Immunostimulatory CpG-oligonucleotidesinduce functional high affinity IL-2 receptors on B-CLL cells:
Costimulation with IL-2 results in a highly immunogenic phenotype
Thomas Decker
a
, Folker Schneller
a
, Manuela Kronschnabl
a
, Tobias Dechow
a
, Grayson B. Lipford
b
, Hermann Wagner
b
, and Christian Peschel
a
a
IIIrd Department of Medicine and
b
Institute of Medical Microbiology and Hygiene, Technical University of Munich, Munich, Germany
(Received 19 November 1999; revised 13 January 2000; accepted 21 January 2000)
Objective.
CpG-oligodeoxynucleotides (CpG-ODN) have been shown to induce proliferation,cytokine production, and surface molecule regulation in normal and malignant human B cells.In the present study, we investigated the potential of CpG-ODN to induce functional high-affinity receptors in leukemic and normal B cells and the effects of costimulation with IL-2 onproliferation, cytokine secretion, and surface molecule regulation.
Methods.
Highly purified B cells from B-CLL patients and normal controls were stimulatedwith CpG-ODN with or without IL-2. Expression of CD25 was determined using FACS, andthe presence of high-affinity IL-2 receptors was determined by scatchard analysis. Costimula-tory effects of IL-2 and CpG-ODN were investigated using proliferation assays, ELISA (IL-6,TNF-
a
), and FACS analysis (CD80, CD86 expression). Reactivity of autologous and allogeneicT cells toward activated B-CLL cells was determined in mixed lymphocyte reactions and Inter-feron-
g
Elispot assays.
Results.
The CpG-ODN DSP30 caused a significantly stronger induction of the IL-2 receptor
a
chain in malignant as compared with normal B cells (
p
5
0.03). This resulted in the expres-sion of functional high-affinity IL-2 receptors in B-CLL cells, but fewer numbers of receptorswith less affinity were expressed in normal B cells. Although addition of IL-2 to CpG-ODN–stimulated cells augmented proliferation in both normal B cells and B-CLL cells, no costimula-tory effect on cytokine production or surface molecule expression could be observed in normalB cells. In contrast, TNF-
a
and IL-6 production was increased in B-CLL cells, and the expres-sion of CD80 and CD86 was further enhanced when IL-2 was used as a costimulus. Autologousand allogeneic immune recognition of B-CLL cells stimulated with CpG-ODN and IL-2 was in-creased compared with B-CLL cells stimulated with CpG-ODN alone.
B-chronic lymphocytic leukemia (B-CLL) is defined as theaccumulation of slow-dividing and long-lived B lympho-cytes closely resembling small resting B lymphocytes [1].The leukemic population exhibits marked hyporesponsive-ness to proliferative signals, which activate normal B cells,including ligation of the CD40 antigen and crosslinking of
the B-cell receptor complex [2,3]. A considerable numberof B-CLL cells express the CD25 molecule and synthesizeDNA in response to IL-2 [4]. The reactivity of lymphocytesto IL-2 is mediated by multimeric high-affinity receptors(Kd
5
10 pM) consisting of two obligate signaling subunits,IL-2R
b
(CD122) and
g
(CD132), and a variably expressedIL-2R
a
(CD25) that regulates affinity for IL-2 [5]. A low num-ber of high-affinity receptors has been described for thoseB-CLL cells that express the CD25 antigen and spontane-ously react to IL-2 [6]. Anti-IgM mAb, phorbol ester, andSAC increase the expression of high-affinity IL-2 receptors,
Offprint requests to: Christian Peschel, M.D., IIIrd Department of Medi-cine, Technical University of Munich, Ismaninger Str. 15, 81675 Munich,Germany; E-mail: [email protected]
T. Decker et al./Experimental Hematology 28 (2000) 558–568
559
and the addition of IL-2 enhances proliferation and immu-noglobulin production [7,8]. CD40 ligation slightly in-creases CD25 expression but fails to induce IL-2 high-affin-ity receptors. However, addition of IL-10 to B cellsstimulated by CD40 ligation strongly upregulates CD25 ex-pression and results in the expression of functional high-affinity IL-2 receptors [6]. Liang et al. [9] reported stronginduction of CD25 expression on human B cells afterstimulation with phosphorothioate oligodeoxynucleotides.Oligonucleotides containing a CpG motif were the most ef-fective [9]. The immunostimulatory properties of bacterialDNA and mimicking synthetic oligonucleotide sequenceshave only recently been recognized. Structural and se-quence usage differences between bacterial and eukaryoticDNA allow immune cells to sense unmethylated CpG aspathogenic in origin (for review see [10]). DNA motifs dis-playing a 5
9
-Pu-Pu-CpG-Pyr-Pyr-3
9
were found to be bio-logically most active [11]. Cell activation requires endoso-mal uptake of CpG-ODN that within minutes results inactivation of the stress kinase pathways [12,13]. As a conse-quence, murine B cells enter the cell cycle, secrete immuno-globulin in vitro and in vivo [11] and are rescued fromspontaneous or Fas-mediated apoptosis [14,15]. Human Bcells proliferate, produce immunoglobulin, and upregulatesurface expression of CD25 and costimulatory molecules onstimulation with active CpG-ODN [9,16]. Recently we havedescribed induction of proliferation, cytokine production,and upregulation of surface molecules, including CD40,CD80, and CD86, in B-CLL cells after activation withCpG-ODN. ODN lacking the CpG motif were less effec-tive. The upregulation of CD40 resulted in a strong costimu-lation in culture systems with CD40 ligand transfected(CD40LF) cells regarding proliferation and cytokine pro-duction, while increased CD80 and CD86 expressioncaused a strong stimulatory capacity toward allogeneic Tcells [17].
