Differential Expression of Antiapoptotic Genes in Human Endometrial Carcinoma: bcl-XL Succeeds bcl-2 Function in Neoplastic Cells

Gynecologic Oncology 77, 419 – 428 (2000) doi:10.1006/gyno.2000.5803, available online at http://www.idealibrary.com on

Differential Expression of Antiapoptotic Genes in Human Endometrial Carcinoma: bcl-XL Succeeds bcl-2 Function in Neoplastic Cells 1 Elvira Crescenzi, Ph.D., Vittoria Criniti, Ph.D., Mannida Pianese, M.D., Mario F. Tecce, M.D., Ph.D.,* and Giuseppe Palumbo, Ph.D. Dipartimento di Biologia e Patologia Molecolare e Cellulare “L. Califano,” and CEOS-CNR Facolta` di Medicina e Chirurgia, Universita` di Napoli “Federico II,” 80131 Naples; and *Istituto di Chimica, Facolta` di Medicina Veterinaria, Universita` di Bari, Italy Received November 4, 1999

homeostasis in multicellular organisms. It is an active process that, making use of both preexisting and de novo synthesized specific proteins, timely provides the removal of cells which are altered, in excess, or unusable. The mechanisms that commit cells to PCD are finely regulated as suggested by the large number of specific genes involved singularly or simultaneously in the positive or negative modulation of the process. For this reason, a comprehensive and exhaustive picture of the apoptotic pathway(s) is always difficult and very often impossible. Notwithstanding, some definite relationships between sensitivity/resistance to apoptotic stimuli and specific genes (for example the genes of the bcl-2 family) have been reported. One important regulator of apoptosis is Bcl-2, a 26-kDa protein that protects cells against apoptosis in a variety of experimental systems, including serum starvation, radiation, genotoxic drugs, heat shock, and UV radiation. The Bcl-2 protein is primarily localized to the nuclear envelope, the endoplasmic reticulum, and the outer mitochondrial membranes. A number of Bcl-2 homologues have been identified. Among these, Bax, a protein that, through dimerization with Bcl-2, antagonizes its action and promotes apoptosis [1] and Bcl-XL, whose sequence, structure, and function remarkably resemble that of the Bcl-2 protein [2]. Indeed, the transcription product of the bcl-X gene is converted, through alternative splicing, into two different mRNA, namely bcl-XL and bcl-XS. These variants give rise to two proteins: the above-mentioned Bcl-XL and Bcl-XS, which functions as an apoptosis-promoting factor. It is noteworthy that the pattern of Bcl-XL expression generally differs from that of Bcl-2, suggesting that the two genes may control PCD independently. It is becoming clear that deregulation of the delicate equilibrium between cellular proliferation and death events may be the cause of the development and progression of several human pathologies including cancer. Indeed, the specific contribution of bcl-2 family genes in tumor pathogenesis and progression has been widely analyzed in neuronal and lymphoid systems. Apoptotic processes in epithelia, are less characterized; in this

Objectives. Previous histochemical observations have suggested a possible involvement of the bcl-2 family genes in the acquisition of neoplastic phenotype of the endometrium. Since knowledge of the type and function of genes controlling the transformed cell may result in new diagnostic, prognostic, and therapeutic approaches, we have investigated at the molecular level the biological role of bcl-2 family genes in endometrial neoplastic cells. Methods. To investigate the relationship between the sensitivity to apoptosis and the expression of the bcl-2 family genes, we set up a model system consisting of four human endometrial carcinoma cell lines. This system constitutes an array of two cell pairs presenting, respectively, endometrioid and adenosquamous phenotypes. G2 and G3 gradings are represented within each pair; in addition, each set contains one cell line that is apoptosis-sensitive and one that is resistant. Transfection of bcl-2 and bcl-XL into apoptosis-sensitive cells was used to monitor the biological function of protective genes. Results. A differential pattern of expression of bcl-2 family genes was observed in apoptosis-sensitive versus resistant cells, independent from the histological subtype. Resistant lines exhibited high amounts of Bcl-XL and low amounts of Bcl-2. Bax expression clearly correlates with cellular susceptibility to apoptosis. Transfection of bcl-XL resulted in a dose-dependent enhancement in resistance toward apoptosis. In contrast, the main effect of bcl-2 constitutive overexpression was to drastically abate the proliferative potential of transfected cells. Conclusions. These data demonstrate, at the molecular level, that bcl-XL is selected as an apoptosis-protective gene in place of bcl-2 while bax retains its dominant proapototic role. © 2000 Academic Press

Key Words: endometrium; adenocarcinoma; apoptosis; bcl-2.

INTRODUCTION Apoptosis or programmed cell death (PCD) is a physiologic mechanism of cellular loss necessary for maintaining tissue 1 This research was supported by a grant from the Agenzia Spaziale Italiana (Rome, Italy) and from the Associazione Italiana per la Ricerca sul Cancro (Milan, Italy).

419

0090-8258/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

420

CRESCENZI ET AL.

