Apoptosis Induced by Serum Deprivation of PC 12 Cells Is Not Preceded by GrowthArrest and Can Occur at Each Phase of the Cell Cycle1
Liora Lindenboim, Rochelle Diamond, Ellen Rothenberg, and Reuven Stein2
Department of Biochemistry. The George S. Wise Faculty of Life Sciences, Tel Aviv University. Ramal Aviv. Tel Aviv 69978, Israel ¡LL, R. SJ, anil Division of Bioloi>\,
California Institute of Technology, Pasadena, California 91125 ¡R.D., E. RJ
ABSTRACT
Previous studies have shown that PC12 cells undergo apoptosis (programmed cell death) when deprived of serum. In the present study, weexamined the relationship of this death process to the cell cycle. PC12 cellpopulations synchronized at different, specific phases of the cell cycleexhibit similar kinetics of cell death following deprivation of serum. Flowcytometry analysis was used to examine the levels of apoptotic death inthese cell populations in relationship to their progression in the cell cycleduring the course of serum deprivation. Such analysis revealed that thecells die during the G(I-G,, S, and perhaps G2-M phases and at the G, to
G, transition.These results, therefore, suggest that the death of synchronized, serum-
deprived PC12 cells occurs throughout the cell cycle and is not dependenton growth arrest. Flow cytometry methodology (acridine orange staining),which determines the RNA content of cells in relationship to their positionin the cell cycle, was used to address these questions in nonsynchronizedcells. These experiments revealed that, upon serum deprivation, an immediate loss of RNA occurred from cells in G,, S, and G ,-M phases. This
loss is accompanied by a slower appearance of cells with degraded DNAcontent. These results show that cells from all phases of the cell cycle aredamaged upon serum deprivation and thus suggest that the apoptotic celldeath of nonsynchronized PC 12 cells may occur from each phase of thecell cycle.
INTRODUCTION
Normal cell death is an important physiological process in whichunnecessary cells are eliminated during both development and adultlife (1-3). The availability of growth factors can play an important
role in controlling normal cell death because most (if not all) cellsdepend on growth factors for their survival (4, 5). Cells that fail toobtain adequate amounts of their survival factors die. For example, inthe nervous system, a massive death of neurons and oligodendrocytesoccurs during development, most likely due to the failure of thesecells to obtain the specific neurotrophic factors on which they depend(6, 7). In vitro studies have shown that both mitotic and postmitoticdifferentiated neurons (8, 9) and oligodendrocytes (5) depend ongrowth factors for their survival. Therefore, the survival effect ofgrowth factors may play an important role in the generation as well asin the differentiation of the nervous system.
One in vitro model system used extensively to study neuronaldifferentiation is the rat pheochromocytoma cell line, PC12. Thesecells proliferate in serum-containing medium. In response to NGF,3
they stop proliferating, differentiate, and acquire a sympathetic, neuron-like phenotype (10, 11). It has recently been demonstrated that
proliferating PCI2 cells (naive) depend upon serum, whereas differentiated postmitotic (neuronal) PC12 cells depend upon serum andNGF for their survival. When serum is withdrawn from naive cells
Received 8/12/94; accepted 1/18/95.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 with18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported by Grant 90-00347 from the United States-Israel Binational
Science Foundation (Jerusalem, Israel).~ To whom requests for reprints should be addressed.1 The abbreviations used are: NGF, nerve growth factor; PI, propidium iodide; AO.
and both serum and NGF from neuronal cells, these cells die andexhibit the characteristic features of the cell death termed apoptosis(programmed cell death; Refs. 12-15). PC12 cells, therefore, can also
be used as a model system for studies on the mechanism of growthfactor-dependent survival in the nervous system. The advantage of
fered by the PC 12 cells is that they can be used in studies of bothmitotic and postmitotic differentiated cells. This is an importantfeature of the cells, since in the nervous system both mitotic andpostmitotic neurons depend on growth factors for their survival. Thedeath of growth factor-deprived PCI 2 cells can be prevented by some
growth factors (e.g., NGF, fibroblast growth factor, and insulin), bythe nuclease inhibitor aurintricarboxylic acid (12, 13), and by exogenous expression of bcl-2 (16). Although some progress has been
made in identifying the survival factors and mechanisms which cankeep PC12 cells alive, very little is known about the intracellularevents involved in the process of apoptosis in these cells. It hasrecently been suggested that a defective cell cycle control is responsible for inducing apoptosis (17, 18). This hypothesis is based interalia on the observations that, in some systems, there is a correlationbetween the cell cycle and apoptosis. For example, in some systems,apoptosis occurs only after cell cycle arrest (19-23), and in other
systems, the cells are preferentially susceptible to death at specificphases of the cell cycle (24-26). The purpose of the present study wasto determine whether the death of serum-deprived PCI2 cells occurs
at a specific phase of the cell cycle and whether it depends on growtharrest.
