Purpose: Dysregulated cytokine/cytokine receptor expression may occur in B-cell lymphoproliferative disorders. Little information is available on interleukin-18 receptor (IL-18R) and IL-18 expression in normal and malignant B cells. Our purpose was to investigate this issue in human naive, germinal center (GC) and memory B cells, and in their neoplastic counterparts.
Experimental Design: We have evaluated IL-18 expression and production in tonsil naive, GC, and memory B cells and in their presumed neoplastic counterparts by reverse transcription-PCR and ELISA. Moreover, IL-18Rα and β expression was investigated in the same cells by reverse transcription-PCR, flow cytometry, and immunohistochemistry.
Results: We found that: (a) IL-18 mRNA was expressed in tonsil naive, GC, and memory B cells. Bioactive IL-18 was secreted by naive and GC, but not by memory B cells; (b) IL-18Rα and β transcripts were expressed in the three B-cell subsets. IL-18Rα was detected on the surface of naive, GC, and memory B lymphocytes, and IL-18Rβ was detected on GC and memory, but not naive, B cells; (c) mantle zone, follicular, marginal zone, Burkitt lymphoma (BL), and B-cell chronic lymphocytic leukemia (B-CLL) cells expressed IL-18 mRNA. B-CLL and BL cells did not produce bioactive IL-18; and (d) lymphoma B cells displayed heterogeneous expression of either or both IL-18R chain mRNA. In contrast, B-CLL cells expressed both IL-18R chains at the mRNA and protein levels.
Conclusions: Dysregulated expression of IL-18 and/or IL-18R in chronic B-cell lymphoproliferative disorders may sometimes contribute to tumor escape from the host immune system.
Interleukin (IL)-18 or IFN-γ inducing factor (IGIF) is a pro-inflammatory cytokine produced by different cell types (as reviewed in Ref. 1 ). IL-18 is synthesized as inactive precursor that is converted into the bioactive molecule after intracellular cleavage by caspase-1 (2 , 3) , or extracellular modification by the proteinase-3 (PR-3) serine esterase (4) .
IL-18 shares a number of functional properties with IL-12, such as amplification of IFN-γ secretion by T, natural killer (NK), and NKT cells (5, 6, 7, 8, 9) , enhancement of granulocyte macrophage colony-stimulating factor production by T lymphocytes (10) , and induction of cytotoxic activity, as well as of Fas ligand expression, in T and NK cells (8 , 11) . IL-18 drives T helper (Th) type 1 (Th1) responses, mostly in association with IL-12 (1 , 5 , 12) . Furthermore, it has been demonstrated that IL-18 can promote Th2 responses in different in vitro and in vivo models (7 , 13) .
Effective antitumor responses are boosted in mice by IL-18 gene transfer into tumor cells or systemic administration of the cytokine, especially in combination with IL-12 or IL-2 (14, 15, 16, 17, 18, 19, 20) . The antitumor activity of IL-18 is primarily mediated by NK- and T-cell activation (16 , 18, 19, 20, 21) .
The IL-18 receptor (IL-18R) is a heterocomplex composed of a constitutive ligand-binding chain, designated as α chain and originally described as IL-1 receptor-related protein (IL-1RrP; Ref. 22 ), and of an inducible accessory chain, the β chain, originally named accessory protein like (AcPL; Ref. 23 ). IL-18Rα binds the cytokine with low affinity, whereas the β chain increases IL-18 binding affinity and is necessary to initiate signal transduction in target cells (23) . IL-18Rα is expressed on the surface of NK cells, CD8+ T lymphocytes, activated CD4+ T cells, and CD19+ peripheral blood B lymphocytes (24) , but information on IL-18Rβ expression is lacking.
Scanty data are thus far available on expression of IL-18 and its receptor in human tumor cells. Thus, IL-18 mRNA has been detected in malignant cells from colon and ovarian carcinoma (25 , 26) , mantle cell lymphoma (27) , and skin cancer (28) . IL-18Rα expression has been reported in different hematopoietic cell lines, part of which were derived from patients with lymphoproliferative or myeloproliferative disorders, but not in primary tumor cells (29) . EBV infection has been shown to up-regulate IL-18 expression in secondary lymphoid tissues from patients with infectious mononucleosis and, at a lesser extent, posttransplant lymphoproliferative disorder (22) .
Previous studies have shown that abnormal expression of cytokine and/or their receptors in neoplastic B cells may contribute to some aspects of the pathophysiology of B-cell lymphoproliferative disorders (as reviewed in Ref. 30 ).
