Bristol Institute for Transfusion Sciences and Royal Hospital for Sick Children, Bristol, UK

Correspondence to: A Blair, Bristol Institute for Transfusion Sciences, Southmead Road, Bristol BS10 5ND, UK; Fax: 011 44 117 991 2002

We have tested the hypothesis that functional dendritic cells (DC) may be generated from patients with acute lymphoblastic leukaemia (ALL). We evaluated the production of DC from blast cells taken at presentation from nine children with ALL. Blast cells were expanded in serum-free medium supplemented with Flt3L, G-CSF, GM-CSF, IL-3, IL-6 and SCF for 7 days and subsequently stimulated with Flt3L, GM-CSF and TGF- for a further 14 days, with the addition of TNF- for the final 48 h of culture. Cultured cells had the morphological appearance of DC and expressed the DC-associated antigens CD1A (range 2-87%) and CD83 (15-44%). Expression of the co-stimulatory molecules CD80 and CD86 was increased and the majority of these cells retained their expression of CD34 (73 ± 4%) and HLA-DR (79 ± 5%). Seven of the nine ALL had a leukaemia-specific abnormality and DC generated from five of these seven cases were derived from the leukaemic clone. Leukaemic DC derived from four HLA-A*02-positive ALL pulsed with CMV-associated peptides could induce significant proliferation of peptide-specific CD8⁺ T cells. This specificity was verified using tetrameric complexes of HLA class I/antigenic peptide. DC could also be generated from cells taken at times of complete remission of ALL and from normal controls using these culture conditions. These findings show that functional DC can be generated both from ALL blasts and from patients in remission; these might be utilised in future for immunotherapeutic strategies in the treatment of ALL. Leukemia (2001) 15, 1596-1603.

acute lymphoblastic leukaemia (ALL); dendritic cells (DC); immunotherapy

Introduction

Dendritic cells (DC) are potent antigen-presenting cells (APC) which are capable of inducing specific immune responses and thus have a potential clinical application in immunotherapy of human malignancies.¹ It is possible to expand populations of DC in vitro and these cells can be pulsed with tumour-specific antigens then used to stimulate clinically relevant immune responses when infused as a vaccine.^2,3,4,5,6 DC have been generated from patients with CML⁷ and AML.^8,9 In these studies the ex vivo generated DC were shown to have the same karyotypic abnormalities as the patients at diagnosis and were assumed to be derived from the leukaemic clone. These leukaemic DC were capable of inducing autologous T cell responses which were cytolytic for the original leukaemic cells but had minimal effect on normal cells.^7,8

While bone marrow transplantation (BMT) is an effective treatment for the majority of patients with ALL, 30-40% of transplant recipients will relapse after the procedure.^10,11 Donor leukocyte infusions (DLI), which induce a graft-versus- leukaemia (GVL) effect, have been used to treat relapse in both chronic and acute myeloid leukaemia following BMT, reported complete remission rates are <70% and 29%, respectively.^{12,13,14,15,16} However, response to DLI in ALL patients has been disappointing (>5%).^12,15 One possible reason for this is that although ALL cells may express tumour-associated antigens, the lack of expression of the co-stimulatory molecules CD80 and CD86 may prevent the generation of clinically significant anti-leukaemic responses.^17,18,19 Adoptive immunotherapy using donor-derived anti-leukaemic T cells could be potentially useful in the treatment of refractory ALL. In the current investigation, we have attempted to generate functional DC from ALL blasts and marrow taken at times of remission.

Materials and methods

Patient cells

Bone marrow (BM) or peripheral blood (PB) cells from nine patients at presentation and BM cells from 18 ALL patients in clinical remission (CR-ALL) were obtained after informed consent and with approval of the Research Ethics Committee of the United Bristol Healthcare NHS Trust. The characteristics of patients with overt leukaemia are shown in Table 1. Clinical remission samples were collected during morphological remission (<5% BM blasts) after therapy. Normal BM (NBM) samples were obtained from donor bone marrow harvests. Cells were Ficoll separated (Sigma-Aldrich, Poole, UK) to obtain a mononuclear cell (MNC) population, then frozen in Iscove's modified Dulbecco's medium (IMDM; GibcoBRL, Paisley, UK) supplemented with 50% fetal calf serum (FCS; GibcoBRL) and 10% DMSO (Manor Park Pharmaceuticals, Bristol, UK) and stored in liquid nitrogen.