MATERIALS AND METHODS
Cell culture-stock PC12 cells obtained from Dr. Gordon Guroff (NIH
Bethcsda, MD) were maintained in DMEM high glucose medium supplemented with 8% heat-inactivated horse serum, 8% fetal calf serum, 25 ng/mlpenicillin, and 2 mM t.-glutamine. For serum-deprivation experiments, conflu
ent stock cultures were washed twice with PBS, detached by 0.5 mM EDTA,and centrifuged. The cells were then resuspended in DMEM and plated ineither 96 flat-well plates for MTT (Sigma Chemical Co.) assay or in dishes for
flow cytometry and DNA fragmentation analyses. For preparation of PCI2cells synchronized in different and specific phases of the cell cycle, subcon-fluent cultures 6 x IO6 cells/140-mm dish were treated with 2 mM thymidine
(Sigma). After 24 h, thymidine was removed, and the cells were washed twicewith PBS and then grown for additional time periods in drug-free medium. Atthe times indicated in "Results" (each time point represents a cell population
synchronized at a different cell cycle phase), the synchronized cells weredetached from dishes and replated in the presence or absence of serum foradditional times as indicated. Cell synchronization was assessed at the varioustimes following thymidine release by flow cytometry.
Flow Cytometry Analysis. Three different methods were used for DNAlabeling. Nuclei labeling was performed as described previously (27). Briefly,each 200 X g centrifuged cell pellet (IO6 cells) was resuspended and fixed in
200 ju.1citrate buffer (250 mM sucrose in 40 mM trisodium citrate) with 5%
dimethyl sulfoxide. At this step, samples can be either directly processed orstored at -70°C. Before flow cytometry analysis, cells were incubated at room
temperature with 900 ju.1of stock solution (3.4 mM trisodium citrate, 0.1%NP40, 1.5 mM Spermine tetrahydrochloride, and 0.5 mM Trisma base) containing 30 f¿g/mltrypsin for 10 min, followed by the addition of 750 jd stocksolution containing 0.5 mg/ml trypsin inhibitor with 0.1 mg/ml RNase A. Afteran additional 10 min of incubation, the cells were stained by the addition of
SERUM-DEPRIVED PC12 CELLS DIE THROUGHOUT THE CELL CYCLE
750 fil of solution containing 208 ng/ml PI (Sigma) with 0.58 mg/ml Sperminetetrahydrochloride for 1-3 h at 4°C.PI fluorescence of individual nuclei was
then measured using a FACS IV flow cytometer (Becton Dickinson, MountainView, CA) excited at the 488 nm wavelength and was collected through a570-nm L.P. filter. A Consort 40 software computer system was used to
analyze the data.Cell labeling was performed as described previously (28). Briefly, each
200 X g centrifuged cell pellet (IO* cells) was resuspended in 1-ml hypotonie
fluorochrome solution (50 ^.g/ml PI in 0.1% sodium citrate plus 0.1% TritonX-100) and stored overnight in the dark at 4°Cbefore the flow cytometric
analysis. The PI fluorescence of individual nuclei was measured using a FACSIV flow cytometer as described previously. The mean of the G, peak wasassigned to channel 100 and, therefore, the events collected between channels5 and 60 were assigned as apoptotic peaks.
AO staining for the bivariate gating analysis was carried out as describedpreviously (29, 30). Briefly, cells were scraped from dishes with a rubber policeman and centrifuged. The cells were then resuspended at IO6 cells/ml in PBS.
Aliquots (200 fil) of cell suspensions were mixed for 30 s with 0.4 ml of chilledsolution containing 0.08 N HC1,0.2% Triton X-100, and 0.15 MNaCl, followed bythe addition of 1.2 ml of solution containing 10 JJ.MAO in an ice-cold citratephosphate (0.2 M; pH 6) buffer. The red (630-nm) and green (535-nm) fluores
cence of individual cells was measured with an Ortho Cytofluorograf 50 H with anargon laser exciting at 488 nm and 250 mW. Single cells were gated from doubletsby measuring the peak width and area of green fluorescence.
DNA Fragmentation Assay. Soluble DNA (27,000 g) was isolated fromIO7 cells/culture exactly as described previously by Hockenbery et al. (31)
except that DNA samples were treated with RNase A (250 fig/ml) before 1/10of the sample was loaded onto the gel. Agarose gel (1.2%) electrophoresis ofDNA was carried out in buffer consisting of 1 ng/ml ethidium bromide, 40 minTris HCl (pH 7.9), 4 mM sodium acetate, and 1 mM EDTA. The DNA wasvisualized under UV.
