 |
INTRODUCTION |
Oligonucleotides can recognize both nucleic acids and proteins
with a high degree of specificity. This is a major reason why they have
been widely investigated as potential therapeutic agents for cancer,
viral infections, and inflammatory diseases. Oligonucleotides can
achieve target recognition by sequence-specific interactions with
nucleic acids or proteins such as in the antisense, antigene, or decoy
approaches (1-4). Alternatively, target recognition can be due to the
specific three-dimensional structure of an oligonucleotide, as in the
aptamer approach (5, 6). These aptameric oligonucleotides often contain
secondary structure elements such as hairpins or G-quartets. The
formation of G-quartet structures is also thought to contribute to
non-antisense growth inhibitory effects of G-rich phosphodiester and
phosphorothioate oligonucleotides (7-9).
Recently, we reported (9) on a novel class of phosphodiester G-rich
oligonucleotides (GROs)1 that
could strongly inhibit the in vitro proliferation of tumor cells derived from prostate, breast, and cervical carcinomas. The
antiproliferative GROs were able to form stable secondary structures
consistent with G-quartet formation. It was determined that these GROs
bound to a specific nuclear protein and, furthermore, that the growth
inhibitory activity of the GROs was positively correlated with their
ability to bind to this protein. The specific GRO-binding protein was
captured using biotinylated GROs and was identified by polyclonal and
monoclonal antibodies to nucleolin. Therefore, we concluded that these
potentially therapeutic oligonucleotides worked by a novel mechanism
that involved binding to nucleolin or a nucleolin-like protein. Our
hypothesis was that binding of GROs causes inhibition of nucleolin
function(s) that results in an arrest of proliferation.
Nucleolin is an abundant 110-kDa phosphoprotein, thought to be located
predominantly in the nucleolus of proliferating cells. Levels of
nucleolin are known to relate to the rate of cellular proliferation
(10, 11), being elevated in rapidly dividing cells such as malignant
cells. Therefore, nucleolin may be an attractive molecular target for
cancer therapy. The remarkable multifunctionality of nucleolin and its
role in cell growth and proliferation have been highlighted in recent
reviews (12-14). The most studied aspects of nucleolin function are
its roles in ribosome biogenesis, which include the control of rDNA
transcription, pre-ribosome packaging, and organization of nucleolar
chromatin (12, 15). It is also thought that nucleolin can act as a
shuttle protein that transports viral and cellular proteins between the cytoplasm and nucleus/nucleolus of the cell (16-18). In addition, nucleolin has been implicated, directly or indirectly, in other roles
including nuclear matrix structure (19), DNA replication (20),
cytokinesis and nuclear division (21), and as a nucleic acid helicase
(12, 22). There have been numerous reports describing the presence of
nucleolin in the plasma membrane of cells (9, 23-28), suggesting a
further function of nucleolin as a cell surface receptor. Clearly,
inhibition of nucleolin function, which we propose is an effect of GRO
binding, would likely result in inhibition of cell proliferation and/or
cell death.
To elucidate further the mechanism of the GRO antiproliferative
activity, we decided to study the effects of GROs on cellular processes, such as nucleic acid and protein synthesis, and cell cycle
progression. We report our findings that GROs specifically inhibit DNA
replication, and we discuss the implications of these results in terms
of potential mechanisms for GRO activity.
| |
EXPERIMENTAL PROCEDURES |
Oligonucleotides--
Except where indicated, oligonucleotides
had phosphodiester backbones and 3'-C-3 aminoalkyl
modifications. They were purchased from Oligos Etc. (Wilsonville, OR)
or synthesized on a Beckman 1000M synthesizer using 3'-C-3-amine
CPG columns from Glen Research (Sterling, VA). Oligonucleotides were
resuspended in water, precipitated with butan-1-ol, washed with 70%
ethanol, dried and resuspended in sterile 10 mM Tris·HCl,
pH 7.5, or phosphate-buffered saline (PBS). They were then sterilized
by filtration through a 0.2-µm filter and diluted with sterile buffer
to give stock solutions of 400 or 500 µM that were stored
in aliquots at 
20 °C. Each oligonucleotide was checked for
integrity by 5'-radiolabeling followed by polyacrylamide gel
electrophoresis. Sequences of oligonucleotides used were as
follows: GRO29A, 5'-TTTGGTGGTGGTGGTTGTGGTGGTGGTGG; GRO15B,
5'-TTGGGGGGGGTGGGT; GRO15A, 5'-GTTGTTTGGGGTGGT; GRO26A, 5'-GGTTGGGGTGGGTGGGGTGGGTGGGG;
GRO26B,5'-GGTGGTGGTGGTTGTGGTGGTGGTGG; and CRO,
5'-TTTCCTCCTCCTCCTTCTCCTCCTCCTCC.
Flow Cytometry Analysis of Cell Cycle--
For the experiments
shown in Figs. 1 and 2B, cells were plated in Dulbecco's
modified Eagle's medium (DMEM, Life Technologies, Inc.) supplemented
with 10% (v/v) fetal calf serum (FCS, Life Technologies, Inc.) and 1%
penicillin/streptomycin solution (Life Technologies, Inc.) at a density
of 105 cells per well in a 6-well plate. After incubation
at 37 °C for 4 h to permit adherence, cells were treated by
direct addition of oligonucleotide to the culture medium to give a
final concentration of 10 µM. For MDA-MB-231 cells only,
a second dose of oligonucleotide was added 24 h after the first.
