IL-12Rβ2 gene functions as a tumor suppressor in human B cell malignancies" />
Advertisement
ArticleOncology Free access | 10.1172/JCI20303
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Airoldi, I. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Di Carlo, E. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Banelli, B. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Moserle, L. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Cocco, C. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Pezzolo, A. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Sorrentino, C. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Rossi, E. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Romani, M. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Amadori, A. in: JCI | PubMed | Google Scholar
1Laboratory of Oncology, G. Gaslini Institute, Genoa, Italy. 2Department of Oncology and Neurosciences, G. D’Annunzio University, Chieti, Italy. 3Laboratory of Tumor Genetics, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. 4Department of Oncology and Surgical Sciences, University of Padua, Padua, Italy. 5Department of Hematology, Azienda Ospedaliera S. Martino, Genoa, Italy.
Address correspondence to: Irma Airoldi, Laboratory of Oncology, G. Gaslini Institute, Largo G. Gaslini 5, 16148 Genoa, Italy. Phone: 39-010-5636342, 39-010-5636524; Fax: 39-010-3779820; E-mail: laboncologia@ospedale-gaslini.ge.it.
Find articles by Pistoia, V. in: JCI | PubMed | Google Scholar
Published June 1, 2004 - More info
The IL-12Rβ2 gene is expressed in human mature B cell subsets but not in transformed B cell lines. Silencing of this gene may be advantageous to neoplastic B cells. Our objective was to investigate the mechanism(s) and the functional consequence(s) of IL-12Rβ2 gene silencing in primary B cell tumors and transformed B cell lines. Purified tumor cells from 41 patients with different chronic B cell lymphoproliferative disorders, representing the counterparts of the major mature human B cell subsets, tested negative for IL-12Rβ2 gene expression. Hypermethylation of a CpG island in the noncoding exon 1 was associated with silencing of this gene in malignant B cells. Treatment with the DNA methyltransferase inhibitor 5-Aza-2′-deoxycytidine restored IL-12Rβ2 mRNA expression in primary neoplastic B cells that underwent apoptosis following exposure to human recombinant IL-12 (hrIL-12). hrIL-12 inhibited proliferation and increased the apoptotic rate of IL-12Rβ2–transfected B cell lines in vitro. Finally, hrIL-12 strongly reduced the tumorigenicity of IL-12Rβ2–transfected Burkitt lymphoma RAJI cells in SCID-NOD mice through antiproliferative and proapoptotic effects, coupled with neoangiogenesis inhibition related to human IFN-γ–independent induction of hMig/CXCL9. The IL-12Rβ2 gene acts as tumor suppressor in chronic B cell malignancies, and IL-12 exerts direct antitumor effects on IL-12Rβ2–expressing neoplastic B cells.
IL-12 is a heterodimeric cytokine bridging innate resistance and antigen-specific adaptive immunity (1). IL-12 is produced predominantly by professional APCs and serves as an important regulator of T cell and NK cell function (2, 3). IL-12 has a powerful antitumor activity related to both IFN-γ–dependent and –independent mechanisms (4–9).
Numerous effects of IL-12 on human B cells from normal individuals have been reported, such as induction of proliferation and differentiation to Ig-secreting cells (10), upregulation of IL-18R (11), and induction of IFN-γ mRNA and protein (11–13).
The biological functions of human IL-12 are mediated by the heterodimeric IL-12R composed of two subunits, the β1 and the β2 chains, conferring high-affinity binding of and responsiveness to IL-12 (14, 15). The β1 chain participates in the formation of the IL-23 receptor, whereas the β2 chain is an unique component of IL-12R (1). The IL-12Rβ2 gene maps on chromosome 1p31.2 and is composed of 16 introns and 16 exons, the first of which is not transcribed (15).
We have previously shown that human lymphoblastoid B cell lines (LCLs) and Burkitt lymphoma (BL) cell lines express constitutively IL-12Rβ1 mRNA but lack IL-12Rβ2 mRNA (16). In contrast, normal human naive, germinal center (GC), and memory B cells express constitutively the transcripts of both IL-12R genes (11). These findings led us to speculate that malignant B cells could benefit from the silencing of the IL-12Rβ2 gene.
Here we first investigated IL-12R gene expression in primary tumor cells from patients with different chronic B cell lymphoproliferative disorders representing the postulated counterparts of the major mature B cell subsets, as no information was available on this issue. IL-12Rβ2 mRNA expression was not detected in any sample. Second, we studied the mechanism(s) involved in and the functional consequences of the silencing of this gene in human primary neoplastic B cells, as well as in transformed B cell lines. Our findings support the conclusion that the IL-12Rβ2 gene functions as tumor suppressor in a wide spectrum of human B cell malignancies.
Expression of IL-12R in normal tonsil B cell subsets and in malignant B cells from different human chronic lymphoproliferative disorders. Figure 1A shows one experiment, representative of the ten performed with identical results, in which the IL-12Rβ1 and β2 transcripts were detected in naive, GC, and memory B cells isolated from tonsil, consistently with a previous study (11). As is apparent, these B cell fractions expressed CD19, but not CD3γ, CD56, or CD68, mRNA.