In the present study, we investigated the potential ofCpG-ODN to induce functional high-affinity receptors inleukemic and normal B cells and the effects of costimula-tion with IL-2 on proliferation, cytokine secretion, and sur-face molecule regulation in B-CLL cells and normal B cells.Because B-CLL cells with upregulated costimulatory mole-cules have been described to induce allogeneic and autolo-gous immune recognition [18,19], the ability of B-CLLcells stimulated with CpG-ODN and IL-2 to activate autolo-gous and allogeneic T cells was also examined.
Materials and methods
Cell samples
After informed consent, peripheral blood was obtained from 11 pa-tients with a diagnosis of B-CLL according to clinical and immu-nophenotypic criteria. Patients were either untreated or had not re-ceived cytoreductive chemotherapy for a period of at least 3months before investigation. At the time of analysis, all patients
were clinically stable, free from infectious complications, and un-dergoing routine clinical outpatient review. For control experi-ments, samples were collected from healthy volunteers or patientswho were undergoing routine tonsillectomy.
Reagents, antibodies, and cell lines
All ODN were used single stranded, phosphorothioate stabilized,and synthesized by TibMolBiol (Berlin, Germany). The followingODN were used: DSP30 (TCGTCGCTGTCTCCGCTTCTTCT-TGCC) [9] and pZ2 (CTCCTAGTGGGGGTGTCCTAT) [13]. AllODN were used at 1
m
M. PE-conjugated anti-CD86 mAb and theappropriate isotype-control were purchased from Pharmingen (SanDiego, CA, USA), FITC-conjugated anti-CD25 mAB, anti-CD80mAb, and appropriate isotype-controls were purchased fromCoulter-Immunotech (Hialeah, FL, USA). IL-2 was obtained fromPepro Tech (London, United Kingdom), and anti-TNF-
a
mABwas obtained from R&D Systems (Wiebaden, Germany).
Separation procedures
Peripheral blood mononuclear cells (PBMNC) were isolated fromheparinized blood samples by centrifugation over a Ficoll-Hypaquelayer (Biochrom, Berlin, Germany) of 1.077 g/mL density. Forseparation of CLL B-cells, PBMNC were incubated with anti-CD2and anti-CD14 magnetic beads (Dynabeads M450; Dynal, Oslo,Norway) according to the manufacturer’s instructions. Such pre-pared B cells from CLL patients were
.
98% pure as assessed bydirect immunofluorescence using a Coulter Epics XL (Coulter,Hamburg, Germany). Nonmalignant B cells did not constitute ameaningful fraction of the total cells isolated because 99% of thesecells coexpressed CD5 and CD19 before and after stimulation withCpG-ODN. B cells from normal controls were separated by posi-tive selection using CD19-coated magnetic beads (Dynal) and De-tachabead (Dynal) according to the manufacturer’s instructions toa purity of
.
98%. Purification of tonsillar B-cell fraction was per-formed by negative selection as described for CLL-B cells. T cellsfor mixed lymphocyte reactions and Elispot assays were isolatedfrom PBMNC by incubation with anti-CD14– and anti-CD19–coated magnetic beads resulting in
.
97% CD2 positive cells. Au-tologous T cells from B-CLL patients were obtained by incubationwith CD4 and CD8-coated magnetic beads (Dynal) and Detacha-bead (Dynal) according to the manufacturer’s instructions. If nec-essary, remaining B-CLL cells were depleted by incubation withCD19 beads.
Culture conditions and proliferation assay
Purified normal and leukemic B cells were cultured in RPMI 1640medium (Biochrom) supplemented with 10% fetal calf serum(Biochrom), penicillin/streptomycin 50 IU/mL, Na-pyruvate 1mM, L-glutamine 2 mM, L-asparagine 20
m
g/mL, 2 mercaptoetha-nol 0.05 mM, HEPES 10 mM, and MEM nonessential amino acids0.7
3
(Biochrom) at 37
°
C and 5% CO
2
in a fully humidified atmo-sphere in 24-well plates at 10
6
cells in a total volume of 1 mL.Purified B cells were cultured at 5
3
10
4
cells/200 uL in round-bottom plates with ODN, IL-2, or both for induction of B-cell pro-liferation. After 72 hours, the cells were pulsed with
3
H-thymidine(DuPont, Paris, France) and harvested in a PHD-Cell Harvester(Dunn Labortechnik) after 16 hours. Thymidine incorporation wasquantified in a B counter (Beckmann, Munich, Germany).
