine phosphoribosyltransferase (HPRT) (EC
2.4.2.8) catalyzes the conversion of hypoxanthine and guanine to their
respective nucleoside monophosphates. Human HPRT deficiency as a result
of genetic mutations is linked to both Lesch–Nyhan disease and gout. In
the present study, we have characterized phosphoribosyltransferase domain
containing protein 1 (PRTFDC1), a human HPRT homolog of unknown
function. The PRTFDC1 structure has been determined at 1.7 A
˚
resolution
with bound GMP. The overall structure and GMP binding mode are very
similar to that observed for HPRT. Using a thermal-melt assay, a nucleo-
tide metabolome library was screened against PRTFDC1 and revealed that
hypoxanthine and guanine specifically interacted with the enzyme. It was
subsequently confirmed that PRTFDC1 could convert these two bases into
their corresponding nucleoside monophosphate. However, the catalytic effi-
ciency (k
cat
⁄ K
m
) of PRTFDC1 towards hypoxanthine and guanine was
only 0.26% and 0.09%, respectively, of that of HPRT. This low activity
could be explained by the fact that PRTFDC1 has a Gly in the position of
the proposed catalytic Asp of HPRT. In PRTFDC1, a water molecule at
the position of the aspartic acid side chain position in HPRT might be
responsible for the low activity observed by acting as a weak base. The
data obtained in the present study indicate that PRTFDC1 does not have
a direct catalytic role in the nucleotide salvage pathway.
Structured digital abstract
l
MINT-7996314: PRTFDC1 (uniprotkb:Q9NRG1 ) and PRTFDC1 (uniprotkb:Q9NRG1) bind
(
MI:0407)byx-ray crystallography (MI:0114)
Abbreviations
DSLS, differential static light scattering; HPRT, human hypoxanthineguanine phosphoribosyltransferase; ImmGP, immucillinGP; PRPP,
a-
D-5-phosphoribosyl 1-pyrophosphate; PRTFDC1, phosphoribosyltransferase domain containing protein 1; TCEP, tris(2-carboxyethyl)phosphine.
4920 FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS
Introduction
Human hypoxanthine guanine phosphoribosyltransfer-
ase (HPRT) (EC 2.4.2.8) is an important enzyme in
the salvage pathway of purine nucleotides. It catalyzes
the transfer of a hypoxanthine or guanine base to
a-d-5-phosphoribosyl 1-pyrophosphate (PRPP), pro-
ducing IMP or GMP and pyrophosphate. Several of
the identified mutations leading to disease are spread
over the entire protein, and are not just restricted to
the active site [1,2]. Among the metabolic consequences
of having HPRT deficiency are an overproduction of
uric acid and elevated levels of PRPP [3]. Patients hav-
ing mutations resulting in partial HPRT deficiency
often suffer from gouty arthritis. However, more
severe deficiencies lead to Lesch–Nyhan syndrome, a
disease with symptoms of self mutilation and mental
retardation [4]. The underlying mechanisms of Lesch–
Nyhan syndrome are still not well understood, and the
absence of HPRT gives rise to a complex altered gene
expression profile [5]. Tissue culture, HPRT knockout
mice and neonatal dopamine lesion models have been
used to elucidate the biochemical and physiological
processes taking place in these diseases [6]. Further-
more, protozoan parasites have no de novo purine
nucleotide synthesis and must rely solely on salvage
enzymes, which therefore makes HPRT a potential
antiparasital drug target [7].
A homolog with 65% identity to HPRT, phos-
phoribosyltransferase domain containing protein 1
(PRTFDC1) is present in the human genome. A recent
study identified this homologue as a gene potentially
involved in the development of ovarian cancer. In a
genome-wide screening for differently methylated pro-
moter islands, PRTFDC1 was transcribed in ovarian
cancer cell lines with unmethylated DNA but not in
cancer cell lines with methylated DNA [8]. A similar
study demonstrated that restoring PRTFDC1 in oral
squamous cell carcinomas cells lacking its expression
inhibited cell growth [9].
HPRT is well characterized using both biochemical
and structural methods, whereas PRTFDC1 is poorly
characterized. Sequence analysis of a wide range of
species indicates that HPRT have eleven completely
conserved amino acids, whereas PRTFDC1 genes do
not show full conservation of these eleven residues
[10,11]. Because one of the differences in PRTFDC1 is
Gly145, which corresponds to the proposed catalytic
residue Asp137 in HPRT [12], it was suggested that
the PRTFDC1 proteins have lost their phosphoribosyl-
transferase activity [10].
The structural characterization of numerous com-
plexes of the human HPRT and several bacterial and
protozoan HPRTs have been undertaken [13–17]. The
structure of human HPRT can be divided into two
domains: a core domain and a hood domain [14,18].
The HPRTs have a flexible loop, referred to as loop
II, that covers the active site in the substrate bound
forms [13,14,17,19]. Human HPRT has been shown to
be a dimer or a tetramer in solution depending on
ionic strength [20].
In the present study, we report the structural and
functional characterization of the human HPRT
homolog PRTFDC1. To shed light on the potential
function of PRTFDC1, the protein was screened
against a nucleotide metabolome library containing 81
potential substrates or regulatory ligands for enzymes
in the human nucleotide metabolism. These were
selected as substrates, intermediates and products of
other enzymes in the human nucleotide metabolism.
The library was screened against PRTFDC1 using dif-
ferential static light scattering (DSLS) thermal-melt
assay. PRPP, IMP and GMP were the top hits and
therefore were identified as potential substrates, prod-
ucts or regulatory ligands for PRTFDC1. The catalytic
efficiency and substrate specificity of PRTFDC1 was
further characterized using a radiochemical assay with
tritium-labeled bases as substrates, whereas the struc-
tural basis for substrate recognition and activity was
revealed by solving the structure of PRTFDC1 in com-
plex with the product GMP at 1.7 A
˚
resolution.
