Skip to main content
NCBI home page
Molecular Therapy logoLink to Molecular Therapy
. 2012 Nov 13;21(2):466–475. doi: 10.1038/mt.2012.227

Graft-versus-leukemia Effect of HLA-haploidentical Central-memory T-cells Expanded With Leukemic APCs and Modified With a Suicide Gene

Monica Casucci 1, Serena Kimi Perna 1,10, Laura Falcone 1, Barbara Camisa 1, Zulma Magnani 1, Massimo Bernardi 2, Alessandro Crotta 2, Cristina Tresoldi 2, Katharina Fleischhauer 3, Maurilio Ponzoni 4, Silvia Gregori 5, Federico Caligaris Cappio 6,7, Fabio Ciceri 2, Claudio Bordignon 7,8, Alessandro Cignetti 9, Attilio Bondanza 1,*, Chiara Bonini 1
PMCID: PMC3594033  PMID: 23299798

Abstract

Allogeneic hematopoietic stem cell transplantation (HSCT) from a human leukocyte antigen (HLA)-haploidentical family donor (haplo-HSCT) is a readily available and potentially curative option for high-risk leukemia. In haplo-HSCT, alloreactivity plays a major role in the graft-versus-leukemia (GVL) effect, which, however, is frequently followed by relapse due to emerging leukemic cell variants that have lost the unshared HLA haplotype as a mechanism of immune escape. We report that stimulation of HLA-haploidentical donor T lymphocytes with leukemic antigen-presenting cells (L-APCs) expands a population of leukemia-reactive T cells, which, besides alloreactivity to unshared HLAs, contain leukemia-associated specificities restricted by shared HLAs. According to a preferential central-memory (TCM) phenotype and to high interleukin (IL)-7Rα expression, these T cells persist in vivo and sustain a major GVL effect in a clinically relevant xenograft model. Moreover, we demonstrate that modifying L-APC–expanded T cells to express the herpes simplex virus thymidine kinase (HSV-tk) suicide gene enables their elimination with the prodrug ganciclovir (GCV), therefore providing a safety switch in case of graft-versus-host disease (GVHD). These results warrant the clinical investigation of L-APC–expanded T cells modified with a suicide gene in the setting of haplo-HSCT.

Introduction

The adoptive transfer of leukemia-reactive T lymphocytes, also known as adoptive T-cell therapy (ACT), is a promising treatment modality for leukemia. The most convincing example of the therapeutic potential of ACT is allogeneic hematopoietic stem cell transplantation (HSCT),1 in which the transfer of donor T cells may induce long-term remission from chemoresistant leukemia, through the so-called graft-versus-leukemia (GVL) effect.

The GVL effect is a complex immunological phenomenon, in which the relative contribution of alloreactive T cells and T cells recognizing hematopoietic lineage-, overexpressed or truly leukemia-specific antigens (here collectively named leukemia-associated antigens or LAAs) is under intense debate. The role of alloreactivity has been confirmed at a peptide-specific level by documenting rising frequencies of circulating cytotoxic T lymphocytes specific for mismatched minor histocompatibility antigens during the GVL effect.2 Conversely, the post-transplant detection of circulating T cells specific for LAAs such as PR13,4 or WT1,4,5 has been shown to correlate with leukemia-free survival.

Although preferable to indiscriminately boost alloreactivity, which may unduly increase the risk of graft-versus-host disease (GVHD), ACT with leukemia-specific T cells poses the considerable challenge of generating sufficient numbers of effectors for subsequent infusion. While to date the indirect presentation of LAAs has been the strategy of choice, there is still considerable debate on the nature of the antigen-presenting cell (APC) and on the antigenic source to be employed to isolate leukemia-reactive effectors. Encouraging results were obtained by using dendritic cells (DCs) from patients6 or from HSCT donors7 after loading with apoptotic leukemic cells. Alternatively, leukemic cells themselves have been exploited as APCs for the direct presentation of LAAs. Leukemic cells from acute myeloid leukemia (AML) patients are poor APCs but, due to a common hematopoietic origin with DCs, after cytokine culturing can differentiate into functional DC-like cells, commonly referred as leukemic DCs.8 Unfortunately, the procedure requires a prolonged culture period (1–2 weeks) and is predictably successful in only a minority of cases (~30%).9 Moreover, the observation that, after cytokine culture, leukemic DCs downregulate WT110 and BCR-abl11 suggests that the process may subvert their antigenic repertoire, therefore jeopardizing the generation of T cells specific for LAAs expressed by the original leukemic cells. Forcing calcium influx in monocytes by pharmacological means reproducibly induces a rapid (1–2 days) and sustained acquisition of many DC features, including the upregulation of molecules involved in antigen presentation and costimulation, and the capacity to stimulate potent T-cell responses.12,13 By analogy, leukemic cells exposed to calcium ionophores (CIs) have been used to generate leukemia-reactive T cells both in the autologous setting14,15 and in the context of HSCT.15,16

Human leukocyte antigen (HLA)-mismatched HSCT from a family donor, which in the case of a full-haplotype mismatch is referred as HLA-haploidentical HSCT (haplo-HSCT), is increasingly applied to high-risk leukemia patients who do not have a readily available HLA-matched donor. We have demonstrated that T cells modified ex vivo to express the herpes simplex virus thymidine kinase (HSV-tk) suicide gene accelerate the immune reconstitution following haplo-HSCT and, in case of GVHD, can be eliminated with the prodrug ganciclovir (GCV).17,18 In haplo-HSCT, however, leukemia often escapes the immunological pressure of alloreactivity by losing the unshared HLA haplotype,19 suggesting that the use of donor T cells sensitized against LAAs restricted by shared HLAs may boost the GVL effect.

In this study, we evaluated the feasibility and the antileukemia efficacy of a strategy based on the ex vivo stimulation of donor T cells with CI-converted leukemic APCs (L-APCs) in the context of haplo-HSCT. Moreover, we explored the implementation of a suicide gene in order to eliminate L-APC–expanded T cells in case of GVHD.

