1276–1281 Nucleic Acids Research, 2000, Vol. 28, No. 5
© 2000 Oxford University Press
Terminal deoxynucleotidyl transferase catalyzes the
reaction of DNA phosphorylation
Andrey A. Arzumanov1, Lyubov S. Victorova1,2,*, Maxim V. Jasko1, Dmitry S. Yesipov3 and
Alexander A. Krayevsky1
1Engelhardt
Institute of Molecular Biology, Russian Academy of Sciences, 32 Vavilov Street, Moscow 117984,
Russia, 2Centre for Medical Studies of University of Oslo, 32 Vavilov Street, Moscow 117984, Russia and
3Moscow State University, Interfaculty Laboratory, Vorob’evy Gory, Moscow 119899, Russia
Received July 7, 1999; Revised October 9, 1999; Accepted January 5, 2000
ABSTRACT
The reaction of phosphorylation and phosphonylation
of an oligodeoxynucleotide 3′-terminal hydroxyl
(oligodeoxynucleotidyl kinase activity) catalyzed by
calf thymus terminal deoxynucleotidyl transferase
(TDT) was found. Triphosphates modified at Pα-, Pα,γor Pα,β,γ-residues served as low-molecular weight
substrates. The reaction was TDT specific; human
DNA polymerases α and β, as well as AMV reverse
transcriptase did not catalyze it. The donor activity of
modified triphosphates or triphosphonates depended
on their structure and was increased with an increase
in their hydrophobicity. The substrate activity of
some modified triphosphates was up to one order of
magnitude higher than that of ddTTP.
INTRODUCTION
The unique DNA polymerase of primitive lymphocytes,
terminal deoxynucleotidyl transferase (TDT; EC 2.7.7.31),
was discovered as an end-addition, template-independent
enzyme of calf thymus extracts and its function has not been
elucidated yet. Its strict limitation in normal animals to primitive
lymphoid cells suggests a role for it in generating the functional
properties of T and B cells. Some experiments strongly support a
role of TDT in insertion of non-germline nucleotides (N segments)
at VDJ joining sites during heavy-chain immunoglobulin gene
rearrangements (1). A similar role for TDT N-segment addition in
T-cell receptor gene rearrangements may also exist (2,3). It has
been shown that the level of TDT in leukocytes of leukemia
patients is very high (4).
In 1994–1995 a new group of very selective inhibitors of
TDT in the cell-free systems was found (5,6). They belong to a
group of modified α-nucleoside 5′-triphosphates of D- and L-series
(I and II, Scheme 1) and their affinity to TDT was comparable
with that of corresponding β-anomers. Based on these results,
we suggested that a dNTP nucleic base did not participate in
the genetically specific interaction with the TDT active center,
and the main contribution to this interaction was made by
dNTP triphosphate residues (7).
Here we describe substrate properties of modified triphosphates of types III–VI (Fig. 1) in the reaction catalyzed by calf
thymus TDT in cell-free solutions.
The structure of potential phosphate donors was selected by
us according to the following chemical criteria: (i) triphosphates
with an α-phosphate or both α- and γ-phosphates esterified by
bulky hydrophobic groups sterically imitating nucleosides, but
lacking chemical functional groups inherent to nucleic bases;
(ii) triphosphates with one or three phosphate residues substituted
by phosphonates; and (iii) dTTP with γ-phosphate substituted
by a phosphonate. The basis for all substitutions will be
discussed below.
Compounds III and VI phosphorylated (phosphonylated)
the 3′-terminus of the oligodeoxynucleotide primers by their
α- or γ-termini, whereas compounds Va–c phosphorylated the
3′-terminus only by their α-phosphonate residues. The intensity of
the incorporation depended both on the structure of the
substrate (Fig. 1) and the TDT preparation. Compounds IVa,b
were completely inactive as phosphorylating agents. The structure
of the reaction products was confirmed by direct comparison
with authentic samples, or by methods of chemical modification.
MATERIALS AND METHODS
The synthesis of compounds III–V will be published elsewhere;
compounds VIa–d were prepared as according toAlexandrova
et al. (8).
