Cancer Letters, 52 (1990) 101-106 Elsevier Scientific Publishers Ireland Ltd.
101
Neurite outgrowth and cell cycle kinetic changes induced by cis-diamminedichloroplatinum II and retinoic acid in a human neuroblastoma cell line D. Di Martino”, C. Avignolob, G.P. Tonini”
B. Marsanob,
“Pediatric Oncology Research Laboratory, for Cancer Research, Genoua (Italy)
G.
Gaslini
Children’s
A. Di Vincib, A. Caraa, W. Giarettib and Hospital and %ophysics
Laboratory,
National Institute
(Received 30 January 1990) (Revision received 13 March 1990) (Accepted 20 March 1990)
Summary
Introduction
The aim of this study was to analyze by flow cytometry the effect of cis-diamminedichloroplatinum II (CDDP) and retinoic acid (RA) on the cell cycle of a neuroblastoma cell line (SK-N-BE (2)C NB) and to correlate the kinetic data with cell morphology. CDDP at 1 pg/ml induced a dramatic G2 + M cell cycle phases block (nearly 200% increase with respect to control) 2 days after treatment. The G2 + M block was spontaneously reuersed starting from the 4th day. The cells treated with 10 fl RA were, instead, induced to irreversibly enter the GO + GI phase of the cell cycle (nearly 20% increase with respect to control) 48 h after treatment. Neurite-like structures were observed for both CDDP and RA treated cells. These data suggest different cell cycle dependent molecular mechanisms and different degrees of differentiation during CDDP or RA treatment of NB cells.
In the last few years new protocols for cancer therapy were emerging in which anticancer drugs were employed at low dosage to induce terminal differentiation of the cancer cells and revert the malignant phenotype [3,4]. In vitro studies demonstrated that aracytin, doxorubicis-diamminodichloroplatinum II =yn and (CDDP) (to cite only a few) were good inducers of cellular differentiation of human leukemia cell lines [ 12,131. CDDP is known to be a potent anticancer drug employed in the chemotherapeutic protocol of solid tumors. We have recently shown that CDDP was able to induce morphological differentiation of human neuroblastoma cells in vitro [9]. Similar observations were previously reported for neuroblastoma cells treated with RA. The aim of the present work was to analyze the cell cycle kinetics induced by CDDP and RA treatments in the human neuroblastoma cell line SK-N-BE [2]C NB and the associated formation of neurite-like structures which are generally considered a marker of terminal differentiation [l]. The cell cycle analysis was performed by high resolution flow cytometry of the nuclear DNA content in suspensions of single nuclei. We have found that the cells
Keywords: neuroblastoma; cisplatin; acid; differentiation; flow cytometry.
retinoic
Correspondence to: G.P.
lini Children’s
Hospital,
0304-3835/90/$03.50 Published and Printed
Tonini, Ped. One. Res. Lab., G. Gasvia 5 Maggio, 39, 16148 Genova, Italy.
0
1990 Elsevier Scientific Publishers
in Ireland
Ireland Ltd.
102
treated with RA were induced to irreversibly enter the GO + Gl phase of the cell cycle, while the CDDP treated cells showed a reversible block in the G2 + M cell cycle phases. Both treatments induced neurite outgrowth in the neuroblastoma cells.
phosphate buffer, 0.146 M NaCl, 1 mM CaCl,, 0.5 mM MgSO, * 7H,O and 0.2% BSA (w/v)) containing 10 mg/l of 4,6-diamidino-2phenylindole-2-hydrochloride (DAPI; Sigma Chemical Co, St. Louis, MO) as specific DNA
Materials and methods Cell culture The human neuroblastoma SK-N-BE [2]C cell line (kindly provided by V. Ciccarone, Frederick, U.S.A.) was cultured in RPM1 1640 (Gibco, Scotland, U.K.) supplemented with 15% fetal calf serum (Gibco), 50 IU/ml sodium penicillin G, 50 pg/ml streptomycin sulfate and 1% glutamine. Approximately 3 x lo6 cells/flask were cultured in 5% CO, contamination, Mycoplasma atmosphere. checked weekly (mycoplasma test kit DAPI, Bohering, Mannheim, F.R.G.) was negative. Differentiation treatment and cell morp.phological analysis The cells were allowed to adhere to the plastic flask for 24 h (time 0). Successive treatments with 10s5 M RA (Sigma Chemical Co., U.S.A.) were performed at 0, 24 and 48 h. Treatment with 1.0 pg/ml of CDDP (BristolMyers, U.S.A.) was performed once at time 0. The cells were harvested after either short or long intervals of time. The medium was renewed every 48 h without further addition of RA and CDDP. Cell morphological changes were assessed daily by phase contrast microscope. Now cytometry The analysis of the cell cycle dependent effects of the CDDP and RA treatments was done by DNA flow cytometry on suspensions of single nuclei [7]. Aliquots of lo6 cells, previously stored at -8OOC in 100 ~1 of citrate buffer (250 mM saccarose, 40 mM trisodic citrate, 5% (v/v) DMSO in distilled water) were used to obtain the suspensions of nuclei in a nuclear isolation medium (0.6% Nonidet P-40 (v/v) in a solution containing 0.01 M
Fig. 1.
