Portal de Periódicos da CAPES

Overexpression of Bcl-2 in Kaposi's Sarcoma-Derived Cells

1998; Elsevier BV; Volume: 111; Issue: 3 Linguagem: Inglês

10.1046/j.1523-1747.1998.00314.x

1523-1747

Chantal Degraef, Jean‐Christophe Noël, Dominique Fokan, Lin Zhou, Olivier Pradier, Martine Ducarme, Liliane Schandené, Jean‐Paul Van Vooren, Thierry Simonart, Dominique Parent, Michel Heenen,

Histiocytic Disorders and Treatments

The pathogenesis of Kaposi’s sarcoma (KS), a tumor of probable vascular origin, remains an enigma. It is still unclear whether KS is a true malignancy or whether it represents a reactive polyclonal process. Using both an immunohistochemical and an immunoblot approach, we found that cells derived from KS lesions express significant levels of Bcl-2, a protein known to prolong cellular viability and to antagonize apoptosis. Bcl-2 expression was found in AIDS-related KS-derived cells, as well as in cells derived from iatrogenic and sporadic KS, indicating that Bcl-2 upregulation may be important in the pathogenesis of KS regardless of its epidemiologic form. By contrast, fibroblasts and dermal microvascular endothelial, cells which are the probable vascular progenitors of KS cells, expressed low levels of Bcl-2. The expression of Bcl-2 in KS-derived cells was associated with a long-term survival in serum-deprived conditions, a situation that has been shown to induce apoptosis in various cell types. Incubation of fibroblasts or of dermal microvascular endothelial cells with KS cell-free supernatants did not enhance Bcl-2 expression, suggesting that Bcl-2 expression is not mediated by an agent released by KS cells. Analogously, KS supernatants failed to promote the viability of fibroblasts and of dermal microvascular endothelial cells cultured in serum-free conditions. Our findings suggest that the spindle cells derived from KS have a survival advantage and may adequately represent the tumor cells of KS. The pathogenesis of Kaposi’s sarcoma (KS), a tumor of probable vascular origin, remains an enigma. It is still unclear whether KS is a true malignancy or whether it represents a reactive polyclonal process. Using both an immunohistochemical and an immunoblot approach, we found that cells derived from KS lesions express significant levels of Bcl-2, a protein known to prolong cellular viability and to antagonize apoptosis. Bcl-2 expression was found in AIDS-related KS-derived cells, as well as in cells derived from iatrogenic and sporadic KS, indicating that Bcl-2 upregulation may be important in the pathogenesis of KS regardless of its epidemiologic form. By contrast, fibroblasts and dermal microvascular endothelial, cells which are the probable vascular progenitors of KS cells, expressed low levels of Bcl-2. The expression of Bcl-2 in KS-derived cells was associated with a long-term survival in serum-deprived conditions, a situation that has been shown to induce apoptosis in various cell types. Incubation of fibroblasts or of dermal microvascular endothelial cells with KS cell-free supernatants did not enhance Bcl-2 expression, suggesting that Bcl-2 expression is not mediated by an agent released by KS cells. Analogously, KS supernatants failed to promote the viability of fibroblasts and of dermal microvascular endothelial cells cultured in serum-free conditions. Our findings suggest that the spindle cells derived from KS have a survival advantage and may adequately represent the tumor cells of KS. endothelial cells human herpesvirus 8 Kaposi's sarcoma Kaposi's sarcoma (KS) is an angioproliferative disease that has aroused considerable interest of late, as it is one of the major clinical manifestations found in patients with the acquired immune deficiency syndrome (AIDS). Until the AIDS epidemic, this tumor was identified in three different settings: classic KS, African-endemic KS, and immunosuppressive drug-related KS. All forms of KS have the same histology characterized by a prominent angiogenesis and by the proliferation of spindle-shaped cells that are considered to be