Introduction
Dendritic cells (DCs) are antigen-presenting cells that regulate the immune response via interactions between major histocompatibility complex (MHC) molecules on DCs and T-cell receptors on T cells alongside costimulatory molecules (1). Cancer vaccines using DCs loaded with anti-genic peptides, tumor lysates, mRNA, DNA or whole tumor cells are an approved approach for enhancing a patient's own immune system to eradicate tumor cells (2). In clinical trials, MHC class I-restricted peptide-loaded DCs have been most commonly used to induce antigen-specific CD8+ cytotoxic T lymphocyte (CTL) responses and, in some cases, have been associated with clinical benefits; however, clinical efficacy has been limited (3,4). It has been reported that activation of CD4+ T helper type 1 (Th1) cell responses via MHC class II molecules on mature DCs (mDCs) is essential for supporting CD8+ CTL priming, memory formation, recruitment to tumor tissue and tumor cell recognition (5-7). Therefore, several long peptides targeting tumor-associated antigen (TAA)-specific CD4+ cells have been developed for the treatment of patients with cancer (5-10). Our previous study reported the clinical benefits of using MHC class I and II peptide-pulsed mDC vaccines to simultaneously induce TAA-specific CD4+ Th1 cells and CD8+ CTLs (4).
At present, numerous TAAs, such as Wilms' tumor 1 (WT1), HER-2/neu, gp100 and tyrosinase, have been isolated and incorporated into cancer vaccine preparations in the form of peptides (3-10). A promising target for cancer vaccines, WT1 was ranked as the top antigen among 75 TAAs due to its high immunogenicity (11). WT1 is highly expressed in various types of tumor and serves oncogenic roles in their formation (12-14). Additionally, WT1 has been successfully applied to induce both CD4+ Th1 cells and CD8+ CTLs (15,16). WT1-specific CD4+ Th1 cells contribute to the induction and maintenance of WT1-specific CD8+ CTLs by establishing the required cytokine milieu via cytokine secretion, including interleukin (IL)-2 and interferon-γ (IFN-γ) (15). In our previous study, mDCs were pulsed with three types of MHC class I- or class II-restricted WT1 peptide: i) 126-134 (amino acid sequence, RMFPNAPYL) for human leukocyte antigen (HLA)-A*02:01/02:06; ii) 235-243 (amino acid sequence, CYTWNQMNL) for HLA-A*24:02; and iii) 332-347 (amino acid sequence, KRYFKLSHLQMHSRKH) for MHC class II (4). Patients with advanced pancreatic ductal adenocarcinoma (PDA) were then treated with pulsed mDCs in combination with standard chemotherapy. In the clinical trial, 4 of 7 patients with PDA exhibited strongly positive WT1-specific immunity, resulting in prolongation of overall survival. However, WT1-specific immunity was not induced in the remaining 3 patients. Therefore, a novel WT1 helper peptide with increased immunogenicity may be essential for the induction and maintenance of antitumor immunity in clinical trials to generate more successful results.
Recently, a novel WT1 helper peptide (WT1 HP34-51; amino acid sequence, WAPVLDFAPPGASAYGSL) was identified (17). Of note, WT1 HP34-51 contains a killer WT1 peptide (WT1 KP37-45; amino acid sequence, VLDFAPPGA) that has affinity for HLA-A*02:01 (18). Accordingly, the novel peptide WT1 HP34-51 may induce not only WT1-specific CD4+ Th1 cells but also CD8+ CTLs simultaneously, at least partially, in an HLA-A*02:01-restricted manner. Thus far, the function of WT1 HP34-51 remains unclear. In the present study, the impact of mDCs pulsed with the novel peptide WT1 HP34-51 (mDC/WT1 HP34-51) on the induction of WT1-reactive antitumor immunity was assessed.
Materials and methods
Prediction of peptide-MHC class I binding or peptide-MHC class II molecules
WT1 peptide-MHC class I binding affinity was predicted by the NetMHC 4.0 server-prediction program (http://www.cbs.dtu.dk/services/NetMHC-4.0). Moreover, WT1 peptide-MHC class II binding affinity was predicted by the NetMHCII-2.3 server-prediction program (https://services.healthtech.dtu.dk/service.php?NetMHCII-2.3).