The [5′-32P]d(CCCAGTCACGACGTOH) tetradecadeoxynucleotide primer was used. The control oligonucleotide
d(CCCAGTCACGACGTOP) (VII) containing a phosphate
group at 3′-position was synthesized from 5′-O-dimethoxytrityluridine coupled to the glass solid phase as described by
*To whom correspondence should be addressed at: Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, 32 Vavilov Street,
Moscow 117984, Russia. Tel: +7 095 135 2255; Fax: +7 095 135 1405; Email: [email protected]
Present address:
Andrey A. Arzumanov, PNAC Division, MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, UK
Nucleic Acids Research, 2000, Vol. 28, No. 5 1277
and purified on an Ultragel AcA54 (LKB) column in the
standard TE buffer (10 mM Tris–HCl pH 7.6, 1 mM EDTA).
For the experiments with DNA polymerases α and β and
reverse transcriptase, the [5′-32P]primer was annealed with
M13mp10 phage DNA, and the primer–template complex was
isolated on a BioGel A-1.5M (Bio-Rad) column in the TE
buffer.
Primer extension assays
Figure 1. Structure of the tested compounds.
Atkinson and Smith (9). After splitting off the support, the 5′substituted oligonucleotide was oxidized with 100 µl of 0.1 M
NaIO4 (1 h), hydrolyzed with 100 µl of 0.1 M NaOH as
according to Krynetskaya et al. (10) and purified by reverse-phase
HPLC. After the removal of the 5′-dimethoxytrityl protecting
group (80% acetic acid, 30 min, 20°C), resulting VII was
precipitated with 5% LiClO4 in acetone and purified by gel
filtration on a Toyopearl HW-40 column eluting with water.
The other control oligodeoxynucleotidyl 3′-methylphosphonate
d(CCCAGTCACGACGTOPMe) (VIII) was synthesized on the
same uridine-modified support by coupling with the 3′-Oimidazolyl(methyl)phosphonyl 5′-O-dimethoxytritylthymidine
according to Miller et al. (11). The product was treated and
purified as described for VII with an additional purification by
gel electrophoresis.
Enzymological procedures
Different calf thymus TDT lots were used: A401-1, A1601-2,
A1801-3 (USB Amersham Life Science), M187A, M187B
(Promega) and HF3403 (Gibco). TDTP+ (lots A401-1, M187A
and M187B) could incorporate an unmodified γ-phosphate
residue of the substrate into oligodeoxynucleotide 3′-termini.
TDTP– (lots A1601-2, A1801-3 and HF3403) lacked this property.
AMV reverse transcriptase was from Amersham; DNA
polymerases α and β were isolated from human placenta in
accordance with the methods of Mozzherin et al. (12) and
Kolocheva and Nevinsky (13), respectively. Tetradecadeoxynucleotide primer as well as oligonucleotides VII–VIII were
labeled at the 5′-termini using [γ-32P]ATP (Radioisotop,
Moscow, Russia) and T4 polynucleotide kinase (Amersham)
For TDT, the assay mixture (6 µl) contained 0.02 µM
[5′-32P]primer, the compounds under study or dNTP, two
activity units of TDT from Amersham or Gibco, 100 mM
sodium cacodylate pH 7.2, 2 mM CoCl2 and 0.05 mM DTT; or
six activity units of the TDT from Promega, 100 mM sodium
cacodylate pH 6.8, 1 mM CoCl2 and 0.05 mM DTT. For AMV
reverse transcriptase, the assay mixture (6 µl) contained
0.01 µM DNA–[5′-32P]primer complex, the compounds under
study or dNTP, two activity units of the enzyme, 10 mM Tris–HCl
buffer pH 8.2, 5 mM MgCl2, 40 mM KCl and 1 mM DTT. In
the case of DNA polymerase α, the assay mixture (6 µl)
contained 0.01 µM template–primer complex, the compounds
under study or dNTP, one activity unit of the enzyme, 10 mM
Tris–HCl buffer pH 7.4, 6 mM MgCl2 and 0.4 mM DTT. For
DNA polymerase β, the assay mixture (6 µl) contained
0.01 µM template–primer complex, the compounds under
study or dNTP, two activity units of the enzyme, 10 mM Tris–HCl
buffer pH 8.6, 6 mM MgCl2 and 0.4 mM DTT. The reactions
were carried out for 30 min at 37°C and terminated by adding
3 µl of deionized formamide containing 0.5 mM EDTA, 0.1%
bromophenol blue and 0.1% xylene cyanol. Reaction products
were separated by electrophoresis in 20% polyacrylamide/7 M
urea sequencing gel, and the gels obtained were autoradiographed. Protein electrophoreses of the TDT samples were
carried out in the separating 0.1% SDS/8% polyacrylamide gel
in 25 mM Tris–HCl, 250 mM glycine, 0.1% SDS buffer
pH 8.3, according to Sambrook et al. (14).
