Leukemia Research 24 (2000) 813–821

Bcl-2 and Bax expression and chlorambucil-induced apoptosis inthe T-cells and leukaemic B-cells of untreated B-cell chronic

lymphocytic leukaemia patients

Alun Thomas a, Chris Pepper a, Terry Hoy b, Paul Bentley a,*a Department of Haematology, Llandough Hospital, Penlan Road, Penarth, Vale of Glamorgan CF64 2XX, UK

b Department of Haematology, Uni6ersity of Wales College of Medicine, Cardiff CF4, 4XN UK

Received 20 November 1999; accepted 8 April 2000

Abstract

Chlorambucil and other cytotoxic drugs kill cells, non-selectively, by inducing apoptosis. In this study, we measured theapoptotic response to chlorambucil in T- and B-cells from untreated B-CLL patients and T-cells, from normal control subjects.We found increased chemosensitivity in the T-cells of B-CLL patients compared to the controls (P=0.0002). The chlorambucilID50 values for T-cells from B-CLL patients showed a direct correlation with Bcl-2 expression (P=0.002) and an inversecorrelation with CD3 cell count (PB0.0001), suggesting a trend of increasing chemosensitivity and decreasing Bcl-2 expressionwith an elevated T-cell count. There was no differential expression of Bcl-2 or Bax between the CD4+ and CD8+ cells of B-CLLpatients, isolated by immunomagnetic separation. We found correlations in the leukaemic B-cells between chlorambucil ID50

values and both Bcl-2 expression (P=0.006), and Bcl-2/Bax ratios (P=0.002), suggesting a role for the Bcl-2/Bax ratio inpredicting the response of untreated CLL patients to cytotoxic treatment. Chlorambucil produced almost identical changes inBcl-2 and Bax expression in normal T-cells and leukaemic B-cells triggered to die by apoptosis, which together with the correlationbetween Bcl-2 and chemosensitivity confirms a pivotal role for Bcl-2 in regulating a distal step in the apoptotic pathway followingcytotoxic cellular damage. © 2000 Elsevier Science Ltd. All rights reserved.

Keywords: B-cell chronic lymphocytic leukaemia; Apoptosis; T-cells; Bcl-2; Bax; Chlorambucil

www.elsevier.com/locate/leukres

1. Introduction

B-cell chronic lymphocytic leukaemia (B-CLL) ischaracterised by a clonal expansion of long-lived neo-plastic B-cells, arrested in the G0/G1 phase of the cellcycle. Although the T-cells represent a minority of thecirculating lymphocytes in B-CLL, their absolute num-bers have been reported to be increased [1]. In additionboth phenotypic and functional abnormalities havebeen described including a decreased CD4:CD8 ratio

[2], probably resulting from a relative expansion of theCD8+ T-cell subset [3]. The abnormal distribution ofthe CD4+, CD8+ subsets together with a loss of T-cellfunction has been correlated with the stage of thedisease [4–6]. More recent work has shown an increasein the number of CD4+ and CD8+ lymphocytes ex-pressing the CD45RO marker with a concomitant in-crease in MHC class II molecules [7]. This suggests thatan increased proportion of T lymphocytes from B-CLLpatients are activated, similar to that seen in chronicantigen stimulation [7]. T-cell activation has beenshown to be more marked in patients with advanceddisease and higher lymphocyte counts, suggesting thatthe source of chronic antigen stimulation could beleukaemic B-cells [4]. Additionally, the co-stimulatorymolecule CD28 and the zeta (j) chain of the T-cellreceptor (TCR) have been reported to have reducedexpression in B-CLL T lymphocytes compared to nor-

Abbre6iations: B-CLL, B-cell chronic lymphocytie leukaemia; FasL,Fas ligand; FCS, foetal calf serum; MESF, molecules of equivalentsoluble fluorochrome; MFI, mean fluorescent intensity; MPC, mag-netic particle concentrator; PBS, phosphate buffered saline; TCR,T-cell receptor.

* Corresponding author. Tel.: +44-29-207-15864; fax: +44-29-20-715399.

E-mail address: paul.bentley@llanhaem.demon.co.uk (P. Bentley).

0145-2126/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.PII: S 0 1 4 5 -2126 (00 )00051 -5

A. Thomas et al. / Leukemia Research 24 (2000) 813–821814

mal T-cells [8]. This together with the activated pheno-type and defects in T-cell function would be expected toresult in anergy and apoptosis in a proportion ofB-CLL patients’ T-cells [8,9].