560
T. Decker et al./Experimental Hematology 28 (2000) 558–568
Mixed lymphocyte reaction
Purified T cells (2
3
10
5
) were cultured with irradiated B-CLL cellsat a 1:2 stimulator/responder ratio for 5–7 days. Before the MLR,B-CLL cells were cultivated with or without ODN, IL-2, or bothfor 48 hours. Cells were washed two times, irradiated (2000 rad),and cultivated with allogeneic or autologous T cells. Cocultureswith autologous T cells were performed in the presence of IL-2 (10U/mL). During the last 16 hours of culture, the cells were pulsedwith 1
m
Ci
3
H-thymidine (DuPont). Cultures were performed inthe presence of human AB serum (PAA Laboratories, Martinsried,Germany)
Elispot assay
Wells of nitrocellulose-bottomed microtiter plates (Millipore, Bed-ford, MA, USA) were coated with 100
m
L IFN
g
mAb (1D1K,Hoelzel Diagnostika) overnight at 4
8
C. After successive washingswith PBS, the coated wells were filled with 100
m
L containing 10
5
T cells and irradiated B-CLL cells as described previously. Cellswere cultured for 24 hours in RPMI 1640 5% human AB serum.Captured cytokines were detected using mAB 7-B6-1 (Hoelzel Di-agnostika) and spots were developed as described previously [20].
125
I-IL-2 binding assay
Equilibrium binding of
125
I-IL-2 was performed as described else-where [21]. Cell aliquots (10
6
) were incubated for 3 hours at 4
8
C inthe presence of various concentrations of
125
I-IL-2 (New EnglandNuclear, Boston, MA, USA) in 200
m
L of culture. Free
125
I-IL-2was separated from surface bound
125
I-IL-2 by rapid centrifugationon binding oil (dibutyl phthalate; Sigma, St. Louis, MO, USA).Nonspecific binding was determined in the presence of 100-foldexcess of cold IL-2.
Cytokine measurements
Cytokine levels were determined using commercially availableELISA kits (TNF-
a
and IL-6 Duoset) according to the instructionsof the manufacturer (Genzyme, Germany). Mean values were cal-culated from duplicate measurements.
Immunphenotyping
Cells were washed in PBS containing 2% FCS and incubated withsaturating amounts of fluorochrome-conjugated mAB. After 30minutes at 4
8
C, the cells were washed with PBS/2%FCS and ana-lyzed via flow cytometry using a Coulter Epics XL cytofluorome-ter, acquiring 5000 events. Data were analyzed using WinMDI 2.8FACS software (Scripps Research Institute, La Jolla, CA, USA).The relative expression of surface antigen is described as the meanfluorescence intensity ratio (MFIR). MFIR equals the MFI of cellsstained with a fluorochrome-conjugated antigen-specific mAb di-vided by the MFI of cells stained with a fluorochrome-conjugatedisotype control mAb. MFIR-induction was calculated using theequation MFIR (stimulated/MFIR [untreated]).
Statistical analysis
Data from individual experiments are presented as mean
6
SEM.Statistical significances were determined using the Wilcoxonsigned rank test and the Mann-Whitney test as appropriate. A
p
value
,
0.05 was considered to be statistically significant.
Results
DSP30 induces the expression ofhigh-affinity IL-2 receptors on B-CLL cells
Purified B-CLL cells and normal peripheral blood B cellsderived from healthy donors were cultured with the immuno-stimulatory CpG-ODN DSP30 [9] or the control ODN pZ2lacking the CpG dinucleotide [13]. After a culture period of48 hours, CD25 antigen expression was analyzed by immu-nophenotyping. Figure 1A shows a typical example of threeindependent experiments. Although pZ2 marginally in-duced the CD25 antigen, maximal upregulation occurred inthe presence of DSP30. To confirm whether the upregula-tion of CD25 resulted in the formation of high-affinity re-ceptors, we performed IL-2 binding assays with
125
I-IL-2 onB-CLL cells after 48 hours of culture with medium alone,DSP30 or pZ2. Figures 1B and 1C show a typical bindingexperiment and scathard plots of the binding data. B-CLLcells activated with DSP30 expressed 1210 receptors percell with a Kd of 5 pM, and no high-affinity receptors couldbe detected on B-CLL cells cultured in the presence of thenon–CpG-ODN pZ2 or medium alone. Similar results wereobtained in two additional experiments.