TABLE 1 Characteristics of Endometrial Cancer Cells Histological phenotype Grade

Endometrioid

Adenosquamous

G2 G3

HEC1B KLE

RL95.2 AN3CA

regard only a few observations in nonmalignant versus malignant tissues have been reported [3–5]. Although endometrial carcinoma is the most common female pelvic malignancy, molecular events involved in the development and progression of this tumor have been in general scarcely investigated. To date, only a limited number of strictly histochemical observations on the expression profiles of the bcl-2 gene family in this system have been reported. Interestingly enough, Chan et al. [6], Chieng et al., [7], and Henderson et al. [8] indicated a bcl-2 down-regulation in carcinomas cells with respect to normal and preneoplastic endometrial lesions. These data suggested that bcl-2 expression could be modulated during the transformation process and suggested the possibility that the bcl-2 gene family was involved in the acquisition of the neoplastic phenotype. No subsequent observations have been reported. The present study describes a direct relationship between the expression of bax and sensitivity to apoptotic stimulus and recognizes bcl-XL as the apoptosis-protective gene in a panel of several endometrial carcinoma cell lines. Our results also provide molecular evidences that in malignant endometrial cells there is a selection against bcl-2. MATERIAL AND METHODS Cell Lines KLE, HEC1B, AN3CA, and RL95.2 endometrial carcinoma cell lines were all obtained from American Type Culture Collection (Rockville, MD). In conformity to the original histological characterizations [9 –13] the cells have been clustered, on the basis of relative histotype, into two groups (refer to Table 1). Cells were cultured according to indications: HEC1B and AN3CA were cultured in MEM (Sigma), KLE cells in a 1:1 mixture of DMEM and Ham’s F12 (Sigma), and RL95.2 cells in plain DMEM (Sigma). All culture media were supplemented with 10% fetal bovine serum, 2 mM glutamine, 0.5 U/ml penicillin, and 0.5 ␮g/ml streptomycin, unless differently specified. For serum withdrawal experiments, semi-confluent cells (i.e., between 1.5 and 2 ⫻ 10 6 cells) were plated and left to attach for 24 h. The next day, after extensive washing, cells were left to grow in serum-free medium.

Stably transfected HEC1B cells were cultured in MEM containing 10% FCS and G418 antibiotic (Calbiochem–Novabiochem) in a final concentration of 400 ␮g/ml. DNA Manipulations and Sequence Analysis Unless differently specified, DNA manipulations were performed according to standard procedure [14]. Cloning plasmid pT7Blue T-vector was obtained from Novagen and used according to the manufacturer’s instructions. Mammalian expression vector pcDNA3, which could be selected by G418 antibiotic, was obtained from Invitrogen. Sequencing was performed with the T7 sequencing kit (Pharmacia Biotech). The analyses of DNA sequences were performed using the University of Winsconsin Genetic Computer Group software package (Version 9.1, Madison, WI [15]). Plasmid Constructions The cDNAs of bcl-X mRNA species were obtained by RT-PCR. Total RNA from the HEC1B cell line was reverse transcribed using SuperScript II RNase H reverse transcriptase (Gibco BRL). Specifically synthesized primers were used to amplify simultaneously the coding regions of both bcl-XL and bcl-XS, giving rise to products of 780 and 591 bp, respectively, as already described [4]. PCR was carried out on 2 ␮l from the reverse transcription reaction in a final volume of 100 ␮l containing 2.5 U Taq polymerase (Promega), 20 ␮M of each deoxynucleotide, 1␮M of each specific primer, in the diluted enzyme reaction buffer for 30 cycles of 40 s at 94°C, 1 min at 60°C, 40 s at 72°C. Finally, samples were kept for 10 min at 72°C. The two products were cloned in the pT7Blue T-vector. The cDNA of bax-␣ was obtained from the HEC1B cell line by RT-PCR by specific primers [4]. The expected 323-bp product was then cloned in pT7Blue T-vector. The cDNA of bcl-2 was obtained from the AN3CA cell line by RT-PCR by specific primers [4]. The product, consisting of a 459-bp species, was also cloned in pT7Blue T-vector. The expression construct for bcl-XL was obtained by subcloning the sequence from pT7Blue T-vector to pcDNA3 expression vector. The pSFFV-neoexpression vector, both empty and containing the sequence encoding bcl-2, was kindly provided by Dr. J. Ashwell (NIH, Bethesda, MD). The validation of all constructs was performed by direct sequencing. RNA Isolation, Northern Blots, and RT-PCR Analysis Total cellular RNA was extracted from cells as previously described [16]. For Northern blots 30 ␮g of RNA was denatured in a formaldehyde/formamide-containing buffer, run on 1.2% (w/v) agarose, 2.2 M formaldehyde gels, and transferred onto Hybond-N membrane (Amersham). The electrophoretic position of the rRNA species was determined by ethidium