At 72 h after the first dose, cells were harvested by
trypsinization, fixed, and stained with propidium iodide using the
Cycle Test Plus kit (Becton Dickinson). Cells were then analyzed using
a FACScan cytometer. The percentage of cells in
G1/G0, S, and G2/M were determined
using the Modfit program or, in a few cases where Modfit was unable to
assign these parameters, by integrating peak areas using the CellQuest
program. For the experiments shown in Fig. 2B, cells were
treated as above and harvested at the times shown and analyzed as described.
Synchronization of Cells--
To synchronize by serum
starvation, cells were plated in complete medium (DMEM, 10% FCS, 1%
antibiotic) at a density of 5 × 104 per well
in 6-well plates and incubated overnight. Cells were washed three times
with sterile PBS, and the medium was replaced with DMEM with 0.1%
(v/v) FCS. After incubation in low serum medium for 48 h, medium
was replaced with complete medium, and GRO29A was added directly to the
medium to give 10 µM final concentration. To synchronize
cells in S phase, cells were incubated for 16 h in complete medium
containing 1 µg/ml aphidicolin. Cells were released by washing three
times with PBS and adding complete medium. GRO29A was then added
directly to the medium to give a final concentration of 10 µM. For each method of synchronization, samples were
prepared in parallel to be harvested for flow cytometric analysis
before release or 24 h after addition of GRO29A. Analysis of cell
cycle was carried out as described above.
DNA Synthesis Assay--
Cells were plated at a density of
1.5 × 103 cells per well, in a 48-well plate. Cells
were incubated for 16 h and then treated by direct addition of
oligonucleotide to the culture medium at a final concentration of 12 µM (or untreated samples received an equal volume of
sterile PBS). A second identical dose was given 60 h after the
first. 72 h after the first treatment, cells were pulsed by
addition of BrdUrd (Sigma, final concentration 10 µM) to the medium for 1 h at 37 °C. Cells were
washed three times with PBS and fixed with methanol for 10 min at room
temperature. Samples were washed with PBS, incubated with 2 M HCl (to denature DNA) for 10 min on ice, washed, and
neutralized with 0.1 M sodium borate. The cells were then
incubated with anti-BrdUrd monoclonal antibody (Becton Dickinson, used
at 1.7 µg/ml) for 30 min at room temperature, washed, and incubated
with Alexa488-labeled goat anti-mouse IgG (Molecular Probes, used at 4 µg/ml) for 20 min at room temperature. After washing, cells were
viewed with a Nikon Eclipse TS100 microscope with ELWD epifluorescence
attachment, and digital images were captured using a Olympus DP10 camera.
RNA Synthesis Assay--
Samples shown in Fig. 4B
were prepared exactly as described for DNA synthesis, and after 72 h of incubation with oligonucleotide samples were pulsed by
addition of BrU (Aldrich) at a final concentration of 1 mM
for 1 h at 37 °C. Cells were fixed with 4% (w/v)
paraformaldehyde in PBS, stained as above (1.7 µg/ml anti-BrdUrd
followed by 4 µg/ml Alexa488-labeled anti-mouse IgG), and
photographed as described above. For the samples shown in Fig.
4A, cells were treated with 0.1 µg/ml actinomycin D for
1 h at 37 °C prior to pulsing with BrU or were incubated with
2.5 mg/ml RNase A for 30 min after fixing.
Protein Synthesis Assay--
Samples were prepared exactly as
described for DNA synthesis. After 72 h of incubation with
oligonucleotide, cells were washed twice with methionine/cysteine-free
medium (Life Technologies, Inc.), incubated for 15 min at 37 °C in
this medium, and then pulsed by addition of
[35S]methionine (final concentration 40 µCi/ml) for
1 h at 37 °C. After washing, proteins were extracted by
addition of 100 µl of citrate saline solution and precipitated by
addition of 400 µl of 10% trichloroacetic acid and 20 mM
pyrophosphate. An aliquot of this solution was used to quantitate total
protein using the BCA Protein Assay (Pierce), and the remainder was
filtered through a glass filter to determine trichloroacetic
acid-precipitable counts. Protein synthesis was expressed as counts/min
per µg of total protein. Experiments were performed in triplicate,
and error bars represent the S.E.
TUNEL Staining and Cell Morphology--
Cells were plated
(1.5 × 104 cells per well for DU145 or 4 × 104 cells per well for MDA-MB-231) in 24-well plates
containing glass coverslips (Fisher) that had been washed and
sterilized. Oligonucleotide was added directly to the medium to a final
concentration of 10 µM at 4 h after plating. Further
identical doses were given at 24 and 72 h subsequent to the first
dose (cells shown in Fig. 6A did not receive the third
dose). At the appropriate time, cells were washed with PBS and fixed
for 15 min at room temperature with 4% (w/v) paraformaldehyde in PBS.
Cells were permeabilized by incubation in 0.1% Triton X-100, 0.1%
sodium citrate for 2 min on ice. Terminal transferase-mediated labeling
of fragmented DNA with dUTP was carried out using In Situ
Cell Death Detection kit (Roche Molecular Biochemicals), according to
the manufacturer's instructions. For morphology analysis cells were
prepared as above and viewed using an Olympus BX60 microscope.