CD25 upregulation is stronger in B-CLLcells as compared with normal B cells and resultsn a higher number of IL-2 receptors with higher affinity
The effect of DSP30 on upregulation of the IL-2 receptor
a
-chain was compared in 11 B-CLL patients and six normalcontrols. One representative experiment for B-CLL patients(Fig. 2A) and normal controls (Fig. 2B) is shown. Using theequation described in the Materials and Methods section,we found a significantly (
p
5
0.03) stronger upregulation ofCD25 in B-CLL patients as compared with normal B cells(Fig. 2C). Because little is known about the level of CD25needed for the formation of high-affinity IL-2 receptors, weperformed binding assays on three B-CLL patients andthree normal controls. Due to the large cell numbers needed,we performed these experiments on tonsillar B cells, whichregulate the CD25 antigen to the same extent as peripheralblood B cells (data not shown). In contrast to B-CLL cells,stimulation with the CpG-ODN DSP30 resulted in the for-mation of fewer IL-2 receptors with less affinity in normalB cells. Receptor numbers per cell were 510, 330, and 702with a Kd of 54, 56, and 190 pM compared with 520, 831,and 1200 receptors per cell with a Kd of 5, 9, and 16 pM inB-CLL cells in three independent experiments with tonsillarB cells and B-CLL cells. Figure 2D shows the scatchardplots of the corresponding binding data.
Effect of costimulation of CpG-ODNand IL-2 on proliferation and cytokine production
To examine the functional consequences of the formation ofhigh-affinity receptors in B-CLL cells vs IL-2 receptors
T. Decker et al./Experimental Hematology 28 (2000) 558–568
561
with reduced affinity in normal B cells, normal and malig-nant lymphocytes were cultured in the presence of IL-2 100U/mL, DSP30 1
m
m, or the combination of both. To deter-mine the proliferative response,
3
H-thymidine was added
for 16 hours after a 72-hour culture. Stimulation withDSP30 alone resulted in a moderate proliferation of normalB cells that was significantly higher (
p
5
0.04) than inB-CLL cells. Although IL-2 alone was not able to stimulate
Figure 1. (A) CD25 expression in B-CLL cells activated by DSP30 and pZ2. The 106 B-CLL cells were stimulated in 24-well plates with the CpG-ODNDSP30 or the control ODN pZ2 for 48 hours. Histograms show fluorescence intensity of the CD25 staining (shaded) superimposed with that of the negativecontrol (unshaded). (B) Illustrates the equilibrium binding data for 125I-IL-2 binding. (C) Represents the Scatchard analysis of the specific binding data.
562
T. Decker et al./Experimental Hematology 28 (2000) 558–568
normal B cells or B-CLL cells, addition to DSP30 led to astrong increase in thymidine incorporation and restored thereduced proliferative response of B-CLL cells to CpG-ODNalone to a normal level (Fig. 3A). This costimulatory effectcould be achieved with concentrations as low as 1 U/mL,with a maximal effect at 100 U/mL (data not shown). Ofnote, stimulation with the non–CpG-ODN pZ2 induced no
proliferation in B-CLL cells, and addition of IL-2 caused amoderate increase in thymidine incorporation in three inde-pendent experiments with B-CLL cells (data not shown).
Cytokine measurements were performed after 48 hoursof culture. TNF-
a
or IL-6 were not detectable in superna-tants of all B-CLL samples and normal B cells cultured inthe presence of IL-2 or medium alone. After stimulation
Figure 2. Change of CD25 antigen expression and formation of high affinity receptors in B-CLL cells and normal B-cells. A representative example ofCD25 upregulation after 48-hour stimulation with DSP30 1 mm in B-CLL cells (A) and normal B cells is shown (B). MFIR induction was calculated asdescribed in the Materials and methods section, and is presented as mean 6 SEM of 11 samples from B-CLL patients and six normal controls (C). Three inde-pendent Scatchard plots of B-CLL cells and human tonsillar B cells are depicted (D).
T. Decker et al./Experimental Hematology 28 (2000) 558–568
563
with DSP30, IL-6 and TNF-
a
secretion could be detected incultures of normal and malignant B cells. Addition of IL-2to normal B cells stimulated with DSP30 did not alter cy-tokine production (Figures 3B and 3C). In contrast, IL-6 se-cretion by malignant cells was significantly (
p
5
0.002) in-creased on addition of IL-2 (Fig. 3B). TNF-
a
productionwas also enhanced, but the difference did not reach statisti-cal difference (
p
5
0.11) (Fig. 3C). Neither stimulation withpZ2 alone nor pZ2 and IL-2 resulted in measurable IL-6 orTNF-
a
levels in three independent experiments (data notshown).