Results
Overall structure
The structure of full-length human PRTFDC1 was
determined at 1.7 A
˚
resolution with two subunits in
the asymmetric unit. Most of the polypeptide chains
could be traced in the electron density, with only a few
residues at the N-terminus and two short flexible loops
left unmodeled. The 3D structures of HPRTs have
been divided into a core and a hood domain [14]. The
core domain of PRTFDC1 contains a six-stranded
twisted parallel b-sheet surrounded by three a-helices.
b4 in the core domain extends into a b-ribbon with b5,
stabilizing loop II. The hood domain is mainly built
up by residues from the C-terminus and consists of a
two-stranded anti-parallel b-sheet composed of b2 and
b9 and an a-helix from the C-terminus (Fig. 1A). The
two subunits in the asymmetric unit, together with two
symmetry-related subunits, form a plausible tetramer
similar to the tetramer in HPRT (Fig. 1B). However,
the gel filtration profile reveals a dimeric protein (data
M. Welin et al. Studies of the human PRTFDC1
FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS 4921
not shown). The addition of PRPP has been observed
to induce tetramerization for HPRTs [21,22] and the
use of 10 mm PRPP in crystallization set-up for
PRTFDC1 could explain the different oligomeric states
for PRTFDC1 in solution and crystal structure.
Nucleotide binding
In the initial electron density maps, a clear difference
density was found in the active site corresponding to a
nucleoside monophosphate, despite no nucleotide
being added to crystallization solutions. The ligand
bound was interpreted as GMP because the N2 of the
base makes a hydrogen bond to a main chain car-
bonyl. Binding of a xanthine base from XMP in the
same position, would lead to the loss of a hydrogen
bond and a less favorable interaction. To investigate
this further, the protein was heated until precipitated,
and the remaining solute was scanned using a spectro-
photometer. The UV-absorption spectrum of the solute
from the precipitated protein displayed a similar
UV-absorption spectrum as a GMP solution, suggest-
ing that GMP was bound to the protein. A similar
GMP bound enzyme was observed in a xanthine phos-
phoribosyltransferase family member in Bacillus subtil-
is [23]. A likely explanation for bound GMP is that it
originates from expression in Escherichia coli cells.
In the structure the base of bound GMP is stacked
between Val143 and Phe194. N1, N2 and O6 of the
nucleotide form hydrogen bonds to main chain atoms,
whereas the side chain of Lys173 interacts with both
O6 and N7 (Fig. 1C). The 2¢OH of the ribose is hydro-
gen bonded to the main chain and the side chain of
Asp142 (Fig. 1C). Both hydroxyl groups of the ribose
are involved in coordinating a putative Ca
2+
ion. This
Ca
2+
is likely to have been exchanged with the active
site Mg
2+
ion normally used by this family as a result
of the high concentration of Ca
2+
present in the crys-
tallization buffer. The geometry is consistent with a
Ca
2+
ion coordinated by Glu141, Asp142, 2¢ and
3¢OH of the ribose and three water molecules with
coordination distances of approximately 2.4 A
˚
. The
phosphate of the GMP is hydrogen bonded to main
chain and side chain atoms of amino acids 145–149
(Fig. 1C).
A second Ca
2+
ion is coordinated by the side chain
of Asp201, a phosphate ion and five water molecules.
The phosphate involved in this interaction is bound by
main chain interactions with Lys76 and Gly77 and
the side chain of Arg207. The phosphate has some
A
C
B
Fig. 1. The structure of PRTFDC1 (A) mono-
mer and (B) tetramer generated using sym-
metry-related molecules. Interactions with
GMP (C). GMP, phosphate and calcium ions
are shown as sticks and spheres colored
pink and black, respectively.
Studies of the human PRTFDC1 M. Welin et al.
4922 FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS
additional density, suggesting that it may be a degra-
dation product of PRPP from the crystallization
solution.
Comparison with HPRTs
Superposition of PRTFDC1 and human HPRT results
in a rmsd of 1.0–1.7 A
˚
for approximately 200 Ca
atoms depending on which HPRT complex is com-
pared (Protein Data Bank code: 1Z7G, 1D6N, 1BZY,
1HMP). The overall structures are very similar with
the exception of loop II, which, in some HPRT com-
plexes with transition state analogs, closes over the
active site. In the B subunit of PRTFDC1, this loop
interacts with a symmetry-related molecule and thus
the whole loop is visible. This open conformation of
the loop has also been observed in Giardia lamblia
guanine phosphoribosyltransferase where the confor-
mation is also stabilized by crystal contacts [24]. The
bound phosphate ion superposes well with one of the
phosphates of pyrophosphate in the transition state
analog complex of human HPRT with immucillinGP
(ImmGP) and pyrophosphate bound (Fig. 2A) [17].
The conserved Lys76 in PRTFDC1 is in a cis-peptide
conformation, which also has been observed in the
human HPRT in complex with the transition state
analog, ImmGP [17].