Morphological differentiation of SK-N-BE [2]C NB cells induced by CDDP (1 pg/ml) and RA (lO+M) after 7 days of treatment. Optical microscope magnification was X 160. A: control cells; B: RA-treated cells; C: CDDP-treated cells.
103
stain. The
measurements were performed using an arc lamp flow cytometer (model ICP22A; Ortho Diagnostic Systems, Westwood, MA) with suitable filters for the DAPI excitation (window between 300 and 400 nm) and for the DAPI fluorescence emission (window between 450 and 490 nm). The fluorescence signals were input to signal processing electronics using 1024 channels for subsequent storage and analysis in a PDP 11-23 computer (Digital Equipment Co., Maynard, MA). Histograms of number of nuclei versus DAPI fluorescence intensity values, in channel number, proportional to DNA content, were obtained. The proportions of cells in the GO +
Gl, S and G2 + M cell cycle phases were evaluated using the method of Baisch et al. 121. Results Cehlar morphology Figure 1 illustrates the changes in cellular morphology of SK-N-BE [2]C NB cells observed after treatment with either RA (B) or CDDP (C) with respect to the control cells growing in absence of these drugs (A). RA induced dramatic alterations in the morphology of the NB cells. Extensive neurites first occurred between 48 and 72 h. The cells treated with CDDP also showed neurite-like
512
N
Control
Control
(48 h)
(7 days)
U
m b e r
CDDP
RA (7 days)
(48 h)
-7
1023
1023
Channel
number
(DNA)
Fig. 2. Cell cycle kinetic analysis by flow cytometry of control cells and of CDDP or RA treated cells. Histograms show number of nuclei plotted versus nuclear DNA content. Top panels: control cells (only medium) at 48 h and 7 days. The main peak corresponds to the GO + Gl cell cycle phase; G2 + M phase cells are at twice values of DNA content; the S phase cells are in between. Bottom panels: cells treated with either CDDP at 48 h after treatment or RA at 7 days after treatment. The total number of nuclei measured in every case was 10,000.
104
structures after 72 h. The maximal elongation of these structures was reached 8 days after CDDP treatment. Cell cycle kinetics Figure 2 shows the most prominent and visually recognizable effects of the CDDP treatment at 48 h and of the RA treatment at 7 days with respect to the corresponding control untreated cells. The induction of a G2 + M block by CDDP at 48h is clearly evident. The control cells, in comparison, show a lower G2 + M peak and a higher GO + Gl peak. The long term effect of RA (at 7 days) consist of visually evident increase of the GO + Gl fraction. of cells. The S and G2 + M cell cycle phase fractions decreased. Quantitative cell kinetic data are reported in Tables 1 and 2. Table 1 reports the cell cycle kinetic short term effects of CDDP and RA. The harvest times for CDDP treated cells are at 1 h, 3 h, 6 h, 9 h, 12 h and 24 h. The control cells and the RA treated cells are shown at 24 h only, since no changes were observed for the
Short term effect of Table 1. human neuroblastoma SKN-BE centage of cells in the GO + Gl, of the cell cycle of control cells, cells. Harvest time
CDDP and RA in the [Z]C NB cell line: perS and G2 + M phases CDDP and RA treated
Cell cycle phases (%) Go + Gl
S
G2 + M
Control 24
55
24
21
CDDP (1 &ml) 1 3 6 9 12 24
52 50 51 48 46 34
29 27 28 32 33 40
19 23 21 20 21 26
RA (lo+ M) 24
64
20
16
Long term effect of Table 2. human neuroblastoma SK-NBE centage of cells in the GO + Gl, of the cell cycle of control cells, cells.
