Leukemia (1999) 13, 289–294  1999 Stockton Press All rights reserved 0887-6924/99 $12.00 http://www.stockton-press.co.uk/leu Differential expression of Bcl-2 in human disorders according to proliferation status and malignancy D Puthier1, C Pellat-Deceunynck1, S Barille´ 1, N Robillard1, M-J Rapp2, N Juge-Morineau2, J-L Harousseau2, R Bataille1 and M Amiot1

1INSERM U463 Institut de Biologie, and 2De´partement d’He´matologie, Nantes, France

Multiple myeloma (MM) is a malignancy characterized by a very Bcl-2 inhibits cell cycle progression by lengthening the G1 slow proliferation of malignant plasma cells leading to their phase and consequently affects the growth rate of cell lines. accumulation within the bone marrow. This suggests that On the other hand some other reports indicate that Bcl-2 resistance to apoptosis may play a critical role both in the 6–8 pathogenesis and resistance to treatment of MM. Bcl-2 is a key favors dormancy over the cycling state. protein for the regulation of apoptosis. However, it has been The influence of Bcl-2 on the cell cycle is in agreement with shown that this protein also regulates the state of proliferation. the regulation of Bcl-2 during activation and differentiation of In the current study, we show that malignant plasma cells from B cells. Indeed, in the compartment, Bcl-2 is highly both the bone marrow and peripheral blood express high levels regulated during activation and differentiation. Proliferative of Bcl-2 and are slowly proliferating cells. In contrast, myeloma cells from extramedullary sites (ie pleural effusion, ascitis, pre-B cells express very low levels of Bcl-2 whereas at the mammary and gastric plasmacytoma) express Bcl-2 weakly mature stage, resting B cells express Bcl-2 highly. In germinal while being highly proliferative. Normal non-dividing bone mar- centers, an important proliferation of B cells (ie centroblasts) row plasma cells express high levels of Bcl-2 protein. In con- occurs while these cells weakly express Bcl-2. Then, con- trast, four highly proliferative reactive plasmacytosis express comitantly to the rescue of germinal centers from apoptosis, weak levels of Bcl-2. We conclude that there is an inverse corre- Bcl-2 is induced on B cells (ie ).9 lation between Bcl-2 expression and the proliferation rate of both normal and malignant plasma cells. These data may be In MM, the expression of Bcl-2 has been investigated only explained by the double function of Bcl-2, ie its well known by immunohistological staining indicating that most of the function as an anti-apoptotic molecule and its intriguing func- patients tested express Bcl-2.10,11 The expression of Bcl-2 has tion as an inhibitory molecule of cell proliferation. been also reported in the bone marrow plasma cells from Keywords: plasma cells; Bcl-2; apoptosis; proliferation; multiple healthy donors11 but no quantitative comparison of this myeloma; reactive plasmacytosis expression with the one observed in MM has been made. In this study, we first determined the expression of Bcl-2 in malignant plasma cells from 49 MM patients in relation to the Introduction localisation of their tumor (bone marrow, peripheral blood and extramedullary sites) using Western blotting or flow cyto- Multiple myeloma (MM) is a malignancy characterized by a metry. Then we compared this expression of Bcl-2 in malig- very slow proliferation of malignant plasma cells leading to nant plasma cells with that of normal plasma cells from bone their accumulation within the bone marrow but by their late marrow and with peripheral blood plasma cells from reactive and high proliferation in extramedullary sites. This accumu- plasmacytosis. We present the first evidence that normal and lation of malignant plasma cells within the bone marrow sug- malignant plasma cells express different levels of Bcl-2 gests that resistance to apoptosis may play a critical role in according to their capacity to proliferate. Our results are inter- the pathogenesis