Synthetic CD4⁺ T Cell–Targeted Antigen-Presenting Cells Elicit Protective Antitumor Responses

Introduction

CD4⁺ T-cell help is critical for the development and the maintenance of protective immune responses. Studies on mouse models of tumor disease ( 1– 3) and cancer patients ( 4– 7) have revealed that although tumor-specific CD4⁺ T-cell responses do arise naturally, various degrees of tolerance eventually develop ( 3, 8– 10). CD4⁺ T cell–directed vaccination has been attempted to promote responses and in some instances shown to augment antitumor responses ( 11– 16). However, cell-based vaccination strategies are currently limited by the need to derive autologous ( 17) and heterologous ( 18) antigen-presenting cells (APC) through lengthy good manufacturing procedures.

To overcome such limitations, synthetic cell-free artificial APC (aAPC) have been developed. These include microbeads coated with anti-CD3/CD28 stimulatory antibodies able to elicit polyclonal T-cell expansion of short-lived effector lymphocytes ( 19) or with whole tumor cell membranes (large multivalent immunogens) capable of eliciting tumor-specific effector CD8⁺ T-cell responses in vitro ( 20, 21). As such, large multivalent immunogens showed some efficacy in a human phase I trial ( 22). After the identification of several MHC class I–restricted tumor-associated antigens (Ag) and the availability of recombinant MHC-I molecules, Ag-specific CD8⁺ T cell–targeted aAPC were also developed by immobilizing MHC-I/peptide complexes on microbeads ( 20, 21, 23, 24) and liposomes ( 25, 26). Such aAPC induced specific CD8⁺ T-cell activation and proliferation in vitro and showed antitumor potential in vivo ( 27– 29).

Fewer attempts have been made to target the aAPC strategy to CD4⁺ T cells in vitro ( 24, 30), and no report exists about their applications in vivo. In recent studies, peptide-HLA–coated beads were used in vitro to expand influenza-specific CD4⁺ T lymphocytes from peripheral blood lymphocytes ( 24, 30), and the expansion was dependent on the presence of additional feeder cells, suggesting the need for costimulation. In the same study, virus-specific CD4⁺ T cells could be expanded from infected individuals and not healthy subjects after polyclonal activation and tetramer-specific sorting of memory CD4⁺ T cells ( 24, 30).

Thus, whether MHC-II–based aAPC might be sufficient to prime Ag-specific CD4⁺ T cells directly from a naive repertoire and to improve antitumor immunity in vivo remains to be determined. To address these issues, we developed synthetic microbead-based aAPC coated with peptide–MHC-II dimers [I-A^d/Leishmania receptor for activated C kinase (LACK); ref. 31] and anti-CD28 stimulating monoclonal antibody (mAb; L/28 aAPC). We report the ability of CD4⁺ T cell–directed aAPC to induce activation and proliferation of Ag-specific naive CD4⁺ T cells in primary T-cell cultures in vitro and in secondary lymphoid organs in vivo and to sustain protective responses against subcutaneous and lung tumors.

Materials and Methods

Proteins and aAPC preparation. I-A^d/LACK molecules consist of recombinant I-A^d–Fc fusion proteins covalently attached to immunodominant peptide from the Leishmania LACK antigen ( 31). Dimers were obtained from stably transfected Drosophila melanogaster cells and purified by protein G affinity chromatography and ionic exchange fast-protein liquid chromatography (Amersham Biotech) as previously described ( 31). aAPC were prepared by incubating 5 μm polystyrene-sulfate latex microparticles (Invitrogen; 20 × 10⁶/mL) in PBS containing different amounts of the proteins of interest: I-A^d/LACK dimers (0–100 μg/mL) or commercially available anti-CD3 (clone 145-2C11, eBioscience; 0.05 μg/mL) and/or anti-CD28 antibodies (clone 37.51, eBioscience; 2 μg/mL). Coating was allowed to occur for 20 min at 4°C on a rotating wheel. Microparticles were then washed twice in PBS with 2% FCS, and free sites were blocked by 30-min incubation at room temperature in complete RPMI with 5% FCS. The following aAPC were prepared: I-A^d/LACK dimers and anti-CD28 mAb (referred as L/28 aAPC), anti-CD3 and anti-CD28 mAb (referred as 3/28 aAPC), and anti-CD28 mAb only (referred as -/28 aAPC). Adsorption of the proteins onto the microparticle surface was monitored by flow cytometry by staining with a FITC-labeled goat anti-hamster mAb (Caltag Laboratories) and a PE-labeled anti–I-E^d/I-A^d mAb (BD Biosciences), which recognize the anti-CD28 mAb and I-A^d/LACK dimers, respectively.

