Deregulated cytokine network and defective Th1 immune response in multiple myeloma

Authors

  • M. A. Frassanito,

    1. Department of Biomedical Sciences and Human Oncology, Section of Internal Medicine and Clinical Oncology, University of Bari Medical School, Bari, Italy
    Search for more papers by this author
  • A. Cusmai,

    1. Department of Biomedical Sciences and Human Oncology, Section of Internal Medicine and Clinical Oncology, University of Bari Medical School, Bari, Italy
    Search for more papers by this author
  • F. Dammacco

    1. Department of Biomedical Sciences and Human Oncology, Section of Internal Medicine and Clinical Oncology, University of Bari Medical School, Bari, Italy
    Search for more papers by this author

Dr Franco Dammacco, DIMO-Section of Internal Medicine and Clinical Oncology, P.za G. Cesare, 11, 70124-Bari-Italy.  E-mail: dimoclin@cimedoc.uniba.it

Abstract

Intracellular cytokine production by peripheral blood mononuclear cells (PBMC) was analysed in 51 patients with multiple myeloma (MM), 22 with monoclonal gammopathy of undetermined significance (MGUS) and 20 healthy subjects, as a parameter of immunological dysfunction in MM. An increased proportion of T cells and HLA-DR+ cells producing IL-6 was observed in MM patients with active disease (at diagnosis and relapsing) compared with patients in remission and with MGUS, whereas no difference of IFN-γ+, IL-2+ PBMC between patients and controls was evident. Determination of serum cytokine levels demonstrated that the imbalanced IL-6 production by T cells and the defective anti-tumour Th1 cell activity were related to elevated levels of IL-6 and IL-12. In vitro studies of PHA- and anti-CD3/anti-CD28 MoAbs stimulation of PBMC demonstrated the ability of lymphocytes from MM patients to differentiate towards the Th1 subset in the presence of rIL-12. By contrast, addition of exogenous rIL-6 impaired IFN-γ production by rIL-12-prompted T cells. Inhibition of Th1 polarization of the immune response by IL-6 was direct on T cells and not mediated by dendritic cells (DC). Evaluation of the ability of MM-derived DC to stimulate cell proliferation of allogenic T lymphocytes and produce IL-12 in vitro, in fact, suggested that MM-derived DC were functionally active. Taken as a whole, these results indicate that a deregulated cytokine network occurs in active MM. They also suggest that increased IL-6 production by peripheral T lymphocytes contributes to the immune dysfunction observed in MM, and enables tumour cells to escape immune surveillance by preventing the anti-tumour Th1 immune response.

Introduction

Immunogenic antigens expressed on the surface of murine and human tumours provide a target for immune-mediated tumour defense mechanisms [1,2] whose effects range from an actual anti-tumour response to tolerance induction. The recent isolation from tumour patients of T lymphocytes reactive with tumour-associated antigens [3–5] underscores the notion that tumours are immunogenic and thus potential targets for immune destruction. Understanding the mechanisms by which tumour cells escape immune surveillance would provide an attractive goal for the prevention of T-cell immunosuppression in cancer patients, and the development of cancer immunotherapy.

Activated T lymphocytes differentiate into two functional subsets, T helper type 1 (Th1) and T helper type 2 (Th2) cells, with different cytokine secretion patterns and effector functions [6]. Th1 cells produce interferon-γ (IFN-γ), interleukin (IL)-2 and tumour necrosis factor (TNF)-β, and promote cell-mediated immunity, whereas Th2 cells produce IL-4, IL-5, IL-6, IL-9, IL-10 and IL-13, and are involved in antibody-mediated immunity. Th1/Th2 polarization depends on several environmental and genetic factors and particularly, on the local concentration of cytokines, such as IL-12 and IL-4, that induce differentiation of naive T lymphocytes to the Th1 and Th2 phenotype, respectively [6]. The biological significance of this dichotomy in human and animal tumours has recently been described [7–9]. A correlation between the type 1 immune response and anti-tumour activity has been suggested, and several experimental studies have demonstrated an anti-tumour effect of IL-12 using gene-modified tumour [10] or antigen-presenting [11] cells, or by systemic administration of IL-12 [9,12]. The mechanism underlying this effect is less clear. IL-12 probably enhances cellular immune responses by favouring a differentiation of Th1 lymphocytes, up-regulating MHC expression on tumour cells, and/or inhibiting tumour angiogenesis [9,13].

