sign in sign up
<<
Home , Cytokine, Cytokine receptor

Leukemia (1998) 12, 1221–1229  1998 Stockton Press All rights reserved 0887-6924/98 $12.00 http://www.stockton-press.co.uk/leu Expression and regulation of c- and response to in childhood malignant T-lymphoblastic cells J Tomeczkowski1, D Frick1, B Schwinzer1, N Wittner1, W-D Ludwig2, A Reiter1, K Welte1 and K-W Sykora1

1Department of Pediatric Hematology and Oncology, Medical School Hannover; 2Department Hematology and Oncology, Virchow Klinikum, Robert Ro¨ssle Klinik, Free University Berlin, Germany

The stem cell factor (SCF) synergizes with IL-7 to that SCF influences the development of T and B cells. Mature enhance the proliferation of . We therefore investi- human peripheral blood express little or no c-kit gated the role of the SCF receptor, the protooncogene c-kit, 14 in the pathogenesis of pediatric T-lineage malignancies. receptors. In combination with IL-7, SCF was found to be a Expression and regulation of c-kit in cells from children with potent stimulus at early stages of T and B lymphoid develop- non-Hodgkin’s lymphoma (T-NHL) or acute lymphoblastic leu- ment.7,15–22 In mice, proliferation occurred in a popu- kemia (T-ALL) and the proliferative effect of SCF on these cells lation of IL2R+/CD4−/CD8−/CD3− thymocytes by the combi- were examined in seven cell lines and 21 biopsy tumor cell nation of IL-7 and SCF.20 A high level of c-kit expression preparations. Inducibility of c-kit receptors by SCF, IL-1␤, IL- ␤ ␣ defines pro-T cells which have not yet rearranged their TCR 2, IL-7, TGF- , TNF- , PMA or calcium ionophore A23187 was 23 − − + + studied by flow cytometry (FCM). C-kit receptors were detected genes. The proliferation of CD4loCD3 CD8 /c-kit /CD25 in three out of seven T-lymphoblastic cell lines and in nine out thymocytes was completely inhibited in the presence of anti- of 21 biopsy tumor cell preparations. Upregulation of c-kit c-kit, which indicated an important role for c-kit/SCF interac- could be induced by cultivation, and to a higher extent by culti- tions during early development.19 C-kit+ cells vation and addition of IL-1␤, TNF-␣, TGF-␤ or A23187. Down- derived from marrow differentiated into CD4CD8 dou- regulation of c-kit occurred in the presence of SCF or PMA. ble-positive cells, and did so approximately 4 days after c-kit+ SCF caused a downregulation of c-kit receptors in eight of nine, and a proliferative response in three of 11 c-kit-positive T- cells derived from did. In the same study, anti-c-kit blocked T cell generation of c-kit+ bone marrow lymphoblastic cell preparations. We conclude that c-kit is able + to transduce a growth stimulatory signal in some T-lympho- cells, but not of c-kit thymus cells.22 blastic cells and that its expression may not be detectable in The growth conditions of T cell lymphoma or leukemia and a resting metabolic or proliferative state. their dependence on growth factors are of interest, because Keywords: c-kit protooncogene; stem cell factor; childhood; acute are being added to protocols for lymphoblastic leukemia; non-Hodgkin’s lymphoma; hematopoiesis treatment of patients with these conditions. The cytokine SCF was tested in phase I clinical trials as an adjunct in the therapy of malignant disease.24–28 Because of its stimulatory effects on Introduction early stem cells, SCF may become a candidate to support not only , but also thrombopoiesis in the setting of Bone marrow precursors are frequent targets of high-dose chemotherapy and may gain significance in bone malignant transformation. The heterogeneous groups of non- marrow transplantation like other widely used cytokines.29,30 Hodgkin’s lymphoma and acute lymphoblastic leukemia can SCF appears to be a potent mobilizer of peripheral blood stem be considered the malignant counterparts of different stages cells (PBSCs),31–34 especially in conjunction with G-CSF.6,29,35 of B and T cell differentiation. These malignant cells originate Little is known about a direct stimulatory or inhibitory effect from lineage-committed hematopoietic progenitor cells and of SCF on childhood malignant T-lymphoblastic cells, from thymocytes which express different markers including although the expression of c-kit in T-lymphoblastic cells has the product of the c-kit protooncogene. This protooncogene been described previously.36–39 The finding that c-kit is encodes the specific surface receptor for the pluripotent hem- involved in the growth regulation of thymocytes makes this 1–3 atopoietic colony-stimulating factor SCF. SCF acts syner- receptor a candidate for involvement in the growth and devel- gistically with other growth factors, including opment of T-lineage leukemia and lymphoma cells. Because (EPO), G-CSF, (M)-CSF, GM-CSF, IL-3, IL-6, IL- of the potential clinical use of SCF, we examined the 4–7 11 and IL-12, and plays an important role in normal and expression and regulation of c-kit receptor on pediatric malig- leukemic hematopoiesis. Mature lymphocytes exhibit little or nant T-lymphoblastic cells and the direct effect of SCF on 1 no responsiveness to the c-kit , SCF. In malignant hem- these cells. atopoietic cells, c-kit is expressed by cells of the erythroid, megakaryocytic and lineages.1 It is also expressed 8 in most cases of acute myeloblastic leukemia (AML) and in Materials and methods some cases of chronic myelogenous leukemia (CML) in blast crisis. By contrast, little or no expression of c-kit has been Cell lines and fresh lymphoblastic cells observed in human leukemia cell lines of the lymphoid 9–11 lineage and in blast cells of acute lymphoblastic leuke- The human T-lymphoblastic cell lines MOLT-3, MOLT-4, Jur- 11–13 mia, although several lines of evidence have indicated kat, CCRF-HSB-2 and CCRF-CEM were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and the cell lines MOLT-16 and SUP-T140 from the Correspondence: K-Walter Sykora, Medical School Hannover, German Collection of Microorganisms and Cells (DSMZ, Department of Pediatric Hematology and Oncology IV, OE 6780, Chairman Prof Dr med Dr hc H Riehm, Carl-Neuberg Straße 1, D- Braunschweig, Germany). 30625 Hannover, Germany Fresh lymphoblastic cells from 21 patients were isolated by Received 25 June 1997; accepted 27 April 1998 density gradient centrifugation (Ficoll–Paque; Pharmacia, C-kit expression in lymphoblastic cells J Tomeczkowski et al 1222 Uppsala, Sweden), washed, and stored in liquid nitrogen with MO, USA), 100 ng/ml A 23187 (Sigma), 10 ng/ml TGF-␤ medium containing 10% dimethylsulfoxide (DMSO) and 20% (Amgen), 1 ng/ml IL-1␤ (Amgen), 10 ng/ml IL-2 (Cetus, fetal calf serum (FCS) until use. Thirteen of the 21 freshly iso- Emeryville, CA, USA), 10 ng/ml IL-7 (Promega), 5 ng/ml TNF- lated tumor cell preparations were obtained from pleural ␣ (Knoll, Ludwigshafen, Germany). effusion, five from tumor or lymph node, and one each from peripheral blood and bone marrow (not expressing the CD117 ). The age at diagnosis ranged from 3 to 15 years. Four of the 21 patients were female and 17 were male. Labeling of SCF and chemical crosslinking with After thawing, the cells were washed, resuspended in digoxigenin-rhSCF (Dig-rhSCF) medium and cultured at 37°C for 1 day. Two of the fresh T- ALL cell preparations isolated from peripheral blood could be One milliliter rhSCF (2.2 mg/ml) in buffer pH 8.5 was labeled cultured over a time period of more than 6 months and were + + by incubation at room temperature for 2 h with digoxigenin- termed MHH-TALL-1 (patient 17: CD3 , CD34 ) and MHH- + − 3-0-methyl-carbonyl-⑀-aminocapronsa¨ure-N-hydroxy-succini- TALL-2 (Patient 4: CD3 , CD34 ). These cells were maintained midester (Dig) (Boehringer Mannheim) dissolved in 150 ␮l at 37°C in humidified 5% CO with weekly replacement of 2 ethanol as described elsewhere.43 Free Dig was removed by half of the medium. All cell lines were cultured in medium a desalting column (Sephadex G-25). MO7e and MHH-TALL- consisting of RPMI 1640 (Gibco BRL, Eggenstein, Germany) 1 cells, 5 × 106 cells in 100 ␮l, were incubated with 15 ␮l containing 10% FCS, and 0.25 ␮g/ml antibiotic-antimycotic (0.13 mg/ml) Dig-rhSCF for 1 h on ice in the presence or solution (Sigma, Deisenhofen, Germany). FCS was tested to absence of a 100-fold excess of unlabeled rhSCF. Cells were support optimal growth of lymphoid cell lines. The mega- washed two times to remove unbound SCF. Subsequently, the karyoblastic cell line MO7e41 was used as a positive control cells were treated with 0.5 mmol/l of the bifunctional for c-kit expression. crosslinker disuccinimidylsuberate (DSS). The cells were lysed and the soluble were denatured and separated in a 7.5% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE). 3H-thymidine incorporation assays The gel was blotted on to a nylon membrane and DIGrh- SCF/c-kit complexes were detected using anti-Dig-AP-conju- The mitogenic activity of rhSCF was determined under serum- gated and the chemiluminescent substrate AMPPD. free conditions in a 3H-thymidine incorporation as Membranes were exposed to Cronex DDS 100 autoradiogra- described elsewhere.42 Significance of stimulation was ana- phy films (Sterling, Newark, DE, USA). lyzed by the paired Student’s test by comparing the treated vs the untreated cells.

