|
|
|
|||||||
|
|||||||||
(Received for publication, September 13, 1994; and in revised form, October 28, 1994) From the
The utilization of dCTP derived from de novo synthesis
through ribonucleotide reductase in exponentially growing CCRF-CEM
cells was compared with the metabolic fate of dCTP produced by the
salvage pathway. Exogenous dCyd was not effectively incorporated into
replicating DNA; instead, dCTP derived from ribonucleotide reductase
(labeled by [5- Replicative DNA synthesis consumes the major portion of cellular
deoxynucleoside triphosphates (dNTPs), whereas other processes such as
repair of damaged DNA and deoxynucleotidyl intermediates of lipid
metabolism probably place lesser demands on the cellular pools.
Evidence from metabolic studies indicates that ribonucleotide
reductase-mediated de novo synthesis of deoxynucleotides is
tightly coupled to replicative DNA synthesis (Nicander and Reichard,
1983; Mathews and Ji, 1992; Reddy and Fager, 1993). Indeed,
multiprotein complexes containing enzymes, including ribonucleotide
reductase, that participate in both dNTP synthesis and DNA replication
have been isolated and characterized (Noguchi et al., 1983;
Harvey and Pearson, 1988; Hammond et al., 1989; Wu et
al., 1994). This concept that the de novo pathway for
dNTP synthesis is linked with DNA replication is further supported by
indications that DNA is replicated at structurally distinct sites in
the nucleus (Nakamura, et al., 1986; Mills et al.,
1989; Cox and Laskey, 1991; Horzak et al., 1993; Coverley and
Laskey, 1994). Kinetic studies of ribonucleoside uptake into the DNA of
whole cells indicate that dNTP pools derived from de novo pathways are likely to be rather small and turn over rapidly
(Nicander and Reichard, 1983). If the dNTP pool utilized for DNA
replication derived from de novo synthesis is localized to the
region surrounding a replication center, it is reasonable to ask
whether dNTPs derived from different sources, such as the salvage
pathways, could also be used for DNA replication. It is clear from
investigations in many cell types that [5- Relatively little
is known of the metabolic fate of dNTPs derived from the salvage of
deoxynucleosides. Evidence exists that salvaged dCyd is utilized in the
formation of thymidylate used in replicative DNA synthesis in cell
lines (Jackson, 1978; Xu and Plunkett 1992). Insight into the
utilization of the dCTP pool derived from the salvage pathway for
purposes other than replicative DNA synthesis was provided by Spyrou
and Reichard(1987, 1989) who demonstrated that dCTP precursors for
deoxyliponucleotide synthesis were derived from salvaged dCyd. Although
some studies have indicated that labeled dCyd may be used to measure
DNA repair (Snyder, 1984; Elliott and Downes, 1986; McKenna and
McKelvey, 1986; Holmberg et al., 1988), it was only recently
suggested that dCTP derived from the salvage pathway might be used as a
specific substrate for DNA repair (Spasokukotskaja et al.,
1992). Following this lead, we sought to identify the source of dCTP
utilized for DNA repair. To this end, we have compared the formation of
dCTP from Cyd and dCyd and have investigated its utilization in human T
lymphoblast CCRF-CEM cells. Our results support the hypothesis that
there are two functionally separate dCTP pools in these cells;
exogenous [5-
Figure 1:
Incorporation
of [5-
Figure 2:
Effect of exogenous dCyd on
[5-
Figure 3:
Incorporation of
[5-
As shown in Fig. 4A, when exponentially growing CCRF-CEM cells were
labeled with [
Figure 4:
Separation of DNA fragments labeled by
[5-
We hypothesize that the
DNA labeled by dCyd that banded at the intermediate density was newly
repaired DNA. Again, this is a relatively low level of incorporation
and probably represents a background level of repair and possibly that
stimulated by the actions of BrdUrd (Ashman et al., 1981;
Shewach et al. 1992). Incorporation of dThd or Cyd into this
peak was too minor to be detected. The overall incorporation of dCyd
was still low compared with either Cyd or dThd in the high density
peak. When both the intermediate density peak and the high density peak
fractions were combined, the DNA specific activity (dpm/UV absorbance
unit) from [
Figure 5:
Effect of BrdUrd on
[5-
Figure 6:
Effects of MMS and UV on
[5-
The functional compartmentation of dCTP derived from the de novo and salvage pathways has important implications for
cellular metabolism. Comparison of the rate of
[5- Our approach involved investigating
incorporation of dCyd and of Cyd into the dCMP of DNA and determining
whether it occurred through DNA replication or DNA repair. The two
pathways were distinguished by differential density labeling;
BrdUrd-labeled replicating DNA banded in CsCl gradients at a high
density, whereas the density of DNA in which BrdUrd was incorporated
during repair was intermediate to replicating DNA and that of parental
DNA without BrdUrd (Fig. 4). Greater than 60% of incorporated
dCyd was found in the intermediate density peak, and so it appeared to
be incorporated through DNA repair. Because some of dCyd identified in
the high density peak was likely due to repair induced by BrdUrd, the
actual proportion of dCyd used for DNA repair was probably even
greater. In contrast, the majority of Cyd was incorporated into the
high density peak of replicating DNA. Due to the much greater amount of
replicating DNA relative to repaired DNA in these growing cells,
estimation of Cyd incorporation into the intermediate density peak was
uncertain, but appeared to be less than 5%. In control cells, the rate
of [5- A compelling body of evidence supports the role of
ribonucleotide reductase as the key enzyme in the functional
compartmentation of dCTP (Moyer and Henderson, 1985; Nguyen and Sadee,
1986; Spyrou and Reichard, 1989; Mathews and Ji, 1992; Reddy and Fager,
1993). It has been calculated that CDP reduction is the most rapid
among the four ribonucleoside diphosphate substrates (Jackson, 1992).