Apoptosis plays an important role in regulating thesize of the of the mature T lymphocyte population [10]since non-functional and autoreactive T-cells are elimi-nated by this process. The best defined regulators ofapoptosis in T-cells are members of the Fas and Bcl-2family of proteins. The Fas (CD95) surface receptorand Fas ligand (FasL) are upregulated upon stimula-tion by antigens. FasL mediates cell death by cross-linking with the Fas receptor in activated T-cells whenthey are repeatedly stimulated by antigen [11]. Thesusceptibility of T-cells to undergo apoptosis is alsocontrolled by the Bcl-2 family of proteins. Overexpres-sion of Bcl-2 has been shown to prevent T-cells fromundergoing apoptosis following deprivation from acti-vating stimuli or growth factors and when they areexposed to cytotoxic drugs or ionising radiation [12]. Incontrast Bax counters the anti-apoptotic effect of Bcl-2and promotes apoptosis [13]. The Bcl-2 family ofproteins have been implicated in the dysregulation ofthe programmed cell death pathway in the leukaemicclone in B-CLL leading to extended cell survival. Insupport of this notion laboratory findings have shownthat Bcl-2 is overexpressed in the leukaemic cells inB-CLL in the apparent absence of any gene rearrange-ments [14]. Gene transfer experiments in which cellswere induced to overexpress Bcl-2 have been shown toincrease resistance of the cells to chemotherapy [15].These experiments would predict a correlation betweenBcl-2 expression and poor clinical outcome in B-CLL.Such a correlation has been reported by Robertson etal. [16]. Other studies have shown that the Bcl-2/Baxratio was higher in the leukaemic cells in patients whohad a poor response to chemotherapy and those withprogressive disease [17,18]. However, conflicting evi-dence does exist [19,20].

In contrast to the leukaemic B-cells, no informationis available regarding the Bcl-2 and Bax status ofnormal lymphocytes from patients with B-CLL. Wedecided to compare the apoptotic responses of T-cellsand B-cells, from previously untreated B-CLL patientsand T-cells from normal controls following in vitroexposure to chlorambucil, a bifunctional alkylatingagent used extensively in the treatment of B-CLL. TheID50 values obtained were correlated with the ex vivoBcl-2 and Bax protein expression. The modulation ofBcl-2 and Bax during apoptosis was compared in T-and B-cells of B-CLL patients and the T-cells of thenormal control subjects, by measuring separately theirexpression in viable and apoptotic cells. In addition,Bcl-2 and Bax expression was quantified and comparedin isolated CD4+ and CD8+ T-cells from B-CLL pa-tients and normal controls. The differential expression

of Bcl-2 and Bax between CD4+ and CD8+ cells ofB-CLL patients was examined to see if it might berelated to the reversed CD4/CD8 ratios seen in somepatients, or the CD4+ subset lymphopenia frequentlyseen after cytotoxic treatment with fludarabine [21].

2. Materials and methods

2.1. Patients and isolation of mononuclear cells

Peripheral blood from B-CLL patients and normalage-matched controls was obtained with their informedconsent. Clinical staging of the B-CLL patients wasperformed according to Binet et al. [22]; all the B-CLLpatients in this study were in Binet stage A and had notpreviously received treatment. Peripheral bloodmononuclear cells were obtained by centrifugation on aFicoll-Hypaque (Sigma, UK) density gradient and werewashed three times in phosphate buffered saline (PBS)prior to incubation or antibody labelling.

2.2. Cell culture and cytotoxic drugs

Freshly isolated lymphocytes (1×106/ml) were incu-bated in Minimum Essential Medium (Gibco, UK)supplemented with 10% heat inactivated foetal calfserum (FCS), penicillin and streptomycin. Cells wereincubated for 48 h at 37°C in an atmosphere of 5% CO2

in the absence or presence of a range of concentrationsof chlorambucil (0–40 mg/ml). Stock solutions of chlo-rambucil (Sigma, UK) were stored at −20°C, and onealiquot was thawed on the day of use.

2.3. Bcl-2 and Bax staining and quantitation

Bcl-2 and Bax staining was analysed by triple colourimmunofluorescence. Briefly, 1×106 cells were incu-bated with 10 ml of either CD19 or CD3, Cy5 PEconjugated antibody together with an isotype control(Dako, UK). The cells were then fixed using the com-mercial ‘Fix and Perm’ kit (Caltag, USA). Followingwashing in PBS the cells were resuspended in perme-abilisation solution and incubated with 10 ml of FITC-labelled Bcl-2 antibody (Dako), and 10 ml of Baxantibody (Santa Cruz, USA) or isotype matched nega-tive controls. After washing the cells were incubatedwith a PE-labelled secondary antibody for the Baxstained cells. A FACScan flow cytometer (Becton Dick-inson, USA) equipped with an argon laser (488 nm)was used to measure fluorescence. Gating was per-formed on CD19+ or CD3+ cells to ensure Bcl-2 andBax were measured in B- or T-cells, respectively. Themean fluorescent intensity (MFI) was calculated forBcl-2 and Bax separately using WinMDI software (J.Trotter, Scripps Research Institute, USA). MFI values

