sign in sign up
<<
Home , B cell, T cell

Leukemia (1999) 13, 98–104  1999 Stockton Press All rights reserved 0887-6924/99 $12.00 http://www.stockton-press.co.uk/leu Inhibition of / interaction by B-CLL cells C Kneitz1, M Goller1, M Wilhelm1, C Mehringer1, G Wohlleben2, A Schimpl2 and H-P Tony1

1Medizinische Poliklinik der Universita¨tWu¨rzburg; and 2Institut fu¨r Virologie und Immunologie der Universita¨tWu¨rzburg, Germany

The course of disease in patients suffering from chronic lym- cytes express cell surface molecules and receptors according phocytic (CLL) is determined by a profound dysregul- to their postulated normal counterparts, CD5+ follicular ation of the . The resulting immune sup- 10 pression is the main cause of death in those patients. In the mantle cell B . Therefore, they can be stimu- present study we addressed the question of whether leukemic lated by or contact-dependent sig- B cells (B-CLL) are able to suppress regular T cell/B cell inter- nals similar to normal B lymphocytes.11–14 T cells, isolated action. Activated CD4+T cell clones induce expression of the from CLL patients, exhibit normal T-helper cell function in early activation CD23 on B lymphocytes in vitro. Under respect of secretion and proliferative responses after conditions used, this B cell activation event was dependent polyclonal activation.15 The predominance of B cells as upon direct T cell contact. Addition of certain bystander B-CLL cells or normal B lymphocytes resulted in a cell number-depen- accessory cells, however, apparently alters the immune func- dent inhibition of B cell induction. This seems to reflect the tion of T helper cells in vitro, resulting in an enhanced competition of B-CLL cells for a cell contact-mediated T cell secretion of -2, whereas -␥ was only helper signal. By using CD40 ligand transfected fibroblasts as slightly induced and IL-4 was not expressed.15 a substitute for T cell help, we show that the same B-CLL cells The communication between T helper cells and B cells also suppress CD40 ligand-mediated B cell activation. B-CLL relies upon the time course of the delivery of multiple signals. cells differ in their ability to inhibit CD40 ligand-mediated B cell activation. Some B-CLL cases (eight out of 14) are unable to The activation of helper T cells depends upon the MHC- compete for the T cell or CD40 ligand-mediated signal, even restricted activation of the T cell by the antigen/MHC though they can functionally interact with CD40 ligand and complex and is regulated by a number of co-stimulatory mol- thereby get activated themselves. In addition, these results ecules such as CD80 and CD86.16 The most important recep- indicate that the observed inhibition is not a result of cell tor system that mediates cell contact-dependent T cell help is crowding by merely reducing the chance of specific B cell/T the CD40/CD40-ligand receptor system.17 The CD40 mol- cell interactions. Collectively, these data indicate that B-CLL ecule has been shown to be present on normal and neoplastic cells are able to inhibit the interaction of activated T lympho- 18 cytes with normal B lymphocytes in vitro. Perturbed T cell/B B lymphocytes and is able to transmit signals for the acti- cell interaction may represent an important mechanism under- vation and differentiation of B cells.13,19 The ligand for CD40 lying the various defects of the specific immune system is expressed on activated T lymphocytes and the interaction observed in patients suffering from B-CLL. of CD40 with its ligand appears to be crucial for normal B Keywords: B-CLL; B cell/T cell interaction; CD40; CD40 ligand differentiation and development.20 Taken together there is circumstantial evidence that B-CLL can interact with helper T cells not only in vitro but also in Introduction vivo. However, the events leading to the observed in these patients are not completely understood. Chronic lymphocytic leukemia is predominantly a B cell neo- Here we describe that in certain, but not all cases of B-CLL, plasia which is characterized by the accumulation of small, the malignant B cells can competitively inhibit T cell contact- + long-lived, slowly dividing monoclonal CD5 B cells. The dependent activation of normal B cells in vitro. Furthermore, clinical course of the disease is predominantly determined by we provide evidence that B-CLL cells can suppress the CD40 a profound dysregulation of the immune system, rather than ligand mediated helper signal in particular. This could be an by infiltration and destruction of organs,1 resulting in various important mechanism by which B-CLL cells perturb normal infections2 and signs of .3 It has been shown that helper T cell/B cell interactions in patients suffering from B- almost 60% of deaths are caused by