Results

CI treatment efficiently converts leukemic cells into immunostimulatory APCs

We collected peripheral blood samples from 20 patients with AML. Patient demographics and disease characteristics are listed in Table 1. Twelve patients had de novo AML. Eight patients had AML secondary to either myelodysplastic syndrome (n = 6) or to previous chemotherapy for other causes (n = 2). Based on clinical parameters and cytogenetic abnormalities, all cases were classified as high-risk (data not shown). Importantly, only 6 out 20 (30%) cases expressed the CD14 molecule, a marker predictive for leukemic-DC differentiation upon cytokine culture.20

Table 1. Patient demographics and L-APC generation.

graphic file with name mt2012227t1.jpg

Open in a new tab

After a short-term exposure to the CI A23187 (48 hours), leukemic cells significantly upregulated the costimulatory molecules CD80, CD86, and CD54, and the antigen-presenting molecule HLA-DR (Figure 1a). Importantly, the expression levels of costimulatory molecules on CI-treated leukemic cells were higher than that of immature DCs from healthy donors, but lower than that measured on mature DCs. Differently from mature DCs, however, leukemic cells exposed to the CI failed to produce the immunosuppressive cytokine interleukin (IL)-10 (Supplementary Figure S1). The effects of the CI on the leukemic-cell phenotype, summarized as increased proportions of cells coexpressing CD80 and CD86, were observed in both de novo and secondary cases (Figure 1b). According to previous reports,14 the efficiency of DC-like conversion after CI treatment (17/19 cases, 89%) was higher than after culturing with granulocyte-macrophage colony-stimulating factor, IL-4, and tumor necrosis factor-α (3/8 cases, 37%, P < 0.01, Table 1) and independent from initial CD14 expression, suggesting a broad effect on multiple FAB (French-American-British classification) subtypes.

Figure 1.

Figure 1

Open in a new tab

Conversion of leukemic cells into leukemic antigen-presenting cells (L-APCs) upon exposure to a calcium ionophore. Leukemic cells from patients with acute myeloid leukemia were exposed for 48 hours to calcium ionophore (CI) and IL-4. (a) The expression of costimulatory (CD80, CD86, and CD54) and antigen-presenting (HLA-DR) molecules on untreated leukemic cells (AML, open circles), leukemic cells exposed to CI and IL-4 (AML+CI, closed circles), control immature DCs (iDC, open squares), and mature DCs (mDC, closed squares) was analyzed by flow cytometry. Results are expressed as MFI ratio (see Methods). Each symbol represents leukemic cells from a single AML patient (n = 19) or DCs from a healthy donor. Results from a paired (AML versus AML+CI) or unpaired (AML versus iDC or mDC) t-test are shown when statistically significant (*P < 0.05, **P < 0.01, ***P < 0.005). (b) Leukemic cells from patients with de novo AML (dnAML) or secondary AML (sAML) were grouped according to their origin. The percentages of leukemic cells coexpressing CD80 and CD86 were measured by flow cytometry in the two groups. Each symbol represents leukemic cells from a single patient (dnAML, n = 11; sAML, n = 8). Results from paired t-test are shown when statistically significant (*P < 0.05, **P < 0.01). (c) AML or AML+CI were irradiated and used as stimulators for the proliferation of allogeneic T lymphocytes at different stimulator:responder (S:R) ratios (x-axis). T-cell proliferation is expressed as stimulation index (y-axis, see Methods). Data from n = 11 patients are shown as means ± SEM. Results from a paired t-test are shown when statistically significant (**P < 0.01, ***P < 0.005). (d) mDC (gray bars), AML+CI (black bars), and iDC (white bars) were used as stimulators as detailed above. Data from AML patients (n = 11) and healthy donor DCs (n = 9) are shown as means ± SEM. Results from unpaired t-test are shown when statistically significant (*P < 0.05, **P < 0.01, ***P < 0.005). (e) The original leukemic cells (AML) and the derived L-APC were analyzed for the leukemia-associated immunophenotype (LAIP) by flow cytometry. Dot plots from a representative AML/L-APC pair depicting coexpression of CD80 (x-axis) and CD34 (y-axis) are shown (left). Histogram (right) shows means ± SEM measured on AML (white bars) and L-APC (black bars) from patients with a CD34− (n = 2) and CD34+ (n = 9) LAIP and patients with CD117− (n = 2) and CD117+ (n = 12) LAIP. (f) WT1 expression on AML (white bars) and L-APC (black bars) was analyzed by qPCR and expressed as WT1/abl copies per cell (y-axis, n = 8). Results on monocytes (Mo) and mature DC from a representative healthy donor are shown in comparison. AML, acute myeloid leukemia; DC, dendritic cells; IL, interleukin; MFI, mean fluorescence intensity; qPCR, quantitative PCR; UPN, unambiguous patient number.

The functional significance of DC-like conversion upon CI-treatment was confirmed by the fact that leukemic cells acquired an increased capacity to promote the proliferation of allogeneic T lymphocytes (Figure 1c). Importantly, when compared with DCs from healthy donors, the allostimulatory capacity of CI-treated leukemic cells was higher than that of immature DCs and comparable to that of mature DCs (Figure 1d). These results suggest that CI treatment efficiently converts leukemic cells into immunostimulatory APCs.

The LAA repertoire of leukemic cells is maintained upon conversion into L-APCs

We next investigated whether the rapidity of L-APC conversion upon CI treatment could associate with the maintenance of the LAA repertoire. We thus compared L-APCs with the original leukemic cells for the expression of surface and intracellular antigens by flow cytometry and quantitative PCR, respectively. The leukemia-associated immunophenotype was remarkably preserved, as L-APCs from CD34+ or CD117+ (c-kit) leukemic cells retained marker expression, while L-APCs from CD34− or CD117− leukemic cells remained negative (Figure 1e). Accordingly, the expression levels of the WT1 antigen, which is an important target for ACT,21,22 were also maintained or even increased (Figure 1f). Besides demonstrating the maintenance of the LAA repertoire, the preservation of WT1, CD34, and CD117 also confirmed the genuine leukemic origin of APCs, as neither monocytes nor DCs express these markers.

Besides unshared-HLA alloreactivity, HLA-haploidentical T cells expanded with L-APCs contain leukemia-associated specificities restricted by shared HLAs

With the aim of validating our strategy in the context of HSCT, we analyzed the capacity of L-APCs to promote the expansion of donor T cells. To this aim, T cells from either HLA-haploidentical donors (n = 6) or from HLA-matched donors (n = 3) were repetitively co-cultured with L-APCs or with the original leukemic cells as control. The HLA typing of all donor/patient pairs is reported in Supplementary Table S1. The expansion rates of HLA-haploidentical T cells co-cultured with L-APCs tended to be higher (at least 10× after three rounds of stimulation in 5/6 cases) compared with that of HLA-matched T cells (>10× expansion observed in 1/3 cases, P = 0.13, Figure 2a), suggesting an important role for alloreactivity.

