inhibition of the growth of glioblastomas by cgp 41251, an ... · vol. 2, 1017-1030, june 1996...

Vol. 2, 1017-1030, June 1996 Clinical Cancer Research 1017

Inhibition of the Growth of Glioblastomas by CGP 41251, an

Inhibitor of Protein Kinase C, and by a Phorbol Ester

Tumor Promoter1

Martin Begemann,2 Sharafadeen A Kashimawo,

Yu-jeong A. Choi, Susan Kim,

Kim M. Christiansen, Gregg Duigou,

Marcel Mueller, Ira Schieren, Subrata Ghosh,

Doriano Fabbro, Nina M. Lampen,

Daniel F. Heitjan, Peter B. Schiff,

Jeffrey N. Bruce, and I. Bernard Weinstein

Columbia-Presbyterian Cancer Center [M. B., S. K., K. M. C., I. S.,

I. B. WI, Departments of Radiation Oncology [S. A. K., P. B. 5.1,

Neurological Surgery [Y. A. C., D. G., S. G., J. N. B.], Genetics and

Development [K. M. C., I. B. W.], and Division of Biostatistics in theSchool of Public Health [D. F. H.], Columbia-Presbyterian Medical

Center, New York, New York 10032; Department for Cancer andInfectious Diseases, Ciba-Geigy, CH-4002 Basel [M. M., D. F.],

Switzerland; and Department of Electron Microscopy, Memorial

Sloan-Kettering Cancer Center, New York, New York 10021[N.M.L.1

ABSTRACT

Protein kinase C (PKC) plays a central role in signal

transduction pathways that mediate the action of certain

growth factors, tumor promoters, and cellular oncogenes.

To explore whether PKC might be an appropriate target forthe chemotherapy of human brain tumors, cell lines were

established from five glioblastomas, one mixed gliosarcoma

and glioblastoma, two astrocytomas, and one choroid plexuscarcinoma. The staurosponine derivative CGP 41251, aninhibitor of PKC, inhibited cell proliferation in all nine cell

lines with an IC50 in the range of 0.4 p�M. Drug withdrawaland clonogenicity assays showed that CGP 41251 induced an

irreversible growth arrest. Three cell lines were examined in

detail: two human glioblastoma cell lines, GB-i and GB-2,

and one gliosarcoma cell line, GS-i. All of these three cell

lines were highly aneuploid and displayed morphologies and

immunohistochemical markers characteristic of the glial lin-

eage. The compound 12-O-tetradecanoylphorbol-13-acetate

(TPA), a tumor promoter and activator of PKC, also inhib-ited the growth of these cell lines. CGP 41251 in combination

with TPA caused further growth inhibition. Cultures treated

with CGP 4i251 displayed an increase in the fraction of cellsin G2-M, a decrease of cells in S phase, and no consistent

effect on G0-G1. Immunohistochemical analyses demon-

strated that growth inhibition by CGP 41251 was associated

with the formation of giant nuclei with extensive fragmen-

tation and apoptotic bodies. These effects of CGP 41251were abrogated by withdrawal of serum from the medium

or by exposure of these cells to aphidicolin, actinomycin D,

cycloheximide, or TPA. In contrast to the effects seen with

the glioblastoma cell lines, nontransformed astrocyte lines

remained viable in the presence of 0.4 and 0.8 �iM CGP41251 and displayed only a slight increase in the fraction of

giant nuclei with fragmentation. The antitumor activity ofCGP 4i25i was demonstrated in vivo against xenografts of

the glioblastoma cell lines U87 MG and U373 MG. These

findings suggest that CGP 41251 might be a useful agent forthe treatment of glioblastomas.

INTRODUCTION

PKC3 is a family of phospholipid-dependent serine and

threonine kinases that has been shown to have the highest

activity in the central nervous system (1-3). About 13 different

isoforms of PKC have been described today, which differ in

their cofactor requirement and are differentially expressed dur-

ing development and in various tissues. Most of these isoforms

are activated by diacylglycerol, which is generated after activa-

tion of phospholipase C. Diacylglycerol binds to the regulatory

domain of various PKC isoforms. Phorbol esters, such as TPA,

also bind to the regulatory domain of these same isoforms and

are potent activators of PKC. Overexpression of PKC in R6 rat

fibroblasts using retroviral-mediated gene transfer demonstrates

that individual isoforms of PKC have different biological ef-

fects. Thus, high level expression of PKC�31 leads to growth

abnormalities, including increased frequency of focus formation

and growth in agar (4); overproduction of PKCa does not cause

any growth defects (5), whereas overexpression of PKC#{128}leads

to malignant transformation and tumorigenicity in athymic mice

(6).

Due to the central role of PKC in growth control, inhibition

of this enzyme might be a particularly promising approach for

cancer therapy (3). Recent studies indicate that the staurosponne

derivative CGP 41251 inhibits PKC in subcellular assays and

also inhibits tumor cell growth in cell culture and in athymic

mice (7-9). Because several isoforms of PKC are normally

expressed at high levels in the central nervous system and

Received 10/24/95; revised 2/12/96; accepted 3/1/96.

I This work was supported by National Cancer Institute Grants CA

02656 (to I. B. W.) and P20-CA60175-02 (to J. N. B.) and an award

from The Brain Tumor Society (to I. B. W.).

2 To whom requests for reprints should be addressed, at Columbia-

Presbyterian Cancer Center, Hammer 1509, 701 West 168th Street, New

York, NY 10032. Phone: (212) 305-6924: Fax: (212) 305-6889.

3 The abbreviations used are: PKC, protein kinase C; TPA, 12-0-

tetradecanoylphorbol-1 3-acetate; BrdUrd, bromodeoxyuridine; MAP2,

microtubulin-associated protein 2; DAPI, 4’,6-diaminidino-2-phenylin-

dole; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bro-

mide: GFAP, glial fibrillary acidic protein; NF16O, neurofilament 160.

Research. on March 29, 2020. © 1996 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

100

10

A

0

U)

a)C)

0

a).0

E

B100

0.

a)

(I)

a)C)

0

a).0

Elz

I

0123456789

Day

-U-CGP41251 .4uM

-�--CGP41251 .8uM

-s’---CGP42700 .4uM

-D--CGP42700 .8uM

-D-TPA Ing/mi

-�‘--TPA lng/ml + CGP41251 .4uM-*--TPA Ing/mi + CGP42700 .4uM

.1

0 1 23456789

1018 Inhibition of the Growth of Glioblastomas by CGP 41251

Day

because glioblastomas are particularly refractory to treatment

with conventional chemotherapeutic agents, the present study

was undertaken to examine the effects of CGP 4 1 25 1 on a series

of human brain tumor-derived cell lines. This study demon-

strates that CGP 41251 is a potent growth inhibitor of glioblas-

toma cells and that this inhibition is, paradoxically, enhanced by

TPA. Growth curves with drug reversal as well as clonogenicity

studies suggest that CGP 4125 1 acts both through a cytocidal

and cytostatic mode. Flow cytometry studies indicate an accu-

mulation of cells in G,-M, and morphological studies indicate

that CGP 4 1 25 1 leads to the formation of giant nuclei and the

formation of apoptotic bodies. Furthermore, the antiproliferative

property of COP 4125 1 was confirmed in xenografts of human

glioblastoma cell lines U87 MG and U373 MG.

MATERIALS AND METHODS

Cell Lines and Cell Culture Conditions. All cell lines

were established in the Brain Tumor Laboratory of the Depart-

ment of Neurosurgery, Columbia-Presbyterian Medical Center.

In brief, fresh tumor specimens were placed in 15% FCS in the

operating room. In the tissue culture laboratory, the tumor cells

were mechanically dispersed with 18-gauge needles or enzy-

matically with trypsin, depending on the tumor consistency.

Dispersed cell explants were then placed in tissue culture dishes

containing DMEM with 20% FCS. The cell lines were also

maintained in DMEM supplemented with 20% FCS (Hybrimax:

Sigma Chemical Co., St. Louis, MO). TPA (LC laboratories,

Woburn, MA) and COP 41251 and COP 42700 (Ciba-Geigy,

Base!, Switzerland) were dissolved in DMSO. To obtain growth

curves, cells were plated at a density of l04/well in 6-well plates

(3-cm diameter) and grown in complete medium with or without

the indicated drugs (Fig. 1). Cell counts were determined on

alternate days in triplicate wells during the subsequent 8 days

with a Coulter Counter. Actinomycin D (Sigma) was dissolved

in DMSO and used at a final concentration of 2.5 pg/ml.

Cycloheximide (Sigma) was dissolved in ethanol and used at a

final concentration of 10 p.g/ml. Aphidicolin (Sigma) was dis-

solved in DMSO and used at a final concentration of I pg/ml.

In all cases, the final concentration of DMSO in the media was

less than 0. 1 %, which had a negligible effect on cell growth.