Flow cytometry (FCM) MTS metabolism assay

T-ALL and T-NHL cells were washed twice with serum-free Fc receptors were blocked by human IgG (Gammagard; medium containing RPMI 1640 supplemented with Iscove’s Baxter, Unterschleissheim, Germany). Cell samples were supplement (Gibco BRL) and -transferrin-sodium selen- incubated for 20 min on ice with saturating amounts of mono- ite supplement (Boehringer, Mannheim, Germany) to a final clonal antibody (MoAb) in PBS, 0.5% BSA, 0.1% sodium concentration of 20 ␮g/ml soy bean lipids, 1 mg/ml bovine azide and were then washed twice in PBS, 0.1% BSA, 0.1% serum albumin fraction V, 21 ␮g/ml transferrin, 1 ␮g/ml insu- sodium azide. The MoAbs reacted with fluorescein isothiocy- Ј lin, 1 ng/ml sodium selenite. Cells were then plated in 96- anate (FITC)-conjugated F(ab )2 rabbit anti-mouse immuno- well flat-bottom microtiter plates at 5 × 104 cells/ml in 200 ␮l globulins (Dako, Hamburg, Germany) or by phycoerythrin Ј serum-free medium per well containing various concen- (PE)-conjugated F(ab )2 goat anti-mouse fragments (Dako) trations (0, 50, 500 ng/ml) of rhSCF and 1 ␮l phenazine used as a second-step reagent. In addition, freshly isolated methosulfate (PMS) and 20 ␮l Owen’s reagent (MTS) cells identified to express high amounts of c-kit were analyzed (Promega Celltiter AQ, Madison, WI, USA). The absorbance by double immunofluorescence. In this procedure, cells were of the formazan at 490 nm was measured using an ELISA plate stained by antibodies against CD117, CD10 or CD34 which Ј reader immediately after assay preparation. The plates were were then detected by PE-conjugated F(ab )2 fragments. An then incubated at 37°C and further color development was excess of irrelevant mouse Ig was used to block free antigen determined during the following days. Significance of stimu- binding sites on the anti-mouse Ig-PE. Fc receptors were lation was tested by the paired Student’s test by comparing blocked by human IgG (Gammagard; Baxter). All samples the treated vs the untreated cells. were analyzed using a FACScan flow cytometer (Becton Dickinson, San Jose, CA, USA). Anti-CD117 (c-kit) MoAbs, supernatant from clone 57A5D8B1, were kindly provided by Stimulation of T-ALL and T-NHL cells Dr H-J Bu¨hring (Tu¨bingen, Germany). The other MoAbs were from different sources: anti-CD3 (UCHT1) FITC-conjugated To assess induced expression of c-kit and to demonstrate c- (DAKO), anti-CD4 (Leu-3A), anti-CD8 (Leu-2A), anti-CD10 kit regulation, the appropriate cytokine or reagent was added (anti-cALLA) (Becton Dickinson), anti-CD19 (Leu-12) FITC- to the cells (5 × 105 cells/ml) in 25 cm2 tissue culture flasks conjugated (Becton Dickinson) and CD34 (anti-HPCA-2) (Becton Dickinson, Meylan, France) containing 10 ml of cul- (Becton Dickinson). ture medium with 10% FCS. The cells were cultured at 37°C Results of FCM are given as mean ratio, calculated as the