Patient characteristics
The present study was jointly reviewed and approved by the Ethics Committee of the Jikei Institutional Review Board, Jikei University School of Medicine and the Clinical Study Committee of Jikei University Kashiwa Hospital [approval no. 29-063 (8679)]. All patients provided written informed consent for the use of their samples in scientific research before samples were collected. To assess the stimulatory ability of mDC/WT1 HP34-51 for the induction of CD8+ CTLs restricted by HLA-A2, 14 patients (7 male, 7 female; aged 37-88 years) with various types of cancer (cancer of the oropharynx, lungs, esophagus, stomach, breast, pancreas, ovary, gallbladder, biliary duct, prostate or uterus) expressing an HLA-A2 allele [02:01 (n=5), 02:06 (n=7) or 02:07 (n=2)] were enrolled in the Tokyo Midtown Center for Advanced Medical Science and Technology between August 2017 and April 2019 (Table I).
Preparation of WT1-pulsed DCs
Peripheral blood mono-nuclear cells (PBMCs) were collected from the 14 patients, none of whom had received cancer vaccines, via leukapheresis and cryopreserved at the Tokyo Midtown Center for Advanced Medical Science and Technology. PBMCs were isolated using a Ficoll-Plaque Premium (Cytiva) density gradient solution, cryopreserved and stored for future experiments. After cryopreserved PBMCs were thawed and washed, their viability was assessed using trypan blue solution (Sigma-Aldrich; Merck KGaA). Trypan blue staining was performed according to the manufacturer's protocols. Cell viability was assessed under a CK40-F100 light microscope (magnification, ×100; Olympus Corporation). Subsequently, the cells were incubated in serum-free AIM-V medium (Gibco; Thermo Fisher Scientific, Inc.) in a 10-cm Primaria cell culture dish with surface-modified polystyrene for enhanced cell culture (Corning Inc.) under 5% CO2 at 37°C in a humidified incubator. After >30 min, plastic-adherent monocytes and nonadherent (NAD) cells were isolated by washing with gentle pipetting. NAD cells were cryopreserved in Bambanker solution (Nippon Genetics Co., Ltd.) and used for induction of WT1-reactive T cells. Adherent monocytes were cultured in serum-free AIM-V medium containing granulocyte macrophage colony-stimulating factor (GM-CSF; 50 ng/ml; Primmune Inc.) and IL-4 (50 ng/ml; R&D Systems, Inc.) under 5% CO2 at 37°C in a humidified incubator for 5 days to generate immature DCs (imDCs). Autologous imDCs were cryopreserved until they were pulsed with WT1 peptides or cocultured with NAD cells.
The viability and function of mDCs were not altered by whether fresh or cryopreserved PBMCs or imDCs were used for mDC production (19). Therefore, cryopreserved imDCs were used to prepare WT1-pulsed DCs in the present study. To generate WT1 peptide-pulsed autologous mDCs, cryopreserved imDCs (1×105) were thawed and incubated under 5% CO2 at 37°C in a humidified incubator for 24 h with 100 µg WT1 HP34-51 or WT1 KP37-45 (all peptides obtained from Greiner Bio-One GmbH) in the presence of GM-CSF (5 ng/ml), prostaglandin E2 (50 ng/ml; Daiichi Fine Chemical Co., Ltd.) and lyophilized preparations of a penicillin-killed, low-virulence strain of Streptococcus pyogenes (OK-432; 10 µg/ml; Chugai Pharmaceutical Co., Ltd.). Autologous imDCs (1×104) were also cultured with 25 µg WT1 HP34-51 or WT1 KP37-45 under 5% CO2 at 37°C in a humidified incubator for 24 h to generate WT1 peptide-pulsed imDCs, which were used as a second stimulator.
