(CANCER RESEARCH 48, 1850-1855, April 1. 1988]
Effects of Antileukemia Agents on Nuclear Matrix-bound DNA Replication in
CCRF-CEM Leukemia Cells1
Daniel J. Fernamtes,2 Cecil Smith-Nanni, Melanie T. Paff, and Toni-Ann M. Neff
Department of Biochemistry, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, North Carolina 27103
ABSTRACT
The effects of various antileukemic agents on DNA replication asso
ciated with the nuclear matrix were investigated in CCRF-CEM leukemia
cells. Residual nuclear matrices were prepared by sequential treatment
of nuclei with 1.5 M NaCI, DNase I, and Triton X-100 and contained I
5, 10, and 37% of the total nuclear DNA, protein, and phospholipid,
respectively. In control cells pulse-labeled for 45 s with |'I I|thymidine,
the specific activity of nascent DNA was four-fold greater in the nuclear
matrix fraction relative to the specific activity of the high salt-soluble
(nonmatrix) DNA fraction. Pulse-labeling and reconstitution experiments
indicated that this enrichment of newly replicated DNA on the nuclear
matrix did not result from aggregation of nascent DNA with the matrix.
A 2-h incubation of tumor cells with either 0.1 MMteniposide (VM-26),
0.2 /«M
VM-26, or 0.5 MMamsacrine (rn-AMSA) reduced the relative
specific activity of nascent DNA on the nuclear matrix by 59, 61, and
54%, respectively, compared to control cells. In contrast hydroxyurea
and c\ tosine arabinoside, at concentrations that markedly inhibited total
nuclear DNA synthesis, did not decrease the relative specific activity of
newly replicated DNA on the matrix. The results provide evidence that
the unt¡proliférât
¡vu
effects of the DNA topoisomerase II inhibitors, VM26 and m-AMSA, are localized on the nuclear matrix of CCRF-CEM
leukemia cells.
INTRODUCTION
Studies on the modes of action of anticancer agents have
revealed that many of the clinically important agents inhibit
tumor cell growth by interfering with DNA replication. How
ever, the actual mechanisms by which anticancer agents inhibit
the synthesis of DNA have often been difficult to define, partly
because of the incomplete knowledge of the complex processes
involved in eukaryotic DNA replication. Recent evidence indi
cated that the nuclear matrix of eukaryotic cells has an impor
tant role in chromatin organization and various nuclear func
tions including DNA replication (1-8). Nuclear matrix proteins
have been shown to bind tightly to DNA and to topologically
constrain the interphase chromatin as supercoiled loops,
thereby organizing the DNA into discrete replicating units (46, 9, 10). In addition, the nuclear matrix also contains specific
binding sites for some of the enzymes involved in DNA repli
cation including DNA polymerase a (11-15), DNA primase
(16, 17), and DNA topoisomerase II (5, 18). The possibility
that these matrix-bound enzymes function within the nucleus
as DNA replication complexes was supported by the observa
tions that nuclear matrices from various types of cells were
highly enriched in DNA replication forks and nascent DNA (2,
13, 19, 20). Moreover, isolated nuclear matrices were capable
of carrying out DNA polymerase a-directed in vitro DNA
synthesis in the absence of an exogenous DNA template (1113).
Previous studies from this laboratory suggest that certain
Received 7/14/87; revised 12/14/87; accepted 12/31/87.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Supported by USPHS Grant CA-44597 and by Grant CH-226 from the
American Cancer Society.
2 Scholar of the Leukemia Society of America, Inc., and to whom requests for
reprints should be addressed.
pyrimidine antagonists interfere with functional interactions
among DNA and the enzymes involved in DNA replication
(21). Because of the importance of the nuclear matrix to DNA
replication, we decided to morphologically and biochemically
characterize the nuclear matrix from CCRF-CEM leukemic
cells. This report describes these results, as well as the effects
of certain antileukemic agents on nuclear matrix-bound DNA
replication.
MATERIALS AND METHODS
Materials. The human CCRF-CEM leukemic cell line was provided
by Dr. Joseph R. Berlino of the Yale University School of Medicine,
and was propagated at 37°Cunder 95% air-5% CO2 in Fischer's
medium supplemented with 10% heat-inactivated horse serum, penicil
lin (20,000 units/liter), and streptomycin (20 mg/liter). SeaKem GTG
agarose was purchased from FMC Corp., Rockland, ME. Tissue culture
medium, serum, and antibiotics were obtained from Grand Island
Biological Company, Grand Island, NY. Cells were checked periodi
cally for Mycoplasma contamination with the Gen-Probe Mycoplasma
ribosomal RNA hybridization kit obtained from Fisher Scientific Co.,
Raleigh, NC. Dithiothreitol, diphenylamine, ethidium bromide, hy
droxyurea, and all other deoxyribonucleosides were obtained from
Sigma Chemical Co., St. Louis, MO. VM-263 was a gift of Dr. William
T. Bradner of Bristol Myers Co., Syracuse, NY, and Ara-C was provided
by Dr. J. Courtland White of Bowman Gray School of Medicine.