421

APOPTOSIS IN ENDOMETRIAL CARCINOMA CELLS

bromide staining. Cross-linking of the RNA on the membrane was carried out by standard UV irradiation. Membranes were prehybridized at 65°C in 0.25 M sodium phosphate, pH 7.2, 7% SDS. Hybridization was performed overnight at 65°C with labeled probe in fresh prehybridization solution. Labeling of cDNAs was done with ␣-[ 32P]dATP and ␣-[ 32P]dGTP by random priming with the DNA labeling kit from Boehringer Mannheim. The cDNAs used for labeling were either PCR amplification products or excised inserts from plasmids containing the sequence of interest. Filters were then washed four times in 20 mM sodium phosphate, pH 7.2, 1% SDS, at 55°C for 30 min. All the filters were autoradiographed (Kodak X-Omat XAR-5 films) for periods ranging from overnight (bax) to 10 days (bcl-2). Quantitative scanning of films was performed by a Discover Pharmacia scanner equipped with a Sun Spark Classic Workstation. For RT-PCR analysis of the two mRNA species of bcl-X (XL and XS), the relative amplification products [4] were separated by agarose gel electrophoresis, blotted onto a Hybond-N membrane (Amersham), and hybridized to a labeled bcl-XL probe. The two hybridizing species (800 and 600 bp, respectively) were finally revealed by autoradiography and quantitatively estimated by scanning. Western Blot Analysis Total cell proteins preparations were obtained lysing subconfluent 100-mm dishes by 1 mM EDTA, 0.2% Triton X-100, 1 ␮g/ml aprotinin, 170 ␮g/ml phenylmethylsulfonylfluoride. Protein concentration was routinely measured by the Bio-Rad protein assay. Polyacrylamide gels (15 or 7.5%) were prepared essentially as described by Laemmli [17]. Molecular weight standards were from Pharmacia. Proteins separated on polyacrylamide gels were blotted onto nitrocellulose filters (Hybond-C pure, Amersham) according to the procedure of Bittner et al. [18]. The actual total amount of proteins transferred on nitrocellulose was estimated by scanning (Discover Pharmacia scanner equipped with a Sun Spark Classic Workstation) after staining with Ponceau Red. Filters were washed and immunologically revealed by incubation with specific antibodies. Specific secondary antisera, conjugated with peroxidase (Amersham; diluted 1:2000), were left to react with primary immunoglobulins. Peroxidated immunoglobulins were finally revealed by ECL Western blotting detection reagent (Amersham) and quantitatively estimated by scanning. Corrections for total protein loading were made when necessary. Antibodies were purchased from Santa Cruz Biotechnologies (namely anti-Bcl-XS/L(S-18), anti-Bax(N-20), anti-Bcl2(C-100), and anti-PARP (H250)).

Cell Transfection HEC1B cells were stably transfected using the lipofectin reagent (Life Technologies) according to the manufacture’s protocol. Selection of transfectant cells was performed by adding, 48 h after addition of each plasmid construct, G418 antibiotic (Calbiochem–Novabiochem) to the medium in a final concentration of 400 ␮g/ml. bcl-XL and bcl-2 transfected cells were selected for geneticin resistance for 2–3 weeks and then pooled. In both cases mixed populations so obtained represent a large number of individual clones. Single clones from bcl-2 transfected cells were obtained by limiting dilution culture. DNA Oligonucleosomal Fragmentation Assay DNA fragmentation analysis was performed according to Lotem and Sachs [19]. Approximately 1 ⫻ 10 6 cells were lysed in 0.5 ml lysis buffer constituted by 10 mM Tris, pH 7.5, 0.6% SDS, 10 mM EDTA. RNase DNase-free solution (Boehringer Mannheim) was added to a final concentration of 15 ␮g/ml, and the lysate was incubated for 20 min at 37°C. After the addition of NaCl (final concentration of 1 M) samples were incubated at 4°C for 2 h. The tubes were centrifuged at 14,000g (30 min, 4°C); the supernatant was carefully aspirated and the pellet discarded. DNA in the supernatant was extracted by phenol– chloroform and precipitated with ethanol at ⫺20°C overnight in Eppendorf tubes. These tubes were centrifuged at 14,000g (30 min, 4°C). The pelletted DNA was air dried and finally resuspended in 20 ␮l of TE (10 mM Tris, 1 mM EDTA, pH 8) buffer. For oligonucleosomal fragmentation assay, 20 ␮g of DNA/lane was customarily loaded on a 1.5% agarose gel containing 0.5 ␮g/ml ethidium bromide. Molecular weight standards were from Boehringer Mannheim (phiX 174 DNA cleaved with HaeIII). Cell Growth Control HEC1B or transfectant cells were seeded in triplicate at 3 ⫻ 10 4 cells/well in a 24-well plate in MEM supplemented with 10% fetal bovine serum. Cell growth was assessed by direct counting 96 h after plating. RESULTS Cellular Susceptibility to Apoptosis Induction Four human cell lines were selected as an experimental system to study programmed cell death in endometrial carcinoma. As shown in Table 1, these lines are grouped into two sets of cells, representative of endometrioid and adenosquamous tumor phenotypes, respectively. Within each set a poorly (G3) and a moderately differentiated (G2) line is represented. To evaluate the role of bcl-2 family genes in the control of apoptotic processes, we first assessed the resistance to serum-

422

CRESCENZI ET AL.

dence of endometrial carcinoma cells from serum growth factors and hormones may largely vary. In addition, no straight correlation with the histological phenotype nor the tumoral grading can be assumed since, within each cell line pair, there is one cell line which is apoptosis-sensitive and one which is resistant. Expression of the bcl-2 Family Genes in Endometrial Carcinoma Cell Lines