Mice, tumor cells, in vivo immunization protocol, and BrdUrd administration. Seven-week-old to 8-week-old BALB/c mice were purchased from Charles River (Charles River Italia). DO11.10 ( 32) and 16.2β Transgenic (Tg; ref. 31) mice were bred in the specific pathogen–free facility of the institute. T lymphocytes derived from 16.2β mice express a Tg TCR β-chain specific for an I-A^d–restricted peptide (LACKp, FSPSLEHPIVVSGSWD) derived from for the Leishmania major–derived Ag LACK. As a result of pairing with endogenous TCR α-chains, only ∼0.5% of CD4 T cells in 16.2β Tg animals are LACK-specific and can be detected with I-A^d/LACK specific multimers (see below). T lymphocytes derived from DO11.10 mice express a Tg αβ-TCR specific for an I-A^d–restricted peptide (OVAp, ISQAVHAAHAEINEAGR) derived from chicken ovalbumin. As a result, 100% of CD4⁺ T cells are ovalbumin-specific.

TS/A-LACK tumor cells were previously described ( 3). Briefly, TS/A cells were transfected with a pcDNA3-derived vector encoding for a truncated form of the L. major-derived model Ag LACK and selected by limiting dilution of G-418–resistant clones ( 3). Tumor cells were grown in vitro at 37°C in complete medium (RPMI 1640 with 10% fetal bovine serum (FBS), 100 units/mL penicillin, 100 units/mL streptomycin, and 2.5 × 10⁻⁵ mol/L 2-ME, G-418 0.5 mg/mL; Invitrogen Life Technologies). In these cells, which do not express MHC-II molecules, LACK is expressed as a cytosolic protein and is representative of a tumor-associated Ag ( 3). Exponentially growing TS/A-LACK tumor cells were either s.c. (4 × 10⁵) or i.v. (2 × 10⁵) injected in 100 μL of PBS. The development of subcutaneous tumors was monitored over time as previously described ( 3). Lungs were recovered 20 d after tumor cell injection, and metastatic foci were dissected and counted.

Where indicated, mice were immunized i.v. with 20 × 10⁶ aAPC resuspended in 100 μL of PBS 20 d before tumor challenge, and boosted 10 d later. As control, mice were immunized s.c. with 2 × 10⁵ lipopolysaccharide (LPS)–matured bone marrow–derived peptide-pulsed dendritic cells (DC/LACK) prepared as previously described ( 3, 33). Briefly, bone marrow precursors were propagated for 7 d in complete Iscove's medium containing 25 ng/mL mouse rGM-CSF and 5 ng/mL mouse rIL-4 (Peprotech). Bone marrow–derived DCs were then allowed to mature in the presence of LPS (1 μg/mL; Sigma-Aldrich) for 8 h and pulsed for 1 h with 2 μmol/L class II–restricted immunodominant peptide LACK. DC maturation and purity were routinely evaluated by flow cytometry after staining with mAb recognizing CD11c, MHC class II, B7.1, B7.2, and CD40 molecules (all from BD Biosciences).

In selected experiments, BrdUrd (Sigma) was injected in the peritoneal cavity at time 0 (4 mg/mL, 300 μL of PBS) and provided in the drinking water (0.8 mg/mL) for the length of the experiment.