CDDP and RA in the [2]C NB cell line: perS and G2 + M phases CDDP and RA treated
Harvest time
Cell cycle phases
(days)
(%)
GO + GlS
G2 + M
Control 7
57
25
18
CDDP (1 pg/ml) 2 3 4 5 7 9
8 13 26 26 38 50
38 32 36 36 34 26
54 55 38 38 28 24
RA (lo+ M) 7
69
18
13
other harvest times. The fraction of control cells within the GO + Gl, S and G2 + M cell cycle compartments were 55%) 24% and 21%. These values were well reproducible in independent experiments and remained approximately constant within 7 days. A decrease of the GO + Gl fraction of the CDDP treated cells from 55% to 34% was observed at 24 h along with an increase of the cells in the S and G2 + M phases. At 2 days after CDDP treatment the percentage of the GO + Gl cells dropped to 8% while the G2 + M cells piled-up to the level of 54% (Table 2). The G2 + M increase was about 200% with respect to control. Later samples (starting from day 4) clearly showed a progressive reduction of the G2 + M fraction to approximately the control level. The cell kinetics of RA treated cells was as follows: the fraction of cells within the GO + Gl phase of the cell cycle increased by nearly 20% (from 55% to 64%) already at 24 h after the treatment (Table 1). This increase remained almost stable within the successive 7 days. Table 2 reports the percentages of cells in the GO + G 1, S and G2 + M
105
cell cycle phases at 7 days (69%) 18% and 13%, respectively) in comparison to the control values (57%) 25% and 18%). Discussion
The property of certain anticancer drugs to induce cellular differentiation has shown great impact in the research for new and more effective strategies for cancer therapy and has elicited many in vitro studies aimed to better understand the coupling of cell proliferation and differentiation. Among neuroblastoma tumors, the capacity of the neuroblasts to revert from malignant cells to mature gangliolike ceils may be of great interest. Malignant phenotype reversion by differentiation probably takes place for neuroblastoma characterized by the stage IVS (51. These patients, in fact, show either spontaneously or by treatment with minimal doses of certain chemical agents, favorable outcome in a high percentage of cases. Among the drugs which are actually employed in clinical trials as differentiating agents there is retinoic acid (RA). Another interesting drug, recently shown to induce cellular differentiation in human neuroblastoma cell lines [9] and in the K562 erythroleukemia cell line [12], is cisplatin (CDDP). Aim of the present study was to better characterize the ceil differentiation induced by CDDP using a human neuroblastoma cell line (NB cells), For this purpose RA and CDDP treated NB cells have been analyzed in their morphological changes (neurite outgrowth as an index of differentiation) and cell kinetic behavior. Cell kinetics, evaluated by means of temporal sequences of DNA histograms obtained by flow cytometry, should be a reliable tool to assess the degree of cell proliferation and differentiation during the course. of RA and CDDP treatments. According to previous literature reports, terminally differentiated cells should, in fact, be expected to be growth arrested with the DNA content of the Gl phase [6,S, 101. NB cells treated with CDDP changed shape
and showed several neurite-like structures protruding from the cell body and interconnected citoplasmic bridges. The treatment with RA appeared to induce a more complex network of neurite-like structures. These morphological observations give further evidence of the capacity of CDDP to induce morphological changes of neuroblastoma cells [9] and well agree with literature reports for the RA differentiating capacity [ 111. CDDP treated NB cells were first blocked in the G2 + M phases of the cell cycle at about 2 days after the treatment. This phenomenon did not occurr in the RA treated cells. It is to be noticed that the CDDP induced G2 + M block was reversed at later times. At 9 days the CDDP treated cells redistributed among the cell cycle phases approximately as in the original controls. These data are not easy to correlate with neither the neurite outgrowth, considered as a marker of end differentiation, nor with the theoretical expectations (accumulation in the Gl phase of the cell cycle). We think that the observed morphology of the CDDP cells does not probably reflect terminal differentiation but previous stages of differentiation still coupled with proliferation. This concept would agree with our kinetic data which do not show any increase of the Gl phase of the cell cycle. Alternatively, one may think that cell differentiation of NB cells occurs also in cell cycle phases other than G 1. This second hypothesis, in our opinion less likely than the first one, requires, however, further investigation. Beside the induction of neurite-like structures in NB cells, cisplatin was clearly cytotoxic as indicated by high debris levels in the DNA measurements. It is likely that the G2 + M block induced by CDDP reflects a reversible cytotoxic effect. The cytotoxic effect is probably different for the individual cells. Beside the sensitive cells which died, it is likely that cells resistant to the drug were still proliferating. In contrast to these effects observed after CDDP treatment, RA induced an increase of the cells into the GO + Gl phase of the cell
106
cycle in parallel with a net decrease of the fraction of cells in the S-phase. The level of RA induced cell differentiation, as judged by the increase of the Gl phase fraction, was above 20%. This fraction is likely to indicate a cell cycle phase of terminal differentiated growth arrested cells (which was indicated as G, by Scott et al. [lo]). Still a fraction of about 30% were in the S and G2 + M cell cycle phases. RA induced cell differentiation, as judged by cell morphology, was instead observed in almost all RA treated cells. The hypothesis, previously suggested for CDDP, i.e., that neurite-like morphology may not indicate cell cycle arrest and terminal differentiation, may also hold for the RA treatment. More work is necessary to better understand the molecular mechanisms underlaying the effects of CDDP and RA in the induction of neuroblastoma cell differentiation and to test the validity of DNA flow cytometry for monitoring the differentiatiing capacity of these drugs in view of therapeutic protocols in cancer patients. Acknowledgements
This work was supported partly by grant 87.17.020 Ricerca Corrente G. Gaslini Children’s Hospital and partly by Associazione Italiana Ricerca sul Cancro (AIRC). The excellent technical contributions of Mr. E. Infusini and Mr. E . Zeraschi are kindly acknowledged.