Screening of nucleotide metabolome library to
identify PRTFDC1 substrates
To investigate the potential function of PRTFDC1 a
DSLS thermal-melt assay was used to screen for inter-
actions with a nucleotide metabolome library contain-
ing 81 compounds of substrates, products and
regulators of other enzymes in the human nucleotide
metabolism (Table S1). The library can then be seen to
constitute the specific metabolome of the nucleotide
metabolism, which would be the most likely source for
a substrate or a regulator of PRTFDC1. The com-
pounds producing the largest increase in calculated
DT
agg
(i.e. the difference in midpoint of the aggrega-
tion process measured by DSLS) were PRPP, GMP
and IMP. Large increase in thermal shift is an indica-
tion of protein-compound binding. The high thermal
shifts for GMP, and IMP indicated that PRTFDC1
could have similar activity as HPRT or, alternatively,
be regulated by these nucleotides. Therefore,
PRTFDC1 was rescreened against nucleobases hypo-
xanthine, guanine, cytidine, uracil, adenine and xan-
thine (X), as well as their respective nucleoside
monophosphates. Although IMP and GMP produced
large shifts, adding only the bases to the enzyme did
not produce any thermal shifts. However, in the pres-
ence of 50 lm PRPP thermal shifts were observed for
hypoxanthine and guanine (Table 1). These results
imply that PRPP is required for the nucleobases to
bind. For comparison the DSLS thermal melt assay
was run for HPRT but, unfortunately HPRT gave
uninterruptable aggregation temperature profiles.
Steady-state kinetic analysis
To examine whether PRTFDC1 possessed any phos-
phoribosyltransferase activity, a radiochemical method
was used with tritium-labeled bases and PRPP as sub-
strates. Reaction products were separated from sub-
strate and quantified by using the DE-81 filter paper
S103
Y112
R207/199
K173/165
G145/D137
D142/134
E141/133
L75/67
D201/193
G77/69
G197/189
G145/D137
A
B
Fig. 2. Superposition of PRTFDC1 with HPRT-ImmGP (Protein Data Bank code: 1BZY) (A) Residues in the PRPP binding motif, nucleotides,
pyrophosphate and phosphate are shown as sticks in brown and pink, respectively. Metals are shown with same coloring scheme. (B)
Superposition of PRTFDC1 with HPRT-GMP (Protein Data Bank code: 1HMP) and HPRT-ImmGP (Protein Data Bank code: 1BZY) the struc-
tures are colored brown, white and pink. The water molecule in proximity to the Asp137 is colored in blue.
M. Welin et al. Studies of the human PRTFDC1
FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS 4923
technique taking the advantage of charge difference
between substrate and product. Phosphoribosyltransfer-
ase activity was detected with hypoxanthine and guanine
and there was no detectable activity with adenine. Sub-
sequently, the kinetic constants were determined.
Human HPRT was characterized in parallel to compare
their catalytic efficiency. The K
m
(hypoxanthine) value
for PRTFDC1 was 23.3 ± 6.8 lm, and the V
max
value
was 0.340 ± 0.037 lmolÆmin
)1
Æmg
)1
, whereas the val-
ues for HPRT were 3.9 ± 1.5 lm and 23.3 ± 2.0 lmolÆ
min
)1
Æmg
)1
, respectively (Fig. 3A,C and Table 2). With
hypoxanthine as a substrate, the K
m
value is approxi-
mately six-fold higher for PRTFDC1, and V
max
is appro-
ximately 70-fold lower compared to those of HPRT.
Thus, the k
cat
⁄ K
m
value with PRTFDC1 is only 0.26%
of that of HPRT. The K
m
(guanine) value for
PRTFDC1 was 36.1 ± 14.3 lm, and the V
max
was value
2.9 ± 0.7 lmolÆmin
)1
Æmg
)1
, whereas the values for
HPRT were 9.9 ± 0.2 lm and 899 ± 117 lmolÆmin
)1
Æ
mg
)1
, respectively (Fig. 3B,D and Table 2). This indi-
cates that, with guanine as a substrate, the K
m
value is
approximately four-fold higher, and the V
max
value is
approximately 310-fold lower compared to HPRT.
Therefore, the catalytic efficiency (k
cat
⁄ K
m
value) of
PRTFDC1 is 0.09% compared to HPRT.
Discussion
Because PRTFDC1 was annotated as a protein with
unknown function, the use of a DSLS thermal-melt
assay proved to be a good initial step for elucidating
which compounds could stabilize the protein. The
increased DT
agg
for IMP, GMP, hypoxanthine ⁄ PRPP
and guanine ⁄ PRPP indicated a similar binding profile as
for HPRT. The DSLS results also suggest a sequential
mechanism where PRPP binds first, inducing a confor-
mation change, then allowing the purine base to bind in
accordance with the mechanism of human HPRT
[19,25]. This conclusion can be made based on the lack
of thermal shifts when only the bases were added,
whereas the addition of PRPP led to a significant ther-
mal shift. Kinetic studies of PRTFDC1 showed that this
enzyme had similar substrate specificity as HPRT, with
guanine as the best substrate. However, the catalytic effi-
ciency (k
cat
⁄ K
m
) of PRTFDC1 is much lower than that
of HPRT. The differences in catalytic efficiency could be
explained by the difference in amino acid residues
involved in catalysis between PRTFDC1 and HPRT.
In the structure, residues interacting with PRPP in
PRTFDC1 are slightly different than those in HPRT
(Fig. 4), with the most striking difference being a Gly
substitution in the position of the HPRT Asp137 [18].
Mutations of Asp137in human HPRT to Asn resulted
in a 290-fold and 18-fold decrease in k
cat
values for
nucleotide formation with guanine and hypoxanthine
bases, respectively, indicating that Asp137 functions as
a catalytic base [12]. However, a tight binding of N7
of the purine base and Asp137 was observed in HPRT
in complex with ImmGP, indicating a role in transition
state stabilization during catalysis [17].
A series of mutations were made of the invariant
Asp in HPRT from Trypanosoma cruzi. The D137G
mutation in this enzyme, corresponding to the situa-
tion observed in PRTFDC1, was shown to have some
residual activity for the forward reaction [26].
When the HPRT in complex with GMP and ImmGP
are superposed with the PRTFDC1 structure, a water
molecule is found at approximately the same distance
from N7 of the guanine base as is the oxygen of the
aspartic acid side chain in the HPRT complex (Fig. 2B).
