Cytokine Response Profiles of Human Myeloid Factor-Dependent Leukemia

Cytokine response proﬁles of human myeloid factor-dependent leukemia cell lines HG Drexler, M Zaborski and H Quentmeier

DSMZ-German Collection of Microorganisms and Cell Cultures, Department of Human and Animal Cell Cultures, Mascheroder Weg 1 B, D-38124 Braunschweig, Germany

Research in cytokine biology has grown exponentially in recent large variety of soluble mediators. Thus, these systems are not years as cytokines (often also termed growth factors) are now always readily amenable for the detailed study of the mech- known to be involved in a wide range of pathological and physiological processes. Continuous human leukemia cell anisms controlling hematopoietic cell processes such as, for lines represent powerful tools to investigate these mech- instance, proliferation. anisms. Most cell lines grow autonomously in standard culture The advent of the cloning and recombinant production of media (containing fetal bovine serum) independent of exter- these soluble effector molecules (variously termed hematopoi- nally added growth stimuli. Over the last 5–10 years a battery etic growth factors or cytokines) and the establishment of of myeloid leukemia-derived cell lines has been established clonal immortalized cell lines have greatly facilitated the that is constitutively dependent on the addition of cytokines to the culture. Such factor-dependent cell lines die rapidly by approaches to in vitro studies. Clearly, continuous leukemia apoptosis when deprived of the appropriate growth factor. We cell lines represent overall a more simple system for the study determined the cytokine response profiles of 19 absolutely of hematopoiesis in vitro. growth factor-dependent leukemia cell lines with myelomono- Until recently all human leukemia cell lines were grown in cytic, erythroid or megakaryocytic phenotypes with regard to medium supplemented only with fetal bovine serum (FBS) and enhanced or suppressed cellular proliferation. Cells were incu- were not dependent on exogenous hematopoietic growth fac- bated in liquid culture with optimal concentrations of various 1 recombinant human cytokines known to have effects on the tors (autonomous growth). Although cell lines dependent on growth of hematopoietic cells. A proliferative or anti-prolifera- interleukin-2 (IL-2) have been previously established, this suc- tive response to these 41 cytokines was assessed by the short- cess was restricted to a particular category of T cell lines, term 3H-thymidine uptake assay. A proliferative response was namely adult T cell leukemia lines. Early attempts to support considered as positive when the stimulation index (SI) was Ͼ2; acute/chronic myeloid leukemia (AML/CML) cell growth in Ͻ inhibition was regarded as significant with an SI 0.5. The vitro with conditioned media from stimulated lymphocytes response profile of each cell line to these 41 cytokines was different and individual. None of the cell lines responded to one succeeded in sustaining growth only for a limited time period or two factors only (minimum to at least five cytokines). Pro- of several weeks. Thus, growth factors were found to be at .but not of all cell lines was sig- least permissive for expansion of myeloid leukemia clones ,(17–13 ؍ liferation of most (n nificantly enhanced by GM-CSF, IL-3, PIXY-321, SCF and IFN- The availability of purified or recombinant cytokines ␥. TGF-␤1 consistently inhibited proliferation (in 11/19 cell enabled further progress in the culturing of leukemia cells. ␣ ␤ ␣ ␤ lines). IFN- , IFN- , TNF- and TNF- had either stimulatory or Over the last 5–10 years a spectrum of factor-dependent cell inhibitory effects. The cell lines responding most often proliferatively (to 15–19 different cytokines) were UCSD/AML1, HU-3, lines has been established from patients with various types of TF-1 and M-07e. In summary, these factor-responsive human leukemia that are absolutely dependent on addition of growth leukemia cell lines represent extremely useful model systems factors to the medium for proliferation and survival.2 The cyto- for the analysis of cytokine effects on hematopoietic cells. The kines involved include some of the interleukins (eg IL-2, IL-3, cytokine response profiles of the individual cell lines provide IL-6) and the so-called colony-stimulating factors (CSF) (GM- guidelines for the