Expression of IL-12R in human tonsil B cells. (A) Expression of IL-12Rβ1 and IL-12Rβ2 mRNA in tonsil B cells and their subsets. From left to right: molecular weight markers (MW); negative control (NC, represented by a Th2 cell clone for CD56, CD19, and IL-12Rβ2 primers; purified tonsil B cells for CD3γ primers; or water in the place of cDNA for IL12Rβ1 and GAPDH primers); positive control (PC, represented by a Th1 cell clone for CD3γ, IL12Rβ1, and IL-12Rβ2 primers; tonsil non_T cells for CD56 primers; or the RPMI 8866 B cell line for CD19 primers). The remaining lanes represent total tonsil B lymphocytes and naive, GC, and memory B cells, respectively, isolated from the same tonsil. On the right, the expected molecular weight of the amplified bands are shown. (B_N) Expression of IL-12Rβ2 protein in frozen tonsil tissue sections. (B) FM, GC, and SE, are boxed (magnification, ∞20). FM staining with CD19 mAb (C, green), anti_IL-12Rβ2 mAb (D, red), and both mAb's in combination with DAPI (E); in E, overlap of green and red colors gives rise to yellow staining, indicating that most cells coexpress CD19 and IL-12Rβ2; control FM staining with DAPI and with isotype- and fluorochrome-matched mAb's of irrelevant specificity (F). Staining of SEs (G_J) and GC areas (K_N) are shown in the same order as FM. Magnification (C_N), ∞100 for all panels.
In subsequent studies, frozen tonsil tissue sections were stained sequentially with phycoerythrin-conjugated (PE-conjugated) anti–IL-12Rβ2 and FITC-conjugated CD19 mAb’s or isotype and fluorochrome matched control mAb’s and then with DAPI. Images were acquired using either a filter selective for fluorochrome (Figure 1, C, G, K and D, H, L) or a triple-band — i.e., DAPI, PE, and FITC–specific — filter (Figure 1, E, I, M and F, J, N).
Figure 1B shows a DAPI staining with the follicular mantle (FM), the GC, and the subepithelial (SE) area in boxes. Figure 1, C, G, K and D, H, L, shows staining with CD19 (green) or anti–IL-12Rβ2 (red) mAb’s, respectively. As is apparent, the large majority of cells positioned in the FM, GC, and SE areas expressed CD19 and IL-12Rβ2. As shown in Figure 1, E, I, and M, a yellow staining developed as a consequence of green and red color overlap, indicating that virtually all CD19+ cells in the FM, GC, and SE areas coexpressed IL-12Rβ2. No staining at all was detected when isotype-matched irrelevant mAb’s conjugated with the same fluorochromes were used in place of the anti–IL-12Rβ2 and CD19 mAb’s (Figure 1, F, J, and N).
Next, we investigated IL-12R mRNA expression in primary malignant B cells. Figure 2 (left panel) shows the results of two representative experiments carried out with mantle cell lymphoma (MCL), marginal zone lymphoma (MZL), and follicular lymphoma (FL) B cells, respectively. As is apparent, all samples constitutively expressed IL-12Rβ1 (upper panel), but not IL-12Rβ2 (lower panel), mRNA. The same results were obtained when malignant B cells purified from four additional MCL, ten additional FL, and two additional MZL cases were tested (not shown). B chronic lymphocytic leukemia (B-CLL) cells always expressed IL-12Rβ1 mRNA but tested consistently negative for IL-12Rβ2 gene expression. Figure 2 (right panel) shows the results obtained with 15 of the 19 B-CLL samples tested.
IL-12R expression in human neoplastic B cells. Left panel (left to right): molecular weight markers; negative control, represented by a Th2 clone; positive control (tonsil B cells); two representative cases (patients 1 and 2 [Pt 1 and Pt 2]) each of MCL, MZL, and FL are shown. Right panel (left to right): molecular weight markers; negative control, represented by a Th2 clone; positive control (tonsil B cells). Fifteen B-CLL cases (Pt 1 to Pt 15) are shown.
These findings indicate that primary malignant B cells isolated from different chronic tumors, analogously to LCLs and BL cell lines, do not express IL-12Rβ2 mRNA (16).
DNA methylation is involved in IL-12Rβ2 gene silencing. The IL-12Rβ2 gene is composed of 16 exons encompassing approximately 110 kb of DNA at 1p31.2. The detailed computer analysis of this region revealed the presence of several CG-rich regions in this genomic segment. However, only one of these regions fulfilled all the stringent criteria to be considered a CpG island. This island maps at position –69/+888 with respect to nucleotide 1 of the reference IL-12Rβ2 cDNA sequence NM_001559 and includes the entire untranslated first exon of IL-12Rβ2. In Figure 3A (top panel), we report the ratio of expected versus observed CpG along with the relative positions of IL-12Rβ2 exon 1 and of the CpG island.
Methylation analysis of nepolastic B cells versus their normal counterparts. (A) Top panel: Plot of the expected versus observed CpG dinucleotides frequency surrounding exon 1 of the IL-12Rβ2 gene. The CpG island and exon 1 are indicated. The broken line indicates the cut-off value of 0.60. Bottom panel: MSP analysis of cells expressing (Th1 clone, unfractionated tonsil B cells, and purified naive, GC, and memory B cells) and not expressing (RPMI 8866 LCL and Raji, BRGM, and Daudi BL cells) the IL12Rβ2 gene. Cells expressing the gene show only the amplification band corresponding to the unmethylated sequence (UM), whereas in nonexpressing cells, this band is present only in trace amounts, and the major product of amplification corresponds to the methylated (M) target sequence. (B) Induction of IL-12Rβ2 gene expression in Raji and RPMI 8866 cell lines treated with 5-Aza-2′-deoxycytidine. One experiment representative of the five performed is shown. From left to right: molecular weight markers; B cells cultured with medium alone (medium) or in the presence of 5ΒM 5-Aza-2′-deoxycytidine for 24 to 96 hours.
