Reversible tetramerization of human TK1 to the high
catalytic efficient form is induced by pyrophosphate,
in addition to tripolyphosphates, or high enzyme
concentration
Birgitte Munch-Petersen
Department of Science, Systems and Models, Roskilde University, Denmark
For decades, it has been the general belief that the
building blocks of DNA, the deoxyribonucleoside
triphosphates (dNTPs), play a central role in maintai-
ning correct DNA synthesis. Recent investigations of
DNA synthetic processes in yeast and human cells
have indicated that initiation and progress of DNA
replication are closely associated with the cellular
dNTP concentration [1–3].
The level of the dNTPs is strictly controlled and
fluctuates during the cell cycle, in close correlation
with the rate of DNA synthesis, with low dNTP levels
in G
1
cells increasing during S phase, generally with
dTTP being the most abundant and dGTP the least
[4–6]. In quiescent cells, dNTP levels are several-fold
lower [7], and in non-proliferating human lymphocytes,
which are G
0
cells, the dTTP pool is many times
smaller than the other dNTP pools [4].
In most cells and organisms except for a few para-
sites, the dNTPs are provided by two main routes, the
de novo and the salvage pathways. The central enzyme
in the de novo route, ribonucleotide reductase, cata-

,
increases in S phase coinciding with the increase in DNA synthesis, and
disappears during mitosis. The fluctuation of TK1 through the cell cycle is
important in providing a balanced supply of dTTP for DNA replication,
and is partly due to regulation of TK1 expression at the transcriptional
level. However, TK1 is a regulatory enzyme that can interchange between
its dimeric and tetrameric forms, which have low and high catalytic effi-
ciencies, respectively, depending on pre-assay incubation with ATP. Here,
the part of ATP that is necessary for tetramerization and how the reaction
velocity is influenced by the enzyme concentration are determined. The
results show that only two or three of the phosphate groups of ATP
are necessary for tetramerization, and that kinetics and tetramerization are
closely related. Furthermore, the enzyme concentration was found to have
a pivotal effect on catalytic efficiency.
Abbreviations
dNTP, deoxyribonucleoside triphosphate; dThd, thymidine; hTK1, human cytosolic thymidine kinase 1; NaP, sodium orthophosphate; NaPP,
sodium dipolyphosphate; NaPPP, sodium tripolyphosphate; rhTK1, recombinant human TK1; TmTK, TK from Thermotoga maritima.
FEBS Journal 276 (2009) 571–580 ª 2008 The Author Journal compilation ª 2008 FEBS 571
level is critical for maintaining a proper balance
between the dNTPs. In addition to the ribonucleotide
reductase-controlled pathway, the dTTP level is con-
trolled by thymidine kinases and TMP nucleotidases,
forming a substrate cycle [8,9].
A crucial step in dTTP synthesis is phosphorylation
of thymidine (dThd) to dTMP. Two thymidine kinases
catalyze this step, the cytosolic TK1 and the mitochon-
drial TK2 (EC 2.7.1.21 for both TK1 and TK2),
encoded by two nuclear genes. TK1 is cell-cycle-specific
and is not expressed in quiescent cells, in which only the
constitutively expressed TK2 is present. The complex

This behavior means that the catalytic efficiency
(k
cat
⁄ K
m
) is approximately 30-fold higher for hTK1 that
had been incubated with ATP. This ‘ATP effect’ on the
kinetics apparently depends on the enzyme concentra-
tion in a linear manner, and no transition to the catalyti-
cally highly active form was observed at concentrations
of hTK1 below 10 ngÆmL
)1
(0.4 nm) [12]. Therefore,
transition does not occur at the low assay concentration
of TK1 (< 3 ngÆmL
)1
). This also explains why both
enzyme forms showed linear progress curves for product
versus time.
It is very likely that the ‘ATP effect’ is a fine tuning
of the hTK1 activity during the cell cycle. When hTK1
is degraded in G
2
⁄ M phase, and given that ATP is
fairly constant during the cell cycle, the initial low
hTK1 concentration in the following G
1
phase implies
predominance of the low-activity dimer form. As the
hTK1 concentration increases during S phase, more