Multiple myeloma (MM) and monoclonal gammopathy of undetermined significance (MGUS) are plasmaproliferative disorders characterized by proliferation of clonal B lymphocytes at various stages of maturation. Myeloma cells produce monoclonal immunoglobulins (Ig) (M component) and express identical Ig gene rearrangements. Several studies have suggested that the idiotypic (id) structure of myeloma Ig is a stable marker that can be regarded as a tumour-specific antigen [14]. MM progresses through a series of stepwise oncogenic events [15]. However, no evidence of intraclonal variations due to mutations in the variable gene sequences has been provided. On the other hand, B cells producing anti-id antibodies, and T cells reactive with the idiotype of the autologous M-component, have been described in murine plasmacytoma and in human MM and MGUS [5,16–18].

MM progression is associated with secondary immnodeficiency and an increased risk of infections, which may be related to alterations in T-cell subsets and cytokine production. An expansion of peripheral CD3+ lymphocytes showing the phenotype of activated memory T cells (HLA-DR+ cells) described in MM patients [19] is associated with an impairment of cell-mediated immune response [20,21]. Several mechanisms have been proposed to explain this immunodeficiency, namely enhanced susceptibility of peripheral T lymphocytes to apoptosis [22,23] and an immunosuppressive effect of CD8+/CD57+ cells [22].

In this study, intracellular cytokine production by peripheral blood mononuclear cells (PBMC) in MM and MGUS patients was analysed as a parameter of immunological dysfunction. In particular, we observed that increased percentage values of IL-6+/CD3+ lymphocytes and a defective Th1 polarization of immune response T1 parallel MM clinical stage. Furthermore, the mechanism by which activated T lymphocytes do not differentiate in vivo to Th1 phenotype was investigated in vitro by evaluating the ability of dendritic cells (DC) from MM patients to induce an effective immune response, and that of activated CD3+ cells to produce type 1 cytokines. Our data indicate that a deregulated cytokine network is involved in MM. Elevated IL-6 production by activated CD3+ lymphocytes may thus be supposed to modulate Th1 polarization of the immune response negatively.

Materials and methods

Patients

Seventy-three patients who fulfilled the South-West Oncology Group (SWOG) diagnostic criteria [24] for MM (51 subjects) and MGUS (22 subjects) were enrolled. According to the Durie and Salmon staging system [25], 13 MM patients were studied at diagnosis, 21 at relapse and 17 in complete/partial remission. Patients at diagnosis and at relapse were classed as having active disease. Twenty healthy subjects were included as controls. Both patients and controls gave their informed consent and the study was approved by the Ethics Committee of the University of Bari.

Lymphocyte preparation and cell culture

PBMC were isolated from heparinized blood by Ficoll-Hypaque gradient centrifugation and resuspended at 1 × 106 cells/ml in RPMI 1640 (Biochrom, Berlin, Germany) containing 10% fetal calf serum (Biochrom), 2 mm glutamine (Biochrom), 100 U/ml penicillin (Biochrom) and 100 µg/ml streptomycin (Biochrom) (culture medium). Cell activation was carried out in 5 day stimulation cultures with 5 µg/ml phytohemagglutinin (PHA; Sigma, St. Louis, MO), or with 1 µg/ml plate-bound anti-CD3 monoclonal antibody (MoAb; Immunotech, Marseille, France) and 1 µg/ml soluble anti-CD28 MoAb (Immunotech). Cultures were additionally supplemented with recombinant human IL-6 (rIL-6, cat. no. 200–06; Peprotech, London, UK) and rIL-12 (Peprotech), either alone or in combination.