Consistent with its central role in supplying dNTPs for DNA
replication, the activity of this enzyme is known to be elevated during
S phase. Because dCTP pools also increase in S phase cells (Liliemark
and Plunkett, 1986; Arner et al., 1988) it is reasonable to
conclude that ribonucleotide reductase is capable of producing an
excess of dCTP beyond that consumed by DNA replication. In contrast,
the activity of deoxycytidine kinase, the rate-limiting step in the
salvage pathway, is fairly constant throughout the cell cycle
(Liliemark and Plunkett, 1986; Arner et al., 1988). This is
consistent with the notion that dCTP generated by the salvage pathway
is not specifically required in S phase. Thus it is unlikely that dCTP
derived from the salvage pathway would compete effectively with dCTP
from the de novo pathway for incorporation into replicating
DNA. If ribonucleotide reductase were localized near the DNA
replication apparatus or were functionally part of that process, the
preferential utilization of dCTP generated by the de novo pathway could be enhanced. Due to technical limitations, the
cellular location of ribonucleotide reductase has remained uncertain.
Although some cellular fractionation studies (Leeds, et al.,
1985) and investigations using immunocytochemistry (Engstrom and
Rozell, 1988) have suggested that ribonucleotide reductase is a
cytosolic enzyme, there is evidence that it may be associated with the
nuclear membrane (Sikorska et al., 1990). Furthermore, several
laboratories have characterized multienzyme complexes that contain
ribonucleotide reductase and other enzymes involved in dNTP synthesis
and DNA replication (Noguchi et al. 1983; Harvey and Pearson,
1988; Hammond et al., 1989, Reddy and Fager, 1993). A possible
resolution to these apparently contradictory findings could be that
ribonucleotide reductase exists both in a multiprotein DNA replication
complex in the nucleus and as a free enzyme in the cytosol. It is
possible that immunocytochemical methodologies may detect the soluble
enzyme in the cytosol, but perhaps not a complexed form in the nucleus
due to blockage of the epitope by other protein components. Evidence
from recent studies demonstrating that DNA replication forks are
arranged in defined spatial patterns within the nucleus provides a
structural context for DNA replication in which functionally
compartmentalized dNTP pools may be a central component. It is now
known that replication forks are tightly clustered in foci within the
nucleus (Nakamura, et al., 1986; Mills et al., 1989;
Cox and Laskey, 1991; Hozak et al., 1993; Coverley and Laskey,
1994). This arrangement is likely to facilitate rapid consumption of
large amounts of dNTPs in each replication focus. It is doubtful that
this rate of DNA synthesis can be supported by the relatively low
concentration of dNTPs estimated assuming that the dNTPs are uniformly
distributed in total cell water. In contrast to the focal nature of
DNA replication, it is likely that DNA repair in response to alkylating
agents or radiation is dispersed throughout the genome, possibly with
foci at transcriptionally active sites (Jackson et al., 1994).
In this situation, DNA synthesis associated with nucleotide excision
repair probably utilizes dCTP from a more general pool. We envision
that this pool is composed of a mixture of dCTP from the salvage
pathway and dCTP which has escaped from its source at ribonucleotide
reductase in the replicating foci. In the process of diffusion, its
concentration has been diluted relative to that within the replicating
focus. When considering the utilization of exogenous dCyd, we propose
that the salvage pathway contributes to this generalized pool of dCTP,
which our experiments have demonstrated comprises the bulk of the
nucleotide used for DNA repair, but makes only a small contribution to
replicative DNA synthesis. The relatively minor utilization of
exogenous dCyd in replicating DNA should not be viewed as contradictory
to the fact that a number of dCyd analogs, such as arabinosylcytosine,
2`,2`-difluoro-2`-deoxycytidine, 5-aza-2`-deoxycytidine, and
2`,3`-dideoxycytidine require the same salvage pathway for activation
prior to incorporation into DNA (Major et al., 1981;
Momparler, 1985; Huang et al., 1991, Starnes and Cheng, 1987).
