This study was approved by the institutional review board of the Aichi Cancer Center. Mononuclear cells from pathologic lymph node or tissue specimens (ten cases), and from peripheral blood (two healthy donors), were obtained (Table 1) after written informed consent in accordance with the Declaration of Helsinki. Mononuclear cells were isolated by the addition of Ficoll and density centrifugation. All patients were diagnosed on the basis of morphological and immunohistochemical findings in accordance with WHO criteria. Categorization as germinal center B cell-like (GCB) or non-germinal center B cell-like (non-GCB) was performed according to the algorithm using the Hans classification [19Hans CP, Weisenburger DD, Greiner TC, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray Blood 2004; 103: 275-82.].

Table 1

Patients’ Characteristics at Diagnosis

The cells were cultured with RPMI 1640 medium containing antibiotics, 10% fatal calf serum (PAA Laboratories GmbH, Austria), 100 nM sodium selenite (Wako Pure Chemical Industries, Osaka, JP), 1 µg/ml cyclosporin A (CsA; Wako Pure Chemical Industries, Osaka, JP) and 2 ng/ml recombinant human IL-4 (PeproTech Inc, Rocky Hill, NJ). NIH3T3 (NIH Swiss mouse embryonic fibroblast cell line) transfected with CD40L (CD40L⁺3T3), which was kindly supplied by Dr. Lee M. Nadler (Dana-Farber Cancer Institute, Boston, MA, USA) [20Schultze JL, Cardoso AA, Freeman GJ, et al. Follicular lymphomas can be induced to present alloantigen efficiently a conceptual model to improve their tumor immunogenicity Proc Natl Acad Sci USA 1995; 92: 8200-4.], was used for stimulating feeder cells and co-cultured with B cells.

CD40L⁺3T3 was cultured for one to three days at 1×10⁵cells/well using a 12-well plate and 2×10⁴ cells/well using a 96-well plate, and then irradiated (100 Gy) on the day B cells were plated. Since one EBV-infected cell is found in about 10⁴-10⁶ B cells, unseparated mononuclear cells from peripheral blood were plated in different numbers per microculture to established EBV-free CD40-stimulated B cell cultures [16Wiesner M, Zentz C, Mayr C, et al. Conditional immortalization of human B cells by CD40 ligation PLoS One 2008 Jan 23; 23: e1464 [about 13 screens]. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2180193/ [[cited 2008 January 23]];]. The cells were cultured on CD40L⁺3T3 in the presence of interleukin-4 and CsA in a humidified atmosphere at 37°C with 5% CO₂ in air for three to seven days [16Wiesner M, Zentz C, Mayr C, et al. Conditional immortalization of human B cells by CD40 ligation PLoS One 2008 Jan 23; 23: e1464 [about 13 screens]. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2180193/ [[cited 2008 January 23]];]. The cell number was counted using the trypan blue dye exclusion test in each well, and the cells were transferred to new stimulating feeder cells at the same conditions and with the same concentration of B cells.

Cell surface antigen expression of B-cell marker was studied by flow cytometry using FITC, PE, PerCP or APC labeled CD5 (Becton, Dickinson and Company; BD, Mountain View, CA, USA), CD10 (BD), CD19 (BD), CD20 (BD), and surface immunoglobulin (sIg) (BD). The CD40L expression of feeder cells and established cell lines was evaluated with anti-CD40L antibody (monoclonal antibody CD154, murine IgG₁, TRAP-1 clone; Immunotech, a Beckman Coulter Company, France) using FITC-labeled sIg (goat, DakoCytomation, Glostrup, Denmark) as a secondary antibody. The events acquired by the cytometer varied from sample to sample, because the events depend on the size of the culture system.