Five to 10 mice per group were used in each experiment. All the in vivo studies were approved by the Ethical Committee of the San Raffaele Scientific Institute and done according to its guidelines.

T-cell primary cultures. Mice were sacrificed and the axillary, brachial, and inguinal lymph nodes (LN) and the spleen were surgically excised. Organs were forced through a nylon screen to make single-cell suspensions, and cells were washed and resuspended in complete medium (RPMI with 5% FBS, 100 units/mL penicillin, 100 units/mL streptomycin, and 2.5 × 10⁻⁵ mol/L 2-ME; Invitrogen Life Technology). Blood was obtained by intracardiac puncture and immediately mixed with heparin (Sigma). Thereafter, blood cells were washed and resuspended in PBS and layered onto Lympholyte-M (2:1 ratio, Cedarlane) to separate mononuclear cells by centrifuging at 2,000 × g for 30 min at room temperature. The lymphocyte layer was then rescued and washed twice in PBS.

2 × 10⁶ LN, 3 × 10⁶ spleen, or 5 × 10⁶ blood-derived cells were plated (24-well plate, 1 mL) with 2.5 × 10⁶ aAPC or alternatively with the same number of irradiated syngeneic splenocytes in the presence of the I-A^d restricted peptides: OVA_323-339 (0.25 μmol/L) or LACK_158-173 (5 μmol/L). Where specified, cells were cultured with aAPC under Th1 or Th2 polarizing conditions respectively by the addition of murine interleukin 12 (IL-12; 5 ng/mL, Peprotech) and anti–IL-4 neutralizing mAb (2.5 μg/mL, clone 11B11, BD PharMingen) or murine IL-4 (5 ng/mL, Peprotech) and anti-IL-12 neutralizing mAb (2.5 μg/mL, clone C17.8, BD PharMingen). In addition, aAPC were combined with murine IL-7 (100 ng/mL, Peprotech).

In selected experiments, cells were labeled with the vital dye, 5-(and-6)-carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Invitrogen), final concentration of 1 μmol/L. Briefly, cells were washed twice with PBS and suspended at 20 × 10⁶/mL in PBS. An equal volume of CFSE at 2 μmol/L was then added. After an 8-min incubation under gentle shaking at room temperature, an equal volume of FCS was added to quench the reaction. Thereafter, the cells were washed twice in RPMI with 10% FCS and plated with aAPC as described above. T-cell proliferation was independently evaluated by [³H]yhymidine incorporation. Briefly, 2 × 10⁵ LN cells were cultured with 2.5 × 10⁵ aAPC in 96-well plate (200 μL complete medium). After 48 h, the cultures were pulsed with [³H]thymidine (1 μCi/well) and harvested after an additional 18 h.

I-A^d/LACK multimer staining and BrdUrd detection. I-A^d/LACK multimer staining was done as previously described ( 3). Briefly, I-A^d/LACK multimers were obtained by incubating I-A^d/LACK dimers (3 μg/sample) with Alexa 488–coupled protein A (Molecular Probes, Invitrogen; 0.3 μg/sample) in PBS for 30 min at RT. Free protein A–binding sites were saturated by the addition of total IgG (1 μg/sample). LN cells (6 × 10⁵) were first incubated with a blocking buffer (5% rat serum and 95% culture supernatant of 2.4G2 anti-FcR mAb-producing hybridoma cells) for 20 min to saturate the FcRs and then stained with I-A^d/LACK multimers for 1 h on ice in PBS supplemented with 0.5% bovine serum albumin. Thereafter, the cells were stained with PE-labeled or PerCP-labeled anti-CD4, anti-CD69, anti-CD44, and anti-CD62L mAb and with allophycocyanin-labeled anti-CD8a, anti-CD11b, and anti-B220 mAb (BD PharMingen). TOPRO-3 (1 nmol/L final concentration; Molecular Probes, Invitrogen) was added to the sample just before flow cytometric analysis to discriminate viable and dead cells. CD8a⁺CD11b⁺B220⁺TOPRO⁺ cells were excluded by electronic gating during the acquisition. Fifty to 100 × 10³ CD4⁺ T cells were acquired using a FACSCalibur flow cytometer (BD Biosciences).

