Leukemia (2005) 19, 1270–1297 & 2005 Nature Publishing Group All rights reserved 0887-6924/05 $30.00 www.nature.com/leu
CORRESPONDENCE Leukemia-derived dendritic cells in acute myeloid leukemia exhibit potent migratory capacity
Leukemia (2005) 19, 1270–1272. doi:10.1038/sj.leu.2403794 Published online 12 May 2005 TO THE EDITOR
A majority of acute myeloid leukemia (AML) patients in complete remission will eventually relapse due to persistence of minimal residual disease (MRD). Development of alternative therapeutic approaches to prevent relapse is a great challenge. Dendritic cell (DC)-based immunotherapy targeting residual leukemic cells might offer an interesting adjuvant therapy. Effective induction of antileukemia responses in vivo requires DCs with high expression levels of costimulatory molecules, capacity to migrate from the vaccination site to T-cell areas of secondary lymph nodes, and in addition, ability to secrete type1 immunity cytokines. Previously, we demonstrated that in vitro treatment of AML cells with cytokines or calcium ionophores results in differentiation towards leukemic DC-like antigen presenting cells (AML-APCs) in approximately 75% of the patients. Moreover, these AML-APCs are able to induce autologous antileukemia T-cell responses in vitro. AML-APCs generated in calcium ionophore-based cultures are more mature with respect to expression of costimulatory molecules, CD83 and HLA-DR, and hence T-cell stimulatory capacity, than AML-APCs from cytokine-based cultures.1 However, little is known about the migratory capacity of leukemia-derived APCs in AML. In this study, we aimed at elucidating the migratory capacity of AML-derived APCs generated by cytokine-based and calcium ionophore-based culture methods. For this purpose, bone marrow or peripheral blood samples from 25 AML patients were obtained after informed consent. Patients were diagnosed as: AML with inv(16) (n ¼ 3), AML with prior myelodysplasia (n ¼ 4), AML without maturation (n ¼ 2), acute myelomonocytic leukemia (n ¼ 5), acute monoblastic leukemia (n ¼ 7), acute monocytic leukemia (n ¼ 3) and acute erythroid/myeloid leukemia (n ¼ 1) according to the World Health Organization classification of myeloid neoplasms. Mononuclear cells containing 480% blast cells were isolated from leukemia samples by Ficoll-density centrifugation (Pharmacia, Uppsala, Sweden). Normal CD34 þ progenitor cells were isolated from leukapheresis material or bone marrow of healthy donors (n ¼ 5) by immuno-magnetic column separation using CD34-microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). Differentiation of AML cells and normal CD34 þ cells towards dendritic cells was induced by 2 weeks of culture with a combination of cytokines (GM-CSF, TNF-a, IL-3, SCF, FLT3-L, IL-4, further referred to as cytokine-cultured AML-APCs or CD34-DCs, Correspondence: TM Westers, Department of Hematology, VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands; Fax: þ 31 204442601; E-mail:
[email protected] Received 14 March 2005; accepted 5 April 2005; Published online 12 May 2005
respectively) as was previously described.1 Maturation of cytokine-cultured AML-APCs and CD34-DCs was induced by additional 2 days of culture with a standard mixture of inflammatory cytokines TNF-a, IL-1b, IL-6 and prostaglandin E2 (PGE2).2 Apart from cytokine-based cultures, AML-APC differentiation was induced by 2 days incubation with calcium ionophore A23187 plus IL-4 (CI-cultured AML-APCs).1 Mean AML-APC percentages after cytokine- and CI-based cultures were 42722% (n ¼ 15, mean7s.d.) and 62720% (n ¼ 10), respectively. Cells were analyzed by flow-cytometry before and after culture for the expression of CD45, CD34, CD1a, CD14, CD40, CD80, CD86, HLA-DR, CD83, CD54, CCR7 and CXCR4. Cells expressing costimulatory molecules (CD40, CD80, CD86), adhesion molecule CD54 and Ag-presentation molecules (HLA-ABC, HLA-DR) were defined as APCs. Expression of CD83 and chemokine receptor CCR7 was used to indicate their maturation