bacterial or viral infec- CLL. tions.4 Mycobacterial or fungal are less frequent.1 Autoimmune-associated phenomena are mainly directed against hematopoietic cells5 and are manifest as autoimmune hemolytic anemia (occurring in 10–25% of patients1)or Materials and methods immune thrombocytopenia (observed in 2% of cases). Remarkably, the are apparently made by ‘nor- mal’ B lymphocytes.6 , occurring in T cell clones up to 60% of B-CLL cases,7 is probably also a result of dys- function of nonclonal B cells. Thus, the intimate interaction The CD4+ T cell clones were obtained by random T cell clon- of leukemic B cells with cells of the normal immune system ing of peripheral lymphocytes (PBL) from healthy is thought to play an important role in the pathogenesis of donors. PBL at a calculated number of 0.5 PBL/well were co- the disease. cultivated with 2 × 104 irradiated feeder cells (PBL mixed with It has been documented in a number of systems that B-CLL EBV-blasts, 3:1) in 96-well round bottom plates (Costar, Cam- cells are able to interact with helper T cells.8,9 B-CLL lympho- bridge, MA, USA). The clones were expanded in RPMI 1640 (Gibco, Karlsruhe, Germany) supplemented with 10% Lym- phocult (Biotest, Dreieich, Germany), 10% human AB-serum Correspondence: H-P Tony, Medizinische Poliklinik, University of Wu¨rzburg, Klinikstr 6-8, D-97070 Wu¨rzburg, Germany; Fax: (BRK-Blutspendedienst, Wu¨rzburg, Germany) and 100 U/ml 49 931 2017068 IL-2 (IL-2 was kindly provided by Prof Dr W Sebald, Received 13 July 1998; accepted 8 September 1998 Wu¨rzburg). T cell/B cell interaction in B-CLL C Kneitz et al 99 CD40 ligand expressing L cells incubated for 90 min at 37°C. Unbound were removed by washing three times with PBS. Cloned T cells CD40 ligand transfectants were prepared as described.21 were irradiated (30 Gy) and cultured at 2 × 104/well with B These cells were trypsinized, washed and plated out in 96- cells at 4 × 104/well. Inhibition of cell contact-mediated well plates at 3 × 103 cells/well, in 24-well plates at induction of normal B lymphocytes (indicator cells) was tested 1.5 × 104/well. Cells transfected only with the pSV2neo plas- by adding increasing numbers of B-CLL cells or normal B lym- mid were used as negative controls. phocytes (competitor cells), stained with PKH26. CD23 expression of indicator B cells was determined after 40 h by double-staining with ␣CD19 and ␣CD23 monoclonal anti- B cell preparations bodies. Necessity of direct cell contact was tested using culture B-CLL cells were isolated from the peripheral blood of inserts with semipermeable membranes in 24-well plates untreated patients. The peripheral blood usually contained 4– (Falcon). T cells were pipetted into culture wells coated with 9 × 1010 leukocytes/l. Cell surface phenotyping revealed a ␣CD3 . In similar experiments, CD40 ligand express- staining pattern typical for B-CLL:22 CD5, CD19, CD23, ing mouse fibroblasts (L cells, 1.5 × 104) were used without CD40, HLA class II positive with low expression of a light stimulation with antibody. B-CLL cells, normal B lymphocytes chain restricted surface IgM. All cells were CD10-negative. or CD40-ligand expressing L cells (as indicated), were added Normal B lymphocytes were isolated from of heal- to the culture wells directly, or they were placed in culture thy patients. For B cell preparation cells or peripheral inserts, thus avoiding direct cellular contact between CD4+ T blood of B-CLL patients were centrifuged through a Ficoll– cells and B-CLL cells. In every case the competitor cells (B- Hypaque gradient at 900 g at 20°C. were CLL cells or normal B lymphocytes) were marked with PKH26 removed from lymphoid cells by adherence on plastic dishes. (Sigma) before cultivation as described above. B lymphocytes were obtained after E-rosetting using AET- treated sheep erythrocytes as described.23 B-CLL cells were usually enriched to 99% CD19-positive cells as shown by FACS analysis FACS analysis. Normal B cell preparations contained 94–97% Three-color fluorescence staining (CD23-FITC (fluorescence CD19-positive cells. Isolated cells were further incubated in 1), PKH26 (fluorescence 2) and CD4 quantum red ml RPMI 1640 (Gibco) with 5 m -glutamine, 100 U/ml penicil- (fluorescence 3)) allowed analysis of CD23 expression on nor- lin G, 100 ng/ml streptomycin (Seromed, Berlin, Germany) ° mal B lymphocytes. The following three gates were combined: and 10% fetal calf serum (FCS; Gibco) at 37 C for the times gate 1: definition of the lymphocyte population according to indicated in the legends. Spleen cells were frozen before iso- cell size and granularity (FSC vs SSC); gate 2: separation of T lation of B lymphocytes in liquid nitrogen to allow the use of lymphocytes by gating on CD4 quantum red negative