the tumor cells of KS. The histogenetic origin of the spindle cells remains controversial. A widely held view is that these cells derive from endothelial cells (EC), though they express macrophage antigens and fibroblast markers. Whether KS is a true malignancy or a polyclonal proliferation is also unknown. Separate lesions from the same patients are derived from an identical clone, which would favor metastatic spread from a common source (Rabkin et al., 1997Rabkin C.S. Janz S. Lash A. et al.Monoclonal origin of multicentric Kaposi's sarcoma lesions.N Engl J Med. 1997; 336: 988-993Crossref PubMed Scopus (206) Google Scholar). On the other hand, KS shows high incidence of spontaneous remission, contains low levels of DNA aneuploidy, and lacks nuclear atypia, which would favor a nonmalignant process (Fukunaga and Silverberg, 1990Fukunaga M. Silverberg S.G. Kaposi's sarcoma in patients with acquired immune deficiency syndrome.Cancer. 1990; 66: 758-764Crossref PubMed Scopus (51) Google Scholar;Simonart et al., 1997Simonart T. Noel J.C. Van Vooren J.P. Parent D. Baghli B. Querton G. Verhest A. DNA diploidy in AIDS-related and steroid-induced Kaposi's sarcoma.Br J Dermatol. 1997; 136: 804Crossref PubMed Scopus (5) Google Scholar). Supporting this view, cells derived from KS biopsies grow transiently in culture and occasionally show malignant features such as karyotypic abnormalities and the ability to produce malignant tumors in immunodeficient mice (Roth et al., 1988Roth W.K. Werner S. Risau W. Remberger K. Hofschneider P.H. Cultured, AIDS-related Kaposi's sarcoma cells express endothelial cell markers and are weakly malignant in vitro.Int J Cancer. 1988; 42: 767-773Crossref PubMed Scopus (72) Google Scholar;Lunardi-Iskandar et al., 1995Lunardi-Iskandar Y. Bryant J.L. Zeman R.A. et al.Tumorigenesis and metastasis of neoplastic Kaposi's sarcoma cell line in immunodeficient mice blocked by a human pregnancy hormone.Nature. 1995; 375: 64-68Crossref PubMed Scopus (217) Google Scholar). Finally, a mixture of cytokines and growth factors secreted by KS cells has been shown to alter the normal physiology of EC, leading to the emergence of activated EC (Fiorelli et al., 1995Fiorelli V. Gendelman R. Samaniego F. Markham P.D. Ensoli B. Cytokines from activated T cells induce normal endothelial cells to acquire the phenotypic and functional features of AIDS-Kaposi's sarcoma spindle cells.J Clin Invest. 1995; 95: 1723-1734Crossref PubMed Scopus (151) Google Scholar). The current hypothesis to explain the formation and development of KS lesions is a multistep pathway involving cytokine dysregulation, active infection by HIV-1 and/or by a new herpesvirus designated human herpesvirus 8 (HHV-8), or KS-associated herpesvirus, overexpression of integrins, and escape from programmed cell death (apoptosis) (Chang et al., 1994Chang Y. Cesarman E. Pessin M.S. Lee F. Culpepper J. Knowles D.M. Moore P.S. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi's sarcoma.Science. 1994; 266: 1865-1869Crossref PubMed Scopus (4974) Google Scholar;Bohan Morris et al., 1996Bohan Morris C. Gendelman R. Marrogi A.J. Lu M. Lockyer J.M. Alperin-lea W. Ensoli B. Immunohistochemical detection of Bcl-2 in AIDS-associated and classical Kaposi's sarcoma.Am J Pathol. 1996; 148: 1055-1063PubMed Google Scholar;Noel et al., 1996Noel J.C. Hermans P. Andre J. et al.Herpesvirus-like DNA sequences and Kaposi's sarcoma. Relationship with epidemiology, clinical spectrum, and histologic features.Cancer. 1996; 77: 2132-2136Crossref PubMed Scopus (79) Google Scholar;Pammer et al., 1996Pammer J. Plettenberg A. Weninger W. et al.CD40 antigen is expressed by endothelial cells and tumor cells in Kaposi's sarcoma.Am J Pathol. 1996; 148: 1387-1396PubMed Google Scholar). The pathogenic links between these steps remain unclear. One of the most important negative regulators of apoptosis is Bcl-2 and its related proteins, which confer a survival advantage to a variety of cell types including lymphoid cells, fibroblasts, melanocytes, epithelial, and neuroneal cells (Tsujimoto et al., 1985Tsujimoto Y. Cossman J. Jaffe E. Croce C.M. Involvement of the bcl-2 gene in human follicular lymphoma.Science. 