in humidified CO2. The final concentrations of the cytokines mean relative immunofluorescence using the specific anti- or reagents and their suppliers were 100 ng/ml rhSCF (Amgen, body, divided by the mean relative immunofluorescence Thousand Oaks, CA, USA), 20 ng/ml PMA (Sigma, St Louis, obtained with an irrelevant, matched control antibody. C-kit expression in lymphoblastic cells J Tomeczkowski et al 1223 Results

C-kit receptor expression and its regulation on T-ALL and T-NHL cells

C-kit receptors were expressed on three out of seven T-lym- phoblastic cell lines (Molt-4, SUPT-1 and CCRF-HSB-2; Table 1) and on nine of 21 freshly isolated T-lymphoblastic cells derived from children with T cell leukemia and lymphoma (Table 1, and Figures 1 and 2). In addition, c-kit was upregulated within 48 h after addition of serum-containing medium to the freshly isolated T-NHL cells from patient 6 (Figure 2) and to the 1–2 month culture cells from patient 4 (MHH-TALL-2; Figure 3). Freshly thawed and cultured cells, which were maintained in serum-free medium, or were main- tained in serum-containing medium but not given fresh medium for at least 5 days, expressed c-kit protein only weakly or not at all. A significant c-kit expression was induced 20 h after addition of fresh medium (Figure 3). After that time- point, c-kit expression again declined until fresh medium was added. Since upregulation of c-kit by the addition of serum- containing medium was observed in cell culture, we asked

Table 1 Expression of CD3, CD10, CD34, CD117 on freshly iso- lated malignant T-lymphoblastic cells and malignant T-lymphoblastic cell lines

Freshly isolated Mean ratio cells (Patient No.) CD3 CD10 CD34 CD117

1 1.3 1 1 54.7 2 1 1 1 4.2 3 1 13.9 30.2 1.2 4 134.4 10.8 1 23.4 5 1 1 11.8 28.8 6 3.1 10.6 1 65.7 7 1.3 3.9 4.6 1.4 8 4.2 1 1 1 9 2.3 NT NT 1 10* 100 2 1 1 11 18.1 8 1 1 12* 12.5 18.5 216 1.1 13 4.3 28.7 1 1 14* 5.8 25.5 3.8 1.1 15* 4.2 7.1 1 1 16* 3.6 4.5 1 1 17 11.3 31.2 157.9 13.3 18* 14.1 20.7 42.5 1.2 19* NT 1.4 NT 2.2 20* NT 2.7 NT 1 Figure 1 C-kit (CD117) expression of freshly isolated T-lympho- 21* 110 53 1 6 blastic cells from patient 17 (a) and patient 4 (b). Cells were thawed Molt-3 1.6 1 NT 1 and cultured for 48 h before staining with mouse anti-CD117, anti- Molt-4 2.4 1.3 141 3.1 CD10, or anti-CD34 antibodies that were detected with PE-conju- Molt-16 9.9 1 NT 1 gated F(abЈ) fragments and counterstained with a FITC-conjugated Jurkat 18.3 1 NT 1 2 monoclonal anti-CD3 antibody. SUPT-1 4.4 1 1 2.4 CCRF-HSB-2 1.2 323.6 1 2.8 CCRF-CEM 1.6 14.1 72.8 1 whether other cytokines or signals were also able to upregul- Cells indicated by asterisks (*) were measured by single immuno- ate c-kit expression. We treated the cells with cytokines likely fluorescence (IF) of the respective . All other cells were to be present in vivo and with second messenger reagents measured by double IF of CD117 vs either CD3, CD10 or CD34. simulating other cytokine and microenvironmental influences. The mean ratios of these antigens are given for the CD117-positive Figures 4 and 5 show the influence of the cytokines IL-1␤, IL- population only. All cells were found to express CD4 and/or CD8, 2, IL-7, TGF-␤, SCF and TNF-␣ and of the second messenger except patient 8 who was CD4 and CD8 negative, and patient 16 who was not tested (data not shown). A mean ratio of 1 signifies reagents PMA and A23187 on the expression of c-kit receptor no detectable antigen expression, mean ratios of Ͼ1.2 were con- on MHH-TALL-2 cells. CD3 and CD10 antigens were meas- sidered positive. ured as estimate of differentiation along the T-lineage under NT, not tested. influence of these cytokines or reagents (data for CD10 not C-kit expression in lymphoblastic cells J Tomeczkowski et al 1224

Figure 2 Upregulation of c-kit (CD117) expression by serum containing medium or by calcium ionophore in T-NHL cells from patient 6. Cells were thawed and analyzed by FCM directly (left), after 48 h in culture (middle) and after 48 h in the presence of A23187 (right) (treatment Ј of the cells with A23187 caused cell death). Cells were stained with a mouse anti-CD117 antibody that was detected with PE-conjugated F(ab )2 fragments and counterstained with a FITC-conjugated monoclonal anti-CD3 antibody.

