Correspondence to: Dr. Arndt Biissing, Krebsforschung Herdecke, Dept. of Applied Immunology, University Herdecke, 0-583l3 Herdecke, Germany. Tel: ++49-2330-623246; Fax: ++49-2330-624003; E-mail: ArBuessQat-online.de.

Key Words: Chronic lymphocytic leukaemia, apoptosis, proliferation, activation markers, interleukin-6, Viscum album L. extracts

.B chronic lymphocytic leukaemia (B-CLL) is characterised by an accumulation of long-lived monoclonal B cells in the peripheral blood and bone marrow (1). Despite their developmental arrest and enhanced longevity in vivo (2), leukemic B cells spontaneously undergo apoptosis when cultured in vitro (3,4). These B cells are protected from spontaneous cell death by cytokines such as interleukin (IL)-4, interferon (IFN)-gamma and IFN-alpha (3-6). Some authors reported that IL-I, IL-2, IL4, IL-13 and tumour necrosis factor (TNF)-alpha can rescue leukemic B cells from apoptosis (3-6), while others observed no effect on spontaneous apoptosis by IL-1, IL-2, and IL-6 (5).

In contrast, proliferation of B-CLL cells is induced by TNF-alpha (7-10), and IL-6 (10). However, in the presence of high TNF-alpha concentrations, IL-6 inhibited TNF-induced B-CLL growth (7,10). In other experiments (7), IL-6 did not induce proliferation of CLL B cells but inhibited B-CLL growth; this effect was restricted to cells from patients who also showed growth stimulation by TNF (7). Also IL-4 inhibited both spontaneous and TNF-induced [³H]-thymidine uptake (9). In fact, IL-4 and IL-6 were suggested to inhibit DNA synthesis by inhibiting potential autocrine growth loops (9,10). These results indicate that the effects of IL-6 on B-CLL cells may be highly dependent on the culture conditions and the investigated cell line.

Given its function as a B cell growth and differentiation factor (11), it was suggestive to link IL-6 to the patho-physiology of certain B cell tumours (12). The fact that IL-6 may act as a growth inhibitor in B-CLL, as observed by Aderka et al (7), shows the complexicity of IL-6 action on B cells and may be relevant to the clinical management of B-CLL. The present study was intended to more clearly define the effects of an in vitro stimulation of leukemic B cells from patients with CLL and B cells from healthy controls with immunomodulators such as IL-6. pokeweed mitogen (PWM), lipopolysaccharides (LPS), and extracts from Viscum album L. (VAL), which were used as an adjuvant in complementary cancer treatment in Europe. These drugs were recognised to induce apoptosis in various tumour cell lines and lymphocytes (13-17), but also to induce a release of pro-inflammatory cytokines such as TNF-alpha, IL-1 and IL-6 (18-24). Due to the fact that patients with B-CLL were also treated with VAL extracts, it is of outstanding importance to exclude possible B-CLL propagating effects. Here, we report that IL-6 and PWM stimulated B-CLL cells in vitro resulting in a strong release of IL-6 and enhanced activation marker expression, while the VAL-induced IL-6 release was associated with cell death via apoptosis and did not enhance proliferation of leukemic B cells.

Materials and Methods

Patients. Thirteen patients wth B-CLL (Table l) and 6 healthy controls, who all gave informed consent to the in vitro investigations, were included in this study. The diagnoses were based on clinical. morphological, and immunophenotypical criteria. None of the patients received chemotherapy or corticosteroids at the time of testing nor had signs of overt inflamation or infections when studied. The mean values of peripheral blood cell subsets of investigated patients were given in Table II. Values from the healthy test persons were in the normal range (data not shown).

Immunomodulators. IL-6 was obtained from Boehringer Mannheim, Germany and was added at 100 and 1000 U/ml, LPS from Salmonella abortus equi (Sigma. Deisenhofen, Germany) was added at 25 and 250 microg/ml, and PWM (Sigma) at 2.5 microg/ml. Aqueous extracts from mistletoe grown on pine (Helixor P, HP) and fir tree (Helixor A, HA) were kindly provided by Helixor, Rosenfeld, Germany (the isotonic solution corresponds to 50 mg of fresh plant material per ml). Thc content of toxic mistletoe Iectins (ML) was 950 ng/ml in HP and 260 ng/ml in HA (galNAc- binding ML II/III: ELLA test); viscotoxins were < 0.5 pg/ml (HPLC). HA and HM were added at lO, l00 and l000 microg/ml (corrcsponding to thc weight of fresh plant material).

