Adhesion to fibronectin via α4 integrin (CD49d) protects B cells from apoptosis induced by serum deprivation but not via IgM or Fas/Apo-1 receptors

Abstract

Apoptosis is a regulated event crucial to the development and proliferation of normal and malignant B cells. We have studied the role of signals delivered via α4 integrin on apoptosis triggered by three different pathways on these cells. For apoptosis induced by serum deprivation, culturing B cells on the recombinant fibronectin fragment H89, a known ligand for α4β1 integrin, resulted in statistically significant (P < 0·005) higher viability values (68%, 65% and 67%) for Ramos, Nalm-6 and EHEB cells, respectively, than culturing cells on poly lysine (42%, 42% and 48%). An antiα4 MoAb reverted the protecting effect, thus confirming that it was due specifically to α4 engagement. Similarly, cells cultured on FN-III4-5, a recently identified fibronectin region which binds activated α4 integrin, also showed statistically significant higher viability than poly lysine cultures. α4 engagement however, did not prevent apoptosis induced on Ramos cells via surface IgM. Adhesion of IM-9 cells, a myeloma cell line carrying functional Fas receptors, to the H89 fragment neither increased cell viability upon triggering apoptosis via Fas when compared to poly lysine. These results indicate that α4 signalling may overcome B cell apoptosis induced by the lack of growth factors but does not seem to affect the IgM or Fas apoptotic pathways, thus suggesting different intracellular mechanisms for these processes.

INTRODUCTION

Apoptosis, or programmed cell death, is an active event in the development and differentiation of B lymphocytes [1]. At the germinal centre, for example, only cells carrying high affinity antigen receptors survive and become memory cells, while the rest die via apoptotic processes [2]. Several cell surface molecules are known to regulate B cell apoptosis including the imunoglobulin component of the B cell receptor (mostly IgM), CD40 and Fas or Apo-1 [1]. Experimental work has shown that cross-linking of surface IgM or Fas with specific monoclonal antibodies (MoAbs) usually results in cell death and that concomitant engagement of CD40 either blocks (anti-IgM) or enhances (anti-Fas) B-cell apoptosis, thus playing a dual role [3–5]. The precise mechanisms by which CD40 control these responses are not well understood and it is likely that other molecules also participate in these processes.

Integrins are a family of adhesion receptors which provide survival signals in many cellular systems (reviewed in [6]). In the case of B cells, it was shown that disruption of the interaction of α4β1 integrin (CD49d/CD29) with its endothelial ligand VCAM-1 (vascular cell adhesion molecule-1) resulted in apoptosis of germinal centre B cells [2]. Similarly, adhesion of chronic lymphocytic leukaemia B cells (B-CLL) to endothelium [7] or bone marrow stroma [8] via α4β1 and αLβ2 integrins prevented their spontaneous apoptosis. We have reported previously that adhesion of B-CLL cells to the extracellular matrix component fibronectin (Fn) or to the recombinant Fn fragment H89, which contains the specific ligands for α4β1 [9], prevents their spontaneous apoptosis and increases the Bcl-2/Bax ratio [10]. α4β1 interaction with VCAM-1 or Fn also prevents chemotherapy-induced apoptosis in lymphoma [11] and myeloma cells [12]. These previous findings suggest a regulatory role for α4β1 in normal and malignant B cell apoptosis.

In this report we have studied whether α4β1 engagement could interfere with three different ways of initiating apoptosis in B cells. We show that α4β1 interaction with the H89 fragment prevents apoptosis induced by serum deprivation. These survival signals, however, do not seem to affect the Fas or surface IgM cross-linking apoptotic pathways.