Figure 2.

Figure 2

Open in a new tab

Expansion and leukemia reactivity of T cells stimulated with L-APCs. T lymphocytes from HSCT donors were repetitively co-cultured with patient leukemic cells (AML) or leukemic APCs (L-APC). (a) The expansion of T cells from HLA-matched (left panel, n = 3) and HLA-haploidentical (right panel, n = 6) donors was measured after three rounds of stimulation and expressed as fold increase. Results from a paired t-test are shown when statistically significant (*P < 0.05). (b) After expansion with L-APCs, T cells from HLA-matched (left histogram) or HLA-haploidentical (right histogram) donors were challenged with donor T cells (donor), PHA, patient T cells (host), and the original leukemia (AML) by IFN-γ ELISpot. Results are expressed as the mean IFN-γ spots/5 × 104 effector T cells. Results from a paired t-test are shown when statistically significant (**P < 0.01). (c) L-APC–expanded T cells were tested in a 51Cr-release assay for the ability to lyse the original leukemia in the presence (AML + unlabeled host) or absence (AML) of unlabeled patient T cells. Results of one representative experiment (n = 3) are shown. (d) T cells from HLA-haploidentical donors sharing the HLA-A2 allele (n = 3) were analyzed by flow cytometry for the presence of T cells specific for WT1, PRAME, PR1, and CMV before (pre L-APC) and after (post L-APC) stimulation with L-APCs. Multimer-binding cells are shown on CD8 T cells. Control multimers were used for background staining (control). Plots from a representative experiment are shown. AML, acute myeloid leukemia; APC, antigen-presenting cell; CMV, cytomegalovirus; E:T ratio, effector to target ratio; HLA, human leukocyte antigen; HSCT, hematopoietic stem cell transplantation; IFN, interferon; PHA, phytohemagglutinin.

To investigate the specificity of L-APC–expanded T cells, we studied the reactivity against different targets by interferon (IFN)-γ ELISpot analysis. L-APC–expanded HLA-haploidentical, and to a lesser extent, HLA-matched T cells specifically reacted against the original leukemic cells (Figure 2b). Importantly, L-APC–expanded HLA-haploidentical T cells reacted against leukemic cells consistently better than against host, nonleukemic targets (P < 0.01), suggesting that, besides alloreactivity to unshared HLAs, this population contains leukemia-associated specificities restricted by shared HLAs. Accordingly, in chromium-release assays, preincubating leukemic cells with an excess of host, nonleukemic targets did not abate cytotoxicity (Figure 2c).

To formally demonstrate that L-APC–expanded HLA-haploidentical T cells were enriched for leukemia-specific T cells, we selected three donor/patient pairs sharing HLA-A2 and measured the frequencies of different HLA-A2–restricted LAA-specific T cells before and after expansion. Following the procedure, WT1-specific CD8+ T cells were significantly enriched in 3/3 cases, whereas PRAME-specific CD8+ T cells in 1/3 cases. In all cases, control cytomegalovirus-specific CD8+ T cells, which at baseline were present at high frequencies, dropped to background levels after L-APC stimulation (Figure 2d), confirming that the enrichment was selective for leukemia-specific T cells.

HLA-haploidentical T cells expanded with L-APCs display a preferential CD8+ TCM phenotype and express IL-7Rα

The success of ACT strictly relies on the in vivo persistence and proliferation of infused T cells. These properties have been associated with the TCM phenotype, i.e., CD45RA−/CD62L+ cells,23,24 and with IL-7Rα expression.25 We therefore analyzed the phenotype of L-APC–expanded T cells. Importantly, LDC-expanded T cells were characterized by a preferential TCM phenotype (Figure 3a) and remarkably maintained IL-7Rα expression (Figure 3b). Of notice, although the majority of L-APC–expanded T cells were CD8+ (Figure 3c), major histocompatibility complex class I-blocking experiments and experiments with sorted T-cell subsets revealed that CD4+ T cells significantly contribute to leukemia reactivity (Supplementary Figure S2).

Figure 3.

Figure 3

Open in a new tab

Differentiation phenotype of T cells expanded with L-APCs. T lymphocytes from HLA-haploidentical donors (n = 5) were analyzed by flow cytometry before (pre L-APC) and after three rounds of stimulation with L-APCs (post L-APC). Representative plots (left) and histograms with mean ± SEM (right) are shown for: (a) the memory T-cell differentiation phenotype (left panel: x-axis, CD45RA, y-axis CD62L; right panel: percentage of CD45RA−/CD62L+ TCM cells), (b) IL-7Rα expression, and (c) CD4/CD8 ratio. Results from paired t-tests are shown when statistically significant (*P < 0.05, **P < 0.01). HLA, human leukocyte antigen; IL, interleukin; L-APC, leukemic antigen-presenting cell; TCM, central memory.

HLA-haploidentical T cells expanded with L-APCs have an antileukemic effect in vivo

To assess the antileukemia efficacy of HLA-haploidentical T cells expanded with L-APCs, we employed a clinically relevant xenograft mouse model of ACT.25,26,27 Starting from two donor/patient pairs, we repetitively stimulated HLA-haploidentical donor T cells with patient's L-APCs (L-APC/T cells). After expansion, reactivity to the original leukemia was confirmed by IFN-γ ELISpot (data not shown). As controls, we generated Epstein–Barr virus (EBV)-reactive T cells (EBV/T cells) by repetitive stimulation of T cells from the same donors with EBV lymphoblastoid cell lines. Immunodeficient mice previously infused with leukemic cells were given either saline or the respective L-APC/T cells or EBV/T cells. Due to the similar kinetics of in vivo leukemia progression (>1% by week 4), the results of the two donor/patient pairs were pooled. In agreement with their TCM phenotype, L-APC/T cells persisted in the circulation of mice for up to 11 weeks, whereas EBV/T cells rapidly disappeared (Figure 4a). T-cell persistence was accompanied by a preserved TCM phenotype and by continued expression of IL-7Rα in vivo (Figure 4b). L-APC/T cells were clearly superior to EBV/T cells in prolonging disease-free survival of mice (P < 0.0045), demonstrating a significant antileukemic effect in vivo (Figure 4c) The observation period could not be extended beyond week 12, due to GVHD occurrence, as we already reported with TCM cells.28

Figure 4.

