sign in sign up
<<
Home , Burkitt lymphoma

(2004) 18, 491–498 & 2004 Nature Publishing Group All rights reserved 0887-6924/04 $25.00 www.nature.com/leu Quantitative analysis of bcl-2 expression in normal and leukemic human B-cell differentiation

P Menendez1,2, A Vargas3, C Bueno1,2, S Barrena1,2, J Almeida1,2 M de Santiago1,2,ALo´pez1,2, S Roa2, JF San Miguel2,4 and A Orfao1,2

1Servicio General de Citometrı´a, Universidad de Salamanca, Salamanca, Spain; 2Departamento de Medicina and Centro de Investigacio´n del Ca´ncer, Universidad de Salamanca, Salamanca, Spain; 3Servicio de Inmunodiagno´stico, Departamento de Patologı´a Clı´nica, Hospital Nacional Edgardo Rebagliati Martins, Lima, Peru´, ; and 4Servicio de Hematologı´a, Hospital Universitario, Salamanca, Spain

Lack of has been linked to prolonged survival of results in the overexpression of the bcl-2 protein and it malignant B cells expressing bcl-2. The aim of the present represented the first clear example of a common step in study was to analyze the amount of bcl-2 protein expressed 9,10 along normal human B-cell maturation and to establish the oncogenesis mediated by decreased cell death. Currently, frequency of aberrant bcl-2 expression in B-cell malignancies. the exact antiapoptotic pathways through which bcl-2 exerts its In normal bone marrow (n ¼ 11), bcl-2 expression obtained by role are only partially understood, involving decreased mito- quantitative multiparametric flow cytometry was highly vari- chondrial release of cytochrome c, which in turn is required for þ À able: very low in both CD34 and CD34 B-cell precursors, high the activation of procaspase-9 and the subsequent initiation of in mature B-lymphocytes and very high in plasma cells. Bcl-2 the apoptotic cascade.11 expression of mature B-lymphocytes from peripheral blood Despite being a characteristic feature of FL bearing t(14;18),7 (n ¼ 10), spleen (n ¼ 8) and lymph node (n ¼ 5) was significantly higher (P 0.02) in CD23À as compared to CD23 þ B cells, overexpression of the bcl-2 protein is not specific of this subtype o 12 independent of the type of tissue analyzed. Upon comparison of non-Hodgkin’s (NHL). High amounts of bcl-2 with normal human B-cell maturation, bcl-2 expression in protein have been reported in other mature peripheral B-cell neoplastic B cells from 144 patients was found to be aberrant neoplasms such as B-cell chronic lymphocytic leukemia (B- in 66% of the cases, usually corresponding to bcl-2 over- CLL),13,14 diffuse large B-cell lymphoma (DLCL),15,16 multiple expression (63%). (FL) carrying t(14;18) myeloma (MM),17–19 of undetermined and MALT lymphoma were the only diagnostic groups con- 19 stantly showing overexpression of bcl-2. Bcl-2 overexpression significance (MGUS) and in cases of B-cell precursor acute 20 was also frequently found in precursor B-acute lymphoblastic lymphoblastic leukemia (BCP-ALL). Such apparent lack of leukemia (84%), typical (77%) and atypical (75%) B-cell chronic specificity of bcl-2 overexpression has contributed to the general lymphocytic leukemia, prolymphocytic leukemia (two of three belief that its assessment is of limited utility for the diagnostic À cases), (55%), but not in t(14;18) FL, subclassification of B-cell malignancies. However, a careful splenic marginal zone lymphoma, and multi- ple myeloma. analysis of the literature shows that information on the levels of Leukemia (2004) 18, 491–498. doi:10.1038/sj.leu.2403231 bcl-2 expressed during normal B-cell differentiation is scanty; Published online 15 January 2004 even more, to the best of our knowledge, no study has been Keywords: bcl-2; flow cytometry; human B-cell differentiation; reported in which quantitative evaluation of bcl-2 expression B-cell leukemic/B-cell lymphoma during normal B-cell maturation has been taken into account prior to establishing whether the amount of this protein in neoplastic cells from different B-cell disorders was abnormally increased or not, as compared to the normal B cells. Introduction The aim of the present study is to analyze quantitatively the amount of bcl-2 protein expressed along the different stages of Control of cell survival is essential in tissues with high cell normal human B-cell maturation, in order to establish the turnover, such as the lympho-hematopoietic system, in order to incidence of aberrant bcl-2 expression in a large series of 144 maintain the normal tissue homeostasis; disruption of the patients suffering from different B-lineage hematopoietic malig- balance between cell production and death is a critical step in nancies. tumorigenesis.1,2 Under physiological conditions, cell death occurs in a programmed way through apoptosis.1–5 The bcl-2 proto- and its homologues play a key role Material and methods in regulating physiological cell death.1,6 Bcl-2 was initially identified by its involvement in the translocation Controls and patients t(14;18)(q32;q21) – a hallmark of follicular lymphoma (FL)7,8 – 9,10 where it acts as a key negative regulator of apoptosis. In For the evaluation of bcl-2 expression along the normal B-cell t(14;18), the bcl-2 gene from is translocated to maturation, a total of 34 normal samples were studied. These the long arm of , into the vicinity of the corresponded to 10 peripheral blood (PB) and 11 bone marrow immunoglobulin heavy-chain (IgH) gene, where its expression 7,8 (BM) samples obtained from an identical number of adult comes under the control of the IgH gene promoter. This healthy volunteers and to eight reactive spleen plus five reactive lymph node samples. The mean age of normal donors was Correspondence: Professor A Orfao, Servicio General de Citometrı´a, 46726 years (range: 23–61 years). Hospital Universitario de Salamanca, Paseo San Vicente 58-182, A total of 47 BM and 97 PB samples from 144 newly Salamanca 37007, Spain; Fax: þ 34 923 294624; E-mail: [email protected] diagnosed untreated patients with B-cell malignancies were Received 7 April 2003; accepted 13 October 2003; Published online studied. Diagnosis of the different B-cell malignances was based 15 January 2004 on clinical, morphologic, immunophenotypic and molecular Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 492 criteria, according to the WHO classification.21,22 According to reactive spleen and reactive lymph node samples (Figure 1). In diagnosis, patients were distributed as follows: B-cell precursor samples from patients with malignant B-cell disorders, bcl-2 ALL, 12 cases (mean age: 24720 years); B-CLL, 64 cases – 56 expression was specifically evaluated for the neoplastic B cells. typical (typ) and eight atypical (atyp) CLL (67714 and 7578 Bcl-2 expression was evaluated as the mean fluorescence years, respectively); prolymphocytic leukemia (PLL), three cases intensity (MFI) obtained after subtracting from the MFI values of (61730 years); (HCL), two cases (55721 bcl-2-stained samples the MFI values obtained for the corre- years); mantle cell lymphoma (MCL), 13 cases (67716 years); sponding isotypic-negative controls for each specific normal and splenic marginal zone lymphoma (SMZL), five cases (68724 neoplastic B-cell subset analyzed. In order to normalize the years); MALT lymphoma, three cases (73718 years); FL, 12 measurements of bcl-2-associated fluorescence, a ratio between cases (6479 years); DLCL, six cases (65721 years); Burkitt the bcl-2 MFI value obtained for each B-cell population and the lymphoma, three cases (44730 years); and MM, 21 cases bcl-2 MFI found for the normal resting T-lymphocytes present in (67710 years). In all cases, informed consent according to the the same sample was established in each sample (bcl-2 B/T Ethics Committee of the University Hospital of Salamanca ratio). For this purpose, resting T cells were identified as those (Spain) was given prior to entering the study. events displaying an FSClow/SSClow pattern typical of mature lymphocytes and that were bcl-2hi, CD38À, CD19À and CD23À.26 Flow cytometry immunophenotypic studies