of MM. esting in understanding both the pathogenesis of MM and Apoptosis is essential for normal development, lymphoid some of the mechanisms leading to the resistance of MM to maturation, homeostasis and defense against viral infections.1 standard treatment. Bcl-2 was first identified in the translocation (14;18) observed in follicular . The Bcl-2 protein has been shown to confer resistance to apoptosis induced in a variety of systems Patients and methods such as growth factor withdrawal, cross link of cell surface receptors, oncogene activation, steroids or DNA damage. Bcl- Patients xL a close homologue to Bcl-2 is also able to promote cell survival.2 Both proteins are thought to protect from apoptosis Forty-nine patients with MM were studied. Their median age by heterodimerizing with proapoptotic members of the Bcl-2 was 64 years (range 35–79 years). The male-to-female ratio family, Bax and Bad.2 They also act by preventing both the was 0.56. A monoclonal serum Ig was found in 38 patients release of cytochrome C from mitochondria and the activation (IgG: 27; IgA: 11; biclonal: one patient). Ten patients had only of caspases leading to the final steps of apoptosis.3 Further- free monoclonal light chains in urine. The ␬:␭ ratio was 1.7. more, there are now several reports consistent with the notion Twenty-eight patients were previously untreated and 21 had that Bcl-2 is involved in the regulation of cell cycle leading progressive disease (ie relapse or primary treatment failure). to a cell cycle inhibitory effect.4–8 Different mechanisms by In the latter, four patients presented with secondary plasma which Bcl-2 acts on the cell cycle have been proposed.4,5 On cell leukemia (PCL) and five with extramedullary involvement the one hand, some findings suggest that overexpression of (EM) during disease progression: ascitis in patient No. 4, pleu- ral effusion in patient Nos 1 and 2, a gastric plasmacytoma for patient No. 3 and a mammary plasmacytoma in patient Correspondence: M Amiot, U463, Institut de Biologie, 9, quai Mon- No. 5. For the patients presenting an extramedullary involve- cousu 44093 Nantes cedex 01, France; Fax: 33 2 40 08 47 78 ment, no involvement either in the peripheral blood or in the Received 30 July 1998; accepted 14 October 1998 bone marrow was observed at the same time. Bcl-2 expression of normal and malignant human plasma cells D Puthier et al 290 Four cases of reactive plasmacytosis were studied. In all procedure. After two washes in PBS, mononuclear cells cases, transient polyclonal circulating plasma cells were (3 × 105 per test) were permeabilized with a commercially observed in the peripheral blood (from 7 to 50% plasma cells) available kit (Intraprep; Immunotech, Gagny, France) accord- in the context of viral infections. Diagnosis of reactive plasma- ing to the manufacturer’s instructions. Cells were then incu- cytosis was established both by cytological and flow cyto- bated with anti-␬/anti-␭ Abs for 20 min at room temperature. metry plasma cell identification (intracellular kappa and After two washes in PBS, cells were stained for 20 min with lambda staining). Bcl-2-FITC or with an IgG1-FITC control mAb in PBS 10% human AB serum. Samples were then analyzed by flow cytometry (FACSCalibur; Becton Dickinson, Le Pont de Claix, Reagents France). Apoptotic cells were excluded by gating on the light scatter region defined above (R1). Data analysis was done in B-B4 monoclonal antibody (mAb) (CD138) was a gift from Dr cIg+/side scatter gate (R2) within the R1 cell population. In all J Wijdenes (Diaclone Research, Besanc¸on, France). Bcl-x rab- cases, at least 4000 viable plasma cells were analyzed. The bit affinity purified polyclonal and Bcl-2 mAb were mean fluorescence ratio (MFR) of Bcl-2 was determined by purchased from Transduction Laboratories (Lexington, KY, dividing the geometric mean of fluorescence of Bcl-2 staining USA) and Dako (Glostrup, Denmark), respectively. by that of the IgG1 control.