Generation of DC from mononuclear cell fractions

Thawed MNC from ALL patients in clinical remission (CR-ALL) and NBM samples were suspended at 10⁶ cells/ml in StemSpan serum-free expansion medium (SFM) consisting of IMDM containing 20% BIT (StemCell Technologies, Vancouver, Canada). Cells were seeded on to six-well tissue culture plates (Becton Dickinson Labware, Cowley, UK) that had been pre-treated with sera for up to 1 h, to enhance adherence of monocytes. Cells were incubated for at least 3 h, then loosely adherent cells were removed by repeated washing with IMDM. Adherent cells were resuspended in either SFM or in RPMI (Sigma-Aldrich) + 10% FCS with the following combinations of cytokines. Recombinant human (rh) Flt3 ligand (Flt3L; 50 ng/ml) + rhGM-CSF (25 ng/ml) + rhTGF- (1.5 ng/ml) or rhIL-4 (25 ng/ml) + GM-CSF + TGF- or Flt3L + GM-CSF + IL-4 + SCF (50 ng/ml) + TGF- (all R&D Systems, Abingdon, UK). Fresh cytokines were added to the cultures three times per week and medium was changed after 7 days. After 14 days incubation at 37°C in a 5% O₂ and CO₂ humidified environment, the cells were allowed to settle to the bottom of the flask and the supernatant was harvested, this was repeated, allowing time for the cells to settle to the bottom of the flask once more. Cells were then resuspended in fresh medium with the same cytokine combination as before, however TNF- (3 ng/ml; R&D Systems) was added instead of TGF- for the final 48 h of culture. The non- and loosely adherent cells were harvested and used for phenotypic and functional analyses.

Generation of DC from CD34-enriched samples

A biphasic culture system was used to generate DC from CD34-enriched CR samples or from ALL blasts. CD34⁺ cells from CR-ALL samples were enriched by positive selection using the MiniMACS CD34 progenitor cell isolation kit (Miltenyi Biotech, Bisley, UK), according to the manufacturer's instructions. The purity of the selected population ranged from 83 to 99% CD34⁺. It was not necessary to enrich diagnostic samples, due to the high content of CD34⁺ ALL blasts (67-99%) in these samples.

Selected CD34⁺ cells or ALL blasts were suspended in SFM at 2 ´ 10⁵ cells/ml supplemented with Flt3L (50 ng/ml), IL-3 (20 ng/ml), IL-6 (20 ng/ml), G-CSF (20 ng/ml), GM-CSF (20 ng/ml) and SCF (50 ng/ml; F36GGMS; all R&D Systems) for 7 days. Expanded cells were then harvested and centrifuged at 250 g for 10 min at 25°C. Cells were resuspended in SFM, supplemented with the three cytokine combinations and maintained for a further 16 days, as described above for generation of DC from unsorted MNC.

Phenotypic analysis of cells by flow cytometry

The phenotype of the freshly isolated and cultured cells was determined by flow cytometric analysis using the Coulter EPICs XL-MCL Flow cytometer (Beckman Coulter, High Wycombe, UK). Cells were resuspended in Hank's balanced salt solution (Sigma-Aldrich) + 2% FCS and stained for 30 min on ice with monoclonal antibodies directly coupled to the fluorochromes, fluorescein isothiocyanate (FITC) or phycoerythrin (PE). CD1A-PE, CD80-PE, CD83-FITC CD86-PE (Coulter-Immunotech Diagnostics, Miami, FL, USA), CD10-FITC, CD19-FITC, CD34-PE and HLA-DR-FITC (Becton Dickinson), were used according to the manufacturer's instructions. Forward and side scatter gates were set up to eliminate debris and analysis gates, which excluded 99.9% of the cells in isotype controls, were used to define positive events. Any positive value (ie 0.1%) for the isotype control was then subtracted from the percentage positive in the relevant stained samples. Figure 1 shows a representative case of ALL blast cells from one patient before and after the culture period.

Peptide-pulsed dendritic cell stimulation of CD8 cells

Five human cytomegalovirus peptides were used to characterise the ability of DC to stimulate MHC class I-restricted antigen-specific responses (Table 2). DC from HLA-A*02-positive ALL patients were pulsed for 24 h with 1 g of peptide (M747, M958, M959, M960 or M965),²⁰ then medium containing non-bound peptide was removed. An HLA-A3 peptide (M969) was used as a control. Purified populations of CD8⁺ responder cells (>98% CD8⁺) were obtained from PBMNC from normal HLA-A*02 homozygote donors by positive selection (MiniMACS). 2 ´ 10⁵ responder CD8⁺ cells were added to the peptide-pulsed DC and the cultures were supplemented with interleukin-2 (IL-2; 50 IU/ml). Cultures were fed at 4 day intervals by exchanging 250 l medium with fresh IL-2 supplemented medium. After 10 days, cells were pulsed with ³H-Tdr and the level of incorporation was determined as described above. Response to each peptide was tested in triplicate wells.