MATERIALS AND METHODS

Cells and cell cultures

The following human B cell lines were purchased from the American Type Culture Collection (ATCC): IM-9, RPMI 8226 and NCI-H929 (multiple myeloma); Ramos, Namalwa and Daudi (Burkitt’s lymphoma); Nalm-6 (pre-B from acute lymphoblastic lymphoma); and WIL2-NS (spherocytic anaemia). The lymphoblastic cell lines RPMI 8866 and JY were obtained from Dr Francisco Sánchez-Madrid (Hospital de la Princesa, Madrid, Spain). The EHEB cell line (chronic lymphocytic leukaemia) was purchased from the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). Cells were maintained in RPMI 1640 supplemented with 10% fetal bovine serum (FBS; ICN Pharmaceuticals, Costa Mesa, CA, USA), 2 mml-glutamine and 40 μg/ml gentamicin (Gibco-BRL, Middlesex, UK).

Antibodies

Monoclonal antibodies (MoAb) P1E6 (antiα2 integrin/ CD49b) and P1B5 (antiα3/CD49c) were purchased from Calbiochem-Novabiochem (La Jolla, CA, USA); LM609 (antiαVβ3/CD51CD61) was purchased from Chemicon International (Temecula, CA, USA); P4C2 (IgG3, antiα4/CD49d) and P1D6 (IgG3, antiα5/CD49e) have been previously described (13); HP2/1 (antiα4), TS2/16 (antiβ1/CD29, integrin activating) and Alex 1/4 (antiβ1) were a gift from Dr F. Sánchez-Madrid; Act-1 (antiα4β7/CDw49d) was a gift from Dr Douglas J. Ringler (Leukosite, Cambridge, MA, USA); CH11 (anti-Fas, apoptosis-inducer) was purchased from Coulter (Miami, FL, USA); EA-5 (anti-CD40) was purchased from Calbiochem-Novabiochem; R1/69 (anti-IgM) was from Dako A/S (Glostrup, Denmark).

Immunofluorescence analyses

Cells, 5 × 10⁵, were incubated for 30 min at 4°C with 100 μl of either purified MoAb (10 μg/ml), hybridoma supernatant (1:2 dilution) or ascitis (1:100 dilution) containing the appropriate MoAb. Cells were washed twice with cold PBS-1% BSA and resuspended in 100 μl of a 1:30 dilution of fluorescein-conjugated F(ab′)₂ fragments of rabbit antibodies to mouse IgG (Dako). After 30 min at 4°C, cells were washed twice, resuspended in 400 μl of cold PBS and analysed by flow cytometry on an EPICS-CS (Coulter).

Recombinant fibronectin fragments

The H89 Fn recombinant fragment was prepared exactly as described [9,10]. This fragment corresponds to the carboxy-terminal region of Fn and contains several sites which bind to α4β1 integrin including the high affinity ligand CS-1 [9,13]. We have shown previously that B cells attach constitutively to H89 and similar fragments via α4β1 integrin in a dose-dependent manner [10,14]. The FN-III4-5 fragment was prepared exactly as previously described [15]. This fragment corresponds to the heparin binding domain III (Hep III) of Fn and contains the H2 site which is a ligand for activated α4β1 and α4β7 integrins [16].

Cell adhesion assays

Flat-bottom eight-well strips with N-oxysuccinimide amine binding surface (Costar Co., Cambridge, MA, USA) were coated overnight at 4°C with increasing concentrations of FN-III4-5 fragment in 0·1m sodium borate pH 8·5. After washing and further coating with 3% bovine serum albumin (BSA), 5 × 10⁴ cells in RPMI-1% BSA were added to each well and incubated for 1–3h at 37°C. Attached cells were stained with 0·1% toluidine blue and quantified by determining the absorbance at 620 nm on a Multiskan Bichromatic plate reader (Labsystems, Helsinki, Finland) as described previously [10]. Integrin activation with TS2/16 MoAb or Mn²⁺ was performed by incubating the cells with 1:10 dilution of hybridoma supernatant or 10 mm Tris, 150 mm NaCl, 1% BSA, 1 mm Mn²⁺, pH 7·2 for 15 min at 37°C prior to the attachment assay.