1985; 228: 1440-1443Crossref PubMed Scopus (1609) Google Scholar;Reed, 1994Reed J.C. Bcl-2 and the regulation of programmed cell death.J Cell Biol. 1994; 124: 1-6Crossref PubMed Scopus (2390) Google Scholar;Lu et al., 1995Lu P.J. Lu Q.J. Rughetti A. Taylor-Papadimitriou J. Bcl-2 overexpression inhibits cell death and promotes the morphogenesis, but not tumorigenesis of human mammary epithelial cells.J Cell Biol. 1995; 129: 1363-1378Crossref PubMed Scopus (94) Google Scholar;Kroemer, 1997Kroemer G. The proto-oncogene Bcl-2 and its role in regulating apoptosis.Nature Med. 1997; 3: 614-620Crossref PubMed Scopus (1710) Google Scholar). The extended cell survival due to Bcl-2 may lead to subsequent acquisition of genetic changes resulting ultimately in tumor progression. Recent studies have shown strong Bcl-2 expression among the spindle-shaped tumor cells of cutaneous KS lesions (Bohan Morris et al., 1996Bohan Morris C. Gendelman R. Marrogi A.J. Lu M. Lockyer J.M. Alperin-lea W. Ensoli B. Immunohistochemical detection of Bcl-2 in AIDS-associated and classical Kaposi's sarcoma.Am J Pathol. 1996; 148: 1055-1063PubMed Google Scholar;Dada et al., 1996Dada M.A. Chetty R. Biddolph S.C. Schneider J.W. Gatter K.C. The immunoexpression of bcl-2 and p53 in Kaposi's sarcoma.Histopathology. 1996; 29: 159-163Crossref PubMed Scopus (41) Google Scholar). Another study has revealed weak Bcl-2 immunoreactivity but overexpression of Bcl-xl, another protein that prevents apoptosis (Foreman et al., 1996Foreman K.E. Wrone-Smith T. Boise L.H. et al.Kaposi's sarcoma tumor cells preferentially express Bcl-xl.Am J Pathol. 1996; 149: 795-803PubMed Google Scholar). In this study, we have investigated the expression of Bcl-2 by cultured KS cells derived from different epidemiologic forms of KS. Fibroblasts and dermal microvascular EC, which are the probable vascular progenitors of KS cells, served as controls. Furthermore, we have tested whether supernatants from KS cell cultures can induce the expression of Bcl-2 in fibroblasts or EC or enhance their survival after serum withdrawal, a stimulus that has been shown to induce apoptosis in various cell types (Hase et al., 1994Hase M. Araki S. Kaji K. Hayashi H. Classification of signals for blocking apoptosis in vascular endothelial cells.J Biochem. 1994; 116: 905-909PubMed Google Scholar). Five independent KS cell cultures were derived from cutaneous KS biopsies from three AIDS patients (KS-1, patch-stage KS; KS-2 and -3, nodular-stage KS), from one renal recipient transplant (KS-4, plaque-stage KS), and from one patient with sporadic KS (KS-5, plaque-stage KS). Cultures were established similarly as previously described (Benelli et al., 1994Benelli R. Repetto L. Carlone S. Parravicini C. Albini A. Establishment and characterization of two new Kaposi’s sacoma cell cultures from an AIDS and a non-AIDS patient.Res Virol. 1994; 145: 251-259Crossref PubMed Scopus (31) Google Scholar). Briefly, fresh lesions were minced inyo small pieces and enzymatically digested with collagenase (1 mg per ml) from Clostridium histolyticum for 90 min. Cells were grown without additional growth factors in minimum essential medium D-Val (Gibco, Paisly, Scotland) containing 10% inactivated fetal calf serum (Gibco), 1% nonessential amino acids, penicillin (100 U per ml), and streptomycin (100 μg per ml). Cells isolated from these cultures were characterized as KS cells morphologically (spindle-shaped morphology) and immunohistochemically (positivity for laminin, vimentin, and collagen IV in the majority of the cells; positivity for α-smooth muscle actin and ICAM-1 in up to 50% of the cells, and negativity for cytokeratin, ICAM-2, ELAM-1, VCAM-1, CD4, CD34, CD40, CD45), in accordance with previous published reports (Roth et al., 1988Roth W.K. Werner S. Risau W. Remberger K. Hofschneider P.H. Cultured, AIDS-related Kaposi's sarcoma cells express endothelial cell markers and are weakly malignant in vitro.Int J Cancer. 