Flow cytometric analysis. Flow cytometric analysis of surface molecu1es from peripheral blood lymphocytcs (EDTA blood) and cultured lymphocytes was performed on EPICS XL-MCL flow cytometer (Coulter. Krefeld, Germany) using monoclonal antibodies (mAb) against the following antigens: CD3, CD5, CDl9, CD25, CD7l (Coulter-Immunotech. Krefeld. Hamburg. Germany). For each sample, lO pl of the mAb was added to l0l pl of ce1I suspension. Forward angle light scatter (FSC) and high perpendicular light scatter (SSC) were used to gate lymphocytes and to exclude cell debris. All quadrants were adjusted to the anti-mouse isotype controls.

Cell culture conditions. Ficoll-isolated peripheral blood mononuclear cells (PBMC, I.5 x lO" cells/ral) from B-CLL patients and healthy test persons were incubated in RPMI FG 1640 medium (Biochrom. Berlin, Germany) supplemented with gentamicin (Refobacin. Merek. Germany. 0.1 mg/ml) and 10% autologous plasma for 7 d in 24-well flat bottom microtiter plates (NUNC, Roskilde, Denmark). in a humidified 5% CO2 atmosphere at 37 C. Differentiation of incubated cells from B-CLL patients revealed lymphocytes at 88.9 +/- 4.8 % and monocytes at 8.0 +/-4.4 %, while FSMC from healthy controls were 84.9 +/- 5.0 % lymphocytes and monocytes at 1l.5 +/- 4.6 %. PBMC. For proliferation studies, the thymidine analogue 5-Bromo-2-deoxyuridine (BrdU. Sigma) was added at a final concentration of 30 microg/ml during the last 20 h of the incubation period. Supernatants were carefully aspirated end kept frozen at -20'C until use.

Viability analysis. To differentiate apoptotic and necrotic cells. Ficoll-isolated lymphocytes cultured for 7 d were analysed for Annexin-V (AV) binding and propidium iodide (PI) uptake as described (l5). Briefly, after washing of cultured cells (1.5 x l0^6 /ml) with PBS and resuspension in binding buffer (l0 nM Hepes/NaOH. pH 7.4, l40 nM NaCl 2.5 nM CaCl2). the cells were stained with 5 pl RTC-labelled Annerin-V (PharMingen, San Diego, CA. USA) and Pl (5 pg/ml: Sigma). After l5 min. of incubation at room temperature in the dark. binding buffer was added and cells were analysed by flow cytometry. Cells undergoing apoptosis, bind Annexin-V to phospatidyl serine translocated from the inner to the outer leaflet of thc cell membrane. and exclude the DNA-intercalating dye propidium iodide (Pl), while necrotic cells (primary necrosis and late,apoptosis) became permeable to Pl. Viable cells are Annexin-V and Pl .

Proliferation analyses by flow cytometry. Detection of proliferation was performed by flow cytometry using the method described by Carayon and Bord (26). In brief, at the end of the culture period, cells were washed and incubated with a monoclonal antibody to CDl9 or the isotype control (Coulter-Immunotech) in the presence of 2 microl normal mouse serum (DAKO. Hamburg) for 3l min. in the dark at room temperature (RT). Subsequently. the cells were fixed in l% paraformaldehyde/PBS/0.01% Tween 20 (Sigma) ovenight. After washing. 1 ml of Ca²+ /Mg²+ -enriched PBS (Biochrom) and 50 Kunitz units/ml ot DNase-I (Sigma) were added to each tube for 30 min. at 37'C (water bath). After washing. cells were resuspended in PBS/lO% BSA,0.5% Tween 20 (150 microl) and anti-BrdU or intracellular isotype control (Becton Dickinson. Heidelberg, Germany). respectively, were added for 45 min. at RT. Subsequently, thc cells were analysed by flow o cytometer by gating cells with adequate FSC and SSC signals.