The extent to which these analogs are incorporated as a result of
repair synthesis relative to DNA replication is unknown, although it
has been assumed that the latter pathway is utilized in the absence of
DNA damaging stimuli. Although the rate of exogenous dCyd incorporation
into replicating DNA (0.36 pmol/10 In
summary, it appears most of the dCTP generated by ribonucleotide
reductase goes to a high throughput, low volume pool; some of this dCTP
appears to become available to a more general pool in the nucleus. The
salvage pathway also contributes dCTP and cytidine nucleotide analogs
to this general pool. One of the functions of this pool in CCRF-CEM
cells is to supply dCTP for repair of DNA. It is possible that this
functional compartmentation could be used to advantage in chemotherapy
by combining cytidine nucleotide analogs with agents and modalities
that evoke a DNA repair response in tumor cells. Recent reports suggest
the utility of this approach (Gregoire et al., 1994; Shewach et al., 1994).
Volume 270,
Number 2,
Issue of January 13, 1995 pp. 631-637
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
PREFERENTIAL UTILIZATION OF SALVAGED DEOXYCYTIDINE FOR DNA REPAIR
IN HUMAN LYMPHOBLASTS (*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
H]Cyd) was the main precursor for
that purpose, apparently because of functional compartmentation of the
dCTP pool in these cells. Studies of the metabolic route of
incorporation of exogenous [5-H]dCyd into DNA of
growing CCRF-CEM cells demonstrated that it was mainly incorporated
through the DNA repair pathway. Incorporation of
[5-H]dCyd into DNA of synchronized cell
populations was maximal in G
cells, whereas
[H]dThd incorporation occurred predominantly in S phase cells. When cellular DNA was density labeled by
incubation with BrdUrd, repaired DNA, which was less dense than
replicated DNA, was preferentially labeled by
[5-H]dCyd. In contrast, replicated DNA was
labeled by both [H]dThd and
[5-H]Cyd. The DNA-damaging agents
methylmethanesulfonate, ultraviolet irradiation, and -irradiation
inhibited [
H]dThd incorporation, whereas they
stimulated the accumulation of [5-H]dCyd in DNA.
Based on these results, we propose that the dCTP pool is functionally
compartmentalized in growing CCRF-CEM cells. dCTP derived from the
salvage pathway is utilized predominantly for DNA repair, whereas the de novo pathway supplies dCTP for DNA replication.
H]dCyd
is a poor substrate for labeling replicating DNA (Plagemann et
al., 1978; Cohen et al., 1983; Karle et al.,
1983; Nicander and Reichard, 1983; Balzarini et al., 1984;
Taljanidisz et al., 1986; Leeds and Mathews, 1987;
Sasvari-Szekely et al., 1989; Xu and Plunkett, 1992; 1993).
This stands in contrast to [5-H]Cyd which
specifically labels dCMP in replicating DNA and to
[H]dThd, which has been taken as the standard for
measuring this process (Reichard, 1988). Incorporation of
[5-H]dCyd and [H]dThd
become comparable only in cells that are not active in DNA replication
(Holmberg et al., 1988) or when de novo synthesis of
dCDP is inhibited (Nicander and Reichard, 1983).H]dCyd labels one dCTP pool, and one
is preferentially used for DNA repair.
Materials
[5-H]dCyd (26
Ci/mmol), [5-H]Cyd (26 Ci/mmol),
[methyl-H]dThd (49 Ci/mmol), and
[8-H]dATP (21.2 Ci/mmol) were obtained from ICN
Biomedicals Inc. (Irvine, CA). Aphidicolin, calf thymus DNA,
methanesulfonic acid methyl ester (MMS), ()BrdUrd, dCyd,
dCTP, and dTTP were obtained from Sigma. DNA polymerase I was a product
of Boehringer Mannheim Corp. Oligonucleotide templates for the assay of
cellular deoxynucleoside triphosphate levels (Sherman and Fyfe, 1989)
were synthesized by Genosys Biotechnologies, Inc. (The Woodlands, TX).Cell Culture and Synchronization
Human lymphoblast
CCRF-CEM cells (Foley et al., 1965) were maintained in RPMI
1640 medium (Life Technologies, Inc.) supplemented with 10% fetal
bovine serum (Life Technologies, Inc.). During exponential growth, the
population doubling time was 22 h. Cells were routinely checked for
contamination with mycoplasma by the American Type Culture Collection
and were consistently negative. Cell number and mean volume were
determined by a Coulter counter (model ZM, Coulter Electronics,
Hialeah, FL). To obtain synchronized cells, a double aphidicolin (2
µM) block method was applied as described by Matherly et al.(1989). The cell cycle phase distribution was determined
by flow cytometry after the cells were treated with pepsin and RNase
and stained with propidium iodide.Determination of Specific Activities of
[5-
Cells in exponential growth were collected and resuspended
in 3 ml of fresh culture medium (0.3-1.0 H]dCTP and [H]dTTP
Pools
10
/ml). Dialyzed fetal bovine serum was used for these
experiments. Cells were labeled with trace amounts of
[5-H]dCyd or [H]dThd (0.2
µCi/ml) and collected by centrifugation at 1500 revolutions/min for
5 min. Nucleotides were extracted in 0.4 N HClO
at
4 °C for 15 min. Following centrifugation the supernatant was
neutralized with KOH, KClO was removed by centrifugation,
and the resulting supernatant was stored at -20 °C. High
pressure liquid chromatographic separation and quantitation of
nucleotides was carried out using a Partisil 10-SAX column (250 mm
4 mm; Whatman Inc., Clifton, NJ) with a gradient generated by
NHH
PO buffer as described
previously (Xu and Plunkett, 1992). The radioactivity associated with
[5-H]dCTP and [H]dTTP was
quantitated with a radioactive flow detector (model A250, Packard
Instrument Co., Meriden, CT). The amounts of dCTP and dTTP in extracts
were quantitated using a DNA polymerase assay with synthetic
oligonucleotide template/primers, as described by Sherman and
Fyfe(1989); determinations were always made in duplicate or triplicate.