In order to evaluate the specificity of the responding CD8⁺ cells, 200 l of cell suspension was removed on days 3, 7, 10 and 14, and pelleted by centrifugation at 400 g. The supernatant was removed and cells re-suspended in 20 l phosphate-buffered saline (pH 7.4). One microlitre CD8-FITC (Becton Dickinson) and 5 l MHC class I/pp65_493-503 tetramer-PE complex (kindly supplied by Prof P Moss, CRC Institute, University of Birmingham, UK) was added to each cell suspension. The proportion of CD8⁺ cells, which bound to the tetramer complex, was evaluated by flow cytometry, together with appropriate isotype controls.

Cytogenetic analysis

Cultured cells from seven presentation samples, with an abnormal karyotype, were evaluated for the leukaemia- specific cytogenetic change by fluorescent in situ hybridisation (FISH). The Aneuvysion kit (Vysis UK, Richmond, UK) was used to detect additional chromosomes 13, 18 or 21 in two patients and one other patient with a complex karyotype. The TEL/AML1 probe (Vysis) was used to detect the presence of t(12;21) in three patients and the EGR1 probe (Vysis) was used to confirm del(5q) in one patient.

Statistics

Statistical analyses were performed using the paired t-test.

Results

Generation of DC from mononuclear cell fractions

Initial experiments were conducted using the plastic adherent cell fraction from eight NBM donors and material from eight ALL patients in CR. After 4-5 days in culture, clusters of cells with very short projections could be observed in suspension. By days 6-8 (NBM) or days 8-12 (CR-ALL), clusters were usually more dispersed and cells with long dendritic projections were evident using a standard inverted microscope. Viability of cells in these cultures and the DC morphology could be maintained for up to 21 days. However, no significant increase in cell number was observed after 14-17 days. The total number of viable cells from CR-ALL samples recovered after the culture period varied among individual patient samples (mean ± s.e.; 7 ± 2 ´ 10⁵ cells/ml, range 7 ´ 10³-2 ´ 10⁶). Interestingly, both viability and the number of cells with DC morphology were lower in cultures where RPMI + 10% FCS was used compared to using SFM (data not shown).

In all cases, the starting population and the cells harvested at the end of the culture period were analysed for surface expression of DC-associated antigens CD1A and CD83, HLA-DR and the co-stimulatory molecules CD80 and CD86. The phenotype of cells generated using the plastic adherence, single stage culture method are shown in Figure 2. It was possible to expand the number of NBM cells expressing CD1A from a mean of 2% at initiation of culture to 28% after culture and the expression of CD83 increased from 0.8% to 16% (Figure 2a). The proportion of cells expressing the costimulatory molecules and HLA-DR also increased (33% CD80⁺, 42% CD86⁺, 67% HLA-DR⁺).

The phenotype of CR-ALL samples before and after the single stage culture are shown in Figure 2b. The mean proportion of cells expressing the DC markers CD1A and CD83 increased from 0.4% to 18% and from 0.3% to 14%, respectively. The number of cells expressing CD80 increased from 4% to 36%. Expression of CD86 and HLA-DR also increased, from 9% to 39% and from 19% to 54%, respectively. A starting population of 10⁶ MNC yielded 2 ± 0.3 ´ 10⁵ CD1A/CD83⁺ DC (range 2 ´ 10³-5 ´ 10⁵). The number of cells with DC morphology and phenotype were lower in cultures using RPMI + 10% FCS compared to SFM (data not shown), demonstrating that the SFM was superior to RPMI + 10% FCS for generating DC in this system.

Evaluation of cytokine combinations for DC generation

Three different cytokine combinations were evaluated to determine the optimal combination for generation of DC from NBM and CR-ALL cells (Figure 3a and b). The combination of Flt3L + GM-CSF + TGF- generated the highest number of cells with DC morphology from both unsorted and CD34⁺- enriched samples. Expression of DC-associated markers, CD1A and CD83, on cultured CR-ALL cells was significantly higher using this combination compared to the other two combinations evaluated (P < 0.001). Similar results were obtained when using NBM cells to generate DC, however, the differences in number and phenotype of cultured cells were not significant. The combination Flt3L + GM-CSF + TGF- was subsequently used in all experiments reported below.