Here we have investigated expression of IL-18, IL-18Rα, and IL-18Rβ in the major B-cell subsets present in secondary lymphoid tissues, i.e., naive, germinal center (GC) and memory B cells, and in their postulated malignant counterparts, based on the hypothesis that dysregulated expression of IL-18, IL-18Rα, and IL-18Rβ may occur in the latter cells in relation to malignant transformation. Furthermore, although IL-18 production by monocytes-macrophages and dendritic cells has been well characterized (1) , little is known on the synthesis of this cytokine in the third type of professional antigen-presenting cells, i.e., B lymphocytes. IL-18 expression in B cells may influence their ability to drive Th-cell differentiation after antigen presentation.
MATERIALS AND METHODS
This investigation was performed after approval by a local Institutional Review Board. Surgically removed tonsils from 10 patients with localized inflammatory disorders were obtained after informed consent.
Lymph node biopsies from six patients with follicular lymphoma (FL), three patients with mantle cell lymphoma (MCL), two patients with marginal zone lymphoma (MZL), and four patients with B-cell chronic lymphocytic leukemia (B-CLL), were obtained from the pathologists after completion of all of the diagnostic procedures. Peripheral blood samples from 12 additional B-CLL patients were obtained after informed consent. Diagnosis was established according to the criteria of the WHO classification (31) . Clonal excess was assessed by the κ/λ or λ/κ immunoglobulin light chain ratio, that ranged from a minimum of 8:1 to a maximum of 20:1 in the different cases. All of the patients (15 males, 12 females; age range, 40–75 years) were untreated at the time of study.
According to the Revised European-American classification of Lymphoid neoplasms (REAL; Ref. 32 ). and the subsequent WHO classifications (31) , FL originates from GC B cells, the majority of MCL cases derive from a CD5+, CD19+ naive B cell homing in the follicular mantle of secondary lymphoid follicles, and MZL represents the abnormal expansion of memory B cells. On the ground of recent studies, B-CLL can be dissected into different subsets according to the mutational status of immunoglobulin variable (IgV) region genes or to the expression of the CD38 or ZAP-70 markers (33, 34, 35) . Although the precise normal counterparts of B-CLL cells have not yet been defined, cases with unmutated IgV genes may originate from a pre-GC B cell, either naive or activated, whereas those carrying IgV gene mutations are of post-GC/memory B-cell derivation (36 , 37) . Because of such heterogeneity, six peripheral blood samples from unmutated and 6 from mutated B-CLL patients were tested in this study. Analysis of IgV gene mutations was carried out by Dr. Franco Fais, Department of Experimental Medicine, University of Genova, as published previously (33) .
Cell Separation and Culture.
Mononuclear cells were isolated from tonsils on Ficoll-Hypaque density gradients and were depleted of T cells by rosetting with neuraminidase-treated sheep erythrocytes. Non-T cells were deprived of macrophages and NK cells by incubation, first with CD68 and CD56 monoclonal antibodies (mAbs), and subsequently with immunomagnetic beads coated with goat antimouse immunoglobulin (Immunotech, Marseille, France; Ref. 38 ). Neoplastic B lymphocytes, either from lymph nodes or peripheral blood, were purified by depletion of T and NK cells and of macrophages as above, and of residual normal B cells bearing the immunoglobulin light chain not expressed by the malignant clone using immunomagnetic beads (38) .
The resulting cell fractions from tonsil contained on average 99% B cells, as assessed by staining with CD19 mAb (see below). The purity of B lymphoma cells was higher than 99%, as assessed by expression of CD19 and of monotypic immunoglobulin light chains. CD3+ T cells, CD56+ NK cells, and CD68+ macrophages were virtually absent (<1%) from these cell suspensions, as assessed by flow cytometry (see below).
Naive B lymphocytes were isolated as IgD+ cells from tonsil B lymphocyte suspensions by immunomagnetic bead manipulation. The IgD− B-cell fractions were further separated into CD38+ (GC) cells and CD38− (memory) cells using the same technique (39 , 40) . All of the above separation procedures were performed at 4°C to prevent spontaneous apoptosis of GC B cells. In some experiments, tonsil B cells and B-CLL cells isolated from peripheral blood were cultured for 48 h (see below).
A panel of Burkitt lymphoma (BL) cell lines, either EBV+ (RAJI, DAUDI, and LAMD2.1) or EBV− (RAMOS, HBL2, and BJAB), were maintained in culture in RPMI 1640 (Seromed) supplemented with 10% FCS (Seromed)
mAbs for Flow Cytometry and Immunohistochemistry.
