Alloantigen specific deletion of primary human T cells by Fas ligand (CD95L)-transduced monocyte-derived killer-dendritic cells

Abstract

Numerous studies have been performed in vitro and in various animal models to modulate the interaction of dendritic cells (DC) and T cells by Fas (CD95/Apo-1) signalling to delete activated T cells via induction of activation-induced cell death (AICD). Previously, we could demonstrate that Fas ligand (FasL/CD95L)-expressing ‘killer-antigen-presenting cells’ can be generated from human monocyte-derived mature DC (mDC) using adenoviral gene transfer. To evaluate whether these FasL-expressing mDC (mDC-FasL) could eliminate alloreactive primary human T cells in vitro, co-culture experiments were performed. Proliferation of human T cells was markedly reduced in primary co-cultures with allogeneic mDC-FasL, whereas a strong proliferative T-cell response could be observed in co-cultures with enhanced green fluorescent protein-transduced mDC. Inhibition of T-cell proliferation was related to the transduction efficiency, and the numbers of mDC-FasL present in co-cultures. In addition, proliferation of pre-activated alloreactive CD4⁺ and CD8⁺ T cells could be almost completely inhibited in secondary co-cultures using mDC-FasL as stimulatory cells, which was the result of induction of apoptosis in the majority of preactivated T cells. The specific deletion of alloreactive T cells by mDC-FasL was confirmed by an unaffected proliferative response of surviving T cells towards allogeneic ‘third-party’ peripheral blood mononuclear cells in a third stimulation, or upon unspecific stimulation with anti-CD3/CD28 beads. The results of this study demonstrate that allospecifically activated T cells are efficiently eliminated by mDC-FasL, supporting further investigations to apply FasL-expressing ‘killer-DC’ as a novel strategy for the treatment of allograft rejection.

Introduction

It has been well established that immunologically naive T cells are most effectively activated by dendritic cells (DC). These cells are continuously produced by haematopoietic stem cells in the bone marrow, and are widely distributed as immature DC into both lymphoid and non-lymphoid tissues.^1–3 After antigen uptake and concomitant pattern recognition a maturation process begins, resulting in immunogenic mature DC (mDC), which express high levels of MHC and co-stimulatory molecules. Dendritic cells are unique in their ability to migrate from peripheral locations to T-cell areas within lymph nodes, where primary T-cell-mediated immune responses are initiated.^4,5 In addition, DC are the most potent presenters of alloantigens, inducing vigorous allospecific immune responses leading to allograft rejection.⁶

On the other hand, DC have also been implicated in maintaining peripheral tolerance. It has been demonstrated that repetitive stimulation of T cells by immature DC leads to anergy and tolerance rather than to T-cell activation.^7,8 However, mDC also have the ability to suppress T cells, leading to tolerance.^8–11 In addition, several different suppressive DC phenotypes have been identified, which have the potential for clinical applications.^12,13 Together, these results suggest that DC represent a heterogeneous cell population with different functions related to their anatomical localization. Furthermore, species-specific characteristics have been described.⁴ Therefore, different mechanisms are operative to influence DC function in vivo.

Numerous studies have demonstrated that activated T cells become sensitive to Fas ligand (CD95L)/Fas (CD95/Apo-1)-mediated apoptosis, allowing autocrine or paracrine apoptosis to occur, a process called activation-induced cell death (AICD).^14,15 It has been proposed that AICD is the mechanism responsible for the limitation of T-cell expansion at the end of an immune response in vivo, as well as for maintaining peripheral tolerance by elimination of autoreactive T cells that have escaped thymic selection.^16–18