Protein Extraction and Immunoblotting. The various

cell lines were grown to subconfluency, washed three times with

PBS, and immediately lysed in a 1% SDS buffer [20 msi

Fig. I Growth curves of GB-I cells in the presence of the indicated

concentrations of TPA (A) and in the presence of the indicated concen-

trations of CGP 41251, CGP 42700, or these drugs plus TPA (B).DMSO, the solvent control for the studies with TPA; D20, the control

obtained in the absence of any drugs. In the statistical analyses, a model

with separate slopes of log cell counts for each TPA dose group was

estimated (A). A test for equality of all slopes was significant at P <

0.001. Slopes of the control group, the 0.01 group. and 0.1 group were

significantly positive, and slopes for the three remaining (high-dose)

groups were not significantly different from 0 at the 0.05 level. Bars,

SD. Data in B, TPA alone, CGP 41251 (at both doses) alone, and CGP

42700 (at 800 nM only) alone, all produced inhibition of proliferation (P< 0.001), although the effect of CGP 42700 was modest. The addition

of COP 41251 to TPA produced additional significant inhibition, as did

the addition of CGP 42700, although again the effect of COP 42700 was

modest. Bars, SD. For additional details, see “Materials and Methods.”



Clinical Cancer Research 1019

Tris-HC1 (pH 7.5) and 2 mrvi EDTA] and boiled for 5 mm. The

protein content in these total extracts was determined according

to the method of Bradford (Bio-Rad). The equivalent of 50 �i.g

protein cell extracts was fractionated by electrophoresis in 7%

polyacrylamide gels, transferred to nitrocellulose, and exposed

to antibodies specific to various isoforms of protein kinase C as

indicated [PKCa, rabbit affinity purified IgO (OIBCO-BRL);

PKCs �, #{128},and �, rabbit polyclonal antisera; Refs. 9 and 10]. The

specificity of the antisera has been demonstrated previously (9).

Clonogenicity. Cells were plated in 100-mm dishes with

10 ml DMEM supplemented with 10% FCS (OIBCO) and

allowed to attach for about 24 h. The numbers of cells to be

plated were chosen such that 50 to 100 colonies would survive

after a specified treatment. COP 41251 or DMSO as control

were then added, and the cells were incubated for an additional

24, 48, 96, 120, and 144 h. Following these incubation periods,

the medium was removed by aspiration and replenished with

DMEM supplemented with 10% FCS without additives. The

cultures were then incubated for 14 days. Cells were fixed in

75% methanoll25% acetic acid and stained with crystal violet.

Collections of 50 or more cells were scored as colonies and

counted for each plate. Each data point was obtained in triplicate

assays. The survival fraction after drug treatment is given as a

percentage of the DMSO control. Control plating efficiencies

were 10% for OB-l, 14.5% for OB-2, 12% for GB-3, and 8%

for 05-1.

BrdUrd Incorporation Studies. BrdUrd labeling (Am-

ersham) was carried for 60 mm at 37#{176}Cin DMEM supple-

mented with 20% FCS. Subsequently, the cells were washed in

PBS and fixed in glacial acetic acid/ethanol for 30 mm. The

endogenous peroxidase activity was quenched by incubation

with methanol/3% peroxide for 30 mm at room temperature.

Subsequently, the dishes were treated with blocking solution

(horse serum, 2% BSA, 0.1% Triton X-100, and PBS) for 1 h.

Then anti-BrdUrd mouse monoclonal antibodies (Amersham)

were applied for 1 h, and the plates were washed in PBS. The

mouse monoclonal antibody was detected with the ABC system

from Vector Laboratories (Burlingame, CA). Staining was car-

ried out with diaminobenzidine using the diaminobenzidine

substrate kit from Vector Laboratories.

Flow Cytometry. Exponentially growing cultures were

trypsinized, collected, and washed twice with PBS. Cell pellets

were suspended in 1 ml PBS and fixed in 5 ml of 70% ethanol

and stored at 4#{176}C.On the day of the analysis, cells were

collected by centrifugation, and the pellets were resuspended in

0.2 mg/mi of propidium iodide containing 0.6% NP-40. RNAse

(1 mg/mI) was added, and the suspension was incubated in the

dark at room temperature for 30 mm. The cell suspension was

then filtered through a 60 jim Spectra mesh filter and analyzed

on a Coulter EPICS 753 flow cytometer for DNA content. The

percentage of cells in different phases of the cell cycle was

determined using a ModFit 5.2 computer program.

Immunohistochemistry. Cells were grown on LabTek

chamber slides that were precoated with poly-D-lysine (Collab-

orative Research) in DMEM (0.1 mg/ml for 1 h). After the

indicated exposure (see “Results”), the cells were fixed with

90% ethanol, 5% glacial acetic acid, and 5% H2O for 30 mm.

The cell preparation was blocked in 2% BSA and incubated

sequentially with various primary and secondary antibodies.

Antibodies against the following proteins were used: OFAP

(rabbit polyclonal antibodies from Sigma and mouse mono-

clonal antibody GAS from Boehninger-Manmheim); neuron-spe-

cific enolase (rabbit polyclonal antibodies; ICN); NF16O (mouse

momoclomal antibody; Accurate Scientific, NN16); neurofila-

ment 200 (rabbit polyclonal antiserum; Sigma); and MAP2

(mouse monoclonal antibody AP420; Sigma). The secondary

antibodies used were goat-antimouse-conjugated FITC (Sigma)

at a 1 :50 dilution and goat-antirabbit FITC-conjugated antibod-

ies (TAOO) at a 1:100 dilution. The nuclei were counterstained

with DAPI (Boehringer-Mannheim) at a final concentration of

1.5 p.g/ml in PBS for 30 mm at room temperature and visualized

with fluorescence microscopy. For quantitation of the nuclear

morphology under various culture conditions, the morphologies

of the nuclei were separated into six categories (oval, frag-

mented, multinucleated, mitotic, pyknotic, and multilobular).

For each data point, four fields, each containing at least 150

nuclei, were evaluated.

Transmission Electron Microscopy. Cells were fixed

as a pellet in 2.5% glutaraldehyde in 1 ,4-piperazinediethanesul-

fonic acid buffer overnight, rinsed in 1 ,4-piperazinediethanesul-

fonic acid buffer, followed by postfixation in 2% 0504 for 1 h.

The samples were then rinsed in distilled H2O, and this was

followed by dehydration in a graded series of alcohol of 50, 75,

and 95% through absolute alcohol, followed by propylene ox-

ide, and then overnight in propylene oxide:PolyfBed 812 (1:1).

The samples were embedded in Poly/Bed 812 in a 60#{176}Coven.

Ultrathin sections were obtained with a Reichert Ultracut S

microtome. The thin sections were photographed using a JEOL

1200-EX electron microscope.

MTT Viability Assay. The MiT assay was performed as

described ( 1 1 ). Cells (2-5 X l0�) were seeded in each well

(0.38 cm2 area) of 96-well plates in DMEM plus 20% FCS. On

the following day, the medium was replenished with the drugs

at the indicated dosages. After various times (see “Results”), the

medium was removed, and 22 �il of MU solution (5 mg/mi in

PBS; Sigma) were added to 200 p.1 of 20% FCS in DMEM in

each well and incubated at 37#{176}Cfor 3 to 4 h. This medium was

then discarded, 100 p.1 of DMSO were added to each well, and

the plates were placed for 15 to 30 mm on a shaker. The color

change was recorded by measuring the absorbance at 570 mm on

an automated multiwell reader. Each data point represents the

mean of quintuplicate assays and was calculated as a fraction of

the control.

In Vivo Experiments. The U87 MG (HTB- 14) and U373

MO (HTB-17) human glioblastoma cell lines were obtained

from the American Tissue Culture Collection (12, 13). Both

tumor lines were maintained by serial passage by implanting

tumor fragments (approximately 25 mg) s.c. into the left flank of

female BALB/c nude mice (Bolmholtgard, Copenhagen, Den-

mark) with a 13-gauge Trokar needle under Forene (Abbott,

Cham, Switzerland) narcosis. Each group of mice contained six

animals. Drug treatment was started on day 5 after transplanta-

tion and continued until day 17. Compounds were prepared as

follows. For iv. applications, COP 41251 was dissolved in

DMSO, Tween 80, and NaC1. In brief, a stock solution contain-

ing 150 mg/ml COP 4125 1 was prepared by adding the appro-

priate amount of compound to DMSO containing 1% Tween 80

dissolved at room temperature. Prior to use, the clear stock



1020 Inhibition of the Growth of Glioblastomas by CGP 41251

Table 1 Description of cell lines”

GB-l GB-2 GS-l

Age (yr)

Gender61

F71

M61

FDiagnosis

Protein expression

GFAP

Glioblastoma multiforme

+

Glioblastoma multiforme

+

Glioblastoma multiforme and gliosarcoma

+

NF16O - - -

MAP2 - -

Doubling time (h)Karyotype

Tumorigenicity in nude mice

58Aneuploid

+

46Aneuploid

+

62Aneuploid

-

a For additional detail, see “M aterials and Methods.”

solution was diluted 1 :20 (v/v) with sterile 0.9% NaC1, resulting

in a milky suspension (final concentration: 7.5 mg/mi COP

41251, 5% DMSO, 0.5% Tween 80, and 0.9% NaC1). For p.o.

applications, COP 4125 1 was prepared using Gelucire 44/14. In

brief, a waxy solid formulation of COP 41251 (18% w/w in

Gelucire 44/14; Oattefosse#{233}, France) was obtained from CIBA

Pharmaceutical. Prior to use, 16.7 mg of the active ingredient

corresponding to 83.3 mg of the waxy formulation were added

to 1 ml of sterile water, and the mixture was sonicated in an

ultrasonic bath for 10 mm, resulting in a milky suspension that

was given to the mice for 13 consecutive days once daily p.o.