Both normal and neoplastic PB and BM samples were collected in tubes containing K3 EDTA as an anticoagulant. BM samples were immediately diluted 1/1 (vol/vol) in phosphate-buffered Statistical analyses saline (PBS). Single-cell suspensions were obtained from spleen and lymph node tissues, by mechanical separation according to The mean values and their standard deviation as well as median, well-established procedures.23 25th and 75th percentiles and range were calculated for each Bcl-2 expression was explored in both the normal and variable. In order to determine the statistical significance of the malignant B-cell subpopulations, by combined immunofluores- differences observed between groups, either the Mann–Whitney cence stainings directed against cell surface and intracytoplas- U and Kruskal–Wallis tests or the Wilcoxon and Friedman tests mic proteins,24 using the following four-color combinations of were used for unrelated and related variables, respectively. monoclonal (MoAbs) directly conjugated with Statistical analyses were performed with the SPSS 9.0 software fluorescein isothiocyanate (FITC), phycoerythrin (PE), PE-cya- program (SPSS Inc, Chicago, IL, USA). P-values p0.05 were nine 5 (PECy5) and allophycocyanine (APC): bcl-2-FITC (clone considered to be statistically significant. 124, DAKO cytomation, Glostrup, Denmark), either CD34-PE (clone HPCA2, Becton Dickinson Biosciences (BDB), San Jose´, CA, USA) – in BM samples – or CD23-PE (clone EBVCS-5, BDB) – in PB, spleen and lymph node samples, CD19-PECy5 (clone SJ25C1, Caltag Laboratories, San Francisco, CA, USA) and Results CD38-APC (clone HB7, BDB). Cell fixation and permeabiliza- tion prior to bcl-2 protein staining was performed using the Bcl-2 expression during normal B-cell maturation Fix&Perm reagent (Caltag Laboratories) according to the methods that have been previously described in detail.25 For Intracellular expression of the bcl-2 protein along the normal each sample analyzed, a second tube containing cells stained BM B-cell maturation as well as in the major subsets of CD23À with the same combination of MoAb directed against the cell and CD23 þ B cells from PB, spleen and lymph node is surface antigens plus an FITC-conjugated, isotype-matched illustrated in Figure 1 and summarized in Table 1. As may be nonspecific mouse immunoglobulin was analyzed in parallel seen in this table, the most immature CD34 þ BM B-cell as negative control. progenitors showed dim bcl-2 expression (bcl-2 MFI of 1176; In all cases, either a two- or three-step data acquisition was bcl-2 B/T ratio of 0.4570.3). The maturation of CD34 þ / performed in a FACSCalibur flow cytometer (BDB) using the CD38 þþ/CD19 þ progenitors into CD34À/CD38 þþ/CD19 þ CellQuest software program (BDB). In the first step, information B-cell precursors was associated with a slight (P40.05) was stored in a minimum of 3 Â 104 total nucleated cells present decreased expression of bcl-2 (bcl-2 MFI 674 and bcl-2 B/T in the sample; this information was used to assess the relative ratio of 0.2770.3). As these cells differentiated into CD34À/ distribution of the different B-cell subpopulations present in CD38 þ /À/CD19 þ B cells, a significant (Po0.005) increase in each normal and neoplastic sample. In the second step, bcl-2 expression was observed (bcl-2 MFI of 44718, and bcl-2 acquisition through an electronic ‘live gate’, drawn on the B/T ratio of 1.6870.6). Differentiated BM plasma cells CD19 þ region in which B cells are located, was carried out in (CD38 þþþ/CD19À/ þ /CD34À) showed the highest bcl-2 levels all samples analyzed; in BM samples, a further third acquisition (bcl-2 MFI of 161748 and bcl-2 B/T ratio of 5.9473). step through a ‘live gate’ drawn on the CD38 þþþ region was In normal PB, CD23 þ B-lymphocytes showed a significantly carried out to analyze plasma cells specifically. In these two (P ¼ 0.008) lower expression of cytoplasmic bcl-2 than CD23À latter steps, information on a minimum of 104 CD19 þ B cells or B-cells (bcl-2 MFI of 40712 vs 57718 and bcl-2 B/T ratio of CD38 þþþ plasma cells was acquired, respectively. For data 0.9670.27 vs 1.2270.35). Likewise, in both reactive spleen analysis, the Paint-A-Gate PRO software (BDB) was used. The and lymph node samples, CD23À B-lymphocytes displayed a expression of cytoplasmic bcl-2 was assessed in the following higher reactivity for bcl-2 than CD23 þ B cells (P ¼ 0.01 and subpopulations of CD19 þ B cells and/or CD38 þþþ plasma 0.02, respectively) (Table 1). Interestingly, no statistically cells: CD34 þ /CD38 þþ/CD19 þ , CD34À/CD38 þþ/CD19 þ , significant differences were found among CD23 þ or among CD34À/CD38À/ þ /CD19 þ and CD34À/CD38 þþþ/CD19À/ þ CD23À B-lymphocytes from PB, spleen and lymph nodes in BM or CD23À/CD19 þ and CD23 þ /CD19 þ in normal PB, (Table 1).