Cell Viability Assay. The number of viable cells was determined by theMTT assay (32). MTT was dissolved in PBS at a concentration of 5 mg/ml.From this stock solution, 10 fil/100 fil medium was added to each well, andplates were incubated at 37°Cfor 4 h. Acid-isopropanol (100 (j.1of 0.04 N HCl
in isopropanol) was then added to the wells and mixed in. After 15 min at roomtemperature, the plates were read on a Microelisa reader, using a test wave
length of 540 nm and a reference wavelength of 690 nm. Data are presented asthe differences between absorbance values at 540 and at 690 nm and asmeans ±SD. The values defined are significant in all cases where Student's t
test yielded P < 0.01.
RESULTS
Serum deprivation has been shown previously to induce the rapiddeath of proliferating PC12 cells (12) accompanied by the DNAfragmentation characteristic of apoptosis (13, 14).
DNA ContentFig. I. DNA histograms from thymidine-synchronized PC12 cells. Cells were treated with 2 mM thymidine for 24 h, and the drug was then removed. At the limes indicated in each
histogram following thymidine release, nuclei were stained with PI and analyzed by flow cytometry as described in "Materials and Methods." N.S., nonsynchronized cells. GQ-G,, S,
and G2-M cell populations are indicated by arrows in the histogram of nonsynchronized cells. The data in insets show the percentage of cells in the different phases of Ihe cell cycle.Results shown are from a representative experiment (of three independent experiments with similar results).
SERUM-DEPRIVED PC12 CELLS DIE THROUGHOUT THE CELL CYCLE
In the present study, we examined whether serum-deprived cells arepreferentiallysusceptible to death at a specific phase of the cell cycle or,alternatively, whether death occurs at all phases of the cell cycle. Toaddress this issue, we prepared PC12 cell populations which had beensynchronized at different, specific phases of the cell cycle so that eachsynchronized population could be deprived of serum and the kinetics ofcell death monitored.To generate populationsof PC12 cells representingdifferentphasesof the cell cycle, PC12 cellswere synchronizedin early-Sphase by treatment with 2 mM thymidine. After 24 h, the drug wasremoved, and the cells were allowed to progress through the cell cycle.Cell populations obtained at various times following thymidine releasecorrespondedto cells present at different,specificphasesof the cell cycle.Cell synchronization was assessed by flow cytometry. As indicated inFig. 1,exponentiallygrowing PCI2 cells were growth arrested at early-Sphase after treatment with thymidine.On removalof the drug, the wholepopulationsynchronouslyprogressedthrough the cell cycle. By 12 h, thecells passed through middle-S, late-S, and G2-M, and by 24 h, into theG,,-G, phase. Thirty to thirty-six h following drug removal, some of thecells began to progress from GirG, into S and G2-Mphases, and the cellsexhibited a similar cell cycle distribution as nonsynchronizedcells.
The different cell cycle populations were deprived of serum for 6,12, and 24 h, and the kinetics of their death was measured. Fig. 2demonstrates that the cell populations corresponding to cells in early-S, middle-S, G2-M, and G0-G, phases (3, 8,12, and 24 h followingdrug release, respectively), as well as nonsynchronized cells, exhibita similar rate of cell death, regardless of their initial position in the cellcycle when serum was withdrawn. The earliest time point examined inthese experiments was 6 h after serum withdrawal, which is theearliest time at which a significant, measurable value of cell death canbe monitored by the MTT assay. In order to determine the ability ofthe different cell populations to undergo apoptosis at earlier times, weexamined the appearance of the characteristic internucleosomal DNAfragmentation. This fragmentation is an early event in the deathprocess of serum-deprived PC12 cells and was clearly detectable 4 hafter serum deprivation, before the appearance of any morphologicalsigns of cell death (13). Soluble DNA was extracted from the cellpopulations corresponding to cells in early-S, middle-S, G2-M, andG0-G, phases (3, 8, 12, and 24 h after drug removal, respectively), aswell as from nonsynchronized cells 4 h after serum deprivation. Asshown in Fig. 3, a clear pattern of internucleosomal DNA fragmen-
Time followingThy. release [h] N.S.
Serum
0S
0.41
O•*""> 0.3
Q00.20.1ITime
followingThy. releasefh]:Iia
'LXI1.I;Ã
;II
:'•.;i;S
iLi',i3
8 12 24 N.S.
