SEARCH

SEARCH BY CITATION

Keywords:

  • GM-CSF;
  • apoptosis;
  • proliferation;
  • malignant plasma cells;
  • cytokine signalling

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. References

It has been shown that granulocyte/macrophage colony stimulating factor (GM-CSF) is able to support myeloma cell propagation in cooperation with interleukin (IL)-6, the major growth factor for malignant plasma cells, although the biological mechanisms involved remain unknown. Therefore we investigated (i) the expression levels of the GM-CSF receptor (GM-CSFR) constituents in three malignant plasma cell lines and in native malignant plasma cells, (ii) the ability of the receptor to mediate common signalling pathways regulating proliferation and cell survival in malignant plasma cell lines, and (iii) the effects of GM-CSF on tumour cell biology.

The GM-CSFRα subunit was detected in the malignant plasma cell lines RPMI-8226, MC/CAR, IM-9 as well as 6/6 native myeloma cell samples derived from the bone marrow of patients with overt disease. Furthermore, GM-CSFR expression was also detected in the CD19+ fraction from 2/3 bone marrow samples and 5/8 peripheral blood samples derived from patients with malignant plasma cell disorders, but not in the CD19+ fraction of peripheral blood from healthy donors. The expressed cytokine receptor α-subunit was able to constitute a functional signalling complex with the ubiquitously expressed GM-CSFRβ subunit, as demonstrated by the fact that GM-CSF induced the p21-ras/mitogen-activated protein kinase (MAPK) signalling cascade in malignant plasma cell lines. Since this signalling cascade plays an essential role in the mediation of both proliferation and cell survival, we investigated the impact of GM-CSF on these two events. Application of GM-CSF led to an increase of DNA-synthesis in MC/CAR, IM-9 and RPMI-8226 cells. Furthermore, it increased longevity of these malignant plasma cell lines by reducing the rates of spontaneous apoptosis. We conclude that (i) the functional GM-CSFR is commonly expressed on malignant plasma cells and that (ii) GM-CSF promotes the clonal expansion of myeloma cells by inhibiting spontaneous apoptosis and promoting DNA synthesis.

The GM-CSF receptor (R) consists of two subunits, a unique α-chain determining cytokine specificity and the ubiquitously expressed β-chain, necessary for signal transduction, which is also present in IL-3 and IL-5 receptor complexes ( Miyajima et al, 1992 ). The GM-CSFR is primarily expressed on myeloid cells during maturation. In response to GM-CSF, it not only mediates proliferation but also activates myeloid cell types to differentiate ( Miyajima et al, 1992 ). Although its function in the myeloid lineage is well known, there is little information about the effects of GM-CSF on the lymphoid lineage of the haemopoietic system.

In neoplastic plasma cells, IL-6 is known to be the major tumour growth factor in vitro and in vivo, acting via auto- and/or paracrine mechanisms on the proliferation of myeloma cells ( Kawano et al, 1989 ). Apart from IL-6, other cytokines such as IL-1, IL-3, tumour necrosis factor (TNF) and GM-CSF have been described as supporting proliferation of native malignant plasma cells and/or myeloma cell lines ( Klein et al, 1989 ; Kawano et al, 1989 ; Bergui et al, 1989 ; Zhang et al, 1990 ). As possible mechanisms for the supportive effects, up-regulation of IL-6 expression in the bone marrow microenvironment ( Klein et al, 1989 ; Kawano et al, 1989 ), synergistic action of these cytokines with IL-6 on the proliferation of tumour cells ( Bergui et al, 1989 ), or, as described for GM-CSF, the interference with IL-6-dependent signalling ( Zhang et al, 1990 ) have been discussed. In fact, it was reported that anti-IL-6 mAbs neutralized the stimulatory effect of GM-CSF on DNA synthesis in myeloma cells ( Zhang et al, 1990 ). This was demonstrated in native malignant plasma cells short-term cultured in their bone marrow microenvironment, as well as in the IL-6-dependent XG-1 myeloma cell line. Therefore an IL-6-mediated mechanism of GM-CSF-action on myeloma cell growth was postulated ( Zhang et al, 1990 ). In addition, GM-CSF not only supported the IL-6-induced proliferation of freshly explanted malignant plasma cells, but, together with IL-6, was also indispensable for the establishment of continuously growing myeloma cell lines ( Zhang et al, 1994 ).

However, a direct biological effect of GM-CSF per se on the expansion of neoplastic plasma cells has not yet been demonstrated. In additon to the above-mentioned impact of the cytokine on development of myeloid cells and, together with IL-6, on proliferation of malignant plasma cells, inhibition of apoptosis by GM-CSF was recently reported as occurring in different myeloid cell types ( Iversen et al, 1996 ; Hara & Miyajima, 1996). This anti-apoptotic effect seemed to depend on a functionally intact ras/MAPK signalling cascade ( Kinoshita et al, 1995 ). These observations raise the question of a potential role of GM-CSF, not only in the survival of myeloid cells ( Iversen et al, 1996 ; Hara & Miyajima, 1996) but also in the mediation of longevity in malignant cells of the B-cell lineage, although no direct evidence for this biological function in myeloma cells has been presented yet.