The specific activity of each dNTP was expressed as
disintegrations/minute/micromole.Measurement of Incorporation of
[5-
CCRF-CEM cells (3
H]dCyd, [H]dThd, and
[5-H]Cyd into DNA 10
in 10 ml) in exponential growth phase were
incubated with either [H]dThd or
[5-H]dCyd (0.2 µCi/ml) for 60 min. The
radioactivity incorporated into HClO-insoluble material was
determined using 25-mm glass fiber discs as described previously (Huang et al., 1990). To quantitate DNA synthesis through the de
novo pathway, 2 10 cells in 10 ml were
incubated with 0.2 µCi/ml [5-H]Cyd for 30
min. Cells were then extracted with 0.4 N HClO.
The acid-insoluble pellet was collected by centrifugation and washed
once with 0.4 N HClO. RNA was removed by
incubation with 0.5 N NaOH at 37 °C for 2 h in the
presence of added calf thymus DNA (0.6 mg/ml). DNA was then
precipitated by adding HClO to a final concentration of
0.75 N. The pellet was collected using centrifugation and
washed twice with 0.4 N HClO. This procedure
eliminated greater than 99% of RNA and recovered more than 95% of the
DNA in the pellet, as demonstrated by [H]Urd and
[H]dThd labeling in control experiments.
Radioactivity was quantitated after the pellet was solubilized with 2 N NaOH at 70 °C for 16 h. Use of
[5-H]dCyd and [5-H]Cyd
assured that the dUMP derived from these nucleosides lost its tritium
at the thymidylate synthase reaction and did not label dTMP in DNA (Xu
and Plunkett, 1993).Separation of Repaired DNA from Replicated
DNA
CCRF-CEM cells were labeled with
[5-H]dCyd, [H]dThd, or
[5-H]Cyd in the presence of 1 µM BrdUrd to density label DNA, which after alkaline CsCl
centrifugation was separated into parental DNA (low density), newly
replicated DNA (high density), and repaired DNA (intermediate density)
according to Ball and Roberts(1971). DNA was isolated as described
previously (Huang and Plunkett, 1992), sheared by eight passes through
a 20-gauge needle, which produced DNA fragments with an average length
of 15 kilobases. The DNA was banded by alkaline CsCl centrifugation as
described by Smith et al.(1981) with minor modifications. Each
centrifugation tube contained a total volume of 13 ml. Alkaline CsCl
density centrifugation was carried out with a Beckman Ti75 rotor at
37,000 revolutions/min at 25 °C for 90 h. Fractions of 0.2 ml were
aspirated from the top of the gradient with a Densiflow IIC apparatus
(Buchler Instruments, Inc., Fort Lee, NJ) and diluted with
HO for measurements of radioactivity by liquid
scintillation counting and of UV absorbance at 260 nm.Effects of DNA-damaging Agents on
[
The differential utilization of
[H]dThd and [5-H]dCyd
IncorporationH]dThd and [5-H]dCyd for
DNA semiconservative replication and DNA repair was studied after cells
were treated with the DNA-damaging agents MMS, UV, or irradiation.
Cells were incubated with the indicated concentrations of MMS for 16 h;
UV treatment at 254 nm or
-irradiation with a
Cs
source (at 0 °C) was carried out immediately before
[5-
H]dCyd or [H]dThd was
added to the cultures. Incorporation of [5-H]dCyd
and [H]dThd into DNA was quantitated after cells
were incubated with these nucleosides for 60 min as described above.