Figure 4

Open in a new tab

In vivo persistence and antileukemic effect of HLA-haploidentical T cells expanded with L-APCs. Three days after the infusion of 5–10 × 106 leukemic cells from UPN#1 or UPN#18, immunodeficient mice received 10–20 × 106 T cells from the respective HLA-haploidentical donors expanded with either EBV LCLs (EBV/T cells, n = 6) or L-APCs (L-APC/T cells, n = 8) and followed for in vivo T-cell persistence and leukemia-free survival by flow cytometry. A cohort of leukemic mice received saline as negative control (n = 7). (a) Representative plots (x-axis, human CD45, y-axis, human CD3) of circulating human T cells measured in mice 11 weeks after the infusion of EBV/T cells or L-APC/T cells are shown on the left. Means ± SEM from all mice are shown on the right. Results from an unpaired t-test are reported (*P < 0.05). (b) Representative plots of the memory differentiation phenotype (x-axis, CD45RA; y-axis, CD62L) and of IL-7Rα expression on circulating L-APC/T cells at week 11 are shown on the left. Histograms with the means ± SEM at weeks 7, 9, and 11 are shown on the right. (c) A Kaplan-Meier survival curve of leukemia-free survival (y-axis) is shown along with the exact P values of a log-rank test comparing the different conditions (insert box). EBV, Epstein–Barr virus; HLA, human leukocyte antigen; IL, interleukin; LCL, lymphoblastoid cell line; L-APC, leukemic antigen-presenting cell; n.e., not evaluated; PBS, phosphate-buffered saline; TCM, central memory; UPN, unambiguous patient number.

Suicide-gene modification enables the conditional elimination of HLA-haploidentical T cells expanded with L-APCs

For translational purposes, it is important to match the antileukemia efficacy of our strategy with the possibility to control GVHD. To this aim, during L-APC co-culture, we exposed HLA-haploidentical T cells to a good manufacturing practice -grade retroviral (RV) supernatant encoding for the HSV-tk suicide gene along with the ΔLNGFR marker gene.17 A single exposure to the RV supernatant resulted in a transduction efficiency of 10–20%, and more than 95% purity was obtained after immunomagnetic sorting for the marker gene (Figure 5a). L-APC–expanded suicide gene-modified T cells retained specific cytotoxicity against the original leukemic cells, which was similar in the HSV-tk+ and HSV-tk− cellular fractions (Figure 5b). Differently from unmodified T cells, L-APC–expanded HSV-tk+ T cells acquired selective sensitivity to GCV, as demonstrated by dose-dependent killing observed upon T cell exposure to the prodrug for 7 days (Figure 5c). Importantly, leukemia-reactivity of L-APC–expanded HSV-tk+ T cells was superior to that of HSV-tk+ cells transduced after activation with OKT3 (Supplementary Figure S3), suggesting an improved GVL activity.

Figure 5.

Figure 5

Open in a new tab

Induction of a conditional suicidal phenotype after genetic modification of T cells expanded with L-APCs. Three days after the third round of stimulation with L-APC, T lymphocytes from HLA-haploidentical donors were transduced with a RV encoding for HSV-tk and ΔLNGFR (n = 2). (a) Representative plots of ΔLNGFR expression (x-axis, CD3; y-axis, LNGFR) after transduction (left) and immunomagnetic sorting of transduced (HSV-tk+) and untransduced (HSV-tk−) T cells are shown (right). Insert numbers represent the percentages of cells in the respective quadrant. (b) HSV-tk+ (left graph) and HSV-tk− (right graph) T cells were tested for the ability to lyse the original leukemic cells (AML) or donor T cells (donor) in a 51Cr-release assay. (c) Killing of HSV-tk+ (black bars) and HSV-tk− (white bars) T cells by increasing concentrations of the prodrug ganciclovir (GCV, x-axis) was assessed. Results are expressed as the relative percentage of survival (y-axis, see Methods). A representative experiment of n = 2 is shown. AML, acute myeloid leukemia; E:T ratio, effector to target ratio; HLA, human leukocyte antigen; HSV-tk, herpes simplex virus thymidine kinase; L-APC, leukemic antigen-presenting cell; RV, retroviral vector.

Discussion

In the present study, we report that stimulating HLA-haploidentical donor T cells with L-APCs is a very efficient procedure for the generation of high numbers of T cells reacting strongly against the original leukemic cells. L-APC stimulation primed and expanded not only alloreactive specificities to unshared HLAs, but also leukemia-associated specificities restricted by shared HLAs, including WT1 and PRAME. Importantly, L-APC–expanded T cells were enriched for CD8+ TCM cells that expressed high levels of IL-7Rα and, when infused in immunodeficient mice, persisted long-term and had a significant antileukemic effect.

Haplo-HSCT is a unique setting in which alloreactivity and leukemia-specific immune responses may cooperate to disease eradication.29 Besides directly participating to the GVL effect, alloreactivity may be crucial at lowering the priming/expansion threshold of LAA-specific T cells, which in vivo is possibly set very high. The majority of LAAs are indeed self-antigens and high-avidity LAA-specific T cells are either deleted in the thymus during negative selection30 or suppressed in the periphery by regulatory networks.31 In agreement with these assumptions, when stimulating HLA-haploidentical T cells with L-APCs, we found that rather than interfering, alloreactivity to unshared HLAs favored the enrichment of T cells specific for multiple, a priori undefined LAAs restricted by shared HLAs, including WT1 and PRAME. Since leukemia relapse following haplo-HSCT is often caused by the emergence of leukemic-cell variants that have lost the unshared HLA haplotype as a mechanism of immune escape,19 but have maintained the shared HLA haplotype, our strategy may greatly increase the GVL effect of the procedure.

The possibility to directly convert primary tumor cells into DC-like APCs is a peculiarity of AML and provides the unique opportunity for the generation of leukemia-reactive T cells without the need of “third player ” APCs. The majority of protocols, however, require prolonged culturing with differentiating cytokines9 and result in the downregulation of important LAAs, including WT1.10 On the contrary, after CI-treatment, we found a remarkably preserved antigenic repertoire and WT1 expression levels that were maintained or even increased compared with the original leukemic cells. Accordingly, stimulating HLA-haploidentical T cells with CI-converted L-APCs enriched for WT1-specific T cells. Since WT1 is abundantly expressed by CD34+/CD38− leukemic progenitor cells,32 it is tempting to assume that our strategy may specifically aid the generation of T cells targeting leukemia-initiating cells.