Flow cytometry determination of intracellular cytokine synthesis

Intracellular immunofluorescence staining of IL-2, IL-4, IL-6 and IFN-γ was performed as described [26]. Briefly, fresh PBMC and cultured lymphocytes were stimulated with 10 ng/ml phorbol myristate acetate (PMA; Sigma) and 1 µm ionomycin (Sigma), or with 1 µg/ml lipopolysaccharide (LPS; Sigma), in round-bottomed culture tubes. In all cases, 3 µm monensin (Sigma) was added to prevent cytokine release. After 5 h incubation, stimulated cells were fixed with 4% paraformaldehyde for 15 min and permeabilized with phosphate-buffered saline (PBS) containing 0·5% bovine serum albumin (Sigma) and 0·5% saponin (Sigma). They were then incubated with phycoerythrin (PE)-conjugated anti-IL-2, anti-IL-4, anti-IL-6 or anti-IFNγ MoAbs (Biosource, Fleurus, Belgium). The surface phenotype of cytokine-producing cells was identified by incubating cell samples with fluorescein isothiocyanate (FITC)-conjugated MoAb (Sigma) to CD3 and HLA-DR antigens. Negative controls were isotype-matched irrelevant antibodies. Lastly, samples were analysed by a FACScan flow cytometer (Becton Dickinson, Mountain View, CA).

IL-6, IL-10 and IL-12 detection

IL-6, IL-10 and IL-12 levels were measured in serum and cell culture supernatant fluids by ELISA. Samples were added to appropriate wells of a microtitre plate coated with a blend of MoAb directed against distinct epitopes of IL-6, IL-10 or IL-12 (Biosource) to ensure a highly sensitive assay. After incubation, appropriate horseradish peroxidase (HRP)-labelled MoAbs, namely anti-IL-6, anti-IL-10 or anti-IL-12, were added. Bound enzyme-labelled antibodies were detected through a chromogenic reaction by the addition of tetramethylbenzidine and H2O2 in acetate citrate buffer. Plates were read at 450 nm in a microenzyme-linked immunosorbent assay reader (Titertek, Flow Irvine, UK). A standard curve was used to quantify cytokine levels.

Generation of DC

PBMC were cultured to generate monocyte-derived DC as described [27]. Briefly, 5 × 106 cells were incubated in culture medium for 2 h at 37°C. Non-adherent cells were then removed and the adherent fraction was cultured in fresh culture medium containing 400 U/ml IL-4 (Peprotech) and 800 U/ml GM-CSF (Peprotech), for 14 days. LPS (1 ng/ml) was added for the final 24 h to stimulate the cells. Harvested DC were analysed for the expression of specific antigens by single- or double-fluorescence staining with a FACScan flow cytometer. FITC- or PE-conjugated MoAb to CD1a (Immunotech), CD14 (Sigma), CD80 (Immunotech), CD86 (Immunotech) and HLA-DR (Sigma) were used. The negative controls were isotype-matched irrelevant antibodies. Cell populations were gated by forward scatter (FSC) versus side scatter (SSC). DC culture supernatant fluids were also collected and tested for cytokine release by ELISA.

Mixed lymphocyte reaction (MLR)

Harvested DC were washed and resuspended at 5 × 106 cells/ml in culture medium containing 0·08 mg/ml mitomycin C (Sigma) to inhibit proliferation. Various concentrations (102 to 104 cells/well) were seeded in triplicate, flat-bottomed, 96-well culture plates. As responder cells, 105 allogeneic peripheral lymphocytes from a healthy donor were added to each well. Plates were incubated for 5 days at 37°C and pulsed with 3H-thymidine (1 µCu/well, Du Pont NEN, Bad Homburg, Germany) for 18 h before harvesting. 3H-thymidine uptake was evaluated in a β-counter (Beckman Instruments, Palo Alto, CA). The ability of DC to stimulate proliferation of responder cells was evaluated from 3H-thymidine uptake, and expressed as stimulation index: cpm responder cells + stimulator cells/cpm responder cells.

