Two thymidine kinases and one multisubstrate deoxyribonucleoside kinase salvage DNA precursors in Arabidopsis thaliana

Article abstract of DOI:10.1111/j.1742-4658.2012.08747.x

Deoxyribonucleotides are the building blocks of DNA and can be synthesized via de novo and salvage pathways. Deoxyribonucleoside kinases (EC 2.7.1.145) salvage deoxyribonucleosides by transfer of a phosphate group to the 5' of a deoxyribonucleoside. This salvage pathway is well characterized in mammals, but in contrast, little is known about how plants salvage deoxyribonucleosides. We show that during salvage, deoxyribonucleosides can be phosphorylated by extracts of Arabidopsis thaliana into corresponding monophosphate compounds with an unexpected preference for purines over pyrimidines. Deoxyribonucleoside kinase activities were present in all tissues during all growth stages. In the A. thaliana genome, we identified two types of genes that could encode enzymes which are involved in the salvage of deoxyribonucleosides. Thymidine kinase activity was encoded by two thymidine kinase 1 (EC 2.7.1.21)-like genes (AtTK1a and AtTK1b). Deoxyadenosine, deoxyguanosine and deoxycytidine kinase activities were encoded by a single AtdNK gene. T-DNA insertion lines of AtTK1a and AtTK1b mutant genes had normal growth, although AtTK1a AtTK1b double mutants died at an early stage, which indicates that AtTK1a and AtTK1b catalyze redundant reactions. The results obtained in the present study suggest a crucial role for the salvage of thymidine during early plant development. 2012 The Authors Journal compilation

Full text of DOI:10.1111/j.1742-4658.2012.08747.x

Two thymidine kinases and one multisubstrate
deoxyribonucleoside kinase salvage DNA precursors in
Arabidopsis thaliana
Anders R. Clausen1, Lenart Girandon1, Ashfaq Ali1, Wolfgang Knecht2, Elzbieta Rozpedowska1,
1
3
1,2
Michael P. B. Sandrini1,2, Erik Andreasson , Birgitte Munch-Petersen and Jure Piskur
ˇ
1 Department of Cell and Organism Biology, Lund University, Sweden
2 BioCentrum-DTU, Technical University of Denmark, Lyngby, Denmark
3 Department of Science, Systems and Models, Roskilde University, Denmark
Keywords
Deoxyribonucleotides are the building blocks of DNA and can be synthe-
sized via de novo and salvage pathways. Deoxyribonucleoside kinases
(EC 2.7.1.145) salvage deoxyribonucleosides by transfer of a phosphate
group to the 5′ of a deoxyribonucleoside. This salvage pathway is well
characterized in mammals, but in contrast, little is known about how
plants salvage deoxyribonucleosides. We show that during salvage, deoxyri-
bonucleosides can be phosphorylated by extracts of Arabidopsis thaliana
into corresponding monophosphate compounds with an unexpected prefer-
ence for purines over pyrimidines. Deoxyribonucleoside kinase activities
were present in all tissues during all growth stages. In the A. thaliana gen-
ome, we identified two types of genes that could encode enzymes which are
involved in the salvage of deoxyribonucleosides. Thymidine kinase activity
was encoded by two thymidine kinase 1 (EC 2.7.1.21)-like genes (AtTK1a
and AtTK1b). Deoxyadenosine, deoxyguanosine and deoxycytidine kinase
activities were encoded by a single AtdNK gene. T-DNA insertion lines of
AtTK1a and AtTK1b mutant genes had normal growth, although AtTK1a
AtTK1b double mutants died at an early stage, which indicates that
AtTK1a and AtTK1b catalyze redundant reactions. The results obtained in
the present study suggest a crucial role for the salvage of thymidine during
early plant development.
Arabidopsis; deoxyribonucleoside kinase;
gene duplication; nucleoside nucleotide
metabolism; thymidine kinase
Correspondence
A. R. Clausen, National Institute of
Environmental Health Sciences, NIH, DHHS,
Research Triangle Park, NC 27709, USA
Fax: +1 919 541 7613
Tel: +1 919 541 5388
E-mail: clausenar@niehs.nih.gov
(Received 12 May 2012, revised 10 July
2012, accepted 13 August 2012)
doi:10.1111/j.1742-4658.2012.08747.x
Database
Sequence data from the present study have been deposited in the EMBL database/GenBank
under accession numbers: AT3G07800.1 (AtTK1a), At5G23070.1 (AtTK1b) and AT1G72040.1
(AtdNK).
Introduction
Living organisms need a balanced supply of DNA pre-
cursors to replicate their DNA and the enzymatic reac-
tions that are involved are vital for any cell.
Deoxyribonucleosides (dNs) and deoxyribonucleotides
(dNts) can be synthesized via the de novo or salvage
pathways. Ribonucleotide reduction by ribonucleotide
reductase (RNR) is the rate limiting step in the
de novo pathway and the salvage pathway is controlled
Abbreviations
BVDU, (E)-5-(2-bromovinyl)-2′-deoxyuridine; dAK, deoxyadenosine kinase; dCK, deoxycytidine kinase; dGK, deoxyguanosine kinase;
dN, deoxyribonucleoside; dNK, deoxyribonucleoside kinase; dNts, deoxyribonucleotides; MS, Murashige and Skoog; RNR, ribonucleotide
reductase; TK, thymidine kinase.
