Artificial antigen-presenting constructs efficiently stimulate minor histocompatibility antigen–specific cytotoxic T lymphocytes

The successful application of donor lymphocyte infusions for the treatment of relapsed leukemia after allogeneic stem cell transplantation illustrates the feasibility of adoptive immunotherapy of hematologic malignancies.¹  To minimize graft-versus-host disease, we earlier proposed the use of minor histocompatibility antigens (mHags) HA-1 and HA-2 as immunotherapeutic reagents.²  HA-1 and HA-2 display hematopoietic-restricted tissue distribution and relevant expression on leukemic cells and their progenitors.^3-6  Moreover, cytotoxic T lymphocytes (CTLs) directed against these mHags do not cause graft-versus-host disease in an ex vivo skin explant model,⁷  and coincide with complete remission of relapsed leukemia and multiple myeloma after HLA-matched HA-1– or HA-2–mismatched donor lymphocyte infusions.⁸ 

HA-1– and HA-2–specific CTLs can be generated in vitro using peptide-pulsed or mHag-transduced autologous dendritic cells (DCs) as antigen-presenting cells (APCs).^9,10  However, expansion of CTLs is difficult due to the limited availability of donor-derived DCs. To date, several reports indicate effective stimulation of T cells by HLA/peptide ligands expressed on artificial antigen-presenting constructs (aAPCs) such as liposomes^11,12  or microbeads.^13-16  We developed aAPCs that can be manufactured under good manufacturing practice conditions and used to expand mHag-specific CTLs for adoptive immunotherapy. Hereto, cell-sized latex microbeads coated with HLA-A2/HA-1 or HLA-A2/HA-2 complexes, CD80, and CD54 were investigated for their potential to efficiently stimulate HA-1– and HA-2–specific CTL clones and lines while keeping their antigen-specific cytolytic properties.

In vivo and in vitro generation of mHag-specific CTL clones, polyclonal CTL lines, and DCs is documented in detail elsewhere.^9,17  CD4⁺ T-helper cells (CD4⁺Th cells) were generated by culturing peripheral blood mononuclear cells (PBMCs) for 14 days in Iscove modified Dulbecco medium (IMDM) containing 10% pooled human serum (HS) and 0.5% standard Dutch diphtheria/pertussis/tetanus/polio vaccine (National Institute of Public Health and the Environment, Bilthoven, The Netherlands), and 20 U/mL interleukin 2 (IL-2; Cetus, Emeryville, CA) from day 6 onward.

All functional assays were performed in IMDM plus 10% HS. Stimulator cells were irradiated (30 Gy) before use. Results of proliferation, interferon-γ (IFN-γ) production, and cytotoxicity assays are expressed as the mean of duplicate samples.

Proliferation and IFN-γ production were determined by coculturing 2.5 × 10⁴ responders with 5 × 10⁴ stimulators. After 48 hours, supernatant was harvested for IFN-γ enzyme-linked immunosorbent assay (ELISA; Sanquin, Amsterdam, the Netherlands); the detection limit was 2 pg/mL. The remaining cultures were labeled with 1.0 μCi (0.037 MBq) ³H-thymidine for 16 hours, after which incorporation was measured using liquid scintillation counting.

Cytotoxicity was evaluated by incubating 2500 ⁵¹Cr-labeled target cells with serial dilutions of effector CTLs for 4 hours; supernatants were harvested for gamma counting: % specific lysis = (experimental release–spontaneous release)/(maximal release–spontaneous release) × 100%.

Cytokine profiles were tested using the Human Th1/Th2 cytokine cytometric bead assay (BD Biosciences, San Diego, CA). Detection limits were IFN-γ, 7.1 pg/mL; tumor necrosis factor α (TNF-α), 2.8 pg/mL; IL-10, 2.8 pg/mL; IL-5, 2.4 pg/mL; IL-4, 2.6 pg/mL; IL-2, 2.6 pg/mL.

