CD40-Activated B-Cell Chronic Lymphocytic Leukemia Cells for Tumor Immunotherapy: Stimulation of Allogeneic Versus Autologous T Cells Generates Different Types of Effector Cells

After informed consent, peripheral blood samples were obtained from patients with B-CLL. The diagnosis of B-CLL was based on standard clinical and laboratory criteria.19 The study included 12 patients (4 women and 8 men; 49 to 82 years of age). Staging was performed according to the Binet classification.20Characteristics of the patients studied are summarized in Table 1.

The CD40L coding region was amplified as previously described.21 Briefly, RNA was isolated from activated human T cells. Reverse transcription was followed by a two-step polymerase chain reaction (PCR). The first-step PCR was performed using sense primers coding for the first 20 nucleotides of the CD40L coding sequence and antisense primers coding for the last 23 nucleotides of the CD40L coding sequence. In a second step, the amplified PCR products reamplified using extended primers. The sense primer (5′-GTA GGA ATT CGT CGA CGC CGC CAC CAT GAT CGA AAC ATA CAA CC-3′) containsEcoRI and Sal I sites, a strong translational start site, and the first 20 nucleotides of the CD40L coding sequence. The antisense primer (5′-GAC TAG TGT CGA CGA ATT CAG AGT TTG AGT AAG CCA AAG-3′) contains the last 23 nucleotides of the CD40L coding sequence, including the stop codon and EcoRI, Sal I, and Spe I sites. For each step, 20 cycles were performed (95°C for 1 minute, 48°C for 30 seconds, and 72°C for 1 minute), followed by one cycle at 72°C for 10 minutes. The 0.8-kb PCR product was digested with EcoRI, gel-purified, and ligated into EcoRI-digested pcDNA3.1 vector (purchased from Invitrogen, NV Leek, The Netherlands). Plasmids containing the CD40L insert were transfected via electroporation in HeLa/SF cells. Transfectants were selected by growth in 250 μg/mL G418 and further subcloned. Biologic activity was determined by costimulation of B-cell proliferation and differentiation.

Mononuclear cells (MNCs) from peripheral blood samples were isolated by centrifugation on a Ficoll/Hypaque (Seromed, Berlin, Germany) density gradient and depleted from monocytes by overnight adherence to plastic tissue culture flasks. Subsequently, the nonadherent lymphocytes were cryopreserved in liquid nitrogen in the presence of 10% dimethyl sulfoxide (DMSO; Sigma, München, Germany). As assessed by flow cytometry, greater than 95% of these cells typically coexpressed CD19 and CD5 surface molecules.

For CD40L-induced activation, freshly isolated B-CLL cells were cultured as previously described.14 Briefly, CD40L or mock-transfected NIH3T3 fibroblasts or HeLa/SF cells were γ-irradiated at 200 Gy, plated at 5 × 10⁵cells/well in 6-well plates in media without G418, and incubated overnight at 37°C in a 5% CO₂ humidified atmosphere. Before addition of the B-CLL cells, the feeder layers were washed twice with phosphate-buffered saline, and tumor cells were cultured at 2 × 10⁶ cells/mL in Iscove’s medium (Seromed) supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mmol/L L-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin. For further studies, culture was performed in presence or absence of interleukin-2 (IL-2; 20 IU/mL), IL-4 (1 IU/mL), and interferon γ (IFNγ; 20 IU/mL). For performing functional studies, tumor cells were harvested after 24, 72, and 120 hours of culture; purified by ficoll density gradient centrifugation; washed; and analyzed by flow cytometry. For T-cell restimulation, the activated B-CLL cells (CD40-CLL) were aliquoted and stored in liquid nitrogen. The CD40L-transfected NIH3T3 cell line was a generous gift from Dr J. Schultze (Dana-Farber Cancer Insitute, Boston, MA). With respect to their stimulatory capacity, no differences between CD40L-transfected NIH3T3 fibroblasts and CD40L-transfected HeLa/SF cells were detected; therefore, both feeder cell lines are referred to as t-CD40L throughout the manuscript.