Statistical analysis

Results are expressed as means ± s.d. Student's t-test and, in several instances, the Wilcoxon non-parametric method were used to compare mean values of specific phenotype expression and in vitro parameters.

Results

Intracellular cytokine production in PBMC of patients and healthy controls

Cytokine-producing precursor cells in the peripheral blood of patients and healthy controls were detected by flow cytometry. Figure 1 shows the percentage values of PBMC producing type 1 (IL-2, IFN-γ) or type 2 (IL-4, IL-6) cytokines. The similar IL-2+, IFN-γ+ PBMC percentages in patients and controls suggests that MM does not involve Th1 polarization of the immune response.

Figure 1.

 Intracellular cytokine production by peripheral blood mononuclear cells (PBMC) in MM and MGUS patients and healthy subjects. Purified PBMC were stimulated with phorbol myristate acetate/ionomycin or with LPS for 5 h, then analysed for intracellular cytokine synthesis. Patients with active MM (at diagnosis and in relapsing disease) showed significantly higher percentage values of IL-6-producing PBMC compared with patients in remission, patients with MGUS and healthy subjects. No difference in Th1-type lymphocytes was observed. (bsl00017) IFN-γ+; (□) IL-2+; (▪□) IL-4+; (▪) IL-6+. *P < 0·005.

By contrast, the percentages of IL-6+ PBMC were significantly increased in MGUS and MM patients (35·1 ± 6·5% and 55·0 ± 18·4%, respectively) compared with the controls (16·9 ± 8·2%). Analysis of IL-6 production by CD4+ and/or CD8+ cells revealed that both lymphocyte subsets were involved (data not shown). In addition, intracellular IL-6 production paralleled MM progression (Fig. 1); patients with active MM showed significantly (P < 0·001) higher values (62·5 ± 13·6%) compared with those in remission (37·7 ± 9·8%) and with MGUS (35·1 ± 6·5%). No relationship between IL-6 and IL-4 production was observed.

Intracellular IL-6 production and CD3+ T cell activation

Since expansion of HLA-DR+ T cells has been described in MM [22], we next evaluated the expression of activation antigen (HLA-DR) on IL-6-producing cells by two-colour fluorescence staining. Figure 2 illustrates a representative cytofluorimetric pattern of CD3+/IL-6+(Fig. 2a, c, e, g) and HLA-DR+/IL-6+(Fig. 2b, d, f, h) lymphocytes in two MM patients at different clinical stages, one MGUS patient and a control subject. Patient no. 13 with relapsing MM (Fig. 2a, b) showed an increased percentage of peripheral CD3+/IL-6+ lymphocytes associated with expansion of HLA-DR+/IL-6+ T cells compared with the patient in remission (Fig. 2c, d) and the patient with MGUS (Fig. 2e, f), respectively. Thus, patients with active MM showed significantly higher values of activated lymphocytes primed to produce IL-6. On the other hand, the percentages of cytokine-producing cells among CD3+ and DR+ populations indicated that the proportion of CD3+/IL-6+ and HLA-DR+/IL-6+ lymphocytes is significantly (P < 0·01) increased in patients with active MM compared with those in remission, those with MGUS and healthy subjects (Table 1). Almost the entire HLA-DR+ population was IL-6-positive.

Figure 2.

Figure 2.

 Representative double-fluorescence staining of peripheral CD3+/IL-6+ and HLA-DR+/IL-6+ lymphocytes in two MM patients at different clinical stages, in one MGUS patient and in a healthy subject. An expansion of IL-6-producing HLA-DR+ lymphocytes was observed in the patient with relapsing MM (a, b) compared with the patient in remission MM (c, d), the MGUS patient (e, f) and the control (g, h).