(7). All placebo-treated animals received the matrix without

COP 41251 (gelucire/water) p.o. Tumor growth was followed

by measuring perpendicular tumor diameters. Tumor volumes were

calculated as described using the formula (‘rr X L X D2/6), where

L is the length and D is the diameter (14, 15). Mean tumor volume

(± SD of six animals) was expressed in cubic centimeters.

Statistical Methods. For experiments on cell prolifera-

tion, the cell counts were analyzed on the log scale. A linear

model was fit with a common intercept (number of cells at day

0) with a separate slope for each treatment group. The equality

of slopes was tested with the regression F test. For BrdUrd

incorporation studies, the data of the percentage of labeled

nuclei were analyzed on the logistic scale. The data were ana-

lyzed with an ANOVA model incorporating a main effect of cell

line, a slope for the drug dose, and an interaction of cell line and

dose. In the clonogenicity studies, the surviving fractions were

transformed to the log scale. The data were analyzed by a linear

model with a main effect of time, a slope for drug dose, and the

interaction of time and dose. We conducted separate analyses

for each cell line. In all analyses, we deleted the data from dose

group 0, because they were scaled to have mean SF = 1 . To

analyze the in vivo tumor growth data, the tumor volumes were

transformed to the log scale. We fit a mixed model with a

random animal effect and a separate slope for each treatment

group (placebo, p.o. administration, and i.v. administration). To

analyze the data on nuclear morphology, a separate two-way

table of treatment by morphology was formed for each of three

replicate assays. To eliminate cells with the smallest counts, our

analyses combined the multinucleated, multilobular, and py-

knotic categories. To compare the treatment groups, we exe-

cuted a Mantel-Haenszel test for general association, stratified

by replication. All analyses were carried out in SAS version

6.09 (SAS Institute, Cary, NC). We fit linear models in Proc

Table 2 Growth inhibition by CGP 41251 in glioblastomas,astrocytomas, and a choroid plexus carcinoma cell line

Patient Diagnosis

IC50 (p.M) for

CGP 41251”

GB-i Glioblastoma multiforme 0.2GB-2 Glioblastoma multiforme 0.4

GB-3 Glioblastoma multiforme IGB-4 Glioblastoma multiforme 0.6GB-S Glioblastoma multiforme 0.8GS-l Glioblastoma multiforme and gliosarcoma 0.2

AT-! Astrocytoma 0.1AT-2 Astrocytoma 0.4

CP-l Choroid plexus carcinoma 0.3

a Determined after a 6- to 8-day exposure of cultures to the drug, as

described in Fig. 1.

Olm, mixed models in Proc Mixed, and executed Mantel-

Haenszel tests in Proc Freq software programs.

RESULTS

Characteristics of Cell Lines. Various properties of the

three major cell lines used in this study are summarized in

Tables 1 and 2. All three cell lines were positive for the

glial-specific protein OFAP and negative for the neural markers

MAP2 and NF16O. Cultures of GB-i and OB-2 cells were

heterogeneous by morphological criteria, displaying cells with

squamous, astrocytic, and polygonal morphologies. OB-l and

GB-2 cells were arranged in fascicular patterns and piled up. In

contrast, cultures of OS- 1 cells formed flat mosaic patterns of

polygonal cells. Occasionally, cells with astrocytic morpholo-

gies were found in 05-1 cultures. All three cell lines were

highly aneuploid and showed continuous chromosomal rear-

rangements during serial passages (data not shown). The cell

lines GB-I and GB-2 were tumorigenic when injected into

athymic mice (nu/nu). The histopathology of the grafted tumors

exhibited highly infiltrative tumor growth, nuclear polymor-

phism, pseudopalisading, and expression of OFAP. Two astro-

cyte lines were established from nontumorous tissue: A-l from

normal appearing brain tissue obtained from a patient with a

glioblastoma multiforme; and A-2 from brain tissue removed

from a patient with epilepsy. A-l was shown to be nontumori-

genic in athymic mice. A-l and A-2 both exhibited astrocytic

morphologies (data not shown).



1 2 3456789101112

PKCa 84 -� � ___ - � -

PKC� 8463-

PKCc 112

84-

k� .i��Ii �

84- � _� �PKC�

63- �


Fig. 2 Immunoblot for PKCs

a, h, #{128},and �. GB-l cells werecultured for 2 days in the pres-

ence of DMEM supplemented

with 20% FCS alone (Lane 1)

or with 0.1% DMSO (Lane 2);

TPA at the following concen-trationsof0.Ol,0.l, 1, 10, and

100 ng/ml (Lanes 3-7): CGP41251 at 0.4 and 0.8 mM

(Lanes 8 and 9); CGP 42700 at

0.8 m�i (Lane /0); and CGP

41251 (0.8 mM) plus TPA at 1

or 10 ng/ml (Lanes 11 and 12).

In the PKC� immunoblot, thesolid arrow denotes the com-

plete PKC�, and the open ar-

row, the PKM form of PKC�.

For additional details, see “Ma-terials and Methods.”

Effects of TPA and CGP 4i25i on Cell Proliferation.

The cell line GB- I was first grown in the presence of increasing

concentrations of TPA (Fig. 1A). TPA markedly inhibited

growth with an IC�() of about 0. 1 ng/ml. Then the effect of the

PKC selective inhibitor COP 4125 1 on the growth of GB-i cells

was tested either alone or in combination with TPA. COP

42700, a close structural analogue of COP 4125 1 that does not

inhibit PKC (7, 9), and the solvent DMSO were used as controls.

As shown in Fig. 1B, proliferation ceased at 0.4 �.tM of the active

compound COP 4125 1, and cell killing was observed with 0.8

p.M of COP 41251 (P < 0.001). COP 42700 was not signifi-

cantly different from growth medium at 0.4 mM; there was a

modest but significant inhibition at 0.8 mivi. Thus, both TPA

(Fig. 1A) and COP 41251, as well as COP 42700 at a modest

level (Fig. I B), inhibit the proliferation of GB- 1 in a dose-

responsive manner. The inhibitory effect of 0.4 p.M COP 412�1

plus 1 ng/ml TPA was greater than the effect of either agent

alone (P < 0.001). The inhibitory effect of TPA plus COP

42700 was modest but significantly greater then that of TPA

alone (Fig. 1B; P < 0.001). This suggested that COP 41251

might be cytocidal when tested in combination with TPA.

Trypan blue exclusion studies, however, did not show increased

cell staining of COP 4 125 1 or COP 4 125 1 plus TPA-treated

cultures when compared to untreated cultures or to COP 42700

or DMSO-treated cultures (data not shown).

The growth-inhibitory effects of COP 4125 1 were analyzed

across a range of concentrations of COP 4125 1 in the GB-i and

eight additional brain tumor-derived cell lines, and the results

are summarized as IC5() (i.e., the concentration required to

produce about a 50% inhibition of growth) in Table 2. The five

glioblastoma-derived cell lines (GB-! to OB-S) displayed IC50s

between 0.2 to 1 p�M. This value was 200 n�i for the OS-i

gliosarcoma cell line; 0. 1 and 0.2 �j.M, for the two astrocytoma

cell line, AT- 1 and AT-2, respectively; and 0.3 �iM for the

choroid plexus carcinoma cell line CP-l (Table 2).

As with the OB-l cell line (Fig. 1B), when TPA was

combined with COP 4125 1, there was further inhibition of

growth of the OB-2 and 05-1 cell lines, and the compound COP

42700 caused a modest growth inhibition in OS- 1 and no

inhibition in OB-2. When the GB-i or 05-1 cells were exposed

to COP 4125 1 at 0.8 p�M for 4 days and the drug then removed,

the proliferation of the cells failed to resume. However, under

similar conditions, the growth inhibition obtained with the OB-2

cells was partially reversed when the drug was removed (data

not shown).