1988; 42: 767-773Crossref PubMed Scopus (72) Google Scholar;Benelli et al., 1994Benelli R. Repetto L. Carlone S. Parravicini C. Albini A. Establishment and characterization of two new Kaposi’s sacoma cell cultures from an AIDS and a non-AIDS patient.Res Virol. 1994; 145: 251-259Crossref PubMed Scopus (31) Google Scholar;Pammer et al., 1996Pammer J. Plettenberg A. Weninger W. et al.CD40 antigen is expressed by endothelial cells and tumor cells in Kaposi's sarcoma.Am J Pathol. 1996; 148: 1387-1396PubMed Google Scholar;Simonart et al., 1996Simonart Th Noel J.C. Liesnard C. et al.Kaposi's sarcoma and herpesvirus 8: a word of caution.Dermatology. 1996; 193: 272Crossref PubMed Scopus (11) Google Scholar). The experiments presented here were performed on cell cultures at the third to the fourteenth passage for immunohistochemistry, at the seventh to the thirteenth passage for western blot analysis, and at the seventh to the eighth passage (doubling time, 4–5 d) for cell survival kinetics. Dermal microvascular EC were cultured by a method similar to the one reported byKluger et al., 1997Kluger M.S. Johnson D.R. Pober J.S. Mechanism of sustained E-selectin expression in human dermal microvascular endothelial cells.J Immunol. 1997; 158: 887-896PubMed Google Scholar. EC were isolated from normal adult breast skin obtained as discarded tissue from reduction mammoplasties (Department of Plastic Surgery, Erasme University Hospital). Informed consent was obtained from the patients. Fresh skin was stretched flat and sectioned horizontally. After a 60 min incubation in 2 mg dispase II per ml (Boehringer, Mannheim, Germany) at room temperature, the epidermis was peeled off and cells from both sides of the underlying epidermis were gently scraped into RPMI (Gibco) and filtered through a 70 μM nylon mesh. The filtrate, containing single cells, was washed once in minimum essential medium D-Val and plated onto tissue culture plastic precoated with 5 μg fibronectin per cm2 (Sigma). EC were allowed to attach ≈4 h before gentle aspiration (to remove cell debris) and addition of fresh media. This medium consists of minimum essential medium D-Val (Gibco) supplemented with 20% fetal calf serum (Gibco), glutamine 2 mM, penicillin (100 U per ml) and streptomycin (100 μg per ml), 100 μg heparin per ml (Braun), 50 μg endothelial cell growth factor per ml (Fluka), 0.5 mM dibutyryl-cAMP, and 3.3 μM isobutyl-methyl-xanthine (both from Sigma). FACS analysis revealed that >90% of the cells were positive for von Willebrand factor. EC were used at passage 3–4 (doubling time, 2–3 d). Human dermal fibroblasts obtained from foreskins were cultured in RPMI (Gibco) containing 10% fetal calf serum (Gibco), 1% nonessential amino acids, penicillin (100 U per ml), and streptomycin (100 μg per ml). These were used at passage 7–8 (doubling time, 3–4 d). Immunohistochemistry using I-19 (Bax) and N-19 (Bcl-2) rabbit polyclonal antibodies (Santa Cruz Biotechnology) and Ab-1 (Bcl-2) (Oncogene Research Products, Calbiochem) monoclonal antibody was carried out on cytospins. Briefly, subconfluent cell cultures were harvested and made into single-cell suspension by trypsinization. For I-19 (Bax) and N-19 (Bcl-2), the cytospins preparations were incubated in 3% (vol/vol) H2O2 in methanol for 1 h at room temperature. After permeabilization with 0.1% saponin, the samples were incubated in normal goat serum for 30 min to block nonspecific interactions, and then anti-Bax or anti-Bcl-2 antibodies were applied at the dilution of 1 μg per ml overnight at 4°C. Normal rabbit immunoglobin fraction (Dako, Glostrup, Denmark) at the same concentration as the primary antibodies served as negative controls. Immunohistochemical staining was achieved with the PK-4001 Vectastain ABC Kit (Vector Laboratories, Burlingame, CA). The slides were counterstained with hematoxylin. For Ab-1 (Bcl-2) monoclonal antibody, the samples were fixed in formol for 1 h at room temperature and then microwaved for 2 × 5 min