Cytokine assay. Cytokines were measured by a double-sandwich ELISA as described (22). using antibody pairs (capture and detector Ab) to human IL4 and IL-6 (R&D systems. Wiesbaden, Germany), recombinant human IL4 and natural human IL-6 from Boehringer Mannheim (Germany). In brief, microtiter plates (Nunc, Wiesbaden, Germany) were coated with the monoclonal anti- cytokine antibodies diluted in hydrogencarbonate buffer (0,1M. pH 9.6) and incubated overnight (4º C). Undiluted supernatants of the cell cultures and cytokine standards at different concentrations were added for 2 h. Afterwards, plates were incubated with the biotinylated detector antibodies for another 2 h, followed by the peroxidase-conjugated streptavidine (Sigma) for 30 min at RT. Detection wns performed by the addition of substratum (o- phenylenediamine) dissolved in citrate buffer (0,1 M. pH 5.0)) using a Spectra ELISA reader (SLT. Crailsheim. Germany). Cytokine concentration was calculated from a standard curve using the SLT EasyFit software. The cut off for IL-6 was 30 pg/ml, and for IL 4 50 pg/ml.

Statistics. Statistical analyses were performed using Wilcoxon's signed rank test or Mann-Whitney test, as indicated. Logarithmic regression analysis of data was performed using the Microsoft Excel 97 software.

Results

Viability analysis. In both patients and controls, addition of the mistletoe extracts HP and HA led to a dose dependent decrease of the viability, as measured by binding of AV and PI uptake. At a final concentration of 1000 pg/ml, the decrease was statistically significant (Tables III and IV). In cultured B-CLL cells, spontaneous cell death was insignificantiy higher (p = 0.087) as compared to lymphocytes from healthy controls. In the presence of IL-6. the viability of leukemic B cells was slightly enhanced, thus, the number of necrotic cells decreased (Table III). Incubation of cells with PWM resulted in activation-induced apoptosis in both. patients and controls (Tables III and IV). In response to HA and HP, the number of AV+ PI+ necrotic cells significantly increased in B-CLL patients (Table III), while, however, in the healthy controls, AV+ PI apoptotic cells significantly increased in response to HP, and in response to HA. both apoptotic and necrotic cells (Table IV). When compared to the B-CLL patients, the number of AV+ PI apoptotic cells was significantly higher in lymphoqtes from healthy controls treated in vitro with HP at 100 and 1000 pg/ml (p < 0.004) and HA at l00 pg/ml (p = 0.04). We observed strong interindividual differences in the susceptibility towards HA and HP, indicating that the same amount of biologically active components (i.e. toxic ML) will not warrant similar responses of individuals. However. the number and relative amount of peripheral B-CLL cells did neither correlate with spontaneous viability. nor IL-6-dependent viability nor HP-induced cell death (R² < 0.5.

Activation marker expression. The activation marker expression (CD25 and CD71) on the surface of 7 d cultured leukemic CD19+ B cells was significantly lower than in healthy controls (p < 0.008), and decreased significantly in response to HA and HP (Table III). In the controls, HA and HP at 1.000 pg/ml severely reduced the number of gated events, thus we were unable to analyse the activation marker expression (Table IV). As expected. PWM, and to a lesser extend IL-6, induced the expression of CD25 and CD71 molecules on B cells in both groups. However, the number of peripheral CD5+ B cells did not correlate with activation marker expression after stimulaition induced by PWM and IL-6 (R² < 0.2). When compared to B-CLL cells, the PWM-induced expression of CD25 and CD71 molecules on B cells from healthy controls was significantly higher (p = 0.007). However, the activation marker expression of patients treated with VC. extracts and those without VAL treatment did not differ (data not shown).

Proliferation of lymphocytes. In response to PWM, a BrdU uptake was recognised predominantly in T cells (Tables III and IV; Figure 1), while only some B cell clones responded in 4 individuals (HE, DS and WE: 1%, VS: 2% of B cells). As compared to healthy controls, the BrdU uptake in B-CLL cells was significantly higher (p = 0.005) in response to PWM, but lower in response to IL-6 at 1,000 pg/ml (p = 0.002). Incubation of lymphocytes from healthy individuals with LPS resulted in a slight increase of BrdU+ T cells, in contrast to leukemic B cells. Only one patient (#9) receiving HP s.c. for 12 months (5 mg 3x per week), responded slightly to IL-6 [1,000 U/ml] and HP [100 pg/ml] with an uptake of BrdU in the B cells (Figure 2); however, these B celis did not express CD25 or CD71 molecules after in vitro stimulation, indicating that these cells did not go into mitosis (Figure 1). The PWM and IL-6-induced BrdU uptake did not correlate with the number of peripheral CD5+ B cells (R2 < 0.1); however, in two patients with CD5+ B cells > 100,000 cells/pl, the PWM-induced BrdU uptake in T cells was lower (1.9% and 6.6%) as compared to the other 11 patients with B-CLL (7.6% to 24.1%)