Effects of CGP 41251 and TPA on the Expression ofSpecific Isoforms of PKC. Fig. 2 displays an immunoblot

analysis of the expression profile of several isoforms of PKC in

GB-I cells and the effects of exposing these cells to various

drugs. PKCs a, �, #{128},and � were detected in total cell extracts but

not the PKC isoforms �3 and -y (data not shown). The molecular

weights of PKCs were as follows: Mr 86,000 (PKCa); Mr

7 1 ,000 (PKC�); a doublet at Mr 96,000 (PKC#{128});and Mr 78,000

(PKC�). We also detected an approximately Mr 58,000 species

cross-reacting with anti-PKC� antisera (Fig. 2, open arrow).

This moiety most likely represents the COOH-terminal catalytic

fragment of PKC� termed PKM (16, 17). The expression of

PKM has been demonstrated in rat brain (16, 17) and might play

a role in long-term potentiation (17). After exposing GB-I cells

for 48 h to TPA at 0.01, 0.1, and 1 ng/ml, there was no

significant effect on the expression levels of the four isoforms

(Fig. 2), although the 1 ng/ml dose caused marked growth

inhibition (Fig. 1A). Exposure to 10 and 100 ng/ml TPA caused

marked down-regulation of PKCs a and e. Exposure to COP

41251 at 0.4 or 0.8 J.LM, or the control compound COP 42700,

caused no appreciable loss of these four isoforms of PKC (Fig.

2, Lanes 8-JO). The combination ofCOP 41251 (0.8 jiM) with

TPA at 1 or 10 ng/ml (Fig. 2, Lanes ii and 12) showed

complete down-regulation of PKCs a and #{128}.Similar profiles of

these PKC isoforms were found with the OB-2 and OS-i cells

(data not shown).

BrdUrd Incorporation Studies. To further analyze the

antiproliferative properties of COP 41251 , the GB-!, OB-2, and

05-1 cell lines were selected for further detailed studies.

BrdUrd incorporation studies were performed to determine pos-

sible effects of COP 4125! on DNA synthesis. Fig. 3 summa-



0

1022 Inhibition of the Growth of Glioblastomas by COP 41251

Ctrl

CGP41251 200 nM

CGP41251 400 nM

CGP41251 800 nM

10 20 30 40

Fig. 3 Effects of COP 4125 1 on BrdUrd incor-

poration. Cultures of OS-I, GB-i, and OB-2 cellswere treated for 4 days with 200, 400, or 800 nMCOP 4125 1 . The cultures were then incubated

with BrdUrd for 1 h, fixed, and prepared for

staining as described in “Materials and Methods.”

The figure indicates the BrdUrd labeling indices,

i.e., the percentage of nuclei positive for BrdUrd.

In the statistical analyses, the slope of the dose of

COP 41251 was significant (P < 0.001), as was

the main effect of the cell line. The statistical

interaction was not significant. Thus, the analyses

suggest a strong inhibitory effect of COP 41251,

which is present in all three of the cell lines

studied. Bars, SD.

% labeled nuclei

rizes BrdUrd labeling indices obtained after a 4-day treatment

with various concentrations of COP 41251. In untreated cul-

tures, the GB-2 cells exhibited the highest labeling index (about

33%), as compared to GB-l and OS-i cells (about 12%). The

addition of COP 41251 (200 to 800 nM) resulted in a dose-

dependent inhibition of BrdUrd incorporation in all three cell

lines. At 800 flM CGP 41251, the labeling index had decreased

by about 80% in all three cell lines when compared to the

untreated cultures. Time course studies with the GB-2 cell line

indicated that inhibition of BrdUrd incorporation began within

24 h after exposure to the drug and continued to decline there-

after (data not shown). Inspection of nuclei revealed that those

nuclei that were labeled were predominantly giant nuclei with a

fragmented morphology (see below), suggesting that these cells

were undergoing re-replication of DNA without entering mito-

sis. Exposure to TPA (100 ng/ml) for 2 days also led to a

decrease in BrdUrd incorporation with both the GB-2 and 05-1

cells, and this inhibitory effect was further augmented by the

addition of COP 41251 (data not shown).

Clonogenicity Assays. To further analyze the antiprolif-

erative effects of COP 41251, clonogenicity studies were per-

formed. GB-i, GB-2, GB-3, and OS-i cells were plated at low

density and incubated for various times with different doses of

COP 41251. Subsequently, the medium was replaced with regular

growth medium, and the cells were cultured in the absence of the

drugs for an additional 2 weeks, after which the surviving fractions

were determined (see “Materials and Methods”). As shown in Fig.

4, all of the four cell lines showed only partial inhibition when

exposed to COP 4125 1 for only 24 h. When exposed to the drug for

48 h or longer, marked inhibition and a clear dose-response rela-

tionship was seen in all of the cell lines, although OB-2 cells were

less responsive then the other three cell lines. In addition, at a given

dose of drug, an increased duration of treatment led to decreased

survival. For example, GB-l cells treated with 1 p.si COP 41251

for 24, 48, 96, and 144 h showed survival fractions of 24, 9, 0.7,

and 0.6%, respectively. The survival of the cells also declined in a

dose-response fashion. Thus, the survival fractions of GB-l cells

after a 96-h exposure to 100, 200, 400, 600, 800, and 1000 ruvi COP

41251 were 53, 43, 13, 1.7, 1.3, and 0.7%, respectively. Finally, the

survival fractions of the particular cell lines were compared at a

given dose for a given duration of exposure. At the longest expo-

sure (120 to 144 h), the LD� (i.e., the concentration required to

produce a 10% survival) for COP 41251 was 360, 740, 320, and

350 flM in GB-i, OB-2, OB-3, and OS-i cells, respectively. At the

maximum dose (1000 rust) and after the longest exposure (120 to

144 h), the surviving fractions were 0.i, 0.6, 3, and 6.2% for the

OB-3, OB-l, OS-l, and OB-2 cells, respectively. These results

demonstrate that OB-l and GB-3 cells are the most sensitive, OS-l

cells are intermediate, and OB-2 cells are least sensitive to inhibi-

tion by COP 4i251. In summary, the effects ofCOP 41251 on cell



z0I-

0

LI.

C,

z>

U)

10

.1

.01

.001

z0I-.

0

LI.

C,z>

U)

0

10

200 400 600 800 1000 1200DOSE (nM)

0 200 400 600 800 1000 1200DOSE (nM)

.1

z0I-

0

LI.

C,z>>

U)

.01

z0I-

0

U.

C,

z

U)

.0010 200 400 600 800

DOSE (nM)1000 1200 0 200 400 600 800 1000 1200

DOSE (nM)


Fig. 4 Clonogenicity assays. Cultures of GB- I . GB-2, GB-3, and GS- 1 cells were plated at low density and then treated with various concentrationsof COP 4125 1 for different periods of times, as indicated, and then assayed for clonogenicity as described in “Materials and Methods.” The graphsindicate the surviving fractions as a function of dose and duration of exposure. There was a statistically significant interaction (P < 0.001 ) between

hours of exposure and the dose of COP 4 125 1 . This means that the slope of the SF dose curve depends on the hour of exposure. Moreover, in each

case, the slope of the SF-dose curve becomes more steep with increasing time of exposure, suggesting a time-dependent sensitivity.

survival depends on both the dosage and the duration of exposure

to the drug and the particular cell line.

Cell Cycle Analysis. To investigate possible effects on

cell cycle progression, exponentially dividing cultures of

GB-I and OB-2 cells were treated with the DMSO solvent

(control) or with COP 41251 (800 nM) for 4 days, and then

the cells were processed for flow cytometry (Table 3). With

both cell lines, the drug caused a 2- to 3-fold increase in the

fraction of the total cell population that was in O,-M and a

decrease in the fraction of cells in S phase. With the GB-i

cells, there was no significant change in the fraction of cells

in O�-O�, and in OB-2 cells, there was a slight decrease in

this fraction. The inactive compound, CGP 42700, had no

effect on cell cycle distribution in cultures of GB- 1 (Table 3)

and OB-2 cells (data not shown). Additional studies, not

shown here, indicated that exposure of GB-I and OB-2 cells

to TPA (100 ng/ml) also led to an increase of cells in O2-M

and a decrease in S phase: the combination of TPA plus COP

41251 further enhanced the accumulation of cells in O,-M.

Table 3 Cell cycle analysis

Exponentially growing cultures of GB-I and GB-2 cells were treated

for 4 days with the indicated drugs and then analyzed for DNA content by

flow cytometry. The data indicate the percentage of the total cell population

in the indicated phase of the cell cycle and are the mean values of duplicate

experiments. For additional details, see “Materials and Methods.”