Regulation of CD80 and CD86expression by CpG-ODN and IL-2
B7 costimulatory molecules are crucial in evoking autolo-gous or allogeneic tumor recognition. Upregulation of bothCD80 and CD86 has been described following stimulationof B-CLL cells with CpG-ODN [6,17]. After a 48-hour cul-ture period with the CpG-ODN DSP30, CD80 and CD86were upregulated in all B-CLL samples and normal B cellswere tested (Figs. 4 and 5). Stimulation with IL-2 alone wasnot sufficient to regulate the expression of these costimula-tory molecules. In contrast, the addition of IL-2 to a CpGstimulus resulted in further upregulation of CD80 and CD86expression in all B-CLL samples with a more pronouncedeffect regarding CD80 regulation. In line with the prolifera-tion data, a costimulatory effect was achieved with 1 U IL-2/mL but 100 U/mL were required for the maximal effect(data not shown). No additional effect was observed whennormal B cells were stimulated with CpG-ODN and IL-2.Representative immunophenotyping experiments from 11B-CLL samples and six normal controls showing CD80 andCD86 expression after culture in medium alone, IL-2,DSP30, or DSP30 and IL-2 are depicted in Figures 4A, 4B,5A, and 5B. The pooled data regarding MFIR induction arepresented in Figures 4C and 5C, showing a statistically sig-nificant upregulation of CD80 (
p
,
0.0003) and CD86 (
p
,
0.004) in B-CLL cells but not in normal B cells. To confirmwhether upregulation of CD80 expression was due to the in-creased TNF-
a
level after stimulation with DSP30 and IL-2,we added saturating amounts of a mAB against TNF-
a
toB-CLL cells stimulated with DSP30 and IL-2 and observedno effect on CD80 expression. Figure 4D shows one ofthree experiments that gave similar results.
B-CLL cells activated with CpG-ODN and IL-2 arepotent activators of allogeneic and autologous T cells
We examined the stimulatory capacity of B-CLL cells cul-tured with IL-2, CpG-ODN, or both and compared it withthe ability of B-CLL cells cultured in medium alone to acti-vate allogeneic or autologous T cells. After 48 hours of cul-ture with the indicated conditions, B-CLL cells werewashed three times, irradiated, and cocultured with autolo-gous or allogeneic T cells at a stimulator–responder ratio of1:2. Cocultures with autologous T cells were performed in
Figure 3. Proliferation and cytokine production of B-CLL cells and nor-mal B cells in response to DSP30 with or without costimulation with IL-2.Purified B-CLL cells and B cells from normal controls were stimulatedwith DSP30 at 1 mM with or without 100 U IL-2/mL. 3H-thymidine wasmeasured after 3 days of culture in triplicate (A). TNFa (B) or IL-6 (C)was determined using an ELISA system. Results are expressed as mean 6SEM of 11 B-CLL patients and six normal controls.
564
T. Decker et al./Experimental Hematology 28 (2000) 558–568
the presence of IL-2 10 U/mL. After 5 days, cell prolifera-tion was assessed by incorporation of
3
H-thymidine. The re-sults of five independent experiments with allogeneic andtwo additional experiments with autologous T cells are de-picted in Figure 6A. While no increase of the proliferativeindex was observed when B-CLL cells were stimulated with
IL-2 alone, activation of B-CLL cells with DSP30 resultedin an enhanced proliferative response in three experiments.Maximal proliferation occurred in the presence of B cellsstimulated with DSP30 and IL-2 in five of six experiments.Of note, stimulation of B-CLL cells with pZ2 or pZ2 andIL-2 did not increase the proliferative index in two indepen-
Figure 4. Expression of CD80 after stimulation with DSP30, IL-2 or both in B-CLL cells or normal B cells. CD80 expression was examined in B-CLL cellsor normal B cells after 48-hour culture with IL-2, DSP30, or the combination of both. Shaded histograms show fluorescence intensity of CD80 in B-CLL cells(Fig. 4A) and normal B cells (Fig. 4B). Unshaded histograms represent isotype controls. MFIR induction was calculated as described in the material andmethods section, and is presented as mean 6 SEM of 11 samples from B-CLL patients and 6 normal controls (Fig. 4C). Figure 4D shows CD80 expression ofDSP30 and IL-2 stimulated B-CLL cells in the presence of TNFa mAB (aTNFa). One representative experiment out of three performed is presented.
T. Decker et al./Experimental Hematology 28 (2000) 558–568 565
dent experiments (data not shown). Elispot assays to deter-mine Interferon-g production were performed after 24-hourco-culture of T cells and stimulated irradiated B-CLL cellson a nitrocellulose membrane. Figure 6B shows the resultsof five experiments with allogeneic T cells and two with au-tologous T cells. The pattern of response closely resembles
the proliferative response of allogeneic or autologous Tcells, but B-CLL cells stimulated with DSP30 and IL-2, orDSP30 alone led to a more consistent increase of spots/105
cells. Even the T cells of CLL sample 2, which showed noproliferative response when stimulated with autologous B-CLLcells, produced a significant number of spots in response to
Figure 5. Expression of CD83 (shaded histograms) after stimulation with DSP30, IL-2, or both in B-CLL cells (A) and normal B cells (B). Unshaded histo-grams represent isotype controls. MFIR induction is presented as mean 6 SEM of 11 B-CLL samples and six normal controls (C).
566 T. Decker et al./Experimental Hematology 28 (2000) 558–568
activated autologous tumor cells. Spots produced by B-CLLcells alone stimulated by DSP30 and IL-2 were less than 15spots/105 cells and were subtracted from spots produced inthe mixed lymphocyte reaction.