The N3→NH2 reduction of 3′-phosphonylated oligonucleotides
by DTT
After the reaction of compounds VIb,c with the oligonucleotide
primer catalyzed by TDT (30 min, 37°C), the samples (6 µl)
were heated at 100°C for 3 min. After adding H2O (2 µl), 25%
NH4OH (1 µl) and 0.5 or 1 M DTT (1 µl), the assays were incubated
for 60 min at 30°C. The reactions were stopped by adding 5 µl
of deionized formamide containing 0.5 M EDTA, 0.1%
bromophenol blue and 0.1% xylene cyanol, and reaction
products were separated by electrophoresis in a 20% polyacrylamide/7 M urea sequencing gel.
Kinetic constants
For kinetic assays, the reaction mixture (6 µl) contained
0.02 µM [5′-32P]primer, two activity units of TDT (lot A1601-2),
ddTTP or compounds Va,c in the 100 mM sodium cacodylate
buffer pH 7.2, 2 mM CoCl2 and 0.05 mM DTT. The reaction
was carried out for 2 min at 37°C and stopped by adding 3 µl
of formamide containing 0.5 M EDTA, 0.1% of bromophenol
blue and xylene cyanol. The products were separated by 20%
PAGE, and the autoradiographs were scanned on a Molecular
Dynamics 300A Computing densitometer. The apparent Km
and Vmax were determined from the rate of the product formation
as a function of substrate concentration.
1278 Nucleic Acids Research, 2000, Vol. 28, No. 5
Figure 2. Primer extension by compounds IIIa–d catalyzed by TDTP– (Series
A, lanes 1–14) and TDTP+ (Series A, lanes 15–20, Series B, lanes 1–4).
(A) Lane 1, primer + enzyme (control); lane 2, as in lane 1 + 0.5 µM ddTTP;
lanes 3–6, as in lane 1 + 2.5, 10, 25 and 100 µM IIIa, respectively; lanes 7–10, as
in lane 1 + 2.5, 10, 25 and 100 µM IIIb, respectively; lanes 11–14, as in lane 1
+ 0.025, 0.1, 0.25 and 1 µM IIId, respectively; lanes 15–17, as in lane 1 +
0.01, 0.1 and 1 µM IIId, respectively; lanes 18–20, as in lane 1 + 1, 10 and
100 µM IIIc, respectively. (B) Lane 1, primer + enzyme (control); lane 2, as
in lane 1 + 0.2 µM ddTTP; lane 3, as in lane 1 + 5 µM IIId; lane 4, 3′-phosphorylated 14mer oligodeoxynucleotide VII.
RESULTS AND DISCUSSION
Introduction of a phosphate or phosphate ester group into
the oligodeoxynucleotide primer
The unique ability of TDT to polymerize dNTPs in a templateindependent manner, as well as to incorporate highly selectively
unusual α-anomers of D- or L-series of carbocyclic nucleotides
I and II allowed us to suggest that the triphosphate moiety of
dNTP is crucial for the interaction with enzyme active center,
and synthetic triphosphates esterified at the α-position by alkyl
or aryl groups can be substrates in TDT-catalyzed extension of
the oligodeoxynucleotide primer. Therefore, we synthesized a
series of α-esterified triphosphates bearing aryl groups
differing in size (phenyl, β-naphtyl) or groups capable of
hydrogen bond formation (p-nitrophenyl), as well as aryl
groups attached to a phosphate through a flexible ethyl arm
(p-nitrophenylethyl). In addition, we synthesized α,γ-disubstituted triphosphates IV to elucidate the role of the γ-substitution
in nucleoside-lacking triphosphates. This series could be
useful for studying some sterical aspects of the substrate–
enzyme interaction.