Having determined that a bona fide CpG island characterizes the known 5′ boundary of the gene, we investigated the contribution of methylation in the silencing of its expression by methylation-specific PCR (MSP). As shown in Figure 3A (bottom panel), methylation of this CpG island was detected in the RPMI 8866 LCL and in the Raji, BRGM, and DAUDI BL cell lines but not in total tonsil B cells or in purified naive, GC, or memory B cells.
Induction of IL-12Rβ2 gene expression in primary neoplastic B cells and transformed B cell lines after in vitro treatment with 5-Aza-2′-deoxycytidine. These experiments were carried out first using the RPMI 8866 LCL and the Raji BL cell line, which do not express the IL-12Rβ2 gene (16). Figure 3B shows that the IL-12Rβ2 transcript was expressed de novo in Raji and RPMI 8866 cell lines after 72 hours incubation with the DNA methyltransferase inhibitor 5-Aza-2′-deoxycytidine. The transcript was no longer detectable in either cell line following 96-hour treatment with 5-Aza-2′-deoxycytidine.
In further experiments, four cases of B-CLL (two with mutated IgV genes, two with unmutated IgV genes) and two cases of FL were studied. B-CLL cells were cultured in the presence or absence of 5-Aza-2′-deoxycytidine, a selective DNA methyltransferase inhibitor (17), for 24 to 96 hours before being tested for IL-12Rβ2 mRNA expression. Due to the limited number of cells available, FL cells were cultured with 5-Aza-2′-deoxycytidine for 72 hours only before being tested. Figure 4A (left panel) shows one B-CLL sample out of the four tested with superimposable results, in which the IL-12Rβ2 gene was expressed de novo 48–96 hours after 5-Aza-2′-deoxycytidine treatment. In Figure 4A (right panel), a representative experiment with purified FL cells is shown, in which the IL-12Rβ2 transcript was induced after 72-hour culture with 5-Aza-2′-deoxycytidine.
Induction of IL-12Rβ2 by treatment with 5-Aza-2′-deoxycytidine in primary tumors. (A) Left panel: Freshly isolated B-CLL cells (T0), B-CLL cells cultured with medium alone (medium) or in the presence of 5 ∝M 5-Aza-2′-deoxycytidine (Aza) for 48 to 96 hours. One sample representative of the four tested is shown. Right panel. Purified FL cells cultured with medium alone (medium) or in the presence of 5-Aza-2′-deoxycytidine for 72h. One sample representative of the two tested is shown. (B) Apoptosis of primary neoplastic B cells incubated with 5-Aza-2′-deoxycytidine for 72 hours and subsequently cultured with hrIL-12 for an additional 48 hours, as assessed by the TUNEL assay. Histogram in the left panel shows the percentages of apoptotic cells detected following culture with medium alone (white bar); with 5-Aza-2′-deoxycytidine for 72 hours and subsequently with medium for additional 48 hours (gray bar); or with 5-Aza-2′-deoxycytidine for 72 hours and subsequently with hrIL-12 for an additional 48 hours (black bar, Aza + IL-12). Two B-CLL and 2 FL samples are shown. In the right panel, one representative experiment performed with B-CLL1 sample is shown. TUNEL positive nuclei stained with FITC are green, whereas interphase nuclei stained with DAPI are blue (magnification, ∞20).
In order to investigate whether 5-Aza-2′-deoxycytidine treatment rendered primary malignant B cells responsive to IL-12, two of the four B-CLL and the two FL samples tested above were first cultured in the presence of the drug for 72 hours and then incubated with human recombinant IL-12 (hrIL-12) or medium for an additional 48 hours before being subjected to apoptosis assays. Controls were the same neoplastic B cell suspensions cultured with medium alone for the whole time interval.
Figure 4B (left panel) shows that a 3-day course of 5-Aza-2′-deoxycytidine approximately doubled the proportion of apoptotic tumor cells from all samples as compared with controls. A subsequent 2-day course of hrIL-12 induced a further increase in the apoptotic rate ranging from 37% to 49% in 2:2 B-CLL and in 1:2 FL cell suspensions (FL2), as assessed by the TUNEL assay. In the remaining FL case (FL1), the percentage of apoptotic cells was not modified by hrIL-12 treatment. Figure 4B (right panel), shows a representative experiment in which a higher proportion of TUNEL-positive B-CLL cells was detected following incubation with hrIL-12, as compared with the same cells treated for 72 hours with 5-Aza-2′-deoxycytidine or with control.
Characterization of IL-12Rβ 2 gene–transfected RPMI 8866 and Raji B cell lines. Additional characterization of the functional consequences of IL-12Rβ2 gene silencing in malignant B cells was carried out using hIL-12Rβ2–transfected B cell lines. The RPMI 8866 LCL and the Raji BL cell line were stably transfected with a plasmid bearing the hIL-12Rβ2 gene or with the empty plasmid. IL-12Rβ2–transfected RPMI 8866 and Raji cells were used for in vitro studies, whereas only the latter cells could be tested for tumorigenicity in SCID-NOD mice (see below). About 95% of G418-resistant transfected cells expressed green fluorescent protein (GFP), as assessed by fluorescence microscopy. However, IL-12Rβ2–transfected cells displayed surface GFP expression, whereas empty vector–transfected cells showed intracellular GFP expression (not shown).