lympho- identical batches of B cells in control experiments. cytes; gate 3: separation of competitor cells from normal B lymphocytes by gating on PKH26 negative lymphocytes. Finally, expression of CD23 was determined on the remain- Reagents ing indicator B cells, defined as lymphocytes, negative for CD4 and PKH26. IL-2 and IL-4 were kindly provided by Prof Dr W Sebald (Theodor-Boveri-Institut, Wu¨rzburg, Germany). Monoclonal antibodies, PE or FITC coupled, were from Becton Dickinson Results (Heidelberg, Germany; Leu 12 (CD19), Leu 1 (CD 5), Leu 4 (CD3)), Dianova (Hamburg, Germany; IOT8 (CD23)) and Induction of CD23 on splenic B lymphocytes depends Coulter (Krefeld, Germany; ␣-IgM, anti-k-chain, anti-l-chain). upon direct cell contact PKH26 (Sigma, Deisenhofen, Germany) was used according to the manufacturer’s instructions. In detail, competitor cells To investigate T cell/B cell interaction in vitro we cultured (B-CLL cells or normal B lymphocytes) were incubated for 2 splenic B lymphocytes together with anti-CD3 activated CD4- min with 1 mm PKH26 for optimal staining. PKH26 persisted positive T cell clones. B cell activation was analyzed by meas- uring inducible expression of the CD23 antigen on primary B in the cells for more than 48 h as shown by FACS analysis. 24,25 Diffusion of PKH26 from marked cells into other co-cultivated cells. CD23 expression on B lymphocytes was determined cells, eg T cells, could not be observed by FACS analysis. by FACS analysis using double staining of B cells with CD19- PKH26 treatment of cells did not induce proliferation or CD23 PE and CD23-FITC monoclonal antibodies. In Figure 1, B cell expression on normal B cells or proliferation of T lympho- activation is shown by induction of the early activation anti- cytes. gen CD23. Under these conditions B cell activation depends upon direct cell contact as separation of T and B lymphocytes by semipermeable membranes in transwell experiments abol- Cell culture techniques ished the effect. Furthermore, IL-4, a main stimulator of CD23, could not replace direct T cell contact even when added in Cell cultures were usually performed in 96-well flat-bottom optimal doses (Figure 1, exp 2). microtiter plates (Falcon, Heidelberg, Germany) in 200 ml RPMI 1640 with 10% FCS and other supplements as indi- + T cell-dependent activation of B lymphocytes is cated. For stimulation experiments with CD4 T cell clones, reduced by competition with B-CLL and normal B wells were precoated with a against lymphocytes CD3 (purified ␣-CD3 antibodies were kindly provided by Prof Dr E Rieber, Dresden). For coating, 30 ml of a 10 ␮g/ml sol- B-CLL patients suffer from diverse defects of the immune sys- ution of ␣-CD3mAb in PBS (phosphate-buffered saline) were tem. One of the most important is the failure to produce anti- T cell/B cell interaction in B-CLL C Kneitz et al 100

Figure 1 Induction of CD23 on normal B cells depends on direct contact with activated CD4+ T cell clones. Purified B cells were cul- tured together with (hatched bars) or separated from anti-CD3 acti- + Figure 2 Reduction of T cell-dependent CD23 induction on nor- vated cells of a CD4 T cell by semipermeable membrane (dark mal B lymphocytes by B-CLL cells. Irradiated CD4+ T cell clones bars) (experiment 1). In experiment 2 an additional 10 ng/ml IL-4 was (2 × 104 cells/well) were cultured with 4 × 104 splenic B cells in anti- added to the culture. CD23 expression on B lymphocytes (MFI) is CD3 coated wells with increasing numbers of PKH26 marked normal given and was measured after 40 h. CD23-positive B lymphocytes in 6 + B lymphocytes (competitor B cells) or B-CLL cells (up to 10 /well). At %: B lymphocytes alone: 3.7%, B lymphocytes activated T cells: the end of the culture period, cells were stained with CD4-Quantum 85.4%, separated: 4%. B lymphocytes + activated T cells + IL-4: 85%, + red and CD23-FITC. CD23 expression on normal B cells was analyzed separated IL-4: 64%. after gating on lymphocytes negative for PKH26 and CD4. The mean fluorescence intensity of CD23 expression on B lymphocytes is shown as percent response seen without competitor cells. One representative experiment out of three is shown. CD23 expression on B lymphocytes unstimulated: 3.6% postive, MFI: 6.7. After cocultivation with acti- bodies in response to infectious agents. A possible explanation vated T cell clones: 39.5%, MFI: 122.5. could be the disturbance of T cell/B cell interaction by com- peting B-CLL cells. To address this question, B-CLL cells were Table 1 Inhibition of T cell-mediated B cell activation by different added to the B cell/T cell co-culture system described above B-CLL cases (Figure 2). Activation of normal B cells, in the following called ‘indicator cells’, was assayed as enhanced