CD40L⁺3T3 cells were plated at 7×10⁴/well using a 12-well plate (Transwell, Costar) on day -1. The medium was changed to 1.5 ml of fresh medium on day 0, and the bottom cup with a 0.4-ìm diameter microporous membrane was put in place. B cells were inoculated into the inner cup with 0.5 ml of cell suspensions containing 0.5×10⁵ cells/ml. The positive control consisted of three conditions: culture with the inner cup without feeder cells, culture without the inner cup with CD40L⁺3T3 cells, and culture without the inner cup with NIH3T3 untransfected with CD40L (CD40L^-3T3) cells. B cells, cultured without the inner cup, were plated with 2 ml of cell suspensions containing 0.5×10⁵ cells/ml using a 12-well plate. B-cell number was counted every four days using the trypan blue dye exclusion test in each well, and the cells were transferred to new plates at the same conditions and with the same concentration of B cells. The culture was continued until day 8. Each condition was repeated in duplicate, and the cell number was counted two times in each well.

DNA was extracted from frozen tissue samples using proteinase K treatment and phenol-chloroform extraction. Oligo Array CGH analysis was performed using the Agilent Human Genomic Microarray Kit 4×44K (Agilent Technologies). This high-resolution 60-mer oligonucleotide-based microarray was used in this study and contains about 44,000 probes with an average spatial resolution of 35 kb (Agilent Technologies, Santa Clara, CA). Genomic DNA from the patient and normal male reference DNA were digested with AluI and RsaI (Promega Corporation) and labeled using an Agilent Genomic DNA Labeling kit according to manufacturer’s instructions. Patient and reference DNA were labeled with Cy3 and Cy5, respectively. Labeled patient DNA and reference DNA were co-hybridized to arrays for 24 hr at 65°C in a rotating oven (Agilent Technologies) at 20 rpm. Washing was performed according to the Agilent protocol. Arrays were analyzed using an Agilent Microarray Scanner and Agilent Feature Extraction software (9.1). Results were presented by Agilent CGH Analytics software (v3.4.27).

Peripheral blood mononuclear cells from two healthy donors were cultured on the stimulator cells of CD40L⁺3T3 with IL-4 containing medium. The cells were plated with 2 ml of cell suspensions containing 0.5×10⁶/ml using a 12-well plate, and with 200 ìl of cell suspensions containing 1.25×10⁵/ml, 2.5×10⁵/ml, 5.0×10⁵/ml or 10.0×10⁵/ml using a 96-well plate. Since the best growth curve was obtained using cell suspensions containing an initial concentration of 1.25×10⁵/ml with a 96-well plate (data not shown), we chose this condition for further experiments. For the first several days, the cells formed a small cluster on CD40L⁺3T3 cells, and after a further five to seven days, most cells became large clusters. The cells, adjusted to a concentration of 1.25×10⁵/ml, were restimulated every 5 to 7 days with fresh stimulator cells. The cells proliferated to about 110 to 120-fold until day 20. Wiesner et al. [16Wiesner M, Zentz C, Mayr C, et al. Conditional immortalization of human B cells by CD40 ligation PLoS One 2008 Jan 23; 23: e1464 [about 13 screens]. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2180193/ [[cited 2008 January 23]];] showed that cultures stimulated with CD40L-expressing murine fibroblasts in the presence of IL-4 can proliferate for more than 900 days. However, the use of our culture system with CD40L⁺3T3 cells and IL-4 resulted in proliferation for about three weeks, and we observed that cell growth through proliferation rapidly ceased (Fig. 1).

Fig. (1) IL4 and CD40L co-stimulation induced short-term proliferation of B cells in vitro from healthy volunteers. CD40L+3T3 stimulating cells and IL-4 induced normal B-cell proliferation for about three weeks. Cells proliferated to about 110 to 120-fold until day 20 and then cell growth ceased on day 26. Abbreviation: IL-4, interleukin-4; CD40L, CD40 ligand; CD40L+3T3, NIH Swiss mouse embryonic fibroblast cell line transfected with CD40L.