FIG. 1. Serum-starvation-nduced apoptosis in endometrial carcinoma cells. (a) DNA internucleosomal fragmentation assay in HEC1B cells. The lane marked MW contains DNA size markers (bp); lanes 3–10, duplicate samples of DNA from cells starved for 1, 2, 3, and 4 days, respectively (as marked). (b) PARP proteolytic cleavage in HEC1B and AN3CA cells at 1, 2, and 3 days of starvation. Note the appearance of the 85-kDa fragment at 2 and 3 days, respectively. (c) Differential resistance to apoptosis in KLE, RL95.2, AN3CA, and HEC1B cell lines. Resistance of cell lines was expressed as the number of days required to detect a distinct DNA electrophoretic ladder and PARP cleavage after serum deprivation. Open and closed bars refer to cells presenting glandular or adenosquamous histotype, respectively. In each group, cells are ordered according to the respective tumor grading (G3 and G2).

starvation-induced apoptosis. To this purpose two independent methods of detection were used, i.e., DNA fragmentation and proteolytic cleavage of poly- (ADP)-ribose-polymerase (PARP). HEC1B and AN3CA cells were highly sensitive to PCD since a clear oligonucleosomal ladder was detected only 2 and 3 days after serum deprivation, respectively (Fig. 1a shows the fragmentation of DNA from HEC1B cells). A perfect temporal coincidence between the internucleosomal fragmentation and the appearance of an 85-kDa PARP fragment was detected in both lines (Fig. 1b). In turn, under similar conditions, RL95.2 and KLE cells survived up to 12 and 15 days. The relative sensitivity to serum-starvation-induced apoptosis of all four cell lines is summarized in Fig. 1c. These observations lead to the conclusion that the depen-

Expression of messengers. To investigate the molecular basis of the different resistance to programmed cell death, we have compared, for each histological subtype, the levels of expression of bcl-2 family genes in resistant and susceptible cells. As shown in Fig. 2a, Northern blot analysis clearly showed a differential expression of bax alpha messenger in sensitive versus resistant cells in both subtypes. HEC1B cells, in fact, demonstrated a level of bax alpha mRNA double that of KLE cells; analogously, sensitive AN3CA cells showed a manifestly more abundant bax alpha expression compared to the RL95.2 cell line. An inverse correlation was found for bcl-2 mRNA: in this case, in fact, both resistant cell lines showed a slightly higher level of bcl-2 expression with respect to their sensitive counterparts (Fig. 2b). The expression of bcl-X messengers, as a whole, did not correlate with cellular susceptibility, but interestingly resulted more abundantly in endometrioid cell lines, with respect to adenosquamous cells (Fig. 2c). Since the two splicing variants of bcl-X can not be discriminated by Northern analysis, the relative amounts of the two mRNA have been evaluated by RT-PCR analysis (Fig. 3a) using the same pair of primers. In all cell lines, polymerase chain reaction produced simultaneously with the two products corresponding to XL and XS forms. In each cell pair, the ratio between proapoptotic XS and protective XL forms, estimated by densitometry, was reproducibly greater in sensitive cells with respect to the resistant ones (Table 2). These results indicate that, in both endometrioid and adenosquamous endometrial cell types, reduced levels of bax and bcl-XS and slightly higher bcl-2 mRNA amounts, compared to apoptosis-sensitive cells, characterize apoptosis-resistant cells. Expression of proteins. To strengthen the above findings, we compared the levels of Bcl-2 family proteins in each set of cells. Western blot analysis confirmed the differential expression of Bax alpha in apoptosis-sensitive versus apoptosis-resistant cells. It appears that cellular susceptibility to PCD correlates with Bax alpha expression in both histological subtypes (Fig. 4a). Surprisingly enough, the analysis of Bcl-2 protein level did not reflect the data obtained by Northern blot (Fig. 4b). In fact, in this case KLE and RL95.2 cells demonstrated, with respect to their sensitive counterparts, significantly lower levels of

423

APOPTOSIS IN ENDOMETRIAL CARCINOMA CELLS

FIG. 3. bcl-X S expression in endometrial carcinoma cells. (a) RT-PCR analysis of bcl-X. The PCR amplification products of the two mRNA species of bcl-X (XL and XS) were separated on a 1.5% agarose gel, transferred to nitrocellulose membrane, and hybridized with a 32P-labeled cDNA probe for bcl-XL. (b) Western blot analysis of Bcl-XS protein (21 kDa). Each well contained about 50 ␮g of proteins from each cell lysate. Corrections for protein loading were made where necessary.

FIG. 2. Expression of bax (a), bcl-2 (b), and bcl-X (c) in endometrial carcinoma cells. (Left) About 30 ␮g total RNA from each individual cell line was analyzed by Northern blot. The 28S ribosomal RNA was used as a control for RNA loading. (Right) The levels of expression, evaluated by densitometry, are represented as a percentage of the amount exhibited by the most apoptosisresistant line (KLE). Open and closed bars refer to cells presenting glandular or adenosquamous histotype, respectively. In each group, cells are ordered according to the respective tumor grading (G3 and G2).