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS
1

Salvage of deoxyribonucleosides in plants
A. R. Clausen et al.
by deoxyribonucleoside kinases (dNKs) (EC 2.7.1.145),
which convert dNs into corresponding 5′-monophos-
phate dNs [1–3]. The dNts can subsequently be con-
verted into corresponding di- and tri-phosphates by
nucleotide kinases and incorporated into the newly-
synthesized DNA strand by DNA polymerases.
two TK1-like genes codes for enzymes with thymidine
kinase activity, whereas the TK2-like gene codes for a
multisubstrate dNK with broad specificity towards
deoxyadenosine, deoxyguanosine and deoxycytidine
(but not thymidine). We also report that mutants in
both TK1-like genes are phenotypically normal,
although a TK1 double mutant died at an early stage.
dNKs are well characterized in unicellular eukary-
otes [4], insects [5] and mammals [2]. However, these
enzymes are not present in yeast and fungi, and the
genes coding for enzymes in the salvage pathway have
presumably been lost in this lineage during evolution
[3]. In the unicellular amoeba Dictyostelium discoide-
um, there are three dNKs: thymidine kinase 1 (TK1)
(EC 2.7.1.21), deoxyadenosine kinase and deoxyguano-
sine kinase [4]. Insects have only a single multisub-
strate enzyme expressing all four dNK activities [5].
Humans have four different dNKs: TK1 and deoxy-
cytidine kinase (dCK), which are mainly located in the
cytosol, and thymidine kinase 2 (TK2) and deoxygu-
anosine kinase (dGK), which are found in the mito-
chondria. The mitochondrial dNKs are essential for
the replication and repair of mitochondrial DNA
because no RNR activity is present in the human
mitochondria. However, although mice deficient in
TK1 [6], TK2 [7] or dCK [8] are severely sick, they
can pass the developmental stage and reach the adult
stage.
Results
A. thaliana phosphorylates all native
deoxyribonucleosides
Initially, we measured the dNK activities in A. thaliana
buds and siliques using radiolabelled dNs as substrates
and ATP as the phosphate donor. The highest activity
was found in the buds with dA and dG. However, dC
and dT were also phosphorylated but to a lower extent
(Fig. 1). A similar relative distribution between the
four activities was observed in green immature siliques,
although the activities were slightly lower. The activi-
ties in seeds were also monitored during culture in
liquid medium and after 24 h, the observed dNK
activities were low (Fig. 2A). Over subsequent days,
the dNK activities increased to almost constant levels.
A peak of the dCK and TK activities could be seen at
day 4. Deoxyadenosine kinase (dAK) activity increased
daily with a peak at day 7, whereas dGK activity was
relatively constant during the time course of the study.
We also monitored dNK activities during the growth
of a cell culture for 8 days after inoculation, although
the activities did not change much, even if the growth
rate of the cells peaked at day 4 (Fig. 2B).
The nucleotide precursor metabolism is poorly
understood in plants. Salvage of pyrimidines has been
described in Pisum sativum [9], Solanum tuberosum
[10], Pinus radiata [11] and Picea glauca [12]. In
A. thaliana, it was demonstrated that the response to
mutagens and proper plant development was depen-
dent of RNR [13] and that RNR and TK1-like genes
were transcribed during genotoxic stress [14]. Salvage
of uracil [15] and uridine [16] was shown to be impor-
tant in early plant development and, furthermore, the
degradation of uridine might be important for germi-
nation and plant development [17]; however, it was
also demonstrated that the same two nucleoside hydro-
lases are unimportant for seedling germination and
development [18]. Recently, a nucleobase importer was
shown to provide substrates for the essential salvage
of pyrimidines [19]. A unique TK1 with the capability
to phosphorylate dN monophosphate was isolated
from tomato [20], and in a preliminary study, one
dNK in A. thaliana was also reported [21].
In the present study, we aimed to characterize the
molecular basis for salvage of dNs in A. thaliana
through biochemical and genetic analysis. We show
that A. thaliana can salvage dNs and also that two
types of dNK genes are present in A. thaliana: two
TK1-like enzymes and a single TK2-like enzyme. The
Fig. 1. Deoxyribonucleoside kinase activities in two plant organs.
Protein extracts of green immature siliques and buds from
1-month-old plants of A. thaliana were analyzed for their ability to
phosphorylate the four natural deoxyribonucleosides (dT, dA, dG
and dC) at 200 lM. Activities with purines dNs are higher than with
pyrimidine dN substrates.
2
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS

A. R. Clausen et al.
Salvage of deoxyribonucleosides in plants
(Fig. S1). An N-terminal deleted mutant with the first
45 amino acids was constructed and named AtTK1b-
dN45. It should be noted that the genomic sequence of
AtTK1b, in contrast to AtTK1a, does not contain any
introns. Multiple alignments of the plant AtTK1a,
AtTK1b-dN45 and AtdNK amino acid sequences,
together with other described dNKs, are shown in Figs
S1 and S2. A phylogenetic analysis of a few character-
istic TK1 and the two A. thaliana TK1s described in
the present study showed that the two plant TK1s
were not duplicated very recently, although this defi-
nitely occurred after the split of animal and plant
TK1s (Fig. 3).
A
TK1 activity is essential
B
In the SALK T-DNA insertion collection library
(Table S1), we identified two putative mutant lines of
AtTK1a and one putative mutant line of AtTK1b that
had the T-DNA insertion in exons. We confirmed the
insertion of the T-DNA by gene specific primers
(Table S1). We could not observe any effect on plant
growth for these three mutants (Fig. 4A–C). In addi-
tion, the measured TK1 and dAK activities were similar
to that observed in wild-type plants (data not shown).