The IL-12Rβ2 transcript was detected by RT-PCR in IL-12Rβ2, but not in empty vector, transfectants (Figure 5A). IFN-γ gene expression is induced de novo in different cell types by incubation with IL-12 (1). Therefore, in the following experiments, IL-12Rβ2–transfected Raji cells that did not express constitutively IFN-γ mRNA, were incubated with or without hrIL-12 and tested for IFN-γ gene expression by RT-PCR. IFN-γ mRNA was induced by hrIL-12 in IL-12Rβ2–transfected cells but not in empty vector transfectant, indicating that the reconstituted IL-12R was functional (Figure 5B). IL-12Rβ2–transfected RPMI 8866 cells were not tested in these experiments, as they expressed constitutively IFN-γ mRNA (not shown).
Characterization of IL-12Rβ2_transfected cells. (A) Expression of IL-12Rβ2 mRNA in IL-12Rβ2 and empty-vector_transfected Raji and RPMI 8866 cells. From left to right: molecular weight markers; negative control, represented by a Th2 clone; positive control (tonsil B cells); Raji cells transfected with the empty vector (vector) or with the IL-12Rβ2_containing vector (IL-12Rβ2); RPMI 8866 cells transfected with the empty vector or with the IL-12Rβ2_containing vector. (B) Induction of IFN-γ mRNA in Raji cells transfected with the empty vector or with the IL-12Rβ2_containing vector after treatment with IL-12 for 36 hours.
Inhibition of cell proliferation and induction of apoptosis in IL-12Rβ 2 gene–transfected Raji and RPMI 8866 B cells upon culture with hrIL-12. The in vitro effects of hrIL-12 on RPMI 8866 and Raji cells transfected with the IL-12Rβ2 gene were next investigated. Figure 6, A and B (left and middle panels), shows the inhibition of the in vitro growth of IL-12Rβ2–transfected cells following incubation with hrIL-12, as assessed by cell count and [H3]-thymidine incorporation. IL-12Rβ2–transfected Raji cells and RPMI 8866 cells showed a significant decrease in cell count following 72 (P = 0.002) and 96 hours (P = 0.01) incubation with hrIL-12, respectively, as compared with cultures performed with medium alone. One representative experiment out of the five carried out for each cell line is shown in Figure 6, A and B (left panel).
Assessment of proliferation and apoptosis of IL-12Rβ2_transfected Raji (A) and RPMI 8866 (B) cells following incubation with hrIL-12. (A and B, left panel) cells were cultured with hrIL-12 for 24 to 96 hours and counted by an electronic counter. One representative experiment out of the five performed is shown. (A and B, middle panel) [H3]-thymidine incorporation by cells incubated with hrIL-12 for 72 hour. The means ± SD from five independent experiments are shown. Results are expressed as cpm (A and B, right panel). Cells were cultured with hrIL-12 for 5 to 96 hours and tested for apoptosis by the TUNEL assay. The means ± SD from three independent experiments are shown. Results are expressed as percent apoptotic cells.
In five different experiments, statistically significant inhibition of [H3]-thymidine incorporation by IL-12Rβ2–transfected Raji cells and RPMI 8866 (P < 0.002 and < 0.0009, respectively) was detected after 72 hours treatment with hrIL-12, as compared with incubation with medium alone (Figure 6, A and B, middle panels).
Figure 6, A and B (right panel), shows the results of three independent experiments in which IL-12Rβ2–transfected Raji and RPMI 8866 cells underwent apoptosis following culture with hrIL-12. As compared with control cells incubated with medium alone, hrIL-12 induced apoptosis of both cell lines was significantly higher after 5 hours culture (P < 0.0002) and reached a plateau from 24 to 96 hours (P < 0.002 for 24, 48, 72 and 96 h).
Cell count, [H3]-thymidine incorporation, and percent apoptotic cells were unchanged at all times tested in IL-12Rβ2 transfectants cultured with medium alone, as well as in wild-type and empty-vector transfected cells incubated with hrIL-12 (Figure 6, A and B).
In summary, these experiments demonstrated that hrIL-12 exerted in vitro antiproliferative and proapoptotic effects on IL-12Rβ2–transfected B cell lines.
hrIL-12 strongly reduces tumorigenicity of IL-12Rβ 2–transfected Raji cells in SCID-NOD mice. In subsequent experiments, the tumorigenicity of IL-12Rβ2–transfected, empty-vector transfected, and wild-type Raji BL cells was investigated in SCID-NOD mice. By the end of the follow-up period, all mice developed tumors that grew in the peritoneal cavity and invaded the ovaries by contiguity; no metastases were detected at distant sites.
Mice injected with IL-12Rβ2–transfected Raji cells and treated with hrIL-12 developed tumors significantly smaller (P < 0.0001) than did mice inoculated with the same cells and treated with PBS (Figure 7A). A representative experiment is shown in Figure 7B. The size of tumors formed in SCID-NOD mice by empty-vector transfected or wild-type Raji cells was unaffected by hrIL-12 or PBS administration (not shown).