Fig. 2. Kinetics of loss of viability of the different cell cycle-specific cell populations.PC12 cells were synchronized by thymidine (Thy) blockade. At 3, 8, 12, and 24 hfollowing thymidine release, the different cell populations (early-S, middle-S, G2-M, andG0-G,, respectively) were transferred to serum-free medium for an additional 0 h S, 6 h
D, 10 h E3, and 24 h W. The number of viable cells was determined by MTT assay asdescribed in "Materials and Methods" and is presented as the difference between absorb-
ance (OD) values at 540 and 690 nm. Values are means ±SD of 32 replicates from onerepresentative experiment. N.S.. nonsynchronized cells. Results shown are from a representative experiment (of four independent experiments with similar results). Bars, SD.
Fig. 3. Agarose gel electrophoresis of soluble DNA extracted from the different cellcycle-specific cell populations. PC12 cells were synchronized as described in Figs. 1 and
2. At the times indicated following thymidine (Thy) release, the cells were transferred toserum-free medium for 4 h, and DNA then was extracted as described in "Materials andMethods." N.S., nonsynchronized cells. Results shown are from a representative experi
ment (of five independent experiments with similar results).
tation is evident in DNA isolated from all these cell populations. Nosoluble DNA was detected in the samples isolated from nonsynchronized cells (Fig. 3) and from the different cell cycle-specificpopulations (data not shown) before serum deprivation.
The results depicted in Figs. 2 and 3 demonstrate that the differentcell populations exhibit a similar rate of cell death after serum deprivation, regardless of their initial position in the cell cycle. Thissuggests that death may occur at each phase of the cell cycle. Alternatively, it is possible that death occurs at a specific phase of thecycle, but this cell cycle-specific death cannot be detected, since thecells are progressing in the cell cycle during the course of the experiment and because the cell cycle-specific populations are not completely homogenous. To discriminate between these two possibilities,treated cells were analyzed by flow cytometry.
DNA content histograms obtained by such analysis indicate boththe distribution of cells in the cell cycle and the appearance ofapoptotic cells (cells with subdiploid DNA content) in each treatment.In addition, this analysis enabled the estimation of the relative levelsof the apoptotic cells in the examined cell population. Serum waswithdrawn from the different PC12 populations (4, 8, 12, and 24 hfollowing thymidine removal) and from nonsynchronized cells for 4and 10 h, and the cells were then stained with PI and analyzed by flowcytometry. The results of these experiments are presented in Fig. 4.
It is evident from the experiments depicted in Fig. 4£that cell deathcan occur directly at the G,rG, phase, since the 0,,-G, cell populationdid not progress in the cell cycle but remained at the G(I-G, phaseduring the course of the experiment while cell death still occurred.Cells can also die at the S-phase, as demonstrated in Fig. 4, B and C.Four h after the beginning of the experiments, both early-S (Fig. 4ß)and middle-S (Fig. 4C) cell populations did not progress from theS-phase into the G()-G, phase (as indicated by the DNA histograms ofthese cell populations in the presence of serum), although a substantialapoptotic cell population is evident after 4 h of serum deprivation.These apoptotic populations, for the most part, must derive from themajority of cells in this population, i.e., the S-phase cells. Cells mayalso die at the G2 to G, transition, as indicated by the failure of theG2-M cells to arrive at G, (compare Fig. 4, B, C, and D, 4 and 10 hafter serum deprivation). The results presented in Fig. 4D indicate that
SERUM-DEPRIVED PC12 CELLS DIE THROUGHOUT THE CELL CYCLE
Treatment time (h)
Fig. 4. DNA histograms from the different cellcycle-specific cell populations following serumdeprivation. PC12 cells were synchronized at thedifferent cell cycle phases as described in Figs. 1and 2. Three h (ß),8 h (Q, 12 h (D), and 24 h (E)following thymidine release, the cells were harvested and replated for the indicated times in serum-free medium ( ) or in medium supplemented with serum (—). The cells were thenstained with PI and analyzed by flow cytometry.DNA histograms from nonsynchronized cells arepresented in (A). Ap, G„-G,,S, and G2-M cell
populations are indicated by arrows. The data inthe insets show the percentage of cells in Ap,GO-G,, S, and G2-M phases, of the different cell
populations grown in the presence (+) or absence(-) of serum. Results shown are from a representative experiment (of four independent experimentswith similar results).
10
ozoo
A_B
7M1-C
7MD
7M-0
—E
7M--76a
-.Ap
1.0Go/G,G^G,57.0\S.Gj/M21.4Ap
11 SGa/M\
Aii/Vw,.,60
120 180240768-,Ap
3+5GO«,
23.9S.Gj/M
66.5Jl.