Leukemia Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 493

Figure 1 Representative dot plots and histograms showing the gating strategy used for the immunophenotypic identification of the different normal CD19 þ B-cell subsets analyzed, and the evaluation of their bcl-2 levels in normal BM (a–d), reactive spleen (e–g), reactive lymph node (h– j) and normal PB (k–m). Total CD19 þ /SSClow cells (painted in black in dot plots A, E, H and K) and CD38 þþþ plasma cells (painted in black in dot plot B), are displayed in dot plots C, F, I and L for further subsetting. Histograms D, G, J and M show bcl-2 expression for each B-cell population identified including plasma cells. In BM (c), four different B-cell subsets were identified: CD34 þ /CD38 þþ/CD19 þ B-cell progenitors (highlighted gray events), CD34À/CD38 þþ/CD19 þ B-cell precursors (blue events), CD34À/CD38 þ /À/CD19 þþ B-cell lymphocytes (black events) and CD34À/CD38 þþþ/CD19 þ plasma cells (red events). In spleen, lymph node and PB samples (f, i and l, respectively) two different subsets were analyzed: CD23À/CD19 þ (gray dots) and CD23 þ /CD19 þ mature B-lymphocytes (black dots). Bcl-2 expression of each of these subpopulations is displayed in histograms D (BM), G (spleen), J (lymph node) and M (PB).

Bcl-2 expression in neoplastic cells from B-cell which could be either CD23 þ or CD23À; and (4) MM patients malignancies showing BM infiltration by neoplastic plasma cells. Neoplastic cells from all cases analyzed, including those from According to their similarities to normal B-cell subsets, Burkitt lymphoma, showed the expression of the bcl-2 protein neoplastic B-cell disorders were classified into four major above the negative control values. As may be seen in Figure 2a groups: (1) those originating in CD34 þ BM B-cell precursors in BCP-ALL cases (n ¼ 12), low amounts of intracellular bcl-2 (BCP-ALL); (2) mature peripheral B-cell neoplasms of either were detected (bcl-2 MFI of 38714). After normalization of bcl- CD23 þ (B-CLL) or CD23À (PLL) mature PB B cells; (3) mature B- 2 measurements, CD34 þ BCP-ALL cases displayed a signifi- cell neoplasms deriving from mature B cells from lymphoid cantly higher bcl-2 B/T ratio than normal CD34 þ B-cell tissues such as the spleen (HCL and SMZL) or the lymph node precursors (0.9670.23 vs 0.4570.3, Po0.001) (Figure2b and (FL, MCL, MALT lymphoma, DLCL and Burkitt lymphoma), 2c and Table 1).