Cell viability assays

Twenty-four-well flat bottom plates (Becton Dickinson, Meylan Cedex, France) were coated overnight at 4°C with either H89, FN-III4-5 (9·8 μg/ml) or poly lysine (p-Lys, 38 μg/ml) dissolved previously in sterile 0·1m sodium borate pH 8·9. Wells were washed with PBS and coated with 1% BSA for 1h at 37°C; 2·5 × 10⁵ cells in 500 μl of RPMI-10% FBS were then added and viability analysed for the different conditions studied using FITC-Annexin V (Bender MedSystems, Vienna, Austria) and/or propidium iodide (PI). For serum-removal experiments, cells were suspended in RPMI or RPMI-10% FBS (control) and viability was measured after 3 days in culture. For apoptosis induced via Fas, cells were incubated with 0·1 μg/ml CH11 MoAb and viability measured after 24 h. In some experiments, cells were first treated with the anti-CD40 MoAb EA-5 (0·1 μg/ml) or with 0·2 μg/ml cycloheximide for 24 h prior to the addition of CH11 and further incubated for 24h. For apoptosis induced via surface IgM, cells treated with 5 μg/ml of R1/69 MoAb for 24h and viability determined as explained. Treatment with P4C2 and P1D6 MoAb was performed by first incubating the cells in 0·5 ml of RPMI with 3 μg of either MoAb for 30 min at room temperature. Cells were then diluted to the appropriate concentration and added to wells coated with p-Lys or Fn fragments, leaving the antibody present for the duration of the assay.

Statistical analyses

Significance of the difference between means was determined by Student’s t-test for non-paired samples using the GraphPad Instat v2·04a program (GraphPad Software, San Diego, CA, USA). Two-tailed statistical significances were determined. A P-value of = 0·05 was considered significant.

RESULTS

Expression of integrins on the cell lines studied

We first analysed the pattern of integrin expression on several B cell lines established form different haemopoietic malignancies. As shown in Table 1, all cells expressed α4 and with the exception of JY, they also expressed the β1 integrin subunit. β7, another partner of α4 integrin, was present in all cell lines except in Ramos and Namalwa. Other integrins were absent (α2) or showed variable expression (α3, α5, αVβ3) among the different cell lines, thus confirming that α4 is a major integrin in B cells. To determine whether α4 integrin could influence the apoptotic response of B cells, we studied the effect of cell adhesion to the recombinant Fn fragment H89, a well-known ligand for α4β1 integrin [9,10,14] in three different B-cell apoptotic pathways: serum deprivation, treatment with anti-IgM or treatment with anti-Fas MoAbs.

Table 1.

Expression of integrin subunits on the cell lines studied¹

Cell line	C (–)	α2	α3	α4	α5	α4β7	αVβ3	β1
IM-9	0·17	0·19	1·53	2·42	0·37	2·02	3·27	4·51
RPMI 8226	0·30	0·32	1·35	0·72	0·63	4·19	2·12	1·51
NCI-H929	0·32	0·47	1·09	8·54	5·41	0·62	0·53	14
Ramos	0·33	0·29	1·02	4·12	0·35	0·30	0·27	5·68
Daudi	0·27	0·36	0·30	1·43	0·31	6·98	0·38	2·58
Namalwa	0·20	0·20	1·45	3·20	0·57	0·35	0·25	3·73
Nalm-6	0·30	0·29	0·35	6·35	1·49	0·56	0·41	12
WIL2-NS	0·20	0·21	0·22	14·4	0·26	9·72	2·02	9·56
RPMI 8866	0·27	0·27	0·26	6·93	0·26	7·68	1·34	0·55
JY	0·30	0·30	0·30	8·20	0·40	6·80	1·40	1·3
EHEB	0·25	n.d.	2·46	8·54	1·40	n.d.	n.d.	6·36

¹Values indicate mean fluorescence intensity determined by flow cytometry. Results from a representative experiment out of two performed are shown.