level. AML-APC-culture with either method significantly upregulated expression of CD40, CD80, CD86, CD54 and HLA-DR as compared to blast cells (data not shown). CD1a expression remained absent on AML-APCs as was described previously.1 In contrast, CD34-derived DC expressed CD1a (46713%). CI-cultured AML-APCs were significantly more mature than cytokine-cultured AML-APCs regarding mean fluorescence intensity of CD83 and HLA-DR (both P ¼ 0.043, Figure 1). Maturation-induction in cytokine-cultured AML-APCs significantly upregulated the expression of costimulatory molecules, CD54, CD83 and HLA-DR (all Po0.05). No significant differences in immunophenotype were observed between CI-cultured AML-APCs and matured cytokine-cultured AML-APCs. Except for CD1a expression, the immunophenotypic profile of matured cytokine-cultured AML-derived APCs did not significantly differ from normal matured CD34-DCs (Figure 1). In general, migration of DCs is closely related to their maturation status. Upon differentiation of immature towards mature DC receptors for inflammatory chemokines (eg CCL3 (MIP1a) and CCL5 (RANTES)) are downregulated and receptors for constitutive chemokines (eg CCL19 (MIP-3b) and CCL21 (SLC)) are increased. Mature DCs are thought to be trapped in lymph nodes via upregulation of CXCR4, the receptor for CXCL12 (SDF-1a). Analysis of chemokine receptor expression revealed that the expression of CCR7, the receptor for lymph node-associated chemokines CCL19 and CCL21, was higher on CI-cultured AML-APCs than on cytokine-cultured AML-APCs. CCR7 expression was most pronounced in matured cytokinecultured AML-APCs (P ¼ 0.028 as compared to cytokinecultured AML-APCs). In addition, matured cytokine-cultured AML-APCs significantly upregulated CXCR4 expression (P ¼ 0.021 as compared to cytokine-cultured AML-APCs). No significant differences in CCR7 and CXCR4 expression were observed for matured cytokine-cultured AML-APCs in comparison with normal matured CD34-DCs (Figure 1). Expression of chemokine receptors might indicate that AMLAPCs are capable of migrating in response to corresponding
Correspondence
1271
Figure 1 Immunophenotypical profile of AML-derived APCs and normal CD34-derived DCs after culture. White bars represent cytokine-cultured AML-APCs (n ¼ 9), gray bars calcium ionophorecultured AML-APCs (n ¼ 6), black bars matured cytokine-cultured AML-APCs (n ¼ 8) and dark gray bars matured CD34-DCs (n ¼ 4). Prior to analysis, cells were gated based on their scatter profile. Panel A depicts percentages (mean7s.d.) represented as percentage of total DC-population defined as CD40bright. CD1a was negative in all AML-APC cultures and left out. Panel B depicts mean fluorescence intensity (MFI) of the tested antibody as compared to its isotype control. Statistical analysis was performed for paired samples by the Wilcoxon signed-ranks test (nPo0.05 as compared to cytokinecultured AML-APC).
chemokines. Therefore, the migratory capacity of AML-APCs was evaluated in a transwell system (5 mm poresize, Corning Costar, Corning, NY, USA). Cells were allowed to migrate towards medium with or without chemoattractants (bacterial peptide formyl-methionyl-leucyl-phenylalanine (fMLP, Sigma, St Louis, MO, USA), CCL5, CCL19 and CXCL12 (R&D Systems, Abingdon, UK). The percentage of migrating cells was quantified flow-cytometrically by means of phycoerythrinlabeled CD40 (Immunotech, Marseille, France) to stain APCs and the addition of a fixed amount of fluorescent beads (FlowCountt Fluorospheres, Coulter, Miami, FL, USA) as a reference. Although significant as compared to spontaneous migration, cytokine-cultured AML-APCs had low motility towards inflammatory chemoattractants (fMLP and CCL5, mean7s.e.m.: 2.671.5 and 1.971.0%, respectively) and constitutive chemokines (CCL19 and CXCL12, mean7s.e.m.: 4.171.6 and 4.971.8%, respectively, Figure 2). CI-cultured AML-APCs displayed higher migration towards CCL5 (mean7s.e.m.: 10.676.2%). Moreover, migration in response to CCL19 was significantly increased (Figure 2, mean7s.e.m.: 30.5711.1%, P ¼ 0.018 as compared to cytokine-cultured AML-APCs). Maturation-induction of cytokine-cultured AML-APCs by proinflammatory cytokines resulted in highest migration towards CCL19 (Po0.001 and P ¼ 0.063 as compared to cytokine-cultured AML-APCs and CI-cultured AML-APCs, respectively). The majority of AML-APCs that migrated towards CCL19 expressed CD83 and CCR7, and had high expression levels of CD80 and CD86 in contrast to nonmigrating APCs