Bcl-2. This finding is suggestive of the existence of posttranscriptional regulative processes in endometrial carcinoma cells. Finally, we analyzed Bcl-XL protein levels (Fig. 4c). As described in the literature for some other female pelvic tissues [20], anti-Bcl-XL antibodies recognize on Western blots two closely migrating bands of Mr ⫽ 31 and 29 kDa, respectively. After normalization for total protein loading, the amount of the slower migrating band appeared to be similar in all cell lines,

while the level of the faster one appeared to vary significantly. A comparison between sensitive and resistant cells of both histological subtypes revealed a differential distribution of the 29-kDa protein. KLE cells demonstrated a level of 29-Bcl-XL higher than that in HEC1B cells; analogously, RL95.2 cells showed a level of 29-Bcl-XL higher than that in AN3CA. Western blot analysis of Bcl-XS isoform fully confirmed the findings obtained on messengers when comparing sensitive to nonsensitive cells (Fig. 3b). Quantitative data from messengers and proteins expression may be better organized according to histological phenotypes (endometrioid and adenosquamous) and, within the histotype, according to tumor grading G3 and G2 (Table 3). From this table, which compares data in terms of relative increased (⫹) or decreased (⫺) expression of messengers and proteins, an interesting symmetric picture ensues. As a whole, these results demonstrate that in two sets of endometrial carcinoma cell lines, representative of two different transformed phenotypes, cellular susceptibility to programmed cell death correlates with Bax alpha protein levels. A possible role for Bcl-XS apoptosispromoting protein is suggested, as well. The protein level of Bcl-2 appears to be significantly reduced in resistant cells of both endometrioid and adenosquamous phenotypes in contrast with their sensitive counterparts. In addition a clear selection for bcl-XL protective gene is distinctly manifest. Bcl-XL overexpression and serum-starvation-induced apoptosis. The data presented above suggested a preferential function of Bcl-XL as an apoptosis-protective protein. To TABLE 2 bcl-XS/XL Ratio in Endometrial Cancer Cells KLE

HEC1B

AN3CA

RL95.2

0.65

1.30

0.95

0.7

424

CRESCENZI ET AL.

FIG. 4. Relative levels of Bcl-2 family proteins in endometrial carcinoma cells. (Left) Western blots of Bax, Bcl-2, and Bcl-XL proteins. Each well contained about 50 ␮g of cellular proteins for Bcl-2 and ⬃10 ␮g for Bcl-XL and Bax. (Right) The levels of protein expression, evaluated by densitometry, are represented as a percentage of the amount exhibited by the most apoptosis-resistant line (KLE). Corrections for protein loading were made where necessary. Open and closed bars refer to cells presenting glandular or adenosquamous histotype, respectively. In each group, cells are ordered according to the respective tumor grading (G3 and G2).

analyze in detail the role of Bcl-XL in endometrial carcinoma cells, the most apoptosis-sensitive cell line HEC1B was transfected with a bcl-XL expression vector. These cells were compared with HEC1B transfected with the neomycin-resistant vector pcDNA3 alone. The low basal level of bcl-2 and the elevated susceptibility of HEC1B cells to serum-starvationinduced apoptosis make these cells particularly suitable for studying the specific effect of Bcl-XL on survival upon serum withdrawal.

After transfection, clones were selected in G418 and pooled. Various controls and mixed populations of stable transfectants were examined for exogenous Bcl-XL expression by Northern (Fig. 5) and Western blots (Fig. 6). Interestingly, Western blot analysis revealed that, after transfection, only the resulting amount of 29-Bcl-XL protein significantly increased, while no changes were detected in the 31-kDa band (Fig. 6). Among others, HEC1B-XLP4, P7, and P8 cell populations displayed, respectively, 29-Bcl-XL levels

APOPTOSIS IN ENDOMETRIAL CARCINOMA CELLS

425

TABLE 3 Relative Expression of Genes and Gene Products of the bcl-2 Family in KLE, HEC1B, and AN3CA and RL95 Cells Cell line

KLE

HEC1B

AN3CA

RL95.2

bcl-2 bax bcl-XL/XS Bcl-2 Bax Bcl-XS Bcl-XL Sensitivity to apoptosis

⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹

⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺

⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺

⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹

⫺

⫹

⫹

⫺

FIG. 6. Bcl-XL protein level in transfected HEC1B cells. Cell lysates were analyzed by immunoblot assay. Lane 1: HEC1B cells; lane 2: vector transfected cells (HEC1B-neo); lanes 3–7, bcl-XL transfected pools (HEC1BXLP9, P6, P7, P8, P4, respectively). Quantitative estimation of 31- and 29-kDa components was performed densitometrically. Corrections for protein loading were made where necessary.

2-, 2.5-, and 3-fold higher than controls (HEC1B-neo). Western blots experiments also demonstrate that the levels of expression of Bcl-XS were unchanged by transfection. As a consequence, all cells overexpressing Bcl-XL protein presented a significantly reduced XS/XL ratio. Furthermore, the amounts of both Bcl-2 and Bax proteins in transfected cells remained definitely unchanged (data not shown). The HEC1B-XLP4, P7, and P8 populations were analyzed for susceptibility to apoptosis induction by serum withdrawal. The P4 pool, which expresses 2-fold more Bcl-XL relative to control cells, is characterized by a slight increase in resistance to programmed cell death, since apoptosis is detectable after 4 days of starvation. The P7 bulk, which expresses 2.5-fold higher Bcl-XL, shows a clear DNA ladder at day 5 (Fig. 7a). Finally, the P8 pool, whose level of expression of Bcl-XL is more than 3-fold higher than control cells, demonstrates higher resistance to serum starvation, surviving up to 14 days. In this pool, the first detectable signs of DNA fragmentation were seen at day 6; such signs became progressively more evident