Generation of DC from CD34-enriched samples

CD34⁺ cells were isolated from the eight CR-ALL samples, used previously, and 10 additional CR-ALL cases. The CD34⁺-enriched cells were expanded for 7 days, followed by the 16 day culture for DC. An overall expansion of cell numbers was observed in these cultures, from an initial inoculum of 2 ´ 10⁵ cells to 3.5 ± 0.2 ´ 10⁶ cells at harvest (day 23). The yield of cells with DC morphology and phenotype was greater using this biphasic system than the plastic adherence method, 10⁵ CD34⁺ cells generated 4 ´ 0.1 ´ 10⁵ DC (range 8 ´ 10³-1.5 ´ 10⁶). In all subsequent experiments, the biphasic culture system was used for generation of DC from CR-ALL and from samples at presentation. The phenotype of CD34⁺ cells isolated from CR-ALL samples cultured in this system is shown in Figure 4a. The proportion of cells which expressed CD1A increased from a mean of 0.03% to 10% and expression of CD83 increased from 0.5% to 15%. Cultured cells expressed CD80 (14%) and CD86 (18%); expression of HLA-DR and CD34 remained close to levels observed at initiation of culture.

Similar results were obtained when ALL blasts from the nine presentation cases detailed in Table 1 were used (Figure 4b). It was possible to generate DC from all the subtypes evaluated; 10⁵ freshly isolated ALL blast cells yielding a mean of 2 ± 0.6 ´ 10⁵ DC, (range 3 ´ 10⁴-4 ´ 10⁵). ALL blasts expressed CD86 (22%) and HLA-DR (81%) and this expression was maintained after the culture period (46% and 79%, respectively). Mean expression of CD1A and CD83 increased from 6.7% and 0.6%, respectively, to 28% for both antigens after differentiation in vitro, accompanied by increased expression of CD80 (33%). The cultured cells retained their expression of CD34 (73 ± 11%) and expression of the lymphoid markers CD10 (64 ± 6%) and CD19 (72 ± 8%). One of the patients studied (pt. 9) was diagnosed as T-ALL. DC generated from this patient grew as well as DC from the B cell ALL patients evaluated and they were not morphologically or phenotypically distinct, apart from their lack of expression of CD10 and CD19.

DC stimulation of peptide-specific responses

DC derived from CR-ALL samples, were pulsed with synthetic peptides and investigated for their ability to stimulate an antigen-specific response using CD8⁺ cells freshly isolated from HLA-A*02 allogeneic donors. Figure 5a shows the proliferative response measured at day 10. All peptide-pulsed DC stimulated a CD8⁺ cell response with the maximum responses being observed using peptides M960 and M747 (pp65_493-503). The antigen specificity of expanded CD8⁺ cell populations was confirmed using MHC class I/tetramer complexes for the M747 (pp65_493-503) peptide. Cells were harvested from cultures stimulated either with M747 or the peptides M960 and M969, as controls, and the proportion of CD8⁺ cells that were stained with the tetramer complex was determined. Tetramer positive-CD8⁺ cells could be detected from day 4 in the cultures which had been stimulated by M747 (pp65_493-503) pulsed DC (Figure 5b). The number of tetramer-positive, antigen-specific, cells detected increased with length of time in culture to reach a maximum of 30% detected at day 16. The CD8⁺ cells stimulated with peptides M960 or M969 did not stain with the tetramer complex, demonstrating that the response to the M747 (pp65_493-503) peptide was antigen specific. In vitro generated DC were also shown to enhance stimulation in mixed lymphocyte reactions (data not shown).

ALL blast DC stimulation of peptide responses

The above methods were used to evaluate the functional capacity of DC derived from ALL blasts from four patients who were HLA-A*02 positive (pts 1-4, Table 1). It was not possible to perform all the functional assays on every overt leukaemia sample studied because of the limited material available from any given patient and lack of information on HLA type at diagnosis. Each ALL patient evaluated in this assay had an abnormal karyotype which could be detected by FISH analysis. DC from each patient were pulsed with peptide M747 (pp65_493-503) for 24 h and used to stimulate CD8⁺ cells from an HLA-A*02 allogeneic donor, results are shown in Figure 6. In each case, a rapid CD8⁺ expansion could be demonstrated from day 0 to day 3. During this time, the absolute CD8⁺ cell number increased by 6.5 ´ 10⁴ cells, from 2 ´ 10⁵ to 2.65 ´ 10⁵ (Figure 6a). Dual staining showed that 34-50% of these cells also stained positive for the pp65_493-503 tetramer. From day 3 to day 7 there was only a marginal increase in absolute CD8⁺ cell numbers. However, 69-79% of these cells now stained positive for the tetramer at this time. The highest CD8⁺ cell numbers were observed between days 10 and 14 in two patients and 82-91% of these cells were tetrameter positive (Figure 6b). These findings demonstrate that DC derived from ALL blasts were capable of stimulating an antigen-specific T cell response.