The following reagents were used for single, double, or triple staining: phycoerythrin (PE)-conjugated anti-IL-18Rα mAb (R&D System, Minneapolis, MN), antihuman IL-18Rβ goat IgG (R&D System, Minneapolis, MN), PE-conjugated swine antigoat IgG (Caltag, Burlingame, CA), horseradish-conjugated antigoat IgG (Santa Cruz Biotechnology, Santa Cruz, CA), PE-conjugated anti-IgD mAb (Dako, Glostrup, Denmark), FITC-conjugated anti-IgD mAb (Caltag), tri-color-conjugated CD38 mAb (Caltag), FITC-conjugate CD3, CD19, CD56, and CD68 mAbs (Becton-Dickinson, San Josè, CA). Controls were fluorochrome-conjugated, isotype-matched mAbs of irrelevant specificity or goat non-immune serum (Caltag). The goat antihuman IL-18Rβ antiserum worked efficiently in immunohistochemistry assays with frozen, but not formalin-fixed and paraffin-embedded, tissue sections, although it did not perform in flow cytometry and Western blot assays. In contrast, the PE-conjugated anti-IL-18Rα mAb worked in flow cytometry experiments, but not in immunohistochemistry assays.
Cells were scored using a FACScan analyzer (Becton-Dickinson, San Jose, CA) and data were processed using CellQuest software (Becton-Dickinson). The threshold line for assessment of percentage of positive cells was based on the maximum staining obtained with irrelevant isotype-matched mAb, used at the same concentration as test mAb. Negative cells were defined such that <1% of cells stained positive with control mAbs. Cells labeled with test antibody that were brighter than those stained with isotypic control antibody were defined as positive. Mean fluorescence intensity (MFI) values of the isotypic control and of test mAbs were used to evaluate whether the differences between the peaks of cells were statistically significant with respect to control. The Kolmogorov-Smirnov test for the analysis of histograms was used, according to the CellQuest software user’s guide
RT-PCR and Sequencing.
RNA was extracted from freshly isolated cells using RNeasy Mini kit from Qiagen (Qiagen GmbH, Hilden, Germany) and was subjected to reverse transcription-PCR (RT-PCR), as reported previously (41) . Primer sequences and profiles of amplification were described in previous studies (41) , with the exception of the following: IL-18 5′-GCTGAAGATGAAAACC and 3′-AGCTAGAAAGTATCCTTC; IL-18Rα 5′-ATT ACC CTT GAC CCT TTG GG and 3′-TCA AAC TCG GCG TTC TTC TT; IL-18Rβ 5′-CCT GCC CTT CAT GGG TAG TA and 3′-ATC CAC TAC GAT TCG GTT GC. Amplification profile was 94°C for 1 min, annealing 55°C (IL-18), 47°C (IL-18Rα), 58°C (IL-18Rβ) for 1 min and extension at 72°C for 1 min. Each cycle of amplification was repeated 35 times.
Ten μl of each sample were electrophoresed through a 1% agarose gel containing ethidium bromide. The specificity of amplification products was checked by confirming the known base-pair sequence length and by sequencing.
Direct sequencing of PCR products was performed with the use of the Dye Terminator Cycle Sequencing kit (ABI PRISM; Perkin-Elmer Applied Biosystem, Norwalk, CT). Sequences were resolved and analyzed on the ABI 373A Sequence Apparatus (Perkin-Elmer Applied Biosystem).
For semiquantitative RT-PCR analysis, nonsaturating conditions were used. In particular, one-fifth of the cDNA concentration tested above (i.e., 1 μl versus 5 μl) was used, and 30 cycles of amplification instead of 35 were performed.
Detection of IL-18 in Culture Supernatants.
Tonsil B-cell subsets were cultured in the presence or absence of different stimuli for 48 h at the concentration of 1 × 106 cells/ml in RMPI 1640 supplemented with 10% FCS. The stimuli tested were: anti-κ (1 μg/ml; Southern Biotechnology Associates, Birmingham, Alabama) in combination with anti-λ (1 μg/ml) immunoglobulin light chain mAbs (Southern Biotechnology Associates); CD40 mAb (1 μg/ml; Immunotech, Marseille, France), alone or in combination with recombinant (r)IL-4 (10 ng/ml; Genzyme, Cambridge, MA); human recombinant (hr)IFN-γ (1,000 units/ml; Boehringer Mannheim, Mannheim, Germany); hrIL-6 (100 units/ml) and hrIL-12 (kindly provided by Genetics Institute; 20 ng/ml); staphylococcus aureus (Calbiochem, La Jolla, CA; 1:10,000); and lipopolysaccharides (Sigma-Aldrich, St. Louis, MO; 10 ng/ml). Because anti-κ, anti-λ, and CD40 mAbs were all of IgG1 isotype, an isotype-matched mAb of irrelevant specificity was tested as control at the final concentration of 1 μg/ml. Cross-linking of surface immunoglobulin light chains was carried out by adding a goat antimouse immunoglobulin antiserum (Caltag) to the B-cell cultures containing anti-κ and anti-λ, or isotype-matched, mAbs. Purified B-CLL cells were cultured 48 h as above in the presence or absence of anti-immunoglobulin light chain mAbs under cross-linking conditions, or of CD40 mAb with or without rIL-4. BL cell lines were cultured for 48 h in medium alone.