α4β1 engagement prevents B cell apoptosis triggered by serum deprivation in some B cells

To establish the best conditions to study induction of B-cell apoptosis by serum removal, we performed kinetic experiments on all cell lines. Cells were cultured in the presence of 10% (control) or 0% FBS-containing medium and samples were taking after 1, 2 and 3 days. Viability was determined by flow cytometry after treatment with annexin-V and/or PI. As shown in Fig. 1, serum withdrawal affected cells differently and three groups were observed after 3-day cultures: the first group comprised sensitive cells (Ramos, RPMI 8866, JY and NCI-H929) with viability values of approximately 25%; the second group (Daudi, Namalwa, IM-9, RPMI 8226, Nalm-6) showed intermediate viability (about 50%) while the third group included resistant cells (WIL2-NS and EHEB) with more than 50% viability at this time. Apoptosis of the most resistant cell line EHEB was monitored up to 6 days when viability decreased to 32% (Fig. 1). For subsequent experiments cells were analysed after 2 (Group 1) or 3 days (Groups 2 and 3) in culture with the exception of Ramos and EHEB cells which were studied after 1 or 4–5 days, respectively.

Fig. 1.

Kinetics of apoptosis induced by serum removal on B cell lines. 5 × 10⁵ cells/ml were cultured on 24-well plates in RPMI 1640 in the presence (closed circles) or absence (open circles) of 10% FBS. At the indicated times, samples were taken and cell viability analysed by flow cytometry using annexin-V and PI. Values represent the average of three different experiments with <10% variablity.

All cells listed in Table 1 adhered to the H89 Fn fragment in a dose-dependent manner and adhesion was inhibited by the antiα4 MoAb HP2/1 (not shown). To test whether B-cell adhesion to the H89 fragment affected the observed induction of apoptosis, cells were cultured on H89-coated wells in the presence or absence of FBS and their viability measured at the appropriate time. As control, cells were also cultured on p-Lys which promotes attachment but not signalling. In the presence of 10% FBS all cell lines had similar viability when cultured on H89 or p-Lys, indicating that p-Lys alone did not induce apoptosis (Fig. 2). In the absence of serum viability dropped in all cells as expected but in three cell lines: Ramos, EHEB and Nalm-6, culturing on H89 resulted in protection from apoptosis yielding higher mean viability values than those of cells cultured on p-Lys (Fig. 2). The increased cell viability observed on H89-cultures was statistically significant for Ramos (difference in means = 25·69; lower 95% confidence interval (CI), 35·43–57·71; upper 95% CI, 48·53–77·63, P = 0·0002), Nalm-6 (difference in means = 22·38; lower 95% CI, 30·9–56·24; upper 95% CI, 53·87–73·31, P = 0·0024), and EHEB (difference in means = 18·79, lower 95% CI, 39·66–59·4; upper 95% CI, 57·29–75·13, P = 0·0021. This protective effect was not observed on other cell lines where viability was similar for H89-or p-Lys-cultures (not shown).

Fig. 2.

Effect of adhesion to the H89 Fn fragment on B cell apoptosis induced by serum deprivation. Ramos, Nalm-6 or EHEB cells (5 × 10⁵/ml) were cultured on (□) p-Lys (38 μg/ml) or () H89 (9·8 μg/ml)-coated wells in the presence or absence of 10% FBS. After the appropriate time cell viability was determined as explained. Values are the average of six (Ramos), seven (EHEB) or eight (Nalm-6) different experiments each performed in triplicate. **P = 0·005; ***P = 0·0005.

To confirm that the antiapoptotic effect was induced via α4 integrin cell viability assays were performed in the presence of the function-blocking P4C2 antiα4 MoAb. The antiα5 P1D6 MoAb, of the same isotype as P4C2, was used as control. As shown in Fig. 3 for Ramos and Nalm-6 cells, P4C2 prevented the protective effect mediated by adhesion to H89 and viability values were similar to those of p-Lys cultures. Incubation with P1D6 had no effect in these assays. Both P4C2 and P1D6 MoAbs had no effect on the viability of control cells incubated in 10% FBS-containing medium (not shown).

Fig. 3.

The P4C2 antiα4 MoAb blocks the antiapoptotic effect induced by α4 ligation. Cells in 0·5 ml RPMI were incubated with 3 μg of either P4C2 or P1D6 MoAbs for 30 min at room temperature. Cells were then diluted with RPMI to 5 × 10⁵/ml, added to p-Lys or H89-coated wells and incubated for 1 (Ramos) or 3 (Nalm-6) days. Viability was measured by PI staining. Values are the average of three different experiments each performed in triplicate. □, pLys;H8 9.