Cytogenetic analysis of cultured cells

Cells from the DC cultures of presentation samples were evaluated for the presence of chromosomal aberrations in patients 1-7, who had a leukaemia specific karyotype (Table 3). Whenever possible, at least 100 interphase cells were scored per sample. No abnormal nuclei were detected in two of the three patients with t(12;21) using the TEL/AML1 probe (pts 5 and 6). However, in the remaining five patients, 66-100% of cells analysed were found to have the karyotypic abnormality present in the patients blasts cells at diagnosis. Since the majority of cultured cells derived from these five patients had typical DC morphology and 40-87% expressed CD1A /CD83, we have calculated that at a minimum 8-53% of these cultured cells were in fact DC derived from the leukaemic clone.

Discussion

In this investigation, we have used a biphasic culture system to generate DC from ALL at presentation and in CR. A significant number of ALL patients relapse following chemotherapy and BMT. This is most likely due to a failure to eliminate MRD and the persistence of refractory leukaemic cells. Therefore, we felt it was important to investigate DC generation from CR-ALL samples. It is possible, although unproven in our experiments, that DC generated from CR samples might have originated from residual cells from the leukaemic clone. In such cases, the DC might have contained tumour antigens and may therefore have the potential to present such antigens to T cells. DC generated in the biphasic culture system we describe here, had the characteristic morphology and immunophenotype consistent with mature APC. Similar proportions of cells with a DC phenotype were generated under these conditions when using CR-ALL, ALL cells at diagnosis or NBM. We found the biphasic system to be superior to the single stage, plastic adherence method for generation of cells with DC morphology and phenotype. Our findings that cells with the characteristic morphology, immunophenotype and functional capacity of DC were detected in the non-adherent fraction of the culture system cells are consistent with those of Choudhury et al.⁸ The tightly adherent cell fraction was found to mainly consist of monocyte/macrophage precursors.

There have been only a limited number of reports on the generation of DC from ALL cells. Robinson et al⁹ generated DC with allostimulatory capacity, using IL-4 + SCF, from four patients with bi-phenotypic leukaemia and Cignetti et al²¹ used IL-4 + CD40L to culture DC from two ALL patients. In contrast, Kohler et al²² were unsuccessful in attempts to generate DC from three patients with Ph⁺ ALL using a number of cytokine combinations. Alternative strategies to generate CTL against ALL cells, using CD40L, have been reported.^23,24 Activation of CD40, the receptor for CD40L, on normal B cells promotes their antigen-presenting capacity by upregulating their expression of intercellular adhesion molecules, CD80 and MHC class I and II. Cardoso et al²³ generated anti- leukaemic CTL from BM in 10 of 15 patients with pre-B ALL by direct co-stimulation with CD40L stimulated pre-B ALL cells with repetitive priming and IL-2 driven expansion. However, it was not possible to generate CTL from PB samples of these patients.²³ Transgenic expression of CD40L has been shown to enhance the anti-leukaemic response of a CD40⁺ murine lymphoblastic cell line.²⁴

Here, we have used Flt3L + GM-CSF + TGF- to generate DC from nine patients at diagnosis of ALL, with pre-B; c-ALL; T-ALL and bi-lineage subtypes, and from 18 patients in CR. It was possible to significantly increase the number of cells which expressed DC-associated markers whilst maintaining or increasing the expression of co-stimulatory molecules. Cultured cells derived from five of seven ALL cases at presentation were found to be derived from the leukaemic clone. The leukaemic origin of the DC is strongly suggested by the following evidence. Taking pt 1 as a representative example; the cultures were initiated with <1.9 ´ 10⁵ blasts, at the end of the culture period the total number of cells had increased to 2.1 ´ 10⁶ of which, 45% expressed the DC markers (9.5 ´ 10⁵ DC). FISH analysis showed that 93% of the cultured cells from this patient had an abnormal karyotype. Thus, 1.5 ´ 10⁵ cultured cells had a normal karyotype by FISH. Given that the total number of DC in this case was 9.5 ´ 10⁵, at least 8 ´ 10⁵ DC were derived from the leukaemic clone. DC were not purified prior to FISH analysis, thus the presence of other cells in the culture may have an effect on the results obtained. However, it is clear from this data that a significant proportion (>8->53%) of the cells which were positive by FISH after culture were in fact DC derived from the leukaemic clone.