2
3
4
5
6
7
8
9
10
11
12
References 13 1
Abemayor, E. and Sidell, N. (1989) Human neuroblastoma cell lines as models for the in vitro study of neoplastic and neuronal cell differentiation. Environ. Health Perspect., 80,3-15.
Baisch, H., Goehde, W. and Linden, W. (1975) Analysis of PCP-data to determine the fraction of cells in the various phases of the cycle. Radiat. Environ. Biophys., 12, 3137. Degos, L., Castaigne, S., Chomienne, C., M.E. Huang, M.E., Tilly, H., Bordessoule, D., Miclea, J.M., Thomas, G., Najean, Y., Abita, J.P. and Wang, Z.Y. (1988) Therapeutic trials of acute myeloid leukemia and myelodysplastic syndromes by LD-ARA C and treatment of promyelocytic leukemia by retinoic acid. In: The States of Differentiation Therapy of Cancer, pp. 361-373. Editors: S. Waxman, G.B. Rossi and F. Takakv. Raven Press, New York. Egorin, M.J. (1988) Clinical trials of polar differentiation agents. In: The States of Differentiation Therapy of Cancer, pp. 389-404. Editors: S. Waxman, G.B. Ross and F. Takakv. Raven Press, New York. Evans, A.E., D’Angio, G.J. and Randolf, J. (1971) A pro- I posed staging for children with neuroblastoma. Cancer, 27,374-378. Giaretti, W., Moro, G., Quarto, R., Bruno, S., Di Vinci, A., Geido, E. and Cancedda, R. (1988) Flow cytometric evaluation of cell cycle characteristics during in vitro differentiation of chick embryo chondrocytes. Cytometry, 9, 281-290. Giaretti, W., Sciallero, S.Bruno, S. Geido, E., Aste, H. and DiVinci, A. (1988) A flow cytometry of endoscopically examined colorectal adenomas and adenocarcinomas. Cytometry, 9,238-244. Nadel-Ginard, B. (1978) Commitment, fusion and biochemical differentiation of a myogenic cell line in the absence of DNA synthesis. Cell, 15,855-864. Parodi, M.T., Varesio, L. and Tonini, G.P. (1989) Morphological change and cellular differentiation induced by cisplatin in human neuroblastoma cell lines. Cancer Chemother. Pharmacol., 25, 114-116. Scott, R.E., Florine, D.L., Wille, J.J., and Yun, K. (1982) Coupling of growth arrest and differentiation at a distinct state in the Gl phase of the cell cycle: G,. Proc. Natl. Acad. Sci. USA, 79,845-849. Sidell, N., Altman, A., Haussler, M.A. and Seeger, R.C. (1983) Effects of retinoic acid (RA) on the growth and phenotype expression of several human neuroblastoma cell lines. Exp. Cell Res., 148, 21-30. Tonini, G.P., Parodi, M.T., Bologna, R., Persici, P. and Cornaglia-Ferraris. P. (1986) Cisplatin induces modulation of transferrin receptor during cellular differentiation in vitro. Cancer Chemother. Pharmacol., 18.92. Tonini, G.P., Radzioch, D., Gronberg, A., Clayton, M., Blasi, E., Benetton, G. and Varesio, L. (1987) Erythroid differentiation and modulation of c-myc expression induced by antineoplastic drugs in the human leukemic cell line K562. Cancer Res., 47,4544-4547.