in a citrate buffer, pH 6. After blocking with normal goat serum 1/20 for 30 min, the cytospins were incubated with 5 μg Bcl-2 (Ab-1) monoclonal antibody per ml and then with Euvision+/HRP (Dako). Peroxidase activity was developed with 3,3′-diaminobenzidine tetrahydrochloride (0.012% in phosphate-buffered saline) and H2O2 (0.1%) and the cells were counterstained with hematoxylin. Confluent cell cultures were washed with complete phosphate-buffered saline and lyzed in sodium dodecyl sulfate buffer (5% β-mercaptoethanol, 10% glycerol, 80 mM sodium dodecyl sulfate, 60 mM Tris, pH 6.8). Lysates from MCF-7 cells were used as a positive control. For each sample, a total quantity of 100 μg of protein was electrophoresed on a 10%–17% sodium dodecyl sulfate-polyacrylamide linear gradient slab gel, in Tris-buffered saline solution. After the electrophoresis, the proteins were electrophoretically transferred to nitro-cellulose sheets. The blots were subsequently incubated for 1 h in 1% fat-free milk in TBS and for 18 h with Bcl-2 (Ab-1) monoclonal antibody at a 2 μg per ml concentration in TBS (20 mM Tris, 125 mM NaCl). After extensive washing in TBS, the immune complex was detected with a biotinylated anti-mouse Ig (diluted 1:250 in TBS), serving as a binding bridge to biotin-streptavidin peroxydase preformed complexes (Amersham, Little Chalfont, U.K.) used at the same dilution. Preformed complex was detected by photographic recording of the chemiluminescence (ECL) emitted by a H2O2 reacting probe, using the Boehringer kit. The immunoblots were scanned with a Microtek Phantom 4800 apparatus. Subconfluent cell cultures were induced to undergo programmed cell death by serum deprivation as described byHébert et al., 1994Hébert L. Pandey S. Wang E. Commitment to cell death is signalled by the appearance of a terminin protein of 30k.Exp Cell Res. 1994; 210: 10-18Crossref PubMed Scopus (37) Google Scholar. In brief, the fetal calf serum containing medium was removed. The cells were then washed two times with serum-free medium in order to remove residual cell-associated fetal calf serum. After the second rinse, the serum-free medium was placed onto the monolayers cultures, which were incubated at 37°C for various periods of time and assayed for viability as described below. For control experiments, the medium was replaced with fresh medium containing 10% fetal calf serum. Conditioned media from KS cells were obtained by washing confluent KS cell monolayers twice with phosphate-buffered saline and incubating them for 48 h with serum-free Mem D-Val. The supernatants were then collected, centrifugated to eliminate cell debris and filtered through 0.45 μm pore size filters to eliminate eventual residual cells. Fibroblasts and EC cultures were washed twice in serum-free medium, and after the second rinse, further incubated with the KS cells conditioned media. Viability after serum deprivation was determined by harvesting the different cell cultures after various time intervals. Cells in individual monolayer cultures were harvested and made into single-cell suspension by trypsinization. The cell suspension was then incubated for 3 min with trypan blue dye. The number of viable cells was estimated as the percentage of cells remaining negative to the blue dye staining after trypan blue exclusion. In a series of experiments we have examined cultured KS cells for the expression of Bcl-2 and Bax and compared them with EC and fibroblasts. We found that Bcl-2 protein was readily immunocytochemically detectable in the majority of KS cells (Figure 1), as was Bax, yet with a much lower staining intensity. The antigen recognized by both Bcl-2 antibodies was cytoplasmic, corresponding to the described localization of the Bcl-2 protein. Interestingly, positive staining for Bcl-2 was observed in AIDS-related KS-derived cells, as well as in cells obtained from iatrogenic and sporadic KS. The expression of Bcl-2 in KS cells was more carefully studied in KS-1, -2, and -3, all cell cultures that were isolated from the same