It is possible that this water molecule could provide
some catalytic power by acting as a weak base in the
reaction.
Both K
m
and V
max
values for HPRT with hypoxan-
thine determined in the present study are in agreement
with values reported in the literature [45]. However, the
V
max
for HPRT with guanine as substrate using the
DE-81 filter assay was 899 lmolÆmin
)1
Æmg
)1
, which is
20-fold higher than reported previously [45]. This
Table 1. DSLS thermal-melt assay. The results are based on two
independent experiments each containing the samples in triplicate
and given as the mean ± SD.
Ligand DT
agg
(°C) 500 lM
a
DT
agg
(°C) 10 mM
b
PRPP 5.8 ± 0.1 12.7 ± 0.8
IMP 5.1 ± 0.7 7.8 ± 0.1
GMP 4.5 ± 0.4 10.5 ± 0.2
CMP )0.2 ± 0.4 0.2 ± 0.2
UMP )0.2 ± 0.4 0.5 ± 0.6
AMP 0.0 ± 0.4 1.4 ± 0.5
XMP 0.6 ± 0.6 3.5 ± 0.4
Ligand DT
agg
(°C) 500 lM
c
DT
agg
(°C) 50 lM
d
PRPP 0.4 ± 0.3
Hypoxanthine ⁄ PRPP 4.7 ± 1.2
Guanine ⁄ PRPP 2.9 ± 0.4
Cytosine ⁄ PRPP )0.1 ± 0.1
Uracil ⁄ PRPP 0.2 ± 0.1
Adenine ⁄ PRPP )0.2 ± 0.1
Xanthine ⁄ PRPP 0.0 ± 0.1
a
Condition; [ligand]: 500 lM in 20 mM Hepes, 300 mM NaCl, 1 mM
MgCl
2
. DT
agg
calculated using control T
agg
; 53.0 ± 0.7 °C.
b
Condi-
tion; [ligand]: 10 m
M in 20 mM Hepes, 300 mM NaCl, 10 mM MgCl
2
.
DT
agg
calculated using control T
agg
; 52.5 ± 0.6 °C.
c
[PRPP]: 50 lM;
[base]: 500 l
M in 20 mM Hepes, 300 mM NaCl, 10 mM MgCl
2
,
1.25% dimethylsulfoxide. DT
agg
calculated using control T
agg
;
52.7 ± 0.4 °C.
d
[PRPP]: 50 lM in 20 mM Hepes, 300 mM NaCl,
10 m
M MgCl
2
, 1.25% dimethylsulfoxide.
Studies of the human PRTFDC1 M. Welin et al.
4924 FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS
discrepancy might be explained by the different methods
used. We used a radiochemical method in which the
amount of products was quantified directly, and there-
fore it is a more accurate and sensitive method com-
pared to the spectrophotometric methods.
Because the enzymatic activity of PRTFDC1
towards hypoxanthine and guanine is only 0.26% and
0.09%, respectively, of the activity of HPRT, the role
of the PRTFDC1 in purine metabolism remains
unclear and it is uncertain whether it has the capacity
to compensate for a deficiency or partial deficiency in
HPRT. Knowledge of the expression pattern of
PRTFDC1 in healthy individuals and patients with an
impaired HPRT gene has not yet provided clues for
whether PRTFDC1 over-expression in individuals car-
rying mutations in HPRT might lead to a milder dis-
ease phenotype. The concentration of hypoxanthine
available for salvage in human tissues is 8.2 ± 1.3 lm
[28,29] compared to the K
m
for PRTFDC1, which is
23 lm, indicated that PRTFDC1 is not turning over
this substrate to a larger extent. Indeed, these data
indicate that PRTFDC1 does not have a direct cata-
lytic role in the nucleotide salvage pathway. When
PRTFDC1 has been shown to interact with HPRT
A
B
C
D
PRTFDC1
PRTFDC1
HPRT
HPRT
0.4
0.3
V (µmol·min
–1
·mg
–1
)
V (µmol·min
–1
·mg
–1
)
V (µmol·min
–1
·mg
–1
)
V (µmol·min
–1
·mg
–1
)
0.2
0.1
0.0
0 50 100 150
–50
200
400
600
800
1000
0 50 100150 200 250
S
–50 0
2
4
6
8
10
12
50 100150 200 250
S
S/V
S/V
0–10
0.10
0.08
0.06
0.04
0.02
10 20 30 40 50 60
S
S
S/V
60
50
40
30
20
10
–40 –20 0 20 40 60 80 100 120
S/V
200 250
[Hx] (µ
M
)
0 50 100 150 200 250
[Hx] (µ
M
)
[G] (µ
M
)[G] (µ
M
)
0
120100806040200
10 20 30 40 50 60
2.0
1200
25
20
15
10
5
0
1000
800
600
400
200
0
1.5
1.0
0.5
0.0
Fig. 3. Characterization of PRTFDC1 and
HPRT with hypoxanthine and guanine.
Michaelis–Menten and Hanes–Woolf plots
illustrate the kinetic pattern. The experi-
ments were repeated three to four times.
(A) PRTFDC1 and hypoxanthine. (B) HPRT
and hypoxanthine. (C) PRTFDC1 and
guanine. (D) HPRT and guanine.
Table 2. K
m
and V
max
values for Hx and G in the presence of 1 mM PRPP, determined using the DE-81 filter paper assay and tritium-labeled
substrates. Experiments have been repeated three to four times and the data are given as the mean ± SD. k
cat
was calculated using
M
w
(HPRT) = 27132 Da and M
w
(PRTFDC1) = 28226 Da. The k
cat
⁄ K
m
for HPRT was set to 100% as a reference for both substrates, and
k
cat
⁄ K
m
for PRTFDC1, shown in parenthesis, was calculated in relation to this.