up to 2 weeks. This fact may reflect the nonhomogeneous character of the cell population and can be related to clonal variation in the mixed pool: cells that integrated more bcl-X acquired the highest resistance, whereas cells with a low number of cDNA copies were more prone to apoptosis. These experiments demonstrate that constitutive overexpression of bcl-XL in endometrial adenocarcinoma cell lines confers resistance to serum growth factors and hormone depletion in a dose-dependent manner (Fig. 7b). Bcl-2 overexpression and serum-starvation-induced apoptosis. Our finding that Bcl-2 protein level is reduced in apoptosis-resistant versus sensitive cell lines suggests a minor role for Bcl-2 in cell protection. On the other hand, in several cell systems, an ability of Bcl-2 to regulate different biological functions, such as differentiation and proliferation, has been demonstrated [21, 22]. In order to determine the role of bcl-2 gene in endometrial carcinoma cells, the apoptosis-sensitive HEC1B cell line was

FIG. 5. Expression of transfected bcl-XL cDNA in HEC1B cells. Total RNA (30 ␮g) from HEC1B cells (lanes 1, 2), control vector transfected pools (HEC1B-neoP1, P2, P3) and from bcl-XL transfected pools (HEC1B-XLP2, P3, P4, P6, P7, P8, respectively) was analyzed by Northern blot. The arrow indicates the transfected bcl-XL mRNA (780 bp). The upper band is determined by endogenous bcl-X mRNA.

FIG. 7. Transfection of bcl-XL in HEC1B cells resolves in increased resistance to apoptosis. (a) Transfected HEC1B-XLP7 cell population is apoptosis-insensitive up to 5 days after serum starvation. (b) The closed bar indicates the latency of apoptosis (days, left scale) in control cells and in transfectants; the open bar is the relative expression of 29-Bcl-XL (right scale). Note the striking correlation between apoptosis latency and the amount of 29-Bcl-XL.

426

CRESCENZI ET AL.

FIG. 8. Bcl-2 protein level in transfected HEC1B cells. About 50 ␮g of total cell lysate proteins/well was analyzed by Western blot. The amount of Bcl-2 protein from mixed populations P3, P13, and P5 and individual clones C1 and C2 is compared with the level of Bcl-2 from parental HEC1B cells. Corrections for protein loading were made where necessary.

transfected with the coding region of bcl-2 inserted in the eucaryotic expression vector pSFFV-neo (a generous gift from Dr. J. Ashwell, NIH, Bethesda, MD). These cells were compared with HEC1B transfected with the neomycin-resistant vector pSFFV alone. After transfection, clones were selected in G418 and either isolated or pooled. Western blot analysis revealed variable increases of Bcl-2 protein in transfectants compared to control cells. In particular, three mixed populations and two single clones, selected for subsequent investigations, displayed Bcl-2 amounts from 3- to 30-fold higher than controls (Fig. 8). Western blots experiments also demonstrated that Bcl-2 overexpression did not alter Bax, Bcl-XL, and Bcl-XS amounts (data not shown). On the basis of the biological behavior, it was possible to subdivide bcl-2 transfectants into two groups: cells overexpressing Bcl-2 up to 10-fold (P13 pool and C2 clone, respectively, in Fig. 8) and cells overexpressing Bcl-2 at a higher extent (up to 30-fold). The first group of cells showed a normal proliferative activity (data not shown), but did not acquire any resistance to serum-starvation-induced apoptosis (Fig. 9). On the other hand, P3 and P5 pools, as well as the C1 clone, demonstrated enhanced resistance to PCD, but showed a concomitant remarkable inhibition of cell proliferation. Cell counts were from ⬃10 to ⬃40% of control at 96 h after plating. This finding suggests that constitutive bcl-2 overexpression either results in ineffective protection of cells from apoptosis or is accompanied by growth inhibitory phenomena. DISCUSSION The expression of bax, bcl-2, and bcl-X genes has been evaluated by immunohistological methods in normal, preneoplastic, and transformed endometrial tissues [6 – 8, 20]. Two

main indications could be derived by such studies: (a) normal endometrial tissue and preneoplastic lesions express bax, bcl-2, and bcl-X; (b) bcl-2 is characteristically down-regulated in neoplastic tissues. These observations suggested a modulation of bcl-2 during the transformation process and indicated a possible involvement of bcl-2 family genes in the acquisition of the neoplastic phenotype. However, except for these evidences, no direct molecular and/or functional data are currently available on the effective biological role of bcl-2 family genes in endometrial neoplastic cells. To understand the biological function of bcl-2 family genes in the control of apoptotic processes in endometrial carcinoma cells, we set up an experimental system constituted by a panel of four human cell lines, representative of two major histological phenotypes (i.e., endometrioid and adenosquamous) of endometrial carcinomas. All cell lines were individually studied for relative expression of bax, bcl-2, and bcl-X genes and gene products and for susceptibility to apoptotic stimuli. Since the most common treatment of endometrial cancer is hormonal therapy (NCI, Cancer Database), we selected serum deprivation as the more appropriate way to induce apoptosis. Two independent methods of detection were used to evaluate the relative resistance of cells to PCD: appearance of PARP proteolytic fragment as an early marker [23, 24] and appearance of the DNA ladder, representative of a massive apoptotic event. The comparative analysis of sensitive and resistant cells demonstrates a characteristic pattern of expression of protective genes: both resistant cell lines, in fact, show high levels of Bcl-XL and low amounts of Bcl-2 proteins, compared to their susceptible counterparts. To evaluate the relative biological functions of the protective genes we have transfected the HEC1B human carcinoma cell line with bcl-2 or bcl-XL expression vectors. We demonstrate that Bcl-XL overexpression confers resistance to serum-starvation-induced apoptosis, in a strictly dosedependent manner. This function is possibly related to the