G�-O

GB-I

S G2-M G�-G

GB-2

S G,-M-�-�

Control 59.2 30 10.8 75.3 15.0 9.7COP 41251 0.8 p.M 61.4 9.6 29 68.5 10.4 21.1

COP 42700 0.8 p.M 52 35.2 12.8 ND” ND ND

,‘ ND, not determined.

Formation of Giant Nuclei with Extensive Fragmenta-tion. To investigate possible changes at the level of cellular

morphology and differentiation, cultures of GB-i, OB-2, and

05-1 cells were incubated with COP 41 25 1, TPA, or a combi-



.�

10

#{149}1�

1024 Inhibition of the Growth of Olioblastomas by COP 4125!

It-

I,

.�




Table 4 Nuclear morphology of OB-2 cells after exposure to COP 41251 and TPA

OB-2 cells were plated on LabTek chamber slides and incubated for 4 days with COP 41251 (0.8 p.M) or TPA (100 nglml) or the combinationof both. Then the cultures were fixed and processed for staining as described in “Materials and Methods.” Denoted are the means of triplicate assays

with standard variation in percent. In all three replications, there was significant association by the x2 test (P < 0.001) of treatment group with

morphology. A Mantei-Haenszel test for general association combining results across the three replicates is significant at P < 0.001 . The proportion

of fragmented cells in the COP 41251-treated group was significantly smaller than the fraction in the group treated with both COP 41251 and TPA

(P < 0.001 by the Mantei-Haenszei test).

Nuclear morphology

Treatment Oval Multi-nucleated Fragmented Mitotic Multi-lobular Pyknotic

Control 90.4 ± 8.4 0.8 ± 0.5 1.7 ± 0.3 3.9 ± 0.5 2.4 ± 0.9 0.8

COP4I251 28.7 ± 2.5 1.5 ± 0.2 65.2 ± 5.4 1.4 ± 1.1 2.1 ± 1.3 1.1

TPA 84.9 ± 3.3 3.4 ± 2.0 2.1 ± 1.5 4.6 ± 2.6 4.1 ± 1.2 0.9

COP41251 + TPA 27.4 ± 9.6 3.6 ± 0.3 54.4 ± 8.3 6.4 ± 4.4 2.2 ± 0.2 6 ± 5.7

nation of both drugs for various periods of times. Subsequently,

the cells were fixed and stained for the expression of various

neural- and glial-specific markers, and the nuclei were counter-

stained with DAPI. Fig. S shows a typical example of a culture

of OB-2 cells. Fig. 5, A and B, show that all of the cells

expressed the glial-specific marker OFAP. However, none of

the cells expressed the neural-specific markers NSE, NF16O, or

MAP2 before or after treatment with COP 4i25l, TPA, or the

two drugs combined (data not shown). Under control conditions

(Fig. 5, A and B), the nuclei were polymorphic and predomi-

nantly oval, with occasional mitoses. After exposure to COP

41251 at 0.8 p.M for 4 days (Fig. 5, C and D) the cells were

flatter and larger in size and acquired giant nuclei with extensive

fragmentation and apoptotic bodies (open arrow). The giant

nuclei in the COP 41251-treated cultures had the longest diam-

eter, which was 2.7-fold that of the untreated control cultures.

The cells that formed apoptotic bodies in this culture no longer

expressed OFAP (Fig. SC), nor did they express NF16O or

MAP2 when stained with the respective antibody (data not

shown). However, not all of the cells underwent these changes.

As shown in Fig. 5, C and D, some of the cells continued to

express OFAP and retained an oval nuclear morphology. In

contrast to the effects of COP 41251, exposure to TPA at 100ng/mi did not cause an increase in fragmented nuclei (Fig. SF).

Combined exposure to COP 4125 1 and TPA tended to resemble

the effects of COP 41251 alone (Fig. SH and data not shown).

To quantitate these observations, the morphologies of the nuclei

were evaluated on a large number of cells and assigned to the

following categories: oval, fragmented, multinucleated, mitotic,

pyknotic, and multilobular. As summarized in Table 4, when

compared to control untreated OB-2 cells. treatment with 0.8 p.M

CGP 41251 for 4 days caused the following changes. The

number of cells containing fragmented nuclei increased from 1.7

to 65%, and the number of mitotic nuclei decreased form 3.9 to

1.4%. On the other hand, following treatment with 100 ng/ml

TPA for 4 days, there was no increase in the number of frag-

mented nuclei (1.7% versus 2.1%), and there was a slight

increase in the number of mitotic cells, from 3.9 to 4.6%, and

multinucleated cells, from 0.8% to 3.4%. Treatment of OB-2

cells with 0.8 p.M COP 41251 plus 100 ng/ml TPA for 4 days

resulted in an increase in the number of fragmented nuclei from

1 .7 to 54% and an increase in the number of mitotic nuclei from

3.9 to 6.4%. Thus, TPA slightly inhibited the formation of

fragmented giant nuclei and protected the cells against the

inhibitory effects of COP 4 125 i with respect to mitosis.

The marked induction of nuclear fragmentation by COP

41251 and the inhibition of this process by TPA were also seen

with the GB-l and OS-i cells. Time course studies with the

GB-! cells showed that the induction of giant nuclei with

fragmentation and apoptotic bodies by 0.8 p.M COP 41251

occurred in about 50% of the cells within 24 to 48 h. Dose-

response studies with OB-2 cells showed that the concentration

of COP 41251 required to produce 50% of the maximum in-

duction of giant nuclei with fragmentation and apoptotic bodies

was about 0.4 p.M. Additional studies indicated that the induc-

tion of nuclear fragmentation by COP 41251 was markedly

inhibited when the cells were grown in 0.2% rather than 20%

FCS, or when the cells were treated with actinomycin D (2.5

p.g/ml), cycloheximide (10 p.g/ml), or aphidicolin (1 p.g/ml),

during the time of treatment with COP 41251 (data not shown).

The treated OB-2 cells were also examined by transmission

electron microscopy (Fig. 6). Various stages of apoptosis with

loss of cell-cell contact, nuclear condensation, and formation of

apoptotie bodies (Fig. 6B. cell b), as well as membrane bleb-

bing, vacuolization, and cellular fragmentation (Fig. 6B, cell a),

were observed. As commonly seen in cells undergoing apopto-

Fig. 5 Morphology of GB-2 cells after exposure to COP 4125! or TPA. The cells were cultured on LabTek chamber slides with COP 41251 or TPA

for 4 days and then fixed and stained with antibodies specific for OFAP (A, C, E, and G). The nuclei were stained with DAPI (B, D, F, and H). Thecontrol cultures (A and B) show OFAP-positive cells (A). Nuclei in the control culture (B) are predominantly oval and polymorphic with occasional

mitoses (arrow). The majority of cells treated with 0.8 mM COP 41251 for 4 days (C and D) form giant nuclei (*) with apoptotic bodies (open arrow).

Some of the COP 41251-treated cells are OFAP negative (C, *). Some OFAP-positive cells with oval nuclei are still present in the culture (C and

D, solid arrow). After treatment with TPA (100 ng/mi) for 4 days (E and F), most of the cells still display an oval nuclear morphology with occasionalmitoses (F, solid arrow). Cells treated with the combination of COP 41251 plus TPA (G and H) display oval and polymorphic nuclei, frequent mitoses(H, solid arrow), and multiple nuclei (H, open arrow). The combination led to a modest decrease of cells with giant nuclei and apoptotic bodies (H)

when compared to cells treated only with COP 41251 (D). H, bar, SO p.m.



: � L’� �

.�Ff ‘ �

,y� �

� . �:��#{231}

I

‘.� - � �

�r:� �.

5��

‘C, ��l/.

.)“

�

rc;.

/

.1.�..

5;

Fig. 6 Transmission electron microscopy of OB-2 cells after exposure to 0.8 mM COP 4125 1 for 4 days. The control culture (A) shows glioblastoma

cells with typical muitilobed nuclei. Exposure to COP 41 25 1 for 4 days (B) leads to apoptosis. Cell b is in the process of apoptosis during the formation

of apoptotic bodies (arrow); cell a is disintegrating and shows extensive vacuolization. A and B, bar, 2 p.m.

1026 Inhibition of the Growth of Olioblastomas by COP 41251

. . : . #{149}� ;� � � �

� � � � . :- � � ‘

� :� ��

. :.

)

/ , � ,

� ��

‘��

. �

sis, organelles like mitochondria, Oolgi apparatus, and endo-

piasmatic reticula were often found to be intact.