selection of the appropriate cell culture for CSF, G-CSF, M-CSF).1,2 Many of these cell lines are dependent such experiments. Keywords: leukemia; cytokines; cell lines; proliferation on either IL-3 or GM-CSF, reflecting the ability of these cytokines to stimulate early progenitor and immature cells. Further studies also suggested that non-lineage specific hormones such as insulin and insulin-like growth factor (IGF-I) can syn- ergistically stimulate growth of factor-dependent leukemia Introduction cells lines.1 Thus, the establishment of factor-dependent cell lines allowed for the continuous growth of leukemic cells that Mature blood cells have a finite lifespan and must be replen- were previously eliminated during selection in standard ished throughout adult life by the continued activity of hema- culture medium. topoietic stem and progenitor cells that are operationally These factor-dependent cell lines are important models for defined by their capacity to repopulate lymphoid and myeloid cytokine signal transduction as well as proliferative responses lineages. For these stem and progenitor cells to proliferate and and differentiation.2 Withdrawal of the supporting growth fac- mature in multiple lineages, coordinated support by a multi- tor commonly leads to cell death within a few days in most tude of cytokines is required. The complexity of hematopo- cell lines. This rapid cell death occurs by an active cellular ietic cells presents difficulties for the identification of cyto- process (apoptosis) that can be suppressed by new addition kines with specific effects on certain types of cells and at the of growth factors.3 Most of these cell lines showed a growth various stages of activation, proliferation or differentiation. response to several different cytokines; quite heterogeneous Cultures seeded with peripheral blood or bone marrow patterns of response were evident with no two lines exhibiting samples develop adherent layers that are composed of a com- the same response phenotype.2 plex mixture of mesenchymal cell types. These stroma cells Here, we have determined the proliferative or anti-prolifer- provide signals mainly via cell–cell contact or secretion of a ative response of a panel of continuous human myeloid cell lines (derived from different types of AML) to a panel of Correspondence: HG Drexler cytokines known to have mitogenic or inhibitory effects on Received 19 September 1996; accepted 20 January 1997 hematopoietic cells. These cytokine response profiles should Cytokine responses of leukemia cell lines HG Drexler et al 702 facilitate the selection of cell lines for specific experiments cells were seeded in triplicate in 100 ␮l medium in flat- with regard to elucidating cytokine-related subjects. bottomed 96-well plates and incubated in the absence or presence of cytokines; for the last 4 h of the 48 h incubation period, 1 ␮Ci [methyl-3H]thymidine (Amersham-Buchler, Braunschweig, Germany) was added to each well. Cells were Materials and methods seeded at 2.5 × 105 cells/ml. As the cell lines were constitutively dependent on externally added growth factors and nor- Culture of leukemia cell lines mally grown in medium containing such factors, the cells were washed extensively immediately prior to the experi- The continuous cell lines were taken from the stock of the cell ments. The stimulation index (SI) was calculated by dividing bank of the DSMZ (German Collection of Microorganisms and the counts per minute (c.p.m.) in the 3H-thymidine uptake Cell Cultures)4,5 or were generously provided for research pur- assay of the growth factor-containing wells by the c.p.m. of poses by the investigators who established the cell lines. The the media alone control wells of the respective cell lines. characteristics of the cell lines, information on the patient from whom they were derived, growth media and references are provided in Table 1. The establishment of the cell line Cytokines MHH-203 (kindly provided by Prof M Freund, Rostock, The following commercially supplied cytokines were used Germany) has not been published yet. The following cell lines and are listed in alphabetical order (in parentheses: full desig- are available from the DSMZ-German Collection of Micro- nation of factor; used at final concentration; specific activity organisms and Cell Cultures: M-07e (DSM ACC 104); MUTZ- where applicable or known; supplier); all cytokines were the 2 (DSM ACC 271); MUTZ-3 (DSM ACC 295); OCI/AML5 human variant: (DSM ACC 247); TF-1 (DSM ACC 334). Cell lines were grown ° • at 37 C in a humidified atmosphere of air containing 5% CO2. 