The combination of Flt3L + GM-CSF + TGF- for the second stage of the culture system, was found to be superior to the other cytokine combinations evaluated for generation of DC in ALL. CD40L has been reported to increase the number of DC generated from both normal marrow and AML.^25,26 We are currently evaluating the use of CD40L and other cytokine combinations in order to optimise our culture system. DC generated in our culture system had the capacity to stimulate primary immune responses measured by the allogeneic mixed leukocyte reaction. In addition, DC pulsed with HLA-A*02 specific-binding peptides could stimulate significant antigen-specific proliferation of CD8⁺ T cells. The specificity of this proliferative response was confirmed using antigen-specific tetramers. The majority of DC generated from these patients cells were derived from the leukaemic blasts and not from normal progenitor cells.

In this study we have described an effective culture system for the generation of DC with functional allostimulatory activity from ALL patients both at diagnosis and in CR. DC generated in this system were capable of presenting CMV antigen to stimulate antigen-specific responses from allogeneic T cells. Further investigation is required to evaluate the ability of the in vitro-generated leukaemic DC to stimulate allogeneic or autologous T cells, which may be capable of inducing a leukaemia-specific immune response.

Acknowledgements

The authors wish to thank the staff of the Stem Cell Processing Laboratory, NBS Bristol and staff of the Regional Cytogenetics Unit, Southmead Hospital, for excellent technical assistance and Dr Marc Williams, Queen Mary and Westfield College, University of London, for expertise in this field.

1 Grabbe S, Beissert S, Schwarz T, Granstein RD. Dendritic cells as initiators of immune responses: a possible strategy for tumor immunotherapy? Immunol Today 1995; 16: 117-121, MEDLINE

2 Hsu FJ, Benike C, Fagoni F, Liles TM, Czerwinski D, Taidi B, Engleman EG, Levy R. Vaccination of patients with B-cell Lymphoma using autologous antigen pulsed dendritic cells. Nat Med 1996; 2: 52-58, MEDLINE

3 Boog C, Kast WM, Timmers HT, Boes J, de Waal LP, Melief CJ. Abolition of specific immune response defect by immunization with dendritic cells. Nature 1995; 318: 59-62,

4 Schuler G, Steinman RM. Dendritic cells as adjuvants for immune-mediated resistance to tumours. J Exp Med 1997; 186: 1183-1187, MEDLINE

5 Wong C, Morse M, Nair SK. Induction of primary, human antigen-specific cytotoxic T lymphocytes in vitro using dendritic cells pulsed with peptides. J Immunother 1998; 21: 32-40, MEDLINE

6 Osman Y, Takahashi M, Zheng Z Toba K, Liu A, Furukawa T, Aizawa Y, Shibata A, Koike T. Activation of autologous or HLA-identical sibling cytotoxic T lymphocytes by blood derived dendritic cells pulsed with tumor cell extracts. Oncol Rep 1999; 6: 1057-1063, MEDLINE

7 Choudhury A, Gajewski JL, Liang JC, Popat U, Claxton DF, Kliche KO, Andreeff M, Champlin RE. Use of leukemic dendritic cells for the generation of antileukemic cellular cytotoxicity against Philadelphia chromosome-positive chronic myelogenous leukemia. Blood 1997; 89: 1133-1142, MEDLINE

8 Choudhury BA, Liang JC, Thomas EK, Flores-Romo L, Xie QS, Agusala K, Sutaria S, Sinha I, Champlin RE, Claxton DF. Dendritic cells derived from acute myelogenous leukemia cells stimulate autologous, antileukemic T-cell responses. Blood 1999; 93: 780-786, MEDLINE

9 Robinson SP, English N, Jaju R, Kearney L, Knight SC, Reid CDL. The in-vitro generation of dendritic cells from blast cells in acute leukaemia. Br J Haematol 1998; 103: 763-771, MEDLINE

10 Oakhill A, Pamphilon DH, Potter MN, Steward CG, Goodman S, Green A, Goulden P, Goulden NJ, Hale G, Waldmann H, Cornish JM. Unrelated donor bone marrow transplantation for children with relapsed acute lymphoblastic leukaemia in second complete remission. Br J Haematol 1996; 94: 574-578, MEDLINE