m-AMSA was obtained from the National Cancer Institute. Distilled
nucleic acid-grade phenol and a 1-kilobase DNA ladder were obtained
from Bethesda Research Laboratories, Gaithersberg, MD. Coomassie
brilliant blue G-250 was purchased from Eastman Kodak Co., Roch
ester, NY. [Methyl-3H]fThd and [2-14C]dThd with specific radioactivi
ties of 7-20 and 0.056 Ci/mmol, respectively, were purchased from
Moravek Biochemicals, Brea, CA. RNase-free pancreatic DNase I with
a specific activity of 2,200 Kunitz units per mg of protein was purchased
from Cooper Biomedicai, Freehold, NJ. Low salt buffer consisted of
10 IHMTris-HCl (pH 7), 1 mM MgCb, 10 mM NaCI, and 1 mM PMSF.
High salt buffer (1.5 M) consisted of 10 mM Tris-HCl (pH 7), 0.6 mM
MgCl2, 1.5 M NaCI, and 1 mM PMSF. High salt buffer (3 M) consisted
of 10 mM Tris-HCl (pH 7), 0.2 HIM MgCl2, 3 M NaCI, and 1 mM
PMSF.
Isolation of the Nuclear Matrix. Logarithmically growing CCRFCEM cells were preincubated for 72 h (approximately three doubling
times) with 30 nM [14C]dThd (final specific radioactivity of 0.056 Ci/
mmol) to uniformly radiolabel the DNA. The tumor cells were then
harvested by centrifugation at 37°C,and the pellets washed at 37°C
with 10 ml of Fischer's medium. The pellet was re-suspended in Fischer's
medium, and 1 ml of the cell suspension containing 2x10" cells was
placed in a microcentrifuge tube and preincubated for 10 min at 37°C
in a shaking water bath. The cells were pulse-labeled for either 45 s or
20 min with 710 nM [3H]dThd (final specific radioactivities of 7 and
1.8 Ci/mmol, respectively). The incorporation of the tri timed nucleoside into DNA was stopped by the addition of 500 ^1 of a 60 IHMEDTA
solution adjusted to pH 7, which was immediately followed by centrif
ugation of the samples for 10s at 13,OOOX£
in a microcentrifuge. The
cell pellets were resuspended in 1 ml of ice-cold 10 mM Tris-HCl (pH
7) buffer containing 1 mM MgCl2 and 1 mM PMSF, allowed to swell
3 The abbreviations used are: VM-26, 4'-demethylepipodophyllotoxinethylidene-/3-D-glucoside (teniposide); m-AMSA, 4'-(9-acridinylamino)-3-methanesulfon-/n-anisidide; Ara-C, l-/3-D-arabinofuranosylcytosine (cytosine arabinoside);
dThd, 2'-deoxythymidine; PMSF, phenylmethylsulfonyl fluoride; TCA, trichloroacetic acid; SDS, sodium dodecyl sulfate.
1850
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.
ANTILEUKEMIA
AGENTS AND THE NUCLEAR MATRIX
matrix DNA were added in triplicate to 2 x 10' unlabeled nuclei.
on ice for 15 min, and then disrupted with 25 strokes in a Dounce
homogenizer. The crude nuclear preparation was layered over 10 ml of Nuclear matrices were prepared from the unlabeled nuclei, and the
45% sucrose (w/v), and then centrifuged at l,900xg for 30 min at 2°C. specific activity of the nascent DNA on the nuclear matrix (3H/"C)
The purification was monitored by phase contrast microscopy following
Giemsa staining. Two hundred structures were counted of which 95%
were intact nuclei with well defined borders free of visible cytoplasmic
contamination, 4% were whole cells, and 1% were fractured nuclei.
The procedure followed for the isolation of the nuclear matrix was a
modification ofthat described by Pardoll et al. (2). Purified nuclei were
gently resuspended in 1 ml of low salt buffer at 4°C.One ml of 3 M
high salt buffer was added over a period of l h with gentle mixing to
yield a final concentration of 1.5 M NaCl. The samples were incubated
for an additional 30 min on ice, and then placed in a water bath at
37°C.The DNA was digested by the addition of 250 units of RNase-
was calculated to determine any nonspecific binding of DNA to the
matrix.
Electron Microscopy. For transmission electron microscopy nuclei
and nuclear matrices were fixed in 3% glutaraldehyde, postfixed in 1%
osmium tetroxide, and stained with lead citrate and uranyl acetate as
previously described (27). Sections of 60-mm thickness were examined
in a Philips EM-400 electron microscope. Nuclei and nuclear matrices
were prepared for scanning electron microscopy by fixation of the
samples on glass coverslips followed by coating of the glass surfaces in
a D. C. sputterer with gold-palladium (28).
free pancreatic DNase I. Preliminary experiments indicated that be
tween 95 and 99% of the total '"C-labeled nuclear DNA was made
aqueous soluble under these conditions, although greater than 97% of
this DNA could still be precipitated with ice-cold 10% TCA. The
samples were then centrifuged and the supernatants (nonmatrix frac
tions) saved on ice for quantitation of DNA. The pellets (nuclear
matrices) were washed once with 1 ml of 1.5 M high salt buffer, once
with 1 ml of low salt buffer containing 1% (v/v) Triton X-100, and then
immediately resuspended and washed with 1 ml of low salt buffer. The
matrix and nonmatrix fractions were precipitated in ice-cold 10% TCA
(w/v) for 1 h, centrifuged, and resuspended in 500 n\ of 0.5 N KOH
for l h at 37°Cto hydrolyze the RNA. All centrifugations were at
7,000xg for 15 min at 4°C.The DNA was solubilized by heating the
samples at 90°Cin 5% TCA as previously described (22). The hot TCA
extraction procedure solubilized greater than 95% of both the 3H and
I4C radioactivities. Aliquots of the solubilized nuclear matrix and
nonmatrix fractions were removed for counting of the 3H and I4Clabels.