Culture supernatants were harvested and tested in triplicate for IL-18 by ELISA, using a quantitative test that detects specifically the biological active form of IL-18 (MBL, Nagoya, Japan). The minimum threshold of this assay is 12.5 pg/ml.
Frozen tissue sections were fixed in ice-cold acetone for 10 min at room temperature and were dried and washed twice with Optimax Wash buffer. Endogenous peroxidase activity was blocked by 30-min incubation at room temperature with methanol containing 3% H2O2. Sections were washed twice in Optimax Wash buffer and were incubated overnight at 4°C with anti-human IL-18Rβ goat IgG at 5 μg/ml or with goat non-immune serum. Sections were washed three times in Optimax Wash buffer and were subsequently incubated with horseradish peroxide conjugate antigoat IgG (5 μg/ml). After washing in Optimax Wash buffer, peroxidase activity was detected by incubating the sections for 6–10 min with DAKO Liquid DAB Substrate Chromogen System (DAKO). Sections were counterstained with Mayer’s hematoxylin (Sigma).
In some experiments, which were complementary to those carried out by immunohistochemistry, double staining of frozen tonsil sections with CD19 mAb and anti-IL-18Rβ goat antiserum was performed by immunofluorescence. Sections were fixed in ice-cold acetone for 10 min at room temperature, were dried, were washed twice with PBS, and were incubated for 20 min at room temperature with blocking solution (PBS + 10% goat serum). Sections were next incubated overnight at 4°C with goat anti-human IL-18Rβ at 5 μg/ml or with goat non-immune serum. After washing in PBS, sections were stained with a PE-conjugated swine antigoat antibody for 45 min at 4°C. Subsequently, sections were washed three times in PBS and were incubated with FITC-conjugated CD19 mAb for 1 h at 4°C. After washing in PBS and drying, sections were mounted with Vectashield Mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI; Vector, Burlingame, CA). Digital images of FITC, rhodamine, and DAPI fluorescence were acquired separately by a CCD camera with highly selective filters using a Nikon Eclipse E1000 microscope (Nikon Instruments, Badhowedorp, the Netherlands).
Assay for Assessment of Apoptosis.
Purified tonsil B lymphocytes were cultured in the presence or absence of hrIL-18 (50 ng/ml) for 4–12 h, and GC apoptotic cells were detected by tricolor staining for CD38, annexin V, and propidium iodide by flow cytometry, using a kit according to the manufacturer’s instruction (Bender MedSystems, Vienna, Austria).
Purity of Neoplastic and Normal B-Cell Suspensions and Study Outline.
Expression of the CD3γ, CD56, CD68, and CD19 genes in purified normal or neoplastic B-cell fractions was investigated by RT-PCR, to exclude that even minute amounts of contaminant cell types were present in the cell suspensions.
CD3γ is selectively expressed in T cells, CD56 is a NK-associated marker, CD68 is a specific macrophage marker and CD19 expression is restricted to the B-cell lineage (41) . Only normal or malignant B-cell suspensions that tested negative for CD3γ, CD56, and CD68 mRNA expression were subjected to additional studies.
Purified normal or neoplastic B cells were tested for IL-18, IL-18Rα, and IL-18Rβ gene expression immediately after isolation by RT-PCR. To check the correspondence between IL-18 mRNA accumulation and protein secretion, B-cell culture supernatants were tested for the presence of bioactive IL-18 by ELISA. IL-18Rα surface expression was studied in naive, memory, and GC B lymphocytes, as well as in B-CLL cells, by flow cytometry. IL-18Rβ expression was investigated by immunohistochemistry on frozen sections from tonsils and on infiltrated B-CLL lymph nodes.
Expression of the IL-18 Gene and Production of the IL-18 Protein in Human Tonsil Naive, GC, and Memory B Lymphocytes.