CD23 expression. Competitor cell Cell number Cell number required In order to determine CD23 expression only on these indicator required for 25% for 50% inhibition cells, the competitor cells, B-CLL cells or normal B cells, were inhibition marked before culture by intracellular staining with PKH26. Gating on PKH26-negative B-lymphocytes allowed us to CLL-1 1.0 × 105 5.8 × 105 determine the induction of CD23 expression on the ‘indicator’ CLL-3 1.0 × 104 2.6 × 105 × 5 × 5 B lymphocytes only. The addition of B-CLL cells resulted in a CLL-4 3.3 10 6.5 10 CLL-8 3.1 × 105 5.0 × 105 cell number dependent inhibition of B cell stimulation as CLL-12 6.0 × 105 1.0 × 106 shown by means of CD23 induction on normal B lympho- CLL-14 2.0 × 105 8.5 × 105 cytes. CLL-2, -5, -6, -7, -9, no inhibition no inhibition The analysis of 14 different B-CLL cases showed a varying -10, -11, -13 degree of inhibition. Figure 2 illustrates the typical result of four B-CLL cases which ranged from a strong inhibition (CLL- Fourteen consecutive CLL patients were studied. The experiments 3) to no measurable inhibition as in CLL-2. Six out of 14 B- were performed according to Figure 2. The values for 25% and 50% CLL cases clearly inhibited parallel ongoing B cell activation. inhibition of T cell-mediated B cell activation were calculated from graphs as shown in Figure 2. Competitor cells ranged from 1 × 104 Table 1 illustrates the required amount of B cell competitor to 1 × 106 cells/well. cells resulting in 25% or 50% inhibition of regular B lympho- cyte activation. In eight out of 14 cases no inhibition could be observed up to the number of 1 × 106 ‘competitor cells’. ing mechanism directly seems to influence the transmission The inhibitory capacity of each CLL case was highly reproduc- of T cell help. ible. This is illustrated by Figures 2, 3 and 4 where CLL-2 and CLL-3 showed the same inhibition profile even in different B cell activation systems. Two additional points can be made CD40 ligand-mediated induction of CD23 on normal regarding the experiments. First, the observed inhibition by B- B cells can be inhibited by B-CLL cells CLL cells is not CLL-specific, since normal B cells are at least as potent in vitro (Figure 2). Second, the observed inhibition The CD40/CD40 ligand system is the central signal transduc- is not simply due to cell crowding since B-CLL cells like CLL- ing complex in cell contact-mediated T cell help. We there- 2, -5, -6, -7, -9, -10, -11 and -13 fail to inhibit the response fore considered it likely that interference with the CD40/CD40 even when added in a 25-fold excess. Therefore the underly- ligand system was a reason for B-CLL-mediated inhibition of T cell/B cell interaction in B-CLL C Kneitz et al 101 was completely inhibited by B-CLL lymphocytes (CLL-3, Figure 3). Corresponding to Figure 2, CLL-2 also failed to inhibit the CD40 ligand-mediated B cell activation, redemon- strating that the observed inhibition is not a consequence of cell crowding. The varying degree of inhibition by different B- CLL cases reflected the results obtained from the experiments using CD4-positive T cell clones. This indicates an individual ability of different B-CLL cases to inhibit normal B cell acti- vation in this system. Inhibition by B-CLL cells depended upon direct contact between competing B-CLL cells with the CD40 ligand trans- fected L cells and normal B lymphocytes. Figure 4 shows that inhibition of CD23 activation in normal B cells was not observed when competing B-CLL cells were separated in transwells. Even the addition of CD40 ligand transfected fibroblasts to B-CLL cells in the transwell to ensure activation of B-CLL cells did not inhibit parallel activation of normal B cells. This argues against a suppressive soluble mediator induced by CD40 ligand-activated B-CLL cells. Figure 3 Some B-CLL cells inhibit CD40 ligand induced CD23 induction of normal B-lymphocytes. CD40 ligand expressing fibro- blasts (3 × 103) were cultured with 4 × 104 normal B cells and increas- ing numbers of PKH26 marked B-CLL cells from patient CLL-2 or −3. B-CLL can interact with CD40 ligand-transfected cells Mean CD23 expression of normal (indicator) B-lymphocytes is shown as percent response of the response without competitor cells. CD23 The experiments using CD40 ligand-transfected L cells indi- expression on B lymphocytes unstimulated: 1.6% positive. After cated that either direct modulation of the CD40L signal by B cocultivation with CD40 ligand expressing L cells: 32.8% positive. cells or B-CLL cells is responsible for the inhibition of parallel B cell activation. Since distinct B-CLL cases (ie CLL-2) repeat- edly failed to show this inhibition, we asked whether these B- CLL cases simply were unable to interact with CD40L trans- fected