Tumorigenicity of IL-12Rβ2_transfected Raji cells in SCID-NOD mice. (A) The mean size ± SD of tumors formed by IL-12Rβ2_transfected Raji cells in SCID-NOD mice treated with hrIL-12 (black bar) or PBS (gray bar) is shown. The difference between mean size of tumors from hrIL-12 treated and PBS treated mice was statistically significant (P < 0.0001). (B) Two representative tumors grown in hrIL-12_treated and PBS-treated mice are shown. (C) Histological and immunohistochemical features of tumors from PBS treated (C, G, E, and I) and hrIL-12 treated (D, F, H, and J) SCID/NOD mice. IL-12Rβ2 gene_transfected Raji tumors from PBS-treated mice are formed by medium- to large-sized neoplastic cells with round to oval nuclei showing two or more nucleoli. The “starry sky” pattern determined by macrophages that have ingested apoptotic tumor cells is also present (arrows) (C). The tumor is well vascularized (E) and does not express detectable level of Mig/CXCL9 (G). An extremely high proliferation rate is evidenced by Ki-67 immunostaining (I). In tumors from hrIL-12 treated mice the above described architectural features are altered by frequent and extensive areas of ischemic-hemorrhagic necrosis (N) (D), which are associated with defective intratumoral microvessel network (F), and expression of Mig/CXCL9 in the cytoplasm and in the proximity of intact tumor cells (H). In addition, the number of proliferating cells in the viable neoplastic tissue surrounding necrotic areas, as assessed by Ki-67 staining, is strongly reduced (J). Magnification, ∞400, for all pictures shown.
Histological and immunohistochemical studies were focused on tumors from mice injected with IL-12Rβ2 gene–transfected Raji cells and treated with either PBS (control) or hrIL12. Tumors from control mice were formed by medium- to large-sized cells showing round to oval nuclei with two or more nucleoli. Mitotic figures were frequent and aspects of spontaneous cell death were also present, as evidenced by the typical “starry sky” pattern imparted by benign macrophages that have ingested apoptotic tumor cells (Figure 7C). Areas of ischemic necrosis were occasionally observed in the larger tumor masses (not shown).
In contrast, in mice treated with hrIL-12, tumors showed multiple and wide areas of ischemic-hemorrhagic necrosis (Figure 7D). These features were associated with defective and damaged intratumoral microvessel network, as assessed by anti–endothelial cell (CD31) staining, which revealed that microvessels were significantly less numerous (P < 0.005; Table 1) in tumors from hrIL-12–treated (Figure 7F) than control (Figure 7E) mice.
Immunohistochemical analyses of tumors developed 4 weeks after i.p. injection of IL-12Rβ2 gene_transfected Raji B cells in SCID/NOD mice treated with PBS or with hrIL-12
A potential mechanism involved in angiogenesis inhibition is suggested by the expression of hMig/CXCL9, an antiangiogenic chemokine, in the viable tumor tissue surrounding necrotic areas from hrIL-12 treated (Figure 7H), but not from control (Figure 7G), mice. This difference was statistically significant (P < 0.005; Table 1). hIP-10/CXCL10 (Table 1) and human IFN-γ (hIFN-γ) (not shown) were not detected in tumors from either hrIL-12 treated or control mice.
Finally, tumors from hrIL-12–treated animals showed a significantly lower proliferation index (P < 0.005; Table 1), as assessed by Ki-67 immunoreactivity (Figure 7, I and J), and a significant increase in the number of apoptotic cells (P < 0.005; Table 1), as assessed by the TUNEL assay, in comparison with tumors from control animals.
This study provides the first demonstration that malignant B lymphocytes from patients with different chronic lymphoproliferative disorders do not express the IL-12Rβ2 gene, as opposed to their normal counterparts in secondary lymphoid organs (11), and suggests that this gene functions as a tumor suppressor in human B cell malignancies. Thus, loss of expression of the IL-12Rβ2 gene in neoplastic B cells was found to be advantageous for tumor growth, survival, and proliferation, both in vitro and in vivo.
An increasing body of evidence indicates that epigenetic mechanisms may be as important as mutations in gene inactivation associated with oncogenesis. Epigenetic transcriptional repression has been documented in a wide variety of tumor types and may affect tumor suppressor genes, cell cycle genes, DNA repair genes, as well as genes involved in invasion and metastasis. Furthermore, reexpression of many of these genes in tumor cells can suppress cell growth or change their sensitivity to chemotherapeutic agents (17, 18). DNA methylation in cancer cells usually affects CpG islands located in the promoter region of target genes (17, 18).
Our results indicate that DNA hypermethylation contributes to the regulation of the expression of the IL-12Rβ2 gene. Indeed, a CpG island was identified within the noncoding exon 1 of the gene, and this island was methylated in neoplastic, but not in normal, B cells. IL-12Rβ2 mRNA was induced de novo in primary tumor cells and in transformed B cell lines cultured in the presence of 5-Aza-2′-deoxycytidine. Finally, the latter cells, purified from patients with B-CLL and FL, underwent apoptosis when exposed to hrIL-12 after pretreatment with 5-Aza-2′-deoxycytidine.