Cell preparation and purification of plasma cells Labeling index of plasma cells

Mononuclear cells were isolated by Ficoll–Hypaque density The labeling index of plasma cells was measured according centrifugation from bone marrow aspirates or peripheral blood to a modification of Carayon and Bord13 using anti-bromo- from patients with MM and healthy donors. The percentage deoxyuridine (BrdU) mAb (BU5-1; Cymbus Bioscience, Euro- of plasma cells was then determined by morphology on May– medex, Souffelweyersheim, France) and analyzed by flow Grunwald-Giemsa smears and check by flow cytometry using cytometry. The percentage of plasma cells in S phase was two-color staining with anti-CD38 and B-B4 mAbs. Only determined as the number of positive cells for BrdU staining samples with more than 8% plasma cells were used for puri- within the population of ␬-or␭-positive plasma cells. Briefly, fication. The purification of myeloma cells using anti-CD138 106 cells were incubated with or without (control) 50 mM ° mAb (B-B4) was performed using a MACS separator (Miltenyi BrdU for 2 h at 37 Cin5%CO2 incubator, washed and per- Biotec, Bergisch Gladbach, Germany). Cells were first labeled meabilized overnight in PBS containing 1% PFA and 0.01% with anti-CD138 mAb then incubated with anti-IgG1 MACS Tween 20 at 4°C. After two washes, cells were treated for microbeads before to proceed to magnetic separation on MS+ 30 min at 37°C with DNAse I (Sigma, St Quentin Fallavier, separation column (Miltenyi Biotec). Only cell populations France). Cells were then washed in 0.5% Tween 20 PBS and with a purity above 99% were used. labeled using PE-coupled anti-light chains mAbs for 30 min at room temperature. Following two washes, both control cells and the incorporated BrdU cells were stained with 10 ␮l FITC- Immunoblot analysis of Bcl-2 and Bcl-x BU5-1 mAb in PBS supplemented with 0.5% Tween 20 and 20% AB serum for 45 min at room temperature. Then the cells Cells (2 × 106) were resuspended in 50 ␮l of lysis buffer were washed, resuspended in PBS, and analyzed immediately (10 mM Tris (pH 7.6) 150 mM NaCI 5 mM EDTA and 1% Triton on a FACScalibur flow cytometer. For this purpose, data X100) containing 2 mM PMSF and 2 ␮g/ml aprotinine. Lysates acquisition was performed on 50 000 cells and then on were cleared by centrifugation at 12 000 g for 30 min at 4°C. 10 000 live gated ␬/␭-positive cells when the plasmacytosis Protein concentration was measured in each cell lysate using represented less than 20% of total cells. bicinchoninic acid (BCA protein assay; Pierce, Rockford, IL, USA). Fifteen micrograms of total proteins were loaded for each blot. The proteins were separated by 12.5% SDS-PAGE Results and transferred on to PVDF membranes. Blots were developed using a chemiluminescent detection system (ECL; Boehringer Bcl-2 is expressed in the bone marrow malignant Mannheim, Meylan, France). plasma cells from all patients with MM

The expression of Bcl-2 was determined by two-color Flow cytometry analysis of Bcl-2 expression immunofluorescence or by immunoblot analysis. For detec- tion of Bcl-2 by two-color immunofluorescence, the plasma For all patients, plasma cell identification was controlled by cells were identified by the expression of cytoplasmic coexpression of CD138, CD38 and intracytoplasmic ␬/␭ in immunoglobulin (cIg). As described in Materials and methods a three-color immunofluorescence assay. Apoptotic cells and and as shown for one patient in Figure 1, apoptotic cells were cellular debris were excluded from the analysis by gating on excluded from analysis by light scatter and cIg/SSC gates. Posi- the light scatter plot (R1). Careful analysis of the three-color tive cIg cells were controlled to be viable plasma cells by their stainings revealed that Ͼ99% of either ␬-or␭-positive cells expression of CD138.14 The immunoblot analysis required a were viable plasma cells that coexpressed CD38 bright and highly purified population of malignant plasma cells. Also, CD138. The expression of CD138 has been shown to be only samples presenting a plasmacytosis greater than 8% were restricted to viable plasma cells but lost on apoptotic plasma purified by an immunomagnetic method leading to a purity cells.12 To study Bcl-2 expression we performed a double of at least 99%. In four patients, both methods were performed immunostaining, cytoplasmic light chain immunoglobulin and gave similar results. This allowed us to study Bcl-2 (cIg) (anti-␬, anti-␭) vs Bcl-2 using an intracytoplasmic staining expression independently of the method of analysis used Bcl-2 expression of normal and malignant human plasma cells D Puthier et al 291

Figure 1 Triple immunostaining showing the identification of plasma cells. Two gates were used to exclude apoptotic cells: light scatter gate is shown on cytogram (a), cIg+/SSC gate is shown on cytogram (b). Analysis of cIg-positive cells (gated on R2 within the R1 population) demon- strates that these plasma cells co-express CD138 and CD38 (c).