A major challenge for ACT is to counteract replicative senescence due to ex vivo manipulation required for generating sufficient numbers of tumor-reactive T cells. When compared to end-stage effectors, TCM cells are characterized by a higher proliferative potential in vitro33 and longer persistence in vivo.23,34 In a xenograft model of GVHD, we have demonstrated that the preservation of the TCM phenotype28 and of IL-7Rα 25,35 expression after ex vivo manipulation associates with the preservation of alloreactivity. By analogy, mouse models of ACT showed that tumor-specific T cells with a less-differentiated phenotype have a higher therapeutic potential than terminally differentiated T cells.36 We found that L-APC–expanded T cells were characterized by a predominant TCM phenotype and high IL-7Rα expression. In a clinically relevant xenograft model, these T cells persisted long term (10–12 weeks), maintained their phenotype and had a significant antileukemic effect.

The major drawback of exploiting alloreactivity for a therapeutic GVL effect is the occurrence of GVHD that, when severe, may jeopardize the clinical outcome of HSCT. We are aware that besides favoring the generation of LAA-specific T cells, stimulation of HLA-haploidentical T cells with L-APCs would predictably preserve, or even augment, the alloreactive potential of the final ACT product. We have therefore implemented a suicide gene as a platform for minimizing the hurdles of our strategy. As shown in elegant mouse studies, the in vivo activation of HSV-tk during the course of GVHD might selectively eliminate alloreactive T cells, while preserving leukemia-specific T cells.37,38 Accordingly, suicide gene therapy of GVHD in the clinic does not preclude the GVL effect.39,40 Alternatively, the suicide gene might be used for the selective depletion of alloreactive T cells ex vivo following a final round of stimulation with host, nonleukemic cells. In this work, we used a good manufacturing practice-grade RV supernatant, already validated in clinical trials,17 to facilitate clinical translation. Novel and promising suicide genes41 will be promptly implemented in the future.

Materials and Methods

Patient and donor samples. This study was approved by the Ethical Committee of our Institution. Nineteen patients with AML and eight HSCT donors were included after written informed consent. Patients' characteristics are listed in Table 1. The HLA typing of patient/donor pairs are reported in Supplementary Table S1. When indicated, peripheral blood mononuclear cells from patients, HSCT donors or healthy donors were isolated by density gradient centrifugation (Lymphoprep; Fresenius, Oslo, Norway) and cryopreserved until use.

Flow cytometry. T-cell and APC phenotype, transduction efficiency, and human chimerism in mouse blood were evaluated by flow cytometry. Cells were stained with FITC, PE, PerCp, PC7, APC or Pacific Blue conjugated mouse monoclonal antibodies (mAbs) (BD Biosciences, San Jose, CA) according to standard practice. Specificities were against human CD33, CD14, CD34, CD117, CD80, CD86, CD54, CD58, HLA-DR, CD3, CD56, CD4, CD8, CD45RA, CD62L, CD127, LNGFR, and mouse CD45. Isotype-matched mAbs were used as negative controls. For detection of antigen-specific T cells, PE-conjugated HLA-A2 dextramers (Immudex, Copenhagen, Denmark) with the following specificities were used: Wilms tumor antigen (WT1, 126-134), Proteinase-3 (PR1, 169-177), PRAME (300-309), and CMV (pp65). PE-conjugated HLA-A2 dextramers with no peptides were used as negative controls. Samples were analyzed with a FACS Calibur cytometer (BD Biosciences) or with a FACS Canto II flow cytometer (BD Biosciences). All data were analyzed with the Flow Jo software (TreeStar, Ashland, OR) and expressed as mean fluorescence intensity ratio. Mean fluorescence intensity ratio is the ratio between the mean fluorescence intensity of the sample and the corresponding isotype control. IL-10 release was analyzed by cytometric bead array assay according to manufacturer's instructions (BD Biosciences).

Generation of L-APCs and DCs. Leukemic cells were cultured in X-vivo-15 (BioWhittaker-Italia, Milan, Italy), while DCs from healthy donors in IMDM (Gibco-Brl, Gaithersburg, MD). In the case of leukemic cells <95%, blasts were sorted to purity by flow cytometry according to marker expression (see below). All media were supplemented with antibiotics, glutamine, and 10% human serum (HS Type AB; BioWhittaker). In the cytokine protocol, leukemic cells were cultured for 7–14 days with granulocyte-macrophage colony-stimulating factor (800 IU/ml; Immunotools, Friesoythe, Germany), IL-4 (500 IU/ml; Immunotools), and tumor necrosis factor-α (50 IU/ml, kindly gifted from Angelo Corti). In the CI protocol, leukemic cells were exposed for 2 days to the A23187 molecule (375 ng/ml; Sigma-Aldrich, St Louis, MO) and to IL-4 (250 IU/ml). Flow-cytometric analysis was performed on cells expressing disease markers (CD34, CD117, CD33). Successful conversion into leukemic APCs was defined as ≥20% of cells coexpressing CD80/CD86 as described.9 To obtain immature DCs from healthy donors, adherent monocytes were cultured for 4 days with granulocyte-macrophage colony-stimulating factor (800 IU/ml) and IL-4 (500 IU/ml). Mature DCs were obtained after 2-day exposure to irradiated mouse fibroblasts expressing human CD40L. DC10 were obtained by adding IL-10 (10 ng/ml) to the culture.42

Thymidine incorporation assay. Irradiated (30 Gy) L-APCs, immature DCs, and mature DCs were compared as stimulators (S) for the proliferation of 1 × 105 peripheral blood mononuclear cells from healthy donors (R) in 96-well round-bottom plates at increasing ratios. After 5 days, overnight incorporation 3H-thymidine (1 µCi; Amersham Biosciences, Piscataway, NJ) was analyzed in a liquid scintillation counter. Results are reported as stimulation index (counts per minute (cpm) of stimulated R cells/cpm of unstimulated R cells).

Molecular analysis of WT1 expression. Total RNA was extracted from 5 × 106 cells and retrotranscribed to cDNA. WT1 cDNA was quantified by end-point real-time PCR using the WT1 profileQUANT kit ELN (Ipsogen, Marseille, France) according to the manufacturer's instructions. Quantitative analysis of WT1 expression was calculated after interpolation with standard curve. The expression of WT1 was normalized for the expression of the abl control gene and expressed as WT1/abl copy number/cell.