^A.60120 180240768-,Ap
2+5G(/G,
24.1S.GÜ/M
69.5AA60
120 180240768-,Ap
2.8G(/G,19.3S.Gj/M
68.9-""
—"^ 1 i 1— o—60
120 180 240
768-Ap
3.1Go/G,70.8,
S, Û2/M9.3n/j
\768
-,•f—Ap
2.616.8¿o/Gì60.055.1.
S.Û2/M 23.519.4i^J
Wt*«^60120 180240768-,Ap
5.821.5G0'G,20.520.4S.Gj/M
67.651.3-KwJ\A^60
120 180240768-.•f
-Ap
3.421.0GO/G,24.220.3S.GÃŽ/M67.451.7k^jl/H60
120 180 240
768-.Ap
6.118.8G(/G,
27.329.1S.Gj/M50.146.0I
1 1 r— 0-
60 120 180240!
+ -7"~ÃŽAp
3.023.BJcyG,76.856.4S,
Gj/M 7.46.2--,+
—Ap
1.528.9GO/G,
68.447.7jj
S.G2/M 20.9 12.3
Iin—\—i —i—i—60
120 180240+Ap
3.849.6GO/G,67.017.9^
S.Û2/M 15.627.5f:Ì
'!|(\¿L*^60
120 180240Ap
6.143.5G(/G,69.729.6S.GJM9221.0/,>»*.
•J\^^60
120 180240•f
-Ap
4.439.8ÕG(/G, 70.541.6iiS.Gj/M
9.711.360
120 180240Ü
Ap 1.242.3jjGo/G,80.344.7i
; S, G¿/M 8.83.3\I'-h\_
ji120 180 240 60 120 180 240
DNA Content
120 180 240
cell death can also occur at the G2-M phase, since in theseexperiments the cells that are present are mainly at late-S and G2-M
phases during the first 4 h of the experiment. However, in thisexperiment, it is not possible to distinguish between late-S and G2-M
phases, so the possibility that the apoptotic cells are derived fromlate-S cells and not G2-M cells cannot be excluded. The conclusion
that PC12 cells may die at each phase of the cell cycle can be furthersupported by the results that the different cell cycle populations die ata similar rate after serum deprivation, i.e., a similar level of apoptoticcells is detected in the different cell cycle populations as well as innonsynchronized cells 4 and 10 h after serum deprivation (about 20and 40% of the total cell populations, respectively; Fig. 4). The results
presented in Fig. 4 also reveal that the cells can die without beinggrowth arrested previously. Accordingly, cell death can occur in cellpopulations which are progressing anywhere in the cell cycle(Fig. 4, B, C, and D).
The experiments described in Fig. 2-4 strongly suggest that thedeath of synchronized, serum-deprived PC 12 cells is not restrictedto one particular point in the cell cycle. To assess whether serum-deprived, non-synchronized PC12 cells also die throughout the cell
cycle, we applied the flow cytometric methodology which simultaneously stains cellular DNA and RNA (AO staining; Refs. 29 and30). This method allowed us to examine the effect of serumdeprivation on the RNA content of cells in relation to their position
SERUM-DEPRIVED PC12 CELLS DIE THROUGHOUT THE CELL CYCLE
Fig. 5. Bivariatc (DNA/RNA) counter plots ofexponentially growing PC12 cells following serumdeprivation. PC 12 cells were transferred to serum-
free medium for the indicated times. The cells werethen stained with AO and anaced by flow cytom-etry as described in "Materials and Methods."
Based on DNA values (green fluorescence), cellswere classified in G,, S, and (1. M phases (asshown by the horizontal lines), whereas G(1and G,cells were distinguished based on RNA values (redfluorescence). G0 cell population is indicated by thearrow. Apoptotic cells (Ap) have lower DNA staining and consist of two populations: Ap-high (Apri), which exhibit high RNA values, and Ap-low(Ap-L), which exhibit low RNA values. The data ininsets show the percentage of cells in each phase ofthe cell cycle and univariate DNA frequency distribution histograms in which the Ap, G0-Gi, S, andG2-M cell populations are indicated. Control, exponentially growing cells. Results shown are froma representative experiment (of two independentexperiments with similar results).
<
in the cell cycle. As shown in Fig. 5, serum deprivation results ina rapid reduction of the RNA content of most cells in the S andG2-M phases and many of G, cells as well. This reduction in the
RNA content may indicate that these cells have initiated theprocess of apoptosis and, therefore, suggest that serum-deprived,
nonsynchronized cells may initiate the apoptotic process in allphases of the cell cycle. In addition to the rapid effect on the RNAcontent of the cells, there is a slow appearance of cells withsubdiploid DNA content (apoptotic cells) in both the high RNAfraction (Ap-high population) and low RNA fraction (Ap-low
population).