This finding indicated, we believe for the first time, that IL-12 was endowed with direct inhibitory effects on primary malignant B cells expressing both chains of the IL-12R. Accordingly, hrIL-12 decreased in vitro proliferation and survival of LCL and BL cells transfected with the IL-12Rβ2 gene. Furthermore, IL-12Rβ2–transfected Raji BL cells formed smaller tumors in SCID-NOD mice treated with hrIL-12 than in control animals. The SCID-NOD mouse model allowed us to assess the direct antitumor activity of hrIL-12 on IL-12Rβ2–transfected Raji BL cells, as hIL-12 does not bind to the murine IL-12R and vice versa (19).
Histological and immunohistochemical studies revealed that hrIL-12 exerted antiproliferative and proapoptotic effects on IL-12Rβ2–transfected tumor cells in vivo, similarly to what we observed in vitro. In addition, strong inhibition of angiogenesis and vascular injury resulting in widespread areas of ischemic-hemorrhagic necrosis were detected in IL-12Rβ2–transfected tumors from hrIL-12 treated mice.
Previously it was shown that the antiangiogenic effects of IL-12 are strictly dependent on IFN-γ–induced IP-10/CXCL10 and Mig/CXCL9 (20, 21). In this study, immunohistochemical analyses demonstrated an accumulation of hMig/CXCL9, but not hIP-10/CXCL10, in IL-12Rβ2–transfected tumor cells from mice treated with hrIL-12. However, hIFN-γ was never detected in these cells. Such findings suggest that the antiangiogenic effects of hrIL-12 in IL-12Rβ2–transfected tumors were related to hIFN-γ–independent induction of hMig/CXCL9 (22). In this respect, it is of note that residual Mig/CXCL9 production has been demonstrated in IFN-γ KO mice (23).
Previous studies in human and murine models have shown that IL-12 stimulates, rather than inhibits, B cell proliferation and differentiation (10, 12). Therefore, it is conceivable that the antiproliferative and proapoptotic effects of IL-12 here reported are related to the transformed state of the target cells. Silencing of the IL-12Rβ2 gene may be an early event in the process of B cell tumorigenesis, as its expression was not detected in LCLs, which are immortal but not tumorigenic (16).
IL-12 is under scrutiny as an antitumor agent in patients with B cell malignancies (24, 25) due to its capacity to activate the host immune system. The present results suggest that IL-12 may be targeted directly to neoplastic B cells in the clinical setting, provided that a combination schedule with 5-Aza-2′-deoxycytidine is designed (26). A note of caution arises from the observation that epigenetic silencing usually recurs in cells that have reexpressed a given gene following incubation with demethylating agents (18). In principle, this problem may be circumvented by administering the cytokine during this window of demethylation. Further studies on these issues are now in progress in our laboratory.
Patient samples. This investigation was performed after approval by a local Institutional Review Board of the G. Gaslini Institute. Lymph node biopsies from 12 patients with FL, six patients with MCL, and four patients with MZL were obtained from the pathologists after completion of all diagnostic procedures. Peripheral blood samples from 19 B cell chronic lymphocytic leukemia (CLL) patients were obtained with informed consent. Diagnosis was established according to the criteria of the WHO classification (27). Clonal excess was assessed by the κ/λ or λ/κ Ig light chain ratio, which ranged from a minimum of 8:1 to a maximum of 20:1 in the different cases. All patients (22 males, 19 females, age range 40–75 years) were untreated at the time of study.
According to the WHO classification (27), FL originates from GC B cells; most 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. Based on findings of recent studies, B-CLL can be dissected into different subsets according to the mutational status of Ig variable (V) region genes or to the expression of the CD38 or ZAP-70 markers (28–30). 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 (31, 32). Due to such heterogeneity, nine peripheral blood samples from unmutated and 10 from mutated B-CLL cases were tested in this study. Analysis of IgV gene mutations was carried out by Franco Fais, Department of Experimental Medicine, University of Genoa, as published (28).
Cell separation and RT-PCR. Total B lymphocytes and the naive, GC, and memory B cell subsets were purified from tonsil as described (11). Neoplastic B lymphocytes were isolated from lymph node or peripheral blood by depletion of T cells, NK cells, and macrophages using immunomagnetic beads, as reported (33). The purity of malignant B cells was higher than 99%, as assessed by expression of CD19 and of monotypic Ig light chains. When surface Ig expression on B-CLL cell suspensions was faint to undetectable, their purity was assessed by double staining for CD19 and CD5 (34). In addition, the purity of normal and malignant B cell suspensions was checked by RT-PCR for CD3γ, CD56, and CD68 gene expression, as reported (11), in order to exclude the presence of even minute amounts of contaminant cell types. Only cell suspensions testing negative for CD3γ, CD56, and CD68 mRNA were subjected to further studies.
RNA was extracted from freshly isolated cells using RNeasy Mini Kit (Qiagen GmbH, Hilden, Germany) and subjected to RT-PCR (11). Expression of IL-12Rβ1, IL-12Rβ2, and IFN-γ mRNA was investigated by RT-PCR using the conditions and the primers published elsewhere (16).
Cell culture, antibodies, and reagents. Both purified primary tumor B cells and the human RPMI 8866 and Raji, BRGM, and DAUDI cell lines were cultured in RPMI 1640 medium (Seromed-Biochrom KG, Berlin, Germany) supplemented with 10% FCS (Seromed-Biochrom KG) for the indicated times. The RPMI 8866 is an EBV-infected LCL, whereas Raji, BRGM, and DAUDI are EBV-infected BL cell lines. HrIL-12 was kindly provided by Stan Wolf, Genetics Institute Inc., Andover, Massachusetts, USA, and used in vitro at the concentration of 2 ng/ml, following preliminary titration experiments. In some experiments, cells were cultured for 24 to 96 hours with 5-Aza-2′-deoxycytidine (5 ∝M) (Sigma-Aldrich, St. Louis, Missouri, USA) (18). Cultures were supplemented with this chemical every 24 hours.