Figure 2 Comparison of migratory capacity of AML-derived APCs and normal CD34-derived DCs in vitro. In every experiment 105 cells were allowed to migrate for 16 h towards medium (RPMI-1640, 1% human serum albumin) with or without chemokines (CCL19 (300 ng/ ml) and CXCL12 (100 ng/ml)) in duplicate. Migration towards CCL19 is depicted on the Y-axis in panel A, migration towards CXCL12 in panel B. The percentages of APCs migrating in the presence of medium alone were subtracted from each data point. Culture methods (cytokine (CK), calcium ionophore (CI), maturation-induction (mat) and origin of APCs (AML or normal CD34 þ ) are depicted on the X-axis. nPo0.05, nnPo0.005 (Mann–Whitney U test).
indicating a mature phenotype (data not shown). Overall, CCL19-induced migration correlated significantly to the expression of CCR7 (n ¼ 25, Po0.001, Spearman’s Rho: 0.692). CXCL12-induced migration also corresponded to the maturation-status of the AML-APCs, however, the increase in migration was less as compared to CCL19-induced migration (Figure 2). No significant differences were observed regarding CCL19induced migration of immature (cytokine-cultured AML-APCs) or more or less mature leukemic APCs (matured cytokinecultured AML-APCs and CI-cultured AML-APCs, respectively) as compared to their normal CD34 þ -derived counterparts (Figure 2). In order to prevent proliferation of residual leukemic cells, AML-APCs cells have to be irradiated before their use as a vaccine in vivo. We evaluated whether this process affects CCL19 and CXCL12-induced migration of matured cytokinecultured AML-APCs. The migratory capacity of the irradiated AML-APCs (30 Gy) was similar to nonirradiated AML-APCs, although irradiation resulted in a small reduction of the number Leukemia
Correspondence
1272 of viable cells (n ¼ 3, 7.070.3% decrease, mean7s.d.) (data not shown). Surface expression of chemokine receptors is predictive for migratory capacity, although other triggers are required. PGE2, for example, is described to be necessary for migration towards CCR7 ligands.3 PGE2 is a long-standing component of DC maturation-inducing cocktails, also in clinical trials. However, it was demonstrated that PGE2 impairs cytokine secretion, in particular IL-12p70, necessary to induce Th1-responses. In contrast, ligation of CD40 prevents the onset of a migratory cell type in favor of the development of DCs that secrete IL-12. Recently, an alternative cocktail for maturation of monocytederived DC was described in which PGE2 and IL-6 were replaced by IFN-a, PolyI:C and IFN-g.2 Furthermore, Cignetti et al4 used CD40-ligation in the presence of IFN-g to mature AML-derived DCs. Although both alternative maturation-inducing cocktails contain substances that are not suitable for clinical use yet, results are promising. DCs treated with these maturation cocktails show migratory as well as IL-12p70secreting capacity. We compared the standard and alternative cocktails in a small number of AML patients and observed no significant differences regarding migration towards CCL19 (n ¼ 3, standard PGE2 cocktail: 66.178.3%, alternative IFN-a cocktail: 74.5720.6%, CD40L/IFN-g cocktail: : 47.5723.5%, mean7s.e.m.). DCs exposed to maturation stimuli appear to be more potent than immature DCs. However, the issue of the maturation state of DCs used as a vaccine in relation to relevant cytokine secretion is still under debate. Kinetic studies have shown that recently activated DCs produce relevant amounts of cytokines while prolonged maturation periods tend