bcl-2, bax, bcl-xL, and bcl-x Expression in Normal and ... · Table 1 Distribution of bcl-2, bax, bcl-xL and bcl-x5 RNA levels in normal and neoplastic ovarian tissues The values

Vol. 4, 517-524, February 1998 Clinical Cancer Research 517

3 The abbreviations used are: PCD, programmable cell death; RT, re-

verse transcriptase; ER, estrogen receptor; PR, progesterone receptor.

bcl-2, bax, bcl-xL, and bcl-x� Expression in Normal and Neoplastic

Ovarian Tissues’

Maria Marone, Giovanni Scambia,

Simona Mozzetti, Gabriella Ferrandina,

Sonia lacovella, Anna De Pasqua,

Pierluigi Benedetti-Panici, and

Salvatore Mancuso2

Laboratory of Antineoplastic Pharmacology, Department of Obstetrics

and Gynecology. Catholic University, 00168 Rome, Italy

ABSTRACT

The bcl-2 family of proteins includes some important

regulators of apoptosis. Among these, bcl-2 and bcl-xL pre-

vent cells from entering apoptosis, whereas bax and bcl-x�can induce cell death. Alterations in the control of this

process can lead to a decrease in cell death, thus contribut-

ing to neoplastic growth. Diminished susceptibility to chem-otherapy has also been attributed, in in vitro systems, toalterations in the levels of bcl-2, bax, or bcl-x. We analyzedthe expression of bcl-2, bax, bcl-xL, and bcl-x� in normal andneoplastic ovarian tissues by reverse transcriptase-PCR andWestern blotting. The RNA and protein levels were signifi-

cantly correlated for all genes. Interestingly, the levels ofthese genes in normal and neoplastic tissues were signifi-

cantly different: bcl-2 was higher in normal tissue (P <

0.002), whereas bax and bcl-xL were higher in carcinoma

(P < 0.018 and P < 0.030, respectively). bcl-x� was present

at low levels in 83% of neoplastic samples and was unde-

tectable in normal tissue. Reverse transcriptase-PCR anal-

ysis of 74 tumors showed no major correlation with chico-pathological parameters or with response to chemotherapy.

Only bax and bcl-xL were correlated with progesterone re-ceptor levels (n = 29, r = +0.44, P < 0.0189, and r = -0.40,

P < 0.035, respectively). No correlation was found withestrogen receptor levels or with p53 immunostaining. Ourdata indicate that the regulation of the bcl-2 family of pro-

teins differs between normal and neoplastic ovarian tissues.Moreover, the modulation of these genes in ovarian carci-

noma is different compared to other tissues; therefore, tissue

specificity is very important in regulation of the bcl-2 family

of proteins.

Received 7/25/97; revised 10/20/97; accepted 1 1/7/97.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked

advertisement in accordance with 18 U.S.C. Section 1734 solely to

indicate this fact.

1 This work was partially supported by Associazione Italiana per la

Ricerca sul Cancro.

2 To whom requests for reprints should be addressed, at Department ofObstetrics and Gynecology, Catholic University, L.go A. Gemelli 8,00168 Rome, Italy. Phone: 39-6-35508736; Fax: 39-6-35508736; E-mail: [email protected].

INTRODUCTION

Cell turnover in normal tissues is regulated by the balance

between the rates of cell proliferation and cell death (1). Con-

sequently, uncontrolled neoplastic growth can be caused not

only by increased proliferation but also by a diminished rate of

cell death, which can result from the failure of cells to undergo

apoptosis or from PCD3 in response to physiological stimuli (1,

2). In addition to this role in the onset or development of cancer,

the modulation of apoptosis can also influence the outcome of

cancer treatment because drug resistance can be attributed partly

to a decreased cellular susceptibility to PCD (3). bcl-2 was the

first characterized gene that was clearly involved in the regula-

tion of PCD by inhibiting apoptosis, thus promoting cell sur-

vival (4). It was shown that transfection of bcl-2 into immature

pre-B cells allowed prolonged survival in the absence of sur-

vival factors (5) and that high levels of bcl-2 protect cells from

apoptosis induced by y irradiation and by a variety of chemo-

therapeutic agents (6).

bcl-2 belongs to a still growing family (7, 8), the members

of which are able to form homo- and/or heterochimers among

themselves; their association and the relative ratio between pro-

and antiapoptotic proteins are responsible for directing the cells

toward death or survival (9, 10). Overexpression of bax was

shown to accelerate PCD by inhibiting the death repressor

activity of bcl-2, probably by forming bcl-2-bax complexes or

by competing with other bcl-2 targets (1 1, 12). The bcl-x gene

mainly gives rise, by alternative splicing, to two mRNA species,

yielding two protein products: bcl-x�, endowed with death re-

pressor activity, and a shorter variant, bcl-x�, which functions as

a dominant inhibitor of bcl-2, thus inducing apoptosis (13).

Although bcl-2 has been described as a molecule involved in a

variety of biochemical pathways (14, 15), the mechanisms by

which the interactions between the various members of the bcl-2

family finally lead to apoptosis are still unknown.

Expression of bcl-2 has been measured in a variety of

human neoplastic tissues, including melanoma (16), non-small

cell lung (17), prostate (18), and breast (19, 21). Several studies,

mostly on breast cancer, reported a direct correlation between

bcl-2 and the steroid hormone receptor status (21, 22), and a

correlation between bcl-2 levels and survival has been reported

but is still controversial (23, 24). Moreover, low bax expression

in breast cancer was shown to be associated with poor response

to chemotherapy and shorter survival (25). Expression of bcl-2

alone has been studied in ovarian cancer as well, suggesting

that, as in breast, high bcl-2 levels may be associated with

longer survival (26, 27).