epidemiologic setting (AIDS associated). Interestingly, the percentage of positive cells was higher in the AIDS-KS cells derived from more advanced lesions (KS-2 and -3, nodular-stage KS) (40%–55%) than in those derived from a macular lesion (KS-1, patch-stage KS) (25%–35%). Further studies are of course needed to establish a possible correlation between the histologic stage of KS and the in vitro expression of Bcl-2. Under similar conditions of culture, EC and fibroblasts demonstrated weak Bcl-2 staining. To further confirm the overexpression of Bcl-2 in KS-derived cells, immunoblot analysis was performed. As depicted in Figure 2, the Ab-1 (Bcl-2) monoclonal antibody detected a protein of ≈24 kDa corresponding with the size expected for the Bcl-2 proto-oncogene in KS-derived cells and in the malignant breast cell line MCF-7, but neither in EC nor in fibroblasts. Previous studies indicated that cell death induced by serum or growth factor deprivation occurs through apoptosis (Hase et al., 1994Hase M. Araki S. Kaji K. Hayashi H. Classification of signals for blocking apoptosis in vascular endothelial cells.J Biochem. 1994; 116: 905-909PubMed Google Scholar). The viability of the cells was examined by staining with trypan blue. As shown in Figure 3, there were considerable differences in the time course of cell death between EC, fibroblasts, and KS cells. Within 48 h of serum withdrawal, all the EC had detached from the plate, whereas the survival remained unaltered in fibroblasts and KS cells. After 15 d of serum deprivation, the majority of KS cells were alive, whereas a substantial proportion of control fibroblastic cells took the dye. KS cell viability was confirmed by the reinitiation of DNA synthesis (data not shown) and cell growth after resuspension in medium supplemented with 10% fetal calf serum. Interestingly, KS-2 and -3, which exhibited high levels of Bcl-2 immunoreactivity (>40% of positive cells) displayed a high resistance to serum deprivation (Table 1). No significant difference could be observed between AIDS and non-AIDS KS cells.Table IComparative levels of Bcl-2 expression in KS cells in respect to their resistance to serum deprivationBcl-2 immunoreactivityaBcl-2 imunoreactivity: +, 35% of positive cells.% of viable cells after 15 d of serum deprivationbMean ± SEM (n = 3).KS-1 (AIDS-related)+65 ± 8KS-2 (AIDS-related)++82 ± 10KS-3 (AIDS-related)++78 ± 10KS-4 (iatrogenic KS)++76 ± 9KS-5 (sporadic KS)++76 ± 8a Bcl-2 imunoreactivity: +, 35% of positive cells.b Mean ± SEM (n = 3). Open table in a new tab As previous studies have shown that KS cells have features of activated EC (Fiorelli et al., 1995Fiorelli V. Gendelman R. Samaniego F. Markham P.D. Ensoli B. Cytokines from activated T cells induce normal endothelial cells to acquire the phenotypic and functional features of AIDS-Kaposi's sarcoma spindle cells.J Clin Invest. 1995; 95: 1723-1734Crossref PubMed Scopus (151) Google Scholar), we next investigated the effect of conditioned media from KS cells on EC. The proliferation of EC treated with culture medium plus conditioned media from KS cells (1:2) was greater than the one of EC treated with culture medium alone (12% ± 2 vs 8 ± 2% positive cells after a 2 h pulse of Bromodeoxyuridine) (mean of triplicates; p < 0.05 by one-tailed Student t test). This is in agreement with previous studies showing that KS cell supernatants support endothelial cell growth (Ensoli et al., 1989Ensoli B. Nakamura S. Salahuddin S.Z. et al.AIDS-Kaposi's sarcoma-derived cells express cytokines with autocrine and paracrine growth effects.Science. 1989; 243: 223-226Crossref PubMed Scopus (418) Google Scholar;Albini et al., 1992Albini A. Repetto L. Carlone S. et al.Characterization of Kaposi's sarcoma-derived cell cultures from an epidemic and a classic case.Int J Oncol. 