Leukemia Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 494 Table 1 Amount of cytoplasmic expression of the Bcl-2 protein during normal B-cell maturation

Tissue type (no. of B-cell population/phenotype Frequency (% of Bcl-2 Bcl-2# B/T cases) nucleated cells) MFI ratio

Bone marrow (n ¼ 11) B-cell CD19+/CD34+/CD38++ 0.0570.05 (0.01–0.2) 10.876 0.4570.3 progenitors (7–16) (0.15–0.75) B-cell precursors CD19+/CD34À/CD38++ 272.5 (0.34–8.5) 6.174 0.2770.3(0– (3–21) 0.57) B-lymphocytes CD19+/CD34À/CD38+/À 1.970.8 (0.2–3.1) 44718a 1.6870.66a,e (22–61) (1.08–2.17) Plasma cells CD19+/CD34À/CD38+++ 0.2370.11 (0.05–0.43) 161748a 5.9473a (44–245) (1.67–16)

Peripheral blood (n ¼ 10) CD23À B- CD19+/CD23À/CD38+/À 0.970.4 (0.4–2.2) 57719b 1.2270.35b,e lymphocytes (34–78) (0.52–1.92) CD23+ B- CD19+/CD23+/CD38+/À 0.770.3 (0.3–1.8) 40712b 0.9670.27b lymphocytes (26–51) (0.42–1.5)

Spleen (n ¼ 8) CD23À B- CD19+/CD23À/CD38+/À 26719 (18–37) 49731c 1.1870.5c,e lymphocytes (30–108) (0.81–1.67) CD23+ B- CD19+/CD23+/CD38+/À 1179 (3–28) 39712c 0.9470.3c lymphocytes (29–92) (0.6–1.19)

Lymph node (n ¼ 5) CD23À B- CD19+/CD23À/CD38+/À 31717 (13–43) 58731d 1.4270.3d lymphocytes (46–70) (0.69–1.81) CD23+ B- CD19+/CD23+/CD38+/À 29721 (6–38) 36732d 0.8970.21d lymphocytes (27–66) (0.75–1.29)

Results are expressed as MFI7s.d. MFI ¼ mean fluorescence intensity. #Ratio between the bcl-2 MFI of each B-cell subset and the bcl-2 MFI of normal resting T-lymphocytes present in the same sample. The range is shown between parentheses. Statistically significant differences were observed between CD23+ and CD23À B-lymphocytes present in PB (bP ¼ 0.008), spleen (cP ¼ 0.01) and lymph node (dP ¼ 0.02). Statistically significant differences were also observed between BM and either PB or spleen B-cells for the bcl-2 B/T ratio (ePo0.006). aPo0.005 between this B-cell population and the other B-cell subsets present in normal BM.

The analysis of bcl-2 expression in mature B-cell neoplasms, the normal B cells, two- to five-fold higher mean levels of bcl-2 including tissue , showed the presence of variable expression were found in all the subgroups analyzed, except but constantly higher levels of bcl-2 on the neoplastic B-cells, as HCL and Burkitt lymphoma (Figure 2c). The highest values compared to normal B cells when the bcl-2 MFI (Po0.04) was obtained for the ratio between bcl-2 expression in neoplastic B considered; Burkitt lymphoma cases were the only exception, cells and in normal B cells were found among patients with since they showed the lowest bcl-2 MFI values (Figure 2a). Once t(14;18) þ FL and MALT lymphomas (mean values of 5.2 and bcl-2 expression by neoplastic B cells was normalized for the 4.1, respectively) as well as among the six MCL cases who had bcl-2 levels found in resting T cells from the same sample (bcl-2 t(11;14) þ , but not within t(11;14 )À MCL (2.8 and 1.5, B/T ratio) only in six out of the 11 groups in which the mature B- respectively). cell neoplasms were distributed, the statistical significance In MM, high levels of bcl-2 MFI were found in clonal plasma (Po0.05) was retained, these including both typical and cells (Figure 2a). However, once these levels were normalized atypical CLL, PLL, MCL, MALT lymphomas and FL with according to bcl-2 expression obtained for the normal resting T t(14;18) þ , but not t(14;18)À FL, HCL, SMZL, DLCL and Burkitt cells present in the same sample, no significant differences lymphomas. However, once this bcl-2 B/T ratio was compared (P40.05) were detected in myelomatous as compared to normal for each diagnostic subgroup with the equivalent ratio found in BM plasma cells (Figure 2b and Table 1). Actually, the ratio

Figure 2 Bcl-2 expression by neoplastic cells from different B-cell malignancies. Bcl-2 levels are expressed either as MFI (a) or as the ratio between the bcl-2 MFI obtained for each neoplastic B-cell population and that found either for the normal residual T cells present in the same sample (b) or the corresponding normal BM B-cell precursors – for BCP-ALLÀ, BM plasma cells – for MM – and either CD23 þ or CD23À B- lymphocytes – for other mature B-cell disorders (c). The boxes extend from the 25th to 75th percentiles and the line in the middle represents median values. Statistically significant differences were observed between the neoplastic cells and their normal counterparts in BM (for BCP-ALL and MM), PB (for typB-CLL, atypB-CLL, MALT lymphoma and PLL), spleen (for HCL and SMZL) and lymph node (for MCL, FL, DLCL and Burkitt lymphoma): nPo0.001, nnPo0.01 and nnnPo0.04. No statistical comparisons with normal B cells were made for data shown in (c), since bcl-2 expression by normal B cells was used to calculate the plotted ratio (bcl-2 levels in neoplastic cells divided by bcl-2 levels in the corresponding normal cells). Abbreviations: MM: ; BCP-ALL: B-cell precursor acute lymphoblastic leukemia; typB-CLL: typical B-cell chronic lymphocytic leukemia; atypB-CLL: atypical B-cell chronic lymphocytic leukemia; PLL: prolymphocytic leukemia; HCL: hairy cell leukemia; MCL: mantle cell lymphoma; FL: follicular lymphoma; SMZL: splenic marginal zone lymphoma; DLCL: diffuse large B-cell lymphoma.