The pattern of the primer extension by III is given in Figure 2A
and B. When the primer extension is catalyzed by TDTP–, new
bands of products (i), whose mobility was lower than that of
the primer were observed. In the case of TDTP+, additional
bands (ii) with higher mobility were found. Bands (ii) for all
compounds under study were located at the same level,
whereas locations of bands (i) were different for different
substrates. The bands (i) belong to the primers extended by the
corresponding phosphate esters, namely, phenyl (IIIa, lanes 3–6),
p-nitrophenyl (IIIb, lanes 7–10), p-nitrophenylethyl (IIId,
lanes 11–14 and 15–17) and β-naphtyl (IIIc, lanes 18–20)
phosphates, respectively (Fig. 2A). β-Naphtyl triphosphate
IIIc showed weak substrate properties in this experiment. We
assign the bands (ii) as belonging to the primer phosphorylated
Figure 3. Primer extension by compounds III–IV. Lane 1, primer + TDTP+
(control); lane 2, as in lane 1 + 10 µM dTTP; lanes 3–5, as in lane 2 + 50, 100
and 300 µM IIIb, respectively; lanes 6–8, as in lane 2 + 50, 100 and 300 µM
IVa, respectively; lanes 9–11, as in lane 2 + 50, 100 and 300 µM IIIc, respectively;
lanes 12–14, as in lane 2 + 50, 100 and 300 µM IVb, respectively.
at the 3′-position, which was confirmed by direct comparison
with the authentic sample VII (Fig. 2B). As seen in Figure 2B,
lane 2, a small amount of product (ii) was also formed when
ddTTP was used as substrate. Evidently, the band intensity was
increased and mobility was decreased with an increase in
hydrophobicity of the incorporated residues. Thus, the
mobility of both modified primers upon gel electrophoresis
corresponded to their structures.
The TDTP+-catalyzed primer extension reaction in the presence
of dTTP and triphosphate derivatives IIIb, IIIc, IVa and IVb
is illustrated in Figure 3. As seen in lanes 3–5 for IIIb and
lanes 9–11 for IIIc, p-nitrophenyl and β-naphtyl phosphate
residues were incorporated both into the initial primer and into
the dTMP-extended oligodeoxynucleotides, which resulted in
the electrophoretic upper- and lower-located bands relative to
those corresponding to the incorporation of the thymidylate
residue (Fig. 3, lanes 3–5, 9–11 and 8 or 14). The locations of
these bands reflected hydrophobicity of the incorporated residues.
Compounds IVa (lanes 6–8) and IVb (lanes 12–14) did not
phosphorylate oligonucleotides, although the dose-dependent
inhibition of the primer extension by dTTP was evident
(compare lanes 6–8 and 12–14 with lane 2).
The incorporation of phosphate and ester phosphate residues
into the oligodeoxynucleotide occur in a different concentrationdependent manner. The formation of the ester phosphorylated
product (i) is parallel to the concentration of IIIb or IIIc,
whereas the amount of phosphorylated product (ii) is changed
in a bell-like mode (compare lanes 3–5 and 9–11). It is worth
noting that formation of product (ii) was not observed with an
Nucleic Acids Research, 2000, Vol. 28, No. 5 1279
increase of the concentration of IIIb,c when the primer was
elongated at least once by a dTMP residue (lanes 5 and 11).
Thus, among the modified triphosphates tested, compound
IIId containing hydrophobic substituent linked through a flexible
ethyl chain showed the best substrate properties. Probably, this
structure best supports the enzyme tuning in substrate binding
to the active center.