Mononuclear cells from 10 DLBCL samples (five stomach, three lymph node and two breast) were plated with 200 ìl of cell suspensions containing 1.25×10⁵/ml on a 96-well plate using the stimulator cells of CD40L⁺3T3 with IL-4 containing medium. Of the ten samples, tumor cells from six patients stopped growing within three weeks. The cells from the remaining four patients have been cultured with continuous proliferation for more than two months, and cell lines were established from two of these cell groups (GDLB-GCB1 and GDLB-GCB2). The cell lines, derived from gastric tissue, showed proliferation for more than three month. The doubling time of the cell lines GDLB-GCB1 (from UPN (unique patient number)1 patient) and GDLB-GCB2 (from UPN2 patient) was 2.9 and 1.7 days, respectively (Fig. 2a, b). These cell lines showed pathological characteristics of bcl2- and CD10+ germinal center B-cell (GCB)-like DLBCL. Phenotypical analysis of the original samples and established cell lines demonstrated the positive expression of CD10, CD19, CD20 and sIgG, and the negative or weakly positive expression of CD40L (Fig. 3). UPN 6 and UPN 10 ceased to proliferate within 2.5 to 3 months.

Fig. (2) IL4 and CD40 co-stimulation induced long-term proliferation of DLBCL B cells in vitro from DLBCL patients. Of the ten DLBCL cases analyzed, tumor cells from four patients showed continuous proliferation for more than two months, and two cell lines (GDLB-GCB1 and GDLB-GCB2) were established from these cells (a, b). These cell lines could proliferate for more than three months and could be restored after cryopreservation. The doubling time of the established cell lines of GDLB-GCB1 (from the UPN1 patient) and GDLB-GCB2 (from the UPN2 patient) was 2.9 and 1.7 days, respectively. Abbreviation: IL-4, interleukin-4; CD40L, CD40 ligand; DLBCL, diffuse large B cell lymphoma, UPN, unique patient number.

Fig. (3) Flow cytometric analyses of the established cell lines (GDLB-GCB1 and GDLB-GCB2). The original tumor samples and the established cell lines (GDLB-GCB1 and GDLB-GCB2) demonstrated the positive expression of CD10, CD19, CD20 and surface immunoglobulin and the negative expression of CD5 (a, b). Surface CD40L expression was the negative expression in GDLB-GCB 1 and weakly positive in GDLB-GCB 2 (c). CD40L+3T3 was used for a positive control of CD40L (c). Abbreviation: CD40L, CD40 ligand; CD40L+3T3, NIH Swiss mouse embryonic fibroblast cell line transfected with CD40L.

We next examined whether the growth of two established cell lines depended on IL-4 and CD40L. The method of CD40L blocking was conducted as our previous experiments that were used for the other antibodies (Kagami et al.) [21Kagami Y, Tobinai K, Kinoshita T, et al. Novel interleukin-2 dependent T-cell line derived from adult T-cell leukemia not associated with human T-cell leukemia virus type I Jpn J Cancer Res 1993; 84: 371-8.]. Removal of IL-4 from the system containing the stimulator cells of CD40L⁺3T3 with IL-4 containing medium resulted in growth inhibition. Proliferation of B cells from the cell lines was also inhibited in medium only or medium containing IL-4 (Fig. 4a, b). When anti-CD40L blocking antibody was added to this culture system, the proliferation of these cell lines was inhibited (Fig. 4a, b,). Proliferation of the cells from GDLB-GCB2 was inhibited in a concentration-dependent manner (Fig.b).