Incorporation of [5-
To
study functions of salvaged dCyd and dThd as precursors of DNA
synthesis, [5-H]dCyd and
[H]dThd into DNA in CCRF-CEM CellsH]dCyd and
[H]dThd were added to exponentially growing
CCRF-CEM cells, and the relationship between the specific activities of
the corresponding triphosphates and the rates of incorporation into DNA
was determined. The size of each pool was undisturbed throughout the
experiment; the dCTP pool was 24 ± 4 pmol/10 cells,
and the dTTP pool was 81 ± 9 pmol/10 cells (mean
± S.D.). The specific activity of the
[H]dTTP pool declined steadily after labeling
with [H]dThd, whereas the specific activity of
the [5-H]dCTP pool remained essentially constant
for 2.5 h after labeling with [5-H]dCyd (Fig. 1A). The initial rate of accumulation of
radioactivity in the acid-insoluble material from
[H]dThd was more rapid than from
[5-H]dCyd (Fig. 1B). Based on
average pool-specific activities of 370 ± 22 dpm/pmol for
[5-H]dCTP and 257 ± 13 dpm/pmol for
[H]dTTP 30-50 min after addition of labeled
nucleoside, the rate of incorporation was 7.8 and 0.36 pmol/10 cells/min for [H]dThd and
[5-H]dCyd, respectively. In fact, the rate
obtained from [H]dThd may underestimate the true
rate of DNA synthesis in these cells, due to the decline in
[H]dTTP pool specific activity of 8% during this
period (Fig. 1A). Our results agree with those in
studies of other cell lines, which demonstrated that the rate of dCyd
incorporation into DNA is slower than that of dThd, indicating that
salvaged dThd, but not dCyd, is readily accessible for use in DNA
replication (Plagemann et al., 1978; Karle et al.,
1983; Nicander and Reichard, 1983; Taljanidisz et al., 1986;
Leeds and Mathews, 1987; Sasvari-Szekely et al., 1989).
H]dCyd and [H]dThd in
exponentially growing CCRF-CEM cells. CCRF-CEM cells were labeled with
0.2 µCi/mL [5-H]dCyd (
) or
[H]dThd (
) to quantitate the rate of
incorporation of each into DNA. Aliquots of cells were withdrawn at
indicated times and were subjected to HClO extraction. The
specific activity of [5-H]dCTP and
[H]dTTP pools of HClO-soluble
extracts (A) and the radioactivity incorporated into
HClO-insoluble material (B) were determined as
described in ``Experimental Procedures.'' The data represent
the mean ± S.D. of three
determinations.
Effect of Exogenous dCyd on
[5-
To investigate
whether exogenous dCyd interfered with the incorporation of
[5-H]Cyd IncorporationH]Cyd into DNA through the de novo pathway regulated by ribonucleotide reductase, cells were
preincubated with different concentrations of dCyd for 2 h before
labeling with [5-H]Cyd for 30 min. The cellular
dCTP pool was elevated by dCyd in a concentration-dependent manner (not
shown), in agreement with our previous studies (Heinemann and Plunkett,
1989; Xu and Plunkett, 1993). This reduced the specific activity of the
[H]dCTP pool to less than one-third of control
cells (Fig. 2). However, addition of dCyd did not affect the
incorporation of [5-H]Cyd into DNA. These results
indicate that the expanded dCTP pool derived from salvaged dCyd did not
mix with the dCTP pool used for DNA synthesis, which was derived from
the de novo pathway via ribonucleotide reductase.
H]Cyd incorporation into DNA. Exponentially
growing CCRF-CEM cells were preincubated with indicated concentrations
of dCyd for 2 h before labeling with [5-H]Cyd
(0.2 µCi/ml) for 30 min. Cells were then subjected to HClO extraction. The specific activity of
[5-H]dCTP pool () and the radioactivity
incorporated into DNA () were quantitated as described under
``Experimental Procedures.''
Incorporation of [5-
DNA repair occurs in all phases of the cell cycle (Russev
and Boulikas, 1992; Kaufmann and Kaufman, 1993). A repair-specific
substrate should demonstrate a different pattern of incorporation into
DNA during the cell cycle from that of a precursor that is incorporated
predominantly through S phase DNA replication. To test the
hypothesis that dCyd may supply a dCTP pool that is specifically
utilized for DNA repair, we compared the incorporation into DNA of
[5-H]dCyd and
[H]dThd into Synchronized CCRF-CEM
CellsH]dCyd and [H]dThd in
different phases of cell cycle. A double aphidicolin block method was
used to synchronize CCRF-CEM cells in the late G phase
(Matherly et al., 1989). As shown in Fig. 3,
[5-H]dCyd incorporation peaked 2 h after cells
were released from the aphidicolin arrest, decreased as cells entered
mid-S phase, and reached a second, lower peak between 12 and
18 h, as cells moved on to G and M phases. In contrast,
[H]dThd incorporation reached its maximum
4-6 h after cells were released from aphidicolin arrest, when
most cells of the population were in S phase. These results are
consistent with the hypothesis that [5-H]dCyd was
not used as a substrate of DNA replication. Rather, we suspect that
dCyd incorporation measures spontaneous DNA repair (Greer and Kaplan,
1986; Lindahl, 1993; Xiao and Samson, 1993) and possibly repair induced
by the synchronizing treatment with aphidicolin.