Figure 3 Time course of c-kit induction on MHH-TALL-2 cells. MHH-TALL-2 cells were maintained in serum-containing medium for 5 days without passaging and were diluted with one volume of fresh medium and immunofluorescence of CD117, CD3 and CD10 was measured after culture periods of 6, 20 and 48 h. The values are given as mean ratios of immunofluorescence.

shown). Also to detect influences other than antigen expression, we estimated size of the cells by flow cytometric methods. Figure 4 summarizes the effects of the reagents on antigen expression after 48 h, and Figure 5 gives the time course of the expression. These results could be confirmed during repeated experiments and the same regulation occurred in the cell line MO7e with SCF, PMA and A23187 (data not shown). Figure 4 Expression of CD117, CD3 and change in and cell size Upregulation of c-kit was not seen after addition of its own of MHH-TALL-2 after treatment with various reagents and cytokines. ligand, SCF, which persisted for up to 48 h (Figure 5). PMA Cells were cultured for 48 h in medium containing the indicated cyto- kines or reagents and mean ratios of fluorescence were determined also downregulated c-kit expression and had the same down- by FCM. The values for antigen expression were plotted as the per- regulatory effect on CD3, but upregulated CD10. The calcium centage increase or decrease of mean ratio in the presence or absence ionophore A23187 and the cytokine TNF-␣ both markedly of the respective reagent. Untreated cells were set to 100%. upregulated c-kit expression, and had different regulatory effects on the CD3 and CD10 antigens used as controls. The upregulation of c-kit by these two reagents was not complete 2 had only minor influence on the expression of the c-kit, at 20 h but continued to increase up to the 48 h investigated. CD3 and CD10 antigens. The cytokines IL-7 and IL1-␤ Upregulation of c-kit by A23187 but not by TNF-␣ was also had little influence on the T-lymphoblastic cells in our seen in the MO7e cells used as controls (data not shown). IL- experiments. C-kit expression in lymphoblastic cells J Tomeczkowski et al 1225

Figure 5 Time course of surface marker expression of MHH-TALL-2 cells after treatment with various reagents and cytokines. Cells were cultured for 6, 20 or 24 and 48 h in the absence or presence of SCF, PMA, A23187, IL-7 (a) and of IL-1␤, TNF-␣, PMA or TGF-␤ (b), respectively, and mean ratios of fluorescence of c-kit and CD3 were estimated in the viable cell population gated (104 cells/gate) in a forward scatter vs sideward scatter FCM analysis. (a) and (b) represent two different experiments. The immunofluorescence values are plotted as mean ratio.