Incorporation of a phosphonate group into the
oligodeoxynucleotide primer
The ability of TDT to introduce a phosphate or ester phosphate
group into DNA may be used for non-radioactive 3′-labeling of
DNA, but this approach is limited due to instability of phosphate
esters in biological media. It is well known that phosphonate
groups are good mimics of phosphates, the stability of the P–C
bond being much higher than that of the P–O–C one (15,16). In
addition, 3′-labeling of DNA needs the presence of a reactive
group. With this in mind, we synthesized compounds V
containing an Fmoc-aminoethyl residue, which is bulky and
highly hydrophobic.
As seen in Figure 4A, the compounds Va–c can phosphonylate the primer at the TDT catalysis. The dose-dependent
primer extension in the presence of Va is observed in lanes 2–6.
As a phosphonate donor, compound Vc (lanes 7–9) was more
active than Va, which is supported by the kinetic data (Table 1).
Compound Vb, differing from Vc by the bulky dibromomethylene
fragment, was by two orders of magnitude less active (lanes 10–14).
When the reaction was catalyzed by TDTP+ in the presence
of dTTP (Fig. 4B), phosphonate Va, unlike compound IIId
(lanes 3–5), extended the primer only by its α-phosphonate
residue (lanes 6–8). Because of the high hydrophobicity of
Fmoc-aminoethyl phosphonate attached to the oligodeoxynucleotide, the electrophoretic mobilities of the products were
smaller than those of the respective oligodeoxynucleotides
elongated with dTMP residues. The high substrate efficacy of
Va can be confirmed by the absence of bands corresponding to
unmodified oligodeoxynucleotides (lanes 6–8). An increase in
the concentration of Va from 0.5 to 10 µM results in the
considerable shortening of oligo sets (compare lanes 6–8 and 2).
Table 1. Kinetic parameters of the TDT-catalyzed primer extension by
ddTTP and compounds Va and Vc
Compound
Km (µM)
Vmax/Vmax for ddTTP
ddTTP
0.032 ± 0.001
1.00
Va
0.033 ± 0.003
0.56
Vc
0.009 ± 0.001
0.56
Kinetic monitoring of the reaction (Table 1) showed that Km
and Vmax of Va,c were similar to or lower than that of ddTTP.
It implies that compounds Va,c have higher affinity to the
enzyme and form the productive substrate–enzyme complex
faster than ddTTP. This is in coincidence with the results
obtained for IIId when a compound bearing a flexible and
hydrophobic substituent was the best TDT substrate.
Phosphate incorporation into the DNA primer in the presence of
ddTTP observed in Figure 2B (lane 2) led us to reassess
substrate properties of previously synthesized thymidine 5′-γphosphonyl α,β-diphosphates VI. These nucleotides were
studied in the primer elongation reaction catalyzed by AMV
Figure 4. Primer extension by compounds V catalyzed by TDTP– (Series A)
and TDTP+ (Series B). (A) Lane 1, primer + enzyme (control); lanes 2–6, as in
lane 1 + 0.0005, 0.001, 0.01, 0.1 and 1 µM Va, respectively; lanes 7–9, as in
lane 1 + 0.001, 0.01 and 0.1 µM Vc, respectively; lanes 10–14, as in lane 1 +
0.001, 0.01, 0.1, 1 and 10 µM Vb, respectively. (B) Lane 1, primer + enzyme
(control); lane 2, as in lane 1 + 5 µM dTTP; lanes 3–5, as in lane 2 + 5, 10 and
20 µM IIId, respectively; lanes 6–8, as in lane 2 + 0.5, 2 and 10 µM Va,
respectively; lane 9, as in lane 1 + 2 µM ddTTP.
RT and DNA polymerases α and β (8,17). Figure 5A illustrates
the results of the primer extension by VIa catalyzed by TDTP+
(lanes 4–6). Sets of doublets can easily be seen in each lane,
and the intensities of these doublets are dose-dependent. The
upper bands of the doublets correspond to the incorporation of
a nucleotide residue, whereas lower ones correspond to the
incorporation of a methyl phosphonate fragment. The incorporation of the methyl phosphonate fragment into the 14mer
oligodeoxynucleotide was confirmed by direct comparison
with the synthetic oligodeoxynucleotide VIII (lane 7). It is
worth mentioning that the 3′-phosphorylated primer obtained
in the presence of ddTTP (lane 2), moved somewhat faster than
the methyl phosphonate-bearing primer because of an additional
negative charge (compare the corresponding bands in lane 2
and lanes 4–7). With TDTP–, compound VIa showed similar
results (data not shown).