dipolyphosphate group is sufficient for inducing transi-
tion to the high-active tetramer, and that kinetics and
oligomerization are closely related. In addition, the
results show a clear relationship between the enzyme
concentration and the catalytically high-active tetra-
meric form, and that the tetramer dissociates into
dimers very slowly.
Results and Discussion
Identification of the group inducing
tetramerization of human TK1
Human TK1 has 234 amino acids and a subunit size
of 25.5 kDa [18]. Several reports have shown by gel
filtration that native as well as recombinant hTK1
elutes as a dimer in the absence of ATP (1–5 mm) and
as a tetramer in its presence [12,13,19,20]. The recently
solved structures of a number of TK1-like enzymes
from human, bacteria and vaccinia virus all show tet-
rameric forms [14,17,21–23]. As the adenosine moiety
does not show electron density in any of the human
TK1 structures, it may be that the adenosine moiety is
of no significance for inducing the reversible dimer–
tetramer transition. Therefore, the present study aimed
to identify the part of the nucleotide molecule
that triggers tetramerization. Figure 1A–C shows the
elution profiles of native TK1 from human lympho-
Enzymatic regulation of human TK1 B. Munch-Petersen
572 FEBS Journal 276 (2009) 571–580 ª 2008 The Author Journal compilation ª 2008 FEBS
cytes in the presence of the ribonucleoside triphos-
phates GTP, CTP and UTP. For comparison, elution
profiles with and without ATP are shown in

from human lymphocytes in a total volume
of 200 lL was injected into a Superdex 200
column (10 · 300 mm) together with 0.1
mg Blue Dextran used as an internal
standard for determination of the void
volume, V
0
, in the individual experiments.
Prior to injection, hTK1 was diluted to 6
lgÆmL
)1
and incubated with 3 mM of the
indicated nucleotides or polyphosphates at
4 °C for 2 h, and stored for at least 2 weeks
at )80 °C. Fractions (200 lL) were collected
into 100 lL column buffer containing 30%
glycerol and 2 m
M ATP. The fractions were
assayed for thymidine kinase activity under
standard assay conditions with 100 l
M
dThd. The molecular markers (vertical bars)
are (from left to right): b-amylase (200 kDa),
BSA (66 kDa), ovalbumin (45 kDa), carbonic
anhydrase (30 kDa) and cytochrome
c (12.4 kDa). V
e
is the elution volume. The
standard variation for V
e

ativity and a K
0.5
(substrate concentration at half-max-
imal velocity) of approximately 15 lm [12]. However,
when hTK1 was incubated with ATP prior to the
assay, it showed hyperbolic kinetics with a K
m
of
approximately 0.5 lm. Both enzyme forms have the
same V
max
, meaning that the catalytic efficiency of
ATP-incubated hTK1 is approximately 30-fold higher
than that of non-incubated hTK1. The two TK1 forms
can therefore be referred to as the high- and low-
efficiency forms. To explain the apparent negative
co-operativity, a model has been proposed whereby the
dimer has high K
m
and the tetramer has low K
m
, and
the ratio between the two forms depends on the dThd
concentration [24]. According to this model, the simul-
taneous presence in the assay of the two forms will
result in the apparent negative co-operative behavior.
To further elucidate this, the relationship between the
oligomerization status and the kinetic behaviour was
investigated, i.e. whether the tetrameric and dimeric
forms in Figs 1 and 2 exhibited low or high catalytic

Phosphate donor specificity
The results from Figs 1–4 showed that inorganic
di- and tripolyphosphates were able to induce tetra-
merization and hyperbolic kinetics with low K
m
values
similar to the nucleoside di- and tri-phosphates, and
Table 1. Native molecular size and kinetic parameters.
Incubation conditions
for hTK1 Mass (kDa) K
0.5
(lM) n (Hill constant)
Phosphate donor
capacity
b
(%)
None 57.5 ± 2.7
a
(5) 16.4 ± 1.0 (10) 0.75 ± 0.04 (10) –
ATP 115 ± 4.5 (5) 0.51 ± 0.03 (10) 1.04 ± 0.04 (10) 100
GTP 101 0.68 ± 0.015 (3) 0.97 ± 0.04 (3) 37 ± 1
CTP 95.5 0.79 ± 0.029 (3) 0.98 ± 0.01 (3) 18 ± 3
UTP 100 0.64 ± 0.072 (3) 1.02 ± 0.04 (3) 19 ± 0.1
ADP 118 0.66 ± 0.11 (3) 0.99 ± 0.01 (3) 4 ± 0.8
AMP 52.7 21.3 ± 4.33 (3) 0.77 ± 0.01 (3) 0
NaPPP 93 0.95 ± 0.04 (2) 1.31 ± 0.08 (2) 0
NaPPP-MgCl
2
98 0.90 ± 0.07 (2) 0.97 ± 0.08 (2) 0
c