resulted in a reduced capacity to bind anti-c-kit antibodies (Table 2). (Different levels of c-kit expression compared to lev- els given in Table 1 are due to different culture conditions.) Time course experiments in SUP-T1 cells showed that about half of the c-kit receptors are already downmodulated after an incubation period of 10 min at 37°C in the presence of rhSCF. At 4°C, no downmodulation of c-kit receptors occurred in SUP-T1 cells during 10 min incubation in the presence of rhSCF (data not shown). These experiments indicated that Figure 6 Chemical crosslinking of Dig-labeled SCF to c-kit recep- competition between ligand and antibody was not the reason tors on MHH-TALL-1 cells. Dig-labeled SCF was crosslinked to T- for decreased receptor detection by antibody. ALL cells and the cellular proteins were separated after lysis by SDS- polyacrylamide gel electrophoreses and blotted on to a nylon mem- brane. The complex consisting of Dig-labeled SCF and c-kit receptors was detected by chemiluminescence (arrow). Lane 1, MO7e cells; Determination of the molecular weight of the c-kit lane 2, MO7e cells in the presence of a 100-fold molar excess of receptor on MHH-TALL-1 cells unlabeled SCF; lane 3, MHH-TALL-1 cells; lane 4, MHH-TALL-1 cells in the presence of a 100-fold molar excess of unlabeled SCF. Molecu- lar weights are given in kDa. To demonstrate SCF binding and to determine the molecular weight of the SCF receptor on a T-ALL cell line we crosslinked digoxigenin-labeled rhSCF to the cell surface of the MHH- Downmodulation of c-kit receptors by SCF on TALL-1 and to the positive control MO7e cell line, respect- T-lymphoblastic cells ively. A cytokine–receptor complex of 165–170 kDa, corre- sponding to the expected molecular weight of the c-kit recep- In all except one of the c-kit-positive T-lymphoblastic cells, tor of approximately 150 kDa was found in MHH-TALL-1 cells preincubation with rhSCF for 3–6 h before FCM analysis and in MO7e cells. Specific displacement of labeled SCF from C-kit expression in lymphoblastic cells J Tomeczkowski et al 1226 Table 2 Downmodulation of c-kit by SCF biopsy tumor cells. We could demonstrate c-kit protein expression by flow cytometry or receptor crosslinking on Cell line (name) or CD117 mean ratio freshly isolated blast cells from nine of 21 patients with T-ALL freshly isolated cells or T-NHL. In addition, three of seven malignant childhood T- (Patient No.) Without rhSCF With rhSCF lymphoblastic cell lines expressed the c-kit protein. In contrast to most cases of myeloid leukemia, little or no Molt-4 3.1 1.8 expression of c-kit had been observed in human leukemia cell SUPT-1 4.8 2.9 lines of lymphoid lineage9–11 and in blast cells of acute lym- CCRF-HSB-2 3 1.9 11–13 1 10.1 4.7 phoblastic leukemic so far. However, an IL-2-dependent 2 1.8 1.4 T-lymphoblastic leukemia cell line that was stimulated to 4 8.5 4.0 secrete GM-CSF by SCF,37 and a case of T-ALL blast cells pro- 5 32.2 12.8 ducing small amounts of c-kit mRNA have been described.36 17 6.0 4.2 More recently, our results38,44 have been confirmed by Kees 21 1.8 1.7 and Ashman45 who found c-kit expression in three of five T- ALL cell lines, and SCF had an synergistic proliferative effect Cells were incubated for 48 h in the presence or absence of SCF with IL-2 on one of these cells. Sperling et al39 discovered CD and c-kit (CD117) expression was determined by flow cytometry. 117 in a small proportion of T-lineage ALL (9%) and Ichino- Results of FCM are given as mean ratio, calculated as the mean 46 relative immunofluorescence using the specific antibody, divided hasama et al found a c-kit expressing T-lymphoblastic lym- by the mean relative immunofluorescence obtained with an irrel- phoma. C-kit expression appears to be a rare event in lym- evant, isotype-matched control antibody. phoblastic leukemia of B-lineage, however.39 Valverde et al,47 found only one of 23 ALL cells of different types positive for c-kit expression and Morita et al48 found no c-kit expression c-kit receptor could be demonstrated by addition of a 100- in 44 lymphoid blast preparations from pediatric leukemia fold molar excess of unlabeled SCF (Figure 6). C-kit receptors patients and 20 cell lines of erythroid and lymphoid lineages. could not be demonstrated on freshly thawed MHH-TALL-1 Whether c-kit is biologically active by transducing a growth cells. signal in the T-lymphoblastic cells was not studied by these investigators. With our experiments we could demonstrate downmodul- Mitogenic activity of rhSCF on freshly isolated ation of the receptor in the presence of rhSCF in all except one T-lymphoblastic cells and cell lines of the c-kit-positive T-lymphoblastic cells tested. Only one of the permanent cell lines, but three of nine freshly isolated The mitogenic activity of rhSCF was tested by MTS metab- blast cell preparations proliferated in the presence of SCF. olism and 3H-thymidine incorporation on three c-kit-positive That most investigators did not detect expression of c-kit in cell lines, and on seven c-kit-positive fresh biopsy cell prep- malignant cells of the lymphoid lineage may have been due arations. MO7e megakaryoblastic cells again served as posi- to the fact that the cells were examined in a resting metabolic tive controls. A significant dose-dependent proliferative state, eg directly after thawing of tissue material for the initial response to rhSCF was detected in three out of 21 freshly iso- diagnostic workup. We made similar observations, but cryo- lated T-lymphoblastic cell preparations. Positive were patients preservation might have damaged the cell surface proteins, 4 (whose cells later became the MHH-TALL-2 cell line), 5 and have changed the protein conformation or have resulted in 17 (later the MHH-TALL-1 cell line) and the MO7e cells shedding of the receptors, making impossible to detect the (Figures 7 and 8). Dose-dependent proliferation is shown in receptor by flow cytometry. However, since upregulation of the range of 1–50 ng/ml rhSCF for patient 17 (fresh cells) c-kit occurred after adding fresh medium to the cells, changes (Figure 8). However, all cells tested were already maximally in their metabolic state seem to be the more likely reason for stimulated by 50 ng/ml rhSCF and no dose–response occurred this observation. Metabolic state or cell cycle activity were despite a 10-fold increase in SCF concentration. One cell line previously reported to have marked effects on c-kit (MHH-TALL-2) was tested in a 3H-thymidine proliferation expression. C-kit expression was upregulated by the c-myb assay side by side with the cryopreserved fresh tumor cells , which plays an important role for pro- (patient 4) from which they were derived (Figure 8). The pro- gression through the cell cycle. Expression of c-kit was essen- liferative response to SCF was similar in both preparations, tially abrogated by arresting the growth of human bone mar- indicating that the SCF responsiveness was not merely row mononuclear cells or by inhibiting myb gene function.49 acquired or lost during prolonged tissue culture. In addition, Consistent with these results, we found that c-kit protein was the three adult T-lymphoblastic cell lines Hut78, Hut102 and only minimally expressed in freshly thawed biopsy T-ALL and MT-1 were investigated. None of these cell lines proliferated T-NHL cells, but was upregulated within 48 h after the in response to SCF and only MT-1 showed c-kit expression addition of fresh medium. and c-kit downmodulation upon treatment with rhSCF (data Leukemic cells in vivo are exposed to many factors and not shown). cytokines that could theoretically induce or influence c-kit receptor expression. Our results, using the T-lymphoblastic cell line MHH-TALL-2, showed that the most potent regulators Discussion in vitro were A23187 and TNF-␣ as upregulators, and PMA and SCF as downregulators of c-kit expression. In our study, The aim of this study was to determine the role of c-kit in TGF-␤ led to an upregulation in T-lymphoblastic cells but to childhood T-NHL and T-ALL cells. Because of the potential a downregulation in MO7e cells, used as a control. TGF-␤ clinical use of SCF in patients undergoing autologous stem treatment of progenitor cell lines and murine primary progeni- cell mobilization,24–29 we investigated c-kit expression in tor cells, which constitutively express functional cell-surface childhood T-lymphoblastic cell lines and in freshly isolated c-kit, resulted in downregulation of cell-surface c-kit C-kit expression in lymphoblastic cells J Tomeczkowski et al 1227