Figure 5B presents the TDTP– catalyzed primer extension in
the presence of thymidine 5′-(γ-alkylphosphonyl)-α,β-diphosphates (VIb–d). These compounds were involved in the reaction
by both thymidylate and alkylphosphonate residues, although
1280 Nucleic Acids Research, 2000, Vol. 28, No. 5
Figure 6. Primer extension by dTTP, ddTTP, IIId and Va, catalyzed by TDTP–
(Series A, B) and TDTP+ (Series C, D) in the presence of 2 mM Co2+ buffer
(Series A, C) or 10 mM Mn2+ buffer (Series B, D). Lane 1, primer + enzyme
(control); lanes 2–5, as in lane 1 + 5 µM dTTP, 0.1 µM ddTTP, 0.1 µM Va and
0.1 µM IIId, respectively.
in lanes 4–6 (for VIb) and lanes 9–11 (for VIc) one can see the
only set of bands. The incorporation of the azidomethyl phosphonate fragment in the primer was not observed because of
the similarity of mobilities of the extended and the starting
primers (lanes 4–6). However, this incorporation became
evident after the azido group of the azidomethyl phosphonateextended primer was reduced with DTT to an amino group
(lanes 7–8, upper bands). Owing to an additional NH3+ charge,
the mobility of the reduced oligodeoxynucleotide was lowered. The
same result was obtained for VIc (lanes 9–11), when the reduced
oligodeoxynucleotide was clearly discriminated at the electrophoresis pattern (lanes 12–13, upper bands). The mobility of the
primer extended with aminoethyl phosphonate of VId (lanes 14–
17) coincided with those of the products resulted from the reaction
with VIc followed by reduction with DTT (lanes 12–13, upper
bands). Similar results were obtained with compounds VIa–d at the
catalysis of the reaction by TDTP+ (data not shown).
Compounds VIa–d also showed weaker substrate properties
than that of Va,c.
Figure 5. Primer extension by compounds VI catalyzed by TDTP+ (Series A)
and TDTP– (Series B). (A) Lane 1, primer+ enzyme (control); lane 2, as in lane
1 + 0.5 µM ddTTP; lane 3, as in lane 1 + 5 µM dTTP; lanes 4–6, as in lane 1
+ 5, 50 and 500 µM VIa, respectively; lane 7, synthetic 3′-methylphosphonylated oligodeoxynucleotide VIII. (B) Lane 1, primer + enzyme (control); lane
2, as in lane 1 + 0.1 µM ddTTP; lane 3, as in lane 1 + 1 µM ddTTP and 10 µM
dTTP; lanes 4–6, as in lane 1 + 100, 300 and 600 µM VIb, respectively; lanes
7 and 8, as in lane 6, primer extension reaction followed by addition of 0.05
and 0.1 M DTT, respectively; lanes 9–11, as in lane 1 + 100, 300 and 600 µM
VIc, respectively; lanes 12 and 13, as in lane 11, primer extension reaction
followed by addition of 0.05 and 0.1 M DTT, respectively; lanes 14–17, as in
lane 1 + 100, 300, 600 and 1000 µM VId, respectively.
Analysis of TDT
We analyzed TDT preparations of six lots available from
Amersham, Promega and Gibco companies in the reactions of
primer phosphorylation and phosphonylation. Although all of
them catalyzed the transfer reaction of substituted phosphates
and phosphonates approximately to the same degree, they
revealed different properties when transferring unmodified γphosphates. TDT from Promega (lots M187A, M187B) and
Amersham (lot A401-1) catalyzed both reactions, whereas
Nucleic Acids Research, 2000, Vol. 28, No. 5 1281
TDT from Amersham (lots A1601-2, A1801-3) and Gibco (lot
HF3403) could not extend the primer with an unmodified
phosphate under various conditions. Figure 6 demonstrates
these properties for TDTP– (lot A1801-3) and TDTP+ (lot A401-1)
either in the presence of 2 mM Co2+ (standard conditions) or
10 mM Mn2+. In the case of 10 mM Mg2+ or 10 mM Zn2+
buffers both enzymes were inactive in this reaction (data not
shown). As substrates, we used dTTP, ddTTP, Vb and IIId.