We crossed two homozygous mutants of AtTK1a
(AtTK1a-1::T-DNA and AtTK1a-2::T-DNA) with a
homozygous mutants of AtTK1b (AtTK1b::T-DNA) to
generate double mutants, (AtTK1ab-1::T-DNA) and
(AtTK1ab-2::T-DNA), lacking both functional TK1
genes. The F1 generation was heterozygous at both
loci, AtTK1a+/AtTK1aꢀ and AtTK1b+/AtTK1bꢀ, and
these lineages behaved similar to the wild-type. We
screened over 100 independent plants from the F2 pop-
ulations on Murashige and Skoog (MS) plates for their
genotype (Table S2). We identified 12 out of 116
plants of the AtTK1ab-1::T-DNA mutant line and
seven out of 110 plants of the AtTK1ab-2::T-DNA
mutant parental line (Fig. 4) that were unable to
develop into a mature plant. These evenly whitish
seedlings did not develop true leaves and only survived
a short period in vitro. PCR analysis of these whitish
small plants (maximum size of 3 mm) showed that
they had homozygous mutations in both the AtTK1a
and AtTK1b genes. Therefore, we characterized this
double mutation as lethal. In addition, we observed 13
plants out of 89 in the AtTK1ab-1::T-DNA cross and
11 out of 81 in the AtTK1ab-2::T-DNA cross that had
variegated leaves (Fig. S3). This variegation was only
visible up to 3 weeks and, subsequently, these leaves
turned green and the plants had a delayed flowering.
These plants produced normal amounts of seeds. PCR
analysis of genomic DNA from these plants revealed
Fig. 2. Time course study on TK, dCK, dAK and dGK activities
during the first days of plant growth. (A) The four dN kinase
activities on seedlings after subculture at day 0. (B) The four dN
kinase activities on
a A. thaliana cell culture of the ecotype
Landsberg after subculture at day 0. dN kinase activities were
determined with 200 lM of the natural substrates dT, dC, dA and
dG. The results are obtained from three independent experiments.
Error bars represent the SD.
Two TK1-like kinases are present in the genome
Based on homology with human TK1, two different
GeneBank entries, accession numbers AAF13097 and
BAB09824, for A. thaliana were identified. AAF13097
(gene locus AT3G07800) predicted an ORF of 717 bp,
resulting in a protein of 238 amino acids with a calcu-
lated molecular mass of 26.1 kDa (AtTK1a).
BAB09824 (gene locus AT5G23070) predicted an ORF
of 834 bp, resulting in a protein of 277 amino acids
with
a
calculated molecular mass of 30.7 kDa
(AtTK1b). The amino acid identity between the
AtTK1a and AtTK1b proteins was 62%. The N-termi-
nal, however, was extended in the AtTK1b protein
and contained a putative organelle localization signal
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS
3

Salvage of deoxyribonucleosides in plants
A. R. Clausen et al.
S.enteritidis TK (DQ384603)
T.maritima TK1 (AAD35486)
B.cereus TK1 (DQ384595)
U.parvum TK (NP_078433)
H.sapiens TK1 (NM_003258)
A.thaliana TK1a (CAF06344)
A.thaliana TK1b dN45 (CAF06346)
B.cereus dGK (DQ384597)
M.mycoides dAK (AJ313086)
B.cereus dAK (DQ384596)
L.acidophilus dAK (AAB09750)
L.acidophilus dGK (AAB09751)
A.thaliana dNK (AAG51141)
H.sapiens dCK (NP_000779)
H.sapiens dGK (AAC50624)
H.sapiens TK2 (O00142)
B.mori dNK (AAK28318)
D.melanogaster dNK (CAB41881)
Fig. 3. Phylogenetic relationship among deoxyribonucleoside kinases. Phylogenetic clustering tree of the dNKs described in the present
study together with other known dNKs from eukaryotes and prokaryotes. dNKs can be separated in two large groups: the TK1-group and
the non-TK1-group. The TK1-group can further be subgrouped into a human-TK1-group and a non-human-TK1-group comprising TK1
enzymes from Gram-negative bacteria. The non-TK1-group can also be subgrouped into a eukaryotic non-TK1-like-group and a bacterial non-
TK1-group. The putative signal sequence containing 45 amino acids were deleted in the N-terminus of AtTK1b. The numbers at the nodes
are the percentage of frequencies with which a given branch appeared in 1000 bootstraps replicates. Note that AtdNK clusters with the
animal dCK and dGK. Further phylogenetic analysis is provided elsewhere [27].
that they were homozygous at the AtTK1b-1::T-DNA
locus but heterozygous at AtTK1a-1::T-DNA.
relationship between the plant dNKs and other dNKs
from eukaryotes, as well as from prokaryotes, is
presented in Fig. 3, and shows that AtdNK groups
together with human dCK and dGK (Fig. 3) but
branches out before the split of the two animal kinases.
Both TK1s have the same substrate specificity
We subcloned, over-expressed and purified the plant
ORFs as glutathione S–transferase-fusions encoding
putative A. thaliana AtTK1a, AtTK1b, AtTK1b-dN45
(which was missing the 45 N-terminal residues) and
AtdNK. Recombinant enzymes were purified by affin-
ity-chromatography. The kinetic parameters of these
purified recombinant enzymes were studied (Table 1).