A. Thomas et al. / Leukemia Research 24 (2000) 813–821 815

were then converted to molecules of equivalent solublefluorochrome (MESF) by using a calibration curve. Thecalibration curve was obtained by measuring beads(Dako) labelled with known amounts of fluorochrome.In addition, the flow cytometer was operated underidentical conditions thus controlling day-to-day varia-tion in fluorescence detection.

2.4. Measurement of in 6itro apoptosis

Cells were cultured as described previously and apop-tosis induced by exposure to a range of five chlorambu-cil concentrations between 0 and 40 mg/ml and 0 and 15mg/ml for CD3+ and CD19+ B-CLL cells, respectively.Spontaneous apoptosis was measured in cultures in theabsence of chlorambucil. After washing twice in PBSthe cells were stained for CD3 or CD19 and Bcl-2 andBax as described previously. Lymphocytes undergoingapoptosis exhibit a reduction in forward light scatterand an increase in right angle light scatter. We havepreviously shown that these characteristic changes canbe used to gate apoptotic cells without the need to labelthe cells with a specific apoptosis marker [23]. Thismethod allowed additional triple-label analysis to beperformed enabling changes in Bcl-2 and Bax to bemeasured in the apoptotic and viable cell populationsof CD3+ or CD19+ gated lymphocytes. Althoughchanges in light scatter were used as the main methodof apoptosis quantitation, Annexin V labelling wasperformed to verify that the gated cells were apoptoticor viable [24].

2.5. Staining for cell surface markers

Freshly isolated mononuclear cells were analysed forCD3 expression prior to incubation with chlorambucil.1×106 cells were resuspended in 100 ml of PBS andincubated for 30 min at 4°C with 10 ml of CD3, Cy5 PEconjugated antibody or an isotype negative control(Dako). Following two washes in PBS, a minimum of5000 events in a CD3+ gate were analysed for eachpatient.

2.6. Immunomagnetic separation of CD4+ and CD8+

T-cells

CD4+ and CD8+ lymphocytes were positively se-lected using Dynabeads M-450 coated with anti-CD4 oranti-CD8, respectively (Dynal, UK). Dynabeads wereadded to the isolated mononuclear cells using a targetto bead ratio of 1:3. The mixture was incubated withmixing for 30 min at 4°C and the rosetted CD4+ orCD8+ cells were then isolated using a magnetic particleconcentrator (MPC). Pure populations of CD4+ orCD8+ T-cells were obtained by washing six times withPBS with 1% FCS. Detachment of the Dynabeads was

achieved by adding 10 ml of Detachabead (Dynal, UK)to the rosetted cells suspended in 100 ml of culturemedium and incubating the mixture, with mixing atroom temperature for 60 min. The tube was placed ona MPC and the detached cells removed by pipetting.The cells were washed three times in culture medium toremove Detachabeads from the solution prior to stain-ing for Bcl-2 and Bax. Additionally, the Dynabead-iso-lated cells were stained using triple colourimmunofluorescence for CD3, CD4 and CD8, to verifythe efficiency of the separation.

2.7. Statistical analysis

Statistical analysis was performed using unpairedStudent’s t-test, Mann–Whitney test and the correla-tion coefficients were obtained from simple linear re-gression analysis.

3. Results

3.1. Chlorambucil induced apoptosis in CD3+

lymphocytes from B-CLL patients and normal controlsand in CD19+ cells from B-CLL patients

A comparison of the chemosensitivity of the Tlymphocytes from 15 B-CLL patients and ten controlsto chlorambucil is shown in Fig. 1; both groups showedheterogeneity in sensitivity to chlorambucil. For theCD3+ cells from B-CLL patients the ID50 ranged from10.9 to 24.2 mg/ml (mean 19.0 mg/ml) and for CD3+

cells from the controls from 15.0 to 36.2 mg/ml (mean28.4 mg/ml). The difference between the two groups wasvery significant (t=4.46, P=0.0002). The CD19+ cellsfrom the B-CLL patients were more susceptible, in eachpatient, to chlorambucil-induced apoptosis than theirCD3+ counterparts. The ID50 for chlorambucil in theCD19+ cells ranged from 0.9 to 14.7 mg/ml (mean 6.1mg/ml). The data are summarised in Table 1.