No spillover of 14Cradioactivity into the 3H channel or vice versa was
observed following calibration of the liquid scintillation counter for
dual label counting. In some experiments the DNA concentrations in
the nuclear matrix and nonmatrix fractions were estimated by the
diphenylamine colorimetrie assay as described by Giles and Myers (23)
using 2'-deoxyadenosine as standard. In these experiments the pre la
beling of DNA for 72 h with [l4C]dThd was omitted. Protein concen
trations were estimated by the Coomassie blue staining method (24)
and lipid phosphorus was quantitated as previously described (25).
Purification and Analysis of Nuclear Matrix-bound DNA. Nuclear
matrices from 6 x IO7 cells were incubated for 2 h at 60°Cin 10 mM
Tris-HCl (pH 8.0) buffer containing 0.2% sodium dodecyl sulfate
(w/v), 20 HIMEDTA, and 100 Mg/ml proteinase K. At the end of the
incubation, sufficient quantities of sodium acetate and magnesium
chloride were added to yield final concentrations of 0.3 M, and 0.01 M,
respectively. Each sample also contained 12.5 UKof polyacrylamide as
coprecipitant. The nucleic acids were precipitated overnight with 2.5
volumes of ethanol at —20°C.
The mixture was then centrifuged at
12,OOOX£
for 30 min at 4°C,and the nucleic acids purified from the
pellet by extraction and back extraction with phenol/chloroform as
previously described (26). The nucleic acids were concentrated by
ethanol precipitation and the pellet was resuspended in 10 mM TrisHCl (pH 8) containing 2 mM EDTA and 10% (w/v) sucrose. The
samples were electrophoresed on 1% agarose gels at 5 V/cm with 89
mM Tris-borate (pH 8), 2 mM EDTA, and 0.5 jig/ml ethidium bromide
as the running buffer. Double-stranded DNA restriction fragments from
75 to 12,000 basepairs were used as molecular weight markers.
Mixing Experiments. Logarithmically growing tumor cells (1.2 x
10") were prelabeled for 72 h with [l4C]dThd and then pulse-labeled for
45 s with [3H]dThd. Purified nuclei were prepared as described above
and triplicate aliquots, each containing 2x10" nuclei, were resuspended
in 500 i/1 of low salt buffer. One of the aliquots was mildly sonicated
with three 15-s bursts at 15 W using a Branson 200 sonifier equipped
with a microtip. This sample was precipitated overnight at —20°C
in
ethanol containing 150 mM NaCI, and the chromatin fragments were
isolated following centrifugation at 13,000xg for 30 min at 4°C.Nu
clear matrix DNA and nonmatrix DNA were isolated and purified from
the remaining aliquots as described above. In the mixing experiments
the radiolabeled chromatin fragments, nonmatrix DNA, and nuclear
RESULTS
Characterization of the Nuclear Matrix. Initial studies in
volved a morphological and biochemical characterization of the
nuclear matrix, since similar studies with CCRF-CEM leuke
mia cells had not been previously reported. In addition, it was
first necessary to demonstrate that newly synthesized DNA was
specifically bound and enriched on the nuclear matrix before
the effects of anticancer agents on nuclear matrix-bound DNA
synthesis could be investigated. Scanning electron microscopy
of whole nuclei revealed nearly spherical structures of about 7
urn in diameter (Fig. \A). Nuclear matrices that were prepared
from CCRF-CEM cells by high salt extraction and DNase I
digestion of purified nuclei contained 1-5, 10, and 37% of the
total nuclear DNA, protein, and phospholipid, respectively.
The removal of most of the nuclear protein, phospholipid, and
DNA permitted the visualization of the residual nuclear frame
work or matrix. The nuclear matrices were of similar size and
shape as unfractionated nuclei but had a sponge-like appearance
due to the presence of numerous pores of varying sizes (Fig.
1B). Comparison of whole nuclei (Fig. 1C) and nuclear matrices
(Fig. ID) by transmission electron microscopy indicated that
the matrix lacked much of the chromatin of the whole nucleus
and contained only small remnants of the nuclei >1us and nuclear
envelope. The matrix consisted mainly of a residual framework
of fibrils and granules that was connected to a single peripheral
lamina. The proteins solubilized from the nuclei and nuclear
matrices were separated by SDS polyacrylamide gel electrophoresis (Fig. 2). Compared to whole nuclei, the nuclear matrices
of CCRF-CEM cells were extensively depleted of histones (A/t
13,000-20,000) but were relatively enriched in M, 60,00070,000 proteins. These results were similar to those previously
obtained with rat liver and other cells (1,3, 7).