In the present study we demonstrated that a functionally active GM-CSFR complex was expressed in neoplastic plasma cells, and that GM-CSF by itself was able to induce an increase in [3H]thymidine incorporation in malignant plasma cells and to sustain their longevity. The effects of GM-CSF correlated with the activation of p21-ras proteins and the MAPK signalling cascade. However, the cytokine had no effect on programmed cell death induced by cytotoxic drugs employed in the treatment of this disease such as the anthracycline doxorubicin or the steroid dexamethasone.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. References

Cell culture

The neoplastic plasma cell lines RPMI-8226, IM-9, MC/CAR, the promyelocytic leukaemia line HL-60 and the T-acute lymphatic leukaemia cell line CEM-C7H2 were used in this investigation. The IM-9 myeloma/lymphoblastoid cell line was obtained from the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany), the MC/CAR plasmacytoma cell line was obtained from the American Type Culture Collection (ATCC, Rockville, Md., U.S.A.) and the RPMI-8226 myeloma line was kindly provided by Dr T. Otani from the Fujisaki Cell Centre, Okayama, Japan. The promyelocytic HL-60 cell line was a kind gift of Dr G. Konwalinka (Department of Internal Medicine, University of Innsbruck Medical School, Austria) and served as a GM-CSFR-positive control. The human T-acute lymphatic leukaemia cell line CEM-C7H2 served as a positive control for drug-induced apoptosis ( Villunger et al, 1997a ). All the cell lines were cultured in RPMI 1640 media (Seromed, Berlin, Germany) supplemented with 10% (v/v) heat-inactivated FCS (Biological Inc., Beth Haemek, Israel), 2 mmol L-glutamine (Seromed, Berlin, Germany), 100 μg/ml gentamicin (GIBCO, Grand Island, N.Y.) at 37°C in a humidified atmosphere containing 5% CO2.

cDNA synthesis and reverse transcription (RT)-PCR

cDNA synthesis and RT-PCR were performed as previously described ( Villunger et al, 1997b ). PCR was performed using Ampli-Taq-Gold DNA-Polymerase (Hoffmann La Roche, Vienna, Austria) for the amplification of the GM-CSFRα subunit (50 cycles) and Taq-DNA polymerase (Boehringer Mannheim, Mannheim, Germany) for all other amplifications (30 cycles) in the relevant buffer systems with 0.2 m M dNTP mixture (Boehringer Mannheim) containing 1 μM of the 5′/3′primers (Microsynth, Balgach, Switzerland). The annealing temperature for the primers used was 60°C for β-actin and GM-CSFRα and 56°C for GM-CSFβ and bcl-xL. β-actin, product size 838 bp: upper 5′ ATC TGG CAC CAC ACC TTC TAC AAT GAG CTG CG 3′; lower 5′ CGT CAT ACT CCT GCT TGC TGA TCC ACA TCT GC 3′; bcl-xL, product size 780 bp: upper 5′ TTG GAC AAT GGA CTG GTT GA 3′; lower 5′ GTA GAG TGG ATG GTC AGT G 3′; GM-CSFRα, product size 682 bp: upper 5′ AGA AAT CGG ATC TGC GAA CAG TGG CAC C 3′; lower 5′ TCC AGG TAC GAC AGC TTC TGA TAG GTC C 3′; GM-CSFRβ, product size 572 bp: upper 5′ CTA CAA GCC CAG CCC AGA TGC 3′; lower 5′ ACC CGT AGA TGC CAC AGA AGC 3′.

Determination of GM-CSFRα expression on native neoplastic plasma cells

Bone marrrow (BM) samples from six patients with multiple myeloma (for clinical data see Table I) and peripheral blood samples from eight patients with multiple myeloma or plasmacytoma, four patients with monoclonal gammopathy of undetermined significance (MGUS) and four healthy donors were analysed. Samples were collected during routinely scheduled examinations after informed consent had been obtained. BM samples were pressed through a 0.2 × 20 mm syringe and diluted in RPMI-1640 medium 1:1 (v/v). BM samples or peripheral blood samples were applied [2:1 (v/v)] onto a FICOLL density gradient (LymphoprepTM, Nycomed, Oslo, Norway) and centrifuged with 500 g for 30 min at room temperature. The mononuclear cell fraction was carefully aspirated, washed twice with PBS and cells (0.5 × 106) were used for GM-CSFRα and isotype staining. After GM-CSFRα staining, the cells were washed three times in PBS and thereafter subjected to CD38 or CD19 immunostaining (30 min on ice) using 20 μl of a PE-conjugated mouse anti-human CD38 mAb or CD19 mAb, respectively (Pharmingen, San Diego, Calif., U.S.A.). The plasma- and B-cell populations were gated according to their unique position in the correlation of forward light scattering, orthogonal light scattering, and immunofluorescent-labelled CD38 ( Villunger et al, 1997b ; Terstappen et al, 1990 ) or CD19, respectively. B cells or plasma cells, stained with the GM-CSFRα mAb, were compared with samples stained with the relevant GM-CSFRα isotype control for determining GM-CSFRα expression on the different B-cell subtypes.

Table 1. Table I. Patient characteristics.* * Abbreviations: BM: bone marrow; PB: peripheral blood; MM: multiple myeloma; PC: plasmacytoma; MGUS: monoclonal gammopathy of undetermined significance. Paraprotein was determined by immune electrophoresis of patients' serum samples during routinely scheduled examinations. Bone marrow plasma cell content (%) was determined by morphologic analysis of cells on smears of BM aspirates.Thumbnail image of

Immunofluorescence

Cells (0.2 × 106) were washed twice in PBS and incubated with 1 μg anti-GM-CSFRα mAb (Genzyme, Cambridge, Mass., U.S.A.) for 60 min at 4°C while being gently shaken. Cells were washed twice with PBS, and incubated with 10 μl of FITC-labelled goat anti-mouse antibody F0479 (Dako, Copenhagen, Denmark) for a further 30 min at 4°C. The pellets were washed twice, resuspended in 100 μl PBS/1% BSA and analysed immediately thereafter by FACScan (5000 cells per sample). Negative controls were carried out simultaneously using a mouse anti-human IgG2a mAb (Dako, Copenhagen, Denmark) instead of the GM-CSFRα mAb. The mean specific fluorescence intensities (MSFIs) were calculated as the ratio [mean fluorescence channels of the specific mAb/isotype-matched control mAb].