Naive, GC, and memory B cells, freshly isolated from tonsils, were first tested for expression of the IL-18 gene. Fig. 1A⇓ shows one representative experiment in which the purity of tonsil B cells was assessed by RT-PCR, according to the protocol described in the previous paragraph. As apparent, the different B-cell fractions expressed CD19, but not CD3γ, CD56, or CD68 mRNA. Fig. 1B⇓ shows two experiments, representative of the 10 performed with superimposable results, in which the IL-18 transcript was found to be expressed in naive, GC, and memory B cells.
To investigate potential differences among the three B-cell subsets in IL-18 mRNA expression, the RT-PCR conditions were modified to allow a semiquantitative evaluation of the results. Fig. 1C⇓ shows one experiment representative of the three performed with comparable results. As apparent, the intensity of IL-18 transcript bands appeared of similar intensity in naïve, GC, and memory B cells.
Next, the same B-cell fractions were cultured for 48 h in the presence or absence of different stimuli (i.e., SAC, LPS, IFN-γ, IL-6, IL-12, CD40 mAb, and/or IL-4, anti-κ and anti-λ mAbs followed by cross-linking with goat antimouse immunoglobulin), and IL-18 was assayed in culture supernatants by an ELISA, which detects exclusively the biologically active form of the cytokine. CD40 mAb and IL-4 mimic in vitro T-dependent B-cell activation, whereas cross-linking of the B-cell receptor mimics T-independent B-cell activation. All of the other stimuli allow us to assess the effects of bacterial components or pro-inflammatory cytokines on B-cell IL-18 production.
Table 1⇓ shows the ranges of the results obtained in eight different experiments. Naive and GC B lymphocytes produced IL-18 in the absence of stimuli, and such production was only marginally decreased by CD40 mAb and anti-immunoglobulin mAbs. GC B lymphocytes produced lower amounts of IL-18 than did naive B cells (Table 1)⇓ . In contrast, memory B lymphocytes did not produce IL-18, either unstimulated or after in vitro stimulation (Table 1)⇓ . Notably, unfractionated tonsil B cells released IL-18 in culture supernatants in the absence of stimuli (1480–2120 pg/ml), indicating that production of the cytokine detected with purified naive and GC B cells was not induced by the separation procedures.
IL-18Rα and IL-18Rβ Expression in Human Tonsil Naive, GC, and Memory B Lymphocytes.
Fig. 2A⇓ shows that IL-18Rα and β transcripts were consistently expressed in freshly isolated naive, GC, and memory B lymphocytes. Two representative experiments of the 10 performed with identical results are shown.
Additional studies were carried out to investigate potential differences in IL-18Rα and β mRNA expression among naive, GC, and memory B cells, using the same semiquantitative RT-PCR technique referred to in the previous paragraph. Fig. 2B⇓ shows one such experiment, representative of the three performed with comparable results. As apparent, the bands corresponding to IL-18Rα and β transcripts appeared fainter in naïve B cells than in GC or memory B cells (Fig. 2B)⇓ , suggesting that the rate of IL-18Rα and β gene transcription was lower in the former than in the latter cells.
Next, surface expression of IL-18Rα was investigated by flow cytometric analysis of freshly isolated tonsil B cells. Tonsil B-cell fractions were stained with anti-IgD, CD38, and anti-IL-18Rα mAbs. Naive B cells were identified as IgD+, CD38− cells; GC B cells as CD38+, IgD− cells; and memory B cells as IgD−, CD38− cells.
Fig. 2C⇓ s hows the results obtained in three different experiments of the six performed. GC and memory B lymphocytes expressed surface IL-18Rα (Fig. 2C)⇓ in the following ranges: 36–53% for GC, and 32–56% for memory B cells. Notably, expression of IL-18Rα was detected both in CD38++ centroblasts and in CD38+ centrocytes (not shown). From 2 to 6% of naive B cells stained for IL-18Rα in the six experiments (Fig. 2C)⇓ .
Immunohistochemical studies were subsequently carried out by staining frozen tonsil sections with anti-IL-18Rβ antibodies. IL-18Rβ-positive cells were detected both in the GC and in the subepithelial area, whereas staining of the follicular mantle was predominantly negative, with the exception of scanty positive cells (Fig. 3A⇓ , left and middle panels). No staining at all was detected when non-immune goat serum was used in the place of the anti-IL-18Rβ antiserum (Fig. 3A⇓ , right panel).
Fig. 3B⇓ , panel 1, shows DAPI staining of a frozen tonsil section in which the mantle zone and the GC of a follicle are highlighted together with the subepithelial area, in which B cells are located (40) . Serial tonsil sections were stained by immunofluorescence with CD19 mAb, anti-IL-18Rβ antibodies, or both. To better visualize coexpression of CD19 and IL-18Rβ, we selected and inspected, at high magnification, microscopic fields enriched for CD19+ cells.