Leukemia Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 495 a 600 * *** 500

400

* ** 300 * ** ***

Bcl-2 MFI ** *** 200 *** ***

100 *

0

BCP-ALLTyp (n B- At PLL(n= HCL ( MCL (n=13)SMZL (n=5)MALTt(14;18)+ (n= t(14;18) FL (n=8)DLCL BURKIT((n= MM(n=21) yp B n C -CLL (n=8) = LL 3) 2) -FL (n T 3 ) 6) lymphoma =12) (n=56) =4) b (n=3

15 )

*** 10

Ratio

** 5 Bcl-2 B/T * * * ***

*

(neoplastic B-cells/resting Tcells) 0 BCP ALL (n=12) MC SMZL (n=5) BURKITT lymphoma (n=3) Typ At PLL (n=HCL (n= MALT (nt(14;18)+t( FL ( DLCL (n=6 MM yp B 14;18) B L (n (n=21) -CLL (n=5-CLL (n =13) 3) 2) =3) -F L ( ) =8) n=8) n=4) 6 c ) 15

10

normal B-cells)

Bcl-2 Ratio 5

(neoplastic B-cells/ 0 M BCP T At PL HCL (n=MCL (nSM MA t(14;18)t(14;18)D BURKIT yp B-CLLyp B-CLL (n=8) LCL M L (n=3) ZL (n LT (n=3) (n=21) -ALL (n=12) (n=6) =13) + FL ( -FL (n=4) 2) =5) T lymphoma (n= (n=5 n=8) 6)

3)

Leukemia Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 496 100 high amounts of the bcl-2 protein have been reported in n=8 neoplastic cells from patients with B-cell NHL carrying n=10 t(14;18) þ as well as in t(14;18)À B-CLL, DLCL, MM, MGUS 80 n=43 n=6 and BCP-ALL.12–20 Such observations have limited the use of n=4 bcl-2 as a diagnostic marker. However, a careful analysis of the 60 literature shows that no study has been reported so far, in which n=6 a quantitative evaluation of bcl-2 expression was performed in neoplastic B cells from BCP-ALL, MM and other mature B-cell 40 malignancies as compared to normal B-cell precursors, BM n=1 plasma cells and mature B-lymphocytes. n=1 20 n=4 The first goal of our study was to analyze quantitatively the expression of the bcl-2 protein along the different stages of 0 normal human B-cell maturation. Our results show that in normal BM, low expression of bcl-2 occurs among the earliest % of cases with an CD34 þ CD19 þ B-cell progenitors; interestingly, as these -20 n=4 CD34 þ B cells differentiate into CD34À/CD38 þþ B-cell precursors and undergo immunoglobulin gene rearrange- Overexpression 1,11 -40 ments, bcl-2 expression decreases to barely detectable Underexpression amounts. In contrast, a high bcl-2 expression is found among