AtTK1a, AtTK1b and the AtTK1b-dN45 phosphory-
lated dU, dT and the antiviral drug 3′-azido-3′-deoxyt-
hymidine (Fig. S4) and hyperbolic Michaelis–Menten
kinetics were observed with all of the dNs tested.
AtdNK is a multisubstrate dNK
Initial attempts to purify recombinant AtdNK were
unsuccessful because all of the dNK activity was lost
during the purification. Therefore, several buffers and
combinations of varying concentrations of Triton
X-100, glycerol, MnCl2, MgCl2, EDTA, CaCl2, benz-
amidine, aminocaproic acid, GTP, ATP, TTP or
NaPPP (sodium pyrophosphate) were investigated for
their effect on the activity and stability (Fig. S5).
Clearly, Tris-buffer was poorer than the phosphate
buffer. The addition of 2 mM ATP and 2 mM MgCl2
to the standard NaCl/Pi buffer increased the activity in
the lyzed extract by approximately fivefold, although it
had no stabilizing effect during storage for 24 h at
ꢀ80 °C. However, the addition of 2 mM MnCl2 to the
NaCl/Pi buffer containing 2 mM ATP and 2 mM
MgCl2 not only increased the activity in the lyzed
Only one non-TK1-like is present
When deposited sequences from A. thaliana were ana-
lyzed for homology to DmdNK, a putative dNK with
a complete ORF of 1086 bp, resulting in a protein of
361 amino acids with a calculated mass of 41.2 kDa,
was found (accession number: AAg51141, gene locus
AT1G72040). We were unable to identify any other
putative dNKs in the genome. The phylogenetic
extract by two-fold, but also had a remarkably
stabilizing effect, which was maintained 100% during
4
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS

A. R. Clausen et al.
Salvage of deoxyribonucleosides in plants
A
E
B
D
C
F
Fig. 4. Seedlings of homozygous mutant lines with mutated kinase genes. Top left to right: (A) AtTK1a-1::T-DNA, (B) AtTK1a-2::T-DNA and
(C) AtTK1b-1::T-DNA homozygous mutant lines, respectively, and (D) wild-type. Flowering time, number of leaves, size of leaves and seed
production were monitored, although there was no difference in appearance and growth between these plants and wild-type. In addition,
the TK activities were close to that of the wild-type. Examples from the F2 populations originating from crosses for the generation of double
TK1 mutant lines: (E) AtTK1ab-1::T-DNA to the bottom left and (F) AtTK1ab-2::T-DNA to the bottom right. The whitish plants are double
mutants, according to PCR analysis, and were unable to develop into adult plants.
storage for at least 6 months at –80 °C. Adding 2 mM
ATP, 2 mM MgCl2 and 2 mM MnCl2 to the chroma-
tography buffer made it possible to purify the enzyme
with a good recovery.
range 14–26 lM, whereas the Km for dU was 561 lM.
Positive cooperativity was found with several sub-
strates (Table 2 and Fig. S4). Despite the efficient
AtdNK had a broad substrate specificity and could
phosphorylate dA, dG and dC with Km values in the
Table 2. Kinetic parameters of the recombinant plant AtdNK. The
data for HsTK2 [35], HsdCK [37] and HsdGK [38] are also shown. NA,
no activity; h, Hill constant. *The two K0.5 values were obtained at
substrate concentrations below or above 8–10 lM substrate.
Table 1. Kinetic parameters of the recombinant plant TK1
enzymes. The data for HsTK1 are also shown [35]. AZT, 3′-azido-3′-
deoxythymidine.
K
0.5 (lM)
kcat /K0.5
(sꢀ1ꢁM
ꢀ1
(
SD)
kcat (sꢀ1
)
)
h
ꢀ1
KM (lM) ( SD)
kcat (sꢀ1) ( SD)
kcat /Km (sꢀ1ꢁM
)
AtdNK
dA
14.2 0.7
16.3 1.0
26.3 1.8
159 20
0.47 0.17
0.29 0.07
0.27 0.08
0.77 0.02
3.3 9 104
1.8 9 104
1.0 9 104
4.8 9 103
1.0
AtTK1a
dU
dC
1.9 0.07
2.0 0.07
1.0
15
2
1.5 0.1
1.5 9 105
7.7 9 105
4.8 9 104
dG
dT
0.62 0.08
2.7 0.8
0.48 0.03
0.13 0.03
dFdC
0.105 0.003 9.3 9 102
1.0
AZT
AtTK1b
dU
BVDU 113 16
dT
dU
0
0
0
NA
6.6 0.5
1.4 0.4
0.38 0.04
0.045 0.006
0.020 0.002
0.009 0.002
6.8 9 103
1.4 9 104
2.2 9 104
561 105
0.23 0.06
4.1 9 102
1.0
dT
HsdCK
dA
AZT
120
1
0.32
0.074
0.41
2.7 9 103
7.4 9 104
2.7 9 103
1
1
1
AtTK1bdN45
dC
dU
dT
5.6 1.3
0.30 0.07
0.20 0.09
0.22 0.03
5.3 9 104
2.4 9 105
1.2 9 105
dG
150
0.84 0.19
1.9 0.4
HsTK2
dT
AZT
HsTK1
dU
0.3; 16*
36
0.29
0.47
9.6 9 105
1.3 9 104
1
1
dC
9.0
0.5
0.6
5.0
4.0
2.1
5.6 9 105
8.0 9 106
3.5 9 106
HsdGK
dA
2.5 9 104
6.4 9 104
1
1
dT
60
36
1.5
2.3
AZT
dG
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS
5

Salvage of deoxyribonucleosides in plants
A. R. Clausen et al.
phosphorylation of dU and (E)-5-(2-bromovinyl)-2′-de-
oxyuridine (BVDU), dT could not be phosphorylated.