To deplete activated T cells by AICD, FasL-expressing murine DC as well as antigen-presenting cells (APC) derived from cell lines, were established, and have been analysed in vitro and in vivo. However, the role of FasL-expressing DC in eliminating alloreactive T cells is still controversial when investigated in haplotype mismatched strains of mice: most groups observe specific deletion of activated T cells by FasL-expressing DC,^19,20 whereas others reported an increased proliferation of T cells.^21,22 More recently, depletion of HLA-A1-specific T cells has been observed following treatment with FasL-expressing APC derived from a Fas-negative mutant of the lymphoblastoid cell line while protective T-cell responses were maintained.²³

We demonstrated previously that FasL/CD95L-expressing killer-APC can be generated from human monocyte-derived mDC using adenoviral gene transfer.^24,25 To further characterize the immunoregulatory properties of these FasL-expressing killer-DC (mDC-FasL), primary and secondary co-cultures were established with primary human allogeneic T cells, and T-cell proliferation and apoptosis were analysed. Our results clearly demonstrate that mDC-FasL eliminate both activated CD4⁺ and CD8⁺ T cells in an alloantigen-specific manner without annihilating other ‘third-party’ T-cell responses. These findings support the concept of applying mDC-FasL as a potential novel treatment to modulate graft-versus-host disease (GvHD).

Materials and methods

Preparation and culture of DC and T cells

Peripheral blood mononuclear cells were isolated from leukapheresis concentrates of healthy donors by density gradient centrifugation and monocytes were separated by counter-current elutriation as previously described.²⁶ Elutriated monocytes were > 90% pure as determined by morphology and antigen expression (CD14, CD3, CD20) measured by flow cytometry. Monocytes were in vitro differentiated to DC for 5–7 days in serum-free CellGro culture media (CellGenix, Freiburg, Germany) supplemented with 500 U/ml interleukin-4 (IL-4; Promocell, Heidelberg, Germany) and 500 U/ml granulocyte–macrophage colony-stimulating factor (Sargramostim; Berlex, Seattle, WA). To induce maturation, DC were additionally stimulated with IL-1β (10 ng/ml), tumour necrosis factor-α (10 ng/ml), IL-6 (1000 U/ml; all from Promocell) and prostaglandin E₂ (1 μg/ml, Minprostin E₂; Pharmacia & Upjohn, Erlangen, Germany) for 2 days.²⁷ The T cells were also separated from mononuclear cells by counter-current elutriation and immediately frozen until further use. CD4⁺ and CD8⁺ T cell subpopulations were enriched using MACS® MultiSort beads (Miltenyi Biotec, Gladbach, Germany) and purity, as determined by flow cytometry, was routinely > 90%.

Recombinant adenoviruses and transduction of DC

The murine FasL was expressed in human mDC as described previously using a Cre/LoxP adenoviral expression system.^28–30 Furthermore, the recombinant adenovirus encoding enhanced green fluorescent protein (AdEGFP) was used as a control vector. Viruses were propagated in HEK 293 cells (Clontech, Heidelberg, Germany) and enriched by ultracentrifugation as described elsewhere.³⁰ The mDC were transduced 7–9 days after initiation of cultures. According to a previously published protocol,³¹ mDC were incubated at a concentration of ≤ 6 × 10⁶/ml serum-free CellGro culture media for 90 min with or without the various recombinant adenoviruses. A multiplicity of infection of 200 was determined as optimal and used throughout all experiments for single transductions, and a multiplicity of infection of 100 for each virus was used in double transduction experiments. Afterwards, cells were resuspended at a concentration of 0·5 × 10⁶/ml in fresh culture media containing 500 U/ml IL-4 and 500 U/ml granulocyte–macrophage colony-stimulating factor. Forty-eight hours after transduction, untreated and transduced mDC were used for further experiments. Expression of murine FasL on human mDC was determined by flow cytometry using phycoerythrin-conjugated clone Kay-10 (BD Pharmingen, Heidelberg, Germany).