Recombinant human IL-2 (rhIL-2), rhIL-4, and rhIFNγ were obtained from Boehringer Mannheim (Mannheim, Germany) and used as indicated. Cytokine measurements were performed using commercial IL-4 and IFNγ enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Wiesbaden-Nordenstadt, Germany) according to the manufacturer’s instructions. The detection limits of the assays were 5 to 2,000 pg/mL.

Immunophenotyping was performed with the following monoclonal antibodies (MoAbs) conjugated with fluorescein isothiocyanate (FITC), phycoerythrin (PE), or phycoerythrin cyanine 5 (PE-Cy5): CD3, CD4, CD5, CD8, CD19, CD20, CD23, CD25, CD28, CD54, CD56, CD58, CD69, CD95, HLA-ABC, HLA-DR, anti-κ, anti-λ (Coulter/Immunotech, Hamburg, Germany), CD40, CD40L, CD80, CD86, and CD95L (PharMingen, Hamburg, Germany). Fluorescence was measured with a Coulter Epics XL-MCL (Coulter Electronics, Miami, FL).

Purification of T cells was performed by negative selection. Briefly, the monocyte-deprived MNCs from healthy donors or B-CLL patients were stained with a cocktail of antibodies (CD11b, CD16, CD19, CD36, and CD56), labeled with goat anti-rat IgG microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany), and selected using magnetic separation columns (Miltenyi Biotec). Isolated cells were greater than 95% pure as determined by immunofluorescent flow cytometry analysis (FACS) and appeared viable by exclusion of trypan blue and forward/side scatter analysis.

Purified T cells from healthy donors or B-CLL patients were stimulated with γ-irradiated (75 Gy) CD40-CLL cells or with native B-CLL cells at different effector to target (E:T) ratios ranging from 10:1 to 5:1. Stimulation was performed on days 0, 7, 14, and 21. Briefly, T cells were harvested weekly, washed, and recultured at a concentration of 1 × 10⁶ cell/mL with γ-irradiated, autologous, or allogeneic stimulator cells in Iscove’s medium (Seromed) supplemented with 5% human heat-inactivated AB-serum (Serva, Heidelberg, Germany), 2 mmol/L L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin. The addition of IL-2 (20 IU/mL) was perfomed 48 hours after (re)stimulation.

Irradiated (75 Gy) B-CLL cells and CD40-CLL cells were used as stimulators, cocultured at 1 × 10⁴ cells/well in a final volume of 200 μL with allogeneic T cells at 1 × 10⁵ cells/well in 96-well round-bottom plates, and incubated for 3 days at 37°C in a 5% CO₂ humidified atmosphere. The culture medium used was Iscove (Seromed) supplemented with 5% heat-inactivated human AB-serum (Serva), 2 mmol/L L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin. All microcultures were performed in triplicate. During the last 12 hours of the 72-hour culture period, cells were pulsed with 0.5 μCi [³H] thymidine (Amersham, Braunschweig, Germany). Cells were harvested onto glass fiber filters and dried, and the [³H] thymidine incorporation was measured by scintillation spectrophotometry in a Wallac Microbeta Plus 1450 scintillation counter (Turku, Finland). The stimulation indexes (SI) were calculated for each individual experiment as follows: SI = cpm_{(T cells + B-CLL cells)}/cpm_{(T cells)}.

T-cell–mediated cytotoxicity was determined using a standard 4-hour [⁵¹Cr] release assay.22 Unstimulated and CD40-stimulated B-CLL cells as well as NK-sensitive K562 cells were used as targets. Target cells were labeled with 100 μCi of [⁵¹Cr] (Na⁵¹CrO₂; Dupont, Bad Homburg, Germany) per 10⁶ cells for 2 hours at 37°C in a water bath. Thereafter, the cells were washed three times in complete medium and seeded in v-bottomed microtiter plates at a concentration of 2.5 × 10³ cells/well. Indicated numbers of effector cells were added in triplicate in 200 μL of complete medium. Supernatants were collected, and the released [⁵¹Cr] was measured in a γ-counter (LKB-Wallac 1282, Uppsala, Sweden). Spontaneous release was determined by incubation of target cells in medium alone, and maximum release was determined by resuspending the wells with 10% Triton X-100. Specific lysis was determined for each individual experiment as follows: specific lysis (%) = [(experimental [⁵¹Cr] release − spontaneous [⁵¹Cr] release)/(maximum [⁵¹Cr] release − spontaneous [⁵¹Cr] release)] × 100.