5637 CM (5637 human bladder carcinoma cell line con- The basal growth media (Table 1; from Gibco BRL, Egg- ditioned medium; 10% vol; containing c. 42 ng/ml G-CSF, enstein, Germany) were supplemented with 5–20% heat- 2.1 ng/ml GM-CSF, 110 pg/ml M-CSF, 120 pg/ml SCF);25 inactivated (at 56°C for 45 min) fetal bovine serum (FBS; • bFGF (basic fibroblast growth factor; 50 ng/ml; 0.4– from Sigma, Deisenhofen, Germany). Freedom of myco- 1.0 × 107 U/mg; R&D Systems, Wiesbaden, Germany); plasma contamination was determined after thawing the cell • CNTF (ciliary neurotrophic factor; 250 ng/ml; R&D); lines by cultivation on agar and by frequent DAPI staining.23 • EGF (epidermal growth factor; 10 ng/ml; Ͼ2–4 × 107 U/mg; Cultures were passaged according to standard procedures: Pharma Biotechnologie, Hannover (PBH), Germany); spent culture medium of these suspension cell lines was • EPO (erythropoietin; 5 U/ml; Ͼ70 000 U/mg; Boehringer exchanged at regular intervals (usually after 24–48 h).5 All cell Mannheim, Mannheim, Germany); lines were examined daily in their culture vessels under an • FLT3L (flt3-ligand; 100 ng/ml; kindly provided by Dr S inverted microscope. Cells were harvested and used for Lyman, Immunex, Seattle, WA, USA); experiments in their logarithmic growth phase with viabilities • G-CSF (granulocyte colony-stimulating factor; 10 ng/ml; exceeding 90% as determined by trypan blue dye exclusion. Ͼ1.0 × 107 U/mg; Boehringer Mannheim); The cell counts and viabilities were examined in standard • GM-CSF (granulocyte–macrophage CSF; 10 ng/ml; hematocytometers. Ͼ1.0 × 107 U/mg; Boehringer Mannheim); As all cell lines used in this project are constitutively growth • HGF (hepatocyte growth factor; 50 ng/ml; R&D); factor-dependent, their culture media had to be supplemented • IFN-␣ (interferon-␣; 10 ng/ml; 2 × 108 U/mg; PBH); either with conditioned medium from the cell line 5637 (at • IFN-␤ (10 ng/ml; 2 × 108 U/mg; PBH); 10–20% v/v) or with recombinant cytokines. The cell line • IFN-␥ (10 ng/ml; Ͼ2.0 × 107 U/mg; Boehringer Mannheim); 5637 (DSM ACC 35)24,25 has been maintained in this institute • IGF-I (insulin-like growth factor I; 100 ng/ml; Boehringer for several years. It is serially passaged by trypsinization in Mannheim); the presence of EDTA and grows rapidly to form an adherent • IGF-II (100 ng/ml; Boehringer Mannheim); monolayer in plastic tissue culture flasks. Routinely, 5637 • IL-1␣ (interleukin-1␣; 10 ng/ml; Ͼ5.0 × 107 U/mg; Boeh- cells are cultured in RPMI 1640 medium supplemented with ringer Mannheim); 10% FBS. After 48–72 h following seeding of freshly thawed • IL-2 (10 ng/ml; Ͼ2.0 × 106 U/mg; Boehringer Mannheim); and washed cells, 5637 cell line-conditioned medium • IL-3 (10 ng/ml; Ͼ1.0 × 107 U/mg; Boehringer Mannheim); (henceforth termed 5637 CM) was harvested, cells and cell • IL-4 (10 ng/ml; Ͼ5.0 × 105 U/mg; Boehringer Mannheim); debris were removed by centrifugation (10 min, 200 g), and • IL-5 (10 ng/ml; Pepro Tech, London, UK); the resulting clear supernatant was passed through a 0.2 ␮m • IL-6 (10 ng/ml; Ͼ1.0 × 108 U/mg; Boehringer Mannheim); filter and was stored at −20°C until use. Recombinant granulo- • IL-7 (20 ng/ml; Ͼ5.0 × 106 U/mg; Boehringer Mannheim); cyte–macrophage colony-stimulating (GM-CSF) and stem cell • IL-8 (100 ng/ml; Ͼ1.0 × 104 U/mg; Boehringer Mannheim); factor (SCF) were used at 1–5 ng/ml and 50 ng/ml, respect- • IL-9 (10 ng/ml; R&D); ively, and were kindly provided by Dr A Mire-Sluis • IL-10 (100 ng/ml; 5 × 105 U/mg; PBH); (London, UK). • IL-11 (10 ng/ml; 2.5–5.0 × 106 U/mg; R&D); • IL-12 (10 ng/ml; R&D); • IL-13 (10 ng/ml; R&D); Proliferation assay • IL-15 (10 ng/ml; R&D); • LIF (leukemia inhibitory factor; 25 ng/ml; Ͼ1.0 × 107 U/mg; The proliferative response of the cell lines to the various effec- Boehringer Mannheim); tors was examined by cell counting in hematocytometers or • MCP-1 (monocyte chemotactic protein-1; 25 ng/ml; R&D); standard 3H-thymidine incorporation and ␤-scintillation coun- • M-CSF (macrophage-CSF; 50 ng/ml; 2.0–5.0 × 106 U/mg; ting (LKB 1209 Rackbeta; Pharmacia, Freiburg, Germany). The R&D); Cytokine responses of leukemia cell lines HG Drexler et al 703 10% +