FIG. 9. Apoptosis in HEC1B cells transfected with bcl-2 Cells from clone C2 (a) and pool P13 (b) were serum starved for the times indicated. Apoptosis was revealed elctrophoretically by DNA internucleosomal fragmentation assay. The first lanes of both panels (marked MW) contain DNA size markers (bp). Note the appearance of a ladder at day 2.

APOPTOSIS IN ENDOMETRIAL CARCINOMA CELLS

Bcl-XL ability to counteract the proapoptotic action of Bax. Furthermore, our data suggest that the main effect of bcl-2 overexpression in HEC1B cell line is to inhibit cell growth. Previous reports described no modulation of Bcl-XL expression during the transformation process in endometrial tissues [6 – 8]. In agreement with these data, we present evidence that the protective function of Bcl-XL gene is retained in endometrial carcinoma cells. In addition, the original observation we now report regards the privileged correlation between the resistance to apoptotic stimuli and the amount of the Bcl-XL species of 29 kDa. As revealed by Western blots, the Bcl-XL protein appears as two closely migrating bands of ⬃29 and ⬃31 kDa. Although this observation has been already made and reported in several tissues, including thymus, lymph nodes, prostate, testes, ovary, oviduct, breast, stomach, duodenum, lung, liver, pancreas, kidney, epidermis, and brain [20], the origin of this doublet has not yet been clarified. Possible explanations of such an electrophoretic behavior include proteolytic processing, covalent modifications, initiation of translation from a downstream signal in the reading frame of BclXL, or even alternative splicing. In this study, quantitative analysis of the two electrophoretical bands in the four cell lines reveals a differential expression of the 29-kDa form in resistant versus sensitive cells. Moreover, bcl-XL transfection results in the exclusive overexpression of the 29-kDa species and in a proportional increase in resistance to apoptosis. The present data are suggestive of a possible different identity and function of these protein isoforms. The Bcl-2 protein level is reduced in resistant cells of both histological types. This observation is in line with a described inverse correlation between bcl-2 and bcl-XL protective genes: cells overexpressing Bcl-XL down-regulate Bcl-2 and cells with sustained expression of Bcl-2 showed a lower level of Bcl-XL [25]. In accordance with previous reports showing a specific ability of bcl-2 to inhibit the growth of solid tumor cells [21, 22], we observe that the main action of bcl-2 overexpression in HEC1B cells is to strongly reduce their proliferative potential. This effect may well account for the reported down-regulation of bcl-2 in neoplastic endometrial tissue and supports the hypothesis that bcl-XL is selected as a PCD-protective gene. In both tumor histotypes, cellular susceptibility to PCD is related to an elevated expression of proapoptotic members of the bcl-2 family, i.e., bax alpha and bcl-XS. However, a preferential role of Bax in the induction of PCD may be hypothesized since the amount of Bax alpha protein is consistently and distinctly more abundant compared to the Bcl-XS protein. In addition, the expression of Bax is manifestly greater in sensitive cells in comparison to the more resistant ones. In contrast, the distribution of the Bcl-XS isoform among the cell lines studied does not vary significantly. A recent observation proposed Bax as a major protein responsible for the regulation of apoptotic processes in normal human endometrium during

427

the menstrual cycle [26]. Our data support this hypothesis and extend it to endometrial carcinoma cells as well. In conclusion, the data presented demonstrate that, in both endometrioid and adenosquamous endometrial neoplastic cells, resistance to apoptosis is related to the 29-kDa Bcl-XL protein, while bcl-2 overexpression effects cell growth inhibition. Since Bax alpha expression correlates with sensitivity to apoptosis, it is possible that, in this system, the mechanism through which Bcl-XL exerts its particular function is to counteract the proapoptotic action of Bax. ACKNOWLEDGMENTS The authors are indebted to Dr. V. E. Avvedimento and to Dr. G. Chinali for thoughtful discussion, to Dr. G. Ashwell for providing the pSFFV-bcl-2 expression vector, and to Mr. F. D’Agnello and F. Berardone for the photographic work.