BrdUrd incorporation was determined as previously described ( 34) by flow cytometry. In particular, after 3 d from aAPC immunization, mice were killed and cells were derived from lymphoid organs recovered as described above. Typically, 1 to 2 × 10⁶ cells were initially surface-stained with PerCP-labeled anti-CD4, PE-labeled anti-CD44, and APC-labeled anti-CD8a, anti-CD11b, and anti-B220 mAb (BD PharMingen). The cells were then fixed with 2% ice-cold formaldehyde, overnight at 4°C. The following day, cells were washed in PBS, permeabilized for 2 to 3 min at 4°C [0.1% Triton-X-100, 0.1% Na-Cytrate (pH 7.2)], washed again in PBS and then in 40 mmol/L Tris-HCl, 10 mmol/L NaCl, and 6 mmol/L MgC1₂ (pH 8.0). Thereafter, the cells were incubated for 1 h at 37°C in the same buffer containing 50 Kunitz units of DNase I (Sigma). After washing in PBS/0.5% Tween 20, FITC-labeled anti-BrdUrd or hamster-IgG-isotype control antibody (BD Biosciences) were added 30 min at RT, washed in PBS/0.5% Tween 20, and washed one last time in PBS. Fifty to 100 × 10³ CD4⁺CD8a⁻CD11b⁻B220⁻ events were collected in the viable gate. The background levels in the staining were measured for each singular sample with the hamster-IgG-isotype–stained control and subtracted in the absolute number calculations.

Cytokine secretion assays and ELISA. For intracellular cytokine detection, 5 × 10⁵ cells were stimulated with 5 × 10⁶ L/28 aAPC (I-A^d/LACK dimers, 20 μg/mL; anti-CD28 mAb, 2 μg/mL) or -/28 aAPC for 5 h at 37°C, of which the last two are in the presence of Brefeldin A (5 μg/mL, Sigma). Thereafter, the cells were surface-stained with PerCP-labeled anti-CD4 mAb, fixed in 2% formaldehyde, and permeabilized in PBS containing 2% FCS, 0.5% saponin, 2% rat serum, and 0.2% sodium azide (permeabilization buffer). The cells were then stained with anti–IL-2, anti–IL-4, anti–IFN-γ, anti–IL-10, anti–IL-17 mAb in permeabilization buffer for 30 min at RT in the darkness. Finally, the cells were washed in permeabilization buffer and then in PBS with 2% FCS. Fifty to 100 × 10³ CD4⁺ events were generally collected in the viable gate on a FACSCalibur. Cytokine release induced by L/28 aAPC was comparable with the one induced by LACK-pulsed syngeneic splenocytes (not shown).

In selected experiments, LN-derived cells (2 × 10⁵) were stimulated with the LACK_158-173 peptide (0–10 μmol/L) and irradiated syngeneic splenocytes (6 × 10⁵, 5,000 rad) in 96-well plate. IL-2 and IFN-γ were measured in 24-h culture supernatants by capture ELISA according to the protocol provided by the manufacturer (BD PharMingen).

Statistical analysis. Statistical analyses were done using unpaired two-tailed Student's t test.

Results and Discussion

Synthetic aAPCs suitable to prime antigen (Ag)–specific CD4⁺ T cells were developed by decorating polystyrene sulfate microspheres (5-μm diameter) with purified peptide–MHC-II dimers (I-A^d/LACK; ref. 31) and anti-CD28 stimulating mAb (clone 37.51; referred to as L/28 aAPC). Such aAPC should thus provide the two major signals needed for T-cell activation: MHC/TCR stimulation together with CD28 costimulation. The amount of Ag immobilized on the beads was controlled titrating the I-A^d/LACK dimers (0–100 μg/mL) used to coat a fixed number of beads. A fixed amount of anti-CD28 antibody (2 μg/mL) was used at the same time. Homogeneous protein adsorption on the beads was confirmed by flow cytometry analyses after staining with specific antibody to detect I-A^d/LACK and anti-CD28 mAb, respectively ( Fig. 1A , representative dot plots). Binding of I-A^d/LACK dimers increased up to coating concentration of 40 μg/mL and reached a plateau at higher concentration ( Fig. 1A, squares in graph). Coating of the anti-CD28 mAb was optimal in the presence of 5 to 20 μg/mL I-A^d/LACK dimers ( Fig. 1A, diamonds in graph).