HA-1 and HA-2 peptides were synthesized according to the reported sequences.^18,19  The biotinylated recombinant HLA-A2/HA-1 and HLA-A2/HA-2 complexes were generated as described and used as monomers for aAPC coating or as tetramers for analysis of HA-1– or HA-2–specific CTLs.²⁰ 

The 5.3-μm polystyrene sulfate latex beads (Interfacial Dynamics, Portland, OR) were incubated sequentially with streptavidin-allophycocyanin (1 μg/10⁷ beads; Molecular Probes, Leiden, The Netherlands), with recombinant human CD80/Fc-chimera and CD54 (0.5 μg and 1.5 μg/10⁷ beads, respectively; R&D Systems, Minneapolis, MN), and with 1% human albumin (Sanquin) for 20 minutes in 0.5 mL phosphate-buffered saline (PBS) per 10⁷ beads at 4° C. The beads were then incubated with biotinylated HLA-A2/mHag complexes (0.5 μg/10⁷ beads) for 40 minutes in 1 mL PBS/10⁷ beads at 4° C. After each incubation step, the beads were washed with 1 mL PBS. See the Supplementary Data on the Blood website (click on the “Data Set” link at the top of the online article) for determination of optimal ligand densities.

The capacity of aAPCs to stimulate mHag-specific CTLs in an efficient and ligand-specific manner was analyzed and compared with that of professional APCs. Three HA-1–specific CTL clones (2.12, 3HA15, 5W38) and 2 HA-2–specific CTL clones (1.7, 1.9) were stimulated with (1) aAPCs coated with the costimulatory molecules CD80 and CD54 only, (2) aAPCs coated with costimulatory molecules and either HLA-A2/HA-1 or HLA-A2/HA-2 complexes, or (3) with HLA-A2⁺ Epstein-Barr virus (EBV)–transformed B-cell lines naturally expressing HA-1 and HA-2 (HA-1⁺/HA-2⁺ EBV-LCLs; Figure 1A-B). All mHag-specific CTL clones showed significant proliferation and IFN-γ production in response to aAPCs coated with the relevant HLA-A2/mHag complexes and costimulatory molecules, comparable to the responses to HA-1⁺/HA-2⁺ EBV-LCLs. No proliferation or IFN-γ production was induced by aAPCs coated with irrelevant HLA-A2/mHag complexes or with costimulatory molecules only.

Improper or partial T-cell receptor signaling by altered peptide ligands or T-cell receptor engagement in the absence of adequate costimulatory signals can lead to changes in cytokine secretion profiles or anergy of T cells.^21-23  Similarly, aAPCs may alter the phenotype and function of CTL clones. Therefore, mHag-specific CTL clones (2.12 and 1.7) were stimulated for 7 days with aAPCs or with HA-1⁺/HA-2⁺ EBV-LCLs. Subsequently, cytotoxic activity and secretion of IFN-γ, TNF-α, IL-10, IL-5, IL-4, and IL-2 were measured. The specific cytotoxicity (Figure 1C) and the cytokine profiles (data not shown) were comparable for either mode of stimulation.

The main goal of this study was to investigate whether aAPCs can replace autologous DCs for the enrichment of polyclonal mHag-specific cultures. Hereto, 5 different polyclonal HA-1–specific CTL lines (1, 2, 3, 4, 5), originally induced and restimulated twice with HA-1 peptide-pulsed autologous DCs, were stimulated as follows. CTL lines 1, 2, 3, and 4 were restimulated for 7 days with (1) aAPCs only, (2) aAPCs and autologous PBMCs, (3) aAPCs and autologous CD4⁺Th cells, or (4) HA-1 peptide-pulsed autologous DCs. HA-1^A2 tetramer staining and cytotoxicity assays were performed at day 0 and at day 7 (Table 1). For all CTL lines, similar or improved enrichment for HA-1–specific CTLs was observed after restimulation with aAPCs and autologous feeder cells when compared to DC-mediated enrichment. The feeder cells could either be total PBMCs or activated CD4⁺Th cells. These responses were driven by antigen presentation via aAPCs because none of the CTL lines showed enrichment after 7 days of incubation with PBMCs or CD4⁺Th cells only (data not shown). In a separate experiment, CTL lines 4 and 5 were restimulated repeatedly with aAPCs and autologous feeder cells. Both lines showed a 5-fold increase in absolute numbers of HA-1^A2 tetramer–positive cells without loss of HA-1–specific cytotoxic activity after 14 days (Figure 2).

In summary, aAPCs coated with HLA-A2/mHag complexes, CD80, and CD54 can be used to selectively enrich mHag-specific CTLs for adoptive immunotherapy. In combination with autologous feeder cells, aAPCs stimulate CTLs as efficiently as peptide-pulsed DCs. The aAPCs can be generated in a nonlaborious way and can be easily removed from the cell culture by a density gradient. All aAPC-constituents are obtainable at clinical grade. Future studies will address the issue of whether aAPCs can be used for the primary in vitro induction of mHag-specific CTLs.

We wish to thank Prof F. Claas and Dr M. J. B. van Stipdonk for critical reading and comments.