3.2. Spontaneous apoptosis in CD3+ cells from B-CLLpatients and normal controls and in CD19+ cells fromB-CLL patients

The amount of spontaneous apoptosis was deter-mined after 48 h incubation in the absence of chloram-bucil and the data are shown in Fig. 1. The CD3+ cellsfrom B-CLL patients and normal controls showed arange of apoptosis between 5.3 and 19.2% (mean11.6%) and 2.4 and 13.9% (mean 9.05%), respectively.The difference was not significant (t=1.58, P=0.13).No correlation was found between the rate of sponta-neous apoptosis and in vitro chemosensitivity to chlo-rambucil for the CD3+ cells from B-CLL (r= −0.36,P=0.2) or control patients (r= −0.16, P=0.16).

A. Thomas et al. / Leukemia Research 24 (2000) 813–821816

Tab

le1

Per

cent

age

spon

tane

ous

apop

tosi

s,B

cl-2

and

Bax

expr

essi

on(e

xvi

vo),

Bcl

-2/B

axra

tios

and

chlo

ram

buci

lID

50

valu

esw

ithi

nth

eC

D3+

and

CD

19+

anti

gen

defin

edpo

pula

tion

sof

15B

-CL

Lpa

tien

ts

CD

3C

D19

Pat

ient

ID5

0(m

g/m

l)B

cl-2

(ME

SF)

Bax

(ME

SF)

Bcl

-2/b

axra

tio

ID50

(mg/

ml)

Spon

tane

ous

Spon

tane

ous

Bcl

-2/B

axra

tio

Bax

(ME

SF)

Bcl

-2(M

ESF

)ap

opto

sis

(%)

apop

tosi

s(%

)

1507

619

307.

83.

612

.337

.215

.65.

816

0792

501

ndnd

nd2

nd10

523

nd16

076.

520

.35.

313

522

2015

6.7

12.7

6.9

15.7

6.5

19.9

311

710

1798

5.6

1065

315

076

1670

9.0

9.6

5.9

2176

4.9

20.0

411

.113

458

1069

222

824.

60.

945

.027

294.

922

.45

1259

526

484.

82.

034

.36.

56

17.2

4.7

2768

1300

475

3513

.713

544

3181

4.3

5.7

29.4

1870

4.0

710

.912

490

1870

6.6

7.1

18.6

11.5

1204

88

24.2

6.9

1730

9.1

nda

ndnd

nd8.

518

.9nd

nd21

.79

16.9

1130

021

589

1846

11.7

14.7

10.8

2176

5.2

18.7

1010

174

2257

4.5

1.1

42.8

8.6

1122

.45.

026

3113

177

1213

417

9943

1710

7.9

9.9

17.0

2120

4.7

14.7

12.2

1402

218

627.

55.

229

.619

.24.

217

.513

8881

2096

14.3

1276

712

832

2762

4.6

3.3

34.1

3098

4.1

19.3

1422

.89.

013

048

2257

5.7

1.7

36.3

1511

775

2582

4.6

and

,no

tde

term

ined

.

A. Thomas et al. / Leukemia Research 24 (2000) 813–821 817

Highly variable rates of spontaneous apoptosis, rangingfrom 5.9 to 45% (mean 26.2%), were observed amongthe CD19+ cells from B-CLL patients and in contrastto the CD3+ cells, a strong correlation was foundbetween spontaneous apoptosis and in vitro chemosen-sitivity to chlorambucil (r= −0.93, PB0.0001). Sig-nificantly, the CD19+ cells also showed a correlationwith ex vivo Bcl-2/Bax ratios (r= −0.69, P=0.009).

3.3. Bcl-2 and Bax measurement

The level of Bcl-2 has been reported to be decreased,and the level of Bax increased in the T-cells of healthyelderly controls compared to young controls [25]. Inview of this, the controls used in these experiments wereage-matched with the B-CLL patients. Bcl-2 and Baxexpression was determined by triple colour immu-nofluorescence using flow cytometry. This allowed themeasurement of Bcl-2 and Bax in the CD19+ andCD3+ cells of each patient. Both Bcl-2 and Bax werefound to be uniformly expressed in the CD3+ cellsfrom the leukaemia patients and controls with no sig-nificant difference in the level of expression between thetwo groups (t=0.07, P=0.95 for Bcl-2 and t=0.06,P=0.95 for Bax). The mean Bcl-2 MESF for theCD19+ cells was 13698.2, higher than that for theCD3+ cells of the B-CLL patients and controls (MESF11144.6 and 11191 respectively), and was responsiblefor the higher mean Bcl-2/Bax ratio seen in the CD19+

cells from the B-CLL patients.