SV40 DNA Replication Assay--
Replication-competent cell
extracts were prepared from HeLa cells by a modification of the method
of Li and Kelly (29, 30). HeLa cells were grown in DMEM (Life
Technologies, Inc.) with 10% supplemented calf serum (HyClone) and
antibiotics and were maintained in logarithmic growth. Briefly, the
cells were released, combined, and washed once with isotonic buffer (20 mM HEPES, 5 mM KCl, 1.5 mM
MgCl2, 1 mM dithiothreitol, 250 mM
sucrose) and twice with hypotonic buffer (isotonic buffer without
sucrose). Following the second hypotonic wash, the cell count was
adjusted to 7 × 107 cells/ml, and the suspension was
placed on ice for 30 min. The cells were gently lysed using a Dounce
glass homogenizer and a tight-fitting plunger and then placed on ice
for 30 min. The suspension was centrifuged at 10,000 rpm for 10 min at
4 °C, and the supernatant was frozen by dripping it through liquid
nitrogen. Replication reactions were carried out using the M13mp2SV
replication template (7406 base pairs), which is an M13mp2 molecule
containing the SV40 origin of replication (31). Replicative form
templates were prepared from infected recA Escherichia coli
by standard methods, and covalently closed circular (form I) DNA was
isolated by two successive CsCl gradients. Template (40 ng) was
preincubated with water or the oligonucleotide (400 nM
final concentration) as indicated for 30 min at 37 °C and then added
to reaction mixes as described previously (32). Replication reactions
(25 µl) contained 30 mM HEPES, pH 7.8, 7 mM
MgCl2, 4 mM ATP, 200 µM each of
CTP, GTP, and UTP, 100 µM each of dATP, dGTP, dCTP, TTP,
and [
-32P]dCTP (6,600 dpm/fmol), 40 mM
creatine phosphate, 2 µg of creatine kinase, and 15 mM
sodium phosphate, pH 7.5. The SV40 large T antigen (Molecular Biology
Resources) was omitted from the negative control (Fig. 7, 1st
lane), and 1 µg was added to all other tubes. After addition of
cell extract (75 µg of protein) to each tube, the reactions proceeded
for 2 h at 37 °C. At this time, an equal volume of stop
solution (2% SDS, 50 mM EDTA, 2 mg/ml proteinase K) was added, and the samples were incubated for an additional 30 min at
37 °C. An aliquot (1/10 volume) was taken for determination of
[-32P]dCTP into acid-insoluble material, and picomoles
of dCTP incorporated was calculated. The relative dCTP incorporation
stated in the text represents the average of three independent
experiments, and standard errors of the data are shown. The DNA was
extracted, and aliquots were electrophoresed on 1% (w/v) agarose gels
containing 0.5 µg/ml ethidium bromide. The dried gels were imaged on
a Storm PhosphorImager (Molecular Dynamics), and the density of the
bands corresponding to covalently closed circular (form I) DNA was
quantified using ImageQuant software.
Cell Proliferation Assay--
HeLa cells were plated at
103 cells per well in 96-well plates. After incubation for
16 h, oligonucleotide was added directly to the culture medium to
give a final concentration of 12 µM. Culture medium was
not changed throughout the duration of the experiment. The relative
number of viable cells in each well was determined 96 h after
addition of oligonucleotides using the MTT assay (33). Briefly, 15 µl
of 5 mg/ml 3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide
(MTT, Sigma) was added to each well. Following 16 h of incubation
at 37 °C, cells were lysed by addition of 75 µl per well of 10%
SDS, 10 mM HCl, and after a further 16 h of incubation at 37 °C, absorbance at 570 nm was determined using a Molecular Devices microplate reader. Experiments were performed in triplicate, and bars represent the S.E.
Helicase Assay--
These assays were carried out
according to a method similar to that of Veaute et al. (34).
Oligonucleotide 55A
(5'-TGAAGGTTTCGAATCAGAGGTAGGTGCCCGGCCTCCAACTTGCCGTATTCCTGGT, unmodified phosphodiester, purchased from Life Technologies, Inc.) was
5'-labeled using T4 kinase and [
-32P]ATP. After
removal of unincorporated [32P]ATP it was annealed
(95 °C for 5 min followed by slow cooling to room temperature) to a
partially complementary oligonucleotide 55B
(5'-ACCAGGAATACGGCAAGTTGGAGGCCGGGCTGGATGGAGACTAAGCTTTGGAAGT, unmodified phosphodiester, purchased from Life Technologies,
Inc.) in order to form the synthetic replication fork substrate
containing a region of 30 base pairs (underlined). Unwinding of this
substrate by recombinant SV40 large T antigen (Chimerx, Milwaukee, WI)
was carried out by incubating 10 fmol of substrate with 100 ng of large
T antigen in the absence or presence of competitor unlabeled oligonucleotides in a final volume of 10 µl of HA buffer (20 mM Tris·HCl, pH 7.5, 7 mM MgCl2,
5 mM dithiothreitol, 2 mM ATP, and 25 µg/ml
bovine serum albumin). Large T antigen was preincubated for 15 min at
37 °C prior to addition. The reaction was allowed to proceed for 15 min at 37 °C before being terminated by addition of STOP buffer (200 mM EDTA, 40% glycerol, 0.6% SDS, 0.15% bromphenol blue,
0.15% xylene cyanol). Samples were analyzed by native polyacrylamide electrophoresis on 8% polyacrylamide gels containing 1× TBE (90 mM Tris borate, pH 8.3, 2 mM EDTA) with 1×
TBE, 0.1% SDS as running buffer. For the assay shown in Fig.
8A, the nonspecific competitor oligonucleotide was (NS)
(5'-AGGACTGTATACTGTCTTGGA, unmodified phosphodiester, purchased from
Life Technologies, Inc.), and 29A is this assay was also unmodified.
For the assay shown in Fig. 8B, all oligonucleotides except
NS were modified with a 3'-C-3 aminoalkyl group.
Nucleolin-Replication Protein A (RPA)
Interaction--
Immunoprecipitation was carried out in 0.5 ml of RIPA
buffer (50 mM Tris·HCl, pH 7.5, 0.5 M NaCl,
0.1 mM EDTA, 100 µM NaF, 1 mM
Na3VO4, 1% Nonidet P-40, 0.5% sodium
deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonyl
fluoride, 1 µM leupeptin, 10 µM aprotinin) for 30 min at 37 °C using 25 µg of HeLa extracts (as used for in vitro replication) and 5 µg of monoclonal antibody to
14-kDa subunit of RPA (Novus Biologicals). Antibodies were captured for 1 h at 4 °C using magnetic beads linked to goat anti-mouse IgG (Magnabind, Pierce), washed three times with 0.5 ml of RIPA, and eluted
by heating the beads for 15 min at 65 °C in a buffer containing 2%
SDS and 5% 2-mercaptoethanol. To investigate the effect of GROs on
nucleolin-RPA interaction, the HeLa extracts were preincubated for 30 min at 37 °C in the absence or presence of oligonucleotides. Precipitated proteins were electrophoresed on 8% polyacrylamide/SDS gels, transferred to polyvinylidene difluoride membrane,
Western-blotted using nucleolin monoclonal antibody (Santa Cruz
Biotechnology, 1:200 dilution) and peroxidase-linked goat anti-mouse
(Santa Cruz Biotechnology, 1:1000), and visualized by chemiluminescence.
| |
RESULTS |
Accumulation of GRO29A-treated Cells in S Phase--
To
investigate cell cycle perturbations induced by antiproliferative
G-rich oligonucleotides, flow cytometry analysis of propidium iodide
stained nuclei was performed. Cell cycle parameters were compared for
untreated cells and cells that had been treated for 72 h with an
active GRO (GRO29A) or a control inactive GRO (GRO15B). The cell lines
examined were DU145 (derived from prostate carcinoma), MDA-MB-231
(breast carcinoma), HeLa (cervical carcinoma), and HS27 (normal
foreskin fibroblasts). In all of the carcinoma lines, we observed a
significant increase in the fraction of cells in the S phase of the
cell cycle for cells treated with GRO29A, as shown in Fig.
1. This was accompanied by a decrease in
the proportion of cells in the G0/G1 and
G2/M phases of the cell cycle, as indicated in Fig. 1.
These changes were specific for GRO29A and not simply due to the
presence of oligonucleotide, because cells treated with GRO15B were
similar to untreated cells.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1.
Flow cytometric analysis of cell cycle
parameters following 72 h of treatment with active GRO29A or
inactive control GRO15B, compared with untreated cells. The
identity of the cell line is indicated on the left, and
numbers above each histogram indicate the ratio of cells in
the G1/G0, S, and G2/M phases of
the cell cycle (data were gated to exclude apoptotic cells for these
calculations). The peak marked by an asterisk
represents apoptotic cells with sub-G1 DNA content.
|
|
We also observed the appearance of a peak corresponding to a population
of cells with sub-G1 DNA content (marked with an
asterisk in Fig. 1). This peak, which is indicative of
apoptotic cells, appeared in the tumor cell samples treated with GRO29A
but was much smaller in the GRO15B-treated or untreated cells. The
induction of apoptosis by GRO29A was examined further and will be
described below.
In contrast to the malignant cell lines, HS27 cells, which are derived
from normal skin fibroblasts, showed no major perturbations in cell
cycle in response to treatment with GRO29A. The HS27 cells were also
found to be considerably less sensitive to the antiproliferative effects of GRO29A than most of the tumor cell lines we have
examined.2 However, it can be
seen from the data (Fig. 1, untreated) that the HS27 cells
have a much lower proportion in S phase (even though they are
sub-confluent), compared with the tumor cell lines, and proliferate at
a slower rate. To test whether the response of cells is related to the
proportion of cells in S phase when GRO is added, we partially
synchronized DU145 cells and added GRO29A when the cells were
predominantly in S phase (after aphidicolin treatment) or
G1/G0 (after serum starvation). Fig.
2A shows that the cells showed
similar S phase accumulation, irrespective of the proportion of cells
in S phase when GRO was added. These data suggest that the differential
responsiveness of HS27 cells is not necessarily related to the lower
proportion of these cells in S phase, but further studies are required
to confirm the tumor selectivity of GRO29A and investigate the
mechanism of selectivity.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 2.
A, flow cytometric analysis of cell
cycle parameters. DU145 cells were partially synchronized by either
serum starvation or aphidicolin treatment (top panel). Cells
were treated with 10 µM GRO29A at the time of release
from synchronization and analyzed after 24 h (bottom
panel). B, graph showing the percentage of cells
accumulated in S phase for asynchronous DU145 cells treated with GRO29A
or control oligonucleotide (GRO15B).
|
|
Studies of cell cycle progression following partial synchronization of
tumor cells showed that even when GRO29A was added in
G1/G0 (following serum synchronization) or
early S phase (following aphidicolin treatment), a considerable
proportion of the cells were able to progress through S phase and
G2/M (data not shown). In fact, the accumulation of cells
in S phase was found to be a gradual process that occurs over several
cell cycles, as shown in Fig. 2B for unsynchronized DU145
cells treated with GRO29A. The gradual arrest of cells in S phase may
be related to the rate of oligonucleotide uptake, which is thought to
be relatively slow. We are currently investigating the relationship
between GRO uptake and cell cycle arrest.
Inhibition of DNA Replication in Cells--
The S phase of the
cell division cycle represents the period in which cells replicate
their DNA (and thus have a DNA content intermediate between that of
G0/G1 and G2/M cells). In this
case, because cell proliferation is inhibited by GRO29A, it seems
unlikely that the increase in the S phase fraction represents an
increase in cells that are actively replicating DNA but more likely an accumulation of cells whose progress is arrested in this phase of the
cell cycle.
To confirm this hypothesis, we analyzed DNA replication in cells
treated with GROs. This was achieved by determining incorporation of
5-bromo-2'-deoxyuridine (BrdUrd), a nucleoside that can be incorporated
into cellular DNA in place of thymidine. A BrdUrd antibody can then
positively stain cells that actively synthesize DNA during the BrdUrd
pulse. In this experiment, DU145 cells were incubated as described in
the absence or presence of GROs for 72 h, at which time BrdUrd was
added to the cell culture medium for 1 h. Cells were then fixed
with methanol, and incorporation of BrdUrd was assessed by indirect
immunofluorescent staining using a BrdUrd antibody. Fig.
3 shows the results of these experiments. In the untreated sample, ~40% of cells was found to be positive for
BrdUrd staining. However, the cells treated with GRO29A showed an
almost complete absence of staining, indicating that no de novo DNA synthesis was occurring in these cells. This effect was specific for GRO29A, because cells treated with the control
oligonucleotide (GRO15B) had similar BrdUrd incorporation to untreated
cells.
View larger version (70K):
[in this window]
[in a new window]
|
Fig. 3.
DNA synthesis in untreated DU145 cells
and cells treated with GRO15B (control oligonucleotide) or GRO29A
(active oligonucleotide). Cells were treated for 72 h and
then pulsed with BrdUrd. Incorporation of BrdUrd was detected by
indirect immunofluorescence using a BrdUrd monoclonal antibody.
|
|
To confirm that the effects of GRO29A on cell cycle and DNA synthesis
occur in parallel, we carried out one further experiment. DU145 cells
were treated in a 6-well plate, pulsed with BrdUrd for 1 h, then
collected by trypsinization, and divided into 2 aliquots. One aliquot
was stained with propidium iodide and analyzed by flow cytometry, and
the other aliquot was used to prepare slides for staining with BrdUrd
antibody. The results indicated that for the same sample of
GRO29A-treated cells, there was an accumulation in S phase concurrent
with an inhibition of DNA synthesis (data not shown).
The time course of inhibition of DNA replication depended on a number
of factors, including the initial cell density and the number and
schedule of GRO29A doses. When DU145 cells were plated at 6.25 × 103 cells per well in a 24-well plate, and treated after
17 h with 10 µM final concentration of GRO, we found
no significant difference between DNA synthesis in treated and
untreated cells at 3, 6, and 9 h after addition of GRO29A.
However, at 12 h after treatment, only 28% of cells stained
positive for BrdUrd in the GRO29A-treated sample, compared with 41 and
44%, respectively, in the untreated and GRO15B treated cells. By
24 h after addition of oligonucleotide, the GRO29A-treated cells
were completely negative for BrdUrd staining.
Effect of GROs on RNA and Protein Synthesis--
Because
inhibition of any one of DNA, RNA, or protein synthesis is expected to
lead eventually to inhibition of the other processes, it is important
to determine which of these functions is arrested first. To investigate
this, we examined both RNA and protein synthesis in parallel with DNA
replication, at a time point at which DNA synthesis was known to be
completely inhibited. As described above, DNA replication was assessed
by BrdUrd incorporation, and a similar method was used to determine RNA
synthesis. This method has been described previously (35, 36) and
involves detection of incorporated 5-bromouridine (BrU) by indirect
immunofluorescent staining with a BrdUrd antibody (which is
cross-reactive for BrU). To ensure that the BrU staining represents
bona fide RNA synthesis, we carried out a preliminary
experiment to assess the effects of treatment with RNase or actinomycin
D, an inhibitor of RNA synthesis. Fig.
4A shows that staining in
these samples was negative, indicating that this technique accurately
reflects RNA synthesis.
View larger version (56K):
[in this window]
[in a new window]
|
Fig. 4.
A, DU145 cells were treated with RNase A
or actinomycin D (an RNA synthesis inhibitor) and compared with
untreated cells to demonstrate that the staining procedure for RNA
synthesis was specific. Cells were pulsed with BrU, and incorporation
of BrU was detected by indirect immunofluorescence using a
cross-reactive BrdUrd monoclonal antibody. B, RNA synthesis
in untreated DU145 cells and cells treated with GRO15B (control
oligonucleotide) or GRO29A (active oligonucleotide). Cells were treated
for 72 h and then pulsed with BrU and stained as described.
|
|
Fig. 4B shows the effects of GRO treatment on RNA synthesis
in DU145 cells. After incubation for 72 h in the absence or
presence of GROs, cells treated with GRO29A showed staining for BrU
incorporation similar to that for untreated and control
(GRO15B)-treated samples. Cells that had been treated in parallel and
pulsed with BrdUrd demonstrated negative staining for DNA synthesis in
the GRO29A-treated sample (shown in Fig. 3). Therefore, it appears that
RNA synthesis is still occurring in GRO29A-treated cells at a time when
DNA synthesis is arrested.
To determine global protein synthesis, the incorporation of
[35S]methionine was examined. Cells were incubated for
72 h in the absence or presence of GROs and pulsed with
[35S]methionine, and radioactivity associated with
trichloroacetic acid-precipitable material was determined by
scintillation counting. To normalize the data for differences in the
number of cells, the total protein content for each sample was
determined by the Bradford assay, and protein synthesis was expressed
as counts per µg of total protein. Parallel samples were pulsed with
BrdUrd in place of [35S]methionine to monitor DNA
synthesis. Fig. 5 shows that there was no
significant difference in the levels of global protein synthesis
between untreated cells and cells treated with GRO29A or GRO15B,
whereas DNA synthesis was completely inhibited by GRO29A (Fig. 3).
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 5.
Protein synthesis in untreated DU145 cells
and cells treated with GRO15B (control oligonucleotide) or GRO29A
(active oligonucleotide). Cells were treated for 72 h,
transferred to methionine/cysteine-free medium, and then pulsed with
35S-labeled methionine. Protein synthesis was determined by
scintillation counting of trichloroacetic acid-precipitable material
and was normalized by determining the total protein content of the
sample and expressing as counts/min per µg of total protein.
|
|
Induction of Apoptosis and Changes in Cell
Morphology--
Induction of apoptosis by GRO29A was suggested by the
appearance of cells with sub-G1 DNA content in the flow
cytometry cell cycle analysis (Fig. 1). We confirmed this finding by
carrying out terminal transferase dUTP nick-end label (TUNEL) staining of GRO-treated cells. Fig. 6A
shows that there was minimal TUNEL staining in untreated DU145 cells
(and in control 15B-treated samples, not shown). The sample treated
with GRO29A showed a punctate pattern of positive nucleoplasmic and
perinuclear staining, consistent with end labeling of fragmented DNA,
an indicator of apoptosis. Although some of the treated cells had the
classic features of apoptosis, at longer treatment times some cells had
morphology characterized by greatly enlarged nuclei and cytoplasm. As
shown in Fig. 6B, this was particularly evident in
MDA-MB-231 breast cancer cells that had been treated for 5 days with
GRO29A. The significance of this proportion of enlarged cells is
unclear at present but is consistent with our observation that protein
synthesis is ongoing, whereas DNA replication and cell division are
inhibited.
View larger version (57K):
[in this window]
[in a new window]
|
Fig. 6.
A, TUNEL staining to show induction of
apoptosis in DU145 prostate cancer cells by GRO29A. Cells were treated
for 72 h and stained as described. Cells treated with GRO15B (not
shown) had similar staining to the untreated samples. B,
phase contrast micrographs of MDA-MB-231 breast cancer cells to show
differences in morphology induced by treatment with an active GRO (29A)
compared with the control GRO (15B).
|
|
Inhibition of DNA Replication in Vitro--
The data presented
suggest that inhibition of DNA replication is a primary event
associated with cell response to treatment with antiproliferative GRO.
In cells, arrest of DNA synthesis can occur as a direct consequence of
interference with the organization or progress of the DNA replication
machinery. Alternatively, DNA replication could be arrested secondary
to other effects, such as alterations in signaling pathways.
To elucidate further the mechanism of GRO29A-induced inhibition of DNA
replication, we examined the effects of GROs on replication in an
in vitro system (29, 30). This system utilizes HeLa cell
extracts to replicate a plasmid containing the simian virus (SV40)
origin of replication and requires addition of SV40 large T antigen,
most probably to facilitate unwinding of the template (37, 38). By
using this assay, 400 nM GRO29A was reproducibly found to
inhibit DNA replication in vitro by about 90%
(incorporation of dCTP was 9.1 ± 1.6% compared with control),
whereas the same concentration of GRO15B had no effect (97.8 ± 7.1% dCTP incorporation compared with control). This assay was then
used to examine a series of GROs with a variety of antiproliferative
activities. These oligonucleotides have been described previously (9), with the exception of GRO26B, which is an active analog of GRO29A that
lacks the three 5'-thymidines. Fig.
7A shows representative results from the in vitro replication assays. The relative
inhibition is shown as a percentage of the control below the gel and
was determined by integration of the bands representing the closed circular (form I) replication product. The ability of these same oligonucleotides to inhibit the proliferation of HeLa cells in culture
is shown in Fig. 7B. Oligonucleotides 29A and 26B were found
to strongly inhibit cell proliferation, 26A showed intermediate inhibitory activity, and 15A, 15B, and CRO had little or no activity. (Although we reported previously (9) some antiproliferative activity in
HeLa cells for multiple doses of 15A, there was no activity when cells
were treated with a single dose.) The relative activities of the six
oligonucleotides in cells were almost identical to those observed in
the SV40 replication assay. Therefore, we conclude that the ability of
oligonucleotides to inhibit cellular proliferation is correlated with
their ability to inhibit DNA replication and that this blockade of
replication is due to a direct effect of GRO on the DNA replication
machinery.
View larger version (46K):
[in this window]
[in a new window]
|
Fig. 7.
A, effect of various oligonucleotides on
the efficiency of DNA replication in vitro. The indicated
oligonucleotide was mixed with the DNA template, and buffer and cell
extracts were added. The reaction was started by the addition of SV40
large T antigen and proceeded for 2 h at 37 °C. The T antigen
was omitted from the negative control (far left lane). The
reaction was stopped after 2 h at 37 °C. The bands indicated
represent covalently closed circular (I), nicked
(II), and linear (III) forms of DNA, and
RI indicates replication intermediates. The density of the
form I was determined with ImageQuant software, and the data were used
to calculate the percentage inhibition of replication compared with the
positive control (no oligonucleotide added), which is shown
below each lane. B, graph indicating the relative
number of viable HeLa cells (as determined by the absorbance at 570 nm
following MTT assay) 96 h after treatment with a single dose (12 µM) of the oligonucleotide indicated.
|
|
Effect of GRO29A on RPA-Nucleolin Interaction and SV40 Large T
Antigen Helicase Activity--
Inhibition of in vitro DNA
replication by GRO29A most probably occurs because the G-rich
oligonucleotide interferes with the assembly or progression of the
complex that carries out the replication process. Many of the proteins
that are required for efficient replication, both in the in
vitro system and in cells, have now been identified (39-41).
Based on the existing literature, we believe the most likely
replication-associated proteins to be modulated GRO29A are replication
protein A (RPA), which is known to bind to nucleolin (20), or the SV40
large T antigen, which can bind to G-quartet structures (42).
RPA is a single-stranded DNA-binding protein that plays a pivotal role
in DNA replication, repair, and homologous recombination. It is
possible RPA could bind directly to GRO29A, but simple single strand
DNA binding would not explain the inability of other oligonucleotides (for example, 15B and CRO) to inhibit replication. There is also the
possibility that GROs can modulate the interaction between RPA and
nucleolin (the putative GRO-binding protein), which was recently
reported by Daniely and Borowiec (20). These authors demonstrated that
this interaction occurs in vitro and found that addition of
exogenous nucleolin protein to the SV40 replication assay strongly
reduced origin unwinding and DNA replication in this system. This was
presumed to be a result of nucleolin binding to and inactivating RPA,
because replication could be rescued by addition of extra RPA. It was
also reported that heat shock of HeLa cells caused a redistribution of
nucleolin from the nucleoli to the nucleoplasm, and the authors (20)
proposed that this redistribution and the resultant binding of
nucleolin to RPA was a mechanism for repression of chromosomal
replication in response to cell stress. If binding of nucleolin to
GRO29A caused a conformational change that enhanced the affinity of
nucleolin for RPA, this would explain the inhibition of replication
both in vitro and in cells. To test this hypothesis, we
carried out immunoprecipitation of HeLa cell extracts with an antibody
to the 14-kDa subunit of RPA in the absence or presence of various
GROs. In accord with Daniely and Borowiec (20), we could detect
nucleolin in the immunoprecipitated proteins by Western blotting (data
not shown). However, the presence of GROs at 100-1000 nM
final concentration (which was sufficient to inhibit in
vitro replication) did not significantly alter the amount of
nucleolin precipitated with anti-RPA antibody. Treatment of HeLa
cells with GRO29A also did not cause significant redistribution of
nucleolin from the nucleoli to the nucleoplasm (as assessed by
immunofluorescent microscopy, data not shown). Therefore, there is no
evidence to support the hypothesis that the nucleolin-RPA interaction
is modulated by GRO29A.
An alternative mechanism for the inhibition of DNA replication in the
SV40 system is the inhibition of large T antigen helicase activity by
GRO29A. The ability of T antigen to unwind structures containing
G-quartets has been documented previously (42); therefore, we
postulated that the presence of an excess of G-quartet containing oligonucleotide could sequester the helicase activity and prevent template unwinding and replication. To test this hypothesis we carried
out T antigen helicase assays in the presence of various oligonucleotides. Because T antigen has a single-stranded DNA binding
domain, it is important to demonstrate that any inhibition by
oligonucleotides is specific. Therefore, we first compared inhibition
of helicase activity by an active GRO (29A) and a nonspecific oligonucleotide (NS). Fig. 8A
shows that GRO29A is significantly more active than NS at inhibiting
helicase activity. To compare the relative ability of the
oligonucleotides used in the replication assay to inhibit helicase
activity, we carried out similar experiments with 50 nM
final oligonucleotide concentration (at which there was a clear
difference between GRO29A and NS). Fig. 8B shows the results
of these experiments. Oligonucleotides 29A, 26B, and 26A were active in
inhibiting unwinding by T antigen, whereas 15A, 15B, and CRO were not.
The inactivity of CRO (which is the same length as 29A) shows that
inhibition of helicase activity was not simply a function of
oligonucleotide length. Comparison of the data in Fig. 8B
with those shown in Fig. 7 indicates a good correlation between
inhibition of helicase activity, inhibition of in vitro DNA
replication, and inhibition of cell growth.
View larger version (79K):
[in this window]
[in a new window]
|
Fig. 8.
A, helicase assay showing unwinding by
SV40 large T antigen of a synthetic oligonucleotide substrate
representing a replication fork. Lanes marked SS and
DS are markers representing unwound (single strand 55A) and
wound (partial duplex 55A/B) substrate in the absence of T antigen.