Here, we determined the protein and RNA expression

levels of bcl-2, bax, bcl-xL, and bcl-x� in a series of normal and

Research. on September 20, 2020. © 1998 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

Table 1 Distribution of bcl-2, bax, bcl-xL and bcl-x5 RNA levels in normal and neoplastic ovarian tissues

The values are the ratio between the relative adsorbance of the RNA of interest and aldolase-A RNA obtained by RT-PCR.

n

bcl-2 bax

Mean Median SD RangeMean Median SD Range

Normal 6 6.36 6.91 2.73 4.07-10.0” 0.90 0.88 0.23 0.59�1,01”

All carcinomas 70 2.23 2.04 1.61 0.01-8.69 2.17 1.75 2.10 0.40-13.00Primary cancers 44 2.18 2.14 1.34 0.10-7.20 2.06 1.81 2.46 0.40-13.00Metastases 7 2.23 2.36 0.91 1.01-3.72 2.37 1.87 1.51 0.83-5.41

Recurrences 19 2.61 1.79 1.59 0.38-8.69 1.73 1.61 0.70 0.45-3.02

n

665

42

6

17

bcl-xL bcl-x�ND cases

(n)

6

11

8

1

2

Range Mean

0. 13�0.20c ND”

0.01-0.76 0.13

0.01-0.71 0.13

0.24-0.47

0.11-0.76 0.16

Median

ND

0.14

0.14

0.08

0.15

SD

ND

0.08

0.08

0.07

0.06

Range

ND

0.02 -0.30

0.02 -0.30

0.07 -0.23

0.04 -0.26

Mean

0. 160.31

0.33

0.32

0.31

Normal SD

0. 15 0.070.29 0.19

0.28 0.18

0.30 0.08

0.28 0.19

518 bcl-2 Family in Normal and Neoplastic Ovary

Normal

All carcinomasPrimary cancers

Metastases

Recurrences

a p < 0.002.bp < 0.018.

C p < 0.030.d ND, not detectable.

cancer ovarian samples and analyzed their correlation with some

clinicopathological and other biological parameters.

PATWNTS AND METHODS

Patients and Tumor Specimens. The patients involved

in this study underwent surgery between 1994 and 1996 at the

Department of Obstetrics and Gynecology of the Catholic Urn-

versity (Rome, Italy). Necrotic, hemorragic and fat tissues were

trimmed out from the tumor samples, which were then cut in

half so that the same histology and grade represented in the

samples was used for histopathological analysis and RNA and

protein extraction. One half was processed for histopathological

examination to determine the percentage of tumor cells in the

samples, which was never lower than 80%. The other half was

used for RNA preparation and, if there was a sufficient amount,

for protein extraction. Tissues were frozen in liquid nitrogen

immediately after surgical resection and stored at -80#{176}Cuntil

processed. Total RNA was extracted from six normal ovarian

tissue biopsies from patients with uterine fibromatosis and from

70 neoplastic specimens (Table 1). Protein lysates were pre-

pared from 1 1 normal ovarian tissue biopsies and from a subset

of 20 neoplastic tissue samples taken from the same population

used for RNA extraction. Due to the limited size of the bioptic

specimens, in most cases, the material was sufficient for RNA

extraction only. Normal tissues came from 1 1 patients with

uterine fibromatosis, who ranged in age from 38 to 52 years old

(mean, 48 years; median, 47 years), 7 of whom were premeno-

pausal and 4 who were postmenopausal. The neoplastic tissues

included mostly primary ovarian cancers and also some meta-

static and recurrent tumors (Table 1). Histological classification

of tumors was carried out according to the WHO system (28)

and tumors were graded as well (Gl), moderately (G2) or poorly

(G3) differentiated. Stage of disease was established according

to the FIGO staging system. All patients received cisplatin-

containing chemotherapy, which was instituted 2-3 weeks after

surgery. Specimens of recurrent disease were obtained from

pelvic or peritoneal tumor masses occurring after a period of

clinical and pathological negative follow-up after intensive sur-

gery and chemotherapy. Recurrent and metastatic tissues shared

the same histological type and grade as their respective primary

tumors. The clinicopathological features of the patients bearing

primary ovarian tumors are described in Tables 2 and 3.

Semiquantitative RT-PCR Analysis. Total cellular

RNA was obtained from tissue biopsies by the method of

Chomczynsky and Sacchi (29). Five p.g of RNA per sample

were separated on 1 % formaldehyde-agarose gels to assess their

integrity. The Perkin-Elmer Gene Amp RNA PCR kit was used

for all the RT-PCRs, which were performed in the Gene Amp

PCR system 9600 (Perkin-Elmer/Cetus, San Diego, CA). After

removal of contaminating chromosomal DNA with DNase I

treatment, 1 p�g of RNA was reverse-transcribed with 25 units of

Moloney murine leukemia virus RT at 42#{176}Cfor 30 mm. Two p.1

of cDNA products were used in each PCR. The PCR were

carried out using different cycling parameters for each set of

primers. The number of cycles and the reaction conditions were

chosen so that none of the target cDNAs reached a plateau and

that the two pairs of primers did not compete with each other.

The sequences of the specific primers for bcl-2 and bax were

described previously (30), and the primers corresponding to the

bcl-x cDNA were chosen so that the two specific products,

bcl-xL and bcl-x�, migrated as distinct bands on a 2% agarose

gel (36). The target cDNAs were coamplified with aldolase-A as

an internal control using the following pair of primers: 5’-

CGCAGAAGGGGTCCTGGTGA-3’ and 5’-CAGCTCCTFCT-

TCTGCTCCGGGGT-3’, which yielded a band of 176 bp. All

oligonucleotide primers were synthesized by Pharmacia Biotech

(Uppsala, Sweden).

bcl-2 was amplified with 1 unit of AmpliTaq DNA polym-

erase in 1 mM MgC12, and a first cycle of 2 mm at 95#{176}C,45 s at

55#{176}C,and 1 mm at 72#{176}Cwas followed by 32 cycles of 45 s at

95#{176}C,45 5 at 55#{176}C,and 1 mm at 72#{176}C.bax was amplified using

1 unit of AmpliTaq DNA polymerase in 5 mr�i MgC12 as



Clinical Cancer Research 519

Table 2 Distribution of bcl-2 and bax RNA levels in primary ovarian carcinomas according to clinico-pathological features

The values are the ratios between the relative adsorbance of the RNA of interest and aldolase-A RNA obtained by RT-PCR.