(patient No. 6 is given as an example in Figure 2). The presence of Bcl-xL protein in all the samples tested but the expression of Bcl-2 on bone marrow plasma cells was deter- absence of the short form of Bcl-x, Bcl-xS. mined in 40 patients with MM. Representative immunoblots of eight patients and cytograms of two patients are outlined in Figure 2. The results show that all malignant plasma cells Weak expression of Bcl-2 in malignant plasma cells from bone marrow express Bcl-2, and flow cytometry analysis from extramedullary sites but not from peripheral allowed quantification of Bcl-2 expression with a mean fluor- blood escence ratio (MFR) of 11.2 ± 5. Moreover, by Western blot analysis, anti-Bcl-x Abs were used as a control showing the Next, we compared Bcl-2 expression in malignant plasma cells from the peripheral blood of four patients with secondary PCL and from extramedullary sites of five patients. In the peri- pheral blood, Bcl-2 expression is similar to that observed in the bone marrow (MFR of Bcl-2 was 8.2 for one patient) (Figure 3b). In contrast, the Bcl-2 expression is weak in all the five extramedullary sites tested. Indeed, myeloma cells from the two pleural effusions (EM Nos 1 and 2) as well as from the gastric plasmacytoma (EM No. 3) presented a low level of Bcl-2 by Western blotting (Figure 3a). Furthermore, in mye- loma cells from the ascitis, Bcl-2 is undetectable by Western blotting and the MFR in flow cytometry is 1.7. Finally, in extramedullary sites the MFR determined by flow cytometry ranged from 1.7 to 4.6 (2.9 ± 1.2). Of interest, the weak expression of Bcl-2 is not related to an absence of IL-6 in extramedullary sites since high levels of IL-6 are found in both ascitis and pleural effusion (0.5, 17, 100 ng/ml; unpublished personal data) In conclusion, we found a high expression of Bcl-2 in all myeloma cells expanding within the bone marrow or circulat- ing transiently in the peripheral blood but a weak expression of Bcl-2 in myeloma cells within extramedullary sites.

Plasma cells from normal bone marrow highly express Bcl-2 whereas plasma cells from reactive plasmacytosis express it very weakly

In order to compare Bcl-2 expression in myeloma cells to that of normal plasma cells, we performed flow cytometry analysis of the bone marrow of five healthy donors. As shown in Figure 4a, the whole population of normal plasma cells express high levels of Bcl-2 (MFR 10.5 ± 3). We next looked

Figure 2 (a) Immunoblot analysis of Bcl-2 and Bcl-xL protein at the expression of Bcl-2 in reactive plasmacytosis. Reactive expression in purified CD138+ myeloma cells from bone marrow plasmacytosis was observed during the course of infectious (BM). The blots were probed with antibodies specific for Bcl-2 and diseases. These plasma cells were polyclonal, very proliferat- Bcl-x simultaneously. (b) Flow cytometry measurement of Bcl-2 ive and short-lived (personal data). By flow cytometry, plasma expression in myeloma cells from two bone marrow samples. Cyto- cells from these cases were found to express very weak levels grams show the cellular population within the light scatter gate. ± Analysis of Bcl-2 expression was done by gating on cIg+ cells within of Bcl-2 (MFR 2.7 0.4) (Figure 4a). Moreover, in two cases, the R1 population. The mean fluorescence ration (MFR) of Bcl-2 in the percentage of plasma cells allowed us to purify them and myeloma cells is given. to perform Western blotting analysis. As shown in Figure 4b, Bcl-2 expression of normal and malignant human plasma cells D Puthier et al 292

Figure 4 (a) Immunoblot analysis of Bcl-2 and Bcl-xL protein expression in purified CD138+ myeloma cells from two reactive plas- macytosis (RPC). (b) Flow cytometry measurement of Bcl-2 expression in plasma cells from one RPC sample and in normal plasma cells from the bone marrow (NPC). Bcl-2 expression is shown in the gated cIg+ cell population because of the very weak percentage of plasma cells in the normal bone marrow (Ͻ1%). Data are presented as a frequency histogram with fluorescence intensity on the horizontal axis and rela- tive cell number on the vertical axis. The dotted line represents stain- ing with IgG1 control Ab. The solid line represents bcl-2 staining.

Figure 3 (a) Immunoblot analysis of Bcl-2 and Bcl-xL protein expression in purified CD138+ myeloma cells from plasma cell leuke- mia (PCL) and extramedullary sites (EM). The blots were probed with antibodies specific for Bcl-2 and Bcl-x simultaneously. (b) Flow cyto- metry measurement of Bcl-2 expression in myeloma cells from one PCL sample and one extramedullary sample. Cytograms show the cellular population within the light scatter gate. Analysis of Bcl-2 expression was done by gating on cIg+ cells within the R1 population. The mean fluorescence ratio (MFR) of Bcl-2 in myeloma cells is given.

by Western blotting, Bcl-2 is almost undetectable in these

plasma cells whereas again Bcl-xL is highly expressed.