We recently reported that the region of Fn encompassing the Hep III domain and represented by the recombinant fragment FN-III4-5 is a ligand for activated α4β1 [15,16]. To test whether α4 interaction with this ligand also protects B cells from apoptosis we first performed adhesion assays using the three cell lines that responded to α4-mediated signalling. As shown in Fig. 4, upon activation with TS2/16 MoAb or Mn²⁺, Ramos, EHEB and Nalm-6 cells attached to the FN-III4-5 fragment in a dose-dependent manner. Incubation of these cells on FN-III4-5 fragment resulted in statistically significant higher viability for Ramos (difference in means = 20·72; lower 95% CI, 36·41–56·35; upper 95% CI, 43·45–69·94, P = 0·0001), Nalm-6 (difference in means = 21·48; lower 95% CI, 44·23–62·94; upper 95% CI, 56·84–81·1, P = 0·0005) and EHEB (difference in means = 11·97; lower 95% CI, 36·87–47·91; upper 95% CI, 51·9–64·82, P = 0·025) than culturing on p-Lys (Fig. 5). As observed for H89, incubation with P4C2 MoAb abolished the antiapoptotic effect of adhesion to FN-III4-5 (shown in Fig. 5B for Nalm-6 cells). These results indicate that the FN-III4-5 region also protects B cells from apoptosis upon binding to α4β1 integrin.

Fig. 4.

Cell adhesion to FN-III4-5 fragment. 5 × 10⁵ cells/ml were treated with TS2/16 MoAb (closed circles) or 1 mm Mn²⁺ (open circles) for 15 min at 37°C and added to 96-well plates coated previously with the indicated concentrations of FN-III4-5 or H89 fragments. After 3h incubation at 37°C attached cells were quantified after staining with 0·1% toluidine blue as described. Values were determined in duplicate and are the average of three different experiments.

Fig. 5.

Adhesion to FN-III4-5 fragment protects B cells from serum removal-induced apoptosis. (a) Cells (5 × 10⁵/ml) were treated with TS2/16 MoAb for 15 min and incubated in 24-well plates coated previously with p-Lys, H89 or FN-III4-5 (9·8 μg/ml) with or without 10% FBS. At the appropriate time cell viability was determined by staining with PI. Values represent the average of seven (Ramos, Nalm-6) or eight (EHEB) different experiments each performed in triplicate. *P = 0·05; ***P = 0·0005. (b) Nalm-6 cells were incubated with P4C2 or P1D6 MoAbs as explained in the legend to Fig. 3. Viability values are the average of two different experiments each performed in triplicate. □, pLys; , FN4-5.

α4β1 binding to the H89 fragment does not affect anti-IgM-induced B cell apoptosis

We also studied the role of α4β1 signalling on apoptosis induced via surface IgM, another effective apoptotic pathway on B cells. Among the three cell lines that were protected from serum removal-induced apoptosis by ligation of α4β1 integrin, only Ramos showed surface expression of IgM and therefore these cells were chosen for these studies. Ramos cells showed a dose-dependent apoptotic response to treatment with anti-IgM MoAb in the range of 0·5–5 μg/ml (not shown). Ramos cells were incubated on H89-or p-Lys-coated wells for 24h in the presence or absence of 5 μg/ml anti-IgM MoAb, a concentration that produced 50% cell apoptosis, and the viability measured by PI staining. Figure 6 shows that adhesion to the H89 fragment via α4β1 integrin had no effect on the apoptotic pathway triggered by the anti-IgM MoAb, in contrast to the observed α4β1-mediated survival effect when apoptosis was induced on these cells by serum removal.

Fig. 6.

Effect of adhesion to the H89 fragment on B cell apoptosis induced by anti-IgM MoAbs. Ramos cells (2·5 × 10⁵/ml) were cultured on wells coated previously with p-Lys or H89 for 1 h, followed by the addition of R1/69 MoAb. Cell viability was determined after 24h by PI staining. Mean values from four different experiments are shown. □, Control;, anti-IgM.