Measurement of [3H]thymidine incorporation

To determine the influence of GM-CSF (Leukomax®, Sandoz, Austria) on the rate of DNA synthesis, 2 × 105 cells/ml were grown in culture medium containing 10% FCS supplemented with 10 ng/ml of the cytokine and pulse-labelled with 7.4 kBq/well [3H]thymidine (specific activity 74 GBq/mmol; Amersham, Bucks., U.K.) for 8 h before harvesting. In order to prove specificity of the cytokine effect, preincubation of cells with 5 μg/ml m(Abs) against the IL-6R (R&D Systems, Minneapolis, Minn., U.S.A., no. AB227 NA) or against the GM-CSFR (Genzyme, Cambridge, Mass., U.S.A. no. 80297701) was performed 60 min prior to stimulation with the cytokine. After 72 h cells were harvested automatically, transferred onto filter papers and washed three times. The incorporated radioactivity was measured with a β-scintillation counter (Beckmann).

Loss of DNA content in single cells (PI-test)

Cells (2 × 105/ml) were incubated with and without 10−6M dexamethasone (Sigma, St Louis, Mo.) or 0.1 μg/ml doxorubicin (EBEWE, Unterach, Austria) in the presence and absence of GM-CSF (10 ng/ml), respectively. Cells were harvested after 96 h of cultivation, washed by centrifugation in PBS and incubated with 300 μl/well propidium iodide (PI)-solution (50 μg/ml PI, 0.1% sodium citrate, and 0.1% Triton X-100) for permeabilization and DNA-staining. Cells were subjected to analysis of cell size and fluorescence intensity in the forward/side scatter program of a FACScan as previously described ( Villunger et al, 1997b ). Apoptosis in PI assays was defined as a decrease in DNA staining and cell size of positive cells as compared to malignant plasma cells and of normal T lymphocytes in G1 phase.

Determination of the ras GTP/GDP ratio

IM-9 cells (5 × 106/well) were washed once in PBS and were metabolically labelled in phosphate-free DMEM media supplemented with 0.5% FCS and 3.7 MBq/ml [32P]orthophosphate (specific activity 333 TBq/mmol; NEN-DuPont, France) for 16 h. Cells were stimulated with 10 ng/ml GM-CSF for 0, 5, 15 and 30 min, washed once in PBS, resuspended in 500 μl lysis buffer (50 m M Tris.Cl, pH 7.5, 20 m M MgCl2, 150 m M NaCl, 1 m M Na3VO4, 20 μg/ml aprotinin and leupeptin, 1% NP-40) and incubated on ice for 30 min. Cellular debris was removed by centrifugation and ras proteins were precipitated with 10 μl of agarose-conjugated anti-pan ras Y13-259 Ab (Oncogene Science, Cambridge, Mass., U.S.A.). 10 μl of rat serum (Serotec, U.K.) served as a negative control. Immunocomplexes were precipitated with 20 μl of Staphylococcus aureus crude extract (Sigma, St Louis, Mo., U.S.A.) and washed three times with lysis buffer without NP-40. Guanine-nucleotides were eluted from the ras proteins by addition of 12 μl Tris.Cl (20 m M) and 4 μl of sodium formiat (4 M) and incubating for 15 min on ice. 6 μl of the samples were separated by thin-layer chromatography on polyethyleneimine-cellulose sheets (Sigma) in 1 M KH2PO4 pH 4.0 at room temperature. Autoradiographs were scanned using Cream® ICT image analysis software.

Measurement of MAP-kinase activity

The in vitro complex kinase assay was performed as previously described ( Egle et al, 1996 ). In brief, cells (2 × 106/ml) were cultured in 0.5% FCS for 24 h, incubated with 10 ng/ml GM-CSF, lysed in 50 m M Tris.Cl pH 7.3, 50 m M NaCl, 5 m M Na2P2O7, 5 m M EDTA, 5 m M NaF, 2% NP-40, 20 mg/ml aprotinin and leupeptin, 5 m M Na3VO4. After clearance of samples by centrifugation, supernatants were incubated with 2 μg/ml of Abs specific for either MAPK1 or MAPK2, respectively (Santa Cruz, St Cruz, Calif., U.S.A., no. SC93/C16, no. SC154/C14) for 2 h at 4°C. The immunocomplexes were precipitated with Staphylococcus aureus crude extract. Samples were washed three times in lysis buffer, followed by three washes in 25 m M Hepes, pH 7.5, 20 m M MgCl2, 2 m M MnCl2 and incubated in 20 μl assay mix [13 μl kinase buffer, 9 μg myelin basic protein (MBP), 10 μM ATP, 7.4 kBq [γ32P]ATP, specific activity 111 TBq/mmol] for 30 min at 30°C. The reaction was stopped by the addition of 6 μl of 5 × SDS-PAGE sample buffer. After SDS/PAGE and blotting, the nylon membranes were exposed to Hyperfilm MP (Amersham, Bucks., U.K.). The application of equal amounts of MAPK-protein loaded on each lane was verified by immunological detection using a polyclonal Ab reacting with both MAPKs 1 and 2 (Zymed, San Francisco, Calif., U.S.A., no. 61-7400). For the evaluation of the MAPK activation status RPMI-8226 cells were stimulated with GM-CSF (10 ng/ml) for up to 30 min, washed in PBS and resupended in 50 μl MAPK lysis buffer. Protein concentrations were determined, equal amounts of proteins were loaded onto the gel and after SDS-PAGE subjected to immunoblotting. The phosphorylation status of MAPK1 and 2 were assessed using an anti-phospho-p42/p44 MAP kinase (Thr 202/Tyr204) Ab (Promega, Madison, Wis., no. NEB-9101) diluted 1:500 in TBSTM. Rabbit anti-mouse secondary horse-radish peroxidase (HRP)-conjugated Ab diluted 1:1000 in TBSTB (Dako, Copenhagen, Denmark) was used for detection of the relevant proteins. Detection was performed using ECL reagents and ECL films (Amersham, U.K.) according to the manufacturer's recommendations. To verify equal loading of the gel, the membrane was stripped and reprobed with a p42 MAPK2 specific mAb (St Cruz, Santa Cruz, Calif., no. D-2) diluted 1:250.