Ref. , disease developed

d → 20% FBS 14 10% FBS 12 20% FBS 11 10% FBS 10 20% FBS 21 10% FBS 20 20% FBS 8 20% FBS 13 + + + + + + + + 20% FBS10% FBS10% FBS10% FBS 15 16 17 18 20% FBS 6 + + + + + 20% FBS 22 + 20% FBS unpublished serum 20% FBS 9 10% FBS 19 10–20% (vol) 5637 CM. Serum: + + + + + 20% FBS 7 + 20% FBS 15 GM-CSF + + GM-CSF/5637 CM GM-CSF GM-CSF/5637 CM GM-CSF GM-CSF/5637 CM GM-CSF/5637 CM GM-CSF/5637 CM GM-CSF GM-CSF/5637 CM GM-CSF/5637 CM GM-CSF + + + + + + + + + + + GM-CSF SCF GM-CSF/5637 CM GM-CSF/5637 CM GM-CSF/5637 CM GM-CSF/5637 CM GM-CSF/5637 CM 50 ng/ml SCF; + + + + + + + + -MEM -MEM -MEM -MEM -MEM -MEM -MEM/IMDM

␣ ␣ ␣ ␣ ␣ ␣ ␣

(patient) 1–5 ng/ml GM-CSF; +

Tissue of Age/Sex Growth medium plus supplement

b AML M7 (D) PB 70/M RPMI 1640 AML M6 (D) PE 68/M RPMI 1640 AML M4 (D) PB 68/F → → → -minimal essential medium. Cytokines:

␣ -MEM,

␣

Year of Original disease (status)

establishment origin

a Leukemia cell lines used Phenotype of the cell lines basedAs on described the in expression the of original publication immunological on surface the markers, respective capability cell to line: AML, differentiate acute upon myeloid stimulation,BM, leukemia (with functional bone marrow; the features, morphological PB, etc.;Media: subtypes peripheral for IMDM, of blood; more Iscove’s M1/M2 PE, for details modified pleural myeloblastic, on Dulbecco’s effusion. M4 medium; the for myelomonocytic, MUTZ-2 myeloid 1993 AML M2 (D) PB 62/M M-MOK megakaryocytic 1989 AML M7 (R) BM 1/F RPMI 1640 MB-02MHH-203 myeloid megakaryocytic/erythroid 1986 MF/MM ? AML PB adult RPMI 1640 M-07e megakaryocytic 1987 AML M7 (D) BM 0.5/F RPMI 1640 HU-3 megakaryocytic/erythroid 1991 AML M7 (D)UT-7 megakaryocytic/pluripotent BM 1988 AML M7 69/F (D) RPMI 1640 BM 64/M phenotypical features of the cell lines, theM5 reader for is monocytic, referred M6MM, myeloid to for metaplasia; Refs erythroid RAEB, 1, refractory andinto 2, anemia M7 malignancy. 4, with 5, for excess 26, of megakaryocytic); 27. blasts. CML-my Status: BC, D, cell chronic sample myeloid obtained leukemia from in patient at myeloid diagnosis blast (prior crisis; to therapy); MDS, R, myelodysplastic at syndromes; relapse. MF, myelofibrosis; Table 1 AML-193ELF-153 myeloid megakaryocytic 1988 AML ? (R) AML M5 (R)a b BMc d ? 48/M 13/F IMDM GM/SO myeloidUCSD/AML1 1988 myeloid CML-my BC (D) BM 1989 AML (R) 50/F RPMI 1640 BM 73/F RPMI 1640 GF-D8 myeloidMUTZ-3 monocyticTF-1 1989 AML M1 (D) erythroid 1993 AML M4 (D) PB 1987 82/M AML M6 (D) PB RPMI 1640 29/M BM 37/M RPMI 1640 human AB serum. Cell lines Phenotype of cell line F-36P erythroid/megakaryocytic 1989OCI/AML1 RAEB OCI/AML5OCI/AML6 myeloidSKNO-1 myeloid myeloid myeloid 1987 1990 1991 AML M4 (D) AML M4 1990 (R) MDS AML M2 (R) PB PB BM 73/F 77/M 22/M RPMI 1640 Cytokine responses of leukemia cell lines HG Drexler et al 704 Table 2a Cytokine response profiles of 19 factor-dependent human leukemia cell lines

Cytokines Cell lines

AML-193 ELF-153 F-36P GF-D8 GM/SO HU-3 M-07e MB-02 MHH-203 M-MOK

5637 CM + + + + +++ +++ +++ +++ + ++ bFGF øøøøøøøøøø CNTF øøøøøøøøøø EGF øøøøøøøøøø EPO øø++ ø + +++ ø ++ øø FTL3L øøøøøøøø+ø G-CSF + ø ø ø ø øøø + ø GM-CSF + + ++ ++ +++ +++ +++ +++ ø +++ HGF øøøøøøøøøø IFN-␣ øøøøø++− øø IFN-␤ øøøøø+++ ++ − ø + IFN-␥ øø+ + +++ +++ + øø ++ IGF-I + ø ø ø ø øøø + ø IGF-II øøøøøøøøøø IL-1␣ øøøø++øø ø ø IL-2 øøøøøø+øø ø IL-3 ++++ø+++ +++ + + +++ IL-4 ø + øø++ +++ +++ øø ++ IL-5 øø+øø+øø ø ø IL-6 øøøøø+++ + øø ø IL-7 øøøøøøøøøø IL-8 øøøøøøøøøø IL-9 øøøøøø++ øø ++ IL-10 øøøøøøøøøø IL-11 øøøøøøøøøø IL-12 øøøøøøøøøø IL-13 øøøø++ øøø ø + IL-15 øøøøøø+++ øø + LIF øøøøø+øø ø ø MCP-1 øøøøøøøøøø M-CSF øøøøøøøøøø MIP-1␣ øøøøøøøøøø NGF øøøøø+++ ++ øø ø OSM øøøøø+øø ø ø PDGF øøøøøøøøøø PIXY-321 + + ++ ++ +++ +++ +++ +++ ø +++ RANTES øøøøøøøøøø SCF ø + øø++ +++ +++ ++ ø +++ TGF-␤1 − øø−−−ø−− − ø −− TNF-␣ øøøøø+++ + ++ ø +++ TNF-␤ øøøøø+++ ø ++ ø + TPO øøøøø+++ +++ ø + +++