Generation of leukemia-reactive T cells and suicide gene modification. Peripheral blood mononuclear cells from HSCT donors were co-cultured with irradiated L-APCs or leukemic cells (stimulator:responder ratio 1:10) in the presence of recombinant human IL-2 (25 IU/ml; Chiron, Emeryville, CA), IL-7 (5 ng/ml; Peprotech, Rocky Hill, NJ), and IL-15 (5 ng/ml; Peprotech). Cells were cultured with cytokines and restimulated with irradiated L-APCs or leukemic cells every 7–10 days. T-cell phenotype and T-cell reactivity against leukemic cells were tested after three rounds of stimulation (about 24 days of culture). When indicated, after the third round, T cells were resuspended in the SFCMM-3 RV supernatant (MolMed, Milan, Italy) and centrifuged at 2,400 rpm for 2 hours at 32 °C. The RV SFCMM-3 encodes for the HSV-tk suicide gene and a truncated form of the low-affinity receptor for the nerve growth factor (ΔLNGFR) as a selection marker.17 T cells expressing HSV-tk were sorted with LNGFR-specific immunomagnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany) and tested for susceptibility to cell death induced by increasing concentrations of the prodrug GCV after polyclonal stimulation. Surviving cells were counted by trypan blue exclusion and the survival percentage calculated as follows: number of living cells in GCV-treated samples/number of living cells in untreated samples × 100.

IFN-γ ELISpot and chromium-release assays. The ELISpot assay was performed according to standard protocol. Briefly, 1 × 105 effector T cells were plated with an equal number of target leukemic cells in 96-well flat-bottom plates (Immobilon-P; Millipore, Billerica, MA) pre-coated with an IFN-γ-specific primary mAb (anti-IFN-γ mAb1-D1K; BD PharMingen, San Diego, CA). After 24 hours, a secondary biotinylated anti-human IFN-γ mAb (mAb 7-B6-1; PharMingen, San Diego, CA) was added, followed by horseradish peroxidase-conjugated streptavidin and the 3-amino-9-ethylcarbazole substrate (PharMingen). After extensive washing, spots were quantified by computer-assisted video image analysis (Zeiss, Heidelberg, Germany). As positive control, effector T cells were stimulated with phytohemagglutinin (2 µg/ml; Sigma-Aldrich). When indicated, phytohemagglutinin T-cell lines of patient or donor origin were used as targets. The lytic ability of T cells was measured by a standard chromium-release assay. Briefly, 51Cr-labeled leukemic cells were co-incubated with T cells at different effector to target ratios and tested for chromium-release in the supernatant after 18 hours (Top Count; PerkinElmer, Waltham, MA). Results were expressed as percentage of specific lysis: 100 × (cpm experimental − cpm spontaneous 51Cr-release)/(cpm maximum 51Cr-release − cpm spontaneous 51Cr-release).

Xenograft model of the GVL effect. After approval from our Institutional Animal Care and Use Committee, 6–8 weeks old immunodeficient NSG mice or NOD/SCID mice pretreated with the rat anti-mouse IL-2/IL-15Rβ mAb for abating residual natural killer cell activity (TMβ1, 1 mg/mouse intraperitoneally), were infused intravenously with 5–10 × 106 leukemic cells from either UPN (unambiguous patient number) #1 or UPN#18. After 3 days, mice received 10–20 × 106 donor T cells generated with the respective patient's haploidentical L-APCs or with EBV-reactive T cells generated as described.43 Briefly, EBV-reactive T cells were generated after three rounds of stimulation with EBV-infected lymphoblastoid cell lines in the presence of IL-2 (100 IU/ml). Circulating T cells and leukemic cells were monitored by weekly bleeding and flow-cytometric analyses of the percentages of human CD45+/CD3+ cells and CD45+/CD33+ cells over total nucleated cells, respectively. When circulating leukemic cells exceeded 1%, mice were considered leukemic and killed to avoid further suffering. Mice were also killed when weight loss >10% due to xenogeneic GVHD.

Statistical analysis. Statistical analysis for comparison of cell distribution in the different phenotypic/functional subsets was performed with a two-tailed Student's t-test for paired and unpaired samples (Prism Software 5.0, GraphPad Software, San Diego, CA). Leukemia-free survival in mice was analyzed with a log-rank (Mantel–Cox) test (Prism).

SUPPLEMENTARY MATERIAL Figure S1. IL-10 production by L-APCs. Figure S2. Contribution of CD4+ and CD8+ L-APC–stimulated T cells to the antileukemic effect. Figure S3. Leukemia recognition by haplo T cells stimulated with OKT3 or L-APCs. Table S1. HLA typing of patients and HSCT donors.

Acknowledgments

This study was supported by the Italian Association for Cancer Research (Investigator grant to C.B. and Special Program Molecular Clinical Oncology, 5 per mille nr. 9965), the Italian Ministry of Health, the Italian Ministry of University and Research (FIRB-Ideas), and the Cariplo Foundation. C.B. is an employee of Molmed S.p.a, whose potential product (HSV-tk) is studied in this work. The other authors declared no conflict of interest.

Supplementary Material

Figure S1.

IL-10 production by L-APCs.

Click here for additional data file. (298.9KB, pdf)
Figure S2.

Contribution of CD4+ and CD8+ L-APC–stimulated T cells to the antileukemic effect.

Click here for additional data file. (378.6KB, pdf)
Figure S3.

Leukemia recognition by haplo T cells stimulated with OKT3 or L-APCs.

Click here for additional data file. (481.8KB, pdf)
Table S1.

HLA typing of patients and HSCT donors.

Click here for additional data file. (55.8KB, pdf)