It was reasoned that if DNA replication occurred at specific
sites on the residual matrix framework, then the insoluble
nuclear matrix should be enriched in nascent DNA compared
to the high salt-soluble (nonmatrix) DNA fraction. As a means
of testing this hypothesis, CCRF-CEM cells were prelabeled
for 72 h (approximately three doubling times) with [14C]dThd
to uniformly label the DNA prior to pulse-labeling of the newly
replicated DNA for 45 s with [3H]dThd. Purified nuclei were
treated with 1.5 M NaCl and DNase I to solubilize 98% of the
DNA. The resultant matrices were isolated and contained frag
ments of bound DNA that were between 0.1 and 1.6 kilobases
in length as determined by agarose gel electrophoresis. The
specific activity of this newly replicated DNA bound to the
nuclear matrix (14.3 3H dpm/14C dpm in parental DNA) was
four-fold higher than the corresponding activity of nonmatrix
DNA (3.6 3H dpm/14C dpm) (Table 1). When the relative DNA
contents of the nuclear matrices and nonmatrix fractions were
1851
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.
ANTILEUKEMIA
AGENTS AND THE NUCLEAR MATRIX
Fig. 1. Electron micrographs of whole nuclei and nuclear matrices from CCRF-CEM leukemia cells. Nuclear matrices were prepared as described in the text and
contained 2,10, and 37% of the total nuclear DNA, protein, and phospholipid, respectively. A and B, scanning electron micrographs of a nucleus and nuclear matrix,
respectively; C and D, transmission electron micrographs of a nucleus and nuclear matrix, respectively. Amm in D, a small remnant of the nuclear envelope. Bars, 1
measured by the diphenylamine colorimetrie assay instead of
by [I4C]dThd prelabeling, a 5.1-fold enrichment of newly syn
thesized DNA on the nuclear matrix (1018 3H dprn/^g of DNA)
was observed compared to the nonmatrix DNA (201 3H dpm/
Mgof DNA). Thus, regardless of the method used to determine
DNA content, a similar degree of enrichment of newly synthe
sized DNA on the nuclear matrix was observed. In all subse
quent studies, the relative amounts of DNA were quantitated
following [l4C]dThd prelabeling because of the much greater
sensitivity of the radiochemical technique compared to the
colorimetrie assay in detecting the small amounts of DNA
bound to the nuclear matrix.
It was important to note that when CCRF-CEM cells were
pulse-labeled for 20 min with [3H]dThd as opposed to 45 s, no
enrichment of the 3H radiolabel was seen on the nuclear matrix
compared to the nonmatrix fraction (Table 1). During the 20min continuous labeling period a large fraction of the [3H]dThd
progressively away from these sites into the nonmatrix (high
salt-soluble) DNA according to the rate of replication fork
movement.
Control Experiments. The results from several types of con
trol experiments indicated that the enrichment of nascent DNA
observed on the nuclear matrix was not the consequence of
aggregation of the nascent DNA fragments with the nuclear
matrix. The repeated isolation of nuclear matrices with a con
stant relative specific activity of nascent DNA (Tables 1 and 3)
suggested that nonspecific association of DNA with the matrix
was unlikely. In addition, the 3H radioactivity in the nuclear
matrix fractions shown in Tables 1 and 3 were measured after
successive washings of the isolated matrices with high salt buffer
containing 1.5 M NaCl, low salt buffer containing 1% Triton
X-100, low salt buffer, and 10% TCA.
Reconstitution experiments indicated that nascent DNA did
not preferentially bind to the nuclear matrix during the matrix
isolation procedure. CCRF-CEM cells were prelabeled with
would have become incorporated into high molecular weight
[14C]dThd for 72 h and then pulse-labeled for 45 s with [3H]DNA that was no longer replicating. Therefore, the results
dThd. The double-labeled chromatin fragments, purified matrix
presented in Table 1 were consistent with the proposed concept
(2, 9, 29) that DNA replication takes place at sites on the DNA, and nonmatrix DNA were prepared and added to sus
nuclear matrix, and that previously replicated DNA moves pensions of unlabeled nuclei. Nuclear matrices were then iso1852
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.
ANTILEUKEMIA
NM
AGENTS AND THE NUCLEAR MATRIX
that dThd was actually incorporated into DNA at replication
sites on the nuclear matrix.
Effects of Antileukemic Agents on Nuclear Matrix-bound
DNA Replication. It was of interest to examine the sensitivity
of nuclear matrix-bound DNA synthesis to various antileukemic
agents. Table 3 shows that brief exposure of CCRF-CEM cells
to relatively low concentrations of either VM-26 or m-AMSA
resulted in preferential inhibition of [3H]dThd incorporation
N
into nuclear matrix DNA compared to nonmatrix DNA. Newly
synthesized DNA in control cells was enriched 4.1-fold in the
nuclear matrix fraction compared to the nonmatrix fraction.