Relationship of Bcl-2 expression in plasma cells to their proliferative rates and malignant status Figure 5 (a) Comparison of Bcl-2 mean fluorescence ratio (MFR) The analysis of Bcl-2 expression indicates that high levels are in normal and malignant plasma cells (PCs). (b) Comparison of labe- found in both normal (MFR 10.5 ± 3) and malignant (MFR ling index (LI) of low Bcl-2 plasma cells (MFR Ͻ5) and high Bcl-2 11.2 ± 5) plasma cells from the bone marrow whereas low lev- plasma cells (MFR у6). LI is defined as the number of BrdU+ cells els are found in malignant plasma cells from extramedullary within the plasma cell population. Statistical analysis was done using sites (MFR 2.9 ± 1.2) and peripheral blood plasma cells from the Wilcoxon rank sum test. BM, bone marrow; PB, peripheral blood. reactive plasmacytosis (MFR 2.7 ± 0.4) (Figure 5a). Because it has been proposed that Bcl-2 is involved in cell cycle regu- trast, both malignant plasma cells from extramedullary sites lation, we studied the proliferation status of these different and plasma cells from reactive plasmacytosis have a high plasma cell populations. The labeling index of plasma cells labeling index (31% ± 11) and express very low levels of Bcl- was measured by flow cytometry, and the results are 2(P Ͻ 0.01). presented in Figure 5b. These results clearly demonstrated that there is an inverse correlation between the level of Bcl-2 and the proliferation rate regardless of their benign or malignant Discussion status. Indeed, we showed that malignant plasma cells from bone marrow or peripheral blood have a very weak labeling This report clearly demonstrates that myeloma cells express index (1.34% ± 0.9) but express high levels of Bcl-2. In con- high levels of Bcl-2 in vivo when they are localized in the Bcl-2 expression of normal and malignant human plasma cells D Puthier et al 293 bone marrow or transiently in the peripheral blood whereas proteins for limiting amounts of Raf-1 as has already been they express it very weakly when they are expanding rapidly proposed by Wang et al.22 outside the bone marrow. Furthermore, by performing quanti- Although the events that regulate the expression of Bcl-2 are tative flow cytometric measurements of Bcl-2 on cIg+ plasma poorly understood, we can assume that: (1) the bone marrow cells we showed that high Bcl-2 expression in bone marrow environment is not involved in the regulation of Bcl-2 since myeloma cells reflects that found in their normal counterparts peripheral blood and bone marrow malignant plasma cells in the bone marrow environment. Finally, we demonstrated express similar level of Bcl-2; (2) IL-6 is not responsible for that Bcl-2 expression is inversely related to the proliferation the observed Bcl-2 regulation because high IL-6 levels are state of the plasma cells regardless of their malignant status. found both in MM and reactive plasmacytosis but also in Bcl-2 expression has already been documented in human pleural effusions and ascitis. plasma cells when homing within the bone marrow,11 but we Our findings are in agreement with the observation of have demonstrated here for the first time that malignant Winter et al23 in intermediate- and high-grade non-Hodgkin’s plasma cells express a normal level of Bcl-2, suggesting that lymphoma, showing that there is an inverse relationship this protein is not likely to be responsible for the emergence between Bcl-2 expression and proliferative activity. Moreover, of malignant clones. In fact, this high expression is consistent the overexpression of Bcl-2 has often been associated with with the notion that such plasma cells are non-dividing, fully both the indolent nature of lymphoma and a better prognosis differentiated and long-lived cells.15,16 However, the in some other types of cancer such as breast cancer and non- extremely long lifespan of plasma cells as recently demon- small cell lung carcinoma.24,25 Moreover, in MM, which is strated by Slifka et al17 could provide an opportunity for mye- characterized by a very slow evolution, it has been reported loma cells to acquire further secondary genetic changes. that there is