REFERENCES

  1. Appelbaum FR. Haematopoietic cell transplantation as immunotherapy. Nature. 2001;411:385–389. doi: 10.1038/35077251. [DOI] [PubMed] [Google Scholar]
  2. Marijt WA, Heemskerk MH, Kloosterboer FM, Goulmy E, Kester MG, van der Hoorn MA.et al. (2003Hematopoiesis-restricted minor histocompatibility antigens HA-1- or HA-2-specific T cells can induce complete remissions of relapsed leukemia Proc Natl Acad Sci USA 1002742–2747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Molldrem JJ, Lee PP, Wang C, Champlin RE., and, Davis MM. A PR1-human leukocyte antigen-A2 tetramer can be used to isolate low-frequency cytotoxic T lymphocytes from healthy donors that selectively lyse chronic myelogenous leukemia. Cancer Res. 1999;59:2675–2681. [PubMed] [Google Scholar]
  4. Kapp M, Stevanovic S, Fick K, Tan SM, Loeffler J, Opitz A.et al. (2009CD8+ T-cell responses to tumor-associated antigens correlate with superior relapse-free survival after allo-SCT Bone Marrow Transplant 43399–410. [DOI] [PubMed] [Google Scholar]
  5. Rezvani K, Yong AS, Savani BN, Mielke S, Keyvanfar K, Gostick E.et al. (2007Graft-versus-leukemia effects associated with detectable Wilms tumor-1 specific T lymphocytes after allogeneic stem-cell transplantation for acute lymphoblastic leukemia Blood 1101924–1932. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Spisek R, Chevallier P, Morineau N, Milpied N, Avet-Loiseau H, Harousseau JL.et al. (2002Induction of leukemia-specific cytotoxic response by cross-presentation of late-apoptotic leukemic blasts by autologous dendritic cells of nonleukemic origin Cancer Res 622861–2868. [PubMed] [Google Scholar]
  7. Montagna D, Daudt L, Locatelli F, Montini E, Turin I, Lisini D.et al. (2006Single-cell cloning of human, donor-derived antileukemia T-cell lines for in vitro separation of graft-versus-leukemia effect from graft-versus-host reaction Cancer Res 667310–7316. [DOI] [PubMed] [Google Scholar]
  8. Cignetti A, Bryant E, Allione B, Vitale A, Foa R., and, Cheever MA. CD34(+) acute myeloid and lymphoid leukemic blasts can be induced to differentiate into dendritic cells. Blood. 1999;94:2048–2055. [PubMed] [Google Scholar]
  9. Cignetti A, Vallario A, Roato I, Circosta P, Allione B, Casorzo L.et al. (2004Leukemia-derived immature dendritic cells differentiate into functionally competent mature dendritic cells that efficiently stimulate T cell responses J Immunol 1732855–2865. [DOI] [PubMed] [Google Scholar]
  10. Li L, Reinhardt P, Schmitt A, Barth TF, Greiner J, Ringhoffer M.et al. (2005Dendritic cells generated from acute myeloid leukemia (AML) blasts maintain the expression of immunogenic leukemia associated antigens Cancer Immunol Immunother 54685–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Lindner I, Kharfan-Dabaja MA, Ayala E, Kolonias D, Carlson LM, Beazer-Barclay Y.et al. (2003Induced dendritic cell differentiation of chronic myeloid leukemia blasts is associated with down-regulation of BCR-ABL J Immunol 1711780–1791. [DOI] [PubMed] [Google Scholar]
  12. Koski GK, Schwartz GN, Weng DE, Czerniecki BJ, Carter C, Gress RE.et al. (1999Calcium mobilization in human myeloid cells results in acquisition of individual dendritic cell-like characteristics through discrete signaling pathways J Immunol 16382–92. [PubMed] [Google Scholar]
  13. Westers TM, Stam AG, Scheper RJ, Regelink JC, Nieuwint AW, Schuurhuis GJ.et al. (2003Rapid generation of antigen-presenting cells from leukaemic blasts in acute myeloid leukaemia Cancer Immunol Immunother 5217–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Houtenbos I, Westers TM, Dijkhuis A, de Gruijl TD, Ossenkoppele GJ., and, van de Loosdrecht AA. Leukemia-specific T-cell reactivity induced by leukemic dendritic cells is augmented by 4-1BB targeting. Clin Cancer Res. 2007;13:307–315. doi: 10.1158/1078-0432.CCR-06-1430. [DOI] [PubMed] [Google Scholar]
  15. Liepert A, Grabrucker C, Kremser A, Dreyssig J, Ansprenger C, Freudenreich M.et al. (2010Quality of T-cells after stimulation with leukemia-derived dendritic cells (DC) from patients with acute myeloid leukemia (AML) or myeloid dysplastic syndrome (MDS) is predictive for their leukemia cytotoxic potential Cell Immunol 26523–30. [DOI] [PubMed] [Google Scholar]
  16. Barbui AM, Borleri G, Conti E, Ciocca A, Salvi A, Micò C.et al. (2006Clinical grade expansion of CD45RA, CD45RO, and CD62L-positive T-cell lines from HLA-compatible donors: high cytotoxic potential against AML and ALL cells Exp Hematol 34475–485. [DOI] [PubMed] [Google Scholar]
  17. Ciceri F, Bonini C, Stanghellini MT, Bondanza A, Traversari C, Salomoni M.et al. (2009Infusion of suicide-gene-engineered donor lymphocytes after family haploidentical haemopoietic stem-cell transplantation for leukaemia (the TK007 trial): a non-randomised phase I-II study Lancet Oncol 10489–500. [DOI] [PubMed] [Google Scholar]
  18. Vago L, Oliveira G, Bondanza A, Noviello M, Soldati C, Ghio D.et al. (2012T-cell suicide gene therapy prompts thymic renewal in adults after hematopoietic stem cell transplantation Blood 1201820–1830. [DOI] [PubMed] [Google Scholar]
  19. Vago L, Perna SK, Zanussi M, Mazzi B, Barlassina C, Stanghellini MT.et al. (2009Loss of mismatched HLA in leukemia after stem-cell transplantation N Engl J Med 361478–488. [DOI] [PubMed] [Google Scholar]
  20. Houtenbos I, Westers TM, Ossenkoppele GJ., and, van de Loosdrecht AA. Identification of CD14 as a predictor for leukemic dendritic cell differentiation in acute myeloid leukemia. Leukemia. 2003;17:1683–4; author reply 1684; discussion 1685. doi: 10.1038/sj.leu.2403014. [DOI] [PubMed] [Google Scholar]
  21. Gao L, Xue SA, Hasserjian R, Cotter F, Kaeda J, Goldman JM.et al. (2003Human cytotoxic T lymphocytes specific for Wilms' tumor antigen-1 inhibit engraftment of leukemia-initiating stem cells in non-obese diabetic-severe combined immunodeficient recipients Transplantation 751429–1436. [DOI] [PubMed] [Google Scholar]