K
m
(lM)
V
max
(lmolÆmin
)1
Æmg
)1
) k
cat
(s
)1
) k
cat
⁄ K
m
(s
)1
ÆM
)1
)
Hypoxanthine
PRTFDC1 23.3 ± 6.8 0.340 ± 0.037 0.16 ± 0.02 7.4 · 10
3
± 1.9 · 10
3
(0.26%)
HPRT 3.9 ± 1.5 23.3 ± 2.0 10.5 ± 0.9 2.9 · 10
6
± 1.0 · 10
6
(100%)
Guanine
PRTFDC1 36.1 ± 14.3 2.9 ± 0.7 1.36 ± 0.34 3.9 · 10
4
± 9.3 · 10
3
(0.09%)
HPRT 9.9 ± 0.2 899 ± 117 406 ± 53 4.5 · 10
7
± 1.0 · 10
7
(100%)
M. Welin et al. Studies of the human PRTFDC1
FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS 4925
[29], one possibility is that it participates in forming
heterooligomers containing subunits of both
PRTFDC1 and HPRT, thereby providing an addi-
tional means of regulating the activity of HPRT.
Recently, Suzuki et al. [9] showed that PRTFDC1
expression often was silenced in oral squamous-cell
carcinoma lines. By reintroducing PRTFDC1 expres-
sion in silenced oral squamous-cell carcinoma cells,
growth was reduced. This indicates that PRTFDC1
might have an important regulatory role, although the
molecular basis for this activity remains to be eluci-
dated. The elucidation of the structure of PRTFDC1
and the definition of its ligand binding specificity from
a large metabolome library provides an initial start
point for defining the molecular function of
PRTFDC1. The availability of a high resolution struc-
ture may also assist efforts aiming to develop chemical
probes that could be used to pinpoint the function of
PRTFDC1 using chemical biology strategies.
Experimental procedures
Cloning, expression and purification
The PRTFDC1 and HPRT gene was obtained from
National Institute of Health Mammalian Gene Collection
(accession numbers: BC008662 and BC000578). The
sequences encoding residues 1–225 (PRTFDC1) and 1–218
(HPRT) were amplified by PCR and inserted into pNIC28-
Bsa4 vector by ligation independent cloning. Constructs
included an N-terminal tag containing a 6-His sequence
(MHHHHHHSSGVDLGTENLYFQSM). The pNIC-Bsa4
vector containing the insert was transformed into E. coli
BL21(DE3) strain and stored at )80 °C for further use.
One hundred and fifty milliliters of LB supplemented with
50 lgÆmL
)1
kanamycin was inoculated and grown at 37 °C
overnight; 40 mL of this culture were used to inoculate 3 L
of TB supplemented with 50 lgÆmL
)1
kanamycin and
approximately 50 lL of Breox antifoam (Cognis Perfor-
mance Chemical UK Ltd) in glass flasks using the large
scale expression system. Cells were grown at 37 °C until
D
600
of 1.2 was reached followed by down-tempering to
18 °C for 1 h in water bath. Expression was induced by the
addition of isopropyl thio-b-d-glactoside with a concentra-
tion of 0.5 mm and growth was allowed to continue over-
night. Cells were harvested by centrifugation at 3500 g for
20 min and frozen at )80 °C. Before purification, the cell
pellet was re-suspended in 50 mm Hepes, 500 mm NaCl,
10% glycerol, 10 mm imidazole and 0.5 mm tris(2-carboxy-
ethyl)phosphine (TCEP) supplemented with one tablet per
50 mL of Complete EDTA-free protease inhibitor tablet
(Santa Cruz Biotechnlogy, Santa Cruz, CA, USA) and
4 lL per 50 mL of benzonase. Cells were disrupted by high
Loop II
PRPP motif
Fig. 4. Sequence alignment of PRTFDC1
(accession number: NP_064585), HPRT
(accession number: AAA36012), Plasmo-
dium falciparum hypoxanthine-guanine-xan-
thine phosphoribosyltransferase (PfHGXPRT)
(accession number: NP_700595), TzHPRT
(accession number: XP_816917) and
GlGPRT (accession number: XP_779753).
Loop II and the PRPP motif are shown in
the alignment. Secondary structure for
PRTFDC1 is shown above the sequence
alignment, g-3
10
-helix. Black and white
boxes refer to identical and similar residues,
respectively.
Studies of the human PRTFDC1 M. Welin et al.
4926 FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS
pressure homogenization run three times at 10 000 p.s.i.
and samples were centrifuged for 40 min at 50 228 g. The
soluble fraction was filtered through 0.2 lm filters and sub-
jected to further purification on an A
¨
KTAprime system
(GE Healthcare).
Columns used were HiTrap Chelating 1 mL and
HiLoadÔ 16 ⁄ 60 Superdex 200 Prep Grade (GE Healthcare).
The 1 mL HiTrap chelating HP column was equilibrated
with IMAC buffer 1 [50 mm Hepes (pH 7.5), 10 mm imidaz-
ole, 500 mm NaCl, 10% glycerol, 0.5 mm TCEP], washed
with IMAC buffer 1 followed by IMAC buffer 2 [50 mm
Hepes (pH 7.5), 30 mm imidazole, 500 mm NaCl, 10% glyc-
erol, 0.5 mm TCEP] and eluated with [50 mm Hepes (pH
7.5), 500 mm imidazole, 500 mm NaCl, 10% glycerol,
0.5 mm TCEP]. Additional purification was conducted on a
Superdex 200 column using gel-filtration buffer [20 mm
Hepes (pH 7.5), 300 mm NaCl, 10% glycerol, 2 mm TCEP].