Clinicopathological

feature n

bcl-2 bax

Mean Median Range SD Mean Median Range SD

Histology

Serous 30 1.98 1.91 0.01-3.55 0.87 2.55 1.54 0.40-13.00 2.78Mucinous 1 2.04

Endometrioid 6 3.78 2.23 0.54-7.20 2.51 1.98 2.15 0.73-2.77 0.75

Undifferentiated 4 2.39 2.48 2.02-7.13 2.82 2.52 2.84 1.46-3.26 0.94

Other 3 2.87 1.73 1.60-3.85 1.26 5.17 5.74 3.99-5.80 1.02

Stage

I-Il 6 2.11 2.20 0.54-3.33 1.78 4.29 3.38” 1.55-9.44 3.02

HI 30 2.28 2.04 0.01-7.13 1.22 2.45 1.60 0.40-13.00 2.47

IV 8 2.45 2.21 0.54-7.20 2.01 1.54 1.50 0.64-2.35 0.65Grading

1-2 9 2.36 2.48 0.54-3.33 0.52 2.65 2.15 1.55-9.44 1.02

3 29 2.21 2.04 0.01-7.20 1.43 2.48 1.64 0.50-13.00 2.78

Ascites

No 24 2.01 2.03 0.54-3.33 0.81 2.69 1.95 0.64-9.44 2.13

Yes 20 2.69 2.19 0.01-7.20 1.78 2.48 1.61 0.50-13.00 2.84

Residual tumor”�2 cm 23 2.35 2.11 0.54-7.13 1.27 1.99 1.60 0.40-6.15 1.33>2 cm 15 2.26 2.02 0.01-7.20 1.67 2.63 1.61 0.50-13.00 3.14

Response to

chemotherapyc

CR 14 2.55 2.19 1.30-7.13 1.52 1.97 1.54 0.40-6.15 1.47PR 12 1.79 1.58 0.54-3.13 0.83 2.07 2.04 0.85-3.96 0.98NC-P 12 2.60 2.24 0.01-7.20 1.79 2.45 1.22 0.50-13.00 3.50

Age (yr)

�60 29 2.26 2.11 0.54-7.13 1.31 2.39 2.04 0.64-6.26 1.52

>60 15 1.84 1.87 0.01-2.97 0.77 3.69 1.60 0.50-13.00 4.07

P < 0.009.a Stage I-Il vs. stage III-IV,

b Only stage III-IV patients.C CR, complete response; PR, partial response; NC-P. no change-progression.

follows: 2 ruin at 95#{176}C,45 s at 65#{176}C,and 1 mm at 72#{176}Cfor the

first cycle and then 45 s at 95#{176}C,45s at 65#{176}C,and 1 mm at 72#{176}C

for 26 cycles. One unit of AmpliTaq Gold in 5 mivi MgC12 was

used for bcl-x using the same cycling parameters applied for

bax, except that the first denaturation step was extended to 10

mm, which was then followed by 28 cycles, as described above.

The PCR products were loaded onto 2% agarose gels and

stained with ethidium bromide. For bcl-2 and bcl-x, images of

the gels were acquired with a Cohu charged coupled device

camera, and quantification was performed with Phoretix ID

(Phoretix International Ltd., Newcastle upon Tyne, United

Kingdom). bax was undetectable by ethidium bromide staining

in most neoplastic samples; hence, the PCR products were

separated on 1 .2% agarose gels, blotted onto nylon membranes

(Duralon; Stratagene, La Jolla, CA), UV-cross-linked and hy-

bndized at 42#{176}Cfor 18 h in 50% formamide, 5 X SSC, 1 X

Denhardt’s solution, 0.2% SDS, and 200 p.g/ml denatured

salmon sperm DNA. About 50,000 cpm/ml of 32P PCR-labelcd

bax and aldolase-A probes were added to the reaction. After

washing, the blots were exposed to an Instant Imager Electronic

Autoradiography Instrument, and band intensity was measured

using the Imager software provided with the instrument (Pack-

ard Instrument Company, Meridian, CT). Each set of reactions

and each blot included the ovarian cancer cell line A2780 as a

positive control and a no-sample negative control. The ratio

between the sample RNA to be determined and aldolase-A was

calculated to normalize for initial variations in sample concen-

tration and as a control for reaction efficiency. All samples were

normalized to their control. Reproducibility of the method was

established by analyzing the level of variability in the control

sample A2780 in different experiments and different gels, which

was never greater than 20%.

Preparation of the Tissue Lysates and Western BlottingAnalysis. Frozen tissues were homogenized in S volumes of

lysis buffer [20 mivi Tris-HC1 (pH 7.4), 0.1 M NaC1, Smsi MgC12,

1% Nonidet P-40, 0.5% sodium deoxycholate, and 2 kallikrein

inhibitor units/mi aprotinin]. The protein concentration was

determined using the Bio-Rad Protein Assay (Bio-Rad Labora-

tories, Hercules, CA). SDS-PAGE and Coomassie blue staining

was performed for all samples as a control for degradation prior

to Western blotting.

For bax and hcl-x detection, 50 p.g of each protein sample

were separated on a 12% SDS-polyacrylamide gel and electro-

blotted onto nitrocellulose membranes (Bio-Rad). After electro-

blotting, the membranes were incubated with 6% nonfat dry

milk in 1 X TBST [0. 1 M Trizma base, 0. 15 M NaC1, and 0.05%

Tween 20 (pH 7.4)] for blocking and then with the primary

antibody in 3% nonfat dry milk in 1 X TBST. Following incu-

bation with an alkaline phosphatase-conjugated goat antirabbit

antibody, visualization of the bound antibody was performed

with the BCIPINBT Phosphatase Substrate System (Kirkegaard

& Perry Laboratories, Gaithersburg, MD). For bcl-2 detection,



Table 3 Distribution of bcl-XL and bcl-x5 RNA levels in primary ovarian carcinomas according to clinicopathological features

The values are the ratios between the relative adsorbance of the RNA of interest and aldolase-A RNA obtained by RT-PCR.