101
Neurite outgrowth and cell cycle kinetic changes induced by cis-diamminedichloroplatinum II and retinoic acid in a human neuroblastoma cell line D. Di Martino”, C. Avignolob, G.P. Tonini”
B. Marsanob,
“Pediatric Oncology Research Laboratory, for Cancer Research, Genoua (Italy)
G.
Gaslini
Children’s
A. Di Vincib, A. Caraa, W. Giarettib and Hospital and %ophysics
Laboratory,
National Institute
(Received 30 January 1990) (Revision received 13 March 1990) (Accepted 20 March 1990)
Summary
Introduction
The aim of this study was to analyze by flow cytometry the effect of cis-diamminedichloroplatinum II (CDDP) and retinoic acid (RA) on the cell cycle of a neuroblastoma cell line (SK-N-BE (2)C NB) and to correlate the kinetic data with cell morphology. CDDP at 1 pg/ml induced a dramatic G2 + M cell cycle phases block (nearly 200% increase with respect to control) 2 days after treatment. The G2 + M block was spontaneously reuersed starting from the 4th day. The cells treated with 10 fl RA were, instead, induced to irreversibly enter the GO + GI phase of the cell cycle (nearly 20% increase with respect to control) 48 h after treatment. Neurite-like structures were observed for both CDDP and RA treated cells. These data suggest different cell cycle dependent molecular mechanisms and different degrees of differentiation during CDDP or RA treatment of NB cells.
In the last few years new protocols for cancer therapy were emerging in which anticancer drugs were employed at low dosage to induce terminal differentiation of the cancer cells and revert the malignant phenotype [3,4]. In vitro studies demonstrated that aracytin, doxorubicis-diamminodichloroplatinum II =yn and (CDDP) (to cite only a few) were good inducers of cellular differentiation of human leukemia cell lines [ 12,131. CDDP is known to be a potent anticancer drug employed in the chemotherapeutic protocol of solid tumors. We have recently shown that CDDP was able to induce morphological differentiation of human neuroblastoma cells in vitro [9]. Similar observations were previously reported for neuroblastoma cells treated with RA. The aim of the present work was to analyze the cell cycle kinetics induced by CDDP and RA treatments in the human neuroblastoma cell line SK-N-BE [2]C NB and the associated formation of neurite-like structures which are generally considered a marker of terminal differentiation [l]. The cell cycle analysis was performed by high resolution flow cytometry of the nuclear DNA content in suspensions of single nuclei. We have found that the cells
Keywords: neuroblastoma; cisplatin; acid; differentiation; flow cytometry.
retinoic
Correspondence to: G.P.
lini Children’s
Hospital,
0304-3835/90/$03.50 Published and Printed
Tonini, Ped. One. Res. Lab., G. Gasvia 5 Maggio, 39, 16148 Genova, Italy.
0
1990 Elsevier Scientific Publishers
in Ireland
Ireland Ltd.
102
treated with RA were induced to irreversibly enter the GO + Gl phase of the cell cycle, while the CDDP treated cells showed a reversible block in the G2 + M cell cycle phases. Both treatments induced neurite outgrowth in the neuroblastoma cells.
phosphate buffer, 0.146 M NaCl, 1 mM CaCl,, 0.5 mM MgSO, * 7H,O and 0.2% BSA (w/v)) containing 10 mg/l of 4,6-diamidino-2phenylindole-2-hydrochloride (DAPI; Sigma Chemical Co, St. Louis, MO) as specific DNA
Materials and methods Cell culture The human neuroblastoma SK-N-BE [2]C cell line (kindly provided by V. Ciccarone, Frederick, U.S.A.) was cultured in RPM1 1640 (Gibco, Scotland, U.K.) supplemented with 15% fetal calf serum (Gibco), 50 IU/ml sodium penicillin G, 50 pg/ml streptomycin sulfate and 1% glutamine. Approximately 3 x lo6 cells/flask were cultured in 5% CO, contamination, Mycoplasma atmosphere. checked weekly (mycoplasma test kit DAPI, Bohering, Mannheim, F.R.G.) was negative. Differentiation treatment and cell morp.phological analysis The cells were allowed to adhere to the plastic flask for 24 h (time 0). Successive treatments with 10s5 M RA (Sigma Chemical Co., U.S.A.) were performed at 0, 24 and 48 h. Treatment with 1.0 pg/ml of CDDP (BristolMyers, U.S.A.) was performed once at time 0. The cells were harvested after either short or long intervals of time. The medium was renewed every 48 h without further addition of RA and CDDP. Cell morphological changes were assessed daily by phase contrast microscope. Now cytometry The analysis of the cell cycle dependent effects of the CDDP and RA treatments was done by DNA flow cytometry on suspensions of single nuclei [7]. Aliquots of lo6 cells, previously stored at -8OOC in 100 ~1 of citrate buffer (250 mM saccarose, 40 mM trisodic citrate, 5% (v/v) DMSO in distilled water) were used to obtain the suspensions of nuclei in a nuclear isolation medium (0.6% Nonidet P-40 (v/v) in a solution containing 0.01 M
Fig. 1.