H]dCyd and [H]dThd in
synchronized CCRF-CEM cells. Cells were synchronized by double
aphidicolin block as described under ``Experimental
Procdures.'' After cells were released from the second aphidicolin
treatment, aliquots of cells were withdrawn at indicated times and were
labeled with either [5-H]dCyd () or
[H]dThd () for 10 min. The radioactivity
incorporated into HClO-insoluble material was quantitated
by liquid scintilation counting. Cell number was determined separately
at the time of assay. The data were expressed as dpm/10 cells. Flow cytometry analysis demonstrated that about 70% of the
cells entered mid-S phase between 4 and 6
h.
Incorporation of [5-
To differentiate the functional
utilization of exogenous dCyd from that of Cyd and dThd, the
incorporation of these nucleosides into DNA was investigated under
experimental conditions that would distinguish between replicative and
repair synthesis of DNA. Cellular DNA was prelabeled with radioactive
nucleosides, followed by density labeling with BrdUrd. After shearing
and separation by alkaline density gradient centrifugation, DNA
replicated during the BrdUrd incubation would band in alkaline CsCl at
a higher density than parental DNA lacking BrdUrd. DNA containing the
low level of BrdUrd incorporated during repair patch synthesis would be
expected to band at an intermediate density.H]dCyd,
[5-H]Cyd, and [H]dThd into
Replicated and Repaired DNAH]dThd for 39 h without BrdUrd and
an additional 13 h in the presence of 1 µM BrdUrd, a
bimodal distribution of [H]dThd-labeled DNA was
observed. Radioactivity was associated with a UV-absorbing peak that
banded at a high density (fractions 32-45) and was also
incorporated into a peak of lesser density (fractions 5-15).
These peaks represent newly replicated and unreplicated DNA after
BrdUrd addition, respectively. The higher density peak shows less UV
absorbance due to the effects of the BrdUrd treatment on DNA
replication. When [H]dThd labeling and BrdUrd
incorporation were carried out simultaneously for half of a cell cycle,
only the high density peak was labeled (Fig. 4B). On
the other hand, when cells were incubated with
[5-H]dCyd, radioactivity was incorporated into a
peak of intermediate density (Fig. 4C, fractions
16-25) in addition to incorporation into the high density peak.
As was the case for dThd, when cells were incubated with
[5-H]Cyd, only the high density peak was labeled (Fig. 4D), indicating that both
[H]dThd and [5-H]Cyd served
mainly as precursors for DNA replication.
H]dCyd and [H]dThd by
alkaline CsCl gradient centrifugation. Each experiment was carried out
with 2 10 exponentially growing CCRF-CEM cells
suspended in 60 ml of media. A, cells were labeled with 1
µM [H]dThd (0.2 µCi/ml) for 39
h. BrdUrd (1 µM) and fresh [H]dThd
were added for the last 13 h of the incubation. B, cells were
incubated with 1 µM [H]dThd (0.2
µCi/ml) for 13 h in the presence of 1 µM BrdUrd. C, cells were labeled with tracer amounts of
[5-H]dCyd (0.2 µCi/ml) for 13 h in the
presence of 1 µM BrdUrd. D, cells were labeled
with tracer amounts of [5-H]Cyd (0.2 µCi/ml)
for 13 h in the presence of 1 µM BrdUrd. Isolation of DNA
and CsCl centrifugation were performed as described under
``Experimental Procdures.'' Fractions of 0.2 ml were
collected and diluted for the measurement of radioactivity () and
UV absorption at 260 nm ().
H]dThd and
[5-H]Cyd labeling was 23- and 13-fold higher,
respectively, than that labeled by [5-H]dCyd.
These results suggest the existence of functionally compartmentalized
dCTP pools for DNA replication and DNA repair in CCRF-CEM cells. The
portion of [5-H]dCyd that was incorporated into
the high density peak may represent the upper limit of mixing of
salvage pathway products with the dCTP pool used in replication. It is
also possible that some DNA repair was taking place in the newly
replicated DNA, probably because of the BrdUrd, which itself evokes DNA
repair processes (Hopkins and Goodman, 1980; Shewach et al.,
1992).BrdUrd-enhanced [5-
We expected the BrdUrd incorporated into
newly replicated DNA in the experiments described above to induce DNA
repair activity because BrdUrd is a mutagen. The DNA fragments that
contain BrdUrd incorporated during both DNA replication and DNA repair
are of high density and cannot be distinguished from DNA replication
fragments using the density sedimentation technique. In agreement with
this view, we found that incorporation of
[5-H]dCyd
Incorporation into DNAH]dCyd in growing CCRF-CEM cells was indeed
enhanced by BrdUrd in a concentration-dependent manner (Fig. 5).
To exclude the possibility that the enhanced
[5-H]dCyd incorporation was due to lowered
cellular dCTP levels (Meuth and Green, 1974; Ashman et al.,
1981; Shewach et al., 1992), we added dCyd to restore the dCTP
pool before [5-H]dCyd labeling. This treatment,
as shown in Fig. 5, did not change the stimulatory effect of
BrdUrd on [5-H]dCyd incorporation. This result
was consistent with the likelihood that at least part of the
[5-H]dCyd incorporation into the high density
peak (Fig. 4) was due to repair rather than replication.