11 Knechtli CJ, Goulden NJ, Hancock JP, Grandage VL, Harris EL, Garland RJ, Jones CG, Rowbottom AW, Hunt LP, Green AF, Clarke E, Lankester AW, Cornish JM, Pamphilon DH, Steward CG, Oakhill A. Minimal residual disease status before allogeneic bone marrow transplantation is an important determinant of successful outcome for children and adolescents with acute lymphoblastic leukemia. Blood 1998; 92: 4072-4079, MEDLINE

12 Horowitz MM, Gale RP, Sondel PM, Goldman JM, Kersey J, Kolb HJ, Rimm AA, Ringden O, Rozman C, Speck B. Graft-versus-leukemia reactions after bone marrow transplantation. Blood 1990; 75: 555-562, MEDLINE

13 Kolb HJ, Mittermuller J, Clemm C, Holler E, Ledderose G, Brehm G, Heim M, Wilmanns W. Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood 1990; 76: 2462-2465, MEDLINE

14 Porter DL, Roth MS, McGarigle C, Ferrara JL, Antin JH. Induction of graft-versus-host disease as immunotherapy for relapsed chronic myeloid leukemia. N Engl J Med 1994; 330: 100-106, MEDLINE

15 Kolb HJ, Schattenberg A, Goldman JM, Hertenstein B, Jacobsen N, Arcese W, Ljungman P, Ferrant A, Verdonck L, Niederwieser D, for the European Group for Blood and Marrow Transplantation Working Party Chronic Leukemia. Graft-versus-leukemia effect of donor lymphocyte transfusions in marrow grafted patients. Blood 1995; 86: 2041-2050, MEDLINE

16 Collins RH Jr, Shpilberg O, Drobyski WR, Porter DL, Giralt S, Champlin R, Goodman SA, Wolff SN, Hu W, Verfaillie C, List A, Dalton W, Ognoskie N, Chetrit A, Antin JH, Nemunaitis J. Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation. J Clin Oncol 1997; 15: 433-444, MEDLINE

17 Foa R, Fierro MT, Cesano A, Guarini A, Bonferroni M, Raspadori D, Miniero R, Lauria F, Gavosto F. Defective lymphokine-activated killer cell generation and activity in acute leukaemia patients with active disease. Blood 1991; 78: 1041-1046, MEDLINE

18 Cardoso AA, Schultze JL, Boussiotis VA, Freeman GJ, Seamon MJ, Laszlo S, Billet A, Sallan SE, Gribben JG, Nadler LM. Pre-B acute lymphoblastic leukaemia cell may induce T-cell anergy to alloantigen. Blood 1996; 88: 41-48, MEDLINE

19 Hirano N, Takahashi T, Takahashi T, Ohtake S, Hirashima K, Emi N, Saito K, Hirano M, Shinohara K, Takeuchi M, Taketazu F,Tsunoda S, Ogura M, Omine M, Saito T, Yazaki Y, Ueda R, Hirai H. Expression of co-stimulatory molecules in human leukaemias. Leukemia 1996; 10: 1168-1176, MEDLINE

20 Solache A, Morgan CL, Dodi AI, Morte C, Scott I, Baboonian C, Zal B, Goldman J, Grundy JE, Madrigal JA. Identification of three HLA-A*0201-restricted cytotoxic T cell epitopes in the cytomegalovirus protein pp65 that are conserved between eight strains of the virus. J Immunol 1999; 163: 5512-5518, MEDLINE

21 Cignetti A, Bryant E, Allione B, Vitale A, Foa R, Cheever MA. CD34⁺ acute myeloid and lymphoid leukemic blasts can be induced to differentiate into dendritic cells. Blood 1999; 94: 2048-2055, MEDLINE

22 Kohler T, Plettig R, Wetzstein W, Schmitz M, Ritter M, Mohr B, Schaekel U, Ehninger G, Bornhauser M. Cytokine-driven differentiation of blasts from patients with acute myelogenous and lymphoblastic leukaemia into dendritic cells. Stem Cells 2000; 18: 139-147, MEDLINE

23 Cardoso AA, Seamon MJ, Afonso HM, Ghia P, Boussiotis VA, Freeman GJ, Gribben JG, Sallan SE, NadlerLM. Ex vivo generation of human anti-pre-B leukemia specific autologous cytolytic T cells. Blood 1997; 90: 549-561, MEDLINE