DISCUSSION
PC12 cells can be induced to differentiate and acquire the pheno-
type of nondividing neurons (10, 11). Both naive and neuronal PC12cells depend on growth factor for their survival. Upon withdrawal ofserum from naive cells or both serum and NGF from neuronal PCI 2cells, these cells die and exhibit the DNA fragmentation characteristicof apoptosis (12-15). These features of PC12 cells resemble the
dependence of mitotic and postmitotic neurons on growth factors inmany aspects and thus make these cells a useful model system forstudying the mechanisms of apoptosis induced in mitotic and postmitotic neurons after growth factor deprivation. The present study concentrates on the death of mitotic PC12 cells after serum deprivation.We asked whether this death is preceded by growth arrest and whetherit occurs at a specific phase of the cell cycle. Our results demonstratethat the death of serum-deprived PC 12 cells is not restricted to one
particular point in the cell cycle and that it is not preceded by growth
arrest. The rationales for these conclusions are: (a) PCI 2 cell populations synchronized at specific and different phases of the cell cycle,as well as nonsynchronized cells, exhibited similar kinetics of celldeath after serum deprivation. Cell death in these experiments wasindicated by three independent parameters, loss of viability (Fig. 2),DNA fragmentation (Fig. 3), and the appearance of the apoptotic cellpopulation in DNA histograms (Fig. 4); (b) the fact that serum-deprived cells can die at some point between the late-S-G2 phase and
the next G, phase is shown in Fig. 4 by the failure of cycling cells toarrive at the G0-G, phase; (c) the fact that serum-deprived cells candie within the S and G2-M phases is shown in Fig. 5 by the abrupt lossof RNA from these cells; and (d) the fact that serum-deprived cells diewithin GQ-G! or when trying to enter or leave G0-G, is shown in Fig.
4 by the appearance of cell death during the time periods in which theG0-Gt cell population did not progress in the cell cycle. Similarly, Fig.
5 shows that there is a cell death excess not accounted for by cells inthe S or G2-M phases, which is represented by an apoptotic fraction
that is either due to cell death within G, or death in cells trying toenter or leave the Gt phase. At this time, we cannot distinguishbetween the latter two possibilities.
Examination of the RNA content of cells in relation to their positionin the cell cycle (Refs. 29 and 30; AO staining) indicates that the death ofserum-deprived exponentially growing cells may occur by at least two
pathways. One involves rapid RNA loss, which affects most of S andG2-M cells and many of G¡cells. This RNA loss may reflect RNA
degradation and shuts off total RNA synthesis as shown previously tooccur in some apoptotic systems (33, 34) or stripping of the cytoplasm.The exact mechanism which leads to the RNA loss is not known,
SERUM-DEPRIVED PC12 CELLS DIE THROUGHOUT THE CELL CYCLE
although it is clear that it succeeds in killing the cells. The secondpathway involves DNA degradation, which results in the appearance oftwo cell populations, Ap-high and Ap-low. The Ap-high cell populationmay derive from each of G,, S, or G2-M cells which contain high RNA
levels; however, since the vast majority of these cells are G[ cells, theAp-high cells are most probably derived mainly from G, cells. The originof Ap-low is not clear; it may derive from any of the cell populations withlow RNA content or from the Ap-high cells.
Serum withdrawal from some cell types like fibroblasts causesgrowth arrest in the G0 state (35). This may also be the case in PCI 2cells, because flow cytometry analysis (Fig. 5) suggests that G0 cells(cells with G, DNA content and low RNA level) accumulate afterserum deprivation. It is not clear from the present study whether G()cells can also undergo apoptosis; however, since these cells accumulate during the course of serum deprivation, and since at a later timefollowing serum deprivation essentially all cells die, it seems likelythat G„cells can also undergo apoptosis.
Apoptosis in many systems is often preceded by growth arrest, mainlybut not only in the G0-G, phase (19-23). The data presented here
demonstrate that, although serum deprivation may drive the cells into theG() state, death may occur also at the G„S, and G2-M phases, i.e., the
apoptotic death of PC12 cells does not require the cells to be first growtharrested at the G() stage. In summary, upon serum deprivation, cell deathoccurs at all phases of the cell cycle. Cells that escaped from deathcontinue to progress in the cell cycle. These cells can either die in the nextphase of the cell cycle or can continue their progression in the cell cycleuntil they are arrested at the G,,-G| state and then die.
Previous studies showed that apoptosis induced by some antitumoragents also occurs in all phases of the cell cycle (36, 37). Our studiesextend these conclusions to a different system in which apoptosis isinduced by growth factor deprivation. The importance of the systemused in the present study is that it represents apoptosis that is likely tooccur in normal cell death.