a trend toward an inverse correlation between The role of Bcl-2 as an anti-apoptotic molecule is now well Bcl-2 expression in bone marrow myeloma cells and known. However, substantial progress have been made in elu- patients survival.26 cidating its role as a regulatory molecule of the cell cycle.4–8 Since most anti-cancer drugs target dividing cells, the high To address such a question we analyzed the percentage of expression of Bcl-2 in bone marrow myeloma cells can par- plasma cells in S phase and demonstrated that we could dis- ticipate in the resistance to treatment observed in MM. tinguish slowly proliferative plasma cells highly expressing Although Bcl-2 can contribute both in resistance to apoptosis Bcl-2 (Bcl-2high) from those highly proliferative weakly and chemoresistance, therapeutical intervention based on a expressing Bcl-2 (Bcl-2low). Indeed, both normal and malig- down-regulation of Bcl-2 alone should be considered care- nant plasma cells in the bone marrow are found as Bcl-2high fully. Indeed, our results suggest that such an intervention whereas malignant plasma cells expanding in extramedullary could favor S phase entry of plasma cells. Consequently, it sites and peripheral blood plasma cells from reactive plasma- will be judicious to use Bcl-2 in combination with S cytosis are found as Bcl-2low. These observations are in agree- and M phase-specific drugs already known to be efficient in ment with the recent demonstration that Bcl-2 can retard tran- the treatment of MM. sition of B cells from the quiescent to the cycling state regardless of the differentiation stage.18 In this way, the high expression of Bcl-2 in both normal and malignant bone mar- Acknowledgements row plasma cells may explain their slow proliferation whereas a weak expression of this protein may contribute in explaining We thank M Etrillard for technical assistance and Laurence the high turn-over observed in malignant plasma cells from Legros (Department of Internal Medecine, Nice) for providing extramedullary sites and in plasma cells from reactive plasma- us with a plasmacytoma sample. This work was supported by cytosis. Whereas this high proliferation rate in extramedullary grants from the Ligue Nationale contre le Cancer et la Ligue contre le Cancer de Loire Atlantique. sites has already been documented19 little is known about this in reactive plasmacytosis.20 However, the high rate of prolifer- ation observed in reactive plasmacytosis is consistent with their fast appearance and disappearance during the clinical References course of infectious diseases. Moreover, in extramedullary 1 White E. Life, death, and the pursuit of apoptosis. Genes Dev sites and in reactive plasmacytosis we observed elevated 1996; 10: 1–15. serum values of lactic dehydrogenase (results not shown) 2 Yang E, Korsmeyer S. Molecular thanatopsis: a discourse on the reflecting an important cells loss and suggesting a weaker Bcl-2 family and cell death. Blood 1996; 88: 386–401. resistance of these plasma cells to apoptosis. Again, one poss- 3 Reed JC. Cytochrome C: can’t live with it – can’t live without it. Cell 1997; 91: 559–562. ible explanation for these observations could be the weak 21 4 Mazel S, Burtrum D, Petrie HT. Regulation of cell division cycle expression of Bcl-2. In a recent work, Ogata et al suggested progression by bcl-2 expression: a potential mechanism for inhi- that myeloma cells from extramedullary sites can also be dis- bition of programmed cell death. J Exp Med 1996; 183: 2219– tinguished from bone marrow and peripheral blood myeloma 2226. cells by their ability to proliferate in response to IL-6, this 5 Borner C. Diminished cell proliferation associated with the death response being associated with the Ras-dependent MAP kin- protective activity of Bcl-2. J Biol Chem 1996; 271: 12695–12698. 6 Vairo G, Innes KM, Adams JM. Bcl-2 has a cell cycle inhibitory ase cascade activation. Again, these findings underline the function separable from its enhancement of cell survival. Onco- high capacity of myeloma cells to proliferate in extramedul- gene 1996; 13: 1511–1519. lary sites. Moreover, it has been shown that Bcl-2 can interact 7 O’Reilly LA, Huang DCS, Strasser A. The cell death inhibitor Bcl- with Raf-1, this interaction being important for the anti-apop- 2 and Its homologues influence control of cell cycle entry. EMBO totic function of Bcl-2.22 Also, Raf-1 appears as a key mol- J 1996; 15: 6979–6990. 