FACS analysis

For detection of apoptosis, cells were washed twice with ice-cold PBS and stained simultaneously with FITC-conjugated Annexin V and propidium iodide (both Pharmingen) for 15 min on ice in the dark with a binding buffer containing 10 mm HEPES/NaOH, 140 mm NaCl and 2·5 mm CaCl₂. Within the next 30 min, cells were analysed for apoptosis using an EPICS XL/MCL (Beckman Coulter, Krefeld, Germany). Data were analysed with WinMDI (shareware Version 2.8, http://facs.scripps.edu/software.html).

Allogeneic mixed leucocyte reaction

For analysis of allospecific activation of T cells, re-stimulation experiments were performed. Primary mixed leucocyte reactions (MLR) were established as bulk cultures in 24-well or 48-well tissue culture plates (Becton Dickinson, Heidelberg, Germany). The T cells of one donor were stimulated with allogeneic mDC or peripheral blood mononuclear cells (PBMC; irradiated at 30 Gy) at a stimulator to responder ratio of 1 : 1 and seeded onto plates at 10⁶ T cells/ml in RPMI-1640 containing 5% autologous plasma or human AB serum. After 7 days T cells were rescued from primary MLR and numbers of living cells were determined by trypan blue exclusion (Sigma, Steinheim, Germany). Aliquots of viable alloreactive T cells were used for secondary MLR, established again as bulk cultures in 12-well plates (Becton Dickinson) with either mDC, mDC-EGFP or mDC-FasL from the primary allogeneic donor at a stimulator to responder ratio of 1 : 7·5. To maintain T cells without stimulator cells, 10⁶ T cells were seeded in RPMI-1640 supplemented with 30 U/ml human recombinant IL-2, and 5% autologous plasma or human AB serum. After 2 days T cells were rescued from secondary MLR and viable cells were subsequently used for a third MLR to determine their proliferative capacity upon different allospecific and polyclonal stimuli. For allogeneic stimulation in the third MLR, PBMC were pooled using equal parts from five different donors (PBMC_mix) and used as a standardized stimulator population in all experiments. To attenuate proliferation of PBMC_mix in MLR, cells were irradiated with 30 Gy. The MLR cultures were incubated in a humidified incubator at 37° and 5% CO₂.

[³H-Methyl]thymidine proliferation assay

Activation of T cells was determined as proliferation of responder cells and was monitored at each stage of the re-stimulation experiments. Therefore, 5 × 10⁴/well naive or rescued T cells were incubated with increasing numbers of allogeneic non-transduced mDC or transduced mDC (mDC-EGFP, mDC-FasL) at stimulator : responder ratios ranging from 1 : 625 to 1 : 1. To determine the remaining proliferative capacity of rescued T cells, additional stimulations were performed with anti-CD3/CD28 beads (Dynal/Invitrogen, Oslo, Norway) or PBMC_mix at stimulator : responder ratios ranging from 1 : 625 to 1 : 1. Cells were incubated in 96-well round-bottom tissue culture plates (Nunc, Roskilde, Denmark) in a total volume of 200 μl RPMI-1640 containing 5% autologous plasma or human AB serum in a humidified incubator (37°, 5% CO₂). Proliferation was determined by adding 1 μCi [³H-methyl]thymidine/well (Perkin Elmer, Boston, MA) for the last 20–24 hr as previously described. All samples were tested in triplicate and values represent means ± SD.

Results

Reduced proliferation of T cells in primary co-cultures with allogeneic FasL-expressing mDC