plants from the F2 generation of plants that were
homozygous at the AtTK1b locus (Fig. S3E) but het-
erozygous at the AtTK1a locus indicates that the loss
of the AtTK1b gene gives a stronger phenotype than
the loss of the AtTK1a gene because we did not detect
any growth phenotypes on plants homozygous for
the AtTK1a mutation and heterozygous for the
AtTK1b mutation. By contrast, double mutants having
mutations in both AtTK1 genes could not develop into
fully growing plants, suggesting that at least one copy
of the TK1 gene is necessary to complete plant
development.
Discussion
The salvage of dNs in plants has been poorly
described. In the present study, we report that A. tha-
liana can salvage dNs and we isolated and character-
ized the responsible enzymes. Extracts of plants from
early growth to mature plant were analyzed for phos-
phorylating capacity towards all four natural dNs
and it is shown that all four dNs can be phosphory-
lated with a preference towards purines over pyrimi-
dines. The dNK activities were at the same level in
developing organs such as buds and green immature
siliques (Fig. 1), plant seedlings (Fig. 2A) and plant
cell cultures (Fig. 2B), indicating that dNK activities
were not limited to a specific cell tissue or develop-
mental stage. The low thymidine kinase activity in
the developing cells and tissues was unexpected com-
pared to the mammalian cells, which have high TK1
activity in dividing cells [22–24]. The dNK in the fruit
fly is also cell cycle regulated, although the demand
for dNTPs and the regulation of the nucleotide pool
in plants is apparently different from that in mam-
mals [25].
TK1 from humans can only phosphorylate dT and
dU but no other natural dN, and similarly recombi-
nant purified AtTK1a and AtTK1b also have narrow
substrate specificity. A 45 amino acid deletion of the
N-terminal of AtTK1b results in an enzyme with the
same Km as the full-length product but an approxi-
mately ten-fold higher kcat, supporting the idea that
those 45 amino acids represent a signal peptide that
needs to be processed before AtTK1b is fully func-
tional.
AtdNK is phylogenetically related to the TK2
family and could represent a model for deducing a
common ancestor to human dCK and dGK enzymes.
AtdNK is larger than the human mammalian counter-
parts and has one extra segment of 32 amino acids
close to the C-terminal and an extended C-terminal of
~ 55 amino acids (Fig. S2). The biological role of these
extra segments remains unknown. Purification of
AtdNK has been complicated as a result of low stabil-
ity during purification. Thus, to be able to purify this
dNK, we set a screen aiming to identify a suitable buf-
fer. We found that the addition of 2 mM ATP, 2 mM
MnCl2, 2 mM MgCl2 and 1% Triton X-100 increased
the activity by almost twenty-fold, and also that the
use of this buffer facilitated the purification of recom-
binant AtdNK (Fig. S5), with MnCl2 apparently being
essential for the in vitro stability for this dNK. AtdNK
has a broad substrate specificity and was able to
phosphorylate deoxyadenosine, deoxyguanosine and
deoxycytidine (but not thymidine), supporting the idea
that AtdNK could represent a common ancestor to
human dGK and dCK, a nonspecialized dNK with a
broad specificity. In addition AtdNK could efficiently
phosphorylate nucleoside analogues such as dif-
luorodeoxycytidine and BVDU. The phosphorylation
of dU and BVDU (but not dT) suggests that there is a
specific steric hinderance for a methyl group on the
base, although apparently not for a bromovinyl group.
A high resolution structure of this enzyme would be
necessary to explain the findings reported in the
present study.
We describe that two TK1-like enzymes (AtTK1a
and AtTK1b) and a single TK2-like enzyme (AtdNK)
are encoded by the A. thaliana genome. Two extra tk1
genes have also been found in mice, although both
psedogenes are inactive in situ [26]. The phylogenetic
relationship with human TK1 suggests that the dupli-
cation took place after the separation of the lineages
leading to plants and mammals (Fig. 3). AtTK1b is
longer in its N-terminal sequence than AtTK1a and its
human homologue. Longer N-terminals fragments are
also found in human TK2 and dGK, which are
located in the mitochondria, suggesting that AtTK1b
could be located in plastids and maybe mitochondria.
Preliminary analysis suggests that A. thaliana mito-
chondria are responsible for a major part of the
observed dNK activity [27].
Single gene mutants of TK1, dCK or TK2 result in
severe phenotypes in mice; however, single gene
mutants of AtTK1a and AtTK1b could grow normally,
indicating that the genes do not encode essential activi-
ties (Fig. 4A–C). The lack of any visible phenotype in
the single gene mutants, either AtTK1a or AtTK1b,
could have two explanations: (a) thymidine salvage
has a minor role in A. thaliana or (b) the genes for
TK1a and TK1b are redundant. To test the second
hypothesis, AtTK1a AtTK1b double mutants were gen-
erated and analyzed. The appearance of variegated
6
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS

A. R. Clausen et al.
Salvage of deoxyribonucleosides in plants
Protein extraction from plant cells and tissues
Materials and methods
Arabidopsis thaliana (ecotype Columbia) for preparations
of siliques and buds were grown in soil under standard
greenhouse conditions at 23 °C with ~ 16 h light per day.