Evidently, for both enzymes, the Mn2+ buffer (Series B and D)
was somewhat less effective than the Co2+ buffer (Series A and
C). Under the standard conditions, as well as in the Mn2+ buffer
(Series A and B, respectively), TDTP– incorporates into the
primer only nucleotide residues from dTTP (lane 2) and
ddTTP (lane 3), a phosphonate fragment from Va (lane 4) and
an esterified phosphate residue from IIId (lane 5). In contrast,
in the Co2+ buffer (Series C), TDTP+ incorporates into the
primer nucleotide and phosphate residues of ddTTP (lane 3,
upper and lower bands, respectively), modified α-phosphonate
residue of Va (lane 4), or p-nitrophenylethyl phosphate and
unmodified γ-phosphate residues of IIId (lane 5, upper and
lower bands, respectively). In the Mn2+ buffer no phosphorylation
occurs in the presence of ddTTP and IIId (Series D, lanes 3
and 5). TDTP+ transfer an unmodified phosphate residue into
the primer 3′-terminus only under standard buffer conditions.
The cause of the difference in the properties of the enzyme lots
tested is unclear. The protein electrophoresis of the lots used
demonstrated that TDTP– contains a major band with a molecular
mass of ∼58 kDa [which correlates with the earlier data (18)],
whereas TDTP+ contains two major proteins with molecular
masses of ∼58 and 62 kDa [previously published by Beach et al.
(19)]. Thus, the difference in the phosphorylating activity can
be accounted for by various enzyme composition. However, we
cannot exclude that incorporation of the non-modified γ-phosphate
residue of the substrate analogues (ddTTP, IIIa–d) is induced
by a contaminant presenting in certain batches of TDT.
From the results obtained, it is evident that all TDT samples
catalyzed the transfer of phosphate ester and phosphonate
residues from the corresponding triphosphate derivatives into
the 3′-terminus of the DNA primer. These reactions are
specific towards triphosphates (triphosphonates); neither
corresponding monophosphates nor diphosphates are active in
it. Moreover, these reactions are specific for TDT. Templatedependent DNA polymerases α and β as well as AMV reverse
transcriptase cannot catalyze such transfer (data not shown).
To conclude, the higher the hydrophobicity of the substituent (at
least among the tested compounds), the higher the efficiency of
transfer of a substituted phosphate (or phosphonate) residue.
On the other hand, the higher the substrate properties, the more
pronounced the incorporation of a ‘standard’ α-phosphate (or
phosphonate) fragment into the DNA chain; and the weaker
the substrate, the more probable is the unusual reaction of
incorporation of both α- and γ-residues.
Since the compounds with increased hydrophobicity and
increased stability may be active phosphonate donors in the
cellular media and in cells, they may be used for the investigation
of TDT cellular functions. Modified triphosphates (especially
those containing reporter and ligand groups) may be useful
tools in biotechnological processes.
One can assign the reaction under discussion to the artifactual
activity of this template-independent enzyme largely because
of the use as substrates of so unnatural compounds. It may be a
relevant explanation, indeed. However, it is also possible that
according to the life cycle of primitive lymphoid or leukemic
cells, any triphosphate (dNTP or rNTP) can serve as a phosphate
donor for a molecular target currently unknown, and this reaction
is catalyzed by TDT (20). Moreover, such a fact was recently
found in the process of the TDT-dependent cell apoptosis (20).
In any case, the results obtained could be interesting for
studying the TDT enzymology as well as the cellular biology
of TDT-dependent cells. The compounds of this type with an
increased stability in human blood may be useful as competitors
of the natural substrates of the enzyme in the treatment of
TDT-accompanied leukemia. In addition, to our knowledge, it
is the only known biochemical method for the incorporation of
a phosphate residue (including phosphates bearing reporter or
ligand groups) to a 3′-hydroxyl of oligodeoxynucleotides.