Fig. (4) Growth curves of the established cell lines (GDLBGCB1 and GDLB-GCB2) in the presence or absence of IL-4 and CD40L stimulation. Removal of IL-4 (●) from the system consisting of CD40L+3T3 and IL-4 containing medium (♦) resulted in cell growth arrest. Proliferation of tumor cells also decreased in medium only (Ɨ) or medium containing IL-4 (Ж). Addition of anti-CD40 blocking antibody to this culture system inhibited the proliferation of these cell lines (▲)(Δ), and proliferation of the cells from GDLB-GCB2 was inhibited in a concentration-dependent manner. Each condition was repeated in duplicate, and the cell number was counted two times in each well on days 3, 6 and 9. One case (GDLB-GCB2) was counted until day 12. Abbreviation: IL-4, interleukin-4; CD40L, CD40 ligand; CD40L+3T3, NIH Swiss mouse embryonic fibroblast cell line transfected with CD40L.

The use of an in vitro transwell assay, which prevents direct cell-to-cell contact, allows confirmation of whether direct contact with CD40L⁺3T3 cells is required for proliferation of the B-cell lines. The cells inoculated into the inner cup with the transwell assay showed cell growth arrest. Cells of GDLB-GCB2 also stopped growing when the cells were cultured with CD40L^-3T3 (Fig. 5b). On the other hand, cells of GDLB-GCB1 stimulated with CD40L^-3T3 showed cell growth on day 8 (Fig. 5a). When these cell lines were cultured with CD40L^-3T3 cells in a 96-well plate, the two cell lines stopped growing as expected.

Fig. (5) Growth curves of the established cell lines (GDLBGCB1 and GDLB-GCB2) without direct contact with feeder cells. When tumor cells and CD40L+3T3 cells were separated with a micropore membrane, the proliferation was inhibited (ж). The positive control consisted of three conditions: culture with the inner cup without feeder cells (Δ), culture without the inner cup with CD40L+3T3 cells (♦), and culture without the inner cup with CD40L-3T3 cells (▲). The cells of GDLB-GCB1 stimulated with CD40L-3T3 showed cell growth on day 8 (Fig. 5a). However, when the cell lines were cultured with CD40L-3T3 cells in a 96-well plate, the both cell lines stopped growing as expected. Each condition was repeated in duplicate, and the cell number was counted two times in each well on days 4 and 8. Data are presented as the mean ± SE. Abbreviation: CD40L, CD40 ligand; CD40L+3T3, NIH Swiss mouse embryonic fibroblast cell line transfected with CD40L; CD40L-3T3, NIH Swiss mouse embryonic fibroblast cell line untransfected with CD40L.

Array CGH analysis revealed that the established cell lines showed the identical genomic imbalances as the original samples, indicating that the established cell lines and original samples were derived from the identical clones (Fig. 6). However, the established cell lines did not have common distinctive genomic imbalances. Array CGH data of the short-term proliferating cells could not be obtained because of the cell death.

Fig. (6) Genomic imbalances using 4x44K oligo array CGH methods. Array CGH analysis revealed that the established cell lines (GDLB-GCB1 and GDLB-GCB2) showed the identical genomic imbalances as the original samples (UPN1 and 2), indicating that the established cell lines and original samples were derived from the identical clones. Abbreviation: CGH, comparative genomic hybridization; UPN, unique patient number.

In this report, we demonstrated that IL-4 and CD40L co-stimulation induced short-term proliferation in DLBCL B cells, and some type of DLBCL B cells showed a long-term proliferation. Interestingly, the three cases including the established two cell lines had common clinicopathological features, presenting with a local extra nodal lesion and a bcl2- CD10+ GCB-like DLBCL phenotype. Moreover, removing IL-4 from the culture system or adding anti-CD40L blocking antibody to the system inhibited proliferation of the cell lines. These results suggested that CD40L and IL-4 co-stimulation played a critical role in long-term proliferation for some type of DLBCL in vitro. However, it is possible that more cell lines could have been established under different conditions. It is also conceivable that tumor cell purity influenced the results.