Green siliques and buds were harvested on 4-week-old
plants. In total, 1–5 g of plant material and ice-cold plant
extraction buffer (1 x NaCl/Pi, 5 mM dithiothreitol, 0.1%
Triton X-100, 6 mM NaF, 2 mM ATP, 2 mM MnCl2,
2 mM MgCl2, 10% glycerol, 0.2 mM phenylmethanesulfo-
nyl fluoride, 0.01 mM leupeptin, 1 nM calyculin and 0.5%
PVP 40) were ground with sand in mortars pretreated
with liquid nitrogen. The cell culture was filtered through a
30-lm nylon net and plant extraction buffer was added.
Subsequently, the culture was thawn and frozen three times
in an ultrasound bath and liquid nitrogen, respectively.
Plant tissues and cell culture were incubated for 10 min at
4 °C with shaking and then centrifuged for 5 min at
18 000 g in a JA-20 rotor (Beckman Coulter, Fullerton,
CA, USA) at 4 °C; supernatants were frozen at ꢀ80 °C.
Biological material
Arabidopsis thaliana ecotype Columbia-0 was used as the
wild-type plant and a green cell culture from A. thaliana eco-
type Landsberg erecta was used for the in vivo measurements
of dNK activities [28]. Escherichia coli XL-1 (Stratagene,
La Jolla, CA, USA) was used for cloning and the TK nega-
tive E. coli KY895 (Fꢀ, tdkꢀ, ilv) was used for expression of
the recombinant dNKs [29]. T-DNA insertion lines were
obtained from the SALK collection [30] and confirmed as
homozygous mutants using gene- and T-DNA-specific prim-
ers (Table S1). We obtained two kind of insertions in the
TK1a gene (referred to as gene A). Therefore, we had a1 and
a2 deletion alleles, designated AtTK1a-1::T-DNA and
AtTK1a-2::T-DNA, respectively. Another insertion line was
obtained, designated AtTK1b-1::T-DNA, in the TK1b gene
(referred to as gene B). The second T-DNA mutant with
insert in AtTK1b (GK-14D03) was also investigated,
although we were only able to recover wild-type from that
seed batch (Table S1). We isolated homozygous mutants for
the TK1a gene (aa) and TK1b gene (bb). The ‘aa’ and ‘bb’
lineages were crossed and generated the F1 generation
(AaBb), which was then self-crossed to generate the F2 popu-
lations (Table S2).
Cloning of genes encoding plant kinases
The ORFs were amplified from a cDNA library obtained
from Stratagene by PCR using specific primers with overhang
for cloning and then cloned into pGEX-2Ts expression vector
(Amersham Pharmacia Biotech/GE Healthcare, Uppsala,
Sweden). In addition, primers were designed to construct
N-terminal deletions of 22, 45 and 63 amino acids. The
constructs harbouring AtTK1a, AtTK1b, AtTK1b-dN22,
AtTK1b-dN45, AtTK1b-dN63 and AtdNK were named
P604, P661, P662, P663, P664 and P605, respectively, and
sequenced using a commercial source (MWG-Biotech, Ebers-
berg, Germany).
Chemicals
Unless stated otherwise, chemicals were of the highest pur-
ity and were obtained either from Sigma (St Louis, MO,
USA) or Merck (Damstadt, Germany). All 3H-labelled
nucleosides were from Moravek Biochemicals Inc. (Brea,
CA, USA), except 3H-labelled dT, which was obtained
from GE Healthcare (Little Chalfont, UK). dFdC was
obtained from Thykn (Mumbai, India). Thrombin was
obtained from Biofac (Kastrup, Denmark).
Phylogenetic analysis
Amino acid sequence alignments were performed using CLU-
STALW [32]. The phylogenetic analysis was performed using
TREECON, version 1.3b [33] by means of clustering and
neighbourhood-joining calculation approaches.
Cultivation of plant cell material
Plant cell culture was subcultured in the medium con-
taining 1 9 Murashige and Skoog (MS) basal salt mix
[31], 1 9 Gamborg’s B5 vitamins and 3% sucrose, with
pH adjusted to 5.7, and, after autoclaving, 0.5 mgꢁLꢀ1
naphthalene acetic acid and 0.05 mgꢁLꢀ1 N6 benzyl ade-
nine was added. In total, 3 mL of the 1-week culture
was transferred in 100 mL of fresh medium and culti-
vated at 22 °C under a 16 : 8 h light:dark cycle. A. thali-
ana seeds (ecotype Columbia) for liquid seedling culture
were inoculated into 100 mL of medium (0.5 9 MS med-
ium, 2% sucrose, pH 5.8) and cultivated at 22 °C under
a 16 : 8 h light:dark cycle. For plant growth experiments,
surface sterilized A. thaliana seeds were sown on MS
plates.
Expression and purification of plant dNKs
Recombinant AtTK1a, AtTK1b and AtTK1b-dN45 were
expressed and purified as described previously [34]. AtdNK
was also purified in accordance with this procedure but
with 2 mM MgCl2, 2 mM MnCl2 and 2 mM ATP added to
all buffers to stabilize the enzyme.
dNK enzyme assays
dNK activities were determined by initial velocity measure-
ments based on four time samples by the DE-81 filter paper
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS
7

Salvage of deoxyribonucleosides in plants
A. R. Clausen et al.
(Whatman International, Maidstone, UK) assay using vari-
ous radiolabelled/labelled dN substrate concentrations [35].