to exhaust DCs resulting in lower cytokine production and hence impaired Th1 responses and priming of tumor-specific cytotoxic T cells.5 Preconditioning the injection site with cytokines may be preferable to in vitro maturation protocols. In many of the current vaccination protocols, GM-CSF and Bacillus CalmetteGuerin (BCG) are added to vaccines as an adjuvant to enhance immune responses. In monocyte-derived DCs, BCG, peptidoglycans and LPS generally upregulate genes for inflammatory cytokines such as TNF-a, IL-1b, IL-6, IL-12p40, IL23, chemokines (CCL20, IL-8), adhesion molecules (ICAM-1) and apoptosis-related proteins.6 BCG-adjuvants seem to induce an inflammatory environment leading to activation and maturation of injected DCs in vivo. This might bypass the need for in vitro maturation protocols and hence the risk of introducing ‘exhausted’ APCs and impaired immune responses. Defective function of leukemia-derived DCs might imply consequences on the induction of antileukemia immune responses. In AML, circulating plasmacytoid and myeloid DCs, originating from the leukemic cell confirmed by a leukemiaassociated chromosomal abnormality, are functionally impaired with respect to maturation and allostimulatory capacity as
Leukemia
compared to their nonleukemic counterparts.7 Functional deficiencies of leukemia-derived DCs are also described in chronic myeloid leukemia. DCs display impaired maturation in response to LPS and PGE2 treatment and impaired migration as compared to monocyte-derived DCs.8 Fortunately, our data showed no significant functional defects in AML-derived APCs with respect to maturation and migration as compared to normal CD34 þ cell-derived DCs. Our data indicate that matured cytokine-cultured AML-APCs and to a lesser extent CI-cultured AML-APCs are potentially able to migrate to lymph nodes. This strengthens the potential of AML-APC-vaccination regimens to induce an antileukemia immune response aiming at the eradication of minimal residual disease in AML. 1 TM Westers1 Department of Hematology, VU University Medical Center, Amsterdam, The Netherlands I Houtenbos1 1 NCL Snoijs AA van de Loosdrecht1 GJ Ossenkoppele1
References 1 Westers TM, Stam AG, Scheper RJ, Regelink JC, Nieuwint AW, Schuurhuis GJ et al. Rapid generation of antigen-presenting cells from leukaemic blasts in acute myeloid leukaemia. Cancer Immunol Immunother 2003; 52: 17–27. 2 Mailliard RB, Wankowicz-Kalinska A, Cai Q, Wesa A, Hilkens CM, Kapsenberg ML et al. alpha-type-1 polarized dendritic cells: a novel immunization tool with optimized CTL-inducing activity. Cancer Res 2004; 64: 5934–5937. 3 Scandella E, Men Y, Gillessen S, Forster R, Groettrup M. Prostaglandin E2 is a key factor for CCR7 surface expression and migration of monocyte-derived dendritic cells. Blood 2002; 100: 1354–1361. 4 Cignetti A, Vallario A, Roato I, Circosta P, Allione B, Casorzo L et al. Leukemia-derived immature dendritic cells differentiate into functionally competent mature dendritic cells that efficiently stimulate T cell responses. J Immunol 2004; 173: 2855–2865. 5 Camporeale A, Boni A, Iezzi G, Degl’Innocenti E, Grioni M, Mondino A et al. Critical impact of the kinetics of dendritic cells activation on the in vivo induction of tumor-specific T lymphocytes. Cancer Res 2003; 63: 3688–3694. 6 Ishii K, Kurita-Taniguchi M, Aoki M, Kimura T, kashiwazaki Y, Matsumoto M et al. Gene-inducing program of human dendritic cells in response to BCG cell-wall skeleton (CWS), which reflects adjuvancy required for tumor immunotherapy. Immunol Lett 2005; 98: 280–290. 7 Mohty M, Isnardon D, Blaise D, Mozziconacci MJ, LafagePochitaloff M, Briere F et al. Identification of precursors of leukemic dendritic cells differentiated from patients with acute myeloid leukemia. Leukemia 2002; 16: 2267–2274. 8 Eisendle K, Lang A, Eibl B, Nachbaur D, Glassl H, Fiegl M et al. Phenotypic and functional deficiencies of leukaemic dendritic cells from patients with chronic myeloid leukaemia. Br J Haematol 2003; 120: 63–73.