• MIP-1␣ (macrophage inﬂammatory protein-1␣; 50 ng/ml; • TPO (thrombopoietin; 100 U/ml = c. 50 ng/ml; kindly pro- 2.0–5.0 × 105 U/mg; R&D); vided by Dr DC Foster, Zymogenetics, Seattle, WA, USA). • NGF (nerve growth factor; 10 ng/ml; R&D); • OSM (oncostatin M; 10 ng/ml; 0.33–0.66 × 107 U/mg; R&D); • PDGF (platelet-derived growth factor; 10 ng/ml; PBH); • PIXY-321 (fusion protein of IL-3 and GM-CSF; 10 ng/ml; Results and discussion kindly provided by Immunex); • RANTES (‘regulated upon activation, normal T cell Autonomous growth of malignant hematopoietic cells is often expressed and secreted’; 50 ng/ml; Pepro); driven by the same factors stimulating normal cell prolifer- • SCF (stem cell factor; 50 ng/ml; 0.5–1.0 × 106 U/mg; R&D); ation. Leukemic cell proliferation may occur in autocrine • TGF-␤1 (transforming growth factor-␤; 50 ng/ml; 1– loops with the neoplastic cells producing their own growth 2 × 107 U/mg; PBH); factors. Growth can also take place in paracrine circuits, • TNF-␣ (tumor necrosis factor-␣; 10 ng/ml; whereby growth factors are released by surrounding cells. Ͼ2.0 × 106 U/mg; Biermann, Bad Nauheim, Germany); Continuous human leukemia cell lines represent powerful • TNF-␤ (= lymphotoxin; 10 ng/ml; Ͼ1.0 × 108 U/mg; Boeh- tools to investigate these mechanisms.26,27 Most lines grow ringer Mannheim) autonomously independent of externally added growth stimuli Cytokine responses of leukemia cell lines HG Drexler et al 705 Table 2b Cytokine response proﬁles of 19 factor-dependent human leukemia cell lines

Cytokines Cell lines

MUTZ-2 MUTZ-3 OCI-AML1 OCI-AML5 OCI-AML6 SKNO-1 TF-1 UCSD-AML1 UT-7

5637 CM + +++ +++ ++ ++ +++ +++ ++ +++ bFGF + øø+øøø + ø CNTF øøøøøø++ø EGF øøøøøøø ø ø EPO øøøøøø+++ ø +++ FTL3L + øø++ + øø ø ø G-CSF + + +++ ++ + +++ øø+ GM-CSF ø +++ + +++ + +++ ++ ++ +++ HGF øøøøøøø + ø IFN-␣ øøøøø+øøø IFN-␤ + + ++ + ø + ø +++ IFN-␥ ++ + ø + + ++ +++ ø +++ IGF-I +++øøøø ø ø IGF-II øøøøøøø ø ø IL-1␣ øøøøøøø ø ø IL-2 øøøøøøø ø ø IL-3 ø +++ ++ +++ + +++ +++ ++ +++ IL-4 ø ++ø+ø+++ + + IL-5 øøøøø++ +++ ø +++ IL-6 ++++ø+ +++ + ++ IL-7 ø + øøøøø ø ø IL-8 øøøøøøø ø ø IL-9 øøøøøøø ø ø IL-10 øøøøøøø ø ø IL-11 øøøøøøø ø ø IL-12 øøøøøøø + ø IL-13 øøøø+ + +++ + ø IL-15 øøøøøøø + ø LIF øøøøøø+++ + ø MCP-1 øøøøøøø + ø M-CSF + +++ + ++ + + ø ++ ø MIP-1␣ ø + øøøøø + ø NGF øøøøøø+++ ø +++ OSM øøøøøø+++ + ø PDGF øøøøøøø + ø PIXY-321 ø +++ ++ +++ ++ +++ +++ +++ +++ RANTES øøøøøøø ø ø SCF ++ +++ + ø ++ + +++ ++ +++ TGF-␤1 øø−−−øø−− −− −− TNF-␣ + ø −− − ø − +++ øø TNF-␤ øø−− øøø+++ øø TPO ø ++øøø+++ øø

Stimulation index (SI) comparing the cytokine-exposed cells directly with the respective control cells (grown without the speciﬁc cytokine, but in standard medium containing FBS) under identical conditions. Stimulatory effects: +, SI 2–5; ++, SI 5–10; +++,SIϾ10. Inhibitory effects: −, SI 0.25–0.5; −−,SIϽ0.25. No effects: ø, SI 0.5–2.