DiscussionThe present study demonstrates that functional high-affinityIL-2 receptors are induced in B-CLL cells after stimulation
with the CpG-ODN DSP30 containing three CG dinucleo-tides. The observed results are sequence-specific and notsimply due to the phosphotrothioate backbone because con-trol experiments using the non–CpG-ODN pZ2 showedonly marginal CD25 regulation that did not result in the for-mation of high-affinity receptors. Stimulation of human Bcells, however, is not strictly dependent on the CpG motifbecause conversion of the CG to GC motifs in the DSP30sequence has been shown to induce still considerable acti-vation of human B cells [16,17]. CD25 upregulation inB-CLL cells was significantly stronger as compared withnormal B cells and resulted in the expression of more IL-2receptors with a higher affinity in B-CLL cells than in nor-mal B cells. Due to the pivotal role of CD25 expression inregulating the affinity of the IL-2 receptor complex [22,23],the differences in CD25 expression after CpG-ODN stimu-lation might account for the higher affinity of IL-2 recep-tors in B-CLL cells. Cocultures with IL-2 and DSP30 wereperformed to examine whether different receptor character-istics result in distinct costimulatory effects in both cellpopulations. Although costimulation with CpG-ODN andIL-2 increased thymidine-incorporation in both normal andmalignant B cells, cytokine production and surface mole-cule regulation was augmented in B-CLL cells but not innormal controls. TNF-a is known to be an autocrine growthfactor for B-CLL cells and is suggested to be important forthe regulation of CD80 expression [24,25]. To investigatewhether the costimulatory effect of IL-2 was mediated bythe increased level of TNF-a in B-CLL cells, we performedCD80 FACS-analysis and proliferation assays in the pres-ence of saturating amounts of mAB against TNF-a and ob-served no effect on proliferation or CD80 expression. Thedifferential effects of IL-2 addition to CpG-ODN activatedB-CLL vs normal B cells, with respect to cytokine produc-tion and surface molecule regulation, might simply reflectthe different numbers and different affinities of the IL-2 re-ceptors expressed. Another explanation is that IL-2 signal-ing in B-CLL cells is altered as compared with normal Bcells once high-affinity receptors are expressed. Frank et al.[26] have reported that, in contrast to normal lymphocytes,all CLL patients tested had STAT1 and STAT3 proteins thatwere constitutively phosphorylated on serine residues 727.It is thought that serine phosphorylation is not sufficient toactivate STATs but can augment gene transcription inducedby tyrosine-phosphorylated STATs [27]. IL-2 and other cy-tokines signaling via the STAT proteins might, therefore,cause maximal activation in B-CLL cells [28].
Most importantly, the costimulatory molecules CD86 andCD80 were further upregulated as compared with stimulationwith CpG-ODN alone in B-CLL cells but not in normal Bcells when IL-2 was used as a costimulus (Figs. 4 and 5).Both molecules have been described to be of crucial impor-tance for the induction of autologous or allogeneic immunerecognition [29], but the level of CD80 seems to be the criti-cal determinant in activation of naive CD8 T cells and media-
Figure 6. Activation of allogeneic and autologous T cells by B-CLLcells. Mixed lymphocyte cultures and calculation of the proliferative indexwere performed as described in the Materials and Methods section. Datashow five individual experiments with allogeneic T cells (open symbols)and two with autologous T cells (Black symbols, Exp. 1; shaded symbolsExp. 2) (A). The mean of each stimulation is indicated by the horizontalline. Elispot assays were performed as described in the Material and Meth-ods section after 24-hour culture and present corresponding data of theexperiments described (B). Using stimulated B-CLL cells alone, spotscould not be detected in cells cultured in medium alone or IL-2. Spots wereless than 5/105 cells using DSP30 as the only stimulus and less than 15/105
cells using DSP30 together with IL-2 and were subtracted from the spotnumbers counted in the cocultures with T cells.
T. Decker et al./Experimental Hematology 28 (2000) 558–568 567
tion of antitumor immunity [30,31]. Although CD86 expres-sion was further increased by approximately twofold whenB-CLL cells were stimulated by CpG-ODN and IL-2, the ef-fect on CD80 expression was much more pronounced. Ac-cordingly, B-CLL cells stimulated with CpG-ODN and IL-2were the most active in inducing proliferation and Interferon-gproduction of allogeneic or autologous T cells (Fig. 6). Be-cause extensive washing was performed to remove residualCpG-ODN or IL-2 and the same effects were observed in au-tologous cultures when IL-2 was present in all stimulations,we consider it unlikely that this effect is mediated by remain-ing IL-2. Interestingly, B-CLL cells stimulated with IL-2 andDSP30 produced a small amount of positive spots in theElispot assay. This might be due to some remaining T cellsactivated by IL-2, but low Interferon-g production by B-CLLcells has been postulated previously [32]. However, the spotnumber shown in Figure 6B has been corrected by subtract-ing spots produced by B cells alone.