Differences between experimental groups were analyzed using the χ² test and the Student’s t-test.

B7 costimulatory molecules play a crucial role in the induction of a T-cell–mediated immune response. Lack to receive costimulatory signals after antigen presentation renders T cells anergic or tolerant.23 24 This may be a mechanism by which CLL cells escape the immune response. Therefore, we determined the cell surface expression of MHC, adhesion, and costimulatory molecules by immunophenotyping of freshly isolated B-CLL cells obtained from 12 patients. The results are summarized in Table 2. The expression level of the surface molecules on B-CLL cells was classified according to their mean fluorescence intensity (MFI). All patients tested expressed high (MFI 1.5 to 2.5 logarithm) or even very high (MFI >2.5 logarithm) levels of MHC class I and II molecules. The adhesion molecules ICAM-1 (CD54) were undetectable in 3 of 12 patients, were expressed at intermediate (MFI 0.5 to 1.5 logarithm) in 8 of 12 patients, and were expressed at high levels in 1 of 12 patients. LFA-3 (CD58) was undetectable in 3 of 12 patients and was expressed at low to intermediate levels in 9 of 12 patients. The costimulatory molecule B7-1 (CD80) was not detectable in 8 of 12 patients or showed only low expression (MFI >0.2 to 0.5 logarithm) in 4 of 12 patients. B7-2 (CD86) was not expressed in 2 of 12 patients, whereas 10 of 12 patients expressed it at low to intermediate levels. CD40 was detectable in all B-CLLs (12/12) at low or intermediate levels. Together, these experiments indicated that B-CLL cells showed a markedly reduced expression of costimulatory molecules, especially of B7-1 (CD80).

In the next step, we stimulated freshly isolated B-CLL cells by t-CD40L and mock-transfected feeder cells in the presence of IL-4 (1 IU/mL; see Materials and Methods). Stimulation by t-CD40L induced a cluster formation of B-CLL cells caused by adhesion to the stimulator cells and a hairy-cell like morphology in some of the B-CLL cells (Fig 1A and B). No aggregation or morphologic changes were observed by stimulation with mock controls. May-Grünwald-Giemsa staining of cytospin smears showed that the size of CD40-CLL cells increased due to an expansion of both nuclei and cytoplasm. Moreover, CD40-CLL showed a relatively high degree of vacuolization, corresponding to an activated state of B-lymphoid cells (Fig 1C). No plasmacytoid differentiation was seen.

To optimize the cytokine cocktail used for stimulation of B-CLL cells, B-CLL cells were stimulated by t-CD40L or mock-transfected feeder cells for 3 days in the presence of various cytokines such as IL-2 (20 IU/mL), IL-4 (1 IU/mL), and IFNγ (20 IU/mL). Figure 2A and B shows that maximal expression of B7-1 and B7-2 was induced by t-CD40L when combined with IL-4. IL-4 alone or in combination with mock stimulation only mediated a slight increase of B7-2 expression on the B-CLL cell surface (data not shown). Stimulation by t-CD40L in combination with IL-2 and IFNγ even reduced the expression of B7-1 when compared with stimulation without cytokines. The addition of IFNγ induced only an upregulation of Fas (CD95) but had no other effects (data not shown). Taken together, the combination of t-CD40L and IL-4 seemed optimal for enhancing the expression of important costimulatory molecules on B-CLL cells and was therefore used in all subsequent experiments.