Phenotypes of DCs and NAD cells
To assess DC phenotypes, imDCs or mDCs (1×105/100 µl) were washed, incubated with Clear Back human Fc receptor blocking reagent (dilution 1:20; cat. no. MTG-001; MBL Life Science) for 5 min at 4°C and then stained with the following monoclonal antibodies (mAbs) at 4°C for 30 min: Fluorescein isothiocyanate (FITC)-conjugated anti-human HLA-ABC (dilution 1:20; cat. no. 311404; clone W6/32), anti-CD80 (dilution 1:20; cat. no. 305206; clone 2D10), anti-CD40 (dilution 1:20; cat. no. 334306; clone 5C3; all from BioLegend, Inc.); FITC-conjugated anti-CD14 (dilution 1:20; cat. no. 11-0149-42; clone 61D3; eBioscience; Thermo Fisher Scientific, Inc.); and phycoerythrin (PE)-conjugated anti-human HLA-DR (dilution 1:20; cat. no. 307606; clone L243), anti-CD11c (dilution 1:20; cat. no. 301606; clone 3.9), anti-CD83 (dilution 1:20; cat. no. 305308; clone HB15e) and anti-CD86 (dilution 1:20; cat. no. 305406; clone IT2.2; all from BioLegend, Inc.). In addition, to assess NAD pheno-types, NAD cells (1×105/100 µl) were washed, incubated with Clear Back human Fc receptor blocking reagent for 5 min at 4°C and then stained with the following mAbs at 4°C for 30 min: FITC-conjugated anti-human CD14 (dilution 1:20; cat. no. 11-0149-42; clone 61D3; eBioscience; Thermo Fisher Scientific, Inc.); FITC-conjugated anti-CD8 (dilution 1:60; cat. no. 301006; clone RPA-T8), PE-conjugated anti-human CD19 (dilution 1:40; cat. no. 302208; clone HIB19), allophycocyanin (APC)-conjugated anti-human CD56 (dilution 1:20; cat. no. 318310; clone HCD56) and APC/cyanine 7 (Cy7)-conjugated anti-human CD4 (dilution 1:60; cat. no. 317418; clone OKT4; all from BioLegend, Inc.). Phenotypes were analyzed using an Attune NxT flow cytometer (Thermo Fisher Scientific, Inc.) and FlowJo analysis software (version 7.6.5; Tree Star, Inc.). Prior to using the analyzer, 4 µg/ml propidium iodide (Sigma-Aldrich; Merck KGaA) was added to the samples to exclude dead cells at 4°C for ~5 min.
Induction of WT1-reactive CD4<sup>+</sup> and CD8<sup>+</sup> T cells
Cryopreserved NAD cells were thawed, washed and incubated under 5% CO2 at 37°C in a humidified incubator for 24 h in 24-well plates with RPMI-1640 medium (Sigma-Aldrich; Merck KGaA) supplemented with 1% minimum essential medium nonessential amino acids, 1 mM sodium pyruvate, 50 µM 2-mercaptoethanol (all from Gibco; Thermo Fisher Scientific, Inc.) and 10% heat-deactivated fetal calf serum (Cytiva) in the presence of IL-2 (20 U/ml; Shionogi & Co., Ltd.) and IL-7 (20 ng/ml; PeproTech, Inc.). The next day (day 1), WT1 HP34-51-pulsed mDCs (mDC/WT1 HP34-51; 1×105) or WT1 KP37-45-pulsed mDCs (mDC/WT1 KP37-45; 1×105) were cocultured with NAD cells (1×106) in RPMI-1640 medium containing 1% minimum essential medium nonessential amino acids, 1 mM sodium pyruvate, 50 µM 2-mercaptoethanol and 10% heat-deactivated fetal calf serum in the presence of low doses of IL-2 (10 U/ml) and IL-7 (10 ng/ml) under 5% CO2 at 37°C in a humidified incubator for 7 days. At day 8, the stimulated cells (1×106) were washed and cocultured again with freshly prepared mDC/WT1 HP34-51 (1×105) or mDC/WT1 KP37-45 (1×105) in the fresh culture medium containing IL-2 (10 U/ml) and IL-7 (10 ng/ml) for 7 days. To generate WT1-reactive T cells effectively, the stimulation was repeated in the same manner for a total of 4 times (day 1, 8, 15 and 22). In addition, as a control (no stimulation group), NAD cells (1×106) were maintained with IL-2 (10 U/ml) and IL-7 (10 ng/ml) in 7-day intervals for a total of 4 times under 5% CO2 at 37°C in a humidified incubator.