Clinicopathological

feature

bcl-xL bcl-x5

n Mean Median Range SD n Mean Median Range SD

Histology

Serous 29 0.33 0.27 0.10-0.71 0.18 23 0.14 0.13 0.02-0.30 0.09

Mucinous I 0.19 1 0.27

Endometrioid 6 0.29 0.29 0.21-0.50 0.12 5 0.10 0.10 0.02-0.19 0.07

Undifferentiated 4 0.34 0.39 0.18-0.46 0.14 3 0.09 0.12 0.02-0.14 0.06

Other 2 0.34 0.34 0.09-0.60 0.36 2 0.14 0.14 0.06-0.22 0.11

Stage

I-I! 5 0.28 0.32 0.21-0.34 0.15 4 0.04 0.04 0.02-0.07 0.08

III 29 0.32 0.27 0.09-0.71 0.17 23 0.14 0.14 0.02-0.30 0.09IV 8 0.33 0.33 0.14-0.71 0.18 7 0.13 0.14 0.02-0.21 0.07

Grading

1-2 9 0.24 0.27 0.14-0.34 0.06 8 0.12 0.12 0.02-0.27 0.09

3 29 0.31 0.28 0.09-0.71 0.18 24 0.11 0.08 0.02-0.30 0.09

Ascites

No 22 0.33 0.25 0.09-0.71 0.20 19 0.1 1 0.10 0.02-0.27 0.09

Yes 20 0.33 0.32 0.10-0.60 0.14 15 0.16 0.16 0.04-0.30 0.08

Residual tumor”�2 cm 22 0.36 0.35 0.09-0.71 0.18 18 0.16 0.19 0.02-0.30 0.09>2 cm 15 0.28 0.19 0.10-0.60 0.15 12 0.12 0.14 0.02-0.22 0.06

Response to

chemotherapy”

CR 13 0.32 0.35 0.09-0.71 0.19 10 0.11 0.12 0.02-0.25 0.06PR 12 0.39 0.43 0.14-0.71 0.18 9 0.18 0.17 0.02-0.27 0.08

NC-P 12 0.28 0.26 0.13-0.60 0.14 10 0.13 0.13 0.02-0.30 0.09Age (yr)

�60 29 0.31 0.28 0.01-0.70 0.17 23 0.13 0.12 0.02-0.27 0.06

>60 13 0.29 0.21 0.09-0.7 1 0. 18 1 1 0.09 0.05 0.02-0.27 0.08

a Only stage III-IV patients.S CR, complete response; PR, partial response; NC-P. no change-progression.


100 p.g of each protein sample were separated on a 15% SDS-

polyacrylamide gel and, after clcctroblotting onto polyvinyli-

dene fluoride (Millipore Co., Bedford, MA), the membranes

were incubated with the primary antibody, as described above.

The membranes were then transferred first to a biotinylated

secondary antibody and then to an avidin-biotin-horseradish

peroxidase complex (ABC Vectastain Elite; Vector Laborato-

rics, Burlingame, CA), and detection was performed with the

3,3-diaminobenzidinc peroxidase substrate kit (Vector Labora-

tories) in 1 x TBST. Images of the blots were acquired with a

Cohu charged coupled device camera, and quantification of the

bands was performed by Phoretix 1 D (Phoretix International).

Band intensity was expressed as relative adsorbance. The fol-

lowing antibodies were used: mouse antihuman bcl-2 (clone

124; DAKO, A/S, Glostrup, Denmark), rabbit antihuman bax,

and rabbit antihuman bcl-x (Santa Cruz Biotechnology, Santa

Cruz, CA). The latter antibody reacts with both bcl-xL and

bcl-x5, which can be distinguished by their different molecular

weights on a SDS-polyacrylamidc gel (Mr �30,000 and 18,000,

respectively).

p53 Analysis. p53 positivity was evaluated by immuno-

histochemistry on formalin-fixed, paraffin-embedded sections

using the antimouse antibody (clone DO-7; Oncogene, Manhas-

set NY; Ref. 31).

ER and PR Analysis. The cytosolic fractions of tissue

samples, prepared as described earlier (32), were used for ER

and PR assays, which were carried out by the dextran-coated

charcoal method according to the EORTC protocol (32).

Statistical Analysis. The Pearson’s correlation test was

used to analyze the relationship between RNA and protein levels

and the correlations among the bcl-2 family members. The

Mann-Whitney and Kruskal-Wallis nonparametric tests were

used to analyze the distribution of bcl-2, bax, bcl-xL, and bcl-x�

levels according to the clinicopathological characteristics of the

cases.

RESULTS

bcl-2, bax, bcl-xL, and bcl-x� RNA Expression Levelsare Different in Normal and Neoplastic Ovarian Tissues.bcl-2 and box RNA expression levels were measured by semi-

quantitative RT-PCR on 76 ovarian samples, which included 6

normal tissues, 44 primary carcinomas, 7 omental metastases,

and 19 recurrent carcinomas. bcl-x RNA levels were assessed on

a total of 71 samples, comprising six normal tissues, 42 primary

tumors, 6 omental metastases, and 17 recurrent tumors. The

reliability of the technique was established by comparing the

values obtained in six different experiments for the ovarian

cancer cell line A2780, which was used as a control. The values

were: bcl-2, mean, 2.11, SD, 0.21; bax, mean, 1.22, SD, 0.24;

bcl-xL, mean, 0.46, SD, 0.03, bc1-x�, mean, 0.09, SD, 0.02. All

the tissue samples showed a detectable band of the expected

molecular size of 385 bp for bcl-2 (Fig. lA) and of 538 bp for

bax (Fig. 1B). bcl-xL and bcl-x5 were amplified by the same pair

of primers and were visible as two distinct bands of 780 and 591

bp, respectively, in the same gel (Fig. lC). Although bcl-xL was



. bcl-2

- . aldolaseA A w-w �

I 23456789

�::=� � � bcl-2

B bax

� �LbcI�x

Fig. 2 Western blot of bcl-2 (A), bax (B). and bcl-x�� (C) in normal(Lanes 1-4) and neoplastic (Lanes 5-8) ovarian tissues. Lane 9, positivecontrols, which are the leukemic T cell line Jurkat (A) and the ovariancancer cell line A2780 (B and C).