To determine the optimal length of t-CD40L stimulation for full activation of B-CLL cells, time course experiments were performed. B-CLL cells were stimulated with t-CD40L for 24, 72, and 120 hours. As soon as 24 hours after stimulation by t-CD40L, a significant upregulation of ICAM-1 (CD54), LFA-3 (CD58), and Fas (CD95) was detectable (Fig 3B). Expression of costimulatory molecules B7-1 and B7-2 reached its maximum on day 3 of t-CD40L stimulation. Mock stimulation (Fig 3C) in the presence of IL-4 (1 IU/mL) caused a slight increase of the adhesion molecules ICAM-1 and LFA-3, as well as of the costimulatory molecule B7-2. No upregulation was detected for the costimulatory molecules B7-1 and Fas (CD95). Table 3 summarizes the data obtained with leukemic cells from 12 patients after a 3-day t-CD40L stimulation in the presence of IL-4 (1 IU/mL). In all cases, CD40-CLL cells showed an increased expression of B7-1 and B7-2 (Table 3). CD40 expression was further increased in 8 cases. ICAM-1 and LFA-3 were also upregulated in most cases, except if they were already expressed at intermediate levels before t-CD40L stimulation (ICAM-1: patients no. 6, 7, and 10; LFA-3: patients no. 7 through 10). Expression of both MHC class I and II molecules was further increased to high or very high levels in those cases in which the expression was lower before CD40 stimulation.

To exclude that treatment with t-CD40L induced the expansion of normal rather than neoplastic B cells, immunophenotypic analyses of light chain restriction and CD5 expression were performed on CD40-CLL cells. As shown in one representative experiment, CD40-CLL cells showed the phenotype of B-CLL cells as demonstrated by coexpression of CD5 and CD19 and by light chain restriction (Fig4).

We then investigated whether CD40-CLL cells provided a proliferative stimulus to T cells. For this purpose, we used highly (>95%) purified allogeneic T cells and incubated them in the presence or absence of IL-2 (20 IU/mL) with γ-irradiated (75 Gy) native CLL and CD40-CLL cells for 72 hours. CD40-CLL cells were prepared by prestimulation with t-CD40L for 24, 72, and 120 hours. T-cell proliferation was assessed by incorporation of [³H]thymidine added during the last 12 hours of the experiment. The highest stimulatory capacity of the CD40-CLL cells was achieved on day 3 of t-CD40L prestimulation (Fig 5). The experiments shown are representative for 3 different allogeneic T-cell donors and different cases of B-CLL examined.

In another set of experiments, we analyzed the proliferative response of highly purified allogeneic CD4⁺ and CD8⁺ T cells. For this purpose, we used γ-irradiated (75 Gy) native CLL and CD40-CLL cells (day 3) and incubated them for 72 hours with different ratios of highly (>95%) purified CD4⁺ and CD8⁺ T-cell subpopulations as well as unpurified peripheral blood mononuclear cells (PBMCs). T-cell proliferation was assessed by [³H]thymidine incorporation during the last 12 hours of the experiment. A representative experiment is shown in Fig 6. CD40-CLL cells but not native CLL cells induced a significant T-cell proliferation, regardless of whether unpurified PBMCs or CD4⁺ or CD8⁺ T cells were used. Moreover, CD4⁺ T cells showed a significant stronger proliferative response than CD8⁺ T cells. The strongest response was seen with PBMCs; this result is readily explained by the presence of some additional immune effector cells (eg, NK cells) or APCs (eg, dendritic cells) in the unpurified PBMC fraction.

In the next step we determined whether CD40-CLL cells could be used to stimulate T-cell differentiation and expansion. For this purpose, we used highly purified allogeneic and autologous T cells and stimulated them weekly (days 0, 7, 14, and 28) with γ-irradiated (75 Gy) native and CD40-CLL cells at a ratio of 5:1. In both settings, it was possible to expand large numbers of T cells in the presence of CD40-activated B-CLLs and exogenous IL-2 (20 IU/mL). In contrast, the expansion was not possible with native B-CLL cells, even in the presence of IL-2.

When T cells were monitored by flow cytometric analysis, we found that only CD40-activated B-CLLs were able to induce activation markers (CD25 and CD95). A marked difference was seen with regard to the CD4/CD8 ratio. In the allogeneic setting, consecutive stimulations with CD40-CLL cells caused a relative and absolute increase of CD8⁺ T cells (up to 50% after 3 restimulations) in 4 of 5 cases investigated. In marked contrast, repetitive stimulations in the autologous setting caused an expansion of CD4⁺ T cells (up to 90% after 3 restimulations) in 6 of 6 cases studied. No increase of CD16⁺/CD56⁺ NK cells was observed.