Figure 7 Proliferative effect of rhSCF on fresh cells and cell lines from patients with T-ALL and T-NHL (MTS reduction) (data from all other negative cells and cell lines not shown). T-lymphoblastic cells were washed, resuspended in serum-free medium and added to the wells of a 96-well plate containing various concentrations of rhSCF (0 = ć,50= b, 500 = ࡯ ng/ml) and the combined MTS/PMS solution. Plates were incubated at 37°C and absorbance at 490 nm was recorded using an ELISA plate reader at the indicated times. Absorbance of MTS/PMS in the absence of cells was subtracted at each timepoint. Data are given as means of triplicates with standard deviation. Significance were tested with Student’s t-test for treated vs untreated cells (*P Ͻ 0.001). expression by a decrease of c-kit message stability starting progenitor cells and in the myelodysplastic and myeloprolifer- after 2 h of TGF-␤ treatment.50 PMA treatment in our study ative disorders53 than in lymphoid cells. led to a downregulation of c-kit on the protein level. The con- We conclude that: (1) c-kit can be expressed on malignant comitant upregulation of the CD10 antigen suggested that T-lymphoblastic cells derived from children with T cell lym- downregulation of c-kit was not due to toxic effects of the phoma or leukemia; (2) that the characteristics of the c-kit pro- PMA treatment. The T cell and thymocyte growth factors IL- tein are similar to the ones described for other cell types. 2 or IL-7 did not significantly influence the expression of c- These characteristics include molecular weight, specific regu- kit. Receptor downmodulation by SCF was seen in eight of lation by different reagents and cytokines, and susceptibility our nine c-kit-positive T-lymphoblastic cell preparations to downmodulation by the receptor’s own ligand SCF; (3) c- resembling the observation of Pietsch et al41 in AML cells. The kit is biologically active by transducing a growth signal in mechanism of downmodulation may be the one described by some of the T-lymphoblastic cells; and (4) that c-kit expression Yee et al51 in mast cells, where SCF was rapidly internalized may not be evident on freshly thawed malignant cells which and receptor degradation was accelerated after binding of SCF are usually used for diagnostic immunophenotyping, but may to its receptor. Our results show that c-kit expression can be only become apparent after brief periods of culture or after influenced by cytokines and other factors. Therefore, c-kit stimulation with calcium ionophore. expression in vivo might be different to the in vitro situation. Mobilization and peripheral stem cell expansion in the This has to be considered when c-kit expression is judged presence of SCF is entering clinical practice. The fact that c- by in vitro experiments, especially in the absence of those kit expression may not be apparent on fresh malignant T cells cytokines. at the initial diagnostic workup, and the possibility of a pro- In addition, our results show that c-kit expression is not liferative effect of SCF on these cells have to be especially restricted to immature, CD34+ bone marrow precursors cells52 considered. Ex vivo expansion of bone marrow cells by SCF and myeloid leukemic cells,13 although c-kit expression is cer- might also lead to an expansion of contaminating malignant tainly much more common in normal early hematopoietic cells. When SCF is used as an adjunct to chemotherapy proto- C-kit expression in lymphoblastic cells J Tomeczkowski et al 1228

Figure 8 Proliferative effect of rhSCF on T-ALL cells (3H-thymidine incorporation). Fresh cells and the cell line (MHH-TALL-2) derived from patient 4, fresh cells from patient 17 (MHH-TALL-1) and control Mo7e cells were maintained for 72 h in serum-free medium in the presence of different amounts of rhSCF; 3H-thymidine was added after 72 h of incubation and cells were harvested 4 h later for liquid scintillation counting. Data are given as means of triplicates with standard deviation.