Table 1.   Percentage of IL-6-producing cells among peripheral CD3+ and HLA-DR+ lymphocytes
 IL-6+-producing
 *CD3+ cells*HLA-DR+ cells
  • *

    Values are expressed as mean ± s.d.

Active MM71·2 ± 18·276·0 ± 10·6
(no. 34)
Remission MM32·4 ± 12·128·0 ± 7·4
(no. 17)
MGUS29·2 ± 8·736·3 ± 5·8
(no. 22)
Healthy subjects23·5 ± 3·926·1 ± 4·9
(no. 20)

Serum IL-6, IL-10 and IL-12 levels

To assess the presence of specific cytokines in the microenvironment, the serum levels of IL-6, IL-10 and IL-12 were measured in patients and controls (Table 2). Serum IL-6 paralleled the intracellular IL-6 production by T cells and the clinical stage of MM. Patients with active MM showed significantly (P < 0·01) higher values of serum IL-6 (35·2 ± 17·4 pg/ml) compared with patients in remission (17·8 ± 8·5 pg/ml) and with MGUS (3·0 ± 2·6 pg/ml). Elevated IL-12 levels were also observed in MM patients, but were not related to MM activity. Serum levels of IL-10 in MM and MGUS patients were comparable with those in controls. The absence of a Th1 immune response in MM was not mediated by the suppressive effect of IL-10 [8], but was associated with increased serum levels of both IL-6 and IL-12.

Table 2.   Serum cytokine levels in multiple myeloma and MGUS patients
 *IL-6*IL-10*IL-12
  • *

    Values are expressed as pg/ml.

  • P < 0·005.

Active MM35·2 ± 17·42·4 ± 0·985·4 ± 38
(no. 34)
Remission MM17·8 ± 8·51·8 ± 0·594·7 ± 46·6
(no. 17)
MGUS3·0 ± 2·63·2 ± 2·448·7 ± 28·7
(no. 22)
Healthy subjects3·5 ± 2·52·1 ± 0·835·3 ± 12·4
(no. 20)

Intracellular IFN-γ production by in vitro stimulated CD3+ T cells

In the second series of experiments, the ability of CD3+ T lymphocytes from MM patients to produce type 1 cytokines was evaluated. Patients' PBMC were stimulated in vitro with PHA or anti-CD3/anti-CD28 MoAbs for 5 days in the presence or absence of rIL-12 and/or rIL-6, and their IFN-γ production was analysed. As shown in Fig. 3, stimulation of PBMC in the absence of cytokines induced intracellular IFN-γ production in only 20% of CD3+ cells. Addition of rIL-12 significantly improved the percentage values of Th1 lymphocytes in stimulated cell cultures, indicating the in vitro ability of CD3+ lymphocytes to differentiate into a Th1 subset. By contrast, addition of rIL-6 impaired IFN-γ production by rIL-12-prompted T cells. Thus, PHA- or anti-CD3/anti-CD28 MoAb stimulation of patients' PBMC in the presence of both rIL-6 and rIL-12 did not induce a strong Th1 polarization. These results indicate the ability of patients' PBMC to produce IFN-γ after stimulation in the presence of rIL-12, and suggested that Th1 polarization of the immune response is negatively modulated by rIL-6.

Figure 3.

Figure 3.

 Effect of rIL-6 and rIL-12 on IFN-γ production by stimulated patients' lymphocytes. PBMC (1 × 106) purified from 32 MM patients at different clinical stages were stimulated in vitro with PHA (5 µg/ml) or with immobilized anti-CD3 MoAb (1 µg/ml) and soluble anti-CD28 MoAb (1 µg/ml) in the presence of culture medium (–), rIL-12 (400 ng/ml) or rIL-6 (100 U/ml). After 5 days, IFN-γ production was evaluated by flow cytometry. Addition of rIL-12 induced a Th1 polarization of cultured PBMC that was significantly inhibited by the presence of rIL-6.