Following a 2-h incubation of cells with either 0.1 fiM VM-26,
0.2 /¿MVM-26, or 0.5 ¿¿M
m-AMSA, the relative specific
activities of nascent DNA on the nuclear matrix were decreased
59% (1.7-fold enrichment), 61% (1.6-fold enrichment), and
54% (1.9-fold enrichment), respectively. Similar results were
obtained when the duration of the [3H]dThd pulse-label was
45
31
changed from 45 s to either 15, 30, or 90 s (data not shown).
It was important to note that these effects of VM-26 and
m-AMSA could not be readily attributable to differences be
tween control and drug treated cells in either dThd transport,
dThd phosphorylation, or deoxyribonucleoside triphosphate
pool sizes. Drug-induced alterations in these parameters would
not be expected to affect the rate of [3H]dThd incorporated into
*
21
14
Fig. 2. SDS polyacrylamide electrophoresis gel of proteins present in whole
nuclei and nuclear matrices of CCRF-CEM cells. Nuclei and nuclear matrices
were isolated as described in "Materials and Methods." The proteins were
electrophoresed on a 12% SDS polyacrylamide gel and stained with Coomassie
blue as described by Laemmli (32). Lanes NM (nuclear matrices) and N (whole
nuclei) were loaded with 31 j»g
and 27 jig of protein, respectively. The apparent
molecular masses (x 1000) of the standard proteins (lane S) are shown at the
right.
lated as described above. It was reasoned that if the newly
replicated [3H]DNA bound preferentially and nonspecifically to
the nuclear matrix during the matrix isolation procedure, then
the 3H-labeled DNA should be enriched in the nuclear matrix
fraction compared to the DNA uniformly labeled with [I4C]dThd. Table 2 shows that regardless of the degree of deproteinization of the newly replicated DNA (i.e., purified matrix DNA
> nonmatrix DNA > chromatin fragments), no enrichment of
newly replicated DNA on the matrix was observed. These
results are in contrast to the four-fold enrichment of nascent
DNA that was measured on the matrix when intact tumor cells
were pulse-labeled with [3H]dThd (Tables 1 and 3). Taken
together, the data presented in Tables 1-3 provided evidence
the matrix DNA fraction compared to that incorporated into
the nonmatrix DNA fraction within the individual cell. In
contrast to the specific effects observed with the DNA topoisomerase II inhibitors, VM-26 and m-AMSA, hydroxyurea and
Ara-C did not reduce the relative specific activity of nascent
DNA on the matrix. A 2-h exposure of CCRF-CEM cells to
200 MMhydroxyurea decreased the incorporation of [3H]dThd
into nuclear matrix DNA 77% compared to control cells (Table
4). However, [3H]dThd incorporation into the nonmatrix DNA
fraction was also decreased 77%, so that no change in the
relative specific activity was observed. Likewise, a 2-h incuba
tion of leukemia cells with 0.1 /ÕMAra-C produced similar
declines in both the matrix and nonmatrix specific activities
compared to control cells (Table 4). Hydroxyurea at 400 fiM
and Ara-C at 0.03 ^M for 2 h also did not change the ratio of
the matrix DNA specific activity to the nonmatrix DNA specific
activity (data not shown). These observations indicated that the
DNA topoisomerase II inhibitors, in contrast to hydroxyurea
and Ara-C, preferentially interfered with DNA replication at
sites on the nuclear matrix.
DISCUSSION
The identification and partial characterization of the nuclear
matrix from various types of eukaryotic cells has provided new
insights into the cellular organization and replication of DNA.
Table 1 Nuclear matrix-bound DNA synthesis
Logarithmically growing CCRF-CEM cells were prelabeled for 72 h in the presence of [14C]dThd, resuspended in fresh medium, and then labeled for either 45 s
or 20 min with [3H]dThd. The nuclei were isolated by centrifugal ion through 45% (w/v) sucrose and the nuclear matrices prepared as described in "Materials and
Methods."
Radiolabeling
procedure45-s
DNA
fractionMatrix
DNA
(14Cdpm)782
synthesized
DNA (3H
dpm)11,218
activity3H
dpmMC
dpm14.3
pulse [3H]dThd
*4.0
0.80.8
±
114,400
32,085
3.6
Nonmatrix
33,121
62129,227Newly
53.3
Matrix
1,945,381Specific
66.6Relative"NonmatrixTotal
* Ratio of the specific activity of DNA in the nuclear matrix fraction to the specific activity in the nonmatrix fraction.
'Mean ±50(^=3).
c P < 0.01 compared to the relative specific activity of the matrix DNA obtained after a 45-s pulse label with [3H]dThd (two-tailed t test).
20-min pulse [3H]dThdNuclear
±0.
Ie1
1853
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.
ANTILEUKEMIA
AGENTS AND THE NUCLEAR MATRIX
replication (1-8). Because of the involvement of the nuclear
matrix in DNA replication, it seemed reasonable that certain
antiproliferative agents would specifically interfere with the
function of DNA replication complexes on the nuclear matrix.
However, for the most part these potentially important effects
have remained unexplored.