(except for FBS). Recently, various myeloid leukemia-derived conditions, cytokine-independent subclones may grow out cell lines have been established that are constitutively depen- within 1–2 months. dent on the addition of cytokines to the culture. Here we ana- In order to examine the mitogenic or inhibitory effects of lyzed 19 absolutely growth factor-dependent leukemia cell hematopoietic cytokines on these cell lines, aliquots of cells lines with myelomonocytic, erythroid or megakaryocytic were exposed to a panel of 41 cytokines in a short-term (48 h) phenotypes which were derived from AML or CML in blast 3H-thymidine uptake assay. Cytokines were used at optimal crisis. concentrations as described in the literature or according to All cell lines used in this study (Table 1) are constitutively own empirical experience with these reagents. A proliferative and absolutely dependent on the exogenous supply of growth response was considered positive (stimulatory effect) when the factors in order to maintain proliferation and survival; in the stimulation index SI was Ͼ2; inhibition of proliferation absence of added growth factors the cells perish undergoing (inhibitory effect) was defined as significant with an SI Ͻ0.5. apoptosis and die within 1–2 weeks (two examples are shown We did not test the effects of combinations of cytokines which in Figure 1). However, under suboptimal culture and feeding might be additive or synergistic, and did not evaluate the Cytokine responses of leukemia cell lines HG Drexler et al 706 effects of these cytokines on cell survival, differentiation, activation or other biological processes. For instance, the enhancement of cell survival actively suppressing the process of apoptosis is distinct from the stimulation of proliferation.3 For the purpose of convenience, the cytokines and the cell lines are arranged in Table 2 in alphabetical order (and not according to common functional aspects of the cytokines or their receptors or to a common origin of the cell lines). Examples of detailed cytokine response profiles are given for the well-known and most often used growth factor-dependent cell lines, namely M-07e and TF-1 (Figure 2). It is of interest that none of the 19 cell lines displayed a cytokine response profile identical with that of another line. Furthermore, none of the cell lines responded to one or two factors only (minimally to at least five cytokines). The cell lines responding most often (positively) to these cytokines were UCSD/AML1 (to 19 different cytokines), HU-3 (n = 18), TF-1 (n = 17) and M-07e (n = 15). As all cell lines are normally Figure 1 Cell death of growth factor-dependent cell lines in the absence of cytokines. Aliquots (0.5 × 106 cells in 1 ml) of M-07e and grown with 5637 CM as a source of cytokines, all of the vari- TF-1 cells were seeded in 24-well plates in standard medium (90% ous cell lines responded proliferatively to incubation with RPMI 1640 + 10% FBS) in the presence or absence of 5 ng/ml GM- 5637 CM, albeit to different degrees. A detailed analysis of the CSF. Total numbers of viable cells were determined every 2nd day composition of 5637 CM with regard to cytokine concen- by cell counting in a hematocytometer chamber. While the GM-CSF- trations had shown earlier that our 5637 CM contained exposed cells grew vigorously, the cells in the cultures devoid of any cytokines (except possibly those contained in the FBS) died within 4– 42 ng/ml G-CSF, 2.1 ng/ml GM-CSF, 0.1 ng/ml M-CSF and 25 8 days. 0.1 ng/ml SCF. While some cytokines elicited a proliferative response in