fibroblasts. All CLL cases showed similar levels of CD40 expression as judged by FACS analysis irrespective of their ability to inhibit normal B cell activation. To address this question further, we analyzed the inducible CD23 expression on the competitor cells themselves stimulated by CD40 ligand transfected L cells. As shown in Figure 5, CD40 ligand trans- fected L cells induced some CD23 expression on normal B cells and B-CLL cells. IL-4 alone was a good inducer of CD23 on normal B cells, whereas IL-4 mediated CD23 expression on B-CLL was usually quite modest.11,26 However, costimu- lation with CD40 ligand and IL-4 resulted in a strong acti- vation of CD23 expression on normal B lymphocytes and B- CLL (Figure 5). These experiments show that B-CLL cells, like normal B cells, are able to receive a stimulating signal by CD40L transfected fibroblasts. In particular, also cells from CLL-2 which fail to inhibit the CD40L signal (Figures 2 and 3) were fully able to receive a CD40L mediated stimulatory sig- Figure 4 Inhibition of CD40 ligand induced CD23 expression on nal. Therefore the potency of inhibiting CD40L cannot be normal B lymphocytes depends on direct cell contact. –᭹–, Normal explained by the inability of certain B-CLL cases to interact × 5 B cells (B; 1 10 /well) with CD40 ligand transfected fibroblasts with CD40L. (LCD40L; 1.5 × 104/well); numbers of B-CLL cells (CLL-3) as indi- cated. –᭺–, Normal B cells (B; 1 × 105/well) with CD40 ligand trans- fected fibroblasts (LCD40L; 1.5 × 104/well); B-CLL cells (CLL-3) separ- ated by semipermeable membrane (transwell). –ᮀ–, Normal B cells Discussion (B; 1 × 105/well) with CD40 ligand transfected fibroblasts (LCD40L; 1.5 × 104/well); B-CLL cells (CLL-3) together with CD40 ligand Chronic lymphocytic are predominantly B cell neo- × 4 transfected fibroblasts (LCD40L; 1.5 10 /well) separated by semi- plasms which are charcterized by the accumulation of slowly permeable membrane (transwell). CD23 expression (MFI) on normal dividing monoclonal CD5-positive B cells.1 The analysis of B lymphocytes was measured after 40 h. cell surface markers and factors favors the assumption that B-CLL cells are representatives of activated T cell-dependent B cell activation. We analyzed if B-CLL cells B cells.27,28 These observations explain only in part the life- can also inhibit the CD40 ligand signal during the activation threatening consequences of the disease such as high fre- of normal B cells. To this end, we activated normal B cells by quencies of hypogammaglobulinemia (60% of the patients), CD40 ligand transfected L cells and titrated increasing num- leading to bacterial infections, or autoimmunity directed bers of B-CLL cells to the cultures. B-CLL cells proved to be against blood cell components (7.7 to 35% of the patients). capable of suppressing the CD40 ligand signal. The CD40 The autoantibodies are in most cases produced by apparently ligand induced CD23 expression on normal B lymphocytes normal B lymphocytes,1,6,29 indicating disturbance of the T cell/B cell interaction in B-CLL C Kneitz et al 102 ecule CD23.24,25 This activation depends on direct cell con- tact and can only partially be substituted for by such as IL-4 (Figure 1). The addition of normal B lymphocytes or increasing numbers of certain B-CLL cells, resulted in a dose- dependent inhibition of T cell contact-mediated B cell stimu- lation (Figure 2). In order to break down the multiple possibilities of B cell/T cell interaction,35 we assessed the role of the CD40/CD40 ligand system as target for the suppressed T cell-dependent B cell activation. Using CD40 ligand-transfected fibroblasts as substitute for activated T cell clones we show that the same B-CLL cells which inhibited the activation of normal B cells by a CD4-positive T cell clone also prevented their activation by CD40 ligand-transfected fibroblasts. This indicates that B- CLL directly interfere with the CD40 ligand-mediated helper signal. This inhibition also required direct cell contact between the inhibitory B-CLL cells and the indicator system of normal B lymphocytes and CD40 ligand expressing fibro- blasts. Separating the inhibitory cells by semipermeable mem- branes shows unimpeded B cell stimulation, suggesting that the observed inhibition is mediated by blocking cell contact- mediated signals, and not because of soluble factors secreted by the B-CLL cells added to the system (Figure 4). Lymphocytes from B-CLL patients showed considerable het- erogeneity in their ability to suppress normal B cell activation. Figure 5 CD40-ligand and interleukin-4 dependent induction of CD23 expression on spleen B lymphocytes and B-CLL cells. Splenic Some even completely failed to block cell contact-mediated B B lymphocytes or B-CLL cells