Nuclear matrices were prepared from CCRF-CEM cells by
sequential
treatment of purified nuclei with low salt buffer, high
activity3H/14C253412916Relative'0.710.510.61
recovered14C
DNA
salt buffer, pancreatic DNase I, and Triton X-100. The mor
RadiolabeledDNA
dpm1112,668794,804899323Hdpm2,80591,1249211,57578314,692Specific
addedChromatin
phological and biochemical characteristics of the resultant mat
frag
rices were generally similar to the type III matrices isolated
mentsNonmatrixDNAPurified
from regenerating rat liver (4). Nuclear matrix preparations
from CCRF-CEM cells were 4-fold enriched in newly synthe
matrixDNANuclearDNAfractionMatrixNonmatrixMatrixNonmatrixMatrixNonmatrixRadiolabeled
sized DNA compared to the high salt-soluble nonmatrix DNA.
The results of the pulse-labeling and reconstitution experiments
" Ratio of the specific activity of DNA in the nuclear matrix fraction to the
indicated that this preferential association of newly replicated
specific activity in the nonmatrix fraction.
DNA with the nuclear matrix did not result from aggregation
of nascent DNA with the matrix. Using similar matrix isolation
and pulse-labeling techniques, others have observed a 2- to 15Table 3 Effects of VM-26 and m-AMSA on nuclear matrix-bound DNA
replication
fold enrichment of nascent DNA on the nuclear matrices of
Logarithmically growing CCRF-CEM cells were prelabeled for 72 h with
F4N Friend cells, HeLa cells, regenerating rat liver, and mouse
[>4C]-dThd, resuspended in fresh medium, and then incubated for 2 h with either
fibroblasts (2, 13,29,30).
no drug (control), VM-26, or m-AMSA. The cells were then pulse labeled for 45
s with [3H]Thd and the nuclear matrices prepared as described in "Materials and
The results obtained from this study with CCRF-CEM cells
Methods."
provide further information for evaluating the role of the nu
Specific activity
clear matrix in DNA replication. A model of DNA replication
Total DNA Newly synthesized
(MCdpm)
DNA
(3H
dpm)
3H/'4C
Relative"'*
has been proposed in which DNA replication complexes and
Condition
replicating DNA loops are attached to the nuclear matrix (2, 9,
ControlMatrixNonmatrixVM-26
0.911.7
±
29). In this model newly replicated DNA moves progressively
away from the matrix-bound replication sites according to the
MM)MatrixNonmatrixVM-26
(0.1
±0.4°11.6
rate of replication fork movement and then becomes incorpo
rated into the high salt-soluble DNA. If this model were appro
MM)MatrixNonmatrixm-AMSA
(0.2
priate for DNA replication in CCRF-CEM cells, one would
±0.4''11.8
expect that a smaller fraction of the newly replicated DNA
MM)MatrixNonmatrix84034,62457632,42865134,45445421,54713,689138,4012,21472,64970822,9171,16230,00516.34.03.82.21.10.72.61.44.1
(0.5
would have migrated away from the matrix-bound replication
±0.2"'
sites during a 45-s pulse label than during a 20 min pulse label.
" Ratio of the specific activity of DNA in the nuclear matrix fraction to the
Thus, the enrichment of newly replicated DNA that was ob
served in the nuclear matrix fraction after a 45-s [3H]dThd pulse
specific activity in the nonmatrix fraction.
* Mean ±SD (/V= 4).
label, but not after the 20-min pulse label, was consistent with
' P < 0.02 compared to the relative specific activity of the control matrix DNA
this model.
(paired t test).
* P< 0.01 compared to the relative specific activity of the control matrix DNA
In order to further characterize nuclear matrix-bound DNA
(paired t test).
replication and investigate the effects of various antileukemic
agents in this system, CCRF-CEM cells were incubated with
Table 4 Effects of hydroxyurea and Ara-C on nuclear matrix-bound DNA
either VM-26, m-AMSA, hydroxyurea, or Ara-C. It was of
replication
considerable interest that only VM-26 and m-AMSA, agents
Logarithmically growing CCRF-CEM cells were prelabeled for 72 h with
114('l-dThd. resuspended in fresh medium, and then incubated with either no drug
which interfered with the activity of DNA topoisomerase II,
(control), Ara-C, or hydroxyurea for 2 h. The cells were then pulse labeled for 45
reduced the relative specific activity of nascent DNA on the
s with [3H]dThd, and the nuclear matrices prepared as described in "Materials
and Methods."
nuclear matrix. This effect was seen after brief exposure of
CCRF-CEM cells to relatively low concentrations of either
Specific activity
VM-26 or m-AMSA. Hydroxyurea and Ara-C, at concentra
Total DNA Newly synthesized
(14Cdpm)
DNA (3H dpm)
3H/14C Relative"'*
Condition
tions which produced marked inhibition of total nuclear DNA
ControlMatrixNon
synthesis, did not preferentially block [3H]dThd incorporation
±14.1
into nuclear matrix-bound DNA compared to nonmatrix DNA.
matrixHydroxyurea(200
We proposed that the effects of VM-26 and m-AMSA on
MM)MatrixNonmatrixAra-C
nuclear matrix-bound DNA replication are related to the local
±14.5
ization of DNA topoisomerase II activity on the nuclear matrix
MM)MatrixNonmatrix84034,62431629,18431444,51313,689138,4011,17525,0722897,94716.34.03.70.90.90.24.1
(0.1
of CCRF-CEM cells. This proposal was supported by the recent
±10.91.01.1
observations that DNA topoisomerase II was a major protein
of the Drosophila nuclear matrix (18), and that VM-26-topo°Ratio of the specific activity of DNA in the nuclear matrix fraction to the
specific activity in the nonmatrix fraction.