H]dCyd incorporation into DNA. CCRF-CEM cells
were incubated with BrdUrd at the indicated concentrations for 16 h.
Cells were then incubated without () or with () 1 µM non-radioactive dCyd for additional 2 h before they were labeled
with [5-H]dCyd for 60 min. The radioactivity
incorporated into DNA was quantitated as described under
``Experimental Procedures.''
Differential Actions of DNA-damaging Agents on
[
To further investigate the possibility
that dCyd may serve specifically as the precursor of DNA repair, we
compared the incorporation of [5-H]dThd and [5-H]dCyd
Incorporation into DNAH]dCyd and
[H]dThd into DNA after cells were treated with
MMS, UV, or -irradiation. As shown in Table 1, these agents
had opposite effects on [5-
H]dCyd and
[H]dThd incorporation; they increased the
incorporation of [5-H]dCyd and decreased the
incorporation of [H]dThd. The lowered
[H]dThd incorporation is consistent with the
inhibition of semiconservative DNA replication by DNA damaging agents
(Swenson and Setlow, 1966; Friedberg, 1985). The good correlation
between treatment with DNA damaging agents and enhancement of
[5-H]dCyd incorporation is consistent with the
hypothesis that the dCTP pool labeled by salvaged
[5-H]dCyd is used for DNA repair but, as
demonstrated in earlier experiments, not for DNA replication.
Effect of DNA-damaging Agents on Labeling of Repaired DNA
by [5-
Density labeling experiments
were conducted to distinguish between utilization of salvaged dCyd for
either replicative or repair DNA synthesis after DNA damage-inducing
treatments. Cells were either incubated with MMS or treated with UV
before their DNA was density labeled with BrdUrd during incubation with
either [5-H]dCydH]dCyd or
[H]dThd. Fig. 6shows representative
profiles of radioactivity associated with sheared DNA after separation
in alkaline CsCl gradients. Labeling of the high density peak
(fractions 35 45) by [H]dThd was decreased to
only 14% of controls in the presence of 0.3 mM MMS (Fig. 6A). In contrast, MMS stimulated
[5-H]dCyd labeling of the intermediate density
peak (fractions 20 33) by 3-fold and the high density peak by 60% (Fig. 6B). Incubations with 0.01-1 mM MMS demonstrated that this effect was concentration dependent up
to 0.3 mM MMS; 1 mM MMS appeared to be too toxic,
producing visible cell lysis (data not shown). A similar but lesser
effect was observed with UV illumination; the 6 J/m dose
increased dCyd incorporation in the intermediate density peak by 54%,
whereas dThd incorporation was decreased to 87% of controls. The
different magnitude of stimulation of dCyd into repaired DNA was due in
part to the fact that MMS was present throughout the labeling, whereas
UV was used only before [5-H]dCyd addition,
perhaps giving UV-treated cells sufficient time to complete DNA repair
early in the 13-h incubation period.
H]dCyd and [H]dThd
incorporation into DNA fragments analyzed by CsCl gradient
centrifugation. Exponentially growing CCRF-CEM cells were treated
without (), or with 6 J/m UV (), or with 0.3
mM MMS (
). Cells were then labeled with
[H]dThd (A) or with
[5-H]dCyd (B) in the presence of 1
µM BrdUrd for 13 h as described in Fig. 4. For
MMS-treated samples, labeling of [H]dThd or
[5-H]dCyd was carried out in the presence of MMS.
Isolation of DNA and CsCl centrifugation were performed as described
under ``Experimental
Procedures.''
H]Cyd incorporation into DNA with that of
[5-H]dCyd indicates that Cyd is efficiently used
for DNA replication in exponentially growing cells, whereas dCyd is a
relatively poor precursor for this purpose. On the other hand, it is
clear that cells utilize salvaged dCyd for the synthesis of dTTP
through the dCMP deaminase pathway (Jackson, 1978; Xu and Plunkett,
1992) and for the synthesis of deoxyliponucleotides (Spyrou and
Reichard, 1987, 1989). The origin of deoxynucleotides used for
mitochondrial DNA synthesis may be derived from a distinct
ribonucleotide reductase associated with this organelle (Young et
al., 1994). The present study provides evidence that
[5-H] dCyd was used selectively as a precursor of
DNA repair in exponentially growing CCRF-CEM cells. Induction of DNA
repair in growing cells with DNA-damaging agents enhanced the
incorporation of [5-H]dCyd into DNA, whereas such
treatments disrupted DNA replication (Fig. 6). Furthermore,
density labeling experiments demonstrated that DNA undergoing repair
was specifically labeled with [5-H]dCyd ( Fig. 4and Fig. 6).H]Cyd incorporation (59 pmol/10 cells/min, Fig. 2) was significantly greater than the rate
of dThd incorporation (7.8 pmol/10 cells/min, Fig. 1). Assuming that dThd incorporation represents the true
rate of DNA replication (Nicander and Reichard, 1983; Reichard, 1988),
the apparent greater rate of [5-H]Cyd
incorporation may be attributed to functional compartmentation of dCTP.