The purity of the protein was estimated on SDS ⁄ PAGE. The
protein concentration was measured using Bradford reagent
[27]. A similar protocol was used for the expression and
purification of HPRT, with the exception that sonication
was used instead of high pressure homogenization, the
purification was conducted on an A
¨
KTAxpress system
(GE Healthcare).
Crystallization, data collection and structure
determination
Crystals were initially obtained from the JCSG screen [28]
[#D10, 100 mm sodium cacodylate (pH 6.5), 200 mm cal-
cium acetate, 40% PEG 300] co-crystallized with 10 mm
PRPP and 10 mm MgCl
2
at room temperature. Crystalliza-
tion experiments were set up using the Phoenix crystalliza-
tion robot (Art Robbins Instrument, Sunnyvale, CA,
USA). In the optimized condition with 100 mm sodium
cacodylate (pH 6.1), 200 mm calcium acetate and 34%
PEG 300, using a protein concentration of 20.5 mgÆmL
)1
,
10 mm PRPP and 10 mm MgCl
2
. 3D crystals grew in a few
days using hanging drop vapor diffusion. A crystal was
transferred into a cryosolution containing 100 m m sodium
cacodylate (pH 6.1), 200 mm calcium acetate and 40%
PEG 300 before being flash frozen in liquid nitrogen.
The human PRTFDC1 crystals belong to space group
P321 and have a solvent content of 56.4%. The asymmetric
unit contained two subunits. Data were collected at ID29
of the European Synchrotron Radiation Facility (Grenoble,
France) using a wavelength of 1.04 A
˚
. The data were inte-
grated using mosflm [29] and scaled using scala from the
ccp4 software suit [30]. The structure was solved with the
molecular replacement software molrep [31] using coordi-
nates from a monomer of HPRT (Protein Data Bank code:
1HMP). After simulated annealing using cns [32], most of
the structure could be manually rebuilt using coot [33].
Refinement was carried out using refmac5 [34]. Data
collection and refinement statistics are shown in Table 3
[35]. Residues 7–225 and 9–225 could be modeled for
subunits A and B, respectively. A covalent modification on
Cys82 was observed likely as the result of a reaction with
cacodylate creating S-(dimethylarsenic)cysteine. Library
files and coordinates for GMP and S -(dimethylarsenic)cys-
teine were obtained from the HIC-UP ligand database [36]
and generated at the prodrg server [37]. All figures were
condtructed using pymol [38] and the alignments were per-
formed using clustalw [39] and ESPript [40]. Superposi-
tions used in the text and figures were made using the ssm
superposition algorithm in coot [33,41]. The coordinates
and structure factors are published in the Protein Data
Bank under accession number 2JBH.
Characterization
Using DSLS thermal denaturation [42–43], PRTFDC1 was
screened against a nucleotide metabolome library consisting
of 81 nucleoside, (deoxy)nucleotide(mono-, di-, tri-phos-
phate), and various other metabolites of both the purine
and pyrimidine pathways (Table S1). The samples were run
twice in duplicate on a StarGazer-384 (Harbinger Biotech-
nology and Engineering Corporation, Markham, Ontario,
Canada). Assay conditions were: 500 lm compound,
0.2 mgÆmL
)1
(7.1 lm) protein, 20 mm Hepes (pH 7.5),
Table 3. Data collection and refinement statistics. Values in paren-
theses refer to the highest resolution shell data.
PRTFDC1-GMP
European Synchrotron Radiation
Facility beam line
ID29
Wavelength (A
˚
) 1.04
Space group P321
Unit cell parameters (A
˚
) a = b = 139.2
c = 52.1
Content of the asymmetric unit Two subunits
Resolution (A
˚
) 1.7 (1.70–1.79)
Completeness (%) 100 (100)
R
meas
(%) 10.0 (48.2)
<I ⁄ r(I)> 19.4 (4.2)
Redundancy 10.8 (11.0)
Number of unique reflections 63939
Refinement
R
work
(%)
a
18.2
R
free
(%)
b
21.5
Rmsd
Bond length (A
˚
) 0.012
Bond angle (
o
) 1.4
Mean B-value (A
˚
2
) 20.2
Ramachandran plot (%)
c
Favored regions 98.04
Additionally allowed regions 1.96
a
R
work
is defined as R||F
obs
| ) |F
calc
|| ⁄ R|F
obs
|, where F
obs
and F
calc
are observed and calculated structure-factor amplitudes, respec-
tively.
b
R
free
is the R factor for the test set (5% of the data).
c
According to MOLPROBITY [35].
M. Welin et al. Studies of the human PRTFDC1
FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS 4927
300 mm NaCl, and 1 mm MgCl
2
. Following initial screen-
ing, a subset consisting of IMP, GMP, XMP, AMP, CMP
and PRPP were screened at 500 lm and 10 mm, with 1 mm
and 10 mm MgCl
2
respectively [20 mm Hepes (pH 7.5),
300 mm NaCl]. A second subset was screened against
50 lm PRPP in combination with 500 lm nucleobase
(hypoxanthine, guanine, adenine, xanthine, cytosine, uracil)
in 20 mm Hepes, 300 m m NaCl, 1 mm MgCl
2
and 1.25%
dimethylsulfoxide. All StarGazer-384 screening was con-
ducted using 384 well optical clear-bottom plates (#242764;
Nunc, Rochester, NY, USA), with an assay volume of
50 lL per well. The plate was heated at 1 °CÆmin
)1
, and
images take every 0.5 °C in the range 25–80 °C. Intensities
(as a measure of light scattering from protein aggregation)
were converted from the images, and the intensities were
plotted as a function of temperature and the midpoint of
transition in aggregation (T
agg
) calculated [42,43] using the
manufactuter’s software (Harbinger Biotech). The reported
DT
agg
represents the calculated difference between T
agg
in
the presence of compound to be tested and a control condi-
tion without compound.