I 2 3 4 5 6 7 9 9101112131415 1817

A � � --

B ��:::�±�:: . -bax

#{149}.i:� � #{149}- �- - -� 6 - aldolase A

I- ‘ - �.bcI-x1I -- � � . .-‘-� . “bcl-x9

� . . aldolase A

Fig. 1 Semiquantitative RT-PCR analysis of bcl-2 (A), box (B), andbcl-x,� (C) in primary ovarian cancers (lanes 1-8), mctastases (lanes

9-11) and recurrent diseases (lanes 12-15). The histological types ofthe tumors were: Lanes 3, 5, and 10, endometrioid; Lane 4, undifferen-tiated. The remaining lanes were all serous. Lane 16, ovarian cancer cellline A2780, used as positive control; Lane 17, negative (no sample)control. A and C, photographs of ethidium bromide-stained gels. B,

autoradiography of the RT-PCR analysis, which was transferred to filter

and hybridized with a bax cDNA probe.

present in all samples, bcl-x5 was detectable in only 83% of the

tumors (54 of 65) and was undetectable in normal samples

(Table 1). The RNA levels of these genes, when analyzed in the

whole population, were not normally distributed and were

slightly skewed to the left (data not shown). Median and range

values of bcl-2, bax, bcl-xL and bcl-s5 did not show any signif-

icant variation according to type of disease (Table 1). bcl-2

levels were significantly higher in normal than in neoplastic

ovarian tissues (P < 0.002); conversely, both bax and bcl-xL

RNA levels were lower in normal than in neoplastic samples

(P < 0.018 and P < 0.030, respectively; Table 1). The RNA

levels of bcl-2, box, bcl-xL, and bcl-x5 were not significantly

correlated with each other (data not shown). The distributions of

bcl-2, box, bcl-xL, and bcl-x� values according to the clinico-

pathological features of the cases examined are shown in Tables

2 and 3. box levels were higher in stage I and II than they were

in advanced-stage tumors (P < 0.009). bcl-2, bcl-xL, and bcl-x5

were not associated with any of the clinicopathological features

considered.

Protein Levels of bcl-2, bax, and bcl-xL Correlate with

RNA Levels. The protein expression levels of bcl-2, bax,

bcl-x�, and bcl-x5 were determined by Western blotting. bcl-2

and bax were measured on 20 ovarian carcinomas and 11

normal ovarian samples, whereas bcl-x expression was meas-

ured on 16 carcinomas and 8 normal ovarian samples. A band of

the expected M1 of 26,000 for bcl-2 (Fig. 2A) and of 21,000 for

bax (Fig. 2B) was detectable in all samples. Both isoforms of

bcl-x, bCl-XL as a Mr 30,000 band and bc1-x� as a Mr 18,000

band, were present in the A2780 control cell line, together with

an additional band of higher molecular weight (Fig. 2C, Lane 9).

In normal samples, bcl-xL was present together with this higher

molecular weight form (Lanes 1-4). Only the Mr 30,000 band

was visible in the neoplastic samples (Fig. 2C, Lanes 5-8),

some of which also contained additional bands that could cor-

respond to different isoforms (as in Lanes 5 and 8). We analyzed

only the Mr 30,000 band. bcl-x� was not present in any of the

tissue samples analyzed. The difference between normal and

neoplastic tissues is shown in the box and whisker plots in Fig.

3. bcl-2 levels were significantly lower in neoplastic (median,

112.0; range, 23.0-328.0) than in normal samples (median,

175.3; range, 97.0-265.9; P < 0.038; Fig. 3A). bax protein

levels followed the opposite trend (Fig. 3B), i.e., they were

higher in neoplastic (median, 1040.0; range, 398.0-1523.0) than

in normal samples (median, 236.0; range, 103.3-1040.0; P <

0.006). bcl-xL (Fig. 3C) was expressed at lower levels in normal

(median, 4,164.9; range, 1,036-8,967) than in neoplastic sam-

ples (median, 14,365; range, 7,544-17,485; P < 0.0033). Anal-

ysis of these proteins in normal and cancer tissues confirmed the

trend observed at the RNA level. The protein levels of the bcl-2

family members analyzed were directly correlated with their

corresponding RNA levels (Fig. 4), as follows: bcl-2, r =

+0.50, P < 0.047; bax, r = +0.44, P < 0.038; bcl-xL, r 0.53,

P < 0.023. bcl-x5, which was present at very low levels, as

measured by RT-PCR, was undetectable by Western blot in all

tissue samples.

Biological Correlation. p53 positivity was determinedby immunohistochemistry on 27 cases, 8 of which were positive

(30%). The RNA levels of bcl-2, bax, bcl-xL, and bcl-x� did not

show any significant correlation with p53 positivity. ER levels,

measured on 29 cases, ranged from 0.01 to 73.6 fmol/mg of

protein (mean, 16.6; median, 10.3; SD, 17.7) and were not

correlated with the RNA level of the bcl-2 family members

analyzed. The PR levels, which were measured on 29 samples

(range, 0.01-159.9 fmol/mg of protein; mean, 19.4; median,

10.6; SD, 10.6), although not correlated with bcl-2, were di-

rectly correlated with box (r +0.44, P < 0.019) and inversely

correlated with bcl-xL (r = -0.40, P < 0.035). In an attempt to

take advantage of the biological information gathered from each

of the parameters examined, we tried to combine bcl-2, box,

bcl-xL, and bcl-x� to define the susceptibility of each sample to

apoptosis. Positivity for the four parameters was defined on the

basis of the median levels for each of them. Cases that were



Normal Neopiastic

A

B

C

.

300

250

c� 200

U

� 150

100

50

0

2000

1600

)( 1200

� 800

400

0

20000

16000

12000

�8000

4000

0

#{149}_.,_-s_-

500

400

I:100

C 20000

16000

.c 12000

4000

Normal Neopiastic

Fig. 3 Box and whisker plots showing the difference in protein levels

between normal and neoplastic ovarian tissues of bcl-2 (A), bax (B), andbcl-x1 (C). -. Median values.

1 2 3 4 5 6

RNA

0 10 20 30 40 50 60 70 80

ANA


Normal Neoplastic

negative for bcl-2 and bcl-xL and positive for box and bcl-xs

were considered to represent a situation such that apoptosis

would be maximally favored. Cases that were positive for bcl-2L

and bcl-xL and negative for box and bcl-x� were considered to

represent a situation of maximal resistance to apoptosis. Sub-

groups representing intermediate situation were also taken into

consideration. No case fell into the category of maximal sus-

ceptibility to apoptosis. Nineteen of 70 cases (27%) fell into the

category of maximal resistance to apoptosis, and 5 1 of 70 cases

(73%) belonged to an intermediate situation, in which any

combination of the four genes is possible, so that apoptosis

would be determined by a mixture of contrasting inputs. These

subgroups were not differently distributed according to response

to chemotherapy.