For determination of dNK activities in E. coli cell extracts,
3 mM NaF was added to the reaction mixture and, when dC
was used as substrate, 0.5 mM of the cytidine deaminase
inhibitor tetrahydrouridine was added. One unit (u) of dNK
activity is defined as the formation of the corresponding
9 Bressan RA, Murray MG, Gale JM & Ross CW
(1978) Properties of pea seedling uracil
phosphoribosyltransferase and its distribution in other
plants. Plant Physiol 61, 442–446.
10 Katahira R & Ashihara H (2002) Profiles of pyrimidine
biosynthesis, salvage and degradation in disks of potato
(Solanum tuberosum L.) tubers. Planta 215, 821–828.
11 Stasolla C, Loukanina N, Ashihara H, Yeung EC &
Thorpe TA (2007) Comparative studies on pyrimidine
metabolism in excised cotyledos of Pinus radiata during
shoot formation in vitro. J Plant Physiol 164, 429–441.
12 Stasolla C, Loukanina N, Ashihara H, Yeung EC &
Thorpe TA (2006) Changes of purine and pyrimidine
nucleotide biosynthesis during shoot initiation from
epicotyl explants of white spruce (Picea glauca). Plant
Sci 171, 345–354.
monophosphate product at at a rate of 1 nmolꢁminꢀ1
.
Kinetic data were evalulated by nonlinear regression analy-
sis using the Michaelis–Menten equation v = Vmax 9 [S]/
h
h
(Km + [S]) or the Hill equation v ¼ Vmax ꢂ ½Sꢃ =ðKh0:5 þ ½Sꢃ Þ
as described previously [36].
Acknowledgements
The authors acknowledge funding from Swedish
Research Council (VR), Cancerfonden (Sweden), Wen-
ner-Gren Foundation, Crafoord Foundation, Sorensen
Foundation, Fysiografen and Jubi Kinase ApS.
13 Wang C & Liu Z (2006) Arabidopsis ribonucleotide
reductases are critical for cell cycle progression, DNA
damage repair, and plant development. Plant Cell 18,
350–365.
14 Chen IP, Haehnel U, Altschmied L, Schubert I &
Puchta H (2003) The transcriptional response of
Arabidopsis to genotoxic stress – a high-density colony
array study (HDCA). Plant J 35, 771–786.
References
1 Arner ESJ & Eriksson S (1995) Mammalian
deoxyribonucleoside kinases. Pharmacol Ther 67,
155–186.
15 Mainguet SE, Gakiere B, Majira A, Pelletier S, Bringel
F, Guerard F, Caboche M, Berthome R & Renou JP
(2009) Uracil salvage is necessary for early Arabidopsis
development. Plant J 60, 280–291.
2 Eriksson S, Munch-Petersen B, Johansson K &
Ecklund H (2002) Structure and function of cellular
deoxyribonucleoside kinases. Cell Mol Life Sci 59,
1327–1346.
16 Chen MJ & Thelen JJ (2011) Plastid uridine salvage
activity is required for photoassimilate allocation and
partitioning in Arabidopsis. Plant Cell 23, 2991–3006.
17 Jung B, Florchinger M, Kunz HH, Traub M,
Wartenberg R, Jeblick W, Neuhaus HE & Mohlmann
T (2009) Uridine-ribohydrolase is a key regulator in the
uridine degradation pathway of Arabidopsis. Plant Cell
21, 876–891.
3 Sandrini MPB & Piskur J (2005) Deoxyribonucleoside
kinases: two enzyme families catalyze the same reaction.
Trends Biochem Sci 30, 225–228.
4 Sandrini MPB, Soderbom F, Mikkelsen NE & Piskur
J (2007) Dictyostelium discoideum salvages purine
deoxyribonucleosides by highly specific bacterial-
like deoxyribonucleoside kinases. J Mol Biol 369,
653–664.
18 Riegler H, Geserick C & Zrenner R (2011) Arabidopsis
thaliana nucleosidase mutants provide new insights into
nucleoside degradation. New Phytol 191, 349–359.
19 Witz S, Jung B, Furst S & Mohlmann T (2012) De
novo pyrimidine nucleotide synthesis mainly occurs
outside of plastids, but a previously undiscovered
nucleobase importer provides substrates for the
essential salvage pathway in Arabidopsis. Plant Cell
Online 24, 1549–1559.
5 Munch-Petersen B, Piskur J & Sondergaard L (1998)
Four deoxynucleoside kinase activities from Drosophila
melanogaster are contained within a single monomeric
enzyme, a new multifunctional deoxynucleoside kinase.
J Biol Chem 273, 3926–3931.
6 Dobrovolsky VN, Bucci T, Heflich RH, Desjardins J &
Richardson FC (2003) Mice deficient for cytosolic
thymidine kinase gene develop fatal kidney disease.
Mol Genet Metab 78, 1–10.
20 Khan Z, Knecht W, Willer M, Rozpedowska E,
Kristoffersen P, Clausen AR, Munch-Petersen B,
Almqvist PM, Gojkovic Z, Piskur J et al. (2010)
Plant thymidine kinase 1: a novel efficient suicide
gene for malignant glioma therapy. Neuro Oncol 12,
549–558.
7 Zhou XS, Solaroli N, Bjerke M, Stewart JB, Rozell B,
Johansson M & Karlsson A (2008) Progressive loss of
mitochondrial DNA in thymidine kinase 2-deficient
mice. Hum Mol Genet 17, 2329–2335.