α4β1 integrin does not protect B cells from apoptosis induced via Fas

We next studied whether adhesion to the H89 fragment via α4β1 integrin would prevent B cell apoptosis triggered by the Fas receptor. Analysis by flow cytometry on the three previously selected B cell lines showed that only Ramos expressed surface Fas while Nalm-6 and EHEB were negative (Table 2). To test whether Fas was functional on Ramos, cells were incubated in suspension with the apoptosis-inducer CH11 anti-Fas MoAb and viability measured after 24h by PI staining. As shown in Fig. 7a, CH11 did not induce apoptosis on Ramos cells even after treatment with an anti-CD40 MoAb (Fig. 7a) or addition of cycloheximide (not shown).

Table 2.

Expression of Fas on B cells^*

Cell line	Control (–)	Fas
Ramos	0·20	1·79
Nalm-6	0·20	0·20
EHEB	0·28	0·31
IM-9	0·20	3·30
RPMI 8226	0·20	4·20

^*Values indicate mean fluorescence intensity.

Fig. 7.

(a) Apoptosis induced via Fas on B cells. The indicated cell lines (2·5 × 10⁵/ml) were incubated in suspension for 24 h with either no antibody (control) or the EA-5 anti-CD40 MoAb (0·1 μg/ml). The CH11 anti-Fas MoAb (0·1 μg/ml) was then added and cells further incubated for another 24-h period. Cell viability was measured with PI. □, No MoAb; , CH11; , aCD40/CH11. (b) Effect of α4β1 engagement on B cell apoptosis induced via Fas. IM-9 cells (2·5 × 10⁵/ml) were incubated on p-Lys or H89-coated wells in the presence or absence of CH11 MoAb. After 24h, viability was measured by staining with PI. Values were determined in duplicate and are the average of two different experiments. □, Control; , CH11.

Fas analysis were then extended to other B cells. As shown in Table 2, the myeloma-derived IM-9 and RPMI 8226 cell lines showed positive Fas surface expression. At difference with the Burkitt lymphoma cell line Ramos, CH11 induced apoptosis on both RPMI 8226 and IM-9 cells and this effect was not significantly increased by an anti-CD40 MoAb (Fig. 7a).

The possible protecting effect of α4β1 was therefore studied on cells which carried functional Fas receptors. IM-9 cells were incubated on H89 fragment-or p-Lys-coated wells during treatment with the CH11 MoAb; after 24h viability was measured by PI staining. As shown in Fig. 7b, adhesion to the H89 Fn fragment did not prevent the apoptotic effect of CH11 as viability of these cells (55·2%) was similar to that of cells incubated on the control p-Lys substrate (49·7%). Similar results (49% on H89 vs. 45% on p-Lys) were obtained with RPMI 8226 cells (not shown).

DISCUSSION

In this report, we have studied the role of α4β1 integrin/ CD49d/CD29 in the regulation of apoptosis in B cells. We show that ligation of α4β1 rescued B cells from serum deprivation-induced apoptosis but had no effect on anti-Fas or anti-IgM-initiated apoptotic pathways.

All B cell lines studied expressed functional α4 integrins as demostrated by flow cytometry and cell adhesion studies. Culturing B cells on the H89 Fn fragment in the absence of serum, clearly resulted in statistically significant protection from apoptosis in three cell lines: Ramos, Nalm-6 and EHEB. This lower percentage of apoptosis was not due to cell proliferation since cells did not proliferate in the absence of serum (results not shown). The α4-mediated survival effect was not observed on other cell lines, suggesting that the ability to respond to α4 signals may be intrinsic to the cell type. Alternatively, the regulatory role of α4 in apoptosis may be important at certain stages of B cell differentiation. In this regard, Nalm-6 (pre-B) and Ramos (Burkitt’s lymphoma) represent lymphoblastic cells which have been used as models for germinal centre B cells, where apoptosis of non-selected B cells is an active process [4]. Indeed interaction of α4β1 and LFA-1 integrins with their respective ligands VCAM-1 and ICAM-1 prevent apoptosis of germinal centre B cells [2]. EHEB cells on the other hand were established from a chronic lymphocytic leukaemia patient and represent mature B cells. We have recently reported that attachment of fresh B-CLL cells to the H89 fragment prevented their spontaneous apoptosis in vitro and thus could constitute one of the mechanisms which maintain abnormal survival of these cells in vivo[10]. Our present results with the EHEB cell line therefore confirm these previous studies and establish that in certain mature B cells α4 integrin effectively induces survival signals.