DISCUSSION

The aim of this study was to elucidate the role of bcl-2 and

some bcl-2-rclated proteins (bax, bcl-xL, and bcl-x5) in ovarian

tissues. bcl-2, bax, and bcl-xL were present to a variable degree

in all samples tested, and their RNA and protein levels were

A 400

350

300C

:B 2500

�200

150

100

50

0 1 2 3 4 5

RNA

Fig. 4 Correlations between the RNA and protein levels of bcl-2 (A),

bax (B), and bd-XL (C).

directly correlated. bcl-x5, the RNA of which was detectable at

low levels in most tumor samples, was undetectable by Western

blotting, which could be indicative of its marginal role in ovar-

ian biology. The significant correlation between RNA and pro-

tein levels suggests that, although further posttranslational pro-

cesses could interfere with the regulation and the activity of

these proteins, their overall levels arc predominantly determined

by their respective RNA expression levels. This allowed us to

perform all the subsequent analyses on the data obtained by

RT-PCR, for which a larger number of samples was available.

In agreement with previous works on breast (22) and ovary

(26), we found that normal ovarian tissues have significantly

higher levels of bcl-2 than do neoplastic ovarian tissues. More-

over, normal tissues expressed lower bax, bcl-XL, and bcl-x�

levels than did neoplastic tissues. These data suggest the possi-

bility that the regulation and, possibly. the role of these genes in

apoptosis may differ in carcinoma, as compared with normal

tissue (33). The increase in bcl-xL levels concomitant with the

decrease in bcl-2 in ovarian tumor supports the hypothesis that

bcl-xL could take over the role ofbcl-2 in carcinoma (33). bcl-xL




has been shown to be a stronger protector against apoptosis than

bcl-2 under certain circumstances (34, 35), and its presence in

tumor is suggestive of an important role in cell turnover regu-

lation. The complex picture, in which the proapoptotic genes

box and bcl-x5 follow the same trend, whereas the antiapoptotic

genes bcl-2 and bcl-xL follow opposite trends in normal and

neoplastic ovary, is very similar to the one recently described in

gastric tissues (33). Conversely, other studies on breast (36) and

prostate (37) described a different trend, which is indicative of

the existence of a tissue-specific mechanism of regulation of the

expression of the bcl-2 family. The distributions of bcl-2, box,

bcl-xL, and bcl-x� levels did not show any variations among

primary, metastatic, or recurrent tumors, suggesting that modi-

fications of these apoptosis related genes are likely to influence

the acquisition of a malignant phenotype but seem to have no

further impact on tumor progression.

bcl-2 has been shown to be inversely associated with tumor

grade or stage in some but not all reports on breast cancer (21,

22), whereas bcl-2 and bax were shown to increase with tumorgrade in prostate cancer (37). In our series, high box levels were

associated with advanced tumor stage, whereas none of the otherparameters showed any major correlation with the clinicopath-

ological features we analyzed.

In vitro studies on ovarian cancer cell lines reported that

overcxprcssion of bcl-2 could prevent cells from cisplatin-in-

duced apoptosis, whereas bax was ineffective (38, 39); however,

no in vivo data was available. Our data showed that the degree

of response to chemotherapy did not vary according to the levels

of the bcl-2 family members analyzed, which emphasizes that

caution is required in transposing in vitro-generated results to an

in vivo setting.

p53, classically known as a tumor suppressor gene, is also

clearly involved in some major apoptotic pathways and has been

shown to regulate both bcl-2 and bax expression in vitro (40).

An inverse correlation between bcl-2 and p53 has been de-

scribed in some carcinomas but has not been unequivocally

reported in ovarian carcinoma (26, 41), and no correlation

between bcl-2 RNA levels and p53 immunostaining was present

in our series of samples.

Several studies on solid tumors (22, 24, 37, 42, 44) mdi-

cated that the regulation of bcl-2 could be hormone dependent,

but no information had yet been reported on ovarian cancer.

bcl-2 levels did not correlate with steroid receptors levels in our

series of samples. Conversely, we observed a direct correlation

between box and PR, which is paralleled by an inverse correla-

tion between bcl-xL and PR. Interestingly, hormonal regulation

seems to affect the two genes box and bcl-xL with high expres-

sion in carcinomas.

Our attempt to combine the data obtained on the four genes

to evaluate the susceptibility to apoptosis did not lead to any

solid conclusions, although, interestingly, we did not find in our

samples, the scenario that maximally favored apoptosis, which

is highly unlikely to be compatible with tissue viability. We are

still at an early stage because more bcl-2 family members are

still being discovered and it has yet to be defined how they are

regulated and to what extent their interaction is modulated in a

tissue-specific manner (45). It had already been hypothesized by

White (45) that it is the global balance between proapoptotic and

antiapoptotic proteins that is relevant for cell survival, and

different bcl-2 family members and different complexes or

different pathways may become predominant in different tis-

sues. Furthermore, we have to consider that various functions

arc being described for each of these genes (8, 46, 47); thus,

thinking of their role exclusively in term of regulation of apop-

tosis may be restrictive. Their expression in tumors could be

directed to pathways that are partially distinct from apoptosis.

The differences we reported in the pattern of expression of

bcl-2, bax, bcl-xL and bcl-x5 in carcinoma compared to normal

ovary point to a role of the bcl-2 family in the acquisition or

maintenance of the transformed features of the tissue, that could

not be drawn from in vitro studies, which arc not representative

of tissue complexity and heterogeneity. More work will, there-

fore, be required in vivo, to fully characterize the expression of

these and other apoptosis-related genes to evaluate their com-

plex interactions, which may influence tumor evolution, and

long-term studies will be required to determine their prognostic

role.

ACKNOWLEDGMENTS

We thank Dr. A. Pontecorvi for providing the aldolase-A primers.

REFERENCES

I . Thompson, C. B. Apoptosis in the pathogenesis and treatment of

disease. Science, 267: 1456-1462, 1995.

2. Reed, J. C. Bcl-2 and the regulation ofprogrammed cell death. J. CellBiol., 124: 1-6, 1994.

3. Hickman, J. A. Apoptosis induced by anticancer drugs. CancerMetastasis Rev., 11: 121-139, 1992.

4. Hockenbery, D. M., Nu#{241}ez,G., Milliman, C., Schreiber, R. D., and

Korsmeyer, S. J. Bcl-2 is an inner mitochondrial membrane protein that

blocks programmed cell death. Nature, 348: 334-336, 1990.