These calculations were based on the average cellular pool-specific
activities; because ribonucleotide reductase activity varies with the
cell cycle (Eriksson et al., 1984) whereas the activity of
dCyd kinase is relatively stable (Liliemark and Plunkett, 1986; Arner et al., 1988), it is possible that the specific activity of
[H]dCTP depends on cell cycle stage and also on
cell type. For example, Chinese hamster ovary cells exhibit a 10-fold
difference in [H]dCTP-specific activity between
G and S phase populations (Leeds and Mathews, 1987); in
contrast, the [H]dCTP-specific activity in
exponentially growing CEM cells was only 60% greater than in S phase
cells (Xu and Plunkett, 1993). Additionally, a compartmentalized dCTP
pool used for DNA replication could have had a much higher specific
activity. If so, the calculated rate of [5-H]Cyd
incorporation into DNA would be decreased to a value comparable to that
of [H]dThd incorporation. Although the average
cellular [H]dCTP-specific activity decreased in
cells that were preincubated with dCyd (Fig. 2), the rate of
[5-H]Cyd incorporation into DNA was apparently
unchanged. This suggests that the de novo metabolic route via
ribonucleotide reductase is largely restricted from mixing with the
dCTP pool generated by the salvage pathway. Furthermore, using
synchronized cells, we found that the peak of
[5-H]dCyd incorporation did not coincide with S
phase DNA replication (Fig. 3), another indication that the dCTP
pool labeled by [5-H]dCyd was excluded from DNA
replication. cells/min) is about 5%
of that of dThd incorporation (Fig. 1), the rate of
2`,2`-difluoro-2`-deoxycytidine incorporation was even less than that
of dCyd (Huang et al., 1991). On the other hand, in comparison
with dCyd, there are factors that favor the incorporation of analogs
via the replication pathway. For example, cytidine nucleotide analogs
accumulate to greater cellular concentrations than does dCTP (Plunkett
and Gandhi, 1993), and the presence of the nucleotides of these drugs
in the cell is prolonged because they are less likely to be eliminated
by dCMP deaminase (Momparler, 1985; Heinemann et al., 1992).
Thus, the incorporation of cytidine analogs may be taken as an
indication of the upper limit of the mixing that occurs between the
general dCTP pool generated by the salvage pathway with the dCTP pool
produced by the de novo pathway for DNA replication.
)
We gratefully acknowledge the expert editorial
assistance of Walter Pagel in the preparation of this manuscript.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Cardoen, E. Van Den Neste, C. Smal, J.-F. Rosier, A. Delacauw, A. Ferrant, G. Van den Berghe, and F. Bontemps Resistance to 2-Chloro-2'-deoxyadenosine of the Human B-cell Leukemia Cell Line EHEB Clin. Cancer Res., November 1, 2001; 7(11): 3559 - 3566. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Taub, X. Huang, Y. Ge, J. A. Dutcher, M. L. Stout, R. M. Mohammad, Y. Ravindranath, and L. H. Matherly Cystathionine-{beta}-Synthase cDNA Transfection Alters the Sensitivity and Metabolism of 1-{beta}-D-Arabinofuranosylcytosine in CCRF-CEM Leukemia Cells in Vitro and in Vivo: A Model of Leukemia in Down Syndrome Cancer Res., November 1, 2000; 60(22): 6421 - 6426. [Abstract] [Full Text]
|
|
|
|
 |
|
 T. G. Petrakis, E. Ktistaki, L. Wang, S. Eriksson, and I. Talianidis Cloning and Characterization of Mouse Deoxyguanosine Kinase. EVIDENCE FOR A CYTOPLASMIC ISOFORM J. Biol. Chem., August 27, 1999; 274(35): 24726 - 24730. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Lall and R. L. Davidson Sequence-Directed Base Mispairing in Human Oncogenes Mol. Cell. Biol., August 1, 1998; 18(8): 4659 - 4669. [Abstract] [Full Text]
|
|
|
|
 |
|
 M. Johansson, S. Brismar, and A. Karlsson Human deoxycytidine kinase is located in the cell nucleus PNAS, October 28, 1997; 94(22): 11941 - 11945. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Iwasaki, P. Huang, M. J. Keating, and W. Plunkett Differential Incorporation of Ara-C, Gemcitabine, and Fludarabine Into Replicating and Repairing DNA in Proliferating Human Leukemia Cells Blood, July 1, 1997; 90(1): 270 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Bianchi, S. Borella, C. Rampazzo, P. Ferraro, F. Calderazzo, L. C. Bianchi, S. Skog, and P. Reichard Cell Cycle-dependent Metabolism of Pyrimidine Deoxynucleoside Triphosphates in CEM Cells J. Biol. Chem., June 27, 1997; 272(26): 16118 - 16124. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |