Download Analysis of Events Associated With Cell Cycle Arrest at G2 Phase

Analysis of Events Associated With Cell Cycle
Arrest at G2 Phase and Cell Death
Induced by Cisplatin
Christine M. Sorenson, Michael A. Barry, Alan Eastman*
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
DNA is the accepted target for cisplatin, but recent evidence
has shed doubt on DNA synthesis as the critical process.
L1210/0 cells incubated for 2 hours with cisplatin progress to
the G2 phase of the cell cycle and are arrested there for several
days. They then either progress in the cell cycle or die. In cells
that eventually die, total transcription, polyadenylated
[poly(A)+] RNA synthesis, and protein synthesis were markedly inhibited only after 48 hours. Nicotinamide adenine
dinucleotide (NAD) and adenosine triphosphate (ATP) levels
decreased after 3 days. Cell membrane integrity was lost after
4 days. These results demonstrate that cells can be lethally
damaged, yet continue to undergo apparently normal metabolic activities for several days. In a previous study, DNA
double-strand breaks were detected after 1 day. We now
show that by 2 days, breaks are visible as fragmentation in the
nucleosome spacer regions of chromatin. This type of damage
is consistent with cell death occurring by the process of
apoptosis. Cell shrinkage and morphology were also consistent with this type of cell death. The slow cell death reported
here appears to occur at the G2/M transition and may involve
events that normally occur at this stage of the cell cycle. These
results demonstrate the importance of DNA degradation as
an early and possibly essential step in cell death. [J Natl
Cancer Inst 82:749-755,1990]
ities to cisplatin, we observed that any inhibition of DNA
synthesis was related to the applied drug concentration and not to
the degree of toxicity. Very sensitive cells progressed at a normal
rate to the G2 phase and subsequently died. In contrast, more
resistant cells could tolerate higher drug concentrations that did
slow DNA synthesis.
These observations led to the two questions addressed here:
why are cells arrested in the G 2 phase, and by what mechanism do
they subsequently die? It should be emphasized that, although G2
arrest appears to be a prerequisite for cell death (except at very
high drug concentrations), all such arrested cells do not die. At
minimally toxic concentrations of cisplatin, cells may eventually
bypass the block and return to normal cycling. Hence, there are
two possible fates for a cell arrested in the G2 phase: survival or
death. It is important to determine how a cell regulates the
outcome.
We initially hypothesized that cells incubated with cisplatin
were arrested in the G2 phase by an inhibition of transcription,
specifically, an inability to produce the full-length messenger
RNA (mRNA) needed for passage to mitosis. Earlier studies had
investigated only the quantity, rather than the quality, of RNA
synthesis (4-6). We therefore investigated inhibition of transcription and other parameters that might be involved in cell death.
Previously, we had shown DNA double-strand breaks to be the
earliest change detected in cells destined to die (7). Here we
demonstrate that these breaks occur in the nucleosome spacer
region of chromatin DNA, giving rise to "nucleosome ladders" by
gel electrophoresis. Such breaks are characteristic of cells dying
by the mechanism of apoptosis (9).
Cisplatin has been shown to be an effective antineoplastic
agent in the treatment of a variety of tumors (/). DNA has been
implicated as the critical target for cytotoxicity (2). Most of the
Materials and Methods
damage is due to DNA-intrastrand cross-links: DNA-interstrand
and DNA-protein cross-links represent less than 1 % of the total Cell Culture
platination of DNA (3). The relative contribution of each lesion to
LI210/0 cells were maintained in an exponential suspension
toxicity is still in contention.
culture at 37 °C in a humidified atmosphere of 5% CO 2 -95% air
Inhibition of DNA synthesis has been observed in a variety of
cells following platination (4-6). DNA synthesis has also been
reported to be more sensitive to cisplatin than either RNA or
Received November 13, 1989; revised January 25, 1990; accepted February 2,
protein synthesis (2). Thus inhibition of DNA synthesis has been
1990.
logically envisioned as the critical step in toxicity. We have
Supported by Public Health Service grant CA-36039 from the National Cancer
previously questioned this assumption (7,8). Rather than being Institute, National Institutes of Health, Department of Health and Human
arrested in the S phase of the cell cycle, as would be expected if Services.
C. M. Sorenson, McArdle Laboratory for Cancer Research, University of
DNA synthesis were inhibited, cells were arrested in the G2 phase
before dying. Although slowed DNA synthesis may occur during Wisconsin, Madison, WI.
M. A. Barry, A. Eastman, Department of Pharmacology and Toxicology,
progression to the G2 phase, this did not correlate with toxicity. In Dartmouth Medical School, Hanover, NH.
Chinese hamster ovary cells either proficient or deficient for
Correspondence to: Alan Eastman, Ph.D., Department of Pharmacology,
DNA repair, and therefore exhibiting markedly different sensitiv- Dartmouth Medical School, Hanover, NH 03756.
Vol. 82, No. 9, May 2, 1990
ARTICLES
749
label ([35S]methionine-cysteine; ICN Biomedicals, Inc., Costa
Mesa, CA) and incubated at 37 °C for 1 hour. The cells were then
centrifuged, rinsed twice with cold PBS containing 200 \yM
methionine, and resuspended in 100 u-L of PBS containing 200
\JA methionine. TCA was added, the precipitate was collected,
and radioactivity was assayed as described above.
DNA Synthesis
ATP and NAD Levels
At each time point, 106 cells were resuspended in 2 mL of fresh
medium containing 0.5 u,Ci of [3H]thymidine (6.7 Ci/mmol;
New England Nuclear Corp., Boston, MA) and incubated for 30
minutes to produce radiolabeled DNA. The cells were then
centrifuged, rinsed twice with cold phosphate-buffered saline
(PBS), and resuspended in 100 \iL of PBS, and 100 y.L of a
solution of salmon sperm DNA [500 jig/mL in 20 mM edetic acid
(EDTA)] was added. To each sample was added 5 mL of 10%
ice-cold trichloroacetic acid (TCA). This was incubated on ice for
15 minutes, and the precipitate was collected by filtration through
2.4-cm-diameter Whatman glass microfiber filters (GF/C; Whatman International Ltd., Maidstone, England). The filters were
rinsed with 10% TCA followed by ethanol, solubilized with 0.5
mL of NCS tissue solubilizer (Amersham Corp., Arlington
Heights, IL) for 30 minutes at 37 °C, and neutralized, and the
radioactivity was determined with a Beckman scintillation
counter.
At each time point, 107 cells were harvested and rinsed with
Hanks' balanced salt solution containing 1 mM sodium phosphate, 1 mM potassium phosphate, and, as phosphatase inhibitors, 5 mM sodium fluoride and 5 mM sodium glycerol phosphate
(11). The cell pellet was then resuspended in 100 \LL of 0.4 M
perchloric acid containing phosphatase inhibitors and incubated
on ice for 30 minutes. The cell lysate was spun for 10 minutes at
4 °C, and 16 jxL of potassium bicarbonate was added to neutralize
the supernatant. Samples were analyzed by anion exchange
high-pressure liquid chromatography on a Whatman Partisil 10
SAX column (12). Quantitation was performed by comparison of
peak heights with a standard curve obtained from injection of
known amounts of nicotinamide adenine dinucleotide (NAD) or
adenosine triphosphate (ATP).
Total RNA Synthesis
Fresh medium containing 0.5 jiCi of [3H]uridine (28.5 Ci/
mmol; New England Nuclear Corp.) in 2 mL was added to 106
cells and incubated for 30 minutes at 37 °C to produce radiolabeled RNA. The cells were then centrifuged, rinsed twice with
cold PBS containing 200 \iM undine, and resuspended in 100 \LL
of PBS containing 200 \iM uridine, and 100 (xL of salmon sperm
DNA (500 M-g/mL) was added. TCA was added, the precipitate
was collected, and the radioactivity was assayed as described
above.
Poly(A)+ RNA Synthesis
A total of 2 x 106 cells were resuspended in 4 mL of fresh
medium containing 2 jiCi of [3H]uridine and incubated at 37 °C
for 2 hours. The cells were centrifuged and rinsed twice with cold
PBS containing 200 \sM uridine. Polyadenylated [poly(A)+]RNA was isolated by a modification of the method of Badley
et al. (70). Unlabeled L1210/0 poly(A)+ RNA (80 ng) was added
to the equilibrated oligodeoxythymidylate [oligo(dT)]-cellulose
in batch fashion and incubated at room temperature for 15
minutes. The L1210/0 cell lysate was added, and the incubation
was continued for an additional 45 minutes. The oligo(dT)-cellulose was briefly centrifuged, the supernatant removed, and
binding buffer added. This sequence was repeated until the
absorbance at 260 nm (A260) of the supernatant was less than
0.05. The poly(A) + RNA was eluted with 1 -mL aliquots of sterile
water and collected by centrifugation, and the radioactivity in
each fraction was determined.
Protein Synthesis
A total of 106 L1210/0 cells were resuspended in 2 mL of
methionine-free McCoy's medium containing 1 [iCi of Tran35S-
750
DNA Degradation
Cells were analyzed by a gel electrophoresis method adapted
from Eckhardt (13). A 125-mL volume of 2% agarose in TBE (89
mMTris, 89 mM boric acid, 2.5 mM EDTA; pH 8.0) was poured
into a minigel support with the comb. After the gel solidified, the
portion of the gel above the comb (2 cm) was removed. This space
was filled with 16 mL of 0.8% agarose-2% sodium dodecyl
sulfate (SDS) in TBE to which 1.25 mg of proteinase K per mL
was added after the agarose had cooled below 50 °C. Each pellet
of 106 cells was suspended in 15 JJLL of sample buffer (10 mg of
ribonuclease/mL, 15% Ficoll 70, 0.01% bromophenol blue in
TBE), and the suspended pellet was transferred to a well. Cell
lysis began in the sample buffer and was completed during
electrophoresis at 20 V for 1 hour. DNA fragments were then
separated by electrophoresis for 3 hours at 90 V. Molecular
weight standards (Msp I-digested pBR322; New England Biolabs, Beverly, MA) were subjected to electrophoresis in adjacent
lanes.
Following electrophoresis, the gel was rinsed in distilled water
and incubated overnight at room temperature with gentle shaking
in 100 mL of TE (10 mM Tris, 1 mM EDTA; pH 8.0) containing
2 mg of ribonuclease A. This procedure removed the SDS and
contaminating RNA that had comigrated with the smaller nucleosome fragments. The gel was rinsed in distilled water and
stained for 30 minutes in 100 mL of water containing 50 ^g of
ethidium bromide. The gel was rinsed, destained in water for 4
hours, and photographed in ultraviolet light.
The approximate amount of DNA detected in these gels was
assessed by laser densitometry of the photographic negatives.
These values were compared with a standard curve produced by
electrophoresis of a known amount of sheared DNA under
identical conditions. Minimally detectable degradation represented about 50 ng, while the maximum degradation detected in
these studies was about 500 ng. The majority of DNA always
remained in the well and could not be quantitated by densitometry. The total DNA applied to each well was about 10-20 u,g; the
Journal of the National Cancer Institute
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
in McCoy's 5a (modified) medium (GIBCO Laboratories, Grand
Island, NY) supplemented with 15% calf serum, penicillin,
streptomycin, and amphotericin B (Fungizone). In all experiments, L1210/0 cells were incubated in various concentrations of
cisplatin (Bristol Laboratories, Syracuse, NY) for 2 hours at
37 °C. Cell size was determined on a Coulter Channelyzer 256.
latter value reflects cells with twice the normal content of DNA as
a result of arrest in the G2 phase.
Scanning Electron Microscopy
Electron microscopy was performed on LI210/0 cells at various time intervals following incubation with cisplatin. Approximately 107 cells were pelleted and resuspended in 1 mL of 2%
glyceraldehyde in 0.1 M cacodylate buffer (pH 7.4) and stored at
4 °C. The cells were centrifuged with a cytocentrifuge at 500 rpm
for 30 minutes onto a round coverslip coated with 0.1% polyL-lysine. The coverslip was postfixed with 1% osmium tetroxide
in 0.1 M phosphate buffer (pH 7.4), dehydrated through an
ascending ethanol series, and critical-point-dried with Freon 13.
The coverslips were mounted on specimen stubs, sputter-coated
with gold, and examined in a Philips 515 scanning electron
microscope.
[3H] THVMIDINE
MO
0
2
4
6
a
0
MCUBATM3N TIME (Days)
Figure 1. Inhibition of macromolecular synthesis in L1210/0 cells at various
times following 2 hr of incubation with cisplatin concentrations of 0.12 (o), 0.25
(•), 0.5 (A), 1 (A), 2 (•), and 4 (•) (jLg/tnL. Incorporation of [3H]thymidine,
[3H]uridine, and [ S]methionine into TCA-precipitable material represents
DNA, RNA, and protein synthesis, respectively.
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
Results
"°C
Cell Survival
The various parameters by which the toxicity of cisplatin has
been measured in L1210/0 cells have been discussed previously
(7). Following 2 hours of drug treatment, the concentration
inhibiting growth by 50% over a 3-day period was 0.7 jjug/mL.
However, this value does not adequately assess cells that are
arrested in the G2 phase for several days and then recover. A more
accurate assessment of toxicity was obtained by measuring trypan
blue dye exclusion for up to 14 days after drug treatment. The
time to maximum loss of membrane integrity was 6-8 days. By
this criterion, a cisplatin concentration of 0.7 jig/mL killed less
than 10% of the cells. Approximately 50% of the cells were killed
at 2 n-g/mL. This was supported by flow cytometry. At 0.7
(jig/mL, the cells were arrested maximally in the G2 phase on day
1 and then recovered over the following 2 days, with no apparent
cell disintegration. At 2 ^g/mL, more than 80% of the cells were
arrested in the G 2 phase by 2 days; after 4-6 days, approximately
half of the cells were observed as debris. Higher concentrations of
cisplatin led to virtually complete disintegration of cells by 6
days. Therefore, drug concentrations of less than 2 (ig/mL are
considered minimally toxic, 2 u-g/mL represents an intermediate
toxicity, and higher concentrations are significantly toxic.
Inhibition of Macromolecular Synthesis
We investigated several parameters to elucidate the sequence
of events occurring in cells destined to die following incubation
with cisplatin. The timing of such events could then be compared
with the maximal G2 arrest on day 2 and the loss of membrane
integrity around day 6. DNA, RNA, and protein synthesis studies
were performed to determine when inhibition of these processes
occurs following platination.
At various time intervals after drug treatment, cells were
incubated with [3H]thymidine, [3H]uridine, or [35S]methionine,
and incorporation was assayed as radioactivity in the acidprecipitable fraction (fig. 1). A significant inhibition of DNA
synthesis was observed even at the earliest time point of 12 hours.
At minimally toxic drug concentrations, DNA synthesis returned
to near-control levels by 2 days; intermediate concentrations
required 5 days. The inhibition of DNA synthesis between days 1
Vol. 82, No. 9, May 2, 1990
and 5 correlated with arrest in the G2 phase. At toxic drug
concentrations, DNA synthesis was permanently suppressed by
2 days.
A significant inhibition of total RNA synthesis was not observed until 2 days following incubation with cisplatin. At
minimally toxic drug concentrations, recovery of total RNA
synthesis was noted by 4 days. This corresponded to the transient
arrest in the G2 phase of the cell cycle, followed by the reemergence of a cycling population. Total RNA synthesis continued to decrease after 2 days at toxic drug concentrations and was
permanently suppressed by 3 days.
The trend of inhibition observed with newly synthesized
protein was similar to that observed with total newly synthesized
RNA. A significant inhibition of protein synthesis was observed
at 2 days. This decrease in synthesis was continuous at toxic
concentrations, while a transient decrease was observed at minimally toxic concentrations.
We hypothesized that cells might be arrested in the G2 phase
due to truncation of newly synthesized mRNA required for
mitosis. However, no reduction in the amount of new poly(A)+
RNA was observed until 2 days (fig. 2). This was after a
significant G2 arrest had occurred (7). The pattern of inhibition of
mRNA synthesis closely followed that observed with total RNA.
At minimally toxic drug concentrations, only a slight decrease of
poly(A)+ RNA synthesis was observed during the 4-day experiment. At toxic drug concentrations, a continual decrease in newly
synthesized poly(A)+ RNA was noted.
DNA Degradation
We had previously observed that formation of DNA doublestrand breaks appeared to be the earliest change detected in cells
destined to die (7). We pursued this observation further by using
gelelectrophoresis to resolve nucleosome fragments (fig. 3). This
modified technique permits rapid and sensitive detection of DNA
degradation but results in some blurring of the smaller bands as
compared with techniques that require an initial purification of
the DNA. The limit of detection on these gels is about 50 ng of
ARTICLES
751
MOr
INCUBATION TIME (DAYS)
A.
Z
<
120
0
<^100
Figure 2. Inhibition
of
synthesis
of
poly(A)+ RNA in
L1210/0 cells at various times following 2
ir of incubation with
the indicated concentrations of cisplatin.
22
ol
82 s^
UJ °
5
a.
\\A\ \ V / /^X.»0.25 >jg/ml
80
Ol 0 5
\
\
60
\
w
z
*»>
\
\
40
t—
Ul
\
2o
20
1
i
2
i
i
3
4
1
I
i
k
m
1
L
^i
i
1
1
k
Ik
1
Figure 4. Changes in
size of LI 210/0 cells at
various times following 2 hr of incubation
with the indicated concentrations of cisplatin.
i
2
oO
^~» 4 pg/ml
i
i
3
i
k
<
o:
^~* 1 pg/ml
\
£
I
2
o
D:
Q
INCUBATION TIME (Days)
0
DNA, or about 0.5% of that added to the wells. At 1 day, little if
any fragmentation was observed at any drug concentration. By 2
days, however, nucleosome fragments were observed following
toxic drug concentrations, and the intensity increased at later
times. At these drug concentrations, a characteristic nucleosome
ladder consisting of multimers of approximately 180 base pairs
was observed. At intermediate concentrations, this ladder became most pronounced at 4 days. At minimally toxic drug
concentrations, no nucleosome fragments were observed. Hence,
the appearance of nucleosome fragments correlated with cell
death and occurred several days before loss of trypan blue dye
exclusion.
14
24
24
24
24
CELL VOLUME (p liter)
size increased markedly at 2 days. This correlated with the
passage of cells into the G2 arrest. By 4 days at the higher drug
concentrations, a significant proportion of the cells had shrunk to
become smaller in size than the control cells. This was particularly noticeable in the cells incubated with the most toxic
concentration of cisplatin.
Cell Size
Since it has been reported that shrinkage occurs in cells dying
by apoptosis (9), we measured cell size following incubation with
cisplatin (fig. 4). At minimally toxic concentrations of cisplatin,
there was only a slight increase in cell size, but this was reversed
by day 3. However, at higher drug concentrations, the distribution of cell size became more heterogeneous, and the mean cell
1 day
2 days
3 days
4 days
pg/ml
bp
• 540
• 360
•»• 180
Figure 3. DNA degradation in L1210/0 cells following 2 hr of incubation with
cisplatin. Cells were harvested after 1-4 days, and DNA degradation was
analyzed by gel electrophoresis. The molecular weights were obtained from
adjacent lanes containing Msp I-digested pBR322 DNA standards.
752
Figure 5. Scanning electron micrographs of LI210/0 cells at various times
following 2 hr of incubation with 8 (ig of cisplatin per mL. A: control; B: I day;
C: 2 days; D: 3 days; E: 4 days; F: 6 days. The white bars at the bottom of each
picture represent 10 (un.
Journal of the National Cancer Institute
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
i
NAD
ATP
0.5 pg/ml
2 pg/ml
B pg/ml
INCUBATION TIME (Days)
Scanning Electron Microscopy
Another factor that has been associated with apoptosis is
changes on the cell surface, such as convolution and formation of
blebs. L1210/0 cells were incubated with 8 \ig of cisplatin per mL
(toxic concentration) for 2 hours and incubated posttreatment for
up to 6 days, and electron microscopy was performed (fig. 5).
Undamaged cells showed characteristic roughened surfaces.
However, following incubation with cisplatin, marked changes
occurred; most pronounced was the porous appearance at 4-6
days and occasionally for cells at earlier time points. This
corresponds to the timing of loss of trypan blue dye exclusion in
these cells (7). Prior to this change, some cells were seen to have
blebs on their surfaces. Such blebs are usually very transient, as
the observable changes in apoptosis can occur in less than 15
minutes (9). Accordingly, very few cells would be expected to
show this phenomenon simultaneously.
ATP and NAD Levels
Increased poly(ADP-ribosyl)ation is commonly associated
with DNA strand breaks (14). (ADP = adenosine diphosphate.)
NAD serves as a substrate in this reaction, thus causing NAD
pools to decrease. ATP levels subsequently decline. Alternatively, loss of NAD and ATP could precede DNA degradation,
indicating loss of osmoregulation as an early step in cell death. At
minimally toxic drug concentrations, no significant decrease in
ATP or NAD levels was observed during the 5-day experiment
(fig. 6). However, at toxic drug concentrations, ATP and NAD
levels dramatically decreased at 3 days. Because these changes
occurred after DNA degradation, they were a result of, and not a
contributor to, the damage.
Discussion
To date, very little is known about the processes involved in
cell death. The general consensus suggests inhibition of DNA
synthesis as the critical step in cisplatin-induced cytotoxicity. Our
earlier work had questioned this dogma (7,8). We therefore
sought an alternative explanation. In this report we have attempted to profile the events that occurred as a result of cisplatin
Vol. 82, No. 9, May 2, 1990
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
Figure 6. Changes in NAD and ATP levels in L1210/0 cells at various times
following 2 hr of incubation with the indicated concentrations of cisplatin.
treatment. In particular, we have investigated events that may be
associated with drug-induced G2 arrest and the subsequent degradation of DNA that appears to be an essential step in cell death.
.. -We initially hypothesized that cells are arrested in the G2 phase
due to an inhibition of transcription, specifically, an inability to
produce the full-length mRNA needed for passage to mitosis.
First, we compared the effect of drug treatment on DNA, RNA,
and protein synthesis. Consistent with previous reports (2), we
observed that DNA synthesis was the most sensitive to drug
treatment in that it was the first suppressed. Since the cells still
survived at the lower concentrations, suppression of DNA synthesis did not predict cell death. These curves show an initial
suppression of DNA synthesis during the S phase. This is
followed by recovery, reflecting passage into the G2 phase, and a
subsequent arrest there. Recovery of DNA synthesis at 3-5 days
represents passage through the next cell cycle. At the highest
concentrations, the cells were dying, and no recovery was
observed. In contrast, neither RNA synthesis nor protein synthesis was suppressed until about 2 days after drug treatment.
These experiments, as well as those previously reported (2),
did not take into consideration that the quality of transcription
could be altered; that is, transcription could terminate on reaching
an adduct in DNA, but then reinitiate from the beginning. The
result could be the same quantity of RNA produced, but with
much of it incomplete. We therefore measured the quantity of
newly synthesized poly(A)+ RNA as an indicator of completion
of transcription. Again, we observed no significant suppression
of synthesis until 2 days after drug treatment. Therefore, G2 arrest
cannot be attributed to detectable changes in transcription.
A recent explanation for drug-induced G2 arrest comes from
the yeast RAD9 mutant (75). Cells with this phenotype were
unable to be arrested in the G2 phase following introduction of
DNA damage; they continued to cycle and died. The RAD9 gene
product was demonstrated to be essential for arrest of cell division
following DNA damage. It is possible that RAD9 is involved in
the surveillance mechanism previously hypothesized by Tobey
(16).