isomerase II complexes were bound primarily to newly repli
* Mean ±SD (Ar= 4).
cated DNA of tumor cells (31). Moreover, an important role
for DNA topoisomerase II on the nuclear matrix was also
suggested by reports that the anchoring of supercoiled DNA
Although the matrix likely functions as a structural component
of the nucleus, numerous studies indicate that it also has an loops to the nuclear matrix induced torsional stress in the DNA
(4, 9).
important role in various nuclear processes including DNA
Table 2 Binding of chromatin fragments, purified matrix DNA, and nonmatrix
DNA to the nuclear matrix
Logarithmically growing CCRF-CEM cells were prelabeled for 72 h with
[14C]-dThd and then pulse labeled for 45 s with [3H]dThd. Chromatin fragments,
nonmatrix DNA, and purified matrix DNA were isolated from the labeled nuclei
and then added to unlabeled nuclei. Nuclear matrices were prepared from the
unlabeled nuclei as described in "Materials and Methods." In each sample the
recovery of either 3H or I4Cradioactivity after nuclear matrix isolation was greater
than 93%.
1854
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.
ANTILEUKEMIA
AGENTS AND THE NUCLEAR MATRIX
It was of interest to consider these different drug effects in
light of the model of nuclear matrix-bound DNA replication
that was discussed above. According to this model all DNA
replication takes place on the nuclear matrix, and the presence
of newly replicated DNA in the nonmatrix fraction is the result
of migration of this DNA from the matrix-bound replication
sites. Therefore, one might expect that other antimetabolite
inhibitors of DNA replication would have an indirect effect like
that observed with hydroxyurea or Ara-C. By blocking to a
similar extent DNA replication and the subsequent migration
of newly replicated DNA away from the matrix, hydroxyurea
and Ara-C would produce no changes in the relative specific
activity of the nascent DNA on the matrix. However, the finding
that VM-26 and m-AMSA actually decreased the relative spe
cific activity of matrix DNA suggested that these drugs induced
direct perturbations in the nuclear matrix DNA replication
system. Gasser and Laemmli (5) have provided evidence that
DNA topoisomerase II is enriched at the attachment sites of
chromatin loops to the Drosophila nuclear matrix. Accordingly,
one might speculate that the interference with DNA topoisom
erase II activity on the nuclear matrix by VM-26 and m-AMSA
would lead to inhibition of the binding of parental and newly
replicated DNA to the matrix. As a result of its decreased
matrix attachment some of the newly replicated DNA would
be more readily extracted during the matrix isolation procedure,
and the relative specific activity of the nascent DNA on the
matrix would thereby be lowered. Another interesting possibil
ity is that following the interaction of VM-26 or m-AMSA with
DNA topoisomerase II on the nuclear matrix, at least some of
the residual DNA synthesis no longer takes place on the matrix.
These potential effects of VM-26 and m-AMSA represent novel
drug actions and are currently under investigation in this labo
ratory. In summary, it appears likely that the nuclear matrix is
an important cellular site at which certain anticancer agents
exert their antiproliferative effects.
ACKNOWLEDGMENTS
Electron microscopy was carried out with the kind assistance of the
personnel in the core electron microscopy laboratory of the Bowman
Gray Oncology Research Center.
REFERENCES
1. Berezney, R., and Coffey, D. S. Identification of a nuclear protein matrix.
Biochem. Biophys. Res. Commun., 60: 1410-1416, 1974.
2. Pardoll, D. M., Vogelstein, B., and Coffey, D. S. A fixed site of DNA
replication in eucaryotic cells. Cell, 19: 527-536, 1980.
3. Bouteille, M., Bouvier, D., and Sève,A. P. Heterogeneity and territorial
organization of the nuclear matrix and related structures. Int. Rev. Cytol.,
«5:135-182, 1983.
4. Berezney, R. Organization and functions of the nuclear matrix, in: L. S.
Hnilica (ed.). Chromosomal Nonhistone Proteins, Vol. 4, pp. 119-180. Boca
Raton, FL: CRC Press, 1984.
5. Gasser, S. M., and Laemmli, U. K. The organization of chromatin loops:
characterization of a scaffold attachment site. EMBO J., 5: 511-518, 1986.
6. Nelson, W. G.. Pienta, K. J., Barrack, E. R., and Coffey, D. S. The role of
the nuclear matrix in the organization and function of DNA. Ann. Rev.
Biophys. Biophys. Chem., 15: 457-475, 1986.
7. Kaufman S. H., Fields, A. P., and Shaper, J. H. The nuclear matrix: current
concepts and unanswered questions. Methods Achiev. Exp. Pathol., 12: 141171, 1986.
8. Razin, S. V. DNA interactions with the nuclear matrix and spatial organi
zation of replication and transcription. BioEssays, 6: 19-23, 1987.
9. Vogelstein, B., Pardoll, D. M., and Coffey, D. S. Supercoiled loops and
eucaryotic DNA replication. Cell, 22: 79-85, 1980.