The ability of L/28 aAPC to support Ag-specific naive T-cell proliferation was determined by the CFSE dilution assay ( 35). We compared CD4⁺ T-cell proliferation induced in cultures derived from 16.2β Tg mice (in which ∼0.5% of CD4⁺ T cells expresses a LACK-specific TCR; see Materials and Methods; ref. 31) and DO11.10 Tg mice (in which all the T cells express an ovalbumin-specific I-A^d–restricted TCR; ref. 32). LN-derived cells were stained with the CFSE vital dye and cultured with L/28 aAPC, and after 5 days CD4⁺ T-cell proliferation was revealed by the appearance of cells with a dimmer CFSE profile. CD4⁺ T-cell proliferation was detectable in cultures of 16.2β Tg mice ( Fig. 1B, middle) and was comparable with that induced by the nominal Ag (LACKp; Fig. 1B, bottom). DO11.10-derived CD4⁺ T cells did not expand in response to L/28 aAPC ( Fig. 1B, middle) but proliferated to the nominal Ag (OVAp; Fig. 1B, bottom). Thus, L/28 aAPC support the expansion of naive CD4⁺ T lymphocytes, and this is Ag-specific.

LACK-specific CD4⁺ T-cell proliferation, both determined by CFSE dilution (day 5; Fig. 1C) or [³H]thymidine incorporation (48 h; Fig. 1C, inset), was maximal in response to L/28 aAPC bearing optimal amounts of I-A^d/LACK dimers and anti-CD28 mAb (5–20 and 2 μg/mL, respectively; Fig. 1C compare with A). Of note, the coimmobilization of anti-CD28 mAb allowed to reduce considerably the amount of I-A^d/LACK dimers needed to elicit optimal T-cell proliferation.

The LACK peptide is covalently linked to the MHC β-chain of I-A^d/LACK dimers and accommodated within the Ag-binding groove in the MHC-II molecules ( 31). Nevertheless, to exclude the possibility that the peptide might be shed from the recombinant molecule and represented by the endogenous APC and verify that L/28 aAPC could directly present Ag to naive CD4⁺ T cells, cultures were set up with highly purified 16.2β CD4⁺ T lymphocytes. Also in this case, L/28 aAPC, but not -/28 APC, elicited the proliferation of a fraction of the CD4⁺ T cell, supporting the ability of peptide/MHC class II tetramers/anti-CD28 mAb aAPC to activate naive T cells ( Fig. 1D).

L/28 aAPC-driven T-cell expansion was accompanied by the enrichment of LACK-specific CD4⁺ T cells also enumerated by flow cytometry analyses using I-A^d/LACK multimer staining ( 3). I-A^d/LACK⁺ CD4⁺ T cells increased both in frequency ( Fig. 2A ) and in total number ( Fig. 2B) in L/28 aAPC or LACK peptide–driven 16.2β, but not DO11.10 LN-derived cultures at day 7 (compare white and black bars in Fig. 2B). In control -/28 and 3/28 aAPC-driven cultures, the frequency (not shown for 3/28 aAPC) and the absolute numbers of LACK-specific CD4⁺ T cells remained comparable with that found ex vivo ( Fig. 2A and B). Peptide stimulation in the presence of irradiated splenocytes as APC led to better yields of Ag-specific CD4⁺ T cells ( Fig. 2B), possibly reflecting the contribution of other accessory molecules, which can be found on the surface of such APC. However, as reported for Flu-specific Ag-primed CD4⁺ T cells ( 30), CD4⁺ T lymphocytes expanded by the nominal Ag presented by feeder cells were less prone to subsequent Ag-driven restimulations than aAPC expanded cells. ³ Thus, L/28 aAPC seems to be an effective and defined tool to increase LACK-specific CD4⁺ T cells by several times starting from a naive repertoire.