Morphological differentiation of SK-N-BE [2]C NB cells induced by CDDP (1 pg/ml) and RA (lO+M) after 7 days of treatment. Optical microscope magnification was X 160. A: control cells; B: RA-treated cells; C: CDDP-treated cells.
103
stain. The
measurements were performed using an arc lamp flow cytometer (model ICP22A; Ortho Diagnostic Systems, Westwood, MA) with suitable filters for the DAPI excitation (window between 300 and 400 nm) and for the DAPI fluorescence emission (window between 450 and 490 nm). The fluorescence signals were input to signal processing electronics using 1024 channels for subsequent storage and analysis in a PDP 11-23 computer (Digital Equipment Co., Maynard, MA). Histograms of number of nuclei versus DAPI fluorescence intensity values, in channel number, proportional to DNA content, were obtained. The proportions of cells in the GO +
Gl, S and G2 + M cell cycle phases were evaluated using the method of Baisch et al. 121. Results Cehlar morphology Figure 1 illustrates the changes in cellular morphology of SK-N-BE [2]C NB cells observed after treatment with either RA (B) or CDDP (C) with respect to the control cells growing in absence of these drugs (A). RA induced dramatic alterations in the morphology of the NB cells. Extensive neurites first occurred between 48 and 72 h. The cells treated with CDDP also showed neurite-like
512
N
Control
Control
(48 h)
(7 days)
U
m b e r
CDDP
RA (7 days)
(48 h)
-7
1023
1023
Channel
number
(DNA)
Fig. 2. Cell cycle kinetic analysis by flow cytometry of control cells and of CDDP or RA treated cells. Histograms show number of nuclei plotted versus nuclear DNA content. Top panels: control cells (only medium) at 48 h and 7 days. The main peak corresponds to the GO + Gl cell cycle phase; G2 + M phase cells are at twice values of DNA content; the S phase cells are in between. Bottom panels: cells treated with either CDDP at 48 h after treatment or RA at 7 days after treatment. The total number of nuclei measured in every case was 10,000.
104
structures after 72 h. The maximal elongation of these structures was reached 8 days after CDDP treatment. Cell cycle kinetics Figure 2 shows the most prominent and visually recognizable effects of the CDDP treatment at 48 h and of the RA treatment at 7 days with respect to the corresponding control untreated cells. The induction of a G2 + M block by CDDP at 48h is clearly evident. The control cells, in comparison, show a lower G2 + M peak and a higher GO + Gl peak. The long term effect of RA (at 7 days) consist of visually evident increase of the GO + Gl fraction. of cells. The S and G2 + M cell cycle phase fractions decreased. Quantitative cell kinetic data are reported in Tables 1 and 2. Table 1 reports the cell cycle kinetic short term effects of CDDP and RA. The harvest times for CDDP treated cells are at 1 h, 3 h, 6 h, 9 h, 12 h and 24 h. The control cells and the RA treated cells are shown at 24 h only, since no changes were observed for the
Short term effect of Table 1. human neuroblastoma SKN-BE centage of cells in the GO + Gl, of the cell cycle of control cells, cells. Harvest time
CDDP and RA in the [Z]C NB cell line: perS and G2 + M phases CDDP and RA treated
Cell cycle phases (%) Go + Gl
S
G2 + M
Control 24
55
24
21
CDDP (1 &ml) 1 3 6 9 12 24
52 50 51 48 46 34
29 27 28 32 33 40
19 23 21 20 21 26
RA (lo+ M) 24
64
20
16
Long term effect of Table 2. human neuroblastoma SK-NBE centage of cells in the GO + Gl, of the cell cycle of control cells, cells.