Figure 2 Cytokine response profiles of cell lines M-07e and TF-1. Aliquots (2.5 × 104 cells in 100 ␮l) of M-07e and TF-1 cells were seeded in flat-bottomed 96-well plates in triplicate in standard medium (90% RPMI 1640 + 10% FBS) in the absence (untreated control) or presence of the various cytokines (at concentrations indicated in Material and methods). For the last 4 h of the 48 h incubation, 3H-thymidine was added for the incorporation assay. Results are expressed as stimulation indices relative to the untreated control cells. Note the strong proliferative response to various cytokines. Cytokine responses of leukemia cell lines HG Drexler et al 707 the majority of cell lines, ie GM-CSF (in 17 lines), IL-3 (eds). DSM Catalogue of Human and Animal Cell Lines, 5th edn. (n = 17), PIXY-321 (n = 17), SCF (n = 14), IFN-␥ (n = 13), IL-4 DSMZ: Braunschweig, Germany, 1995. (n = 11), IFN-␤ (n = 10), IL-6 (n = 10), other factors were not 6 Lange B, Valtieri M, Santoli D, Caracciolo D, Mavilio F, Gemper- lein I, Griffin C, Emanuel B, Finan J, Nowell P, Rovera G. Growth at all effective in enhancing cellular proliferation of any of the factor requirements of childhood acute leukemia: establishment of cell lines, ie EGF, IGF-II, IL-8, IL-10, IL-11, RANTES. The most GM-CSF-dependent cell lines. Blood 1987; 70: 192–199. consistent inhibitory molecule was TGF-␤1 inhibiting the 7 Mouthon MA, Freund M, Titeux M, Katz A, Guichard J, Breton- growth of 11 cell lines. Four factors (IFN-␣, IFN-␤, TNF-␣, Gorius J, Vainchenker W. Growth and differentiation of the human TNF-␤) with inhibitory effects on some cell lines stimulated megakaryoblastic cell line (ELF-153): a model for early stages of the proliferation of other lines. megakaryopoiesis. Blood 1994; 84: 1085–1097. 8 Chiba S, Takaku F, Tange T, Shibuya K, Misawa C, Sasaki K, Miya- These factor-dependent cell lines can be used for a multi- gawa K, Yazaki Y, Hirai H. Establishment and erythroid differen- tude of assays in which cytokine biology is analyzed. The tiation of a cytokine-dependent human leukemic cell line F-36: a large number of different cytokines now identified and the parental line requiring granulocyte–macrophage colony-stimulat- pleiotropic biological effects they exhibit result in a plethora ing factor or interleukin-3, and a subline requiring erythropoietin. of scientific topics that can be explored. In the past, the most Blood 1991; 78: 2261–2268. prevalent application of cytokine-dependent cell lines was the 9 Rambaldi A, Bettoni S, Tosi S, Giudici G, Schiro` R, Borleri GM, biological (proliferative or anti-proliferative) assay to provide Abbate M, Chiaffarino F, Colotta F, Barbui T, Biondi A. Establish- 28 ment and characterization of a new granulocyte–macrophage col- estimates of levels of individual cytokines. However, the use ony-stimulating factor-dependent and interleukin-3-dependent of cell lines for research in the cytokine field is certainly not human acute myeloid leukemia cell line (GF-D8). Blood 1993; restricted to quantitative cell line-based bioassays. For 81: 1376–1383. example, the use of cell lines facilitates elucidation of cell 10 Oez S, Tittelbach H, Fahsold R, Schaetzl R, Bu¨hrer C, Atzpodien signalling processes induced by the growth factors. Detailed J, Kalden JR. Establishment and characterization of a granulocyte– descriptions of the biological effects of these groups of cyto- macrophage colony-stimulating factor-dependent human myeloid cell line. Blood 1990; 76: 578–582. kines on hematopoietic normal and malignant cells are pro- 11 Morgan D. Differential cytokine response of human lineage spe- vided in the large number of review articles and books on cific progenitor cell lines. Blood 1993; 82: 373a. these topics (for example, Refs 29 and 30). We did not test 12 Avanzi GC, Lista P, Giovinazzo B, Miniero R, Saglio G, Benetton cytokine receptor expression by these cell lines and only G, Coda R, Cattoretti G, Pegoraro L. Selective growth response to occasionally results are reported in the literature on some of IL-3 of a human leukaemic cell line with megakaryoblastic fea- the cell lines and selected cytokine receptors. Future work is tures. Br J Haematol 1988; 69: 359–366. required to determine the exact cytokine receptor expression 13 Morgan DA, Gumucio DL, Brodsky I. Granulocyte–macrophage colony-stimulating factor-dependent growth and erythropoietin- status on these cell lines which might also clarify some of the induced differentiation of a human cell line MB-02. Blood 1991; disparate response patterns observed. 