abnormal mean amount of Bcl-2/ neplastic B-cells the more mature B-lymphocytes, especially in plasma cells that -60 BCP-ALL Typ Atyp MCL SMZL t(14;18)+ t(14;18)- DLCL MM are long-living cells that need regulatory mechanisms to prevent (n=12) B-CLL B-CLL (n=13) (n=5) FL FL (n=6) (n=21) 29,30 (n=56) (n=8) (n=8) (n=4) apoptosis. High bcl-2 protein levels were also found among the normal mature B-lymphocytes present in PB, spleen and Diagnostic Group lymph nodes. Interestingly, no significant differences were Figure 3 Frequency distribution of the patients studied who observed in any of these tissues with regard to bcl-2 expression showed abnormal bcl-2 expression among the B-cell malignancies in CD5 þ vs CD5À B cells (data not shown). In contrast, bcl-2 for the diagnostic groups in which at least four cases were studied. expression was significantly higher in the CD23À B cells as Abbreviations: MM: multiple myeloma; BCP-ALL: B-cell precursor compared to the CD23 þ compartment of mature B-lympho- acute lymphoblastic leukemia; TypB-CLL: typical B-cell chonic lymphocytic leukemia; atypB-CLL: atypical B-cell chronic lymphocy- cytes, independent of whether PB, spleen or lymph node tic leukemia; MCL: mantle cell lymphoma; FL: follicular lymphoma; samples were considered. It should be noted that CD23 is SMZL: splenic marginal zone lymphoma; DLCL: diffuse large B-cell expressed in mature B cells upon antigen recognition and lymphoma. subsequent cell activation;31 accordingly, the lower amounts of the bcl-2 protein found in CD23 þ B cells may reflect an between bcl-2 expression in myelomatous and normal plasma increased susceptibility to death of B-lymphocytes, following cells was slightly lower than one (Figure 2c), indicating that bcl- antigen-driven cell activation. Overall, these results confirm and 2 overexpression is not a common feature of MM plasma cells. extend previous observations showing that the expression of bcl- 2 largely varies in normal individuals along the B-cell differentiation, the amount of this protein per cell correlating Frequency of aberrant bcl-2 expression in B-cell with the expected cellular lifespan.27,32,33 Especially interesting malignancies according to diagnosis is the observation that during the rearrangement of the immunoglobulin genes in BM B-cell precursors, bcl-2 expres- Overall, 66% of all cases studied showed bcl-2 expression at sion is extremely low probably reflecting that at this stage, B-cell either abnormally high (63%) or low (3%) levels, as compared to precursors are particularly prone to apoptosis that would the amount of bcl-2 found in the corresponding normal B cells. facilitate the prevention of survival of nonfunctional B cells. As shown in Figure 3, bcl-2 overexpression was found in Once the levels of bcl-2 expression in the different normal variable proportions in the different diagnostic subgroups, the BM, PB, lymph node and spleen B-cell compartments were highest frequencies corresponding to t(14;18) þ FL (eight of eight established, our major interest focused on establishing the cases) and to MALT lymphoma (three of three cases), whereas frequency of abnormal bcl-2 expression in B-cell malignancies. the lowest frequencies were observed in MM (19%), SMZL (one Overall, neoplastic B cells from each of the cases from all of five cases) and t(14;18)À FL (one of four cases). Among the diagnostic groups studied showed expression of bcl-2, including other diagnostic groups, intermediate frequencies of bcl-2 patients with Burkitt lymphoma; these findings suggest that flow overexpression were found: BCP-ALL, 84%; typ B-CLL, 77%; cytometry may be more sensitive than conventional immuno- atyp B-CLL, 75%; PLL, two of three cases; HCL, one of two histochemistry for the detection of low levels of bcl-2.34 cases; MCL, five of six cases with t(11;14) þ and one of five Typically, once B-cell neoplasms were grouped according to without t(11;14)À; and DLCL, four of six patients. Conversely, their maturation stage (B-cell precursors, mature B-lymphocytes the underexpression of bcl-2 was only found among MM and plasma cells), bcl-2 expression in neoplastic B cells varied patients (19%), Burkitt lymphoma (two of three cases) and in in a similar way to what occurs in the normal B-cell maturation, one of the two HCL patients studied, but in none of the other with the exception of Burkitt lymphoma that had unexpectedly diagnostic subgroups. low bcl-2 levels. In order to standardize the quantitative measurements of bcl-2 by flow cytometry, a ratio between the MFI value of the target Discussion cell population and that of resting T cells present in the same sample was established both for normal and neoplastic B cells. The bcl-2 protein plays a key role in preventing apoptosis both This allowed to carry out not only a direct comparison between in normal and malignant hematopoietic cells.1,27,28 To date, the bcl-2 B/T ratio found in neoplastic cells and that found in