Patient-derived DC are functionally active

IL-12, the critical cytokine in induction of a Th1 immune response, is produced upon activation of DC. In the next experiments, the ability of patient-derived DC to activate T lymphocytes and produce IL-12 was investigated. DC were generated in vitro from the adherent fraction of PBMC of 15 MM patients, 11 MGUS patients and seven healthy subjects using GM-CSF and rIL-4. After 14 days, LPS-stimulated cells showed morphological features of monocyte-generated DC, and their phenotypic analysis demonstrated the expression of CD1a, HLA-DR, CD80 and CD86 antigens on 75% of cells (data not shown).

The immunostimulatory activity of patient-derived DC was evaluated by MLR assay. Proliferation of allogeneic responder cells was induced using different numbers of LPS-stimulated DC (Fig. 4a). Patient-derived DC induced cell proliferation of alloreactive T lymphocytes, and the stimulation index was dependent on the number of DC used as stimulator cells. The responses induced by DC from patients and controls were similar, even when few stimulator cells were used. Lastly, determination of IL-12 levels in DC culture supernatant fluids demonstrated that LPS-stimulated, patient-derived DC produced large amounts of this cytokine (Fig. 4b). These data clearly demonstrate that patient-derived DC are functionally active.

Figure 4.

Figure 4.

 Functional studies of monocytes-derived dendritic cells (DC). DC were generated from adherent PBMC cultured with GM-CSF (800 U/ml) and rIL-4 (400 U/ml) and stimulated with LPS (1 ng/ml) for the final 24 h of culture. (a) T cell stimulatory function (MLR assay) of patient- (●) and healthy subject-derived (▴) DC. LPS-stimulated DC from 15 MM patients, 11 MGUS patients and seven healthy subjects were treated with mitomycin C and used at different concentrations. Responder cells (1 × 105 per well) were obtained from an allogeneic donor. No difference in allostimulatory activity between patient- and control-derived DC was evident. (b) Supernatant fluids of cultured DC generated from patients and healthy subjects were collected and IL-12 content was evaluated by ELISA. Patient-derived DC released large amounts of this cytokine, comparable with control-derived DC.

Discussion

The present study focused on defective Th1 polarization of the immune response in relation to IL-6 production by activated peripheral CD3+ lymphocytes in MM patients. Flow cytometry of PBMC demonstrated expansion of activated T cells producing IL-6, but not IFN-γ, IL-2 and IL-4. Chronic activation of the immune system, which has been clearly shown in MM [19], is associated with several T cell dysfunctions [20–22]. We related the HLA-DR+ T cell expansion to imbalanced IL-6 production by CD3+ cells and defective anti-tumour Th1 cell activity.

Th1/Th2 cytokine balance plays an important regulatory role in the immune system, and its de-regulated ratio is responsible for several immunopathological conditions [28]. Th1/Th2 polarization depends on the presence in the microenvironment of specific cytokines, such as IL-12 and IL-10 [28,29]. Specifically, IL-12 production by activated DC drives differentiation and proliferation of Th1 lymphocytes, and induces its own positive, as well as negative regulation by IFN-γ and IL-10 cytokine release, respectively [30]. IL-12 is the most potent anti-tumour cytokine [9–12,31]. Several studies have demonstrated that it induces an effective Th1-type cell-mediated immune response against established tumours [9]. A striking result of our study was the absence of a Th1 immune response in MM patients that occurred in the presence of high serum levels of IL-6 [32] and IL-12 cytokines, and normal values of IL-10 [33]. Serum IL-12 failed to drive a Th1 polarization of MM activated T lymphocytes in vivo, and this was not due to inhibition by IL-10 [8,29]. These data suggest that a deregulated cytokine network involving a defective, anti-tumour, Th1-type immune response occurs in MM patients.

One point of apparent discrepancy is the lack of parallelism between the increased percentage values of IL-6-producing and the normal values of IL-4-producing PBMC. Although further experiments are in progress to clarify this point, it can be envisaged that serum IL-12 and/or an as yet unidentified IL-4 inhibitory factor prevents the expansion of the corresponding population [6].