Enzyme activities were determined by initial velocity
measurements based on four time samples using the DE-81
(DEAE-cellulose) filter paper assay adopted from the de-
oxynucleoside kinase assay method [44] with tritium-labeled
bases as substrates. The standard assay mixture contained
in a reaction volume of 50 lL: 100 mm Tris-HCl (pH 8.0),
10 mm MgCl
2
, 0.5 mgÆmL
)1
BSA, 1 mm PRPP, 1–200 lm
[
3
H]hypoxanthine (13.8 CiÆmmol
)1
; Perkin Elmer, Boston,
MA, USA), 1–100 lm [
3
H]guanine (10.7 CiÆmmol
)1
; Mor-
avek, Brea, CA, USA) and [
3
H]adenine (27.2 CiÆmmol
)1
;
Perkin Elmer). The assay mix was preheated at 37 °C for
5 min, and the reactions were started by adding 10 lLof
enzyme (HPRT: 0.07 or 0.7 ng per assay; PRTFDC1: 39 ng
per assay), and a 10 lL aliquot was withdrawn and spotted
onto DE-81 filter paper at 0, 4, 8 and 12 min. The nega-
tively-charged products will bind to the positively charged
DE-81 filter papers. Nonreacted substrates were removed
by washing the filter papers 3 · 5 min in 5 m m ammonium
formate solution and once in deionized water. The tritium-
labeled products on DE-81 filter paper were eluted for
30 min in 0.5 mL of 0.2 m KCl ⁄ 0.1 m HCl; subsequently,
2 mL of scintillation liquid (Optiphase ‘hisafe’ 3; Perkin
Elmer) was added to each vial, and radioactivity was
counted in a liquid scintillation counter (Beckman Coulter,
Fullerton, CA, USA). The kinetic data were fitted into the
Michaelis–Menten equation v = V
max
Æ[S] ⁄ (K
m
+[S]) using
the sigmaplot enzyme kinetic model, version 2.1 (SPSS
Inc., Chicago, IL, USA).
Acknowledgements
This work was supported by grants from the Swedish
Research Council for Environment, Agricultural
Sciences and Spatial Planning (to L.W. and S.E.), the
Swedish Research Council (to H.E. and P.N.) and
the Swedish Cancer Foundation (to H.E. and P.N.).
The Structural Genomics Consortium is a registered
charity (number 1097737) that receives funds from the
Canadian Institutes for Health Research, the Canadian
Foundation for Innovation, Genome Canada through
the Ontario Genomics Institute, GlaxoSmithKline,
Karolinska Institutet, the Knut and Alice Wallenberg
Foundation, the Ontario Innovation Trust, the Ontario
Ministry for Research and Innovation, Merck & Co.,
Inc., the Novartis Research Foundation, the Swedish
Agency for Innovation Systems, the Swedish Founda-
tion for Strategic Research and the Wellcome Trust.
References
1 Jinnah HA, De Gregorio L, Harris JC, Nyhan WL &
O’Neill JP (2000) The spectrum of inherited mutations
causing HPRT deficiency: 75 new cases and a review of
196 previously reported cases. Mutat Res 463, 309–326.
2 Jinnah HA, Harris JC, Nyhan WL & O’Neill JP (2004)
The spectrum of mutations causing HPRT deficiency:
an update. Nucleosides Nucleotides Nucleic Acids 23,
1153–1160.
3 Sculley DG, Dawson PA, Emmerson BT & Gordon RB
(1992) A review of the molecular basis of hypoxanthine-
guanine phosphoribosyltransferase (HPRT) deficiency.
Hum Genet 90, 195–207.
4 Lesch M & Nyhan WL (1964) A familial disorder of
uric acid metabolism and central nervous system func-
tion. Am J Med 36, 561–570.
5 Song S & Friedmann T (2007) Tissue-specific aberra-
tions of gene expression in HPRT-deficient mice: func-
tional complexity in a monogenic disease? Mol Ther 15,
1432–1443.
6 Jinnah HA (2009) Lesch-Nyhan disease: from mechanism
to model and back again. Dis Model Mech 2, 116–121.
7 Wang CC (1984) Parasite enzymes as potential targets
for antiparasitic chemotherapy. J Med Chem 27, 1–9.
8 Cai LY, Abe M, Izumi S, Imura M, Yasugi T & Ushijima
T (2007) Identification of PRTFDC1 silencing and
aberrant promoter methylation of GPR150, ITGA8 and
HOXD11 in ovarian cancers. Life Sci 80, 1458–1465.
9 Suzuki E, Imoto I, Pimkhaokham A, Nakagawa T,
Kamata N, Kozaki KI, Amagasa T & Inazawa J (2007)
PRTFDC1, a possible tumor-suppressor gene, is
frequently silenced in oral squamous-cell carcinomas by
aberrant promoter hypermethylation. Oncogene 26,
7921–7932.
10 Keebaugh AC, Sullivan RT & Thomas JW (2007) Gene
duplication and inactivation in the HPRT gene family.
Genomics 89 , 134–142.
11 Nicklas JA (2006) Pseudogenes of the human HPRT1
gene. Environ Mol Mutagen 47, 212–218.
Studies of the human PRTFDC1 M. Welin et al.
4928 FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS
12 Xu Y & Grubmeyer C (1998) Catalysis in human hypo-
xanthine-guanine phosphoribosyltransferase: Asp 137
acts as a general acid ⁄ base. Biochemistry 37, 4114–4124.
13 Balendiran GK, Molina JA, Xu Y, Torres-Martinez J,
Stevens R, Focia PJ, Eakin AE, Sacchettini JC & Craig
SP III (1999) Ternary complex structure of human
HGPRTase, PRPP, Mg
2+
, and the inhibitor HPP
reveals the involvement of the flexible loop in substrate
binding. Protein Sci 8, 1023–1031.