CD34+lin− cells, and the generation of colony-forming cell pro- cols, a direct negative influence of treatment with SCF on chil- + − dren suffering from T-lymphoblastic malignancies appears to geny from CD34 lin cells cultured with -3, be a possibility. Whether SCF has a direct influence on the colony-stimulating factor, or granulocyte–macrophage colony- stimulating factor. Blood 1991; 77: 2316–2321. growth of leukemia or lymphoma cells in vivo resulting in a 7 McNiece IK, Langley KE, Zsebo KM. The role of recombinant stem clinically relevant change in the outcome of patients can only cell factor in early development. Synergistic interaction with be answered by clinical studies. IL-7. J Immunol 1991; 146: 3785–3790. 8 Ikeda H, Kanakura Y, Tamaki T, Kuriu A, Kitayama H, Ishikawa J, Kanayama Y, Yonezawa T, Tarui S, Griffin JD. Expression and Acknowledgements functional role of the proto-oncogene c-kit in acute myeloblastic leukemia cells. Blood 1991; 78: 2962–2968. 9 Hu ZB, Ma W, Uphoff CC, Quentmeier H, Drexler HG. c-kit This work was supported by Deutsche Krebshilfe grant expression in human megakaryoblastic leukemia cell lines. Blood W88/91/Sy1. We thank Dr L Souza (Amgen, USA) for supply 1994; 83: 2133–2144. of cytokines, Dr H-J Bu¨hring (Tu¨bingen) for supply of CD117 10 Andre C, d’Auriol L, Lacombe C, Gisselbrecht S, Galibert F. C-kit MoAb, and Prof H Riehm for support. mRNA expression in human and murine hematopoietic cell lines. Oncogene 1989; 4: 1047–1049. 11 Kanakura Y, Ikeda H, Kitayama H, Sugahara H, Furitsu T. References Expression, function and of the proto-oncogene c-kit product in human leukemia cells. Leuk Lymphoma 1993; 10: 1 Galli SJ, Zsebo KM, Geissler EN. The kit ligand, stem cell factor. 35–41. 12 Bu¨hring HJ, Ullrich A, Schaudt K, Mu¨ller CA, Busch FW. The pro- Adv Immunol 1994; 55: 1–96. c-kit 2 Zsebo KM, Wypych J, McNiece IK, Lu HS, Smith KA, Karkare SB, duct of the proto-oncogene c-kit (P145 ) is a human bone mar- Sachdev RK, Yuschenkoff VN, Birkett NC, Williams LR et al. row surface antigen of hematopoietic precursor cells which is Identification, purification, and biological characterization of expressed on a subset of acute non-lymphoblastic leukemic cells. factor from buffalo rat liver-conditioned Leukemia 1991; 5: 854. medium. Cell 1990; 63: 195–201. 13 Kubota A, Okamura S, Shimoda K, Harada M, Niho Y. The c- 3 Yarden Y, Kuang W-J, Yang-Feng T, Coussens L, Munemitsu S, kit molecule and the surface immunophenotype of human acute Dull TJ, Chen E, Schlessinger J, Francke U, Ullrich A. Human leukemia. Leuk Lymphoma 1994; 14: 421–428. proto-oncogene c-kit: a new 14 Papayannopoulou T, Brice M, Broudy VC, Zsebo KM. Isolation of for an unidentified ligand. EMBO J 1987; 6: 3341–3351. c-kit receptor-expressing cells from bone marrow, peripheral 4 Tsuji K, Lyman SD, Sudo T, Clark SC, Ogawa M. Enhancement blood, and fetal liver: functional properties and composite anti- of murine hematopoiesis by synergistic interactions between steel genic profile. Blood 1991; 78: 1403–1412. factor (ligand for c-kit), interleukin-11, and other early acting fac- 15 Billips LG, Petitte D, Dorshkind K, Narayanan R, Chiu CP, Land- tors in culture. Blood 1992; 79: 2855–2860. reth KS. Differential roles of stromal cells, interleukin-7, and kit- 5 Hirayama F, Katayama N, Neben S, Donaldson D, Nickbarg EB, ligand in the regulation of B . Blood 1992; 79: Clark SC, Ogawa M. Synergistic interaction between interleukin- 1185–1192. 12 and steel factor in support of proliferation of murine lym- 16 Funk PE, Varas A, Witte PL. Activity of stem cell factor and IL-7 phohematopoietic progenitors in culture. Blood 1994; 83: 92–98. in combination on normal bone marrow B lineage cells. J Immu- 6 Bernstein ID, Andrews RG, Zsebo KM. Recombinant human stem nol 1993; 150: 748–752. cell factor enhances the formation of colonies by CD34+ and 17 Inui S, Sakaguchi N. Establishment of a murine pre-B cell clone C-kit expression in lymphoblastic cells J Tomeczkowski et al 1229 dependent on interleukin-7 and stem cell factor. Immunol Lett 36 Ferrari S, Grande A, Manfredini R, Tagliafico E, Zucchini P, Torelli 1992; 34: 279–288. G, Torelli U. Expression of 1, 3, 6, stem cell factor 18 Steel CM, Hutchins D. Soluble factors and cell-surface molecules and their receptors in acute leukemia blast cells and in normal involved in human B lymphocyte activation, growth and differen- peripheral lymphocytes and . Eur J Haematol 1993; 50: tiation. Biochim Biophys Acta 1989; 989: 133–151. 141–148. 19 Godfrey DI, Zlotnik A, Suda T. Phenotypic and functional charac- 37 Kiss C, Cesano A, Zsebo KM, Clark SC, Santoli D. Human stem terization of c-kit expression during intrathymic T cell develop- cell factor (c-kit ligand) induces an autocrine loop of growth in a ment. J Immunol 1992; 149: 2281–2285. GM-CSF-dependent megakaryocytic leukemia cell line. Leukemia 20 Morrissey PJ, McKenna H, Widmer MB, Braddy S, Voice R, Charr- 1993; 7: 235–240. ier K, Williams DE, Watson JD. Steel factor (c-kit ligand) stimulates 38 Tomeczkowski J, Frick D, Schwinzer B, Reiter A, Welte K, Sykora the in vitro growth of immature CD3− CD4−/CD8− thymocytes: K-W. Expression and regulation of c-kit receptor and proliferative synergy with IL-7. Cell Immunol 1994; 157: 118–131. response to stem cell factor in childhood lymphoma and leuke- 21 Godfrey DI, Kennedy J, Suda T, Zlotnik A. A developmental path- mia. Acta Hematol 1995; 93: 152 (Abstr.). way involving four phenotypically and functionally distinct sub- 39 Sperling C, Schwartz S, Buchner T, Thiel E, Ludwig WD. sets of CD3−CD4−CD8− triple-negative adult, mouse thymocytes Expression of the stem cell factor receptor C:KIT (CD 117) in acute defined by CD44 and CD25 expression. J Immunol 1993; 150: leukemias. Haematologica 1997; 82: 617–621. 4244–4252. 