  22. Xue SA, Gao L, Hart D, Gillmore R, Qasim W, Thrasher A.et al. (2005Elimination of human leukemia cells in NOD/SCID mice by WT1-TCR gene-transduced human T cells Blood 1063062–3067. [DOI] [PubMed] [Google Scholar]
  23. Zheng H, Matte-Martone C, Jain D, McNiff J., and, Shlomchik WD. Central memory CD8+ T cells induce graft-versus-host disease and mediate graft-versus-leukemia. J Immunol. 2009;182:5938–5948. doi: 10.4049/jimmunol.0802212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Berger C, Jensen MC, Lansdorp PM, Gough M, Elliott C., and, Riddell SR. Adoptive transfer of effector CD8+ T cells derived from central memory cells establishes persistent T cell memory in primates. J Clin Invest. 2008;118:294–305. doi: 10.1172/JCI32103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Bondanza A, Hambach L, Aghai Z, Nijmeijer B, Kaneko S, Mastaglio S.et al. (2011IL-7 receptor expression identifies suicide gene-modified allospecific CD8+ T cells capable of self-renewal and differentiation into antileukemia effectors Blood 1176469–6478. [DOI] [PubMed] [Google Scholar]
  26. Nijmeijer BA, Mollevanger P, van Zelderen-Bhola SL, Kluin-Nelemans HC, Willemze R., and, Falkenburg JH. Monitoring of engraftment and progression of acute lymphoblastic leukemia in individual NOD/SCID mice. Exp Hematol. 2001;29:322–329. doi: 10.1016/s0301-472x(00)00669-x. [DOI] [PubMed] [Google Scholar]
  27. Nijmeijer BA, Willemze R., and, Falkenburg JH. An animal model for human cellular immunotherapy: specific eradication of human acute lymphoblastic leukemia by cytotoxic T lymphocytes in NOD/scid mice. Blood. 2002;100:654–660. doi: 10.1182/blood.v100.2.654. [DOI] [PubMed] [Google Scholar]
  28. Bondanza A, Valtolina V, Magnani Z, Ponzoni M, Fleischhauer K, Bonyhadi M.et al. (2006Suicide gene therapy of graft-versus-host disease induced by central memory human T lymphocytes Blood 1071828–1836. [DOI] [PubMed] [Google Scholar]
  29. Kolb HJ. Graft-versus-leukemia effects of transplantation and donor lymphocytes. Blood. 2008;112:4371–4383. doi: 10.1182/blood-2008-03-077974. [DOI] [PubMed] [Google Scholar]
  30. Sohn SJ, Thompson J., and, Winoto A. Apoptosis during negative selection of autoreactive thymocytes. Curr Opin Immunol. 2007;19:510–515. doi: 10.1016/j.coi.2007.06.001. [DOI] [PubMed] [Google Scholar]
  31. Mueller DL. Mechanisms maintaining peripheral tolerance. Nat Immunol. 2010;11:21–27. doi: 10.1038/ni.1817. [DOI] [PubMed] [Google Scholar]
  32. Inoue K, Ogawa H, Sonoda Y, Kimura T, Sakabe H, Oka Y.et al. (1997Aberrant overexpression of the Wilms tumor gene (WT1) in human leukemia Blood 891405–1412. [PubMed] [Google Scholar]
  33. Sallusto F, Lenig D, Förster R, Lipp M., and, Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999;401:708–712. doi: 10.1038/44385. [DOI] [PubMed] [Google Scholar]
  34. Gett AV, Sallusto F, Lanzavecchia A., and, Geginat J. T cell fitness determined by signal strength. Nat Immunol. 2003;4:355–360. doi: 10.1038/ni908. [DOI] [PubMed] [Google Scholar]
  35. Kaneko S, Mastaglio S, Bondanza A, Ponzoni M, Sanvito F, Aldrighetti L.et al. (2008IL-7 and IL-15 allow the generation of suicide gene-modified alloreactive self-renewing central memory human T lymphocytes Blood 1131006–1015. [DOI] [PubMed] [Google Scholar]
  36. Gattinoni L, Zhong XS, Palmer DC, Ji Y, Hinrichs CS, Yu Z.et al. (2009Wnt signaling arrests effector T cell differentiation and generates CD8+ memory stem cells Nat Med 15808–813. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Litvinova E, Maury S, Boyer O, Bruel S, Benard L, Boisserie G.et al. (2002Graft-versus-leukemia effect after suicide-gene-mediated control of graft-versus-host disease Blood 1002020–2025. [DOI] [PubMed] [Google Scholar]
  38. Kornblau SM, Aycox PG, Stephens C, McCue LD, Champlin RE., and, Marini FC. Control of graft-versus-host disease with maintenance of the graft-versus-leukemia effect in a murine allogeneic transplant model using retrovirally transduced murine suicidal lymphocytes. Exp Hematol. 2007;35:842–853. doi: 10.1016/j.exphem.2007.02.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ciceri F, Bonini C, Marktel S, Zappone E, Servida P, Bernardi M.et al. (2007Antitumor effects of HSV-TK-engineered donor lymphocytes after allogeneic stem-cell transplantation Blood 1094698–4707. [DOI] [PubMed] [Google Scholar]
  40. Traversari C, Marktel S, Magnani Z, Mangia P, Russo V, Ciceri F.et al. (2007The potential immunogenicity of the TK suicide gene does not prevent full clinical benefit associated with the use of TK-transduced donor lymphocytes in HSCT for hematologic malignancies Blood 1094708–4715. [DOI] [PubMed] [Google Scholar]
  41. Di Stasi A, Tey SK, Dotti G, Fujita Y, Kennedy-Nasser A, Martinez C.et al. (2011Inducible apoptosis as a safety switch for adoptive cell therapy N Engl J Med 3651673–1683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Gregori S, Tomasoni D, Pacciani V, Scirpoli M, Battaglia M, Magnani CF.et al. (2010Differentiation of type 1 T regulatory cells (Tr1) by tolerogenic DC-10 requires the IL-10-dependent ILT4/HLA-G pathway Blood 116935–944. [DOI] [PubMed] [Google Scholar]
  43. Savoldo B, Goss JA, Hammer MM, Zhang L, Lopez T, Gee AP.et al. (2006Treatment of solid organ transplant recipients with autologous Epstein Barr virus-specific cytotoxic T lymphocytes (CTLs) Blood 1082942–2949. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1.

IL-10 production by L-APCs.

Click here for additional data file. (298.9KB, pdf)
Figure S2.

Contribution of CD4+ and CD8+ L-APC–stimulated T cells to the antileukemic effect.

Click here for additional data file. (378.6KB, pdf)
Figure S3.

Leukemia recognition by haplo T cells stimulated with OKT3 or L-APCs.

Click here for additional data file. (481.8KB, pdf)
Table S1.

HLA typing of patients and HSCT donors.

Click here for additional data file. (55.8KB, pdf)

Articles from Molecular Therapy are provided here courtesy of The American Society of Gene & Cell Therapy

RESOURCES