8 Huang D, O’Reilly L, Strasser A, Cory S. The anti-apoptosis func- ecule for both the apoptotic and proliferative pathways. Alto- tion of Bcl-2 can be genetically separated from its inhibitory effect gether, these observations lead us to speculate that on cell cycle entry. EMBO J 1997; 16: 4628–4638. competition could exist between Ras and Bcl-2 family 9 Liu Y, Mason D, Johnson G, Abbot S, Gregory C, Hardie D, Gor- Bcl-2 expression of normal and malignant human plasma cells D Puthier et al 294 don J, Maclennan I. cells express bcl-2 protein turnover of developing B lymphocytes in vivo. J Immunol 1997; after activation by signals which prevent their entry into apoptosis. 159: 2301–2311. Eur J Immunol 1991; 21: 1905–1910. 19 Drewinko R, Alexanian H, Boyer B, Barlogie B, Rubinow S. The 10 Hamilton M, Barker H, Ball J, Drew M, Abbot S, Franklin I. Nor- growth fraction of human myeloma cells. Blood 1981; 57: 333– mal and neoplastic human plasma cells express bcl-2 . 338. Leukemia 1991; 5: 768–771. 20 Jourdan M, Bataille R, Seguin J, Xg Z, Chaptal P, Klein B. Constitut- 11 Pettersson M, Jernberg-Wiklund H, Larsson L, Sundstro¨m C, Givo ive production of interleukin-6 and immunologic features in car- I, Tsujimoto Y, Nilsson K. Expression of the bcl-2 gene in human diac myxomas. Arthris Rheum 1990; 33: 398–402. multiple myeloma cell lines and normal plasma cells. Blood 1992; 21 Ogata A, Chauchan D, Teoh G, Treon S, Urashima M, Schlossman 79: 495–502 R, Anderson K. IL-6 triggers myeloma cell growth via the ras- 12 Jourdan M, Ferlan M, Legouffe E, Horvathova M, Liautard J, Rossi dependent mitogen-activated protein kinase cascade. J Immunol JF, Widjenes J, Klein B. The myeloma cell antigen syndecan-1 is 1997; 159: 2212–2221. lost by apoptotic myeloma cells. Br J Haematol 1998; 100: 22 Wang H, Rapp U, Reed J. Bcl-2 targets the protein kinase Raf-1 637–646. to mitochondria. Cell 1996; 87: 629–638. 23 Winter J, Anderson J, Reed JC, Krajewski S, Variakojis D, Bauer 13 Carayon P, Bord A. Identification of DNA-replicating lymphocytes KD, Fisher RI, Gordon LI, Oken MM, Jiang S, Jeffries D, Domer subsets using a new method to label the bromo-deoxyurudine P. Bcl-2 expression correlates with lower proliferative activity in incorporated into the DNA. J Immunol Meth 1992; 147: 225–230. the intermediate- and high-grade non-Hodgkin’s : an 14 Pellat-Deceunynck C, Bataille R, Robillard N, Harrousseau J-L, Eastern Cooperative Oncology Group and Southern Oncology Rapp M-J, Juge-Morineau N, Widjenes J, Amiot M. Expression of Group Cooperative Laboratory Study. Blood 1998; 91: 1391– CD28 and CD40 in human myeloma cells: a comparative study 1398. with normal plasma cells. Blood 1994; 84: 2597–2603. 24 Lipponen P, Pietilainen T, Kosma V, Aaltomaa S, Eskelinen M, 15 Barlogie B, Epstein J, Selvanayagam P, Alexanian R. Plasma cell Syrjanen K. Apoptosis suppressing protein Bcl-2 is expressed in myeloma – new biological insights and advances therapy. Blood well-differentiated breast carcinoma with favourable prognosis. J 1989; 73: 865–879. Pathol 1995; 177: 49–55. 16 Hallek M, Bergsagel L, Anderson KC. Multiple myeloma: increas- 25 Pezzela F, Turley H, Kuzu I, Tungekar M, Dunnill M, Pierce C, ing evidence for a multistep transformation process. Blood 1998; Harris A, Gatter KC, Masson DY. Bcl-2 protein in non-small-cell 91: 3–21. lung carcinoma. New Engl J Med 1993; 329: 690–694. 17 Slifka MK, Antia R, Whitmire JK, Ahmed R. Humoral immunity 26 Ong F, Nieuwkoop JV, Groot-Swings GD, Hermans J, Harvey M, due to long-lived plasma cells. Immunity 1998; 8: 363–372. Kluin P, Kluin-Nelemans J. Bcl-2 protein expression is not related 18 O’Reilly L, Harris A, Tarlinton D, Corcoran L, Strasser A. to short survival in multiple myeloma. Leukemia 1995; 9: Expression of a bcl-2 transgene reduces proliferation and slows 1282–1284.