14 Eads JC, Scapin G, Xu Y, Grubmeyer C &
Sacchettini JC (1994) The crystal structure of human
hypoxanthine-guanine phosphoribosyltransferase with
bound GMP. Cell 78, 325–334.
15 Focia PJ, Craig SP III, Nieves-Alicea R, Fletterick RJ
& Eakin AE (1998) A 1.4 A crystal structure for the
hypoxanthine phosphoribosyltransferase of Trypanoso-
ma cruzi. Biochemistry 37, 15066–15075.
16 Shi W, Li CM, Tyler PC, Furneaux RH, Cahill SM,
Girvin ME, Grubmeyer C, Schramm VL & Almo SC
(1999) The 2.0 A structure of malarial purine phospho-
ribosyltransferase in complex with a transition-state
analogue inhibitor. Biochemistry 38, 9872–9880.
17 Shi W, Li CM, Tyler PC, Furneaux RH, Grubmeyer C,
Schramm VL & Almo SC (1999) The 2.0 A structure of
human hypoxanthine-guanine phosphoribosyltransferase
in complex with a transition-state analog inhibitor. Nat
Struct Biol 6, 588–593.
18 Scapin G, Grubmeyer C & Sacchettini JC (1994) Crys-
tal structure of orotate phosphoribosyltransferase. Bio-
chemistry 33, 1287–1294.
19 Keough DT, Brereton IM, de Jersey J & Guddat LW
(2005) The crystal structure of free human hypoxan-
thine-guanine phosphoribosyltransferase reveals exten-
sive conformational plasticity throughout the catalytic
cycle. J Mol Biol 351, 170–181.
20 Johnson GG, Eisenberg LR & Migeon BR (1979)
Human and mouse hypoxanthine-guanine phosphoribo-
syltransferase: dimers and tetramers. Science 203, 174–
176.
21 Gayathri P, Sujay Subbayya IN, Ashok CS, Selvi TS,
Balaram H & Murthy MR (2008) Crystal structure of
a chimera of human and Plasmodium falciparum
hypoxanthine guanine phosphoribosyltransferases
provides insights into oligomerization. Proteins 73,
1010–1020.
22 Strauss M, Behlke J & Goerl M (1978) Evidence against
the existence of real isozymes of hypoxanthine phospho-
ribosyltransferase. Eur J Biochem 90, 89–97.
23 Arent S, Kadziola A, Larsen S, Neuhard J & Jensen
KF (2006) The extraordinary specificity of xanthine
phosphoribosyltransferase from Bacillus subtilis
elucidated by reaction kinetics, ligand binding, and
crystallography. Biochemistry 45, 6615–6627.
24 Raman J, Sumathy K, Anand RP & Balaram H (2004)
A non-active site mutation in human hypoxanthine
guanine phosphoribosyltransferase expands substrate
specificity. Arch Biochem Biophys 427, 116–122.
25 Xu Y, Eads J, Sacchettini JC & Grubmeyer C (1997)
Kinetic mechanism of human hypoxanthine-guanine
phosphoribosyltransferase: rapid phosphoribosyl trans-
fer chemistry. Biochemistry 36, 3700–3712.
26 Canyuk B, Focia PJ & Eakin AE (2001) The role for
an invariant aspartic acid in hypoxanthine phospho-
ribosyltransferases is examined using saturation muta-
genesis, functional analysis, and X-ray crystallography.
Biochemistry 40, 2754–2765.
27 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein
utilizing the principle of protein-dye binding. Anal
Biochem 72, 248–254.
28 Page R & Stevens RC (2004) Crystallization data min-
ing in structural genomics: using positive and negative
results to optimize protein crystallization screens.
Methods 34, 373–389.
29 Leslie AGW (1992) Joint CCP4 + ESF-EAMCB News-
letter on Protein Crystallography, No. 26.
30 CCP4 (1994) The CCP4 suite: programs for protein
crystallography. Acta Crystallogr D Biol Crystallogr 50,
760–763.
31 Vagin A & Teplyakov A (2000) An approach to multi-
copy search in molecular replacement. Acta Crystallogr
D Biol Crystallogr 56, 1622–1624.
32 Brunger AT, Adams PD, Clore GM, DeLano WL,
Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J,
Nilges M, Pannu NS et al. (1998) Crystallography &
NMR system: a new software suite for macromolecular
structure determination. Acta Crystallogr D Biol Crys-
tallogr 54, 905–921.
33 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr D Biol
Crystallogr 60, 2126–2132.
34 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the maxi-
mum-likelihood method. Acta Crystallogr D Biol Crys-
tallogr 53, 240–255.
35 Lovell SC, Davis IW, Arendall WB III, de Bakker PI,
Word JM, Prisant MG, Richardson JS & Richardson
DC (2003) Structure validation by Calpha geometry:
phi,psi and Cbeta deviation. Proteins 50, 437–450.
36 Kleywegt GJ & Jones TA (1998) Databases in protein
crystallography. Acta Crystallogr D Biol Crystallogr 54,
1119–1131.
37 Schuttelkopf AW & van Aalten DM (2004) PRODRG:
a tool for high-throughput crystallography of protein-
ligand complexes. Acta Crystallogr D Biol Crystallogr
60, 1355–1363.
38 DeLano WL (2002) The PyMOL Molecular Graphics
System. DeLano Scientific, Palo Alto.
39 Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ,
Higgins DG & Thompson JD (2003) Multiple sequence
M. Welin et al. Studies of the human PRTFDC1
FEBS Journal 277 (2010) 4920–4930 ª 2010 The Authors Journal compilation ª 2010 FEBS 4929