Immunoblotting

Immunoblottings were performed as previously described ( Egle et al, 1996 ). Anti-bcl-2 mAb (0.5 μg/ml) (Cambridge Research Biochemicals, London, no. ON-11925), polyclonal anti-bax Ab or anti- bcl-xL Ab (0.5 μg/ml) (Santa Cruz, St Cruz, Calif., U.S.A., no. SC493/N-20) as well as rabbit anti-mouse- and swine anti-rabbit secondary horse-radish peroxidase (HRP)-conjugated Abs diluted 1:1000 in TBSTB (Dako, Copenhagen, Denmark) were used for detection of the relevant proteins. Diaminobenzidine (Sigma, St Louis, Mo.) diluted in Tris-buffered saline (TBS) served as a substrate solution.

Statistics

Statistical calculations were performed using the statview 4.5® program. Levels of significance for the comparison of GM-CSF-effects on [3H]thymidine incorporation, apoptosis, and activation of p21-ras were assessed using Fisher's PLSD test with a 95% confidence limit.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. References

GM-CSFR constituents are commonly expressed on neoplastic plasma cell lines and native malignant plasma cells but not on B cells from healthy individuals

Expression of the GM-CSFRα- and β-subunits in malignant plasma cell lines was determined at the RNA and protein level. For this purpose, total RNA was extracted from RPMI-8226, IM-9 and MC/CAR cells and subjected to RT-PCR analysis using primers specific for the GM-CSFRα subunit (Fig 1a). Another set of primers was generated to amplify the GM-CSFRβ subunit necessary for signal transduction, which is commonly expressed within the haemopoietic system (Fig 1a). RNA derived from the HL-60 promyelocytic leukaemia cell line served as a positive control. PCR products corresponding in size to the α- and β subunit, respectively, were detected in the RPMI-8226, IM-9 and MC/CAR cell lines and matched the product detected in the HL-60 cell line in size (Fig 1a). Using flow cytometry, the expression of the GM-CSFRα subunit was also confirmed at the protein level in all cell lines ( Table IIa). Furthermore, analysis of native malignant plasma cells revealed positive results (6/6), confirming that expression of the GM-CSFRα subunit in myeloma cells occurs also in vivo (Fig 1b). We also investigated whether the CD19+ B-cell fraction which contains myeloma cell precursors ( Bergsagel et al, 1995 ) also expressed the GM-CSFRα protein. The CD19+ fraction was derived from bone marrow or peripheral blood of patients with overt disease and analysed by flow cytometry for GM-CSFRα expression ( Table IIb). The CD19+ cells of two out of three samples derived from bone marrow of the patients were GM-CSFRα positive ( Table IIb). In addition, CD19+ peripheral blood cells from five out of eight patients with myeloma or plasmacytoma expressed GM-CSFRα. In clear contrast, the CD19+ B-cell fractions derived from the peripheral blood of four healthy donors were negative (Fig 1 b; Table IIb).

image

Figure 1. 9+ fraction of normal peripheral blood lymphocytes revealed negative results.

Download figure to PowerPoint

Table 2. Table II(a). Flow cytometric* analysis of the GM-CSFRα expression in malignant plasma cell lines and HL-60 cells. Table II(b). Flow cytometric* analysis of the GM-CSFRα expression in malignant and non-malignant B cells.†,‡ * Determined by immunofluorescence as the ratio [specific mAb/isotyope-matched control mAb]. A ratio < 1.5 was considered as negative.* Determined by immunofluorescence as the ratio [specific mAb/isotyope-matched control mAb]. A ratio < 1.5 was considered as negative. Malignant plasma cells, malignant or non-malignant B cells were identified according to their unique forward scattering/side scattering profile after CD38 and CD19 immunostaining, respectively ( Terstappen et al, 1990 ). Abbreviations: BM: bone marrow; PB: peripheral blood; MM: multiple myeloma; PC: plasmacytoma; MGUS: monoclonal gammopathy of undetermined significance; n.d.: not determined.§ CD4+ cells served as a GM-CSFR negative control population.Thumbnail image of

In view of the difference between CD19+ B cells from patients with myeloma and healthy donors as far as GM-CSFRα expression is concerned, we analysed the GM-CSFR expression in blood samples from four patients with monoclonal gammopathy of undetermined significance (MGUS), a premalignant monoclonal disorder. We found an expression of GM-CSFRα in two out of four samples of CD19+ blood cells derived from these patients ( Table IIb).