It was possible that this phenomenon was a particular feature of the Tg model used here, having a higher frequency of LACK-specific naive precursors. To answer this question, we did similar experiments using peripheral blood mononuclear cells (PBMC) of BALB/c mice, which have a normal T-cell repertoire. LACK-specific CD4⁺ T cells accumulated to sizable numbers in naive BALB/c blood-derived cultures ( Fig. 2C, bottom), whereas remained undetectable in control DO11.10 (ovalbumin restricted) cultures ( Fig. 2C, top). Altogether, these data indicate that L/28 aAPC can prime and expand LACK-restricted CD4⁺ T cells from both 16.2β Tg and normal (BALB/c) naive repertoires.

The surface and functional phenotype of aAPC-primed CD4⁺ T cells was next investigated. PBMCs of 16.2β mice were CFSE labeled and cultured for 6 days with -/28 aAPC or L/28 aAPC and analyzed by flow cytometry ( Fig. 2D) for the expression of activation markers. As expected, CD4⁺ T cells cultured with -/28 aAPC retained the original CFSE content, did not increase in size, and revealed CD4, CD44, CD62L, and CD25 (not shown) surface levels comparable with those found on naive animals ex vivo. In L/28 aAPC-driven cultures, a fraction of CD4⁺ T cells diluted its CFSE content, increased in size, and up-regulated the expression of CD4, CD44 ( Fig. 2D), CD69, and CD25 (not shown). Proliferating cells were distributed among CD62L^high and CD62L^low cells ( Fig. 2D), as also in the case of LACK peptide–primed lymphocytes (not shown).

The ability of cultured cells to produce cytokines was analyzed after control [-/28 or phorbol 12-myristate 13-acetate (PMA) and Ionomycin] or LACK-specific restimulation (L/28; Fig. 3 ). Upon control (-/28) restimulation, negligible amounts of cytokine secretion were found ( Fig. 3A). At difference, upon LACK-specific stimulation (L/28), a sizable frequency of CD4⁺ T cells in L/28 aAPC-driven and LACK peptide–driven cultures, but not in control -/28 aAPC–driven cultures, expressed IL-2 ( Fig. 3A). A small fraction of IL-2⁺/IFN-γ⁺ and IFN-γ⁺ CD4⁺ T cells was also found, whereas IL-4 secretion remained within background levels ( Fig. 3A and B). Thus, aAPC induce activation and proliferation of Ag-specific CD4⁺ T cells, without favoring their polarization. Nonetheless, as it might be relevant for both adoptive and active immunotherapy applications to generate Ag-specific CD4⁺ T cells with different helper capability, we investigated whether it was possible to combine L/28 aAPC with recombinant cytokines and favor CD4⁺ helper T-cell polarization. Again, in the presence of the L/28 aAPC alone, most of the LACK-specific CD4⁺ T cells produced only IL-2 upon LACK-specific stimulation (∼11% in Fig. 3B). However, when cultures were done with L/28 aAPC and recombinant IL-12 (Th1 polarization), LACK-specific IFN-γ–producing CD4⁺ T cells were instead induced (∼11%; Fig. 3B). Conversely, culturing the cells with L/28 aAPC and recombinant IL-4 (Th2 polarization) favored the differentiation of LACK-specific IL-4 producing CD4⁺ T cells ( Fig. 3B). This suggests that L/28-driven T-cell activation does not drive terminal differentiation of activated cells and rather preserve the lymphocyte susceptibility to exogenous cytokines, which drive their differentiation into Th1-like or Th2-like effectors. In addition, L/28 aAPC were combined with IL-7, a known survival factor for T cells ( 36), the accumulation of multifunctional IL-2⁺, IL-2⁺/IFN-γ⁺, and IFN-γ⁺ IL-4⁻ CD4⁺ T cells ( Fig. 3B). Of note, up to 50% of L/28 and L/28+IL-7 cultured CD4⁺ T cells expressing high levels of CD44 maintained the surface expression of the LN homing molecule CD62L ( Fig. 3C).