DC are antigen-presenting cells endowed with the greatest efficiency in induction of a Th1-type immune response [34]. They are extremely few in vivo but can now be generated in vitro. We found that DC derived from MM patients were functionally active [35] in that they produced large amounts of IL-12 and stimulated cell proliferation of allogeneic CD3+ lymphocytes. These results are consistent with several clinical trials in which antigen-pulsed autologous DC were used as a vaccine to elicit id-specific Th1 cell responses in the treatment of minimal residual disease or relapsed MM [36,37]. Encouraging results were obtained in these trials. The T cell response was transient, however, and the clinical benefits of id vaccination remain to be determined. Identification of the mechanism by which active DC and serum IL-12 correlate with an in vivo inability to mount an effective Th1 response would greatly improve immunotherapeutic strategies for MM.

In vitro studies of PHA and anti-CD3/anti-CD28 MoAb-stimulated PBMC demonstrated the ability of MM lymphocytes to differentiate towards the Th1 phenotype in the presence of rIL-12. By contrast, addition of rIL-6 inhibited in vitro Th1 polarization in both patients and healthy subjects. These results prompted us to speculate that elevated serum IL-6 levels may inhibit the Th1 immune response in MM patients. Increased serum levels of IL-6 have been documented in active MM patients [32]. However, the data are discordant due to differences in assay sensitivity [38]. By intracellular cytokine staining, a sensitive assay that directly analyses the phenotype of cytokine-producing cells, we clearly demonstrated that IL-6 production by activated T cells paralleled serum IL-6 levels and the clinical stage of MM.

IL-6 is an immunoregulatory cytokine that plays an important role by modulating cytokine production by several types of lymphoid cells expressing its specific receptor, IL-6R, and may be a key factor in the choice between a Th1 or Th2 immune response [39–41]. The cytokine produced by DC initiates polarization of naive CD3+ T cells to the Th2 phenotype by inducing the production of IL-4 that antagonizes the IL-12-mediated Th1 differentiation [39]. In our study, IL-6 inhibited IFN-γ production by IL-12-induced CD3+ cells, but we did not observe an appreciable IL-4-producing cell population. Furthermore, the inhibitory effect of IL-6 was direct on T cells and not mediated by DC. Addition of exogenous human rIL-6 to anti-CD3/anti-CD28 MoAb-stimulated cell cultures drastically reduced the percentage of IFN-γ-producing cells. On the other hand, in vitro IL-12 production by stimulated DC was not altered by the addition of IL-6 (data not shown), suggesting that IL-6 directly inhibited Th1 polarization via T cells. A recent study on murine CD4+ Th1 cells supports this hypothesis by providing the molecular mechanism by which IL-6 inhibits Th1 differentiation; in particular, IL-6 is able to activate STAT3 and to up-regulate suppressor of cytokine signalling 1 (SOCS1) expression that inhibits STAT1 phosphorylation and IFN-γ gene expression [42]. Increased percentage values of PBMC producing IL-6, but not IL-4 and/or IL-5, were observed in MM patients and were related to elevated serum levels of IL-6 and IL-12. Our overall results suggest that IL-6 production by PBMC is a reliable prognostic marker in MM and may represent a mechanism to induce T cell immunosuppression.

We have recently demonstrated that myeloma plasma cells produce IL-6, and that this production is related to their resistance to drug- and Fas-induced apoptosis [43,44]. Thus, it is speculated that IL-6 plays a central role in the progression of MM in that it favours two mechanisms of tumour immune escape, namely resistance of tumour cells to apoptosis and a defective anti-tumour Th1 immune response.

Acknowledgements

The study was supported by grants from Associazione Italiana della Ricerca sul Cancro (AIRC), Milan, and the Ministry of University and Scientific and Technological Research, Rome, Italy.

Ancillary