The present study examined the relationship between apoptosis andthe cell cycle in proliferating PC12 cells. It is interesting to comparethese studies to nondividing neuronal PC12 cells. It has recently beenreported that the death of growth factor-deprived neuronal PC 12 cells
is accompanied by a burst of DNA synthesis (38) and activation ofP34cdc2 kinase (39). The interpretation of these results implied that the
postmitotic neuronal PC12 cells die because of aborted attempts toreenter the cell cycle. These results may, therefore, suggest thatapoptosis is induced differently in growth factor-deprived PC 12 and
neuronal PC12 cells. However, the observation that every growthfactor or agent which protects naive PC 12 cells from serum deprivation also protects neuronal PC12 from serum and NGF deprivationsuggests that the death process of these two cell systems is verysimilar. Clearly, further work is needed to address the issue of howapoptosis is induced in growth factor-deprived naive and neuronal
PC12 cells.
REFERENCES
1. Wyllie, A. H., Kerr, J. F. R., and Currie, A. R. Cell death: the significance ofapoptosis. Int. Rev. Cytol., 68: 251-305, 1980.
2. Cohen, J. J. Apoptosis. Immunol. Today, 14: 126-130, 1993.
3. Franklin. J. L.. and Johnson, E. M.. Jr. Suppression of programmed neuronal death bysustained elevation of cytoplasmic calcium. Trends Neurosci., /5: 501-508, 1992.
4. Raff, M. C. Social controls on cell survival and cell death. Nature (Lond.), 356:397-399, 1992.
5. Raff, M. C., Barres, B. A., Burne, J. F., Coles, H. S., Ishizaki, Y., and Jacobson, M. D.Programmed cell death and the control of cell survival: lessons from the nervoussystem. Science (Washington DC), 262: 695-700, 1993.
6. Oppenheim. R. W. Cell death during development of the nervous system. Annu. Rev.Neurosci., 14: 453-501, 1991.
7. Barres, B. A., Hart, I. K., Coles, H. S. R., Burne, J. F., Voyvodic, J. T., Richardson,W. D., and Raff, M. C. Cell death and control of cell survival in the oligodendrocytelineage. Cell, 70: 31-46, 1992.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
Martin. D. P., Schmidt, R. E., Distefano. P. S., Lowry, O. H., Carter, J. G., andJohnson, E. M. Inhibitors of protein synthesis and RNA synthesis prevent neuronaldeath caused by nerve growth factor deprivation. J. Cell Biol., 106: 829-844, 1988.
Birren, S. J., Lo, L., and Anderson, D. J. Sympathetic neuroblasts undergo a developmental switch in trophic dependence. Development (Camb.), 119: 597-610. 1993.Greene, L. A.. Aletta, J. M.. Rukenstein, A., and Green, S. H. PC12 phcochromocy-
toma cells: culture. NGF treatment and experimental exploitation. Methods Enzymol.,147: 207-216, 1987.Fujita. K., Lazarovici. P.. and Gurol'f. G. Regulation of the differentiation of PC12
pheochromocytoma cells. Environ. Health Perspect., 80: 127-142, 1989.
Rukenstein, A., Rydel, R. E., and Greene, L. A. Multiple agents rescue PC12 cellsfrom serum-free cell death by translation and transcription-independent mechanisms.J. Neurosci., //: 2552-2563, 1991.
Batistatou. A., and Greene, L. A. Aurintricarboxylic acid rescues PC12 cells andsympathetic neurons from cell death caused by nerve growth factor deprivation:correlation with suppression of endonuclease activity. J. Cell Biol., //5: 461-471,
1991.Batistatou, A., and Greene. L. A. Internucleosomal DNA cleavage and neuronal cellsurvival/death. J. Cell Biol.. 722: 523-532, 1993.Pittman. R. N., Wang, S.. Dibenedetto, A. J.. and Mills, J. C. A system for characterizing cellular and molecular events in programmed neuronal cell death. J. Neuro-sci., ¡3:3669-3680, 1993.
Mah, S. P., Zhong, L. T., Liu. Y., Roghani, L. A., Edwards, R. H., and Bredesen, D. E.The protooncogenc hcl-2 inhibits apoptosis in PC12 cells. J. Neurochem., 60: 1183-
1186, 1993.Ucker, D. S. Death by suicide: one way to go in mammalian cellular development?New Biol., J: 103-109, 1991.Rubin, L. L., Philpott, K. L., and Brooks, S. F. The cell cycle and cell death. Curr.Biol., 3: 391-394, 1993.Sabourin. L. A., and Hawley, R. G. Suppression of programmed death and G, arrestin B-cell hybridomas by interleukin-6 is not accompanied by altered expression ofimmediate early response genes. J. Cell. Physiol., 145: 564-574. 1990.Chen, Y. Y., and Rosenberg. N. Lymphoid cells transformed by Ahelson virus requirethe V-abl protein-tyrosine kinase only during early G,. Proc. Nati. Acad. Sci. USA,89: 6683-6687, 1992.