To analyse whether mDC-FasL could inhibit the proliferation of alloreactive T cells, mDC were first transduced simultaneously using AxCANCre and AdloxPFasL or AdEGFP, and transduction efficiency was determined after 48 hr by flow cytometry. Primary co-cultures of transduced mDC-FasL, mDC-EGFP or mock-treated mDC were then established with different numbers of primary human allogeneic T cells, and proliferation of T cells was determined after 6 days by [³H]thymidine incorporation. FACS analysis demonstrated highly efficient transduction of mDC by AdEGFP, as up to 88% of transduced mDC expressed EGFP. In contrast, different frequencies of FasL-expressing mDC could be observed following simultaneous transduction with AxCANCre and AdloxPFasL, ranging from 8 to 53%. However, an inhibition of > 50% of the T-cell proliferative response could be determined in primary co-cultures of T cells with mDC expressing high levels of FasL compared with co-cultures with mock-treated mDC or mDC-EGFP at an effector to target (E : T) ratio of 1 : 5 (Fig. 1a). Furthermore, inhibition of the T-cell proliferative response at an E : T ratio of 1 : 5, was dependent on FasL expression on transduced mDC. Thereby, increasing amounts of FasL-expressing mDC resulted in an increasing inhibition of allogeneic T-cell proliferative responses (Fig. 1b). Together these results demonstrate that mDC-FasL reduces proliferation of allogeneic T cells in allogeneic MLR, which was directly dependent on the level of FasL expression as well as on the E : T ratio (Fig. 1).

Figure 1.

Inhibition of T-cell proliferation is dependent on transduction efficiency of Fas ligand (FasL) expression by mature dendritic cells (mDC). (a) After transduction of mDC with AxCANCre and AdloxPFasL or adenovirus encoding enhanced green fluorescent protein (AdEGFP), transduction efficiency was determined after 48 hr using FACS and is shown by dotplots (lower panel). Subsequently, primary co-cultures of transduced mDC-FasL, mDC-EGFP or mock-treated mDC were established with increasing numbers of primary human allogeneic T cells, and proliferation of T cells was determined after 6 days by [³H]thymidine incorporation. All samples were tested in triplicate and values are indicated as means ± SD. Graphs are derived from one representative experiment out of > 10 independent experiments. (b) Inhibition of proliferation is directly dependent on FasL expression on mDC after transduction. Data were collected from secondary mixed leucocyte reactions at an mDC : T-cell ratio of 1 : 5. Inhibition of proliferation was calculated as 100 − proliferation_mDC–FasL × 100/proliferation_mDC = % inhibition of proliferation and FasL-expression determined by flow cytometry. Data of 27 independent experiments are displayed (r-value = 0·773).

Reduced proliferation of pre-activated T lymphocytes co-cultured with allogeneic mDC-FasL

Activated alloreactive T cells are critically involved in the rejection of allografts. Therefore, we analysed, whether mDC-FasL are able to eliminate previously activated alloreactive T cells in vitro. Co-cultures of mDC, mDC-EGFP or mDC-FasL were established with allogeneic T cells previously activated in a primary MLR with mDC of the same monocyte donor. Proliferation of T cells was determined after 48 hr. Proliferation of allospecifically activated T lymphocytes was markedly reduced in co-cultures with mDC-FasL, whereas a strong proliferative T-cell response could be observed in cultures with mDC-EGFP or untreated mDC. Inhibition of T-lymphocyte proliferation was directly related to the frequency of mDC-FasL present in co-cultures, and at an mDC-FasL to T-cell ratio of 1 : 1 proliferation was almost completely abolished (Fig. 2a). To determine whether inhibition of T-cell proliferation by allogeneic mDC-FasL was the result of apoptosis induction in responding T cells, FACS analysis was performed 48 hr after initiation of the secondary MLR at a DC : T cell ratio of 1 : 5. Annexin V/propidium iodide staining revealed strong induction of apoptosis in T lymphocytes co-cultured with mDC-FasL (82%), whereas only minimal T-cell apoptosis was observed in control cultures with mDC (20%) and mDC-EGFP (25%), which is comparable to background death of T cells cultured alone (Fig. 2b). These data demonstrate that pre-activated, alloreactive T cells can be efficiently eliminated by FasL-expressing ‘killer-DC’.

Figure 2.