Interleukin-6 release. In the culture supernatants of cells from patients with B-CLL, a significant induction of IL-6 after incubation with HA, HP, IL-6, LPS and PWM was recognised, but both VAL extracts were the least potent (Table III). We observed an IL-6 release in response to activation induced by IL-6 and PWM, but also in response to cell death induced by HA and HP at 1,000 pg/ml. Similar responses were observed in healthy individuais (Table IV), but with higher IL-6 levels and stronger interindividual differences. This effect may be attributed to the higher production of endogenous IL-6 by monocytes and T cells, which is the main source of IL-6, as compared to the leukemic B cells, Again, the number of peripheral CD5+ B-CLL cells did not correlate with PWM and IL-6-induced IL-6 release (R² < 0.3). Surprisingly, patient #9 who was the only with an advanced stage (Rai III) showed a high spontaneous IL-6 release (248 pg/ml) as compared to the other patients (mean: 28.6 + 64.9 pg/ml) and exhibited an increased BrdU uptake in response to IL-6 [1000 U/ml] and HP [100 microg/ml] and a very high level of peripheral HLA-DR+ T cells, which are often observed in patients with high destructive turnover of cells (A. Bussing and M. Schietzel, unpublished observations).

In B-CLL cells, no spontaneous IL-4 release was observed, while in response to PWM, an induction of 11-4 was recognised in 3 out of 13 patients (range: 81 to 907 pg/ml). However, 3 of the healthy test persons, showed a high spontaneous IL-4 level (range: l61 to 1,345 pg/ml), without significant enhancyment of IL-4 release by the stimulators (data not shown).

Discussion

There is growing evidence that some cytokines promote B cell survivai, while others enhance cell death (5). In contrast to B-cell leukaemia and lymphoma, there is much evidence that IL-6 is essentially involved in the in vivo generation of plasma cell neoplasia. The role of the autocrine production of IL-6 in B-CLL is unclear. but it may influence proliferation of progenitor cells (5). It was suggested that an induction of IL-6 may have some importance in vivo because endogenously produced IL-6 is seen to limit the proliferation of leukemic B cells (7). In our experiments, exogenous IL-6 marginaliy protected the B-CLL cells from spontaneous apoptosis, while the VAL-induced release of IL-6 was associated with reduced viability of B-CLL cells. In PWM-treated leukemic B cells we observed both. high IL-6 level and activation-induced apoptosis, indicating that an IL-6 release will neither warrant protection against spontaneous apoptosis nor guarantee an induction of B cell proliferation.

Consistent with the findings of others (6), human cytokines when used alone have minimal effect on the in vitro growth of leukemic B cells and induced proliferation in only some B-CLL populations. In fact. the B-CLL cells are in the G0 phase of the cell cycle and respond poorly to extemal stimuli that induce proliferation in normal B cells (5). IL-4 was shown to enhance survival of B cells in CLL in vitro by inhihiting apoptosis (27), which was associated with an increase of Bcl-2 expression. Jewell et al (3.4) observed protection from apoptosis and increase of Bcl-2 in B-CLL cells by INF-alpha. In our experiments, however, IL-4 was induced only in 3 out of 13 cultures stimulated with PWM. In these cases, no protection was observed as the cells underwent activation-induced apoptosis. Since the galNAc-binding ML III, which is the dominating lectin in the VAL extracts used in our experiments. was recognised to induce apoptosis and to decrease intracellular Bcl-2 proteins (l5-I7), one may speculute that a decrease of Bcl-2 proteins may also be involved in apoptosis of B-CLL cells induced by VAL in vitro, which is however probably independent from an IL-4 and lL-6 production. Moreover, while ML lll selectively killed CD8+ CD62L "memory" cells in vitro, the toxin did not significantly affect CD19+ B cells up to a final concentration of 100 ng/ml (16).

In conclusion. the in vitro administration of IL-6 did not increase or accelerate the proliferation of leukemic B cells. Moreover. VAL incubation did neither result in an enhancement of B-CLL cell proliferation nor protection against spontaneous apoptosis. A release of IL-6 was associated with both. stimulation of B-CLL cells and cell death. These results do not support theoretical objections that VAL-induced cytokines may stimulate B cell neoplasia. However, the clinical relevance of the treatment of B-CLL patients with VAL remains to be clarified.

Table 1. Description of 13 patients with B-CLL