Because B-CLL cells express MHC I and II moleculesand surface Ig at variance to Ig expressed by normal B cells[33] and normal function of T cells from B-CLL patientshas been described [34], the leukemic cells should be sus-ceptible to host-immune recognition. However, B-CLL cellsare poor stimulators of T cells even in allogeneic mixed lym-phocyte reactions because they lack important costimulatoryaccessory molecules necessary for efficient T-cell activation[17,35]. Prior studies have demonstrated the potential forT-cell mediated immunotherapy for B-cell lymphoma (forreview see [36]). CD40 activated malignant B cells havebeen shown to stimulate allogeneic and autologous T cellsin B-CLL [18,19], follicular lymphoma [37], and acute lym-phoblastic leukemia [38]. We have recently reported thatactivation of B-CLL cells with CpG-ODN was as efficientin stimulating allogeneic T cells as CD40Ligation [17]. Herewe show that costimulation with CpG-ODN and IL-2 fur-ther increases the potential of B-CLL cells to activate allo-geneic and autologous T cells. However, future studies willshow whether generation of cytotoxic T cells is possible us-ing this approach and extensive in vitro testing has to beperformed to define the optimal conditions required. Poten-tial clinical applications include the generation of tumor vac-cines or stimulator cells for adoptive immune therapy. Sys-temic application of CpG-ODN is tested in ongoing Phase Istudies (H. Davis, personal communication), and IL-2 appli-cation has been demonstrated to be feasible in the treatmentof cancer and leukemia [39,40]. Taken together, stimulationof B-CLL cells with CpG-ODN and IL-2 might be an attrac-tive immunotherapeutic strategy for B-CLL and seems to besuperior to stimulation with CpG-ODN alone.
AcknowledgmentsThis work was supported by a research grant from the TechnicalUniversity of Munich (KKF H30-97). We thank Ch. Meyer zumBüschenfelde for helpful discussions.
References1. O’Brien S, Del Giglio A, Keating M (1995) Advances in the biology
and treatment of B-cell chronic lymphocytic leukemia. Blood 85:3062. Fluckiger AC, Rossi JF, Bussel A, Bryon P, Bancherau J, Defrance T
(1992) Responsiveness of chronic lymphocytic leukemia B cells acti-vated via surface Igs or CD40 to B-cell tropic factors. Blood 80:3173
3. Caligaris-Cappio F, Hamblin TJ (1999) B-cell chronic lymphocyticleukemia: a bird of a different feather. J Clin Oncol 17:399
4. Hivroz C, Grillot-Courvallin C, Brouet JC, Seligman M (1986) Heter-ogeneity of responsiveness of chronic lymphocytic leukemic B cells toB cell growth factor IL-2. Eur J Immunol 16:1001
5. Nelson BH, Willerford DM (1998) Biology of the interleukin-2 recep-tor. Adv Immunol 70:1
6. Fluckiger AC, Garrone P, Durand I, Galizzi JP, Bancherau J (1993) In-terleukin 10 (IL-10) upregulates functional high affinity receptors onnormal and leukemic B lymphocytes. J Exp Med 178:1473
7. Karray S, Dautry-Varsat A, Tsudo M, Merle Beral H, Debre P,Galanaud P (1990) IL-4 inhibits the expression of high affinity IL-2 re-ceptors on monoclonal human B cells. J Immunol 145:1152
8. Begley CG, Burton JD, Tsudo M, Brownstein BH, Ambrus JL, Wald-mann TA (1990) Human B lymphocytes express the p75 component ofthe interleukin 2 receptor. Leuk Res 14:263
9. Liang H, Nishioka Y, Reich CF, Pisetsky DS, Lipsky P (1996) Activa-tion of human B cells by phosphorothioate oligonucleotides. J Clin In-vest 98:1119
10. Lipford GB, Heeg K, Wagner H (1998) Bacterial DNA as immune cellactivator. Trends Microbiol 6:496
11. Krieg AM, Yi AK, Matson S, Waldschmidt TJ, Bishop GA, TeasdaleR, Koretzky DA, Klinman DM (1995) CpG motifs in bacterial DNAtrigger direct B-cell activation. Nature 374:546
12. Yi AK, Krieg AM (1998) Cutting edge: rapid induction of mitogen-activated protein kinases by immune stimulatory CpG DNA. J Immu-nol 161:4493
13. Häcker H, Mischak H, Miethke T, et al. (1998) CpG-DNA-specific ac-tivation of antigen-presenting cells requires stress kinase activity andis preceded by non specific endocytosis and endosomal maturation.Embo J 17:6230
14. Yi AK, Chang M, Peckham DW, Krieg AM, Ashman RF (1998) CpGOligodeoxyribonucleotides rescue mature spleen B cells from sponta-neous apoptosis and promote cell cycle entry. J Immunol 160:598
15. Wang Z, Karras JG, Colarusso TP, Foote LC, Rothstein TL (1997) Un-methylated CpG motifs protect murine B lymphocytes against FAS-mediated apoptosis. Cell Immunol 180:162