Mature dendritic cell–Fas ligand (mDC-FasL) efficiently inhibits proliferation of allogeneic pre-activated T lymphocytes by induction of apoptosis. (a) T cells were rescued from a first mixed leucocyte reaction (MLR) with allogeneic mDC and co-cultured with mDC, mDC-enhanced green fluorescent protein (EGFP) or mDC-FasL of the same donor in a secondary MLR for 48 hr and proliferation of T cells was determined by [³H]thymidine incorporation. All samples were tested in triplicate and values are indicated as means ± SD. Results are derived from one representative experiment out of > 10 independent experiments. (b) To determine whether inhibition of T-cell proliferation by allogeneic mDC-FasL was the result of apoptosis induction, FACS analysis was performed 48 hr after initiation of the secondary MLR (DC : T-cell ratio 1 : 5). Frequencies of apoptotic cells were determined as described by Annexin V/propidium iodide staining. Numbers above each dotplot represent percentage of total cells. Results are derived from one representative experiment out of five independent experiments.

Reduced proliferation of pre-activated CD4⁺ and CD8⁺ T lymphocytes co-cultured with allogeneic mDC-FasL

Differences towards FasL/Fas-mediated apoptosis have been observed between CD4⁺ and CD8⁺ T cells.^32,33 To investigate whether differences in susceptibility to FasL-mediated cell death also occurred using allogeneic activated T-cell subsets by mDC-FasL, T cells were allospecifically activated and rescued, viable T cells were separated into CD4⁺ and CD8⁺ subpopulations before the secondary MLR. Proliferation of allospecific CD4⁺ and CD8⁺ T cells was significantly impaired after 48 hr following co-culture with mDC-FasL, whereas a strong proliferative T-cell response was observed in control cultures (Fig. 3a). To test the sensitivity upon Fas-mediated apoptosis in both subpopulations separately, frequencies of apoptotic CD4⁺ and CD8⁺ T cells were determined by FACS analysis 24 and 48 hr after initiation of the secondary MLR. The high proportion of apoptotic CD4⁺ and CD8⁺ T cells co-cultured with mDC-FasL, demonstrated that both T-cell subpopulations acquire similar sensitivity towards Fas-mediated apoptosis during allospecific activation. In contrast, apoptosis observed in control cultures was within the background level (T cells cultured in the absence of DC). Despite the elimination of pre-activated apoptotic CD4⁺ and CD8⁺ T cells, a substantial percentage of viable T cells remained in co-cultures with mDC-FasL (Fig. 3b). These data demonstrated that allogeneic activated CD4⁺ and CD8⁺ T cells are equally sensitive to elimination via FasL expressed by ‘killer-DC’.

Figure 3.

Reduced proliferation of pre-activated CD4⁺ and CD8⁺ T lymphocytes co-cultured with allogeneic mature dendritic cell–Fas ligand (mDC-FasL). (a) Allospecifically pre-activated and rescued, viable T cells were separated into CD4⁺ and CD8⁺ subpopulations before a secondary mixed leucocyte reaction (MLR) with mDC, mDC-enhanced green fluorescent protein (EGFP) and mDC-FasL. Proliferation of CD4⁺ and CD8⁺ T cells was determined by [³H]thymidine incorporation. All samples were tested in triplicate, and values are indicated as means ± SD. Results are derived from one representative experiment out of three independent experiments. (b) Reduced proliferation of T cells co-cultured with allogeneic mDC-FasL was the result of apoptosis induction as confirmed by FACS after 24 and 48 hr following initiation of the secondary MLR (DC : T-cell ratio 1 : 5). Data represent percentage of AnnexinV⁺/propidium iodide (PI)⁺ and AnnexinV⁺/PI⁻ gated T cells (means ± SD). Results are derived from three representative experiments out of three independent experiments.