from different patients (1 × 105/well, CLL cell stimulation, even when high numbers of cells were added 1–4) were cultured for 40 h in medium only (B), in the presence of (Figures 2 and 3, Table 1). Eight out of 14 CLL cases showed 3 × 103 control neo-transfected fibroblasts (B+L) or with 3 × 103 CD40 no inhibition, even when used in 25-fold excess (Table 1). The ligand transfected fibroblasts (B+LCD40L). IL-4 (10 ng/ml) was added variability of different B-CLL cases to inhibit CD40L-mediated + + + as indicated (B IL-4, B LCD40L IL-4). CD23 expression on normal B cell activation is highly reproducible even when different B cells or B-CLL cells is given as mean fluorescence intensity (MFI). sources of T cell help were used (CD4 positive T cells vs CD40L transfected fibroblasts). This is illustrated in Figures 2 immune system in CLL patients. The events leading to the and 3 where CLL-3 effectively inhibits normal B cell activation described phenomena are as yet unclear. whereas CLL-2 repeatedly fails to do so. Since eight out of 14 Normal B lymphocytes require help from activated T cells B-CLL cases did not inhibit normal B cell activation, this for antigen-specific activation and differentiation. This is strongly argues against a ‘space problem’ by merely reducing mediated by cytokines such as IL-2, -4 and -5 and signals the probability of specific interactions between the normal involving direct cell contact.30 The most important cell con- ‘indicator’ B cells and the CD40L transfected fibroblasts tact mediated helper signal for B cells seems to be received (Figure 3) or CD4-positive T cell clones (Figure 2, Table 1). via CD40 from CD40 ligand expressing activated T lympho- So far the difference between inhibitory and not inhibitory cytes.14,19,31 The signal directly activates resting B cells B-CLL cases is not clear. The analysis of conventional B-CLL (Figure 5) and enhances the responsiveness to other signals associated cell surface like CD19, CD5, CD23 and such as cytokines.32 It is known that blocking the CD40/CD40 particularly CD40 showed no quantitative differences as ligand interaction in vitro by monoclonal antibodies results in judged by FACS analysis. Furthermore all tested B-CLL impaired B cell responses20 and CD40 ligand-deficient indi- responded to signals by CD40L transfected fibroblasts viduals fail to undergo switching and (Figure 5). Thus the failure of B-CLL-2 to inhibit parallel B cell formation (hyper-IgM syndrome).33 B-CLL cells express similar activation is not caused by the inability of these cells to inter- cell surface receptors and signal transducing molecules as do act with CD40L (Figure 5). We can only speculate that in normal B-lymphocytes.34 These surface receptors display a addition to CD40 other surface molecules, eg adhesion mol- regular function and B-CLL cells respond to external signals ecules, are required for effective competition for T cell help. such as lymphokines or direct cell contact-mediated T cell Antigen-derived B cell differentiation takes place in germi- help in a way similar to normal B lymphocytes (Figures 1 and nal centers of nodes. B lymphocytes get activated in T 5).34 Thus, it seems plausible that B-CLL cells and cells of the cell-rich zones in association with interdigitating cells and T normal immune system communicate with each other in vivo cell help with the CD40L-mediating signal playing a central leading to the observed immune dysregulation in B-CLL role.14,36,37 Disruption of the follicular microenvironment by patients. Therefore we decided to investigate, whether B cell/T B-CLL could disturb the tightly regulated cellular interactions cell interaction in B-CLL is disturbed. Since B-CLL leukemic which normally lead to the development of mature antibody- lymphocytes are found in excess in most compartments of the secreting plasma cells or memory cells. Therefore it can be immune system, we addressed the question whether B-CLL envisaged that inhibition of the CD40 ligand transmitted cells are able to inhibit a T cell contact dependent in vitro helper signal by B-CLL cells, as demonstrated in this paper, model of B cell activation by competing for the cell contact- may be one important mechanism for the observed immune mediated helper signals. suppression in B-CLL patients. Normal B lymphocytes are activated by anti-CD3 stimu- It is important to point out that inhibition of cell contact- lated CD4-positive T cell clones to express the activation mol- mediated T cell help is not specific to B-CLL cells since nor- T cell/B cell interaction in B-CLL C Kneitz et al 103 mal B cells inhibit at least as well in vitro (Figure 2). However 5 Hamblin TJ, Oscier DJ, Young BJ. Autoimmunity in chronic lym- it has to be taken into account that in vivo the