3.4. Correlation of Bcl-2 and Bax le6els with drugsensiti6ity

The Bcl-2 and Bax protein levels were measured inthe patients and controls prior to incubation (ex vivo)and were compared with their ID50 values for chloram-

bucil. A correlation between Bcl-2 and ID50 values wasnoted for the CD3+ cells of the B-CLL patients (r=0.76, P=0.002). The ID50 values of the CD19+ cells ofthe B-CLL patients exhibited a correlation betweenBcl-2 expression (r=0.72, P=0.006) and a strongercorrelation with the Bcl-2/Bax ratios (r=0.78, P=0.002).

3.5. Expression of Bcl-2 and Bax in CD4+ and CD8+

isolated lymphocytes from B-CLL patients and normalcontrols

Lymphocytes isolated by magnetic separation werestained for CD3 and CD4 or CD8 and more than 95%purity was found in each case. The Bcl-2 and Baxexpression of CD4+ and CD8+ separated lymphocytesfrom 15 untreated CLL, patients and ten age-matchedcontrols was determined by dual colour immunofluores-ence and the results expressed as MESF. As shown inFig. 2, there was no significant difference in the levelsof Bcl-2 and Bax expression or the Bcl-2/Bax ratio inCD4+ and CD8+ cells from the B-CLL patients com-pared with normal controls. Additionally, differences inBcl-2 and Bax expression between CD4+ and CD8+

cells in the B-CLL patients were not significant (t=1.53, P=0.14 for Bcl-2 and t=0.43, P=0.67 for Bax).

3.6. Correlation of CD3 numbers with T-cellchemosensiti6ity

Total peripheral blood CD3 numbers were measuredin all 15 B-CLL patients and controls prior to incuba-tion (ex vivo). The mean (9SD) absolute CD3 countfor the B-CLL patients was 2.9×109/l (91.9) and1.5×109/l (90.3) for the controls. This was a signifi-cant difference (Mann–Whitney U=18.5, P=0.008).Interestingly, a close inverse correlation was found in

Fig. 1. Mean chlorambucil ID50 values and spontaneous apoptosis values (9SD) in CD19+ and CD3+ lymphocytes from B-CLL patients andCD3+ lymphocytes from normal controls following 48 h incubation in vitro. Open bars, chlorambucil ID50; solid bars, spontaneous apoptosis.

A. Thomas et al. / Leukemia Research 24 (2000) 813–821818

Fig. 2. Bcl-2 and Bax expression in CD4+ and CD8+ lymphocytes from 15 B-CLL patients and ten normal controls. (A) Mean Bcl-2 expression(9SD), (B) mean Bax expression (9SD) and (C) mean Bcl-2/Bax ratios (9SD). Open bars, CLL patients; solid bars, controls.

the T-cells of the B-CLL patients between chlorambucilID50 values and absolute CD3 numbers (r= −0.93,PB0.0001) (Fig. 3). A significant inverse correlationwas also found between CD3 numbers and Bcl-2 ex-pression (r= −0.76, P=0.002).

3.7. Bcl-2 and Bax modulation during apoptosis

Viable and apoptotic cells were gated based on theirforward and side scatter characteristics. Bcl-2 and Bax

expression was then quantified separately for both pop-ulations of cells. Both drug-induced and spontaneousapoptosis produced similar changes in Bcl-2 and Baxlevels in the CD3+ cells from controls and B-CLLpatients and in the CD19+ cells from B-CLL patients.Fig. 4 shows histograms of Bcl-2 and Bax expressionfor the CD3+ and CD19+ cells of one B-CLL patientand the CD3+ cells of a control, following incubationwith chlorambucil. Bcl-2 expression was found to bereduced and Bax expression increased in the apoptotic

A. Thomas et al. / Leukemia Research 24 (2000) 813–821 819

cells relative to the viable cells. Bcl-2 expression wasreduced by approximately a half whereas Bax expres-sion showed a 2–3-fold increase. The changes in Bcl-2and Bax expression following chlorambucil-inducedapoptosis were quantitatively similar regardless of drugconcentration or whether apoptosis was triggered inCD3+ or CD19+ from B-CLL patients or CD3+ cellsfrom controls.