Fas-mediated deletion of T cells by allogeneic FasL-expressing mDC is alloantigen specific

To analyse whether mDC-FasL eliminated T cells in an alloantigen-specific fashion, T lymphocytes of donor A (T cells_A) were first allospecifically stimulated in co-cultures with irradiated allogeneic mDC obtained from donor B (mDC_B) in a primary MLR. The T cells_A were rescued from primary MLR after 7 days, and co-cultured in a secondary MLR with allogeneic mDC-FasL obtained from donor B (mDC-FasL_B). After 48 hr, T cells_A were rescued again, and re-stimulated in a third MLR with either mDC_B, a mixture of PBMC obtained from five different donors (PBMC_mix) or anti-CD3/anti-CD28 beads. After 72 hr, proliferation of T cells_A was determined. A strong proliferative T-cell_A response was observed following stimulation with previously unknown donor PBMC_mix or anti-CD3/anti-CD28 beads. In contrast, proliferation remained within background levels in T cells_A re-stimulated with mDC_B, demonstrating that mDC-FasL_B eliminated T lymphocytes indeed in an alloantigen-specific manner during secondary MLR. In addition, these results indicate that treatment with mDC-FasL does not result in complete immunosuppression (Fig. 4).

Figure 4.

The suppression of T cells by mature dendritic cell–Fas ligand (mDC-FasL) is alloantigen specific. T lymphocytes of donor A (T cells_A) were stimulated with irradiated mDC obtained from donor B (mDC_B) for 7 days, and subsequently co-cultured with allogeneic mDC-FasL obtained from donor B in a secondary mixed leucocyte reaction. After 48 hr, T cells_A were rescued, and re-stimulated with either mDC_B, peripheral blood mononuclear cells obtained from five different donors (PBMC_mix), or anti-CD3/anti-CD28 beads. After 72 hr, proliferation of T cells_A was determined by [³H]thymidine incorporation. All samples were tested in triplicate, and values are indicated as means ± SD. Results are derived from one representative experiment out of five independent experiments.

Discussion

Alloreactive T cells specifically activated by APC are critically involved in the pathogenesis of allograft rejection and GvHD, respectively. Therefore, several strategies have been evolved to manipulate the interaction of APC and T cells to induce anergy, or to delete reactive T cells in an antigen-specific manner^13,34 as T cells acquire sensitivity to FasL-mediated apoptosis following activation. Within the body, activated T cells are eliminated at so-called immune privileged sides such as the eye, testes or ovaries, where epithelial cells express FasL.³⁵

Conflicting results were reported from studies using FasL-expressing DC in murine models because some investigators claimed antigen-specific elimination of activated T cells by FasL signalling, whereas others observed increased proliferation rather than deletion of activated T cells.^19–21 In addition, studies using primary human ‘killer-DC’ are still lacking. Therefore, experiments were performed using MLR cultures to analyse whether human CD4⁺ and CD8⁺ T cells could be allospecifically deleted by FasL-expressing mDC generated from primary human monocytes.

The present results demonstrate that proliferation of CD4⁺ and CD8⁺ T cells was significantly reduced in co-cultures with allogeneic ‘killer-DC’ because of induction of apoptosis. These results are consistent with previous reports demonstrating that T cells acquire sensitivity to FasL-mediated apoptosis following co-culture with allogeneic stimulator cells.²⁵ Previously, depletion of HLA-A1-specific T cells has been observed following treatment with FasL-expressing APC.²³ However, the APC used in this study were derived from a Fas-negative mutant of the lymphoblastoid cell line, limiting the clinical potential of this approach. In contrast, the present results were obtained from experiments using primary human monocytes to generate FasL-expressing mDC displaying the full set of the individual major and minor MHC. As a result of the profound MHC-mismatch, large numbers of allospecifically activated T cells were depleted by ‘killer-DC’. Importantly, elimination was not restricted to alloreactive CD8⁺ T cells but alloreactive CD4⁺ T cells were also efficiently removed by ‘killer-DC’. In contrast, irrelevant T cells apparently were not affected during co-cultures with ‘killer-DC’ because re-stimulation with allogeneic ‘third-party’ PBMC_mix, as well as polyclonal activation via anti-CD3/anti-CD28, resulted in a strong proliferative T-cell response demonstrating alloantigen-specific elimination of CD4⁺ and CD8⁺ T cells by ‘killer-DC’. Several different factors might have contributed to the slightly reduced proliferative T-cell response observed following stimulation with the PBMC_mix compared with the anti-CD3/anti-28 beads: Anti-CD3/anti-CD28 beads are potent polyclonal stimulators of T cells that result in the activation of virtually all T cells in the co-cultures. In contrast, the allogeneic PBMC_mix does not display such a magnitude of stimulatory signals because the numbers of professional APC providing strong co-stimulation are limited. In addition, HLA-typing of donors has not been performed, and therefore overlapping HLA molecules of the priming donor and the PBMC_mix donors were not excluded, diminishing the proliferative T-cell response. Taken together, these results confirm the observation that HLA-A1⁺ FasL-expressing APC delete HLA-A1-specific T cells while protective T-cell responses are maintained.²³ Furthermore, the elimination of around 50% of all pre-activated CD4⁺ or CD8⁺ T cells is in line with findings that demonstrate a frequency of 10–50% alloantigen responding T cells in vivo. Bystander activation and subsequent proliferation could be excluded, revealing an exquisite specificity of an alloresponse.^36,37