Leukemia Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 497 normal cells, but also to establish the frequency at which Acknowledgements abnormally high or low bcl-2 expression occurs in the different diagnostic subgroups. P Mene´ndez was supported by a grant from Fondo de As expected, patients with FL carrying t(14;18) (q32;q21) Investigaciones Sanitarias, Madrid, Spain (FIS, BEFI 98/9669). showed the highest bcl-2 B/T ratio, all cases studied over- This work was supported in part by a grant from the Red Tema´tica expressing bcl-2. However, it should be noted that neoplastic B ‘‘Mieloma mu`ltiple y otras gammapatias monoclonales’’. Fondo cells from most BCP-ALL, B-CLL, PLL, MALT lymphoma, MCL de Investigacio´n Sanitaria, Ministerio de Sanidad y Consumo, and DLCL, despite lacking t(14;18), also displayed overexpres- Madrid, Spain. sion of bcl-2; moreover, no cases showing an abnormally low expression of bcl-2 were found in any of these diagnostic subgroups. This is in clear agreement with previous studies References reporting that despite the close association between bcl-2 expression and the presence of t(14;18) (q32;q21), high bcl-2 1 Cory S. Regulation of lymphocyte survival by the bcl-2 gene levels may frequently be found in the absence of this family. Annu Rev Immunol 1995; 13: 513–543. 8,35–37 2 Williams GT. Programmed cell death: apoptosis and oncogenesis. translocation. Nevertheless, it should be noted that bcl-2 Cell 1991; 65: 1097–1098. levels were clearly higher among FL patients carrying t(14;18) 3 Vaux DL, Cory S, Adams JM. Bcl-2 gene promotes hematopoietic and that similarly increased bcl-2 levels were found only in the cell survival and cooperates with c- to immortalize pre-B cells. three MALT lymphomas and a few MCL cases, from all patients Nature 1988; 335: 440–442. studied. Interestingly, among MCL cases, bcl-2 expression was 4 Hockenbery D, Nun˜ez G, Milliman C, Schreiber RD, Korsmeyer two-fold higher among those cases carrying t(11;14)(q13;q32) as SJ. Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 1990; 348: 334–346. compared to those who did not, suggesting that overexpression 5 Nun˜ez G, London L, Hockenbery D, Alexander M, Mckean JP, of might be associated with an upregulation of bcl-2 Korsmeyer SJ. Deregulated bcl-2 gene expression selectively as previously pointed out.38 Further studies in a larger series of prolongs survival of growth factor-deprived hematopoietic cell patients are necessary to elucidate the exact relationship lines. J Immunol 1990; 144: 3602–3610. between cyclin D1 and bcl-2 overexpression in MCL. Alto- 6 Oltvai ZN, Milliman CL, Korsmeyer SJ. Bcl-2 heterodimerizes gether, these results indicate that extremely high bcl-2 expres- in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell 1993; 74: 609–619. sion in the context of a phenotype compatible with FL could be 7 Tsujimoto Y, Finger LR, Yunis J, Nowell PC, Croce CM. Cloning of used to screen for t(14;18); in contrast, normal bcl-2 expression the chromosome breakpoint of neoplastic B-cells with the t(14;18) in neoplastic B cells bearing an FL phenotype would rule out the chromosomal translocation. Science 1984; 226: 1097–1099. presence of this cytogenetic abnormality. 8 Ngan BY, Chen-Levy Z, Weiss LM, Warke RA, Cleary ML. Despite the fact that the mutational status of the IgH gene was Expression in non-Hodgkin’s lymphoma of the bcl-2 protein not analyzed in most cases included in this study, no statistically associated with the t(14;18). N Engl J Med 1988; 318: 1638–1644. 9 McDonnell TJ, Deane N, Platt FM. Bcl-2-immunoglobulin significant differences were found with regard to clinical, transgenic mice demonstrate extended survival and biological and survival features of B-CLL cases showing over- follicular lymphoproliferation. Cell 1989; 57: 79–88. expression of bcl-2 and those who did not, suggesting that bcl-2 10 McDonnell TJ, Korsmeyer SJ. Progression from lymphoid hyper- overexpression in B-CLL does not confer a different clinical plasia to high-grade malignant lymphoma in mice transgenic for behavior of the . the t(14;18). Nature 1991; 349: 254–256. In MM, plasma cells typically displayed high levels of bcl-2. 11 Adams JM, Cory S. The bcl-2 protein family: arbiters of cell survival. Science 1998; 281: 1322–1326. Despite this, bcl-2 expression in MM plasma cells was similar to 12 Pezzella F, Tse AGD, Cordell JL, Pulford KAF, Gatter KC, Masson that of normal BM plasma cells, abnormally high or low bcl-2 KY. Expression of the bcl-2 protein is not specific for the 14;18 levels being found in only a minor proportion of cases. To the chromosomal translocation. Am J Pathol 1990; 137: 225–232. best of our knowledge, this is the first report in which bcl-2 13 Schena M, Larsson LG, Gottardi, Gaidano G, Carlsson M, Nilsson expression is compared in normal and myelomatous plasma K et al. Growth and differentiated associated expression of bcl-2 cells. Previous studies have suggested that abnormally increased in B chronic lymphocytic leukemia cells. Blood 1992; 79: 2981–2989. bcl-2 expression could occur in plasma cells from most MM 14 Adachi M, Tefferi A, Greipp PR, Kipps TJ, Tsujimoto Y. Preferential 17–19 patients. Here, we clearly show in a relatively large series linkage of bcl-2 to immunoglobulin light chain gene in chronic of newly diagnosed untreated MM patients that this would only lymphocytic leukemia. J Exp Med 1990; 171: 559–564. occur in a minor proportion of cases, indicating that bcl-2 15 Piris MA, Pezzella F, Garcia-Montero JC, Orradre JL, Villuendas R, expression is not enhanced in most myelomatous plasma Sanchez-Beato M et al. p53 and bcl-2 expression in high grade B cells. These results could contribute to explain the lack of lymphomas: correlation with survival time. Br J 1994; 69: 337–341. correlation reported in MM between bcl-2 expression and both 16 Hill M, MacLenan A, Cunningham DC, Vaudhan D, Burke M, 19,39 response to treatment and survival. In addition, no Clarke P et al. Prognostic significant of bcl-2 expression and bcl-2 correlation was found in the present study between bcl-2 major breakpoint region rearrangement in diffuse large cell non- expression and neither the proportion of cycling plasma cells : A British National Lymphoma Investigation nor the presence of 13/13qÀ (data not shown), Study. Blood 1996; 88: 1046–1051. further supporting the lack of prognostic impact for bcl-2 17 Hamilton MS, Barker HF, Ball J, Drew M, Abbot SD, Franklin IM. 40,41 Normal and neoplastic plasma cells express the bcl-2 antigen. expression in MM. Leukemia 1991; 5: 568–571. In summary, our results show that the quantitative expression 18 Ladanyi M, Wang S, Niesvizky R, Feiner H, Michaeli J. Proto- of bcl-2 largely varies during normal human B-cell differentia- oncogene analysis in multiple myeloma. Am J Pathol 1992; 141: tion; once such variations are taken into account, bcl-2 949–953. overexpression appears to be a constant finding only in FL with 19 Miguel-Garcı´a A, Orero T, Matutes E, Carbonell F, Miguel-Sosa A, t(14;18) þ and MALT lymphomas. Despite this, bcl-2 over- Linares M et al. Bcl-2 expression in plasma cells from neoplastic gammopathies and reactive : a comparative study. expression is also frequently observed in BCP-ALL, B-CLL, PLL Haematologica 1998; 83: 298–304. À and MCL, but not in t(14;18) FL, SMZL, Burkitt lymphoma and 20 Campana D, Coustan Smith E, Manabe A, Buschle M, Raimondi MM. SC, Behm FG et al. Prolonged survival of B-lineage lymphoblastic