CD19, CD3, CD56, CD68, CD5, as well as anti-κ and anti-λ Ig light chain monoclonal antibodies, were purchased from Becton-Dickinson (San Jose, California, USA). Cells were scored using a FACScan analyzer (Becton-Dickinson), and data were processed using CellQuest software (Becton-Dickinson). Immunofluorescence studies with tonsil tissue section (see below) were carried out using FITC-conjugated CD19 mAb (Caltag, Burlingame, California, USA) and PE-conjugated anti–IL-12Rβ2 mAb (Pharmingen, San Diego, California, USA). In all experiments, controls were fluorochrome-conjugated, isotype-matched mAb’s of irrelevant specificity (Caltag).
Methylation analysis. Identification of CpG islands in the genomic segment containing the IL-12Rβ2 locus was conducted with CPGPLOT (http://www.ebi.ac.uk/emboss/cpgplot/) and with the GRAIL/CpG software embedded in the NIX package (http://www.hgmp.mrc.ac.uk/) as described (35). The cut-off criteria utilized to consider a CpG-rich region as an island were: length greater than 500 bp, C + G content greater than 50%, and observed CpG/expected CpG greater than 0.60. The target sequence was retrieved from the GenBank by aligning the IL-12Rβ2 cDNA reference sequence (NM_001559) with the human genome sequence. The analysis was conducted on a fragment of approximately 110 kb between nucleotides 29,324,925 and 29,436,845 of the reference contig NT_032977.6. This sequence includes all 16 exons of the IL-12Rβ2 gene.
The methylation status of the target sequence was assessed by MSP (36). This technique relies upon the chemical conversion of unmethylated (but not of the methylated) cytosine into thymine by sodium bisulfite treatment. Since the methylated C preceding G are protected from this conversion, the methylated and unmethylated sequences will differ only at the CpG doublets. The chemically modified sequences are then amplified by PCR utilizing primers that recognize selectively the methylated or the unmethylated target.
The chemical conversion was carried out as follows: 1 ∝g of DNA was diluted in 50 ∝l of water and denatured in 0.3 M NaOH for 15 minutes at 37°C. Thirty microliters of 10 mM hydroquinone and 520 ∝l of 3 M sodium bisulfite at pH 5, both freshly prepared, were then added to the samples and incubated in mineral oil at 50°C for 16 hours. The DNA samples were purified with Microcon4 microconcentrators (Millipore, Milan, Italy) according to the manufacturer’s instructions. The chemical modification was completed by treatment with 0.3 M NaOH for 15 minutes at 37°C and neutralized by adding CH3COONH4 at the final concentration of 3 M, followed by ethanol precipitation. Modified DNA was resuspended in 30 ∝l of water and stored at –20°C.
Two microliters of modified DNA were utilized for PCR amplification with the following sets of primers: MetSN, ATCGTTTAGTTTCGATTTTCGTTTC; MetASN, GCTCTCCGC GCTCTCTACC; UnMetSN, GTTATTGTTTAGTTTTGATTTTTGTTTT; UnMetASN, CCACACTCTCTACCAACACTC. PCR fragments were separated onto a 3% Metaphor gel (Biowhittaker, Rockland, Maine, USA).
Tissue studies. Frozen tonsil sections were fixed in ice-cold acetone for 10 minutes at room temperature, dried, washed twice with PBS, and incubated for 20 minutes at room temperature with blocking solution (PBS + 10% human serum). Sections were next incubated overnight at 4°C with PE-conjugated rat anti-human IL-12Rβ2 mAb’s, with FITC-conjugated CD19 mAb’s, or with isotype- and fluorochrome-matched mAb’s of irrelevant specificity. For double-staining experiments, sections that had been treated with the anti–IL-12Rβ2 mAb’s were washed three times in PBS and incubated with FITC-conjugated CD19 mAb’s for 1 hour at 4°C. After washing in PBS and drying, sections were mounted with Vectashield Mounting Medium containing DAPI (Vector, Burlingame, California, USA). Digital images of FITC, PE, 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).
Transfection of human B cell lines. The hIL-12Rβ2 cDNA (kindly donated by Lars Rogge, Pasteur Institute, Paris, France) was cloned in the SmaI site of pEGFP-N1 plasmid (Clontech Laboratories Inc., Palo Alto, California). Both RPMI 8866 and Raji cell lines were transfected with 10 ∝g of the IL-12Rβ2/pEGFP-N1 or with the empty vector by electroporation (250 mV, 950 ∝F). Stable transfectants were selected by growing the transfected cells in RPMI 1640 medium (Seromed-Biochrom KG), supplemented with 10% FCS (Seromed-Biochrom KG), containing 750 ∝g/ml G-418 (Calbiochem, La Jolla, California, USA). Efficiency of transfection was monitored by visualization of GFP using a fluorescence microscope (Nikon E1000; Nikon Instruments, Tokyo, Japan). Raji and RPMI 8866 IL-12Rβ2–transfected cells were checked for expression of the IL-12Rβ2 gene by RT-PCR (11). In some experiments, IL-12Rβ2– and empty vector–transfected Raji and RPMI 8866 cell lines were cultured with hrIL-12 for 36 hours before being tested for IFN-γ mRNA expression by RT-PCR (37).