1992; 1: 723-730PubMed Google Scholar). No Bcl-2 induction could still be detected in EC treated with conditioned media from KS cells. Analogously, no Bcl-2 induction could be observed in dermal fibroblasts. Because cytokines and growth factors play an important role in KS and because some growth factors have been found to promote the survival and proliferation of EC (Kondo et al., 1994Kondo S. Yin D. Aoki T. Takahashi J.A. Morimura T. Takeuchi J. Bcl-2 gene prevents apoptosis of basic fibroblast growth-factor-deprived murine aortic endothelial cells.Exp Cell Res. 1994; 213: 428-432Crossref PubMed Scopus (61) Google Scholar), we considered the possibility that the supernatant from KS cell cultures might protect EC from cell death. KS cells supernatants still had no effect on the survival of EC cultures (Figure 4). Analogously, conditioned media from KS cells did not promote the cell survival of fibroblasts. Using an immunohistochemical approach, we have shown that KS-derived cells overexpress Bcl-2, a protooncogene known to antagonize apoptosis. This high level of Bcl-2 was in contrast with low expression of Bax, which may be considered as the counterpart of Bcl-2, promoting apoptosis. Two previous studies have shown an in vivo overexpression of Bcl-2 in KS spindle-shaped cells (Bohan Morris et al., 1996Bohan Morris C. Gendelman R. Marrogi A.J. Lu M. Lockyer J.M. Alperin-lea W. Ensoli B. Immunohistochemical detection of Bcl-2 in AIDS-associated and classical Kaposi's sarcoma.Am J Pathol. 1996; 148: 1055-1063PubMed Google Scholar;Dada et al., 1996Dada M.A. Chetty R. Biddolph S.C. Schneider J.W. Gatter K.C. The immunoexpression of bcl-2 and p53 in Kaposi's sarcoma.Histopathology. 1996; 29: 159-163Crossref PubMed Scopus (41) Google Scholar). Another study found that KS spindle cells in vivo and in vitro essentially expressed Bcl-xl with little or no Bcl-2 (Foreman et al., 1996Foreman K.E. Wrone-Smith T. Boise L.H. et al.Kaposi's sarcoma tumor cells preferentially express Bcl-xl.Am J Pathol. 1996; 149: 795-803PubMed Google Scholar). The expression of Bcl-2 was confirmed here by a western blot analysis. Interestingly, we have found high levels of Bcl-2 in AIDS-related KS-derived cells as well as in cells derived from iatrogenic and sporadic KS. These findings suggest that Bcl-2 upregulation may be important in the pathogenesis of KS and are in line with the view that the different epidemiologic forms of KS represent a uniform disease spectrum with common pathogenic denominators. Although the origin of the KS tumor cells remains unclear, it is commonly believed that they derive from EC. This is based on a variety of data including expression of similar surface antigens as well as phenotypic conversion of normal endothelial cells into spindle-shaped KS-like cells when grown in KS media (Rutgers et al., 1986Rutgers J.F. Wieczorek R. Bonetti F. Kaplan K.E. Posenett D.N. Friedman-Kien A.E. Knowles D.M. The expression of endothelial cell surface antigen by AIDS-associated Kaposi's sarcoma: evidence for a vascular endothelial cell origin.Am J Pathol. 1986; 122: 493-499PubMed Google Scholar;Fiorelli et al., 1995Fiorelli V. Gendelman R. Samaniego F. Markham P.D. Ensoli B. Cytokines from activated T cells induce normal endothelial cells to acquire the phenotypic and functional features of AIDS-Kaposi's sarcoma spindle cells.J Clin Invest. 1995; 95: 1723-1734Crossref PubMed Scopus (151) Google Scholar). In this respect, we tested the effect of supernatant from KS cells on EC; however, we detected no effect of cell-free products secreted by KS cells on either the expression of Bcl-2 by EC or their survival in serum deprivation conditions. This suggests that the overexpression of Bcl-2 in KS-derived cells is mediated by a mechanism that is not cytokine induced. Our findings suggest that the spindle cells derived from KS lesions have a survival advantage and may adequately represent the tumor cells of KS. Supporting this view, we found that KS-derived cells were more resistant than EC or fibroblasts to cell death mediated by serum deprivation. Although the constitutive expression of Bcl-2 has been proposed to be involved in cell protection from apoptosis mediated by growth factor deprivation (Garland and Halestrap, 1997Garland J.M. Halestrap