GM-CSF induces activation of the ras/MAPK signalling cascade in malignant plasma cells

To verify the establishment of a functional GM-CSFR complex in malignant plasma cells after cytokine application, we tested the ability of GM-CSF to induce the receptor-linked ras/MAPK signalling cascade in IM-9 cells. For the determination of ras activity, cells were metabolically labelled with [32P]orthophosphate, stimulated with the cytokine and analysed for changes in the GDP/GTP ratios of p21-ras which reflects the activation status of small G-proteins ( Figs 2a and 2b). Treatment of IM-9 cells with GM-CSF led to an increase in the activated GTP-bound form of ras molecules (P = 0.04). Looking for molecules involved in the mitogenic signal transduction downstream of ras, we analysed the impact of GM-CSF on the activity of MAPKs 1 and 2 by monitoring changes in the in vitro substrate-phosphorylating potential of these proteins. Stimulation of IM-9 cells with GM-CSF led to a rapid increase in the phosphorylation of MBP by these two kinases ( Figs 2c and 2d). To confirm the activation of MAPKs in RPMI-8226 plasma cells, we analysed the phosphorylation status of MAPKs 1 and 2 after GM-CSF treatment using a phosphospecific MAPK antibody. Phosphorylation of MAPKs on Tyr and Thr residues, performed by the dual specific kinase MEK1, causes MAPK activation and occurs as a consequence of triggering the ras pathway ( Su & Karin, 1996). The phosphorylation pattern of MAPKs therefore reflects their activation status. Immunoblotting revealed an increase in the amount of phosphorylated MAPKs 1 and 2 in response to cytokine treatment of RPMI-8226 cells (Fig 2e and 2f), demonstrating enhanced activation of the kinases.

image

Figure 2. (hatched bars) activities.

Download figure to PowerPoint

GM-CSF increases [3H]thymidine incorporation in neoplastic plasma cells

Since activation of the mitogenic ras/MAPK signalling pathway is essentially involved in cell proliferation, we analysed the impact of GM-CSF on DNA synthesis. Cultivation for 72 h of the RPMI-8226, MC-CAR and IM-9 plasma cells in regular culture medium supplemented with 10 ng/ml GM-CSF led to a significant increase in DNA synthesis rates when compared with untreated controls (Fig 3a). To prove that the observed effects were specific for the cytokine, inhibition experiments using a neutralizing mAb against GM-CSFRα were performed in IM-9 cells. A 60 min preincubation of cells with 5 μg/ml of anti-GM-CSFRα mAb efficiently decreased the stimulatory effect of GM-CSF on [3H]thymidine incorporation (Fig 3b). It has been suggested that GM-CSF-induced DNA synthesis is dependent on a functional IL-6 signalling cascade ( Zhang et al, 1990 ). To test this hypothesis, we preincubated IM-9 and RPMI-8226 cells with an anti-IL-6R Ab, which blocked the IL-6-induced increase of DNA synthesis in IM-9 and in RPMI-8226 cells (Fig 3b and data not shown). However, application of this IL-6R Ab did not abrogate the GM-CSF-induced increase in [3H]thymidine incorporation either in IM-9- or RPMI-8226 cells (Fig 3c).

image

Figure 3. [H]thymidine for the last 6 h. Bars represent mean values ±SEM (n = 8).

Download figure to PowerPoint

GM-CSF supports longevity by inhibiting spontaneous apoptosis in malignant plasma cells

A reduced sensitivity towards programmed cell death might partially contribute to the increases in [3H]thymidine incorporation observed after cytokine treatment of malignant plasma cells. Therefore we analysed the impact of GM-CSF on apoptosis by determining the percentage of apoptotic cells using the PI assay, a well-established method for the quantification of apoptotic cells ( Nicoletti et al, 1991 ). The observed increase of DNA synthesis in MC-CAR, RPMI-8226 and IM-9 cells treated with GM-CSF was accompanied by a reduction in spontaneous apoptosis (Fig 4a). Again, specificity of this effect was proved by inhibition experiments in IM-9 cells using the anti-GM-CSFR mAb (Fig 4b). In contrast to the effect on spontaneous apoptosis, cell death induced by the anti-tumour drug doxorubicin, commonly used in the treatment of plasma cell neoplasms, was not inhibited by GM-CSF treatment of the malignant plasma cell lines ( Table III). In this context, the impact of GM-CSF on steroid-induced apoptosis was also tested, but all plasma cell lines analysed proved to be resistant to treatment with dexamethasone ( Table III). In order to determine whether the anti-apoptotic effect of GM-CSF on spontaneous apoptosis was caused by the alterations of the expression levels of bcl-2, bcl-xL or bax, immunoblottings as well as RT-PCR experiments were carried out. Analysis of the protein levels of bcl-2 and bax revealed their constitutive expression in all three cell lines investigated and stimulation of these cells with GM-CSF for up to 96 h did not significantly alter the expression levels of these two proteins. Furthermore, RT-PCR analysis of bcl-xL mRNA levels revealed no changes after treatment with GM-CSF (data not shown). However, changes in the phosphorylation status or dimerization behaviour of these proteins cannot be excluded.

image

Figure 4. ). (b) After preincubation with or without GM-CSFR mAb (5 μg/ml) for 60 min, IM-9 cells (2 × 105/ml) were cultivated in the presence of 10 ng/ml of GM-CSF (solid bars) for 96 h and rates of spontaneous apoptosis were determined using the PI assay and compared with untreated controls (open bar). Bars represent mean ±SD (n = 2).