CDDP and RA in the [2]C NB cell line: perS and G2 + M phases CDDP and RA treated
Harvest time
Cell cycle phases
(days)
(%)
GO + GlS
G2 + M
Control 7
57
25
18
CDDP (1 pg/ml) 2 3 4 5 7 9
8 13 26 26 38 50
38 32 36 36 34 26
54 55 38 38 28 24
RA (lo+ M) 7
69
18
13
other harvest times. The fraction of control cells within the GO + Gl, S and G2 + M cell cycle compartments were 55%) 24% and 21%. These values were well reproducible in independent experiments and remained approximately constant within 7 days. A decrease of the GO + Gl fraction of the CDDP treated cells from 55% to 34% was observed at 24 h along with an increase of the cells in the S and G2 + M phases. At 2 days after CDDP treatment the percentage of the GO + Gl cells dropped to 8% while the G2 + M cells piled-up to the level of 54% (Table 2). The G2 + M increase was about 200% with respect to control. Later samples (starting from day 4) clearly showed a progressive reduction of the G2 + M fraction to approximately the control level. The cell kinetics of RA treated cells was as follows: the fraction of cells within the GO + Gl phase of the cell cycle increased by nearly 20% (from 55% to 64%) already at 24 h after the treatment (Table 1). This increase remained almost stable within the successive 7 days. Table 2 reports the percentages of cells in the GO + G 1, S and G2 + M
105
cell cycle phases at 7 days (69%) 18% and 13%, respectively) in comparison to the control values (57%) 25% and 18%). Discussion
The property of certain anticancer drugs to induce cellular differentiation has shown great impact in the research for new and more effective strategies for cancer therapy and has elicited many in vitro studies aimed to better understand the coupling of cell proliferation and differentiation. Among neuroblastoma tumors, the capacity of the neuroblasts to revert from malignant cells to mature gangliolike ceils may be of great interest. Malignant phenotype reversion by differentiation probably takes place for neuroblastoma characterized by the stage IVS (51. These patients, in fact, show either spontaneously or by treatment with minimal doses of certain chemical agents, favorable outcome in a high percentage of cases. Among the drugs which are actually employed in clinical trials as differentiating agents there is retinoic acid (RA). Another interesting drug, recently shown to induce cellular differentiation in human neuroblastoma cell lines [9] and in the K562 erythroleukemia cell line [12], is cisplatin (CDDP). Aim of the present study was to better characterize the ceil differentiation induced by CDDP using a human neuroblastoma cell line (NB cells), For this purpose RA and CDDP treated NB cells have been analyzed in their morphological changes (neurite outgrowth as an index of differentiation) and cell kinetic behavior. Cell kinetics, evaluated by means of temporal sequences of DNA histograms obtained by flow cytometry, should be a reliable tool to assess the degree of cell proliferation and differentiation during the course. of RA and CDDP treatments. According to previous literature reports, terminally differentiated cells should, in fact, be expected to be growth arrested with the DNA content of the Gl phase [6,S, 101. NB cells treated with CDDP changed shape
and showed several neurite-like structures protruding from the cell body and interconnected citoplasmic bridges. The treatment with RA appeared to induce a more complex network of neurite-like structures. These morphological observations give further evidence of the capacity of CDDP to induce morphological changes of neuroblastoma cells [9] and well agree with literature reports for the RA differentiating capacity [ 111. CDDP treated NB cells were first blocked in the G2 + M phases of the cell cycle at about 2 days after the treatment. This phenomenon did not occurr in the RA treated cells. It is to be noticed that the CDDP induced G2 + M block was reversed at later times. At 9 days the CDDP treated cells redistributed among the cell cycle phases approximately as in the original controls. These data are not easy to correlate with neither the neurite outgrowth, considered as a marker of end differentiation, nor with the theoretical expectations (accumulation in the Gl phase of the cell cycle). We think that the observed morphology of the CDDP cells does not probably reflect terminal differentiation but previous stages of differentiation still coupled with proliferation. This concept would agree with our kinetic data which do not show any increase of the Gl phase of the cell cycle. Alternatively, one may think that cell differentiation of NB cells occurs also in cell cycle phases other than G 1. This second hypothesis, in our opinion less likely than the first one, requires, however, further investigation. Beside the induction of neurite-like structures in NB cells, cisplatin was clearly cytotoxic as indicated by high debris levels in the DNA measurements. It is likely that the G2 + M block induced by CDDP reflects a reversible cytotoxic effect. The cytotoxic effect is probably different for the individual cells. Beside the sensitive cells which died, it is likely that cells resistant to the drug were still proliferating. In contrast to these effects observed after CDDP treatment, RA induced an increase of the cells into the GO + Gl phase of the cell
106
cycle in parallel with a net decrease of the fraction of cells in the S-phase. The level of RA induced cell differentiation, as judged by the increase of the Gl phase fraction, was above 20%. This fraction is likely to indicate a cell cycle phase of terminal differentiated growth arrested cells (which was indicated as G, by Scott et al. [lo]). Still a fraction of about 30% were in the S and G2 + M cell cycle phases. RA induced cell differentiation, as judged by cell morphology, was instead observed in almost all RA treated cells. The hypothesis, previously suggested for CDDP, i.e., that neurite-like morphology may not indicate cell cycle arrest and terminal differentiation, may also hold for the RA treatment. More work is necessary to better understand the molecular mechanisms underlaying the effects of CDDP and RA in the induction of neuroblastoma cell differentiation and to test the validity of DNA flow cytometry for monitoring the differentiatiing capacity of these drugs in view of therapeutic protocols in cancer patients. Acknowledgements
This work was supported partly by grant 87.17.020 Ricerca Corrente G. Gaslini Children’s Hospital and partly by Associazione Italiana Ricerca sul Cancro (AIRC). The excellent technical contributions of Mr. E. Infusini and Mr. E . Zeraschi are kindly acknowledged.