4. Discussion

In this study, we have demonstrated that T-cells fromuntreated B-CLL patients showed increased susceptibil-ity to chlorambucil-induced apoptosis as comparedwith T-cells from age-matched normal controls. Thestrong correlations in the B-CLL patients between CD3numbers and the chlorambucil ID50 (PB0.0001), andBcl-2 protein expression and CD3 numbers (P=0.002)suggests a trend of increasing chemosensitivity anddecreasing Bcl-2 expression with increasing T-cellcount. In addition to activation changes in the T-cellsof B-CLL compatible with chronic antigen stimulation[4,7], evidence also exists for T-cell clonal expansiondirected against leukaemia-related antigens. This in-cludes the presence of aberrant TCR repertoires, withindividual patients often expressing unique Va or Vbtranscripts that are stable over time and over-repre-sented compared to normal controls [26,27]. More thanone expanded T-cell clone has been reported in somepatients, suggesting multiple T-cell expansions in re-sponse to B-CLL leukaemic cells [28]. Chemotherapywith chlorambucil has been shown to produce a pro-portionally greater fall in T-cell numbers in patients

possessing the most aberrant TCR repertoires [29];following treatment the TCR repertoire was found tobe more similar to normal controls. It is interesting tospeculate that the T-cells from B-CLL patients may bemore sensitive to chlorambucil due to the presence ofT-cell clones expanded in response to leukaemia relatedantigens. If so our results would suggest that suchT-cell clones would be more sensitive to chlorambucilbecause of a lower Bcl-2 protein expression. Futurework might evaluate the chemosensitivity and Bcl-2expression from normal subjects who have a chronicantigen stimulation, e.g. chronic infection, in order todetermine whether comparable changes occur to thoseobserved in T-cells from B-CLL patients.

CD4+ and CD8+ cells from B-CLL have been re-ported to express increased amounts of Fas receptor incomparison to normal controls [30]. Only the CD4+

cells showed increased Fas mediated apoptosis in vitroand this sensitivity correlated in individual patients withthe CD4/CD8 ratio, suggesting a possible explanationfor the reversed CD4/CD8 ratio reported in some CLLpatients. Additionally, there is evidence that Fas-in-duced apoptosis is not inhibited by Bcl-2, at least inT-cells, with Fas and Bcl-2 regulating different path-ways of apoptosis [31,32]. We found no difference inBcl-2 and Bax protein expression or Bcl-2/Bax ratiosbetween the CD4+ and CD8+ T-cells of the B-CLLpatients. Therefore, the CD4+ lymphopenia found invivo following treatment with fludarabine or the re-versed CD4/CD8 ratios reported in many CLL patientscannot be explained by dysregulation of the Bcl-2 fam-ily. Consoli et al. (1998) found no difference in the invitro apoptotic response of CD4 and CD8 cells tofludarabine suggesting a slower recovery of the CD4+

relative to the CD8+ cells post chemotherapy [33].

Fig. 3. Correlation between the T-cell chlorambucil ID50 values and CD3 numbers from 15 B-CLL patients.

A. Thomas et al. / Leukemia Research 24 (2000) 813–821820

Fig. 4. Histograms showing Bcl-2 and Bax expression in viable(shaded) and apoptotic cells (clear) following 48 h incubation withchlorambucil. In (A) and (B) CD3+ gated cells are shown from anormal control and B-CLL patient respectively, following incubationwith 15 mg/ml chlorambucil. In (C) CD19+ gated cells are shownfrom a B-CLL patient following incubation with 5 mg/ml chlorambu-cil.

chlorambucil ID50 and the Bc-2/Bax ratios in theleukaemic B-cells would imply a role for the ratio inpredicting the response of untreated patients tochemotherapy. Molica et al. extended the role of theBcl-2/Bax ratio, to suggest it may be more sensitivethan current prognostic indicators in assessing the riskof disease progression, especially in early B-CLL [37].

We found no differences in the changes in Bcl-2 andBax expression in the apoptotic cells of T-cells andleukaemic B-cells. Additionally, the changes were thesame whether the cells were induced to die by chloram-bucil or spontaneous apoptosis. This together with thestrong correlation between Bcl-2 protein expression andchlorambucil ID50 values suggests that the process ofapoptosis, following exposure to cytotoxic agents, con-verges into a final common pathway used by both theT-cells and leukaemic B-cells, which is probably regu-lated by Bcl-2.

In summary the data presented suggests that the levelof Bcl-2 protein expression is an important determinantof the relative sensitivity or resistance of T-cells and theleukaemic B-cells in B-CLL. Furthermore, Bcl-2/Baxratios are of additional prognostic significance in thecase of the leukaemic B-cells.

Acknowledgements

This work was supported in part by a grant from theWelsh Bone Marrow Transplant Research Fund.

References

[1] Catovsky D, Miiani E, Okos A, Galton D. Clinical significanceof T-cells in chronic lymphocytic leukaemia. Lancet 1974;2:751.