oligomerization of hTK1
The above-described experiments were all performed
with the native enzyme purified from human lympho-
cytes to a final concentration of approximately
5 lgÆmL
)1
[25], and the concentration of the applied
enzyme in the gel filtration experiments in Figs 1 and
2 was 50 ngÆmL
)1
(10 ng applied). Using recombinant
techniques, concentrations of pure hTK1 more than
1000–10 000-fold higher can be obtained, enabling
considerably higher concentrations during gel filtra-
tion. This may explain the appearance of both dimer
and tetramer peaks during gel filtration of non-incu-
bated recombinant human TK1 (rhTK1), although
the tetramer peak is the smallest [19,20]. In these
studies, TK1 was applied at a concentration of
approximately 3 lgÆmL
)1
. Recently, it was reported
that human TK1 elutes exclusively as a tetramer
when applied at a concentration range of 0.4–
20 mgÆmL
)1
[26]. The authors suggest that the high-
level expression of TK1 obtained in their work may
influence the oligomerization pattern of the enzyme.
However, the more than 100-fold higher concentra-

max
was
obtained before and after the treatment. As seen
from Fig. 5C, only a minor part of the enzyme is in
the tetramer form. This elution profile is very similar
to those previously reported by Berenstein et al. and
Frederiksen et al. [19,20]. In their gel-filtration
Fig. 2. Effect of orthophosphate and di- and tri-polyphosphates on
oligomerization of native hTK1. hTK1 was diluted and incubated
with 3 m
M of the indicated nucleotides or phosphate compound
without MgCl
2,
injected onto the Superdex 200 column, eluted with
column buffer without MgCl
2
containing 2 mM of the respective
nucleotide or phosphate compound, and assayed as described for
Fig. 1.
B. Munch-Petersen Enzymatic regulation of human TK1
FEBS Journal 276 (2009) 571–580 ª 2008 The Author Journal compilation ª 2008 FEBS 575
experiments on recombinant human TK1, Li et al.
[13] diluted and treated the enzyme as in Fig. 5C and
also found a similar elution profile. Together, these
observations show that rhTK1 behaves as a tetramer
even in the absence of phosphate groups when
applied at concentrations of 200 lgÆmL
)1
or higher,
Fig. 3. Effect of nucleotides on hTK1 dThd

in behaviour after dilution to 6 lgÆmL
)1
indicate that
dissociation of the tetramer to the dimer is a slowly
progressing process. This is supported by the kinetic
behaviour of the recombinant enzyme as shown in
Fig. 5D, where the enzyme was diluted from
0.5 mgÆmL
)1
immediately before the kinase assay.
Under these conditions, the kinetic behaviour was
essentially like that of the tetramer form, exhibiting
hyperbolic kinetics with a K
m
of 0.7 lm. This also
indicates slow dissociation of the tetramer form, and
may explain why linear progress curves are always
obtained with all forms of the enzyme and under all
incubation conditions. When rhTLK1 is diluted from
high storage concentrations to low assay concentra-
tions of 2–3 ngÆmL
)1
, which is below the limit for
the ATP tetramerization effect, the enzyme would be
expected to dissociate to the dimer form with higher
K
m
during the assay, and this would result in non-
linear progress curves. However, slow dissociation
from tetramer to dimer will result in linear progress