immunologic phocytic leukemia. J Clin Pathol 1986; 39: 713–716. organs are highly compartmentalised and normal B cells do 6 Kobayashi R, Picchio G, Kirven M, Meisenholder G, Baird S, Car- son DA, Mosier DE, Kipps TJ. Transfer of human chronic lympho- not regularly enter the regions where a coordinated T cell- cytic leukemia to mice with severe combined immune deficiency. dependent B cell activation takes place. In contrast, in B-CLL Leukemia Res 1992; 16: 1013–1023. we have high B cell numbers which do not observe the regu- 7 Dighiero G. An attempt to explain disordered and hypo- lated structures of lymph nodes. Under these circumstances it gammaglobulinemia in CLL. Nouv Rev Fr Hematol 1988; 30: is conceivable that inhibition of CD40L signals becomes rel- 283–288. evant and leads to the observed immune suppression. Thus 8 Umetsu DT, Essermann L, Donlon TA, DeKryuff RH, Levy R. Induction of proliferation of human follicular (B type) cell number and homing pattern, rather than a specific cellular cells by cognate interaction with CD4+ cell clones. J Immunol signal, determine the ability of B-CLL cells to suppress normal 1990: 144: 2550–2557. immune responses. 9 Duan X, Nerl C, Janssen O, Kabelitz D. B cell maturation in Recently it has been reported that B-CLL cells physically chronic lymphocytic leukemia. IV. T cell-dependent activation of downmodulate CD40L expression on helper T cells.38,39 In leukemic cells by staphylococcal enterotoxin ‘’. these in vitro experiments excess numbers of B-CLL cells were Immunol 1992; 75: 420–426. necessary to downmodulate CD40L expression on T cells. We 10 Gobbi M, Caligaris-Cappio F, Janossy G. Normal equivalent of B cell malignancies: analysis with monoclonal antibodies. Br J can reproduce these results using B-CLL or normal B cells. Haematol 1983; 54: 393–403. Thus, our results extend these studies by demonstrating a func- 11 Tony HP, Lehrnbecher T, Merz H, Sebald W, Wilhelm M. Regu- tional effect of the observed physical downmodulation of lation of IL-4 responsiveness in lymphoma B cells. Leukemia Res CD40L. 1991; 15: 911–919. It seems likely that multiple causes lead to the immune dys- 12 Tony HP, Bru¨ckner A, Wilhelm M. Differential signal requirement regulation in B-CLL cells in vivo. It has been suggested that for upregulation of HLA-class-II molecules in lymphoma B cells. Hematol Pathol 1993; 7: 79–90. B-CLL cells are able to modulate T lymphocyte activation. 13 Hermann P, Van-Kooten C, Gaillard C, Banchereau J, Blanchard Cocultivation of leukemic B cells with T lymphocytes pro- D. CD40 ligand positive CD8 T cell clones allow B cell growth foundly influenced the cytokine profile of the non-malignant and differentiation. Eur J Immunol 1995; 25: 2972–2977. T cell population. Furthermore the secretion of cytokines by 14 Blanchard D, Gaillard C, Hermann P, Banchereau J. Role of CD40 B-CLL cells themselves could also influence parallel specific antigen and interleukin-2 in T cell dependent human B lympho- immune responses.40,41 Nevertheless, the inhibition of T cell/B cyte growth. Eur J Immunol 1994; 24: 330–335. cell interaction by B-CLL described here, seems not to involve 15 Decker T, Flohr T, Trautmann P, Aman MJ, Holter W, Majdic O, Huber C, Peschel C. Role of accessory cells in cytokine production cytokine secretion (Figure 4). by T cells in chronic B cell lymphocytic leukemia. Blood 1995; Hence, there are multiple mechanisms by which B-CLL 86: 1115–1123. cells can downregulate specific immune responses in vivo. 16 Allison JP. CD28- interactions in T cell activation. Curr Opin Competitive inhibition or downmodulation of CD40 ligand- Immunol 1994; 6: 414–419. mediated helper signals, as described in this paper, seems to 17 Durie FH, Foy TM, Masters SR, Lamann JD, Noelle, RJ. The role be one plausible mechanism underlying the impairment of B of CD40 in the regulation of humoral and cell-mediated immunity. Immunol Today 1994; 15: 406–411. cell function during disease progression in patients suffering 18 Banchereau J, Bazan F, Blanchard D, Brie`re F, Galizzi JP, van from B-CLL. Kooten, C, Liu JY, Rousset F, Saeland S. The CD40 antigen and its ligand. Ann Rev Immunol 1994; 12: 881–922. 19 Banchereau J, Rousset F. Growing human B lymphocytes in the CD40 system. Nature 1991; 353: 678–679. Acknowledgements 20 Noelle RJ, Ledbetter JA, Aruffo A. CD40 and its ligand, an essential ligand-receptor pair for -dependent B cell activation. The gift of IL-2 and IL-4 from Prof Dr W Sebald (Wu¨rzburg) Immunol Today 1992; 13: 431–433. and purified ␣CD3 antibody from Prof Dr E Rieber (Dresden, 21 Wohlleben G, Gray D, Schimpl A. In vitro of naive Germany) is very much appreciated. We thank M Clark for mouse B cells: establishment of IgM secreting hybridomas specific for soluble or from B cells cultured on CD40 ligand critical reading of the manuscript. This work was supported transfected mouse fibroblasts. Int Immunol 1996; 8: 343–349. by the Wilhelm-Sander-Stiftung Projekt-No. 92.033.2 and the 22 Caligaris-Cappio F, Janossy G. Surface markers in chronic lympho- Interdisziplina¨res Zentrum fu¨r Klinische Forschung cytic leukemias of B cell type. Semin Hematol 1985; 22: 1–12. (Bundesministerium fu¨r Forschung und Technologie). 