To further characterize the different effector functions of allogeneic versus autologous T cells, we performed 4-hour standard chromium release tests (see Materials and Methods). Native B-CLL cells, CD40-CLL cells, and NK-sensitive K562 cells were used as targets. These different effector cells were expanded until day 28 and then restimulated with CD40-CLL cells and native CLL cells in the presence of IL-2. A significant, cytolytic activity of T cells was only seen if allogeneic T cells from healthy donors were used (Fig 7B). At this time point, allogeneic T cells lysed both native B-CLL cells and CD40-CLL cells. No response against K562 cells was detected. Furthermore, no T-cell response was observed with T cells before stimulation with CD40-CLL (Fig 7A). In marked contrast, autologous T cells expanded by repetitive stimulation with CD40-CLL cells showed only a slight lytic activity against CD40-CLL cells and no lytic activity against native B-CLL cells or the NK-sensitive K562 cell line (Fig 7D). Taken together, the stimulation of allogeneic versus autologous T cells by CD40-CLL cells induced a fundamentally different response in that CD8⁺ cells with cytolytic activity could only be expanded in the allogeneic system.

In the next step we tried to characterize the autologous, CD4⁺ T cells with respect to their cytokine release pattern. For this purpose, we restimulated 1 × 10⁵autologous T cells with 2 × 10⁴paraformaldehyde-fixed (1%) CD40-CLL cells. The supernatant was collected 48 hours later and analyzed for IL-4 and IFNγ. Figure 8 summarizes the data of 6 patients. In 4 of 6 patients tested, we detected predominantly IFNγ, suggesting a Th1-like immune response.

This report investigates the potential of CD40L-stimulated CLL (CD40-CLL) cells to activate effector T cells for tumor vaccination. The essential finding of this report is that the quality of the autologous T-cell response against CD40-CLL cells differs dramatically from the allogeneic T-cell response against these cells. So far, it was known that CD40-ligation could be used in B-CLL cells to upregulate adhesion and costimulatory molecules.25-27 Our study confirmed these findings by showing that stimulation of B-CLL cells by t-CD40L resulted in a significant upregulation of both adhesion and costimulatory molecules in B-CLL cells in a time-dependent manner, regardless of previous treatment with cytostatic drugs. CD40-CLL cells were able to stimulate both the allogeneic and the autologous T-cell proliferation. The addition of IL-4 further enhanced the expression of both B7-1 and B7-2, whereas IFNγ or IL-2 reduced it compared with t-CD40L alone (Fig 2).

A high expression of B7-1 is critical for the induction of a CTL response.28-30 To induce an effective immune response against B-CLL cells, consecutive T-cell stimulations were performed with CD40-CLL cells. Allogeneic CD40-CLL cells strongly stimulated both CD4⁺ as well as CD8⁺ T cells, as previously described.27 In marked contrast, stimulation of autologous CD40-CLL cells strongly favored the outgrowth of CD4⁺, but not CD8⁺ T cells. With respect to the effector T-cell function, there were also marked differences. A cytolytic activity was only induced with allogeneic T cells stimulated by CD40-CLL cells. This resulted in a significant alloantigen-specific cytotoxic activity against both CD40-CLL and naive B-CLL cells, similar to previous findings.27 In marked contrast, the stimulation of autologous T cells by CD40-CLL cells induced the expansion of a predominantly CD4⁺, Th1-like effector cell population without cytolytic activity. Moreover and in contrast to findings in follicular lymphoma and acute lymphoblastic leukemia, CD40-CLL cells did not allow to expand autologous T cells with cytolytic activity.31 32 