A. Energy metabolism during apoptosis.J Biol Chem. 1997; 8: 4680-4688Crossref Scopus (139) Google Scholar), the mechanism of this resistance remains unclear. Future functional experiments, such as microinjection of anti-sense cDNA could probably give more information regarding the relationship between the high expression of bcl-2 and the resistant nature of KS cells to death. One possible mechanism, which is consistent with the long doubling time of KS cells reported here and in other studies (Benelli et al., 1994Benelli R. Repetto L. Carlone S. Parravicini C. Albini A. Establishment and characterization of two new Kaposi’s sacoma cell cultures from an AIDS and a non-AIDS patient.Res Virol. 1994; 145: 251-259Crossref PubMed Scopus (31) Google Scholar), is the prolongation of the G1 phase of the cell cycle, with subsequent facilitation of a prolonged metabolically dormant state from which cells recover with high efficiency (Borner, 1996Borner C. Diminished cell proliferation associated with the death-protective activity of Bcl-2.J Biol Chem. 1996; 271: 12695-12698Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). The pathogenic mechanisms leading to Bcl-2 overexpression are not fully understood. Previous studies have found Bcl-2 upregulation and consequent blocking of apoptosis by viral gene products. For example, Epstein–Barr virus may induce direct activation of Bcl-2 in infected B cell lines (Finke et al., 1992Finke J. Fritzen R. Ternes P. et al.Expression of bcl-2 in Burkitt’s lymphoma cell lines: induction by latent Epstein–Barr virus genes.Blood. 1992; 80: 459-469PubMed Google Scholar) and herpes simplex virus-2 may act cooperatively with IFN-γ in the induction of Bcl-2 and the associated survival of infected neurons (Geiger et al., 1995Geiger K.D. Gurushanthaiah D. Howes E.L. Lewandowski G.A. Reed J.C. Bloom F.G. Sarvetnick N.E. Cytokine-mediated survival from lethal herpes simplex virus infection: role of programmed neuronal death.Proc Natl Acad Sci USA. 1995; 92: 3411-3415Crossref PubMed Scopus (45) Google Scholar). Whether a similar mechanism is operative in KS is conjectural. Recently, a new herpesvirus, designated HHV-8, has been detected in all forms of KS; however, its pathogenic role remains to be determined. Interestingly,Cheng et al., 1997Cheng E.H. Nicholas J. Bellows D.S. Hayward G.S. Guo H.G. Reitz M.S. Hardwick J.M. A Bcl-2 homolog encoded by Kaposi's sarcoma-associated virus, human herpesvirus 8, inhibits apoptosis but not heterodimerize with Bax or Bak.Proc Natl Acad Sci USA. 1997; 94: 690-694Crossref PubMed Scopus (398) Google Scholar found that HHV-8 encodes a Bcl-2 homolog that inhibits apoptosis but neither homodimerizes nor heterodimerizes with other Bcl-2 family members. It seems very unlikely that this homolog could react with the antibody we used because we and others did not maintain HHV-8 in KS cell cultures (Lebbé et al., 1995Lebbé C. de Crémoux P. Rybojad M. Costa da Cunha C. Morel P. Calvo F. Kaposi's sarcoma and new herpesvirus.Lancet. 1995; 345: 1180Abstract PubMed Google Scholar;Simonart et al., 1996Simonart Th Noel J.C. Liesnard C. et al.Kaposi's sarcoma and herpesvirus 8: a word of caution.Dermatology. 1996; 193: 272Crossref PubMed Scopus (11) Google Scholar). Whether HHV-8 plays a role in the expression of bcl-2 in cultured KS cells is conjectural. If this is so, the absence of HHV-8 DNA sequences in KS-derived cells may suggest a “hit and run” mechanism. In conclusion, we have found consistent, strong immunoexpression of Bcl-2 in cells derived from cutaneous KS lesions, irrespective of epidemiologic subtype. The consequences of upregulated Bcl-2 expression in KS cells may be a prolonged spindle cell viability that, when coupled with cytokine- or viral-induced cellular proliferation, likely contributes to the pathogenesis and progression of KS. The authors would like to thank Michel Decock for his excellent technical assistance and Pr. Paul Galand for his critical review of the manuscript. This work was supported by the Erasme Foundation.