Download figure to PowerPoint

Table 3. Table III. Influence of GM-CSF treatment of malignant plasma cell lines on drug-induced apoptosis.* * Cells were incubated with or without GM-CSF (10 μg/ml) and dexamethasone 10−6M or doxorubicin (0.1 μg/ml) for 72 h. The percentage of apoptotic cells was determined using the PI assay ( Nicoletti et al, 1991 ) The percentage of drug-induced cell death was determined as [total percentage of apoptotic cells − percentage of apoptotic cells in untreated controls]. Mean values of three experiments are given. The human T-acute lymphatic leukaemia cell line CEM-C7H2 served as a drug-sensitive control cell line ( Villunger et al, 1997a ). The percentage of apoptosis was assessed after 48 h of drug treatment. Mean values of three experiments are given. Thumbnail image of

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. References

As yet very little information on the function of GM-CSF in the biology of late-stage B cells is available. Although the signal-transducing β-subunit of the receptor is ubiquitously expressed within the haemopoietic system, the α-subunit was, until recently, found exclusively on cells derived from the myeloid lineage ( Miyajima et al, 1992 ). In vitro data suggested a supportive function of GM-CSF on myeloma cell growth, but these effects were discussed as being indirect and mediated via IL-6-dependent signalling in malignant plasma cells ( Zhang et al, 1990 ). Recently, expression of the GM-CSFR constituents was described as occurring during terminal neoplastic B-cell differentiation ( Till et al, 1996 ). Therefore, the possibility of GM-CSF having a direct effect on the expansion of malignant plasma cells had to be considered. This question is of great interest because, in the treatment of multiple myeloma, GM-CSF has been used not only for stem cell mobilization, in order to support recovery after high-dose chemotherapy ( Barlogie et al, 1990 ; Dimopoulos et al, 1993 ; Tarella et al, 1993 ), but also to accelerate cell-cycle transit of tumour cells to sensitize them to therapeutic drugs ( Smith et al, 1995 ).

Analysis of three malignant plasma cell lines and native malignant myeloma cell samples confirmed the constitutive expression of the GM-CSFRα subunit in all cell lines (3/3) and in all native malignant plasma cell samples analysed (6/6). Furthermore, GM-CSFR expression was observed not only on the CD38+ plasma cell fraction of bone marrow but also on CD19+ B cells in the peripheral blood of 5/8 patients with multiple myeloma or plasmacytoma, the majority of whom later develop myeloma. This GM-CSFRα-positive B-cell fraction may contain malignant precursors ( Bergsagel et al, 1995 ). This finding is underlined by the fact that the CD19+ fraction of MGUS patients displayed considerably lower levels GM-CSFR expression whereas the CD19+ fraction of healthy donors were consistently negative ( Table IIb). These findings not only confirm the specificity of data obtained in myeloma cell samples in association with the neoplastic plasma cell fraction but also might indicate that acquisition of GM-CSFR expression may be part of myeloma pathogenesis. However, analysis of a larger patient population is necessary to further support this hypothesis.

In order to determine whether the detected GM-CSFRα subunit was able to directly mediate signals involved in the regulation of cell growth and survival, we tested whether the application of GM-CSF was able to induce the ras/MAPK signalling cascade in malignant plasma cell lines (Fig 2). Activation of this signalling cascade was demonstrated to occur in response to GM-CSF, IL-3 and IL-5, all of which utilize the same β-subunit for signal transduction ( Miyajima et al, 1992 ). In addition to the mitogenic signals mediated along this pathway, GM-CSFR-mediated signalling was demonstrated to be an essential event in the regulation of apoptosis in cells of the myeloid lineage ( Iversen et al, 1996 ; Hara & Miyajima, 1996). Therefore we tested the impact of GM-CSF on DNA synthesis and spontaneous as well as drug-induced apoptosis in these cells. Cultivation of neoplastic plasma cell lines in the presence of GM-CSF led to enhanced DNA synthesis in MC/CAR, IM-9 and RPMI-8226 cells (Fig 3a). This observation confirmed prior reports about a stimulatory effect of GM-CSF on proliferation of native malignant plasma cells, cultured in their native microenvironment, as well as of the IL-6-dependent XG-1 myeloma cell line ( Zhang et al, 1990 ). The effect of GM-CSF on XG-1 myeloma cell proliferation was thought to be mediated by IL-6, since preincubation of the cells with an anti-IL-6 mAb abrogated the GM-CSF-induced [3H]thymidine incorporation in this cell line. In addition, the proliferative effect of GM-CSF on native malignant plasma cells was partially inhibited by application of anti-IL-6 mAb to in vitro primary cultures ( Zhang et al, 1990 ). Enhanced IL-6-responsiveness was postulated to account for the GM-CSF-induced increase in DNA-synthesis, but the nature of this effect remained obscure ( Zhang et al, 1990 ). Our observation, however, is in clear contrast to the postulated IL-6-dependence of GM-CSF activity in myeloma cells, since (i) the cell lines investigated here did not require IL-6 for propagation in vitro ( Villunger et al, 1996 ), (ii) preincubation of RPMI-8226 or IM-9 cells with anti-IL-6R Abs did not lead to the abrogation of GM-CSF-induced DNA synthesis (Fig 3c), while specifically blocking the effects of IL-6 on [3H]thymidine incorporation (Fig 3b), and (iii) expression levels of the α- and β-subunits of the IL-6R remained unchanged in IM-9, MC/CAR as well as RPMI-8226 cells after application of GM-CSF as monitored by RT-PCR (data not shown), thus ruling out enhanced sensitivity to IL-6, due to up-regulation of IL-6R components. It was therefore demonstrated that GM-CSF was able to act on neoplastic plasma cells directly, without involvement of IL-6 signalling. After all, the failure of GM-CSF to induce DNA synthesis in the presence of IL-6 Abs in the XG-1 cell line or freshly explanted native myeloma cells might only reflect the fact that IL-6-dependent myeloma cells, deprived of the essential growth factor, become primed for apoptosis ( Billadeau et al, 1995 ) and, therefore, may no longer be able to respond to other cytokines. However, induction of apoptosis in IL-6-deprived myeloma cells could be prevented by overexpression of constitutively activated ras proteins ( Billadeau et al, 1995 ). As we demonstrated here, GM-CSF was able to activate ras proteins in neoplastic plasma cells and to significantly reduce spontaneous apoptosis in RPMI-8226, MC/CAR and IM-9 cells (Fig 4). The inhibitory effect of GM-CSF on programmed cell death in neoplastic plasma cells is in line with a report about anti-apoptotic signalling along the β-subunit of the IL-3/GM-CSFR in the IL-3-dependent murine pro-B-cell line Ba/F3, where the inhibition of apoptosis mediated by the IL-3/GM-CSFR was shown to be transmitted by the activation of ras molecules ( Kinoshita et al, 1995 ). IL-3-induced inhibition of apoptosis was recently demonstrated to depend on the ability of ras to activate phosphoinositide-3 kinase (PI-3 kinase), leading to the activation of PKB/Akt and subsequently to the inactivation of the pro-apoptotic factor Bad by serine-phosphorylation ( del Peso et al, 1997 ).