Recently, we have generated killer artificial APC (KaAPC) by coupling an apoptosis-inducing an a-Fas-IgM monoclonal antibody together with an HLA-A₂-immunoglobulin dimer molecules onto beads. In support of the recent findings, these KaAPC were able to deplete antigen-specific T cells in a Fas/FasL-dependent manner.³⁸ However, identification and characterization of the relevant autoantigeneic peptide as well as the HLA subtype of the patient are prerequisites to develop the KaAPC strategy as a potential treatment, which is not required, if ‘Killer-DC’ are used to delete alloreactive or autoreactive T cells. Thereby, host-derived ‘killer-DC’ might interact with allo-specific donor cells via direct and indirect pathways, resulting in the alloantigen-specific, Fas/FasL-dependent depletion of GvHD-inducing T cells. However, because of the high variability of FasL expression observed in transduced DC and the consequence of variable elimination rates of allo-specific T cells, it will be of great importance to ensure an effective ‘killer-DC’-phenotype. One might envision the controlled generation of functional host ‘killer-DC’, which will be phenotypically and functionally characterized before ex vivo purging of donor transplant products like haematopoietic stem cells, mobilized blood stem cells or bone marrow. The transplantation of stem cell products efficiently depleted from allo-reactive T cells ex vivo will most probably lead to a reduced morbidity and mortality of GvHD.³⁹

In the light of these considerations and the promising results obtained so far, numerous questions have to be addressed before considering ‘killer-DC’ as a potential treatment strategy. (i) To ensure a reproducible high level of DC functionality, the FasL expression should be optimized using matrix metalloproteinase (MMP) inhibitors,⁴⁰ exogenous coating of FasL²³ or MMP cleavage site-depleted FasL constructs as possible strategies. (ii) The role of FasL-expressing DC in the depletion and inhibition of allo-specific regulatory T cells and T helper type 17 T-cell function should be extensively studied before applying these strategies ex vivo or in vivo⁴¹ to ensure maximum efficiency. As appropriate function of regulatory T cells has been described as an important tolerance mechanism, elimination of regulatory T cells by ‘killer-DC’ might substantially diminish the treatment potential of the ‘killer-DC’ strategy. (iii) Finally, the role of a neutrophil-mediated inflammatory response related to enhanced FasL expression has to be further investigated in vivo.⁴²

Much research has demonstrated allo-specific deletion of activated T cells in mice using FasL-expressing DC^19,20 but here we demonstrate for the first time efficient allo-specific T-cell depletion by human ‘killer-DC’. This represents an important further step to their potential use in the treatment of allograft rejection and GvHD, although many further questions have still to be addressed in future studies.

Disclosures

The authors declare no competing financial interests.