The biological behavior, tumor progression and prognosis of patients may be influenced by the tumor microenvironment for some types of malignant lymphomas [22Kagami Y, Jung J, Choi YS, et al. Establishment of a follicular lymphoma cell line (FLK-1) dependent on follicular dendritic cell-like cell line HK Leukemia 2001; 15: 148-56.-24Monti S, Savage KJ, Kutok JL, et al. Molecular profiling of diffuse large B-cell lymphoma identifies robust subtypes including one characterized by host inflammatory response Blood 2005; 105: 1851-61.]. CD40L is expressed mainly in activated CD4+ T lymphocytes. CD40L binds to CD40 of normal B cells, and induces growth or survival signals of B cells [15Banchereau J, de Paoli P, Vallé A, et al. Long-term human B cell lines dependent on interleukin-4 and antibody to CD40 Science 1991; 251: 70-2., 16Wiesner M, Zentz C, Mayr C, et al. Conditional immortalization of human B cells by CD40 ligation PLoS One 2008 Jan 23; 23: e1464 [about 13 screens]. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2180193/ [[cited 2008 January 23]];]. CD40L expression has also been shown in several types of B-cell lymphomas [11Younes A, Kadin ME. Emerging applications of the tumor necrosis factor family of ligands and receptors in cancer therapy J Clin Oncol 2003; 21: 3526-4., 12Pham LV, Tamayo AT, Yoshimura LC, et al. A CD40 signalosome anchored in lipid rafts leads to constitutive activation of NF-kappaB and autonomous cell growth in B cell lymphomas Immunity 2002; 16: 37-50.]. Pham et al. reported that endogenously expressed CD40L in aggressive B-cell lymphomas binds to the CD40 receptor within cell membrane lipid rafts, and that this CD40 signalosome constitutively activates the canonical NF-êB signaling pathway [12Pham LV, Tamayo AT, Yoshimura LC, et al. A CD40 signalosome anchored in lipid rafts leads to constitutive activation of NF-kappaB and autonomous cell growth in B cell lymphomas Immunity 2002; 16: 37-50.]. In the present study, endogenous CD40L expression was negative in GDLB-GCB1, and weakly positive in GDLB-GCB2. Removing CD40L⁺3T3 feeder cells from the CD40L and IL-4 co-stimulation system or adding anti-CD40L blocking antibody to the system inhibited proliferation of these cell lines. An in vitro transwell assay also showed cell growth arrest. These

experiments suggested that direct CD40L stimulation from the tumor microenvironment, but not endogenously expressed CD40L, is important for the proliferation of some type of DLBCL.

IL-4 is a pleiotropic cytokine regulating differentiation and growth of normal B cells. IL-4 is known to be a potent survival factor for many cell types, and the effect on neoplastic B cells is different for each lymphoma type [25Paul WE, et al. Interleukin-4 a prototypic immunoregulatory lymphokine Blood 1991; 77: 1859-70.-28Kadoch C, Dinca EB, Voicu R, et al. Pathologic correlates of primary central nervous system lymphoma defined in an orthotopic xenograft model Clin Cancer Res 2009; 15: 1989-97.]. Lu et al. [29Lu X, Nechushtan H, Ding F, et al. Distinct IL-4-induced gene expression, proliferation, and intracellular signaling in germinal center B-cell-like and activated B-cell-like diffuse large-cell lymphomas Blood 2005; 105: 2924-32.] reported that in GCB-like DLBCL, IL-4 induces expression of its known target genes by activation of STAT6, resulting in cell proliferation. In contrast, in ABC-like DLBCL, IL-4 activates the Akt pathway and results in a decrease of cell proliferation by cell-cycle arrest [29Lu X, Nechushtan H, Ding F, et al. Distinct IL-4-induced gene expression, proliferation, and intracellular signaling in germinal center B-cell-like and activated B-cell-like diffuse large-cell lymphomas Blood 2005; 105: 2924-32.]. In our study, IL-4 was also an important factor for promoting proliferation of the established cell lines; however, the stimulation of IL-4 in the absence of CD40L did not show any significant cell growth. For some GCB-like DLBCL patients, CD40L and IL-4 co-stimulation might be associated with lymphoma cell proliferation.