40 Smith SD, Morgan R, Link MP, McFall P, Hecht F. Cytogenetic 22 Matsuzaki Y, Gyotoku J, Ogawa M, Nishikawa S, Katsura Y, Gach- and immunophenotypic analysis of cell lines established from elin G, Nakauchi H. Characterization of c-kit positive intrathymic patients with T cell leukemia/lymphoma. Blood 1986; 67: 650– stem cells that are restricted to lymphoid differentiation. J Exp Med 656. 1993; 178: 1283–1292. 41 Pietsch T, Kyas U, Steffens U, Yakisan E, Hadam MR, Ludwig WD, 23 Godfrey DI, Kennedy J, Mombaerts P, Tonegawa S, Zlotnik A. Zsebo K, Welte K. Effects of human stem cell factor (c-kit ligand) Onset of TCR-beta gene rearrangement and role of TCR-beta on proliferation of myeloid leukemia cells: heterogeneity in expression during CD3−CD4−CD8− thymocyte differentiation. J response and synergy with other hematopoietic growth factors. Immunol 1994; 152: 4783–4792. Blood 1992; 80: 1199–1206. 24 Kurtzberg J, Meyers F, McGuire B, Crawford J. Mobilization of 42 Tomeczkowski J, Yakisan E, Wieland B, Reiter A, Welte K, Sykora peripheral blood progenitor cells in patients given recombinant K-W. Absence of G-CSF receptors and absent response to G-CSF methionyl human stem cell factor. Proc Am Assoc Res in childhood Burkitt’s lymphoma and B-ALL cells. Br J Haematol 1993; 34: 211 (Abstr.). 1995; 89: 771–779. 25 Tong J, Gordon MS, Srour EF, Cooper R, McNiece I, Hoffman R. 43 Tomeczkowski J, Beilken A, Frick D, Wieland B, Ko¨nig A, Falk Effects of the in vivo administration of recombinant methionyl M, Reiter A, Welte K, Sykora K-W. Absence of c-kit receptor and human stem cell factor (SCF) on human hematopietic stem cells absent proliferative response to stem cell factor in childhood Burk- (HSC). Proc Am Assoc Cancer Res 1993; 63: 216 (Abstr.). itt’s lymphoma cells. Blood 1995; 86: 1469–1480. 26 Demetri GD, Gordon M, Hoffman R, Hayes DF, Sledge G, Sulli- 44 Tomeczkowski J, Frick D, Schwinzer B, Reiter A, Welte K, Sykora van S, Edwards R, Merica E, Battiato L, Griffin JD. Effects of recom- K-W. Expression and regulation of c-kit receptor and proliferative binant methionyl human stem cell factor on hematopoietic pro- response to stem cell factor in childhood lymphoma and leukemia genitor cells in vivo: preliminary results from a phase I trial. Proc cells. In: Abraham N, Asano S, Brittinger G, Maestroni G, Shad- Am Assoc Cancer Res 1993; 34: 217 (Abstr.). duck R (eds). Molecular Biology of Hematopoiesis 5. Plenum 27 Orazi A, Gordon MS, Neiman RS, Sledge G, Battiato L, McNiece Press: New York, 1996, pp 173–189. I, Hoffman R. Bone marrow effects of recombinant methionyl 45 Kees UR, Ashman LK. Synergistic action of interleukin-2 and Steel human stem cell factor (SCF) in patients with normal hematopiesis. factor (SLF) on a human T lymphoblastoid cell line. Leukemia 1995; 9: 1046–1050. Proc Am Assoc Cancer Res 1993; 34: 465 (Abstr.). 46 Ichinohasama R, Endoh K, Ishizawa K, Okuda M, Kameoka J, 28 Henschler R, Brugger W, Kanz L, Mertelsmann R. Haematopoietic Meguro K, Myers J, Kadin ME, Mori S, Sawai T. Thymic lympho- growth factors in clinical medicine (news). Eur J Cancer 1994; blastic lymphoma of committed precursor origin: 30A: 118–119. a case report. Cancer 1996; 77: 2592–2603. 29 Appelbaum FR. The use of colony stimulating factors in marrow 47 Valverde LR, Matutes E, Farahat N, Heffernan A, Owusu-Ankomah transplantation. Cancer 1993; 72: 3387–3392. K, Morilla R, Catovsky D. C-kit receptor (CD 117) expression in 30 Lyman SD, Williams DE. Biological activities and potential thera- acute leukemia. Ann Hematol 1996; 72: 11–15. peutic uses of steel factor. A new active on multiple 48 Morita S, Tsuchiya S, Fujie H, Itano M, Ohashi Y, Minegishi M, hematopoietic lineages. Am J Pediatr Hematol Oncol 1992; 14: Imaizumi M, Endo M, Takano N, Konno T. Cell surface c-kit 1–7. receptors in human leukemia cell lines and pediatric leukemia: 31 Brandt J, Briddell RA, Srour EF, Leemhuis TB, Hoffman R. Role of selective preservation of c-kit expression on megakaryoblastic cell c-kit ligand in the expansion of human hematopoietic progenitor lines during adaptation to in vitro culture. Leukemia 1996; 10: cells. Blood 1992; 79: 634–641. 102–105. 32 Brugger W, Mocklin W, Heimfeld S, Berenson RJ, Mertelsmann + 49 Ratajczak MZ, Luger SM, DeRiel K, Abrahm J, Calabretta B, R, Kanz L. Ex vivo expansion of enriched peripheral blood CD34 Gewirtz AM. Role of the KIT protooncogene in normal and malig- progenitor cells by stem cell factor, interleukin-1 beta (IL-1 beta), nant human hematopoiesis. Proc Natl Acad Sci USA 1992; 89: IL-6, IL-3, -gamma, and erythropoietin. Blood 1993; 81: 1710–1714. 2579–2584. 50 Dubois CM, Ruscetti FW, Stankova J, Keller JR. Transforming 33 Andrews RG, Bartelmez SH, Knitter GH, Myerson D, Bernstein growth factor-beta regulates c-kit message stability and cell-sur- ID, Appelbaum FR, Zsebo KM. A c-kit ligand, recombinant human face protein expression in hematopoietic progenitors. Blood 1994; + stem cell factor, mediates reversible expansion of multiple CD34 83: 3138–3145. colony-forming cell types in blood and marrow of baboons. Blood 51 Yee NS, Langen H, Besmer P. Mechanism of kit ligand, phorbol 1992; 80: 920–927. ester, and calcium-induced down-regulation of c-kit receptors in 34 Andrews RG, Bensinger WI, Knitter GH, Bartelmez SH, Longin K, mast cells. J Biol Chem 1993; 268: 14189–14201. Bernstein ID, Appelbaum FR, Zsebo KM. The ligand for c-kit, stem 52 Reuss Borst MA, Buhring HJ, Schmidt H, Muller CA. AML: immu- cell factor, stimulates the circulation of cells that engraft lethally nophenotypic heterogeneity and prognostic significance of c-kit irradiated baboons. Blood 1992; 80: 2715–2720. expression. Leukemia 1994; 8: 258–263. 35 Haylock DN, To LB, Dowse TL, Juttner CA, Simmons PJ. Ex vivo 53 Siitonen T, Savolainen ER, Koistinen P. Expression of the c-kit expansion and maturation of peripheral blood CD34+ cells into proto-oncogene in myeloproliferative disorders and myelodysplas- the myeloid lineage. Blood 1992; 80: 1405–1412. tic syndromes. Leukemia 1994; 8: 631–637.