2
3
4
5
6
7
8
9
10
11
12
References 13 1
Abemayor, E. and Sidell, N. (1989) Human neuroblastoma cell lines as models for the in vitro study of neoplastic and neuronal cell differentiation. Environ. Health Perspect., 80,3-15.
Baisch, H., Goehde, W. and Linden, W. (1975) Analysis of PCP-data to determine the fraction of cells in the various phases of the cycle. Radiat. Environ. Biophys., 12, 3137. Degos, L., Castaigne, S., Chomienne, C., M.E. Huang, M.E., Tilly, H., Bordessoule, D., Miclea, J.M., Thomas, G., Najean, Y., Abita, J.P. and Wang, Z.Y. (1988) Therapeutic trials of acute myeloid leukemia and myelodysplastic syndromes by LD-ARA C and treatment of promyelocytic leukemia by retinoic acid. In: The States of Differentiation Therapy of Cancer, pp. 361-373. Editors: S. Waxman, G.B. Rossi and F. Takakv. Raven Press, New York. Egorin, M.J. (1988) Clinical trials of polar differentiation agents. In: The States of Differentiation Therapy of Cancer, pp. 389-404. Editors: S. Waxman, G.B. Ross and F. Takakv. Raven Press, New York. Evans, A.E., D’Angio, G.J. and Randolf, J. (1971) A pro- I posed staging for children with neuroblastoma. Cancer, 27,374-378. Giaretti, W., Moro, G., Quarto, R., Bruno, S., Di Vinci, A., Geido, E. and Cancedda, R. (1988) Flow cytometric evaluation of cell cycle characteristics during in vitro differentiation of chick embryo chondrocytes. Cytometry, 9, 281-290. Giaretti, W., Sciallero, S.Bruno, S. Geido, E., Aste, H. and DiVinci, A. (1988) A flow cytometry of endoscopically examined colorectal adenomas and adenocarcinomas. Cytometry, 9,238-244. Nadel-Ginard, B. (1978) Commitment, fusion and biochemical differentiation of a myogenic cell line in the absence of DNA synthesis. Cell, 15,855-864. Parodi, M.T., Varesio, L. and Tonini, G.P. (1989) Morphological change and cellular differentiation induced by cisplatin in human neuroblastoma cell lines. Cancer Chemother. Pharmacol., 25, 114-116. Scott, R.E., Florine, D.L., Wille, J.J., and Yun, K. (1982) Coupling of growth arrest and differentiation at a distinct state in the Gl phase of the cell cycle: G,. Proc. Natl. Acad. Sci. USA, 79,845-849. Sidell, N., Altman, A., Haussler, M.A. and Seeger, R.C. (1983) Effects of retinoic acid (RA) on the growth and phenotype expression of several human neuroblastoma cell lines. Exp. Cell Res., 148, 21-30. Tonini, G.P., Parodi, M.T., Bologna, R., Persici, P. and Cornaglia-Ferraris. P. (1986) Cisplatin induces modulation of transferrin receptor during cellular differentiation in vitro. Cancer Chemother. Pharmacol., 18.92. Tonini, G.P., Radzioch, D., Gronberg, A., Clayton, M., Blasi, E., Benetton, G. and Varesio, L. (1987) Erythroid differentiation and modulation of c-myc expression induced by antineoplastic drugs in the human leukemic cell line K562. Cancer Res., 47,4544-4547.