The question still remains as to the cause of a cell's demise
following incubation with cisplatin. There are two known mechanisms of cell death: necrosis and apoptosis (9). The latter
pathway, also known as programmed cell death, occurs during
metamorphosis, differentiation, and general cell turnover. Apoptosis is also the cause of death in thymocytes exposed to
glucocorticoids (77) or x rays (18). One of the characteristics of
apoptosis is chromatin condensation associated with DNA degradation, giving rise to internucleosomal cleavage products. The
endonuclease involved is of nonlysosomal origin, because the
membranes remain intact. In addition, new protein synthesis is
required. In the LI210/0 cells studied here, we detected the
formation of nucleosome ladders at toxic concentrations within
2 days of drug treatment. The intensity of the nucleosome ladder
increased with time and drug concentration. At nontoxic drug
concentrations, no such degradation was observed. It is probable
that the DNA degradation previously observed as DNA doublestrand breaks by neutral elution at 24 hours (7) progresses and by
48 hours is observable as internucleosomal cleavage products.
These events occurred long before loss of membrane integrity.
The cells did not begin to take up trypari blue until 4 days after
drug treatment (7).
ARTICLES
753
We also monitored NAD and ATP levels following incubation
of the cells with cisplatin. NAD is a substrate for poly(ADPribosyl)ation of proteins. This protein modification is stimulated
by DNA strand breaks and generally suppresses protein activity
(20). Extensive or unrepaired DNA strand breaks reportedly
cause poly(ADP-ribose) polymerase to be continuously active.
This causes depletion of NAD and, subsequently, depletion of
ATP during attempts to replenish the NAD pool. Such alterations
are thought to account for the rapid cell death that occurs before
DNA repair takes place (20,21). However, we were able to detect
significant decreases in NAD and ATP only at toxic concentrations and 3 days following drug treatment. This was 1 day after
the detection of DNA double-strand breaks. These data suggest
either that significant poly(ADP-ribosyl)ation was a delayed
response or that these breaks did not stimulate poly(ADPribosyl)ation. An endonuclease potentially involved in apoptosis
is itself inhibited by poly(ADP-ribosyl)ation (22). Reduced NAD
could therefore lead to its activation. However, since the changes
we observed occurred after DNA degradation, they were an effect
of, and not a contributor to, the damage.
The results presented here affirm the importance of DNA
degradation as an early and presumably essential step in cell
death. The inhibition of transcription and protein synthesis
occurred at about the same time as DNA degradation, so it is not
possible to confirm which is cause or effect. Loss of NAD and
ATP occurred later, presumably as a consequence of these
changes. These results clearly demonstrate that cells can be
lethally damaged, yet continue to undergo apparently normal
metabolic activities for several days. These results are also
different from those reporting rapid cell death. However, similar
events appear to occur in both types of cell death; the difference is
presumably in the signal transduction pathway. The slow cell
death reported here appears to occur at the G/M transition and
may therefore involve events that normally occur at this stage of
the cell cycle. This is. consistent with the idea that apoptosis is
related to chromatin^ondensation.
Tremendous progress has been made recently in understanding
754
the events essential for passage of cells into mitosis. The major
component appears to be the kinase coded for by the cdc2 gene
(histone HI kinase) that is activated and inactivated by specific
phosphorylations and by a variety of associated proteins (23,24).
One essential protein is cyclin B, the only protein whose synthesis
is required for passage into mitosis (14). Two regulators of
histone HI kinase are defined in yeast by the cdc25 and weel
genes; the former activates the kinase, while the latter suppresses
it. Mutants with an imbalance in these genes can undergo a
"mitotic catastrophe" reminiscent of premature chromatin condensation (25). It seems highly probable that these events are
involved as part of the signal transduction pathway during cell
death by apoptosis. The events reported in this paper are not
restricted to DNA-damaging agents, as many other toxic agents
induce the same events (26). Therefore, there will probably be
multiple steps within the pathway at which drug effects can be
manifested.
References
(1) LOEHRER PJ, EINHORN LH: Cisplatin. Ann Intern Mcd 100:704-713, 1984
(2) ROBERTS JJ, THOMSON AJ: The mechanism of action of antitumor platinum
compounds. Prog Nucl Acid Res Mol Biol 22:71-133, 1979
(3) EASTMAN A: The formation, isolation and characterization of DNA adducts
produced by anticancer platinum complexes. Pharmacol Ther 34:155-166,
1987
(4) HARDER HC, ROSENBERG B: Inhibitory effects of anti-tumor platinum
compound on DNA, RNA and protein synthesis in mammalian cells in vitro.