5. Nu#{241}ez,G., London, L., Hockenbery, D., Alexander, G. M., and

McKeam, J. P. Deregulated bcl-2 gene expression selectively prolongssurvival of growth factor-deprived hemopoietic cell lines. J. Immunol.,

144: 3602-3610, 1990.

6. Miyashita, T., and Reed, J. C. Bcl-2 oncoprotein blocks chemother-

apy-induced apoptosis in a human leukemia cell line. Blood, 81: 151-

157, 1993.

7. Yin, X., Oltvai, Z. N., and Korsmeyer, S. J. BHI and BH2 domains

of Bcl-2 are required for inhibition of apoptosis and heterodimerization

with Bax. Cell, 74: 607-619, 1993.

8. Rao, L., and White, E. Bcl-2 and the ICE family of apoptotic

regulators: making a connection. Cun. Opin. Genet. Dcv., 7: 52-58,

1997.

9. Sedlak, T. W., Oltvai, Z. N., Yang, E., Wang, K., Boise, L. H.,

Thompson, C. B., and Korsmeyer, S. J. Multiple Bcl-2 family membersdemonstrate selective dimerization with Bax. Proc. NatI. Acad. Sci.USA, 92: 7834-7838, 1995.

10. Sato, T., Hanada, M., Bodrug, S., Irie, S.. Iwama, N., Boise, L. H..

Thompson, C.B., Golemis, E., Fong, L.. Wang, H., and Reed, J. C.,Interactions among members of the bcl-2 protein family analyzed with

a yeast two-hybrid system. Proc. Natl. Acad. Sci. USA, 91: 9238-9242,

1994.

11. Oltvai, Z. N., Milliman, C. L., and Korsmeyer, S. J., Bcl-2 het-

erodimenzes in vivo with a conserved homolog. Bax, that accelerates

programmed cell death. Cell, 74: 609-619, 1993.

12. Yin, C., Knudson, C. M., Korsmeyer, S. J., and Van Dyke, T. Baxsuppresses tumorigenesis and stimulates apoptosis in vivo. Nature

(Lond.), 385: 637-640, 1997.

13. Boise, L. H., Gonzalez-Garcia, M.. Postema, C. E.. Ding. L..Lindsten, T., Turka, L. A., Mao, X., Nunez, G., and Thompson, C. B.

bcl-x, a bcl-2-related gene that functions as a dominant regulator of

apoptotic cell death. Cell, 74: 597-608, 1993.




14. Linette, G. P., Li, Y., Roth, K., and Korsmeyer, S. J. Cross talkbetween cell death and cell cycle progression: BcI-2 regulates NFAT-mediated activation. Proc. Natl. Acad. Sci. USA. 93: 9545-9552, 1996.

15. Distelhorst, C. W., Lam, M., and McCormick, T. S. Bcl-2 inhibitshydrogen peroxide-induced ER Ca� ‘ pool depletion. Oncogene, 12:

2051-2055, 1996.

16. Cerroni, L., Soyer, H. P., and Kerl, H. Bcl-2 protein expression incutaneous malignant melanoma and benign melanocytic nevi. Am. J.Dermopathol.. 17: 7-1 1, 1995.

17. Pezzella, F., Turley, H., Kuzu, I., Tungekar, M. F., Dunnill, M. S.,Pierce, C. B., Harris, A., Garter, K. C., and Mason, D. Y. bcl-2 protein

expression in non-small-cell lung carcinoma. N. Engl. J. Med., 329:

690-694, 1993.

18. McDonnell, T. J., Troncoso, P., Brisbay, S. M., Logothetis, C.,Chung, L. W. K., Hsieh, J., Tu, S., and Campbell, M. L. Expression ofthe protooncogene bcl-2 in the prostate and its association with emer-gence of androgen-independent prostate cancer. Cancer Res., 52: 6940-6944, 1992.

19. Silvestrini, R., Veneroni, S., Daidone, M. G., Benini, E., Boracchi,P., Mezzetti, M., Di Fronzo, G., Rilke, F., and Veronesi, U. The Bcl-2protein, a prognostic indicator strongly related to p53 protein in lymphnode-negative breast cancer patients. J. Nati. Cancer Inst. (Bethesda),

86: 499-504, 1994.

20. Krajewski, S., Bodrug, S., Gascoyne, R., Berean, K., Krajewska,

M., and Reed, J. C. Immunohistochemical analysis of McI-i and Bcl-2proteins in normal and neoplastic lymph nodes. Am. J. Pathol., 145:

515-525, 1994.

21. Ceccarelli, C., Santini, D., Chieco, P., Taffurelli, M., Marrano, D.,and Mancini, A. M. Multiple expression patterns of biopathological

markers in primary invasive breast carcinoma: a useful tool for eluci-

dating its biological behaviour. Ann. Oncol., 6: 275-282, 1995.

22. Sierra, A., Lloveras, B., Castellsagu#{233}, X., Moreno, L., Garcia-

Ramirez, M., and Fabra, A. Bcl-2 expression is associated with lymph

node metastasis in human ductal breast carcinoma. Int. J. Cancer, 60:

54-60, 1995.

23. Hellemans, P., van Dam, P. A., Weyler, J., van Oosterom, A. T.,Buytaert, P., and van Marck, E. Prognostic value of bcl-2 expression ininvasive breast cancer. Br. J. Cancer, 72: 354-360, 1995.

24. Hurlimann, J., Larrinaga, B., and Vala, D. M. L. bcl-2 protein in

invasive ductal breast carcinomas. Virchows Arch., 426: 163-168,1995.

25. Krajewski, S., Blomqvist, C., Franssila, K., Krajewska, M., Wase-nius, V. M., Niskanen, E., Nordling, S., and Reed, J. C. Reduced

expression of proapoptotic gene box is associated with poor response

rates to combination chemotherapy and shorter survival in women withmetastatic breast adenocarcinoma. Cancer Res., 55: 4471-4478, 1995.

26. Henriksen, R., Wilander, E., and Oberg, K. Expression and prog-

nostic significance of Bcl-2 in ovarian tumors. Br. J. Cancer, 72:

1324-1329, 1995.