Leukemia Bcl-2 expression in normal and neoplastic human B cells P Menendez et al 498 leukemia cells is accompanied by overexpression of bcl-2 protein. cells and its over expression in mast cell leukemia. Am J Hematol Blood 1993; 81: 1025–1031. 1999; 60: 191–195. 21 Harris N, Jaffe E, Diebold J, Flandrin G, Muller-Hermelink H, 30 MacLennan IC. Germinal centers. Annu Rev Immunol 1994; 12: Vardiman I et al. World Health Organization classification 117–139. of neoplastic of the hematopoietic and lymphoid 31 Bonnefoy JY, Lecoanet-Henchoz S, Aubry JP, Gauchat JF, Graber tissues: report of the Clinical Advisory Committee Meeting-Airline P. CD23 and B-cell activation. Curr Opin Immunol 1995; 3: House, Virginia, November 1997. J Clin Oncol 1999; 17: 355–359. 3835–3849. 32 Nun˜ez G, Hockenbery D, McDonnell TJ, Sorensen CM, Korsmeyer 22 Harris N, Jaffe E, Stein H, Banks PM, Chan JKC, Clearly ML et al. A SJ. Bcl-2 maintains B cell memory. Nature 1991; 353: 71–73. revised European–American classification of lymphoid neoplasms: 33 Rolink A, Melchers F. Molecular and cellular origins of B a proposal from the International Lymphoma study Group. Blood lymphocyte diversity. Cell 1991; 66: 1081–1094. 1994; 84: 1361–1392. 34 Falini B, Mason DY. Proteins encoded by genes involved in 23 Benito GE, Sanchez ML, Del Pino-Montes J, Calvo JJ, Menendez P, chromosomal alterations in lymphoma and leukemia: clinical Garcia-Marcos MA et al. A new cytometric method for the value of their detection by immunocytochemistry. Blood 2002; 99: immunophenotypic characterization of bone-derived human 409–426. osteoclasts. Cytometry 2002; 49: 261–266. 35 Pezzella F, Gatter KC, Mason DY, Bastard C, Duval C, Krajewski A 24 Mene´ndez P, Caballero MD, Prosper F, Can˜izo MC, Pe´rez-Simo´n et al. Bcl-2 protein expression in follicular lymphomas in absence JA, Mateos MV et al. The composition of the leukapheresis of 14;18 translocation. Lancet 1990; 336: 1510–1511. products impacts on the hematopoietic recovery after autologous 36 Seite P, Hillion J, d’Agay MF, Gaulard P, Cazals D, Badoux F et al. transplantation independently of the mobilization regimen. Trans- Bcl- gene activation and protein expression in a follicular fusion 2002; 42: 1159–1172. lymphoma: a report of 64 cases. Leukemia 1993; 3: 410–417. 25 Kappelmayer J, Gratama JW, Karaszi E, Mene´ndez P, Ciudad J, 37 Sanchez-Beato M, Sanchez-Aguilera A, Piris MA. Cell cycle Rivas R et al. Flow cytometry detection of intracellular myeloper- deregulation in B-cell lymphomas. Blood 2003; 101: 1220–1235. oxidase, CD3 and CD79a: interaction between monoclonal 38 Hofmann WK, de Vos SK, Wachsman W, Pinkus GS, Said JW, clones, fluorochromes and sample preparation protocols. Koeffler HP. Altered apoptosis pathways in mantle cell lymphoma J Immunol Methods 2000; 242: 53–65. detected by oligonucleotide microarray. Blood 2001; 98: 26 Lima M, Teixeira Mdos A, Dos Santos AH, Queiro´s ML, Justic¸a B. 787–794. Decreased expression of bcl-2 (p26) in CD8(+) lymphocytes of 39 Ong F, Nieuwkoop JA, Groot-Sswings GMJS, Hermans J, Harvey patients with T-cell lymphoproliferative disorders of large gran MS, Kluin PM et al. Bcl-2 protein expression is not related to short lymphocytes. Hematol Oncol 1997; 15: 81–91. survival in multiple myeloma. Leukemia 1995; 9: 1282–1284. 27 Hockenbery DM, Zutter M, Hickey W, Nahm M, Korsmeyer S. 40 Pere´z-Simo´n JA, Garcı´a-Sanz R, Tabernero MD, Almeida J, Bcl-2 protein is topographically restricted in tissues characterized Gonzalez M, Fernandez-Calvo J et al. Prognostic value of by apoptotic cell death. Proc Natl Acad Sci USA 1991; 88: numerical chromosome aberrations in multiple myeloma: a FISH 6961–6965. analysis of 15 different chromosomes. Blood 1998; 91: 28 DiGiuseppe JA, Lebeau P, Augenbraun J, Borowitz MJ. Multi- 3366–3371. parameter flow cytometric analysis of bcl-2 and Fas expression in 41 Tricot G, Barlogie B, Jagannath S, Bracy D, Mattox S, Vesole DH normal and neoplastic hematopoiesis. Am J Clin Pathol 1996; 106: et al. Poor in multiple myeloma is associated only with 345–351. partial or complete deletions of or abnormalities 29 Cervero´ C, Escribano L, San Miguel JF, Diaz-Agustin B, Bravo P, involving 11q and not with other karyotype abnormalities. Blood Villarrubia J et al. Expression of bcl-2 by human bone marrow mast 1995; 86: 4250–4256.

Leukemia