8 Toy G, Austin WR, Liao H-I, Cheng D, Singh A,
Campbell DO, Ishikawa T-o, Lehmann LW,
Satyamurthy N, Phelps ME et al. (2010) Requirement
for deoxycytidine kinase in T and B lymphocyte
development. Proc Natl Acad Sci USA 107, 5551–5556.
21 Clausen AR, Girandon L, Knecht W, Survery S,
Andreasson E, Munch-Petersen B & Piskur J (2008)
ˇ
A multisubstrate deoxyribonucleoside kinase from
plants. Nucleic Acids Symp Ser 52, 489–490.
8
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS

A. R. Clausen et al.
Salvage of deoxyribonucleosides in plants
22 Johnson LF, Rao LG & Muench AJ (1982) Regulation
of thymidine kinase enzyme level in serum-stimulated
mouse 3T6 fibroblasts. Exp Cell Res 138, 79–85.
23 Munch-Petersen B & Tyrsted G (1977) Induction of
thymidine kinases in phytohaemagglutinin-stimulated
human lymphocytes. Biochim Biophys Acta 478, 364–375.
24 Sherley JL & Kelly TJ (1988) Regulation of human
thymidine kinase during the cell cycle. J Biol Chem 263,
8350–8358.
deoxyribonucleoside kinases in a species-specific
manner. J Antimicrob Chemother 60, 510–520.
35 Munch-Petersen B, Cloos L, Tyrsted G & Eriksson S
(1991) Diverging substrate specificity of pure human
thymidine kinases 1 and 2 against antiviral
dideoxynucleosides. J Biol Chem 266, 9032–9038.
36 Knecht W, Bergjohann U, Gonski S, Kirschbaum B &
Loffler M (1996) Functional expression of a fragment
of human dihydroorotate dehydrogenase by means of
the baculovirus expression vector system, and kinetic
investigation of the purified recombinant enzyme.
Eur J Biochem 240, 292–301.
25 Legent K, Mas M, Dutriaux A, Bertrandy S, Flagiello
D, Delanoue R, Piskur J & Silber J (2006) In vivo
analysis of Drosophila deoxyribonucleoside kinase
function in cell cycle, cell survival and anti-cancer drugs
resistance. Cell Cycle 5, 740–749.
37 Bohman C & Eriksson S (1988) Deoxycytidine kinase
from human leukemic spleen: preparation and
characteristics of homogeneous enzyme. Biochemistry
27, 4258–4265.
26 Seiser C, Beck G & Wintersberger E (1990) The
processed pseudogene of mouse thymidine kinase is
active after transfection. FEBS Lett 270, 123–126.
27 Clausen AR (2009) Deoxyribonucleoside kinases in
bacteria, plants and humans, PhD thesis. Lund
University, Sweden.
38 Wang L, Karlsson A, Arner ES & Eriksson S (1993)
Substrate specificity of mitochondrial 2′-deoxyguano-
sine kinase. Efficient phosphorylation of
2-chlorodeoxyadenosine. J Biol Chem 268, 22847–22852.
28 Lenman M, Sorensson C & Andreasson E (2008)
Enrichment of phosphoproteins and phosphopeptide
derivatization identify universal stress proteins in
elicitor-treated Arabidopsis. Mol Plant-Microbe Interact
21, 1275–1284.
Supporting information
The following supplementary material is available:
Fig. S1. Multiple amino acid sequence alignments of
the AtTK1a and AtTK1b described in the present
study together with some other known TK1s originat-
ing from eukaryotes and prokaryotes.
29 Igarashi K, Hiraga S & Yura T (1967) A
deoxythymidine kinase deficient mutant of Escherichia
coli. II. Mapping and transduction with phage phi80.
Genetics 57, 643–654.
Fig. S2. Multiple amino acid sequence alignment of
characterized eukaryotic non-TK1 kinases, including
AtdNK.
30 Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen
HM, Shinn P, Stevenson DK, Zimmerman J, Barajas P,
Cheuk R et al. (2003) Genome-wide Insertional
mutagenesis of Arabidopsis thaliana. Science 301,
653–657.
Fig. S3. Fourteen-day-old plants.
Fig. S4. Kinetic parameters of the recombinant plant
kinases with different substrates.
31 Murashige T & Skoog F (1962) A revised medium for
rapid growth and bio assays with tobacco tissue
cultures. Physiol Plantarum 15, 473–497.
Fig. S5. Stability of AtdNK in different buffers.
Table S1. SALK lines with mutated TK1a and TK1b
used in the present study.
32 Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F
& Higgins DG (1997) The CLUSTAL_X windows
interface: flexible strategies for multiple sequence
alignment aided by quality analysis tools. Nucleic Acids
Res 25, 4876–4882.
Table S2. Genotyping of F2 populations from Parental
crosses between homozygous single mutants at the
TK1 loci (the TK1a and TK1b gene).
This supplementary material can be found in the
online version of this article.
33 Van de Peer Y & De Wachter R (1994) TREECON
for Windows: a software package for the construction
and drawing of evolutionary trees for the Microsoft
Windows environment. Comput Appl Biosci 5,
569–570.
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 reorganized 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.
34 Sandrini MPB, Clausen AR, On SLW, Aarestrup FM,
Munch-Petersen B & Piskur J (2007) Nucleoside
analogues are activated by bacterial
FEBS Journal (2012) ª 2012 The Authors Journal compilation ª 2012 FEBS
9