24 Dilloo D, Brown M, Roskrow M, Zhong W, Holladay M, Holden W, Brenner M. CD40 ligand induces an antileukemia immune response in vivo. Blood 1997; 90: 1927-1933, MEDLINE

25 Mackey MF, Gunn JR, Maliszewski C, Kikutani H, Noelle RJ, Barth RJ. Dendritic cells require maturation via CD40 ligand to generate protective antitumor immunity. J Immunol 1998; 161: 2094-2098, MEDLINE

26 Labeur MS, Roters B, Pers B, Mehling A, Luger TA, Schwarz T, Grabbe S. Generation of tumor immunity by bone marrow-derived dendritic cells correlates with dendritic cell maturation stage. J Immunol 1999; 162: 168-175, MEDLINE

Figure 1 Immunophenotype of freshly isolated and cultured ALL blast cells. ALL blast cells were expanded in SFM supplemented with F36GGMS for 7 days then cultured for a further 14 days in the presence of Flt3, GM-CSF and TGF-, with the addition of TNF- for the final 48 h of culture. Freshly isolated (a) and cultured cells (b) were stained with fluorescent antibodies and analysed by flow cytometry. The results of one representative experiment are shown.

Figure 2 Phenotype of cells cultured from MNC fraction of NBM and CR-ALL. NBM cells (a) and CR-ALL cells (b) were set up in culture, following the removal of non-adherent fraction and grown in serum-free medium supplemented with FLt3L, GM-CSF and TGF- for 14 days. Cells were analysed for a number of DC-associated antigens and co-stimulatory molecules by flow cytometry prior to the initiation of culture and at the end of the culture period. Results represent the mean ± s.e. of eight NBM and eight CR-ALL samples. () Pre-culture cells; () day 14 cultured cells.

Figure 3 Effect of cytokine combinations on the generation of DC. Three different cytokine combinations were evaluated for the ability to induce DC differentiation from expanded CD34⁺ NBM cells (a) and CD34⁺ CR-ALL cells (b). Flt3 + GM-CSF + TGF- generated the highest number of cells with DC morphology and phenotype from both ALL and NBM cells. Expression of CD1A and CD83 was significantly higher using this combination to stimulate DC differentiation in ALL cases compared to the other two combinations evaluated (P < 0.001). () Flt3L + GM-CSF + TGF-; () IL-4 + GM-CSF + TGF-; () Flt3L + IL-4 + GM-CSF + SCF + TGF-. Results represent the mean ± s.e. of seven NBM samples and 10 CR-ALL samples.

Figure 4 Expression of DC-associated markers on cells from ALL patients in CR or ALL blasts cultured in the bi-phasic system. CD34⁺-enriched cells were expanded in SFM + F36GGMS for 7 days then grown in SFM supplemented with Flt3L, GM-CSF and TGF- for 14 days. TNF- was added for the final 48 h of culture. Cells were analysed by flow cytometry prior to the initiation of culture () and at the end of the culture period (). Results represent the mean ± s.e. of 18 CR-ALL samples (a) or mean ± s.e. of nine patient samples at presentation of disease (b).

Figure 5 DC stimulation of peptide-specific responses. DC were pulsed with a number of HLA-A*02 or HLA-A3-specific peptides and evaluated for their ability to stimulate CD8⁺ T cells. Proliferation was measured by ³H-Tdr incorporation at day 10 of culture (a). Proliferation in control cultures of target cells alone or DC alone was <500 c.p.m. The specificity of the proliferative response was confirmed using a tetrameric complex of HLA class 1/M747(pp65) (b). The tetramers only bound to CD8⁺ cells which had been stimulated by DC pulsed with M747(pp65).

Figure 6 Ability of peptide pulsed ALL DC to stimulate antigen-specific responses. DC generated from four ALL patients at presentation (pts 1-4) and undifferentiated blasts from patient 4 were pulsed with M747 (pp65), or an irrelevant peptide M960, for 24 h then evaluated for their ability to stimulate CD8⁺ cells from normal donors over a 14 day period (a). The specificity of the response was confirmed by flow cytometric analysis of HLA class 1/M747(pp65) tetramer binding on CD8⁺ cells (b). --, Patient 1; --, patient 2; --, patient 3; --, patient 4; - -- -, undifferentiated blasts, patient 4; ⋅ ⋅ ⋅ ⋅, irrelevant peptide controls M960.

Table 1 Clinical characteristics of ALL patients with overt disease

Table 2 Human cytomegalovirus peptides used to pulse DC

Table 3 Cytogenetic analysis of cultured ALL blasts