[2] Platsoukas CD, Galinski M, Kempin S, Reich L, Clarkson B,Good RA. Abnormal T lymphocyte subpopulations in patientswith B-cell chronic lymphocytic leukemia: an analysis by mono-clonal antibodies. J Immunol 1982;129:2305.

[3] Kay NE, Johnson JD, Stanek R, Douglas SD. T-cell subpopula-tions in chronic lymphocytic leukemia: abnormalities in distribu-tion and in vitro receptor maturation. Blood 1979;54:540.

[4] Totterman TH, Carlsson M, Simonsson B, Bengtsson M,Nilsson K. T-cell activation and subset patterns are altered inB-CLL and correlate with the stage of the disease. Blood1989;74:786.

[5] Peller S, Kaufman S. Decreased CD45RA T cells in B-cellchronic lymphatic leukemia patients: correlation with diseasestage. Blood 1991;78:1569.

[6] Zaknoen SL, Kay NE. Immunoregulatory cell dysfunction inchronic B-cell leukemias. Blood Rev 1990;4:165.

[7] Frovlova EA, Richards SJ, Jones RA, Rawstron A, Master PS,Teasdale J, et al. Immunophenotypic and DNA genotypic analy-sis of T-cell and NK-cell subpopulations in patients with B-cellchronic lymphocytic leukaemia (B-CLL). Leuk Lymphoma1995;16:307.

[8] Rossi E, Matutes E, Morilla R, Owusu-Ankomah K, HeffernanM, Catovsky D. Zeta chain and CD28 are poorly expressed onT lymphocytes from chronic lymphocytic leukemia. Leukemia1996;10:494.

We have previously shown normal B-cells from age-matched controls to have lower Bcl-2 expression andlower Bc-2/Bax ratios, and to be more sensitive tochlorambucil than B-CLL cells [34]. In the presentstudy, we found a correlation in the leukaemic B-cellsbetween Bcl-2 levels and the ID50 value for chlorambu-cil (P=0.006) and a stronger correlation between theID50 value and the Bcl-2/Bax ratio (P=0.002). Thepatients expressing the least amount of spontaneousapoptosis were found to be the most resistant to chlo-rambucil. The strong correlation of the Bc2/Bax ratioto both the chlorambucil ID50 and spontaneous apop-tosis shows the importance of the ratio in determiningthe susceptibility of the leukaemic cells to apoptoticsignals. There was a heterogeneity in response to chlo-rambucil among the patients, unrelated to treatment orclinical stage, with some patients markedly more resis-tant than others. This was in agreement with the resultsof Bentley et al. and Silber et al. who suggested thismay represent de novo resistance in some untreatedpatients [35,36]. The strong correlation between the

A. Thomas et al. / Leukemia Research 24 (2000) 813–821 821

[9] Bartik MM, Welker D, Kay NE. Impairments in immune cellfunction in B cell chronic lymphocytic leukemia. Semin Oncol1998;25:27.

[10] Strasser A. Life and death during lymphocyte development andfunction: evidence for two distinct killing mechanisms. CurrOpin Immunol 1995;7:228.

[11] Wong B, Choi Y. Pathways leading to cell death in T cells. CurrOpin Immunol 1997;9:358.

[12] Cory S. Regulation of lymphocyte survival by the Bcl-2 genefamily. Annu Rev Immunol 1995;13:513.

[13] Oltvai ZN, Milliman CL, Korsmeyer SJ. Bcl-2 heterodimerizesin vitro with a conserved homolog, Bax, that accelerates pro-grammed cell death. Cell 1993;74:609.

[14] Hanada M, Delia D, Aiello A, Stadtmauer E, Reed JC. Bcl-2gene hypomethylation and high level expression in B-cell chroniclymphocytic leukemia. Blood 1993;82:1820.

[15] Reed J. Molecular biology of chronic lymphocytic leukemia:implications for therapy. Semin Haematol 1998;35:3.

[16] Robertson LE, Plunkett W, McConnell Y, Keating MJ, McDon-nell TJ. Bcl-2 expression in chronic lymphocytic leukemia and itscorrelation with the induction of apoptosis and clinical outcome.Leukemia 1996;10:456.

[17] Aguilar-Santelises M, Rottenberg ME, Lewin N, Mellstedt H,Jondal M. Bcl-2, Bax and p53 expression in B-CLL in relation toin vitro survival and clinical progression. Int J Cancer1996;69:114.

[18] Pepper C, Hoy T, Bentley P. Bcl-2/Bax ratios in chroniclymphocytic leukaemia and their correlation with in vitro apop-tosis and clinical resistance. Br J Cancer 1997;76:935.