incubated for 2 h at 4 °C, and stored at )80 °C for more than
2 weeks. (D) dThd substrate kinetics with recombinant human TK1
(0.1 ng in 50 lL assay reaction volume) diluted from 0.6 mgÆmL
)1
to 0.01 lgÆmL
)1
immediately before assay. Inset, Hofstee plot of
the data.
B. Munch-Petersen Enzymatic regulation of human TK1
FEBS Journal 276 (2009) 571–580 ª 2008 The Author Journal compilation ª 2008 FEBS 577
origin [21,22]. In a recent study, the bi-substrate inhi-
bitor P1-(5¢-adenosyl)P4-[5¢-(2¢-deoxy-thymidyl)] tetra-
phosphate (AP4dT) was crystallized together with
hTK1 and TmTK [17,23]. In both structures, thy-
mine and the three phosphates were bound in the
lasso motif, essentially as for dTTP in the previous
structures. The authors conclude that the fourth
phosphate, which is analogous to the a-phosphate in
ATP, is observed in both structures, whereas electron
density is obtained only with the adenine group in
the TmTK structure [23]. Moreover, with the ana-
logue bound, the TmTK structure appears more
open than the hTK1 structure. This indicates that
the adenine group in the hTK1 structure makes only
a few, if any, contacts with the enzyme. It may also
explain at least partly why the kinetic and oligomeric
effects can be exhibited by only two phosphate
groups, which probably are analogous to the a and
b phosphate groups in the nucleotide ADP. On the
other hand, the large difference in phosphate donor

values is displayed by
the dimer forms. These observations strengthen the
previous hypothesis that the dimer ⁄ tetramer inter-
change of TK1 with low ⁄ high catalytic efficiency is a
fine-tuning mechanism that may serve to provide a bal-
anced supply of dTTP throughout the cell cycle,
adjusted to the need for DNA synthesis [12,13,24]. As
dTTP is a key regulator of ribonucleotide reductase,
higher dTTP concentrations will result in unbalanced
dNTP pools, which are known to be mutagenic [27–
29]. In the light of these effects, the complex regulatory
and structural properties of hTK1 may be important
for maintaining a balanced supply of the DNA precur-
sor. This underlines the importance of elucidating the
molecular and structural background of the enzymatic
and catalytic properties of human thymidine kinase.
Experimental procedures
Superdex 12, Glutathione–Sepharose, pGEX-2T vector,
thrombin, [methyl-
3
H]dThd (25 CiÆmmol
)1
) and the Esc-
herichia coli strains XL Gold and BL21 were purchased
from Amersham Biosciences (now part of GE Healthcare
Bio-Sciences, Hillerod, Denmark). Strains XL Gold and
BL21 were used to propagate and express, respectively,
the recombinant thymidine kinase. Chaps was purchased
from Roche A/S (Copenhagen, Denmark). Triton X-100,
dithiotreitol, non-radioactive nucleosides and molecular

in Superdex column buffer (50 mm
imidazole ⁄ HCl pH 7.5, 5 mm MgCl
2
, 0.1 m KCl, 2 mm
Chaps and 5 mm dithiothreitol), incubated with or without
3mm of the respective nucleotide or phosphate compound
for 2 h at 4 °C, and stored for at least 2 weeks at )80 °C
before use for kinetic and molecular mass analyses. The
activity at saturating conditions was similar before and
after dilution, incubation and storage.
Native molecular size
The apparent molecular size was determined by gel filtra-
tion on a Superdex 12 (10 · 300 mm) column connected to
a Gradifrac automatic sampler (Amersham Biosciences) as
Enzymatic regulation of human TK1 B. Munch-Petersen
578 FEBS Journal 276 (2009) 571–580 ª 2008 The Author Journal compilation ª 2008 FEBS
described previously [19]. The column was pre-equilibrated
in column buffer (50 mm imidazole ⁄ HCl pH 7.5, 5 mm
MgCl
2
, 0.1 m KCl, 2 mm Chaps and 5 mm dithiothreitol)
containing two milimolar of the respective nucleotide or
phosphate compounds. In each experiment, 0.2 mL enzyme
dilution containing 0.1 mg Blue Dextran 2000 (Sigma-
Aldrich) was applied. Blue dextran was used as an internal
standard for determination of the void volume V
0
of the
column. This value was used for calculation of V
e