10. Pienta, K. J., and Coffey, D. S. A structural analysis of the role of the nuclear
matrix and DNA loops in the organization of the nucleus and chromosome.
J. Cell Sci. Suppl., /: 123-135, 1984.
11. Smith, H. C., and Berezney, R. Nuclear matrix-bound deoxyribonucleic acid
synthesis: an in vitro system. Biochemistry, 21: 6751-6761, 1982.
12. Smith, H. C., and Berezney, R. Dynamic domains of DNA polymerase a in
regenerating rat liver. Biochemistry, 22: 3042-3046, 1983.
13. Foster, K. A., and Collins, J. M. The interrelation between DNA synthesis
rates and DNA polymerases bound to the nuclear matrix in synchronized
HeLa cells. J. Biol. Chem., 260: 4229-4235, 1985.
14. Tubo, R. A., and Berezney, R. Pre-replicative association of multiple repli
cative enzyme activities with the nuclear matrix during rat liver regeneration.
J. Biol. Chem., 262: 1148-1154, 1987.
15. Tubo, R. A., and Berezney, R. Identification of 100 and 150 S DNA
polymerase a-primase megacomplexes solubilized from the nuclear matrix
of regenerating rat liver. J. Biol. Chem., 262: 5857-5865, 1987.
16. Wood, S. H., and Collins, J. M. Preferential binding of DNA primase to the
nuclear matrix in HeLa cells. J. Biol. Chem., 261: 7119-7122, 1986.
17. Tubo, R. A., and Berezney, R. Nuclear matrix-bound DNA primase. Eluci
dation of an RNA priming system in nuclear matrix isolated from regener
ating rat liver. J. Biol. Chem., 262: 6637-6642, 1987.
18. Berries, M., Osheroff, N., and Fisher, P. A. In situ localization of DNA
topoisomerase II, a major polypeptide component of the Drosophila nuclear
matrix fraction. Proc. Nati. Acad. Sci. USA, «2:4142-4146, 1985.
19. Tubo, R. A., Smith, H. C., and Berezney, R. The nuclear matrix continues
DNA synthesis at in vivo replication forks. Biochim. Biophys. Acta, 825:
326-334, 1985.
20. Valenzuela, M. S., Mueller, G. C., and Dasgupta, S. Nuclear matrix-DNA
complex resulting from EcoRl digestion of HeLa nucleoids is enriched for
DNA replicating forks. Nucleic Acids Res., //: 2155-2165, 1983.
21. Fernandes, D. J., and Cranford, S. K. Dissociation of thymidylate biosyn
thesis from DNA biosynthesis by 5-fluoro-2'-deoxyuridine and 5,8-dideazaisofolic acid. Cancer Res., 46: 1741-1747, 1986.
22. Hryniuk, W. M., Fischer, G. A., and Berlino, J. R. 5-Phase cells of rapidly
growing and resting populations: differences in response to methotrexate.
Mol. Pharmacol., 5: 557-564, 1969.
23. Giles, K. W., and Myers, A. An improved diphenylamine method for the
estimation of deoxyribonucleic acid. Nature (Lond.), 206: 93, 1965.
24. Bradford, M. M. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding.
Anal. Biochem., 72: 248-254, 1976.
25. Chalvardjian, A., and Rudnicki, E. Determination of lipid phosphorus in the
nanomolar range. Anal. Biochem., 36: 225-226, 1970.
26. Maniatis, T., Fritsch, E. F., and Sambrook. J. Molecular Cloning: A Labo
ratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory,
1982.
27. Moran, J. R., and Lewis, J. C. The effects of severe zinc deficiency on
intestinal permeability: an ultrastructural study. Pediatr. Res., 19: 968-973,
1985.
28. Lewis, J. C., White, M. S., Prater, T., Taylor, R. G., and Davis, K. S.
Ultrastructural analysis of platelets in nonhuman primates. Exp. Mol. Pa
thol., 37: 370-381, 1982.
29. Berezney, R., and Buchholtz, L. A. Dynamic association of replicating DNA
fragments with the nuclear matrix of regenerating rat liver. Exp. Cell Res.,
/J2.-1-13, 1981.
30. Djondjurov, L., Ivanova, E., Markov, D., Bardarov, S., and Sachsenmaier,
W. Is the nuclear matrix the site of DNA replication in eukaryotic cells?
Exp. Cell Res., 164: 79-96, 1986.
31. Nelson, W. G., Liu, L. F., and Coffey, D. S. Newly replicated DNA is
associated with DNA topoisomerase II in cultured rat prostatic adenocarcinoma cells. Nature (Lond.), 322: 187-189, 1986.
32. Laemmli, U. K., Cleavage of structural proteins during the assembly of the
head of bacteriophage T4. Nature (Lond.), 227: 680-685, 1970.
1855
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.
Effects of Antileukemia Agents on Nuclear Matrix-bound DNA
Replication in CCRF-CEM Leukemia Cells
Daniel J. Fernandes, Cecil Smith-Nanni, Melanie T. Paff, et al.
Cancer Res 1988;48:1850-1855.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/48/7/1850
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1988 American Association for Cancer Research.