REFERENCES 1. Oltvai ZN, Milliman CL, Korsmeyer SJ: Bcl-2 heterodimerizes in vivo with a conserved homologue, bax, that accelerates programmed cell death. Cell 74:609 – 619, 1993 2. Boise LH, Gonza`lez-Garcı´a M, Postema CE, Ding L, Lindsten T, Turka LA, Mao X, Nun˜ez G, Thompson CB: Bcl-x, a bcl-2 related gene that functions as dominant regulator of apoptotic cell death. Cell 74:579 – 608, 1993 3. Tenniswood MP, Guenette RS, Lakins J, Mooibroek M, Wong P, Welsh JE: Active cell death ACD in hormone-dependent tissue. Cancer Metast Rev 11:197–220, 1992 4. Bargou RC, Daniel PT, Mapara MY, Bommert K, Wagener C, Kallinich B, Royer HD, Do¨rken B: Expression of the bcl-2 gene family in normal and malignant breast tissue: low bax-a expression in tumor cells correlates with resistance towards apoptosis. Int J Cancer 60:854 – 859, 1995 5. Tu H, Jacobs SC, Borkowski A, Kyprianou N: Incidence of apoptosis and cell proliferation in prostate cancer: relationship with TGF-beta and bcl-2 expression. Int J Cancer 69:357–363, 1996 6. Chan WK, Mole MM, Levison DA, Ball RY, Lu QL, Patel K, Hanby AM: Nuclear and cytoplasmic bcl-2 expression in endometrial hyperplasia and adenocarcinoma. J Pathol 177:241–246, 1995 7. Chieng DC, Ross JS, Ambros R.A: bcl-2 expression and the development of endometrial carcinoma. Mod Pathol 9:402– 406, 1996 8. Henderson GS, Brown KA, Perkins SL, Abbott TM, Clayton F: bcl-2 is down-regulated in atypical endometrial hyperplasia and adenocarcinoma. Mod Pathol 9:430 – 438, 1996 9. Richardson GS, Dickersin GR, Atkins L, MacLaughlin DT, Raam S, Merk LP, Bradley FM: KLE: a cell line with defective estrogen receptor derived from undifferentiated endometrial cancer. Gynecol Oncol. 17:213–230, 1984 10. Way DL, Grosso DS, Davis JR, Surwit EA, Christian CD: Characterization of a new human endometrial carcinoma (RL95-2) established in tissue culture. In vitro 19:147–158, 1983 11. Dawe CJ, Banfield WG, Morgan WD, Slatick MS, Curth HO: Growth in continuous culture, and in hamsters, of cells from a neoplasm associated with acanthosis nigricans. J Natl Cancer Inst 33:441– 456, 1964 12. Kuramoto H: Establishment of a cell line of human endometrial adenocarcinoma in vitro. Acta Obst et Gynaec Jap 19:47–58, 1972 13. Kuramoto H, Tamura S, Notake Y: Studies of the growth and cytogenetic

428

14. 15. 16.

17. 18.

19.

20.

CRESCENZI ET AL.

properties of human endometrial adenocarcinoma in culture and its development into an established line. Am J Obstet Gynecol 114:1012–1019, 1972 Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, Cold Spring Harbor Laboratory Press, 1989 Devereux J, Haeberli P, Smithies O: A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res 12:387–395, 1984 Chomczynski P, Sacchi N: Single step method of RNA isolation by acid guanidinium thiocyanate–phenol– chloroform extraction. Anal Biochem 162:156 –159, 1987 Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 – 685, 1970 Bittner M, Kupferer P, Morris CF: Electrophoretic transfer of proteins and nucleic acids from slab gels to diazobenzyloxymethyl cellulose or nitrocellulose sheets. Anal Biochem 102:459 – 471, 1980 Lotem J, Sachs L: Regulation of bcl-2, bcl-XL and bax in the control of apoptosis by hematopoietic cytokines and dexamthasone. Cell Growth Differ 6:647– 653, 1995 Krajewski S, Krajewska M, Shabaik A, Wang HG, Irie S, Fong L, Reed JD: Immunohistochemical analysis of in vivo patterns of bcl-X expression. Cancer Res 54:5501–5507, 1994

21. Pietenpol JA, Papadopoulos N, Markowitz S, Willson JKV, Kinzler KW, Vogelstaein B: Paradoxical inhibition of solid tumor cell growth by bcl-2. Cancer Res 54:3714 –3717, 1994 22. Vairo G, Innes KM, Adams JM: Bcl-2 has a cell cycle inhibitory function separable from its enhancement of cell survival. Oncogene 13:1511–1519, 1996 23. Kaufmann SH, Desnoyers S, Ottaviano Y, Davidson NE, Poirier GG: Specific proteolytic cleavage of poly (ADP) ribose polymerase: an early marker of chemoterapy apoptosis. Cancer Res 53:3976 –3985, 1993 24. Browne SJ, MacFarlane M, Cohen GM, Paraskeva C: The adenomatous polyposis coli protein and retinoblastoma protein are cleaved early in apoptosis and are potential substrates for caspases. Cell Death Diff 5:206 – 213, 1998 25. Poruchynsky MS, Wang EE, Rudin CM, Blagosklonny MV, Fojo T: Bcl-XL is phosphorylated in malignant cells following microtubule distuption. Cancer Res 58:3331–3338, 1998 26. Tao XJ, Tilly KI, Maravei DV, Shifren JL, Krajewski S, Reed JC, Tilly JL, Isaacson KB: Differential expression of members of the bcl-2 gene family in proliferative and secretory human endometrium: glandular ephitelial cell apoptosis is associated with increased expression of bax. J Clin Endocrinol Metab 82:2738 –2746, 1997