Cell proliferation and apoptosis assays. To test the effects of hrIL-12 on proliferation and growth of IL-12Rβ2–transfected RPMI 8866 and Raji cell lines, 105 cells/0.2 ml were cultured in triplicate for 24–96 hours in the presence or absence of hrIL-12. Wild-type RPMI 8866 and Raji cell lines and empty vector transfectants were used as controls. Every 24 hours, cells were either counted using an electronic counter (Beckman Coulter Inc., Brea, California, USA) or pulsed with 0.5 ∝Ci/well [H3]−thymidine (ICN Biomedicals, Costa Mesa, California, USA) for 18 hours before harvesting. Cell-associated radioactivity was determined by liquid scintillation counting. Apoptosis was assessed by the TUNEL assay (ApopTag Apoptosis Detection Kit; Serologicals Corporation, Norcross, Georgia, USA), following the instructions of the manufacturer.
Mouse studies. Four- to 5-week-old female SCID-NOD mice were purchased from Harlan Laboratories (Udine, Italy) and housed in sterile enclosures under specific pathogen-free conditions (38). All procedures involving mice and their care were in accordance with institutional guidelines in compliance with national and international laws and policies (European Economic Community Council Directive 86/109, OJL 358, December 1, 1987, and NIH Guide for the Care and Use of Laboratory Animals). Two groups of seven animals each were injected intraperitoneally (i.p.) with 107 parental Raji or empty vector–transfected Raji cells, respectively. Two groups of eleven mice each were injected i.p. with 107IL-12Rβ2–transfected Raji cells. One group of mice for each combination was treated with three weekly doses of hrIL-12 (1 ∝g/mouse/dose) starting from 72 hours after injection of tumor cells. The other group of mice from each combination was injected with PBS (controls) according to the same time schedule.
Four weeks after tumor cell inoculation, mice were sacrificed by excess ethyl-ether anesthesia, and autopsies were carried out. Maximum length and width of the tumor mass were measured with a caliper, and tumor volume (mm3) was calculated according to the following formula: 0.523 ∞ length ∞ width2.
Morphologic and immunohistochemical analyses. Tissue samples were processed for histological evaluation or for immunohistochemistry, as previously described (39). The following mAb’s were used: anti-mouse CD31 (clone mEC-13.324; provided by A. Vecchi, Istituto M. Negri, Milan, Italy); anti–hIP-10/CXCL10 (clone 6D4; Abcam, Cambridge, United Kingdom), anti-hMig/CXCL9 (PeproTech, London, United Kingdom), anti–hIFN-γ (clone C-19; Santa Cruz Biotechnology, Santa Cruz, California, USA), anti–hKi-67 (clone MIB-1; Dako, Glostrup, Denmark). Apoptosis was detected using the ApopTag Kit. Quantitative studies of stained sections were performed independently by three pathologists in a blind fashion. Microvessels were counted in 10 randomly chosen fields under a microscope at a total magnification of ∞400 (∞40 objective and ∞10 ocular lens; 0.180 mm2 per field). Expression of IFN-γ, IP-10/CXCL10, and Mig/CXCL9 was scored as absent (−), low (±), moderate (+), or frequent (++). The rates of proliferating cells, as assessed by immunoreactivity for Ki-67, and of apoptotic cells, as assessed by the ApopTag assay, were determined by counting the number of positive cells and number of total cells in the viable neoplastic tissue, excluding areas of tissue necrosis, under a microscope at a total magnification of ∞600 (∞60 objective and ∞10 ocular lens; 0.120 mm2 per field).
Statistical analysis. Differences in cell growth and apoptosis, tumor volume, number of tumor microvessels, Ki-67 positive cells, and apoptotic cells were evaluated by Student’s t test.
This study has been supported by grants from Associazione Italiana per la Ricerca sul Cancro (AIRC) to A. Amadori and V. Pistoia, Istituto Superiore di Sanità, AIDS Project to A. Amadori, Ministero della Saluta, Ricerca Corrente to V. Pistoia, and Ministero dell’Istruzione, Università e Ricerca, 60%, to A. Amadori. The authors thank Walter Habeler for help in animal studies, Lars Rogge for the gift of IL-12Rβ2 cDNA, Mauro Truini for providing pathological samples, Franco Fais for the analysis of IgV region mutations in B-CLL samples, Annunciata Vecchi for the gift of the CD31 monoclonal, Ignazia Prigione for help in cell separation and flow cytometry experiments, Stefano Regis for advice in molecular studies, and Chiara Bernardini for excellent secretarial assistance.
Nonstandard abbreviations used: B chronic lymphocytic leukemia (B-CLL); Burkitt lymphoma (BL); follicular lymphoma (FL); follicular mantle (FM); germinal center (GC); green fluorescent protein (GFP); human IFN-γ (hIFN-γ); human recombinant IL-12 (hrIL-12); lymphoblastoid cell line (LCL); mantle cell lymphoma (MCL); marginal zone lymphoma (MZL); methylation-specific PCR (MSP); phycoerythrin (PE); subepithelial (SE).
Conflict of interest: The authors have declared that no conflict of interest exists.