Taken together, the data indicate that synthetic peptide-MHC class II and anti-CD28 mAb-coated aAPC elicit the Ag-specific activation and expansion of CD4⁺ T cells from a naive normal T-cell repertoire and in combination with recombinant cytokines can be used to differentiate effector lymphocytes in vitro. Such versatility of aAPC might be particularly relevant in adoptive immunotherapy applications whenever multifunctional memory-like CD4⁺ T cells or unpolarized CD4⁺ T cells, which might have a superior therapeutic activity when compared with terminally differentiated effector T cells ( 37– 39) are needed.

We next investigated whether L/28 aAPC could elicit CD4⁺ T-cell responses in vivo in 16.2β Tg mice. This model system was previously exploited to trace L. major ( 31), tumor ( 3), and DC vaccine–induced CD4⁺ T-cell responses ( 10). Control, -/28, and L/28 aAPC (2 × 10⁷) were injected in the tail vein. BrdUrd was used to evaluate in vivo cell proliferation. LACK-pulsed bone marrow–derived mature DCs ( 3, 10) were injected into a separate group of mice to compare aAPC to professional APC. AAPC i.v. administration resulted in systemic redistribution as evidenced by beads accumulation in well-vascularized organs, such as the spleen and the lungs, but did not affect the health of the animals for the time of observation (not shown). Three days after injection, a group of mice was sacrificed and cells derived from the peripheral LN (axillary, brachial, and cervical) were analyzed by flow cytometry after staining with the I-A^d/LACK⁺ multimers, anti-CD4, anti-CD44, anti-CD69, and anti-CD62L ( Fig. 4 ). At day 3, the frequency of I-A^d/LACK⁺ CD4⁺ T cells increased in the peripheral LN of mice immunized with L/28 aAPC, whereas in control, mice immunized with -/28 aAPC remained comparable with the one found in naive mice ( Fig. 4A). Furthermore, a fraction of I-A^d/LACK⁺ CD4⁺ T cells in L/28 aAPC immunized mice increased in size, expressed high levels of CD69, CD44 ( Fig. 4B), and CD25 (not shown), and had down-regulated the surface expression of CD62L ( Fig. 4B). At day 3, L/28 aAPC-mediated CD4⁺ T-cell activation was comparable with that induced by LACK-loaded DC ( Fig. 4A and B). By this time, L/28 and not -/28 aAPC immunized mice also showed a higher frequency of CD4⁺ CD44^high T cells, which had incorporated BrdUrd after immunization ( Fig. 4C). These data strongly suggest that Ag-specific CD4⁺ T cells were activated (CD44^high) and divided (BrdUrd⁺) in L/28 aAPC immunized mice. Ten days after L/28 aAPC injection, the frequency of I-A^d/LACK⁺ CD4⁺ T cells had returned to baseline levels and was comparable with that found in control mice. However, a fraction of LACK-specific CD4⁺ T cells expressed high levels of CD44 indicative of previous Ag encounter ( Fig. 4D). To analyze the Ag-specific cytokine secretion profile, LN cells were stimulated with LACK peptide for 24 hours and cytokines were measured in culture supernatants by ELISA. LACK-specific IL-2 release was detectable in LN cultures of L/28 aAPC-LACK ( Fig. 5A) and DC-LACK day 3 immunized mice ( 3, 10), but not in those of naive and -/28 aAPC immunized mice ( Fig. 5A ). Similarly to what we found in vitro ( Fig. 3A and B), LACK-specific IFN-γ secretion was not found in culture supernatants of L/28 aAPC immunized mice ( Fig. 5B). This is different from DC/LACK immunized mice, which generally show sizable frequencies of IFN-γ–producing cells ( 3, 10). IL-4, IL-10, and IL-17 remained within background detection in all samples (not shown). By day 10, LACK-specific IL-2 secreting cells were no longer detectable (not shown), possibly due to its low frequency, in spite of the presence of CD44 high cells ( Fig. 4D).