[19] Kitada S, Andersen J, Akar S, Zapata JM, Takayama S, Kra-jewski S, et al. Expression of apoptosis-regulating proteins inchronic lymphocytic leukemia: correlations with in vitro and invivo chemoresponses. Blood 1998;91:3379.

[20] Johnston JB, Daeninck P, Verburg L, Lee K, Williams G, IsraelsLG, Mowat M, Begleiter A. P53, mdm2, Bax and Bcl-2 and drugresistance in chronic lymphocytic leukemia. Leuk Lymphoma1997;26:435.

[21] Bergmann L, Fenchel K, Jahn B, Mitrou PS, Hoelzer D. Im-munosuppressive effects and clinical response of fludarabine inrefractory chronic lymphocytic leukemia. Ann Oncol 1993;4:371.

[22] Binet JL, Auquier A, Dighiero G, Chastang C, Piguet H,Goasguen J, et al. A new prognostic classification of chroniclymphocytic leukemia derived from multivariate survival analy-sis. Cancer 1981;48:198.

[23] Pepper C, Thomas A, Hoy T, Bentley P. Chlorambucil resistancein B-cell chronic lymphocytic leukaemia is mediated throughfailed Bax induction and selection of high Bcl-2-expressing sub-clones. Br J Haematol 1999;104:581.

[24] Vermes I, Haanen C, Steffens-Nakken R, Reutelingsperger C. Anovel assay for apoptosis: flow cytometric detection of phos-phatidylserine expression on early apoptotic cells using fluores-cein labelled Annexin V. J Immunol Methods 1995;184:39.

[25] Aggarwal S, Gupta S. Increased apoptosis of T cell subsets inageing humans: altered expression of Fas (CD95), Fas ligand,Bcl-2 and Bax. J Immunol 1998;160:1627.

[26] Wen T, Mellstedt H, Jondal M. Presence of clonal T cellpopulations, in chronic B lymphocytic leukemia and smoulderingmyeloma. J Exp Med 1990;171:659.

[27] Farace F, Orlanducci F, Dietrich P-Y, Gaudin C, Angevin E,Courtier M-R, et al. T cell repertoire in patients with B chroniclymphocytic leukemia. J Immunol 1994;153:4281.

[28] Alatrakchi N, Farace F, Frau E, Carde P, Munck J-N, TriebelF. T-cell clonal expansion in patients with B-cell lymphoprolifer-ative disorders. J Immunother 1998;21:363.

[29] Rassenti LZ, Bessudo A, Tsai A, Kipps TJ. Analysis of therepertoire expressed by the T cells of patients with chroniclymphocytic leukemia pre and post chemotherapy. Blood1998;92(Suppl 1):432a.

[30] Tinhofer I, Marschitz I, Kos M, Henn T, Egle A, Villunger A, etal. Differential sensitivity of CD4+ and CD8+ T lymphocytes,to the killing efficacy of Fas (Apo- 1/CD95) ligand+ tumourcells in B chronic lymphocytic leukemia. Blood 1998;91:4273.

[31] Van Parijs L, Biuckians A, Abbas AK. Functional roles of Fasand Bcl-2-regulated apoptosis of T lymphocytes. J Immunol1998;160:2065.

[32] Moreno MB, Memon SA, Zacharchuk CM. Apoptosis signalingpathways in normal T cells. Differential activity of Bcl-2 andIL-1b-converting enzyme family protease inhibitors on glucocor-ticoid- and Fas-mediated cytotoxicity. J Immunol 1996;157:3845.

[33] Consoli U, El-Tounsi I, Sandoval A, Snell V, Kleine H-D,Brown W. Differential induction of apoptosis by fludarabinemonophosphate in leukemic B and normal T cells in chroniclymphocytic leukemia. Blood 1998;91:1742.

[34] Pepper C, Bentley P, Hoy T. Regulation of clinical chemoresis-tance by bcl-2 and bax oncoproteins in B-cell chroniclymphocytic leukaemia. Br J Haematol 1996;95:513.

[35] Bentley DP, Blackmore JA. The inhibition of DNA synthesis inchronic lymphocytic leukaemia cells by chlorambucil in vitro. BrJ Cancer 1992;65:171.

[36] Silber R, Degar B, Costin D, Newcomb EW, Mani M, Rosen-berg CR, et al. Chemosensitivity of lymphocytes from patientswith B-cell chronic lymphocytic leukemia to chlorambucil,fludarabine, and camptothecin analogs. Blood 1994;84:3440.

[37] Molica S, Dattilo A, Giulino C, Levato D, Levato L. Increasedbcl-2/bax ratio in B-cell chronic lymphocytic leukemia is associ-ated with a progressive pattern of disease. Haematologica1998;83:1122.

.