Apparently, GM-CSF supports cell survival not only in the myeloid lineage of haemopoiesis ( Iversen et al, 1996 ; Hara & Miyajima, 1996; Kinoshita et al, 1995 ) but also, as we have demonstrated in the present study, in mature neoplastic plasma cells (Fig 4). Our observations are also in line with a report that GM-CSF, like IL-6, is indispensable for the establishment of primary myeloma cell lines from native malignant plasma cells ( Zhang et al, 1994 ). Paul & Baumann (1990) described a similar potential of GM-CSF regarding the establishment of EBV+ B-cell clones. Although two of the cell lines used in the present study were EBV+, the EBV RPMI-8226 cell line also expressed functional GM-CSFR. So, irrespective of EBV status, GM-CSF action may be a signal which controls the longevity of B cells, thereby contributing to the establishment of a malignant clone in multiple myeloma. In contrast to IL-6 ( Hardin et al, 1994 ), GM-CSF was not able to inhibit apoptosis induced by the anti-tumour drug doxorubicin ( Table III). Therefore our data argue against the possibility of negative interference of GM-CSF with tumour cell kill in vivo during chemotherapy-induced aplasia, but it cannot be ruled out that the cytokine improves survival of malignant plasma cells and thereby enhances contamination of CD34+ haemopoietic progenitor cells mobilized with GM-CSF.

The present study demonstrated that the GM-CSFR detected on malignant late-stage B cells is functionally active and mediated a positive effect on the growth and survival of neoplastic plasma cells. Whether GM-CSFR expression is acquired during cancerogenesis or is a constant feature of plasma cells contributing to the pathogenesis of multiple myeloma, and what the mechanism of GM-CSFR regulation is, remain areas for future investigations.

Footnotes
  1.  Present address: Department of Medical Chemistry and Biochemistry, University of Innsbruck Medical School, Fritz-Pregl-Strasse 3, A-6020 Innsbruck, Austria.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. References

Supported by the Austrian Research Countil Project FWF P 8947 med and the Province of Tyrol.