Interestingly, attachment to the FN-III4-5 fragment also protected B cells from apoptosis in the absence of serum. This fragment represents a region of Fn which we recently identified as a ligand for activated α4 integrins [15,16]. We now report for the first time that α4 interaction with the FN-III4-5 ligand is able to deliver intracellular survival signals in B cells. This is therefore a novel function for this region of Fn not characterized previously.

Apoptosis was also effectively induced in Ramos cells by monoclonal antibodies to surface IgM and cross-linking of α4β1 integrin by the H89 fragment ligand did not prevent or reduce this effect. The fact that on the same cells α4 integrin ligation was able to overcome the lack of growth factors (serum removal) suggests that both apoptotic pathways follow different mechanisms and involve different regulatory molecules. Alternatively, it is possible that α4 alone is not sufficient to protect from apoptosis in this case and requires other signals. It is well known, for example, that CD40 MoAbs, r-TNF-α and IFN inhibit IgM-induced apoptosis in some Burkitt’s lymphoma cells [[5,,17,18,19]. Interestingly, the protective effect of CD40 appeared to be enhanced by co-ligation of β2/CD18 integrins in the DND-39 cell line [17]. Although in the present study we did not test the possibility of a similar cooperative effect between CD40 and β1/CD29 integrins, our results clearly show that signals induced by α4β1 engagement exclusively may substitute for the lack of growth factors but are unable to overcome the IgM-induced apoptotic pathway.

Triggering the death receptor Fas/Apo-1 is an important apoptotic mechanism operating mainly on mature T and B lymphocytes [1,20]. In the present study we have found heterogeneous expression and function of this receptor depending on the origin of the cell line. Albeit Fas was present on all myeloma and Burkitt’s lymphoma cells tested, an anti-Fas MoAb induced apoptosis in myeloma but not in lymphoma cells. These results are in agreement with previous reports in which anti-Fas MoAbs failed to induce apoptosis of tonsillar B cells and Ramos cells [4,5]. In these studies, however, addition of an anti-CD40 MoAb increased Fas expression and Fas-mediated apoptosis. Although in the present report ligation of CD40 also increased Fas expression (results not shown), we did not observe apoptosis on lymphoma cells upon anti-Fas treatment. While we do not have an explanation for this discrepancy, it is possible that the different experimental conditions used in those previous studies and ours (48h versus 24h treatment) account for the different results.

Regardless of the lack of functionality of Fas on lymphoma cells observed in our study, attachment of myeloma cells carrying functional Fas receptors to the H89 fragment did not affect the levels of apoptosis induced by anti-Fas MoAbs. This indicates that survival signals initiated via α4β1 integrin alone are apparently unable to overcome the Fas apoptotic pathway in myeloma cells. However, this may not apply to all cell types and all integrins since it was shown recently that ligation of α2β1 integrin by its ligand collagen or by antibodies, down-regulated expression of Fas ligand and prevented apoptosis of Jurkat T cells [21].

It has been suggested that the IgM and Fas apoptotic pathways operate at different times of B cell differentiation with IgM acting at initial stages and Fas at more mature stages [5]. Our present results indicate that α4β1 integrin engagement alone apparently has no effect on either pathway. α4β1, however, can substitute or collaborate with growth factors during normal B cell development; α4β1 could also contribute to malignant cell survival in environments lacking proper B cell growth factors and thus affect progression of the haematological disorder.