ACKNOWLEDGEMENTS
The authors are grateful to Dr Elena Shirokova and Dr Marina
Kukhanova for valuable discussion. This work was supported by
the Russian Foundation for Basic Research, projects 98-03-32930a
and 99-04-48315; by the Program ‘Leading Scientific
Schools’, grant 96-15-97646; and by the ISTC, grant 1244.
REFERENCES
1. Alt,F. and Baltimore,D. (1982) Proc. Natl Acad. Sci. USA, 79, 4118–4121.
2. Komori,T., Okada,A., Stewart,V. and Alt,F.W. (1993) Science, 261, 1171–1175.
3. Gilfillan,S., Dierich,A., Lemeuur,M., Beniost,C. and Mathis,D. (1993)
Science, 261, 1175–1178.
4. McCaffrey,R.P., Harrison,T.A., Parkman,R. and Baltimore,D. (1975)
N. Engl. J. Med., 292, 775–781.
5. Semizarov,D.G., Victorova,L.S., Dyatkina,N.B., von Janta Lipinsky,M.
and Krayevsky,A.A. (1994) FEBS Lett., 354, 23–26.
6. Dyatkina,N., Semizarov,D., Victorova,L., Krayevsky,A., Theil,F. and
von Janta-Lipinski,M. (1995) Nucl. Nucl., 14, 723–726.
7. Krayevsky,A.A. and Chernov,D.N. (1996) J. Biomol. Struct. Dyn., 14,
225–230.
8. Alexandrova,L.A., Skoblov,A.Yu., Jasko,M.V., Victorova,L.S. and
Krayevsky,A.A. (1998) Nucleic Acids Res., 26, 778–786.
9. Atkinson,T. and Smith,M. (1984) In Gait M.J. (ed.), Oligonucleotide Synthesis.
A Practical Approach. IRL Press Limited, Oxford, UK, pp. 31–85.
10. Krynetskaya,N.F., Zayakina,G.V., Oretskaya,T.S., Volkov,E.M. and
Shabarova,Z.A. (1986) Nucl. Nucl., 5, 33–43.
11. Miller,P., Reddy,M., Muracami,A., Blake,K., Lin,S. and Agris,C. (1986)
Biochemistry, 25, 5092–5097.
12. Mozzherin,D.Yu., Atrazhev,A.M. and Kukhanova,M.K. (1992) Mol. Biol.
Russian, 26, 999–1010.
13. Kolocheva,T.A. and Nevinsky,G.A. (1993) Mol. Biol. Russian, 27, 1368–1379.
14. Sambrook,J., Fritsch,E.E. and Maniatis,T. (1989) Molecular Cloning:
A Laboratory Manual. 2nd Edn. Cold Spring Harbor Laboratory Press,
Cold Spring Harbor, NY.
15. Blackburn,G.M., Taylor,G.E., Tattershall,R.U., Thathcher,G.R.J. and
McLennan,M.G. (1986) In Bruzik,K.S. and Stec,W.I. (eds),
Biophosphates and their Analogues—Synthesis, Structure, Metal and
Activity. Elsevier, Amsterdam, The Netherlands, pp. 451–464.
16. Hamilton,C.J., Roberts,S.M. and Shipitsin,A.V. (1998) Chem. Commun.,
1087–1088.
17. Arzumanov,A.A., Semizarov,D.G., Victorova,L.S., Dyatkina,N.B. and
Krayevsky,A.A. (1996) J. Biol. Chem., 271, 24389–24394.
18. Johnson,D. and Morgan,A.R. (1976) Biochem. Biophys. Res. Commun.,
72, 840–849.
19. Beach,C.M., Chan,S.K., Vanaman,T.C. and Coleman,M.S. (1985)
Biochem. J., 227, 1003–1007.
20. Koc,Y., Urbano,A.G., Sweeney,E.B. and McCaffrey,R. (1996) Leukemia,
10, 1019–1024.