References

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. References
  • 1
    Barlogie, B., Jagannath, S., Dixon, D.O., Cheson, B., Smallwood, L., Hendrickson, A., Purvis, J.D., Bonnem, E., Alexanian, R. (1990) High-dose melphalan and granulocyte-macrophage colony-stimulating factor for refractory multiple myeloma. Blood, 76, 677 680.
  • 2
    Bergsagel, P., Smith, P., Szczepek, A., Mant, M., Belch, A., Pilarski, L. (1995) In multiple myeloma, clonotypic B lymphocytes are detectable among CD19+ peripheral blood cells expressing CD38, CD56, and monotypic Ig light chain. Blood, 85, 436 447.
  • 3
    Bergui, L., Schena, M., Gaidino, G., Riva, M., Caligaris-Cappio, F. (1989) Interleukin-3 and interleukin-6 synergistically promote the proliferation and differentiation of malignant plasma cell precursors in multiple myeloma. Journal of Experimental Medicine, 1710, 613 620.
  • 4
    Billadeau, D., Jelinek, D.F., Shah, N., LeBien, T.W., Van Ness, B. (1995) Introduction of an activated N-ras oncogene alters the growth characteristics of the interleukin 6-dependent myeloma cell line ANBL6. Cancer Research, 55, 3640 3646.
  • 5
    Del Peso, L., Gonzalez-Garcia, M., Page, C., Herrera, R., Nunez, G. (1997) Interleukin-3-induced phosphorylation of BAD through the protein kinase Akt. Science, 278, 687 689.
  • 6
    Dimopoulos, M.A., Delasalle, K.B., Champlin, R., Alexanian, R. (1993) Cyclophosphamide and etoposide therapy with GM-CSF for VAD-resistant multiple myeloma. British Journal of Haematology, 83, 240 244.
  • 7
    Egle, A., Villunger, A., Marion, K., Böck, G., Gruber, J., Auer, B., Greil, R. (1996) Modulation of Apo-1-induced programmed cell death in myeloma cells by interferon alpha2. European Journal of Immunology, 26, 3119 3126.
  • 8
    Hara, T. & Miyajima, A. (1996) Function and signal transduction mediated by the interleukin 3 receptor system in hematopoiesis. Stem Cells, 14, 605 618.
  • 9
    Hardin, J., MacLeod, S., Grigorieva, I., Chang, R., Barlogie, B., Xiao, H., Epstein, J. (1994) Interleukin-6 prevents dexamethasone-induced myeloma cell death. Blood, 84, 3063 3070.
  • 10
    Iversen, P.O., To, L.B., Lopez, A.F. (1996) Apoptosis of hemopoetic cells by the granulocyte-macrophage colony-stimulating factor mutant E21R. Proceedings of the National Academy of Sciences of the United States of America, 93, 2785 2789.
  • 11
    Kawano, M., Tanaka, H., Ishikawa, H., Nobuyoshi, M., Iwato, K., Asaoku, H., Tanabe, O., Kuramoto, A. (1989) Interleukin-1 accelerates autocrine growth of myeloma cells through interleukin-6 in human myeloma. Blood, 73, 2145 2148.
  • 12
    Kinoshita, T., Yokota, T., Arai, K., Miyajima, A. (1995) Suppression of apoptotic death in hematopoietic cells by signalling through the IL-3/GM-CSF receptors. EMBO Journal, 14, 266 275.
  • 13
    Klein, B., Zhang, X.G., Jourdan, M., Content, J., Houssiau, F., Aarden, L., Piechaczyk, M., Bataille, R. (1989) Paracrine rather than autocrine regulation of myeloma-cell growth and differentiation by interleukin-6. Blood, 73, 517 526.
  • 14
    Miyajima, A., Kitamura, T., Harada, N., Yokota, T., Arai, K. (1992) Cytokine receptors and signal transduction. Annual Review of Immunology, 10, 295 331.
  • 15
    Nicoletti, I., Migliorati, G., Pagliacci, M.C., Grignani, F., Riccardi, C. (1991) A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry. Journal of Immunological Methods, 139, 271 279.
  • 16
    Paul, C. & Baumann, M. (1990) Modulation of spontaneous outgrowth of Epstein-Barr virus immortalized B-cell clones by granulocyte-macrophage colony stimulating factor and interleukin-3. Blood, 75, 54 58.
  • 17
    Smith, H., Hussein, M.A., Sandstrom, K., Hays, A. (1995) Diverse biologic activity of GM-CSF in multiple myeloma. Blood, (Abstract). 86, 3359.
  • 18
    Su, B. & Karin, M. (1996) Mitogen-activated protein kinase cascades and regulation of gene expression. Current Opinion in Immunology, 8, 402 411.
  • 19
    Tarella, C., Boccadoro, M., Omede, P., Bondesan, P., Caracciolo, D., Frieri, R., Bregni, M., Siena, S., Gianni, A.M., Pileri, A. (1993) Role of chemotherapy and GM-CSF on hemopoietic progenitor cell mobilization in multiple myeloma. Bone Marrow Transplantation, 11, 271 277.
  • 20
    Terstappen, L., Johnson, S., Segers-Nolten, I., Loken, M.R. (1990) Identification and characterization of plasma cells in normal human bone marrow by high resolution flow cytometry. Blood, 76, 1739 1745.
  • 21
    Till, K.J., Burthem, J., Lopez, A., Cawley, J.C. (1996) Granulocyte-macrophage colony-stimulating factor receptor: stage-specific expression and function on late stage B-cells. Blood, 88, 479 486.
  • 22
    Villunger, A., Egle, A., Kos, M., Hartmann, B.L., Geley, S., Kofler, R., Greil, R. (1997a) Drug-induced apoptosis is associated with enhanced Fas (CD95/Apo-1) ligand expression but occurs independently of Fas (Apo-1/CD95) signaling in human T-acute lymphatic leukemia cells. Cancer Research, 57, 3331 3335.
  • 23
    Villunger, A., Egle, A., Kos, M., Hittmair, A., Maly, K., Greil, R. (1996) Constituents of autocrine IL-6 loops in myeloma cell lines and their targeting for suppression of neoplastic growth by antibody strategies. International Journal of Cancer, 65, 498 505.
  • 24
    Villunger, A., Egle, A., Marschitz, I., Kos, M., Boeck, G., Ludwig, H., Geley, S., Kofler, R., Greil, R. (1997b) Constitutive expression of Fas (Apo-1/CD95) ligand on multiple myeloma cells: a potential mechanism of tumor-induced suppression of immune surveilllance. Blood, 90, 12 20.
  • 25
    Zhang, X.G., Bataille, R., Jourdan, M., Saeland, S., Banchereau, J., Mannoni, P., Klein, B. (1990) Granulocyte-macrophage colony-stimulating factor synergizes with interleukin-6 in supporting the proliferation of human myeloma cells. Blood, 76, 2599 2605.
  • 26
    Zhang, X.G., Gaillard, J.P., Robillard, N., Lu, Z.Y., Gu, Z.J., Jourdan, M., Boiron, J.M., Bataille, R., Klein, B. (1994) Reproducible obtaining of human myeloma cell lines as a model for tumor stem cell study in human multiple myeloma. Blood, 83, 3654 3663.