Fig. 3B⇓ , panel 2, shows an enlarged GC area, as detected by DAPI staining. Panels 3 and 4 show single staining with CD19 and anti-IL-18Rβ, respectively, whereas panel 5 shows triple staining with DAPI, CD19, and anti-IL-18Rβ. As apparent, in the latter panel, a yellow staining developed as a consequence of green (panel 3) and red (panel 4) color overlap, indicating that virtually all CD19+ cells in the GC coexpressed IL-18Rβ.
Fig. 3B⇓ , panels 6, shows DAPI staining of an enlarged subepithelial area, in which memory B cells settle (40) . Fig. 3B⇓ , panels 7 and 8, show single staining of the same area with CD19 and anti-IL-18Rβ, respectively, whereas panel 9 shows triple staining with DAPI, CD19, and anti-IL-18Rβ. Most CD19+ cells in the subepithelial area were found to coexpress IL-18Rβ, as indicated by the development of a yellow staining. Panel 10 shows the negative staining, obtained with fluorochrome-conjugated isotypic control of CD19 mAb, of a DAPI-counterstained tonsil tissue section.
Finally, most CD19+ B cells in the follicular mantle were IL-18Rβ negative (not shown). This is consistent with the immunohistochemical results shown in Fig. 3A⇓ .
When goat non-immune serum was tested in the place of anti-IL-18Rβ antibodies, no staining was observed (not shown).
These results indicated that the complete IL-18R, composed of the α and β chains, was expressed by GC and memory B cells, but not by naive B lymphocytes.
Because GC B cells were the only subset that produced constitutively the IL-18 protein and that expressed IL-18R on the cell surface, they were tested for propensity to undergo apoptosis in vitro after 4–12 h culture of purified tonsil B lymphocytes with or without rIL-18. The proportions of apoptotic GC B cells that were detected as CD38+, annexin V+, or propidium iodide+ cells were similar in the presence and in the absence of rIL-18 at all times tested, indicating that the cytokine was devoid of any pro- or antiapoptotic activity (not shown).
Expression of the IL-18 Gene in Neoplastic B Cells from MCL, FL, and MZL.
IL-18 gene expression in MCL, FL, and MZL B cells was next investigated. Fig. 4A⇓ shows the purity of the malignant B-cell suspensions from one each, representative MCL, FL, and MZL cases, in which only the CD19 transcript was consistently found (Fig. 4A)⇓ .
Fig. 4B⇓ shows IL-18 gene expression in B cells from two representative cases of MCL, FL, and MZL. Identical results were obtained with one additional MCL and four additional FL cases (not shown).
MCL B cells, the presumed counterpart of naive B lymphocytes, constitutively expressed the IL-18 transcript. The same result was obtained with B cells from FL, which originates from GC B lymphocytes, and from MZL, the postulated counterpart of memory B cells (Fig. 4B)⇓ . Because of the paucity of purified neoplastic B cells, production of IL-18 in culture supernatants could not be addressed.
Expression of the IL-18 Gene and Production of the IL-18 Protein by B-CLL Cells.
We next investigated IL-18 gene expression and production of bioactive IL-18 by malignant B cells purified from the peripheral blood of 12 B-CLL patients, 6 with unmutated and 6 with mutated IgV genes.
Fig. 4A⇓ shows the purity of one representative B-CLL case, as assessed by molecular analysis. Fig. 4B⇓ shows constitutive expression of IL-18 gene transcript in B cells from 2 representative B-CLL cases (one with unmutated, the other with mutated IgV genes), of the 12 tested with superimposable results.
To investigate the production of bioactive IL-18 in B-CLL culture supernatants, malignant B cells, purified from four different B-CLL cases (two with unmutated, two with mutated IgV genes), were cultured with medium alone, with CD40 mAb with or without rIL-4, or with anti-immunoglobulin light chain mAbs under cross-linking conditions for 48 h. IL-18 was never detected in cell culture supernatants (data not shown).
Expression of the IL-18 Gene and Production of the IL-18 Protein by BL Cell Lines.
Fig. 4B⇓ shows that IL-18 gene transcript was detected in six BL cell lines, which represent malignant expansions of GC B lymphocytes (32) , irrespective of their EBV+ (three cell lines) or EBV− (three cell lines) status. Constitutive production of the bioactive IL-18 protein was evaluated by ELISA, using supernatants from BL cell lines cultured for 48 h in the absence of stimuli. The assay was negative for all of the supernatants tested.