IntJ Cancer 6:207-216, 1970
(J) HOWLE JA, GALE GR: Cw-Dichlorodiammineplatinum(II): Persistent and
selective inhibition of deoxyribonucleic acid synthesis in vivo. Biochera
Pharmacol 19:2757-2762, 1970
(6) SALLES B,
BUTOUR JL,
MACQUET JP: a'.s-Pt(NH3)2Cl2
and trans-
Pt(NH3)2Cl2 inhibit DNA synthesis in cultured LI210 leukemia cells.
Biochem Biophys Res Commun 112:555-563, 1983
(7) SORENSON CM, EASTMAN A: The mechanism of c«-diamminedich]oroplat-
inum(II) induced cytotoxicity: The role of G2 arrest and DNA double-strand
breaks. Cancer Res 48:4484-4488, 1988
(8) SORENSON CM, EASTMAN A: Influence of c«-diamminedichloroplatinum(II)
on DNA synthesis and cell cycle progression in excision repair proficient
and deficient Chinese hamster ovary cells. Cancer Res 48:6703-6707,1988
(9) WYLUE AH, KERR JFR, CURRIE AR: Cell death: The significance of
apoptosis. Int Rev Cytol 68:251-306, 1980
(10) BADLEY JE, BISHOP GA, ST. JOHN T, ET AL: A simple, rapid method for the
purification of poly A + RNA. BioTechniques 6:114-116, 1988
(//) WESTWOOD JT, CHURCH RB, WAGENAAR EB: Changes in protein phospho-
rylation during the cell cycle of Chinese hamster ovary cells. J Biol Chem
260:10308-10313, 1985
(12) LJEBES LF, KRIGEL RL, CONKLYN M, ET AL: Ribonucleotide content of
(13)
(14)
(15)
(16)
(17)
(18)
mononuclear cells from normal subjects and patients with chronic lymphocytic leukemia: Increased nicotinamide adenine dinucleotide concentration
in chronic lymphocytic leukemia lymphocytes. Cancer Res 43:5608-5617,
1983
ECKHARDT T: A rapid method for the identification of plasmid deoxyribonucleic acid in bacteria. Plasmid 1:584-588, 1978
MURRAY AW, KIRSCHNER MW: Cyclin synthesis drives the early embryonic
cell cycle. Nature 339:275-280, 1989
WEINERT TA, HARWELL LH: The RAD9 gene controls the cell cycle
response to DNA damage in Sacchwomyces cerevisiae. Science 241:
317-322,1988
TOBEY RA: Different drugs arrest cells at a number of distinct stages in G 2 ,
Nature 254:245-247, 1975 .
COHEN JJ, DUKE RC: Glucorticoid activation of a calcium-dependent
endonuclease in thymocyte nuclei leads to cell death. J Immunol 132:38—42,
1984
YAMADA T, OHYAMA H: Radiation-induced interphase death of thymocytes
is internally programmed (apoptosis). Int J Radiat Biol 53:65-75, 1988
(19) DUKE RC, CHERVENAK R, COHEN JJ: Endogenous endonuclease-induced
DNA fragmentation: An early event in cell-mediated cytolysis. Proc Nad
Acad Sci USA 80:6361-6365, 1983
(20) GAAL JC, PEARSON CK: Covalent modification of proteins by ADPribosylation. Trends Biochem Sci 11:171-175, 1986
Journal of the National Cancer Institute
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
Several other characteristics of apoptosis have been investigated. Following incubation of L1210/0 cells with cisplatin, we
observed an initial increase in cell size associated with arrest in
the G2 phase (fig. 4). At minimally toxic concentrations, the cell
size had recovered by 4 days, but at toxic concentrations, a
significant proportion of the cells demonstrated a reduction in
volume to less than control cells. Cell shrinkage is often associated with apoptosis, but the increased size has not previously been
reported. This increase may not have been observed before for
several reasons. First, many of the observations have been made
in whole tissues, in which apoptotic cells are typically quite
isolated and infrequent (9). Because of this relative infrequence,
it would be difficult to observe cells undergoing early stages of
apoptosis, particularly since such swollen cells might not be
recognized as abnormal. Second, work in thymocytes shows a
much more rapid induction of apoptosis in response to dexamethasone (17,19). Generally, in vitro studies have not been
performed under conditions that would permit detection of delayed cell death. We also detected occasional cell surface blebs by
scanning electron microscopy (fig. 5). The holes subsequently
observed correlated with the timing of trypan blue uptake. The
holes probably resulted from the bursting of the blebs.
(2/) BERGER NA: Poly(ADP-ribose) in the cellular response to DNA damage.