Ishida, Y., Furukawa, Y.. Decaprio, J. A.. Saito. M., and Griffin, J. D. Treatment ofmyeloid leukemic cells with the phosphatase inhibitor okadaic acid induces cell cyclearrest at either G,S or G,-M depending on dose. J. Cell. Physiol., ISO: 484-492,
1992.Kim, I-K., Lee, J-H., Sohn, H-W., Kim, H-S., and Kim, S-H. Prostaglandin A, andA12-prostaglandin J, induce apoptosis in L1210 cells. FEBS Lett., 321: 209-214.
1993.Wu, F. Y-H., Chang, N-T., Chen, W-J., and Juan, C-C. Vitamin K,-induced cell cyclearrest and apoptotic cell death are accompanied by altered expression of C-/O.Yandc-myr in nasopharyngcal carcinoma cells. Oncogene. K: 2237-2244, 1993.Gorczyca, W., Gong. J., Ardelt, B., Tráganos, F.. and Darzynkiewicz. Z. The cellcycle related differences in susceptibility of HL-60 cells to apoptosis induced byvarious antitumor agents. Cancer Res., 53: 3186—3192, 1993.Boehme. S. A., and Lenardo, M. J. Propriocidal apoptosis of mature T lymphocytesoccurs at S phase of the cell cycle. Eur. J. Immunol., 23: 1552-1560, 1993.Yonish-Rouach, E., Grunwald, D., Wilder, S., Kimchi. A., May, E., Lawrence, J-J.,May, P.. and Oren. M. p53-mediatcd cell death: relationship to cell cycle control.Mol. Cell. Biol., 13: 1415-1423, 1993.Vindelov, L.. Christensen. I. J.. and Nissan. N. I. A detergent-trypsin method for thepreparation of nuclei for flow cytometric DNA analysis. Cytometry, 3: 323-327,
1983.Nicoletti, I., Migliorati, G., Pagliacci. M. C., Grignani. F., and Riccardi, C. A rapidand simple method for measuring thymocyte apoptosis by propidium iodide stainingand flow cytometry. J. Immunol. Methods, 139: 271-279, 1991.Darzynkiewicz, Z.. Sharpless, T.. Staiano-Coico, L., and Melamed. M. R. Subcom-partments of the G[ phase of cell cycle detected by flow cytometry. Proc. Nati. Acad.Sci. USA, 77: 6696-6699. 1980.Diamond, R. A. Separation and enrichment of cell populations by centrifugal elutria-tion. Methods: A Companion to Methods Enzymol., 2: 173-182, 1991.
Hockenbery, D.. Nunez, G., Milliman, C.. Schreiber. R. D.. and Korsmeyer, S. J.Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death.Nature (Lond.), 348: 334-336, 1990.
Mosmann, T. Rapid colorimetrie assay for cellular growth and survival: applicationto proliferation and cytotoxicity assays. J. Immunol. Methods, 65: 55-63, 1983.Cidlowski, J. A. Glucocorticoids stimulate ribonucleic acid degradation in isolated ratthymic lymphocytes in vitro. Endocrinology, ///: 184-190, 1982.Wyllie, A. H., and Morris, R. G. Hormon-induced cell death. Am. J. Pathol., 109:78-87, 1982.Evan, G. I., Wyllie, A. H., Gibert, C. S., Littlewood, T. D.. Land, H., Brooks, M.,Waters, C. M.. Penn. L. Z.. and Hancock. D. C. Induction of apoptosis in fibroblastsby c-myc protein. Cell, 69: 119-128, 1992.Cotter. T. G., Glynn, J. M., Echeverri, F., and Green. D. R. The induction of apoptosisby chemotherapeutic agents occurs in all phases of the cell cycle. Anticancer Res., 12:773-780, 1992.Piacentini, M., Fesus, L.. and Melino. G. Multiple cell cycle access to the apoptoticdeath programme in human neuroblastoma cells. FEBS Lett., 320: 150-154, 1993.
Farinelli, S. E., Ferrari, G.. and Greene, L. A. A possible link between the cell cycleand apoptosis of PC12 cells. Soc. Neurosci. Abst., 19: 885, 1993.Brooks, S. F., Gibson, A., and Rubin, L. L. Apoptosis induced by NGF withdrawalfrom differentiated PC12 cells involves activation of P34"k2 kinase. Soc. Neurosci.