IL-18Rα and -β Gene Expression in Neoplastic B Cells from MCL, FL, and MZL.
Next, the expression of IL-18Rα and β transcripts in freshly isolated malignant B cells from three MCL, six FL, and two MZL cases was investigated. Variable expression of IL-18Rα and β, as assessed by RT-PCR, was detected in the individual cases. Thus, B cells from two of three MCL patients expressed the transcripts of both IL-18R chains, whereas, in the third case, neither transcript was found. Tumor cells from three of six FL patients showed expression of IL-18Rα and β mRNA. IL-18Rβ mRNA only was detected in two of six cases, and IL-18Rα transcript only was observed in one of six cases. Malignant B cells from one of two MZL patients expressed the transcripts of both IL-18R chains, whereas, in the other case, mRNA of either IL18Rα or β was not detected. Fig. 5A⇓ shows the results obtained with two each of representative MCL, FL, and MZL cases.
IL-18Rα and IL-18Rβ Expression in B-CLL.
Next, IL-18Rα and β expression was investigated at the mRNA level in the same 12 B-CLL cases already tested for IL-18 expression (see above). Fig. 5A⇓ shows that, in two representative B-CLL cases (one with unmutated, the other with mutated IgV genes), malignant B cells expressed the transcripts of the genes encoding both IL-18R components. The same pattern of expression was consistently detected in the remaining 10 B-CLL cases (not shown).
In subsequent experiments, IL-18Rα expression on B-CLL cells was studied by flow cytometry. Fig. 5B⇓ shows two representative cases (one with unmutated, the other with mutated IgV genes), in which IL-18Rα was detected on the surface of malignant cells. Expression of IL-18Rα was consistently observed in six additional B-CLL cases, three with unmutated and three with mutated IgV region genes (range of positive cells, 18–39%).
Finally, immunohistochemical studies with frozen sections from four B-CLL lymph nodes were carried out to investigate expression of the IL-18Rβ protein on the surface of tumor cells. All of these samples displayed effacement of the lymph node architecture. Fig. 5C⇓ shows the results obtained with two B-CLL lymph nodes, representative of the four studied. As apparent, the IL-18Rβ protein was detected on the surface of large cells, often referred to as paraimmunoblasts (42) , that appeared clustered in pseudofollicles. Most tumor cells with small lymphocyte morphology did not express IL-18Rβ (Fig. 5C)⇓ .
IL-18Rα and IL-18Rβ Gene Expression in BL Cell Lines.
In a final series of experiments, IL-18Rα and IL-18Rβ mRNA expression was investigated in the six BL cell lines. The results obtained with one representative EBV+ and one representative EBV− BL cell line are shown in Fig. 5A⇓ . The transcripts of the IL-18Rα and β chains were not detected in these, nor in the four remaining, BL cell lines.
In this study, we have investigated the expression of IL-18, IL-18Rα, and IL-18Rβ in human tonsil naive, GC, and memory B lymphocytes, and in their presumed neoplastic counterparts.
Expression of the IL-18 gene was consistently detected in freshly isolated normal B-cell subsets and in malignant B cells from MCL, FL, MZL, B-CLL, and BL. Only naive and GC B cells produced the bioactive IL-18 protein in culture supernatants, whereas this was not the case for memory B cells, B-CLL, and BL tumor cells.
Expression of IL-18Rα and IL-18Rβ transcripts was always observed in tonsil naive, GC, and memory B cells. Cytofluorimetric studies showed that IL-18Rα was expressed in GC and memory B cells but was virtually undetectable in naive B cells. Likewise, the bulk of tonsil GC and subepithelial, but not follicular mantle, B cells were found, by immunohistochemistry, to express IL-18Rβ. The finding that the IL-18Rα and IL-18Rβ proteins were poorly expressed in naive B cells may be related to transcriptional and/or posttranscriptional events. Semiquantitative RT-PCR experiments, indeed, suggested that the transcriptional rate of IL-18Rα and IL-18Rβ genes was lower in naive than in GC or memory B lymphocytes.
Finally, variable positivity for mRNA expression of IL-18Rα or β was found in MCL, FL, and MZL samples. B-CLL cases showed consistent expression of IL-18Rα and β transcripts, whereas neither transcript was detected in six BL cell lines. A fraction of malignant B cells in the individual B-CLL clones expressed IL-18Rα and β proteins.
The finding that GC B cells had the unique feature of producing IL-18 and expressing the two chains of the IL-18R on the cell surface, raised the possibility that the cytokine could modulate programmed cell death in a paracrine and/or autocrine manner. However, no evidence in support of this hypothesis was obtained from in vitro experiments performed in the presence of rIL-18.