Viability and Nuclear Morphology of Nontransformed

Astrocyte Cultures. To investigate the effects of COP 41251

on nontransformed astrocytes, primary human astrocyte lines

were established and examined for viability and nuclear mor-

phology after exposure to COP 41251. One nontransformed

astrocyte line (A-l) was obtained from a noninfiltrated brain

sample of a patient undergoing resection of a glioblastoma

multiforme, and the other (A-2) was derived from a brain

sample of a patient with epilepsy. The nuclear morphology of

these cell lines was evaluated after a 5-day exposure to various

concentrations of COP 41251 (Fig. 7). Only about 1% of the

A-l cells displayed apoptotic bodies and giant nuclei in the

absence or presence of 0.8 p.M COP 41251. The A-2 cells also

showed about 1% apoptotic cells in the control culture and 3, 4,

and 6% giant nuclei with fragmentation and apoptotic bodies

after a 5-day exposure to COP 41251 at 0.2, 0.4, and 0.8 p.M,

respectively. In contrast, OB-2 cells treated in parallel with

similar concentrations of COP 41 25 1 showed an increase from

5.2% under control conditions to 21, 34, and 85% after exposure

to 0.2, 0.4, or 0.8 p.M COP 41251, respectively. The selective

increase in nuclear fragmentation in OB-2 cells was confirmed

after treatment for only 24 h with COP 41251 (data not shown).

Furthermore, the viability of glioblastoma cells was compared to

that of the nontransformed cells after treatment with COP

41251, using the MU assay. Cultures of GB-!, OB-2, OS-i,

A-I , and A-2 cells were treated with COP 4125 1 at 0.8 p.M and

compared to DMSO-treated control cultures. There was a sig-

nificant decrease in viability among the tumors cell lines, and

30, 50, and 60% in the GB- 1 , GB-2, and OS- 1 cells, respec-

tively. In contrast, the two astrocyte lines A-i and A-2 showed

a viability of greater than 90%. Thus, normal astrocyte cell lines

are only minimally affected by exposure to COP 41251 in

contrast to glioblastoma cell lines.

Antitumor Activity. The antiproliferative effect of COP

4125 1 was further investigated in xenografts. Two previously

described human glioblastoma cell lines, U87 MO and U373

MO, were used for the tumorigenicity studies since they have

been shown to grow rapidly in nude mice in contrast to the

above-described GB-i and OB-2 lines. In this study, serially

passaged tumor fragments of U87 MO as well as U373 MO cells

were grafted s.c. into athymic mice, and the mice were observed

for tumor formation for the following 18 or 25 days, respec-

tively. The mice were treated once daily with COP 41251, with

either 75 mg/kg body weight p.o. or 10 mg/kg body weight iv.,

beginning on day 5 after transplantation and continuing to day

17. As shown in Fig. 8A, the growth ofthe transplanted U87 MO

tumor was inhibited by about 50% when the athymic mice were

treated with COP 41251, either when administered p.o. or iv.

Orowth of the U373 MO grafts was completely inhibited when

the animals were treated with COP 4125 1 (Fig. 8B). Further-

more, in the group of mice carrying the U373 MO tumor, the

drug was removed at day 1 7, and the animals were observed for



DMSO

CGP41251 200nM

CGP41251 400nM

CGP41251 800nM


Fig. 7 Nuclear morphology of nontransformed

astrocyte lines, A-l and A2, and the GB-2 glio-

blastoma-derived cell line after exposure to COP

4 125 1 . The cultures were seeded on LabTekchamber slides and maintained in the presence of

the indicated concentrations of COP 41251 for S

days. They were then fixed, and the DNA was

stained with DAPI and analyzed as described in

“Materials and Methods” and Table 4. Bars, SD.

the subsequent 9 days. There was negligible regrowth of the

tumor during the latter period of time. Thus, COP 41 25 1 is also

a potent inhibitor of the tumorigenicity of human glioblastoma

cells in athymic mice.

DISCUSSION

Olioblastoma multiforme, also designated grade IV astro-

cytoma, is a common form of primary human brain tumors (I 8).

It is associated with a poor prognosis, even with the most

advanced current types of multiniodal therapy (19). Previous

studies have demonstrated the expression of several isoforms of

PKC in astrocytomas of various grades (20-24) and provided

evidence for a role of PKC in regulating the proliferation of both

astrocytes and astrocytoma cells in vitro (25, 26). This study

demonstrates that both a potent activator of PKC, TPA, as well

as a potent inhibitor of PKC, COP 41251, inhibit the in vitro

proliferation of human glioblastoma cells and certain other types

of human brain tumor-derived cell lines (summarized in Table

2). Furthermore, the inhibition obtained with TPA plus COP

4125 1 is greater than that obtained when each agent is used

separately, suggesting that they inhibit growth by different

mechanisms. Consistent with this conclusion is the finding that

0 10 20 30 40 50 60 70 80 90 100

% of nuclei displaying fragmentation

cultures treated with COP 41251 form giant nuclei, whereas

cultures treated with TPA contain multinucleated cells.

At the present time, it is not apparent why both TPA, an

activator of PKCs, and the PKC inhibitor COP 4125 1 cause

growth inhibition of the glioma cell lines (Fig. I ). This is

apparently not simply because TPA causes down-regulation of

PKCs, since TPA-mediated growth inhibition was seen at doses

of TPA (0. 1 to 1 ng/ml; Fig. 1A) that did not produce detectable

down-regulation of PKCs a, �, #{128},or � (Fig. 2). Alternatively,

stimulation of the activity of one or more specific isoforms of

PKC in these cells by TPA might lead to growth inhibition.

Thus, it has been observed that overexpression of specific iso-

forms of PKC either stimulate or inhibit growth in the presence

of TPA, depending on the cell type (4, 27). Furthermore, over-

expression of PKCS in CHO cells results in growth inhibition

(28), and when these derivatives were treated with TPA, the

cells arrested in O,-M and there was an accumulation of cells in

telophase (28). It is of interest that although TPA is a tumor

promoter in mouse skin, TPA inhibits the in vitro growth of not

only human glioblastoma cells, as indicated in the present study,

but also the growth of a variety of human melanoma (29), breast

(30), and colon (27) cancer cell lines. In the latter cases, the



I -.-- Placebo treated controls� -ern- 75 mg/kg p.o. once dailyL-�--� 10 mg/kg i.v. once daily

- I I I I

5 8 11 14 18

days after tumor transplantation

-.- Placebo treated controls-0- 75 mg/kg p.o. once daily-k- 10 mg/kg iv. once daily

1028 Inhibition of the Growth of Olioblastomas by COP 41251

4 M. Begemann, unpublished results.

A 2.5’

2.0’

C’)

E’’C)

a)N(I)

0

E� 1.0’

B

0.5�

0.0

1.00

0.75

C’)

EC)

a)

.� 0.50

0

E

0.25

5 10 15 20

days after tumor transplantation

Fig. 8 Antitumor activity of COP 41251. Antitumor activity was

tested using xenografts of the human glioblastoma multiforme cells U87

MG (A) and U373 MO (B), as described in “Materials and Methods.”

The compound was administered from day 5 to day 17 following s.c.transplantation of the tumor cells, as indicated. In both in vivo tumorgrowth experiments, admmistration of COP 41251 significantly inhib-

ited tumor growth (P < 0.001). In both cases, the placebo and treated

curves were significantly different, and the two modes of administration

of COP 41251 were not significantly different (A, P = 0.9; B, P - 0.6).Bars, SD.

precise mechanism of inhibition is also not known. Other acti-

vators of PKC that lack tumor-promoting activity, for example

the bryostatins (30-33), might therefore, be useful agents in

cancer therapy.

The present studies establish the fact that the staurosporine

derivative COP 4125 1 is a potent inhibitor of the in vitro growth

of nine early-passage cell lines derived from the following

human tumors: five glioblastoma multiforme, one mixed glio-

blastoma and gliosarcoma, two astrocytomas, and one choroid

plexus carcinoma. Using growth-curve studies, the IC�0 was in

the range of 0.4 p.M, although the OB-3 gliobiastoma cell line

was relatively resistant, since inhibition of growth required 0.8

to 1 p.M COP 4125 1 (Table 2). However, in clonogenicity

studies, this line was highly sensitive (Fig. 4). More detailed

studies suggest that CGP 4 125 1 inhibited the proliferation of

these cell lines through both cytostatic and cytocidal mecha-

nisms. However, although COP 41251 has a high selectivity for

inhibition of PKC (7), our studies do not exclude the possibility

that its growth-inhibitory effects in the above-described cell

types is due, at least in part, to inhibition of other protein

kinases. Of particular interest might be inhibitory effects on the

functions of the platelet-derived growth factor and epidermal

growth factor receptors (8), since these receptors have been

shown to be frequently activated in glioblastomas through mu-

tations and amplifications of the corresponding genes (34-37).