78: 2860–2871. Such an extensive cytokine response profile using a large 14 Itano M, Tsuchiya S, Minegishi N, Fujie H, Minegishi M, Morita number of cytokines (n = 41) and cell lines (n = 19) has not S, Yambe T, Ohashi Y, Masuda T, Koike T, Konno T. Establishment been provided before. Thus, these data clearly may be useful and characterization of a novel human immature megakaryoblas- for scientists interested in investigations focusing on cytokine- tic leukemia cell line, M-MOK, dependent on fibroblasts for its viability. Exp Hematol 1995; 23: 1301–1309. induced events in both malignant and normal hematopoietic 15 Hu ZB, Ma W, Zaborski M, MacLeod RAF, Quentmeier H, Drexler cells. We could not detect any apparent correlation between HG. Establishment and characterization of two novel cytokine- the various responses to the cytokines (or of the overall cyto- responsive acute myeloid and monocytic leukemia cell lines, kine response profile for that matter) and both the cell lineage, MUTZ-2 and MUTZ-3. Leukemia 1996; 10: 1025–1040. or maturational stage as well as the specific phenotypical or 16 Wang C, Curtis JE, Minden MD, McCulloch EA. Expression of a functional characteristics of the leukemic cell lines studied. retinoic acid receptor gene in myeloid leukemia cells. Leukemia In summary, the present cytokine response profiles of 19 1989; 3: 264–269. 17 Wang C, Koistinen P, Yang GS, Williams DE, Lyman SD, Minden continuous human leukemia cell lines to some 41 different MD, McCulloch EA. Mast cell growth factor, a ligand for the factors encompass a wide spectrum of factors known to influ- receptor encoded by c-kit, affects the growth in culture of the blast ence hematopoietic cells. These data, while restricted to the cells of acute myeloblastic leukemia. Leukemia 1991; 5: 493–499. stimulatory and inhibitory effects of cytokines on the prolifer- 18 Tohda S, Yang GS, Ashman LK, McCulloch EA, Minden MD. ation of a panel of unique cell lines, will help to select the Relationship between c-KIT expression and proliferation in acute appropriate cell culture for in vitro cytokine experiments. myeloblastic leukemia cell lines. J Cell Physiol 1993; 154: 410– 418. 19 Matozaki S, Nakagawa T, Kawaguchi R, Aozaki R, Tsutsumi M, Murayama T, Koizumi T, Isobe RNT, Chihara K. Establishment of a myeloid leukaemic cell line (SKNO-1) from a patient with t(8;21) who acquired monosomy 17 during disease progression. Br J References Haematol 1995; 89: 805–811. 20 Kitamura T, Tange T, Terasawa T, Chiba S, Kuwaki T, Miyagawa 1 Drexler HG, Gignac SM, Minowada J. Hematopoietic cell lines. K, Piao YF, Miyazono K, Urabe A, Takaku F. Establishment and In: Hay RJ, Park JG, Gazdar A (eds). Atlas of Human Tumor Cell characterization of a unique human cell line that proliferates Lines. Academic Press: Orlando, 1994, pp 213–250. dependently on GM-CSF, IL-3, or erythropoietin. J Cell Physiol 2 Oval J, Taetle R. Factor-dependent human leukemia cell lines: 1989; 140: 323–334. new models for regulation of acute non-lymphocytic leukemia cell 21 Oval J, Jones OW, Montoya M, Taetle R. Characterization of a growth and differentiation. Blood Rev 1990; 4: 270–279. factor-dependent acute leukemia cell line with translocation 3 Williams GT, Smith CA, Spooncer E, Dexter TM, Taylor DR. (3;3)(q21;q26). Blood 1990; 76: 1369–1374. Haemopoietic colony stimulating factors promote cell survival by 22 Komatsu N, Nakauchi H, Miwa A, Ishihara T, Eguchi M, Moroi suppressing apoptosis. Nature 1990; 343: 76–79. M, Okada M, Sato Y, Wada H, Yawata Y, Suda T, Miura Y. Estab- 4 Drexler HG, MacLeod RAF, Borkhardt A, Janssen JWG. Recurrent lishment and characterization of a human leukemic cell line with chromosomal translocations and fusion genes in leukemia– megakaryocytic features: dependency on granulocyte–macro- lymphoma cell lines. Leukemia 1995; 9: 480–500. phage colony-stimulating factor, interleukin 3, or erythropoietin 5 Drexler HG, Dirks W, MacLeod RAF, Quentmeier H, Steube K for growth and survival. Cancer Res 1991; 51: 341–348. Cytokine responses of leukemia cell lines HG Drexler et al 708 23 Uphoff CC, Brauer S, Grunicke D, Gignac SM, MacLeod RAF, 27 Hassan HT, Drexler HG. Interleukins and colony stimulating fac- Quentmeier H, Steube K, Tu¨mmler M, Voges M, Wagner B, tors in human myeloid leukemia cell lines. Leuk Lymphoma 1995; Drexler HG. Sensitivity and specificity of five different myco- 20: 1–15. plasma detection assays. Leukemia 1992; 6: 335–341. 28 Mire-Sluis AR, Page L, Thorpe R. Quantitative cell line based 24 Fogh J, Wright WC, Loveless JD. Absence of HeLa cell contami- bioassays for human cytokines. J Immunol Meth 1995; 187: nation in 169 cell lines derived from human tumors. J Natl Cancer 191–199. Inst 1977; 58: 209–214. 29 Callard R, Gearing A (eds). The Cytokine FactsBook. Academic 25 Quentmeier H, Zaborski M, Drexler HG. The human bladder car- Press: London, 1994. cinoma cell line 5637 constitutively secretes functional cytokines. 30 Nicola NA (ed). Guidebook to Cytokines and their Receptors. Leukemia Res 1996 (in press). Oxford University Press: Oxford, 1994. 26 Lu¨bbert M, Herrmann F, Koeffler HP. Expression and regulation of myeloid-specific genes in normal and leukemic myeloid cells. Blood 1991; 77: 909–924.