Alternative functions of IL-18 produced by naive and GC B cells might be stimulation of immunoglobulin and IFN-γ synthesis by the B cells themselves, as shown in human and murine models (43 , 44) , or interaction with other cell types present in secondary lymphoid organs, such as T cells or follicular dendritic cells.
As mentioned, tonsil naive B cells displayed low-to-absent surface expression of IL-18Rα and β. However, we have previously demonstrated that IL-12 up-regulates the expression of both IL-18R chains in tonsil B cells, and synergizes with IL-18 in the induction of IFN-γ production by IgD+ naive B cells (43) . These latter findings, together with the results of the present study, suggest that resting naive B cells are unresponsive to IL-18 but become responsive to it after interaction with IL-12, e.g., in the course of an inflammatory process.
Despite consistent expression of IL-18 mRNA, memory B cells did not produce the bioactive cytokine in culture supernatants, possibly as a consequence of posttranscriptional events.
Studies carried out in different murine models have demonstrated that B cells presenting antigenic peptides to specific T cells do not drive generation of Th1 cells (45 , 46) . Although the major cytokines involved in the induction of Th1 cell polarization are IFN-γ and IL-12 (47, 48, 49) , it has been shown that also IL-18, especially in synergism with IL-12, participates in this process (5 , 8 , 11 , 12) . Thus, it is tempting to speculate that failure of memory B cells, the most efficient APCs within the B-cell lineage, to produce IL-18 may contribute to their inability to promote Th1 differentiation.
The investigation of IL-18, IL-18Rα, and IL-18Rβ expression in B-cell lymphoproliferative disorders disclosed remarkable differences in comparison with their postulated normal counterparts. RT-PCR experiments showed that MCL, FL, and MZL B cells displayed heterogeneous expression of IL-18Rα and -β mRNA, with approximately one-half of the cases devoid of either or both transcripts. Furthermore, six cell lines from BL, a tumor originating from GC B cells (32) , did not express IL-18Rα or -β mRNA.
As mentioned, B-CLL is composed of at least two major subgroups based on the absence of presence of IgV gene mutations, possibly originating from pre- and post-GC B cells, respectively (33, 34, 35) . The B-CLL cases here investigated showed consistent expression of IL-18Rα and -β mRNA, irrespective of the IgV gene mutational status. Surface expression of IL-18Rα was detected in B-CLL cells isolated from the peripheral blood, whereas IL-18Rβ protein expression was investigated in infiltrated B-CLL lymph nodes. The latter experiments demonstrated that the majority of small B-CLL lymphocytes were unreactive with the anti-IL-18Rβ antibody, whereas larger malignant B cells, morphologically identifiable as paraimmunoblasts, displayed consistent staining.
IL-18Rα and -β expression on the surface of B-CLL cells were obtained from discrete samples because of technical limitations of the antibodies used. However, neither chain of the IL-18R was detected on the totality of malignant B cells from the individual cases, suggesting that expression of the complete heterodimeric receptor was not clonally distributed.
Neither B-CLL cells nor BL cell lines produced bioactive IL-18, despite the consistent expression of IL-18 mRNA. Because the B-CLL samples included mutated and unmutated cases, the negative results obtained with the latter are discordant in relation to those observed with their possible normal counterparts, i.e., naive B cells. Failure of BL tumor cells to produce IL-18 is clearly at variance with the behavior of their normal postulated counterparts, i.e., GC B cells.
In summary, the present study suggests that malignant B lymphocytes from different B-cell chronic lymphoproliferative disorders display heterogeneous expression of IL-18 and of the two components of the IL-18R, as compared with their normal counterparts. It is tempting to speculate that deranged expression of the cytokine and/or its receptor in neoplastic B cells may in some cases contribute to the facilitating of tumor growth, e.g., by impairing antitumor effector mechanisms because of down-regulation of IL-18 production.
We thank Dr. Annalisa Pezzolo for help in the immunofluorescence experiments with tonsil tissue sections, Dr. Franco Fais for the analysis of IgV gene mutations in the B-CLL samples, Marta Camoriano for help in immunohistochemistry studies, and Chiara Bernardini for excellent secretarial assistance.
Grant support: Grants from Associazione Italiana per la Ricerca sul Cancro (to V. P. and S. R.) and from Ministero della Salute (to V. P.).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342; Fax: 39-010-3779820; E-mail:
- Received July 11, 2003.
- Revision received September 18, 2003.
- Accepted September 18, 2003.