We hypothesize that COP 4125 1 inhibits the growth of

glioblastoma cells, at least in part, by arresting cells at the O2-M

transition point, and to a variable extent during � . This

hypothesis is consistent with the flow cytometry data (Table 3),

the inhibition of BrdUrd labeling (Fig. 3), the markedly reduced

number of mitotic figures, and the increased number of giant

nuclei with fragmentation and apoptotic bodies seen in the

treated cultures (Fig. 5). Studies in yeast have shown that PKC

is required for the 02-M transition (38). Furthermore, yeast cells

deficient in the cdc2/cyclin B complex are unable to transit from

the 02 to the M phase and form giant nuclei (39). In recent

studies, we have found that glioblastoma cells treated with COP

41251 exhibit a decrease in cdc2/cyclin B histone Hi kinase

activity.4 Other inhibitors of PKC have been found to arrest cells

in 02 or result in inhibition of cdc2 kinase activity (40-45).

High concentrations of staurosporine lead to the arrest of trans-

formed cells in 02 but arrest nontransformed cells in both G�

and 0, (� 45). However, further studies are required to deter-

mine the precise effects of COP 41251 on cell cycle progression

and cyclin-CDK activity.

A striking finding in the present study is that COP 412�1

induces nuclear fragmentation and other morphological changes

26 characteristic of apoptosis (Fig. 5). A common feature of apop-tosis is internucleosomal DNA fragmentation, which was inves-

tigated by preparation of high molecular weight DNA from

treated cultures as well as through various in situ procedures.

However, DNA fragmentation was not observed after treatment

with COP 41251 in these gliobiastoma cells (data not shown).

Nevertheless, apoptotic cell death has been shown to occur

without DNA fragmentation (46-48). The absence of DNA




fragmentation might reflect the absence of endonucleases in

these particular cell types or, alternatively, different pathways

leading to cell death. The COP 41251-induced nuclear fragmen-

tation with apoptotic bodies was partially inhibited by TPA (Fig.

5), which is consistent with the ability of certain phorbol esters

to inhibit apoptosis in thymocytes (49). Serum deprivation,

actinomycin D, and cycloheximide markedly inhibited COP

4125 1-induced nuclear fragmentation, suggesting that de nos’o

RNA and protein synthesis and cell cycle progression are re-

quired for COP 41 25 1 to induce apoptosis, which presumably

occurs in 02. Another staurosporine derivative, K252a, has been

shown to uncouple cell cycle control in rat diploid fibroblasts,

leading to tetraploidization, which was inhibited also by serum

deprivation, cycloheximide, and actinomycin D (43). Other in-

vestigators (50) have recently reported that tamoxifen, which

also inhibits PKC, also induces apoptosis in human malignant

gliomas, but they did not examine cell cycle-related effects.

Nuclear fragmentation and micronucleation, as well as 02 ar-

rest, were described in leukemic lymphocytes after exposure to

staurosporine, the parental compound, from which COP 41251

was derived (44). Immunohistochemical studies (data not

shown) provide evidence that OB-l, OB-2, and OS-! cell lines

used in the present study express a mutant p53 protein. This

suggests that COP 4125 1 induces apoptosis by a p53-independ-

ent pathway and, therefore, may be useful against a variety of

human tumors. Furthermore, cultures of normal astrocytes were

relatively resistant to the toxic effects of COP 41251, including

the induction of nuclear fragmentation (Fig. 7).

Finally, COP 4125 1 when given either by the oral route or

i.v. markedly inhibited the tumorigenicity of two previously

characterized human glioblastoma cell lines, U87 MO and U373

MO, following s.c. transplantation into athymic mice. This

finding extends previous studies indicating that COP 41251

inhibits the growth of other types of human cancer cells in

athymic mice (7). Thus, COP 41251 might be a useful agent

when administered either alone or in combination with other

agents for the therapy of human brain tumors.

ACKNOWLEDGMENTS

We are deeply grateful to Philippe Rousselot and Christine Neyt fortechnical advice, as well as valuable comments on the manuscript. We

thank Jingmei Xu for assistance with the statistical analyses.

REFERENCES

1 . Nishizuka, Y. Studies and perspectives of protein kinase C. Science

(Washington DC), 233: 305-312, 1986.

2. Nishizuka, Y. The molecular heterogeneity of protein kinase C and

its implications for cellular regulation. Nature (Lond.), 334: 661-665,1988.

3. Weinstein, I. B., Borner, C. M., Krauss, R. S., O’Dnscoll, K., Choi,

P. M., Moritomi, M., Hoshina, S., Hsieh, L-L., Tchou-Wong, K-M.,Nichols-Guadagno, S., Ueffing, M., and Ouiliem, J. Pleiotropic effectsof protein kinase C and the concept of carcinogenesis as a progressivedisorder in signal transduction. In: J. Brugge, T. Curran, E. Harlow, andF. McCormick (eds.), Origins of Human Cancer: A Comprehensive

Review, pp. 1 13-124. NY: Cold Spring Harbor Laboratory, 1991.

4. Housey, 0. M., Johnson, M. D., Hsiao, W. W. L., O’Brian, C. A.,

Murphy, J. P., Kirschmeier, P., and Weinstein, I. B. Overproduction ofprotein kinase C causes disordered growth control in rat fibroblasts.

Cell, 52: 343-354, 1988.

S. Borner, C., Filipuzzi, I., Weinstein, I. B., and Imber, R. Failure of

wild-type or a mutant form of protein kinase C-a to transform fibro-blasts. Nature (Lond.), 353: 78-80, 1991.

6. Cacace, A. M., Guadagno, S. N., Krauss, R. S., Fabbro, D., and

Weinstein, I. B. The e isoform of protein kinase C is an oncogene when

overexpressed in rat fibroblasts. Oncogene, 8: 2095-2104, 1993.

7. Meyer, T., Regenass, U., Fabbro, D., Aiteri, E., Roesel, J., Mueller,M., Caravatti, 0., and Matter, A. A derivative of staurosporine (COP41,251) shows selectivity for protein kinase C inhibition and in vitro

anti-proliferative as well as in vivo anti-tumor activity. mt. J. Cancer,43: 851-856, 1989.

8. Andrejauskas-Buchdunger, E., and Regenass, U. Differential inhibi-

tion of the epidermal growth factor-, platelet-derived growth factor-, and

protein kinase C-mediated signal transduction pathways by the stauro-

sporine derivative COP 41251. Cancer Res., 52: 5353-5358, 1992.

9. Marte, B. M., Meyer, T., Stabel, S., Standke, 0. J. R., Jaken, S.,Fabbro, D., and Hynes, N. E. Protein kinase C and mammary celldifferentiation: involvement of protein kinase Ca in the induction of

�3-casein expression. Cell Growth & Differ., 5: 239-247, 1994.

10. Ausubel, F. A., Brent, R., Kingston, R. E., Moore, D. D., Seidman,J. 0., Smith, J. A., and Struhl, K. (eds.), Current Protocols in MolecularBiology. New York: Greene & Wiley-Interscience, 1990.

11 . Mosmann, T. Rapid colonmetric assay for cellular growth and

survival: application to proliferation and cytotoxicity assays. J. Immu-

nol. Methods, 65: 55-63, 1983.

12. Ponten, J., and Macintyre, E. H. Long-term culture of normal andneoplastic glia. Acta Pathol. Microbiol. Scand., 74: 465-486, 1968.

13. Beckman, 0., Beckman, L., Ponten, J., and Westermark, B. 0-6-PDand PGM phenotypes of 16 continuous human tumor cell lines. Hum.Hered., 21: 238-241, 1971.

14. Buchdunger, E., Trinks, U., Mett, H., Regenass, U., Mueller, M.,

Meyer, T., McGlynn, E., Pinn, L. A., Traxler, P., and Lydon, N. B.4,S-Dianilinophthalimide: a protein-tyrosine kinase inhibitor with selec-

tivity for the epidermal growth factor receptor signal transduction path-

way and potent in vivo antitumor activity. Proc. Natl. Acad. Sci. USA,91: 2334-2338, 1994.

IS. Evans, B. D., Smith, I. E., Shorthouse, A. J., and Miliar, J. L. A

comparison of the response of human lung carcinoma xenografts to

vindesine and vincristine. Br. J. Cancer, 45: 466-468, 1982.

16. Inoue, M., Kishimoto, A., Takai, Y., and Nishizuka, Y. Studies ona cyclic nucleotide-independent protein kinase and its proenzyme inmammalian tissues. J. Biol. Chem., 252: 7610-7616, 1977.

17. Sacktor, T. C., Osten, P., Valsamis, H., Jiang, X., Naik, M. U., and

Sublette, E. Persistent activation of the � isoform of protein kinase C inthe maintenance of long-term potentiation. Proc. NatI. Acad. Sci. USA,90: 8342-8346, 1993.

18. Ringertz, N. Grading of gliomas. Acta Pathol. Microbiol. Scand.,

27: 51-64, 1950.

19. Black, P. M., Schoene, W. C., and Lampson L. (eds.), Astrocyto-mas: Diagnosis, Treatment, and Biology. Boston: Blackwell Scientific

Publications, 1993.