The inability of CD40-CLL cells to stimulate CD8⁺ T cells in the autologous system can be explained by several alternative mechanisms. First, CD40 expressed on B-CLL cells may costimulate CD4⁺ T cells rather than CD8⁺ T cells.33 Second, autologous peripheral blood T lymphocytes might be less efficient in mounting a cytolytic response against lymphoma cell antigens than T cells derived from the bone marrow or tumor-infiltrating lymphocytes.32 Third, most if not all the anti-idiotype immune responses reported to date have been by CD4⁺ cells (both Th1- and Th2-type).34-36Fourth, a distinct pattern of costimulatory molecules expressed on CD40-CLL cells may induce the preferential stimulation of Th1-like, CD4⁺ T cells; for example, it is known that CD8⁺ T cells require higher densities of B7-1 to attain an equivalent level of activation as CD4⁺ T cells, and CD40-CLL cells may just not express enough B7-1 and/or B7-2 to activate CD8⁺ T cells.28 Sixth, the cytokine secretion by stimulated T cells themselves may contribute to the modulation of costimulatory molecules on CD40-CLL cells. These different mechanisms act probably in concert in regulating the threshold by which an ongoing T-cell response is maintained.

However, the exact factors preventing the outgrowth of autologous CD8⁺ effector T cells with antileukemic activity remain to be elucidated. It seems highly unlikely that autologous, peripheral blood T lymphocytes are intrinsically incapable of mounting a CTL response in B-CLL patients, because the generation of tumor cell-specific CTLs from the peripheral blood of B-CLL patients has been recently demonstrated.37 

It will be of interest to learn whether the CD4⁺ Th1 cells stimulated by CD40-CLL cells are able to induce B-CLL cell death, eg, by triggering a Fas-dependent apoptotic pathway, as shown in human tonsillar and Burkitt’s lymphoma B cells.38-40 Preliminary experiments in our laboratory showed that CLL cells were indeed rapidly eliminated in coculture with these CD4⁺ T cells (R. Buhmann, unpublished data).

The CD40-CD40L interaction is crucial for the immune response. The ability to trigger and to regulate the induction and expression of CD40L on CD4⁺ cells appears to be of paramount importance to the generation of the T-cell–dependent antigen response.41,42 An excess of CD40-expressing leukemia cells might interfere with these cognate T-cell responses and provoke an aquired CD40L deficiency syndrome in B-CLL.43 Previous studies and our data suggest that these defects can be restored at least in part by an effective T-cell activation, with lymphoma or leukemia cells expressing costimulatory molecules at sufficient density.6,7,31 Activation of naive T cells can be brought by any APC, as long as sufficient levels of one or more accessory molecules are expressed. Later, during the process of T-cell activation and differentiation, the requirements for full T-cell stimulation decrease, as T cells become more responsive to a particular antigen.42,44 45 Thus, an APC that is only weakly stimulatory for naive T cells can become an efficient stimulator for preactivated T cells. Accordingly, native B-CLL cells might have properties of weakly stimulating APCs that need the concomitant presence of strong APCs (such as CD40-CLL cells or dendritic cells) to fully activate T cells.

Taken together, our results suggest that CD40 activation of B-CLL cells may reverse T-cell anergy against the neoplastic clone. The results might provide new perspectives for the immunotherapy of B-CLL, similar to previous observations in follicular lymphoma7,31 and pre-B acute lymphoblastic leukemia (ALL).6Most importantly, the clinical application to produce tumor vaccines or to generate stimulator cells for adoptive immune transfer strategies in B-CLL by this approach will meet less practical limitations than in other lymphoid malignancies, because B-CLL cells are readily obtained from peripheral blood. Moreover, the identity of CD40-CLL cells can be rapidly tested by measuring κ or λ light chain restriction and CD5/CD19 coexpression on the tumor cell surface by flow cytometry. This will facilitate the practical implementation of these approaches in the adjuvant therapy of B-CLL.

The authors are indebted to many individuals for their help in the preparation of this report: Dr Joachim Schultze, Dr Gabriella Pichert, Doris Schmitter, and Dr John Gribben for their instruction in the initial stage of the study; Dr Susanne Danhauser-Riedl for her unconditional assistance in flow cytometry; Dr Heidi Feldmann and Bettina Meier for their excellent expertise in computer aided design of the figures; and, finally, Karin Schulz for her support in molecular biology.