3
H]dThd in a final volume of 50 lL. The reaction
was started by adding approximately 0.1 ng enzyme
diluted from 6 lgÆmL
)1
in ice-cold enzyme dilution buffer
(50 mm Tris ⁄ HCl pH 7.5, 1 mm Chaps, 3 mgÆmL
)1
BSA)
immediately before the start of the reaction. During the
first 15 min of the reaction, four samples of 10 lL each,
taken at various time points 3, 6, 9 and 12 min after the
start of the reaction, were applied to the DE-81 filters.
The filters were washed three times for 5 min each in
5mm ammonium formate and once for 5 min in water,
and the nucleotides were eluted from the DE-81 filters
by shaking for 30 min in 0.2 m KCl ⁄ 0.1 m HCl, after
which the radioactivity was determined by scintillation
counting.
Analysis of kinetic data
Kinetic data were fitted by non-linear regression analysis to
the Michaelis–Menten equation v ¼ V
max
½S=ðK
m
þ½SÞ
or the Hill equation v ¼ V
max
½S
n

in early S-phase. J Cell Sci 114, 747–750.
2 Chabes A & Stillman B (2007) Constitutively high
dNTP concentration inhibits cell cycle progression and
the DNA damage checkpoint in yeast Saccharomyces
cerevisiae. Proc Natl Acad Sci USA 104, 1183–1188.
3 Koc A, Wheeler LJ, Mathews CK & Merrill GF (2004)
Hydroxyurea arrests DNA replication by a mechanism
that preserves basal dNTP pools. J Biol Chem 279,
223–230.
4 Munch-Petersen B, Tyrsted G & Dupont B (1973) The
deoxyribonucleoside 5¢-triphosphate (dATP and dTTP)
pools in phytohemagglutinin-stimulated and non-stimu-
lated human lymphocytes. Exp Cell Res 79, 249–256.
5 Tyrsted G (1975) The pool size of deoxyguanosine
5¢-triphosphate and deoxycytidine 5¢-triphosphate in
phytohemagglutinin-stimulated and non-stimulated
human lymphocytes. Exp Cell Res 91, 429–440.
6 Chabes A & Thelander L (2003) DNA building blocks
at the foundation of better survival. Cell Cycle 2, 171–
173.
7 Snyder RD (1988) Consequences of depletion of cellular
deoxyribonucleoside triphosphate pools on the exci-
sion–repair process in cultured human fibroblasts.
Mutat Res 200, 193–199.
8 Reichard P (1988) Interactions between deoxyribonucle-
otide and DNA synthesis. Annu Rev Biochem 57, 349–
374.
9 Pontarin G, Ferraro P, Valentino ML, Hirano M, Rei-
chard P & Bianchi V (2006) Mitochondrial DNA deple-
tion and thymidine phosphate pool dynamics in a

Biol 10, 513–519.
16 Johansson K, Ramaswamy S, Ljungcrantz C, Knecht
W, Piskur J, Munch-Petersen B, Eriksson S & Eklund
H (2001) Structural basis for substrate specificities of
cellular deoxyribonucleoside kinases. Nat Struct Biol 8,
616–620.
17 Segura-Pena D, Lichter J, Trani M, Konrad M, Lavie
A & Lutz S (2007) Quaternary structure change as a
mechanism for the regulation of thymidine kinase 1-like
enzymes. Structure 15, 1555–1566.
18 Bradshaw HD Jr & Deininger PL (1984) Human thymi-
dine kinase gene: molecular cloning and nucleotide
sequence of a cDNA expressible in mammalian cells.
Mol Cell Biol 4, 2316–2320.
19 Berenstein D, Christensen JF, Kristensen T, Hofbauer
R & Munch-Petersen B (2000) Valine, not methionine,
is amino acid 106 in human cytosolic thymidine kinase
(TK1). Impact on oligomerization, stability and kinetic
properties. J Biol Chem 275, 32187–32192.
20 Frederiksen H, Berenstein D & Munch-Petersen B
(2004) Effect of valine 106 on structure–function rela-
tion of cytosolic human thymidine kinase. Kinetic prop-
erties and oligomerization pattern of nine substitution
mutants of V106. Eur J Biochem 271, 2248–2256.
21 Birringer MS, Claus MT, Folkers G, Kloer DP, Schulz
GE & Scapozza L (2005) Structure of a type II thymi-
dine kinase with bound dTTP. FEBS Lett 579, 1376–
1382.
22 Kosinska U, Carnrot C, Eriksson S, Wang L & Eklund
H (2005) Structure of the substrate complex of thymi-

of dTTP pool size by anaphase promoting complex
⁄
cyclosome is essential for the maintenance of genetic
stability. Genes Dev 19, 1920–1933.
Enzymatic regulation of human TK1 B. Munch-Petersen
580 FEBS Journal 276 (2009) 571–580 ª 2008 The Author Journal compilation ª 2008 FEBS