RadiatRes 101:4-15, 1985
(22) YOSHIHARA K, TANIGAWA Y, BURZIO L, ET AL: Evidence for adenosine
diphosphate ribosylation of Ca 2 + , Mg24"-dependent endonuclease. Proc
Natl-Acad Sci USA 72f289-293, 1975
(23) RIABOWOL K, DRAETTA G, BRIZUELA L, ET AL: The cdc2 kinase is a nuclear
~
protein that is essential for mitosis in mammalian cells. Cell 5,7:393-401,
1989"
"*"'" ' "
(24) PINES J, HUNTER T: Isolation of a human cyclin cDNA: Evidence for cyclin
mRNA and protein regulation in the cell cycle and for interaction with
p34cdc2. Cell 58:833-846, 1989
(25) RUSSELL P, NURSE P: cdc25 + Functions as an inducer in the mitotic control
- o f fission yeast. Cell 45:145^153, 1986
(26) BARRY MA, EASTMAN A: Induction of programmed cell death by anticancer
drugs, toxins and hyperthermia. Proc Am AssocCancer Res 3O:553f 1989
Jeffrey S. Weber, Steven A . Rosenberg
render it susceptible to destruction by cytolytic T cells, whereas
the absence of class I antigens may result in increased tumor
Tumor-infiltrating lymphocytes (TILs) are T cells that can be
growth and metastatic spread due to immune evasion (5-7).
grown from enzyme-digested murine or human tumors.
Cytolytic T cells can be found within tumors in humans and mice
When adoptively transferred to tumor-bearing hosts concur(8-10). These tumor-infiltrating lymphocytes (TILs) can be
rent with the administration of recombinant interleukin-2
isolated from a variety of enzyme-digested murine and human
(rIL-2), TILs can mediate significant regression of tumor. To
tumors. When grown in the presence of interleukin-2 (IL-2), they
examine whether expression of class I major histocompatibilexhibit a remarkable degree of lytic specificity in vitro for the
ity complex on tumor cells influenced the generation and
tumor of origin [(5,9); Barth R, Bock S, Mul6 J, et al.: manuscript
antitumor activity of TILs, we used clones of murine B16BL6
submitted for publication]. In the mouse, these Thy-1 + , Lyt-2 + ,
melanoma either transfected with or lacking the class I gene
L3T4~ cells mediate significant tumor regression when as few as
K" to generate TILs at a high dose (1,000 U/mL) or at a low
106 cells are adoptively transferred to tumor-bearing mice condose (20 U/mL) of human rIL-2. TILs grew from both tumors
current with the administration of low doses of recombinant IL-2
in high-dose rIL-2, but they grew from the class I-expressing
(rIL-2) (8). In humans, the same type of approach was effective in
tumor only in low-dose rIL-2. TILs from the class
the treatment of advanced melanoma as well (11). The factors that
I-deficient tumor did not lyse any target tested in vitro, nor
determine susceptibility to therapy with TILs have-not yet been
did they demonstrate any therapeutic effect in vivo on estabelucidated.
lished tumors that lacked or expressed class I. In contrast,
Since the levels of class I MHC expression differ on human and
TILs from the class I-expressing tumor specifically lysed the
tumor of origin in vitro and caused it to regress in vivo. mouse tumors, we hypothesized that variation in class I antigen
Further, these TILs demonstrated activity in vitro against the expression on tumor cells may influence the ability to generate
non-class I-expressing melanoma treated with the combina- prolonged cultures of TILs and their therapeutic efficacy in vivo.
tion of murine recombinant interfering and human recom- TILs were isolated from tumors derived from MHC-deficient
binant tumor necrosis factor a; in vivo, when administered clones of B16BL6 melanoma as well as from clones that were
b
with recombinant interferon 7 and recombinant tumor ne- transfected with a gene encoding the class I gene K . We tested
crosis factor a, TILs from the class I-expressing tumor whether TILs from class I-deficient and class I-expressing B16
mediated regression of non-class I-expressing pulmonary tumor transfectants could recognize and lyse their autologous
metastases, presumably by augmenting class I expression. [J targets in vitro and whether augmented class I expression after
Natl Cancer Inst 82:755-761,1990]
The level of expression of class I major histocompatibility
complex (MHC) antigens on virally infected cells or tumor cells is
an important determinant of their interaction with cytolytic T
lymphocytes (1,2). Since recognition of viral and tumor antigens
occurs in an associative reaction with MHC molecules (3,4), the
presence of class I antigens on the surface of a tumor cell may
Vol. 82, No. 9, May 2, 1990
Received January 4, 1990; revised February 9, 1990; accepted February 15,
1990.
Surgery Branch, National Cancer Institute, Bldg. 10, Rra. 2B42; National
Institutes of Health, Bethesda, MD 20892.
We thank Drs. James Mull and Bernard Fox for stimulating and helpful
discussions, Joan Harris for typing this manuscript, and Gilbert Jay for providing
transfectant cells and encouragement.
ARTICLES 755
Downloaded from http://jnci.oxfordjournals.org/ at Pennsylvania State University on September 11, 2016
Effects of Murine Tumor Class I Major
Histocompatibility Complex Expression on
Antitumor Activity of Tumor-Infiltrating
Lymphocytes