27. Herod, J. J., Eliopoulos, A. G., Warwick, J., Niedobitck, G., Young,L. S., and Kerr, D. J. The prognostic significance of Bcl-2 and p53expression in ovarian carcinoma. Cancer Res., 56: 2178-2184, 1996.

28. WHO. WHO Handbook for Reporting Results ofCancer Treatment,Publication No. 48, pp. 16-21. Geneva: WHO, 1979.

29. Chomczynsky, P., and Sacchi, N. Single step method of RNAisolation by guanidium-thiocyanate-phenol-chloroform extraction. Anal.Biochem., 162: 156-159, 1987.

30. Wang, T. T. Y., and Phang, J. M. Effects of estrogen on apoptoticpathways in human breast cancer cell line MCF-7. Cancer Res., 55:

2487-2489, 1995.

31. Elledge, R. M., Clark, G. M., Fuqua, S.. Yu, Y., and Allred, D. C.

p53 protein accumulation detected by five different antibodies: relation-

ship of prognosis and heat shock protein 70 in breast cancer. Cancer

Res., 54: 3752-3757, 1994.

32. EORTC Breast Cancer Cooperative Group. Revision of the stand-ards for the assessment of hormone receptor in human breast cancer.

Eur. J. Cancer, 16: 1515-1517, 1980.

33. Krajewska, M., Fenoglio-Preiser, C. M., Krajewski, S., Song, K.,

Macdonald, J. S., Stemmerman, G., and Reed, J. C. Immunohistochem-ical analysis of Bcl-2 family proteins in adenocarcinomas of the stom-

ach. Am. J. Pathol., 149: 1449-1457, 1996.

34. Gottschalk, A. R., Boise, L. H., Thompson, C. B., and Quintans, J.Identification of immunosuppressant-induced apoptosis in a murineB-cell line and its prevention by bcl-x but not bcl-2. Proc. Natl. Acad.Sci. USA, 91: 7350-7354, 1994.

35. Shimizu, S., Eguchi, Y., Kosaka, H., Kamiike, W., Matsuda, H., andTsujimoto, Y. Prevention of hypoxia-induced cell death by Bcl-2 andBcl-xL. Nature (Lond.), 374: 811-813, 1995.

36. Bargou, R. C., Daniel, P. T., Mapara, M. Y., Bommert, K., Wage-ncr, C., Kallinich, B., Royer, H. D., and Dorken, B. Expression of the

bcl-2 gene family in normal and malignant breast tissue: low bax-ca

expression in tumor cells correlates with resistance towards apoptosis.Int. J. Cancer, 60: 854-859, 1995.

37. Krajewska, M., Krajewski, S., Epstein, J. I., Shabaik, A., Sauva-geot, J., Song, K., Kitada, S., and Reed, J. C. Immunohistochemicalanalysis of bcl-2, bax, bcl-x, and md- 1 expression in prostate cancers.

Am. J. Pathol., 148: 1567-1576, 1996.

38. Eliopoulos, A. G., Kerr, D. J., Herod, J. J. 0., Hodgkins, L.,

Krajewski, S., Reed, J. C., Young, L. S., and Herod, J. The control ofapoptosis and drug resistance in ovarian cancer: influence of p53 and

Bcl-2. Oncogene, 11: 1217-1228, 1995.

39. Perego, P., Giarola, M., Righetti, S. C., Supino, R., Caserini, C.,Delia, D., Pierotti, M. A., Miyashita, T., Reed, J. C., and Zunino, F.Association between cisplatin resistance and mutation of p53 gene andreduced bax expression in ovarian carcinoma cell systems. Cancer Res.,56: 556-562, 1996.

40. Miyashita, 1., Krajewski, S., Krajewska, M., Wang, H. G., Lin,H. K., Liebermann, D. A., Hoffman, B., and Reed, J. C., Tumorsuppressor p53 is a regulator of bcl-2 and box gene expression in vitroand in vivo. Oncogene, 9: 1799-1805, 1994.

41. Diebold, J., Barcuon, G., Felchner, M., Meier, W., Dopfer, K.,Schmidt, M., and Lohrs, U. bcl-2 expression, p53 accumulation and

apoptosis in ovarian carcinomas. Am. J. Clin. Pathol., 105: 341-349,1996.

42. Gee, J. M. W., Robertson, J. F. R., Ellis, I. 0., McClelland, R. A.,Hoyle, H. B., Kyme, S. R., Finlay, P., Blamey, R. W., and Nicholson,R. I., Immunohistochemical localization of bcl-2 protein in humanbreast cancers and it relationship to a series of prognostic markers and

response to chemotherapy. Int. J. Cancer, 59: 619-628, 1994.

43. Gompel, A., Sabounn, J. C., Martin, A., Yaneva, H., Audouin, J.,

Decroix, Y., and Poitout, P. Bcl-2 expression in normal endometriumduring the menstrual cycle. Am. J. Pathol., 6: 1195-1202, 1994.

44. Saegusa, M., Kamata, Y., Isono, M., and Okayasu, I. Bcl-2 expres-sion is correlated with a low apoptotic index and associated with

progesterone receptor immunoreactivity in endometrial cancer.J. Pathol., 180: 275-282, 1996.

45. White, E. Life, death, and the pursuit of apoptosis. Genes Dcv., 10:

1-15, 1996.

46. Minn, A. J., Velez, P., Shcndel, S. L., Liang, H., Muchmore, S. W.,Fesik, S. W., Fill, M., and Thompson, C. B. Bcl-XL forms an ionchannel in synthetic lipid membranes. Nature (Lond.), 385: 353-357.

47. O’Reilly, L. A., Huang, D. C. S., and Stasser, A. The cell deathinhibitor Bcl-2 and its homologues influence control of cell cycle entry.

EMBO J., 15: 6979-6990, 1996.



1998;4:517-524. Clin Cancer Res M Marone, G Scambia, S Mozzetti, et al. neoplastic ovarian tissues.bcl-2, bax, bcl-XL, and bcl-XS expression in normal and

Updated version

http://clincancerres.aacrjournals.org/content/4/2/517

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/4/2/517To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



bcl-2, bax, bcl-xL, and bcl-x Expression in Normal and ... · Table 1 Distribution of bcl-2, bax, bcl-xL and bcl-x5 RNA levels in normal and neoplastic ovarian tissues The values

Documents