16. Bauer M, Wagner H, Heeg K, Lipford GB (1999) DNA activates hu-man immune cells through a CpG sequence dependent manner. Immu-nology 97:699
17. Decker T, Schneller F, Sparwasser T, Lipford GB, Wagner H, PeschelC (2000) Immunostimulatory CpG-Oligonucleotides cause prolifera-tion, cytokine secretion and an immunogenic phenotype in B-CLL pa-tients. Blood 95:999
18. Buhmann R, Nolte A, Westhaus D, Emmerich B, Hallek M (1999)CD40-activated B-Cell chronic lymphocytic leukemia cells for tumorimmunotherapy: stimulation of allogeneic versus autologous T cellsgenerate different types of effector cells. Blood 93:1992
19. Kato K, Cantwell MJ, Sharma S, Kipps TJ (1998) Gene transfer ofCD40-ligand induces autologous immune recognition of chronic lym-phocytic leukemia B cells. J Clin Invest 101:1133
20. Herr W, Schneider J, Lohse AW, Meyer zum Büschenfelde KH,Wölfel T (1996) Detection and quantification of blood-derived CD81
T lymphocytes secreting tumor necrosis factor alpha in response toHLA-A2.1-binding melanoma peptide antigens. J Immunol Methods191:131
21. Robb RJ, Greene WC, Rusk CM (1984) Low and high affinity cellularreceptors for interleukin 2. Implications for the level of Tac antigen. JExp Med 160:1126
568 T. Decker et al./Experimental Hematology 28 (2000) 558–568
22. Roessler E, Grant A, Ju G, Tsudo M, Sugamura K, Walmann TA(1994) Cooperative interactions between the interleukin-2 receptor al-pha and beta chains alter the interleukin-2 binding affinity of the re-ceptor subunits. Proc Natl Acad Sci U S A 91:3344
23. Kondo M, Takeshita T, Ishii N, et al. (1994) Functional participationof the IL-2 receptor-gamma chain in IL-7 receptor complexes. Science263:1453
24. Foa R, Massaia M, Cardona S, et al. (1990) Production of tumor necro-sis factor-alpha by B-cell chronic lymphocytic leukemia cells: a possi-ble regulatory role for TNF in the progression of the disease. Blood76:393
25. Ranheim EA, Kipps TJ. Tumor necrosis factor-alpha facilitates induc-tion of CD80 (B7-1) and CD54 on human B cells by activated T cells:complex regulation by IL-4, IL-10 and CD40L. Cell Immunol 161:226
26. Frank DA, Mahajan S, Ritz J (1997) B lymphocytes from patients withchronic lymphocytic leukemia contain signal transducer and activatorof transcription (STAT)1 and STAT3 constitutively phosphorylated onserin residues. J Clin Invest 100:3140
27. Wen Z, Zhong Z, Darnell JE (1995) Maximal activation of transcrip-tion by STAT1 and STAT3 requires both tyrosine and serine phospho-rylation. Cell 82:241
28. Meinhardt G, Wendtner CM, Hallek M (1999) Molecular pathogenesisof chronic lymphocytic leukemia: factors and signalling pathways reg-ulating cell growth and survival. J Mol Med 77:282
29. Lenschow DJ, Walunas TL, Bluestofne JA (1996) CD28/B7 system ofT cell costimulation. Ann Rev Immunol 14:233
30. Fiels PE, Finch RJ, Gray GS, et al. (1998) B7.1 is a quantitativelystronger costimulus than B7.2 in activation of naive TCR-transgenic Tcells. J Immunol 161:5268
31. Stremmel C, Greenfield EA, Howard E, Freeman, GJ, Kuchroo VK(1999) B7-2 expressed on EL4 Lymphoma suppresses antitumor Im-munity by an interleukin-4-dependent mechanism. J Exp Med 189:919
32. Buschle M, Campana D, Carding SR, Richard S, Hoffbrand AV, Bren-ner MK (1993) Interferon gamma inhibits apoptotic cell death in B cellchronic lymphocytic leukemia. J Exp Med 177:213
34. Decker T, Flohr T, Trautmann P, et al. (1995) Role of accessory cellsin cytokine production by T cells in Chronic B-cell lymphocytic leuke-mia. Blood 86:1115
35. Dazzi F, D’Andrea E, Biasi G, et al. (1995) Failure of B cells ofchronic lymphocytic leukemia in presenting soluble and alloantigens.Clin Immunol Immunopathol 75:26
36. Schultze JL, Nadler LM (1999) T cell mediated immunotherapy for Bcell lymphoma. J Mol Med 77:322
37. Schultze JL, Seamon MJ, Michalak S, Gribben JG, Nadler LM (1997)Autologous tumor infiltrating T cells cytotoxic for follicular lym-phoma can be expanded in vitro. Blood 89:3806
38. Cardoso A, Schultze JL, Boussiotis VA, et al. (1996) Pre-B acute lym-phoblastic leukemia cells may induct T cell anergy to alloantigen.Blood 88:41
39. Keilholz U, Conradt C, Legha SS, et al. (1998) Results of Interleukin-2-based treatment in advanced melanoma: a case record-based analysisof 631 patients. J Clin Oncol 16:2921
40. Cortes JE, Kantarjian HM, O’Brien S, et al. (1999) A pilot study of in-terleukin-2 for adult patients with acute myelogenous leukemia in firstcomplete remission. Cancer 85:1506