23 Kaplan ME, Clark C. An improved rosetting assay for detection of human T lymphocytes. J Immunol Meth 1974; 5: 131–135. 24 Kikutani H, Suemura M, Owaki H, Nakamura H, Sato R, Yamsaki K, Barsumian EL, Hardy RR, Kishimoto T. Fc epsilon receptor, a References specific differentiation marker transiently expressed on mature B cells before isotype switching. J Exp Med 1986; 164: 1455–1469. 1 Dighiero G, Travede P, Chevret S, Fenaux P, Chastang C, Binet 25 Bonnefoy JY, Leconaet-Henchoz S, Aubry JP, Gauchat JF, Graber J-L and the French Cooperative Group on CLL. B-cell chronic P. CD23 and B cell activation. Curr Opin Immunol 1995; 7: leukemia: present status and future directions. Blood 1991; 355–359. 78: 1901–1914. 26 Fournier S, Tran ID, Suter U, Biron G, Delespess G, Sarfati M. The 2 Chapel H, Buch C. Mechanisms of in chronic lympho- in vivo expression of type b CD23 mRNA in B chronic lympho- cytic leukemia. Semin Hematol 1987; 24: 291–296. cytic leukemic cells is associated with an abnormally low CD23 3 De Rossi G, Granati L, Girelli G, Gandolfo G, Arista MC, Martelli upregulation by IL-4: comparison with their normal cellular M, Conti L, Marini R, La Tagliata R, Leone R, Mandelli F. Inci- counterparts. Leukemia Res 1991; 15: 609–618. dence and prognostic significance of autoantibodies against 27 Schuh K, Avots A, Tony HP, Serfling E, Kneitz C. Nuclear NF-ATp erythrocytes and in chronic lymphocytic leukemia (CLL). is a hallmark of unstimulated B cells from B-CLL patients. Leuk Nouv Rev Fr Hematol 1988; 30: 403–406. Lymphoma 1996; 23: 583–592. 4 Hansen MM. Chronic lymphocytic leukemia: clinical studies 28 Sarfati M. and the two isoforms of CD23 antigen: based on 189 cases followed for a long time. Scand J Haematol their possible contribution to the physiopathology of chronic lym- 1973; 18: 1–286. phocytic disease. Nouv Rev Fr Hematol 1993; 35: 109–113. T cell/B cell interaction in B-CLL C Kneitz et al 104 29 Kipps TJ, Carson DA. Autoantibodies in chronic lymphocytic leu- 36 MacLennan ICM. Germinal center. Annu Rev Immunol 1994; 12: kemia and related systemic . Blood 1993; 81: 17–39. 2475–2487. 37 Holder MJ, Wang H, Milner AE, Casamayor M, Armitage R, Sprigs 30 Clark EA, Ledbetter JA. How B and T cells talk to each other. MK, Fanslow WC, MacLennan ICM, Gregory CD, Gordon J. Sup- Nature 1994; 367: 425–428. pression of in normal and neoplastic human B lympho- 31 Nishioka Y, Lipsky PE. The role of CD40–CD40 ligand interaction cytes. Eur J Immunol 1993; 23: 2368–2371. in human T cell/B-cell collaboration. J Immunol 1994; 153: 38 Yellin MJ, Sippel K, Inghirami G, Covey LR, Lee JJ, Sinning J, Clark 1027–1036. EA, Chess L, Lederman S. CD40 molecules induce down-modu- 32 Spriggs MK, Armitage RJ, Strockbine L, Clifford KN, MacDuff KM, lation and endocytosis of T cell surface T cell–B cell activating Sato TA, Maliszewski CR, Fanslow WC. Recombinant human molecule/CD40-L. J Immunol 1994; 152: 598–608. CD40 ligand stimulates B cell proliferation and immunoglobulin 39 Cantwell M, Hua T, Pappas J, Kipps TJ. Acquired CD40-ligand E secretion. J Exp Med 1992; 176: 1543–1550. deficiency in chronic lymphocytic leukemia. Nature Med 1997; 33 Korthauer U, Graf D, Mages HW, Briere E, Padayachee M, Mal- 3, 984–989. colm S, Ugazio AG, Notarangelo LD, Levinsky RJ, Kroczek RA. 40 Biondi A, Rossi V, Bassan R, Barbui T, Bettoni S, Sironi M, Manto- Defective expression of T-cell CD40 ligand causes X-linked vani A, Rambaldi A. Constitutive expression of the interleukin-6 immunodeficiency with hyper-IgM (see comments). Nature 1993; gene in chronic lymphocytic leukemia. Blood 1989; 73: 1279– 361: 539–541. 1284. 34 Fu SM, Chiorazzi N, Kunkel HG. Differentiation capacity and other properties of the leukemic cells of chronic lymphocytic leu- 41 Foa R, Massaia M, Cardona S, Tos AG, Bianchi A, Attisano C, kemia. Immunol Rev 1979; 48: 23–44. Guarini A, di Celle PF, Fierro MT. Production of tumor 35 Van Kooten C, Gaillard C, Galizzi J-P, Hermann P, Fossiez F, Ban- factor alpha by B-cell chronic lymphocytic leukemia cells: a poss- chereau J, Blanchard D. B cells regulate expression of CD40 ible role of TNF in the progression of the disease. Blood 1990; ligand on activated T cells by lowering the mRNA level and 76: 393–400. through the release of soluble CD40. Eur J Immunol 1994; 24: 787–792.