Having determined that L/28 aAPC can activate LACK-specific CD4⁺ T cells in vivo, we investigated whether L/28 aAPC active vaccination might confer resistance to LACK-expressing TS/A tumors ( 3). BALB/c mice were immunized i.v. with -/28 aAPC or L/28 aAPC or s.c. with LACK-pulsed DC, boosted 10 days later, and challenged s.c. with 0.4 × 10⁶ TS/A-LACK cells. Whereas solid TS/A-LACK tumors rapidly developed in control, PBS injected (not shown), and -/28 aAPC immunized mice ( Fig. 6A ), tumor growth was delayed in L/28 aAPC and DC-LACK vaccinated mice ( Fig. 6A). By day 10, the volume of TS/A-LACK tumors in these mice was half of that found in control -/28 aAPC immunized mice ( Fig. 6A). Interestingly, depletion of CD4⁺ T cells after L/28 aAPC vaccination abolished the resistance to tumor growth, supporting a direct role for L/28 aAPC-activated CD4⁺ helper T cells in tumor protection ( Fig. 6A). By day 30, up to 40% of L/28 aAPC-immunized mice remained tumor-free (not shown) and survived up to 1 year in the absence of major clinical signs. It should be noted that, although encouraging, results obtained with L/28 aAPC were inferior to those obtained by DC-LACK immunization, which conferred long-term tumor-free survival in up to 80% of the mice ( 40).

As we found that aAPC localized also in the lungs after i.v. administration (not shown), we asked whether L/28 aAPC could protect against lung tumor metastasis. BALB/c mice were left untreated or immunized i.v. with -/28 aAPC or L/28 aAPC or s.c. with DC-LACK, boosted 10 days later, and challenged i.v. with 2 × 10⁵ TS/A-LACK cells. After i.v. injection TS/A-LACK tumor cells rapidly generated lung metastases in control (PBS) and -/28 aAPC treated mice ( Fig. 6B and C). In contrast, both L/28 aAPC and DC-LACK immunized mice proved resistant to TS/A-LACK lung tumor metastasis to a comparable extent ( Fig. 6B and C). Together, these data indicate that CD4⁺ T-cell targeted aAPC induce specific CD4⁺ T-cell activation and proliferation in vivo and support antitumor responses.

Concluding Remarks

We report here that latex bead–based aAPC presenting a single MHC class II peptide derived from a tumor-associated model antigen (LACK) in the context of CD28-mediated costimulation is able to activate Ag-specific CD4⁺ T cells, in vitro and in vivo, and also confer resistance to subcutaneous tumors and lung metastasis. Although less potent than bone marrow–derived DC, synthetic MHC-II/peptide-based aAPC might provide a suitable alternative to favor helper functions and the development of protective immune responses in vitro and in vivo. Such aAPC might provide the opportunity to bypass the need to derive autologous and heterologous APC under good manufacturing procedure conditions in clinic and obtain defined APC preparation (i.e., optimal Ag densities and costimulatory abilities) suitable for large-scale production. Latex-based microspheres had no significant side effects when given i.v. in mice ( 27, 29) and s.c. or i.d. in humans ( 22). Nevertheless, the issue of safety remains to be fully determined. The development of biodegradable supports and the increasing availability of defined MHC class I/class II–restricted tumor-specific peptides ( 41) and HLA-class I/class II recombinant molecules ( 42), together with the possibility of combining class I–restricted and class II–restricted antigenic determinant and different costimulatory ( 43, 44) and inhibitory ( 45) receptors, or immunomodulating agents (i.e., cytokines) might provide the unique possibility of generating “ad-hoc” and “off-the-shelf” reagents suitable for different clinical purposes.