NI is a "no incubation" control lane containing
substrate and T antigen without incubation at 37 °C. In all other
lanes, the helicase reaction was allowed to proceed for 15 min at
37 °C without competitor (0) or in the presence of
unlabeled competitor oligonucleotides 29A or NS at the concentrations
shown. B, helicase assay showing inhibition of substrate
unwinding by T antigen in the presence of 50 nM
concentration of competitor oligonucleotide, as indicated.
|
|
| |
DISCUSSION |
GROs are a new type of antiproliferative oligonucleotide with
considerable potential as therapeutic agents for cancer. Although the
activity of GROs is known to correlate with their ability to bind to
nucleolin protein, the precise mechanism by which they exert their
antiproliferative effects is unknown. Because nucleolin is involved in
many aspects of cell growth, proliferation, and apoptosis, knowledge of
a putative target protein does not necessarily identify the processes
that are affected by GROs. Further information regarding the pathways
affected by GROs would facilitate the design of agents that act by a
similar mechanism that may be even more active or have improved
pharmacological features, compared with oligonucleotides. Therefore, in
this study we proceeded to explore the mechanism of GROs by
examining their effects on cellular processes.
The conclusions of our studies are that treatment of cells with
antiproliferative GRO29A causes an arrest of cell cycle progression in
S phase and an inhibition of DNA replication. Because DNA synthesis is
affected before RNA and protein synthesis, we have concluded that this
is a primary cause of proliferation inhibition. Furthermore, because
GROs can also inhibit DNA replication in a cell-free assay, we infer
that the action of the GROs results directly from an effect on a
protein (or proteins) involved in DNA replication. To investigate the
hypothesis that GRO effects are mediated by an inhibition of DNA
synthesis, we have used an in vitro DNA replication assay.
This assay is based on the replication of a 7.4-kilobase pair circular
genome that contains an SV40 origin of replication (ori),
originally described by Li and Kelly (29, 30) and modified by Roberts
and Kunkel (31). The reaction is initiated by SV40 large T antigen,
which is required to recognize and unwind the ori. Proteins
in human cell extracts carry out all subsequent steps. Inhibition of
replication by GRO29A could be mediated either by damaging the template
or by modulating the activity of a protein (or proteins) that is
required for replication. The former mechanism is unlikely, because the
replication products of damaged templates show severe inhibition of
form I DNA but relatively greater amounts of nicked and linear DNA (32,
43). The GRO29A examined in this study severely inhibited the synthesis
of all forms of DNA, suggesting interference with the proper function
of the complex that carries out DNA replication (known as the replisome
or synthesome). Therefore, we investigated the effects of GRO29A on the
replication-associated proteins most likely to be modulated. These were
the SV40 large T antigen, which is known to bind to and unwind
G-quartet structures (42), and RPA, which is known to bind to nucleolin
(20), the putative GRO-binding protein (9). Although we could observe the interaction between nucleolin and RPA, we found that this was not
significantly affected by the presence of GROs. On the other hand, the
ability of T antigen to unwind a synthetic substrate representing a
replication fork was strongly inhibited by GRO29A. Moreover, for a
series of six oligonucleotides, the relative activity in inhibiting T
antigen helicase mirrored the relative activity in inhibiting DNA
replication in vitro and also in inhibiting tumor cell
proliferation. Therefore, it would appear that the antiproliferative
effects of GROs on cancer cells are mediated by inhibition of DNA
replication, which in turn may be related to inhibition of helicase
activity. Of course, SV40 T antigen is not normally present in human
cells, but GRO29A could also be an inhibitor of a human replicative
helicase, which will most likely share similar features with the viral
T antigen. The identity of the human replicative helicase is still not
certain, but a hexameric complex of proteins known as minichromosome
maintenance has been reported to have helicase activity in
vitro and is generally thought to be a good candidate (44, 45).
The ability of this complex to bind to or unwind G-quartets has not yet
been reported.
G-quartet unwinding has been described previously (46-48) for a number
of helicases, but to our knowledge, this is the first report that the
presence of G-quartet structures can prevent a replicative helicase
unwinding its double-stranded substrate. Inhibition of helicase
activity has been reported for several DNA-binding agents, including
many antitumor antibiotics such as anthracyclines (49-56). In these
cases, it is most likely that inhibition is caused by the formation of
a strong complex between the DNA-binding ligand and the template DNA,
which impede the action of the helicase (49). There is considerable
evidence that inhibition of helicase activity by such compounds may
play some role in their anticancer activity (57), but the effect of
helicase inhibition in cancer cells has not been extensively studied.
Nucleolin has been identified by us as a GRO-binding protein (9), and
by several other groups (58-61) as a G-quartet-binding protein.
Although the results in this paper do not clearly define a link between
the molecular effects of GROs and their binding to nucleolin, they
point to an effect by inhibition of helicase activity and DNA
replication. Therefore, it is interesting to note that nucleolin itself
has been reported to have helicase activity and is also known as DNA
helicase IV (12, 22). In addition, nucleolin has been shown to interact
with at least three components of the DNA replisome complex, namely RPA
(20), topoisomerase I (62), and poly(ADP-ribose) polymerase
(63).
Our future studies will focus on identification of cellular helicases
that may be inhibited by GRO29A, as well as clarification of the role
of nucleolin in GRO activity. Such studies could lead to valuable
insights into the role of nucleolin in DNA replication, as well as
further elucidation of the molecular mechanisms of GRO effects.
,
,
TOP
ABSTRACT
INTRODUCTION