20. Benzil, D. L., Finkeistein, S. D., Epstein, M. H., and Finch, P. W.

Expression pattern of a-protein kinase C in human astrocytomas mdi-

cates a role in malignant progression. Cancer Res., 52: 2951-2956,

1992.

21. Baltuch, 0. H., Couldwell, W. T., Villemure, J-G., and Yong, V. W.

Protein kinase C inhibitors suppress cell growth in established and low

passage glioma cell lines: a comparison between staurosporine andtamoxifen. Neurosurgery, 33: 495-501, 1993.

22. Couldwell, W., Antel, J. P., and Yong, V. W. Protein kinase Cactivity correlates with the growth rate of malignant gliomas. Part II.

Effects of glioma mitogens and modulators of protein kinase C. Neu-rosurgery, 31: 717-724, 1992.

23. Misra-Press, A., Fields, A. P., Samols, D., and Ooldthwait, D. A.

Protein kinase C isoforms in human glioblastoma cells. Glia, 6: 188-197, 1992.



1030 Inhibition of the Growth of Olioblastomas by COP 412�1

24. Ahmad, S., Mineta, T., Martuza, R. L., and Olazer, R. I. Antisenseexpression of protein kinase Ca inhibits the growth and tumorigenicityof human glioblastoma cells. Neurosurgery, 35: 904-909, 1994.

25. Couidwell, W. T., Uhm, J. H., Antel, J. P., and Yong, V. W.Enhanced protein kinase C activity correlates with the growth rate ofmalignant gliomas in vitro. Neurosurgery, 29: 880-887, 1991.

26. Bhat, N. R. Role of protein kinase C in glial cell proliferation. J.Neurosci. Res., 22: 20-27, 1989.

27. Choi, P. M., Tchou-Wong, K. M., and Weinstein, I. B. Overexpres-sion of protein kinase C in HT29 colon cancer cells causes growthinhibition and tumor suppression. Mol. Cell. Biol., 10: 4650-4657,1990.

28. Watanabe, T., Ono, Y., Taniyama, Y., Hazama, K., Igarashi, K.,Ogita, K., Kikkawa, U., and Nishizuka, Y. Cell division arrest inducedby phorbol ester in CHO cells overexpressing protein kinase C-b sub-species. Proc. Nati. Acad. Sci. USA, 89: 10159-10163, 1992.

29. Arita, Y., O’Driscoil, K. R., and Weinstein, I. B. Orowth inhibitionof human melanoma-derived cells by l2-O-tetradecanoyl phorbol 13-acetate. mt. J. Cancer, 56: 229-235, 1994.

30. Kennedy, M. J., Prestigiacomo, L. J., Tyler, 0., May, W. S., andDavidson, N. E. Differential effects of bryostatin 1 and phorbol ester onhuman breast cancer cell lines. Cancer Res., 52: 1278-1283, 1992.

31. Hornung, R. L., Pearson, J. W., Beckwith, M., and Longo, D. L.Preclinical evaluation of bryostatin as an anticancer agent against sev-eral murine tumor cell lines: in vitro versus in vivo activity. Cancer Res.,

52: 101-107, 1992.

32. Schuchter, L. M., Es, A. H., May, W. S., Laulis, M. K., Pettit, 0. R.,

and Hess, A. D. Successful treatment of murine melanoma with bryo-statin 1. Cancer Res., 51: 682-687, 1991.

33. Steube, K. 0., and Drexler, H. 0. Differentiation and growthmodulation of myeloid leukemia cells by the protein kinase C activating

agent bryostatin-l . Leukemia Lymphoma, 9: 141-148, 1993.

34. Muleris, M., Almeida, A., Dutrillaux, A. M., Pruchon, E., Vega, F.,

Delattre, J. Y., Poisson, M., Malfoy, B., and Dutrillaux, B. Oncogene

amplification in human gliomas: a molecular cytogenetic analysis. On-

cogene, 9: 2717-2722, 1994.

35. Nister, M., Libermann, T. A., Betsholtz, C., Peuersson, M., Claes-

son-Welsh, L., Heidin, C-H., Schessinger, J., and Westermark, B. Ex-pression of mRNAs for platelet-derived growth factor and transforminggrowth factor-a and their receptors in human malignant glioma celllines. Cancer Res., 48: 3910-3918, 1988.

36. Ekstrand, A. J., James, C. D., Cavanee, W. K., Seliger, B., Petters-son, R. F., and Collins, V. P. Genes for epidermal growth factorreceptor, transforming growth factor a, and epidermal growth factor and

their expression in human gliomas in vivo. Cancer Res., 51: 2164-2172,

1991.

37. Chaffanet, M., Chauvin, C., Lame, M., Berger, F., Chedin, M.,

Rost, N., Nissou, M-F., and Benabid, A. L. EGF receptor amplificationand expression in human brain tumors. Eur. J. Cancer, 28: 1 1-17, 1992.

38. Levin, D. E., Fields, F. 0., Kunisawa, R., Bishop, J. M., and

Thorner, J. A candidate protein kinase C gene, PKCJ, is required for the

S. cerevisiae cell cycle. Cell, 62: 213-224, 1990.

39. Hayles, J., Fisher, D., Woollard, A., and Nurse, P. Temporal orderof S phase and mitosis in fission yeast is determined by the state of thep34cdc2-mitotic B cyclin complex. Cell, 78: 813-822, 1994.

40. Oadbois, D. M., Hamaguchi, J. R., Swank, R. A., and Bradbury,E. M. Staurosporine is a potent inhibitor of p34cdc2 and p34cdc2-iikekinases. Biochem. Biophys. Res. Commun.. 184: 80-85, 1992.

41. Th’ng, J. P. H., Ouo, X-W., Swank, R. A., Crissman, H. A., andBradbury, E. M. Inhibition of histone phosphorylation by staurosporineleads to chromosome decondensation. J. Biol. Chem., 269: 9568-9573,

1994.

42. Hofmann, J., O’Connor, P. M., Jackman, J., Schubert, C., Ueberail,

F., Kohn, K. W., and Orunicke, H. The protein kinase C inhibitorilmofosine (BM 41440) arrests cells in 02 phase and suppresses cdc2kinase activation through a mechanisms different from that of DNAdamaging agents. Biochem. Biophys. Res. Commun., 199: 937-943,

1994.

43. Usui, T., Yoshida, M., Abe, K., Osada, H., Isono, K., and Beppu, T.

Uncoupled cell cycle without mitosis induced by a protein kinaseinhibitor, K-252a. J. Cell Biol., 115: 1275-1282, 1991.

44. Bruno, S., Ardelt, B., Skierski, J. S., Traganos, F., and Darzynk-

iewicz, Z. Different effects of staurosporine, an inhibitor of proteinkinases, on the cell cycle and chromatin structure of normal and leuke-

mic lymphocytes. Cancer Res., 51: 470-473, 1992.

45. Crissman, H. A., Oadbois, D. M., Tobey, R. A., and Bradbury,E. M. Transformed mammalian cells are deficient in kinase-mediatedcontrol ofprogression through the O� phase ofthe cell cycle. Proc. NatI.Acad. Sci. USA, 88: 7580-7584, 1991.

46. McConkey, D. J., Hartzell, P., Jondal, M., and Orrenius, S. Inhibi-tion of DNA fragmentation in thymocytes and isolated thymocyte nuclei

by agents that stimulate protein kinase C. J. Biol. Chem., 264, 13399-

13402, 1989.

47. 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 cellsurvival in the oligodendrocyte lineage. Cell, 70: 31-46, 1992.

48. Cohen, 0. M., Sun, X-M., Snowden, R. T., Dinsdale, D., andSkilieter, D. N. Key morphological features of apoptosis may occur in

the absence of internucleosomal DNA fragmentation. Biochem. J., 286:

331-334, 1992.

49. Ucker, D. S., Obermiller, P. S., Eckhart, W., Apgar, J. R., Berger,N. A., and Meyers, J. Oenome digestion is a dispensable consequence of

physiological cell death mediated by CTLs. Mol. Cell. Biol., 12: 3060-

3069, 1992.

50. Couldwel!, W. T., Hinton, D. R., He, S., Chen, T. C., Sebat, I.,

Weiss, M. H., and Law, R. E. Protein kinase C inhibitors induceapoptosis in human malignant glioma cell lines. FEBS Lett., 345:

43-46, 1994.



1996;2:1017-1030. Clin Cancer Res M Begemann, S A Kashimawo, Y A Choi, et al. promoter.inhibitor of protein kinase C, and by a phorbol ester tumor Inhibition of the growth of glioblastomas by CGP 41251, an

Updated version

http://clincancerres.aacrjournals.org/content/2/6/1017

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/2/6/1017To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



inhibition of the growth of glioblastomas by cgp 41251, an ... · vol. 2, 1017-1030, june 1996...

Documents