Microscopic analysis of cell clusters
On day 4 of the final 7-day simulation interval (day 26), cell cluster formation was examined under a EVOS™ XL core imaging system (magnification, ×4; cat. no. AMEX1000; Thermo Fisher Scientific, Inc.).
Detection of CD4<sup>+</sup> or CD8<sup>+</sup> T cells producing IFN-γ in response to WT1-KP<sub>37-45</sub>
To assess the impact of mDC/WT1 HP34-51 on the activation of WT1-specific T cells, WT1-reactive T cells (1×105) induced by 4 total stimulations with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 were restimulated once with WT1 HP34-51-pulsed imDCs (imDC/WT1 HP34-51), WT1 KP37-45-pulsed imDCs (imDC/WT1 KP37-45) or imDCs (1×104) alone in 96-well U-bottomed plates (Corning Inc.) under 5% CO2 at 37°C in a humidified incubator for 6 h using BD GolgiStop (BD Biosciences). After stimulation, the cells were washed, incubated with Clear Back human Fc receptor blocking reagent for 5 min at 4°C and then stained with a PE/Cy5-conjugated anti-human CD8 mAb (dilution 1:20; cat. no. 15-0088-71; clone RPA-T8; eBioscience; Thermo Fisher Scientific, Inc.) and an APC/Cy7-conjugated anti-human CD4 mAb (dilution 1:60; cat. no. 317418; clone OKT4; BioLegend, Inc.). Thereafter, they were washed, fixed at 4°C for 20 min and then permeabilized using a BD Cytofix/Cytoperm Plus Fixation/Permeabilization kit with BD GolgiStop (BD Biosciences). The cells were then incubated with an APC-conjugated anti-human IFN-γ mAb (dilution 1:25; cat. no. 506510; clone B27; BioLegend, Inc.). Lymphocytes were gated on an forward scatter-side scatter (FSC-SSC) dot plot, and CD4+ or CD8+ T cells were regated on a CD4/IFN-γ or CD8/IFN-γ dot plot, respectively. The percentages of CD4+IFN-γ+ cells and CD8+IFN-γ+ cells among the CD4+ or CD8+ T cell population, respectively, were analyzed using the same equipment and software as were used to analyze DC/NAD phenotypes.
WT1-specific memory cells in CD4<sup>+</sup> or CD8<sup>+</sup> T cell populations induced by mDC/WT1 HP34-51
To assess the memory cell subsets of activated WT1-reactive CD4+ or CD8+ T cells, NAD cells were first stimulated with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 for 4 times, followed by restimulation with imDC/WT1 KP37-45 at a ratio of 10:1 under 5% CO2 at 37°C in a humidified incubator. NAD cells without any stimulation were maintained and used as a control. After stimulation, the cells were washed, incubated with Clear Back human Fc receptor blocking reagent as previously described and then stained with an FITC-conjugated anti-human exon 4 splice variant of the tyrosine phosphatase (CD45RA) mAb (dilution 1:40; cat. no. 304106; clone HI100; BioLegend, Inc.), a PE-conjugated anti-human chemokine receptor type 7 (CCR7) mAb (dilution 1:10; cat. no. FAB-197P; clone 150503; R&D Systems, Inc.), or a PE/Cy5-conjugated anti-human CD8 mAb and an APC/Cy7-conjugated anti-human CD4 mAb at 4°C for 30 min. Lymphocytes were gated on an FSC-SSC dot plot, CD4+ or CD8+ T cells were regated, and CD45RA and CCR7 were then analyzed using the same equipment and software as were used to analyze DC/NAD phenotypes. Activated CD4+ or CD8+ T cells can be classified into 4 groups based on CD45RA and CCR7: CD45RA+CCR7+ naïve, CD45RA−CCR7+ central memory (CM), CD45RA−CCR7− effector memory (EM) and CD45RA+CCR7− terminally differentiated effector memory (EMRA) cells.
Statistical analysis
Data are presented as the mean ± SD. Comparisons of the difference in the percentage of cell surface markers between imDCs and mDCs were performed with a paired t-test or a Wilcoxon signed-rank test. P<0.05 was considered to indicate a statistically significant difference. Statistical analyses were performed using Microsoft Office Excel 2011 (Microsoft Corporation) with the add-in software Statcel 3 (OMS Publishing Inc.).
Results
Binding of WT1 peptides to MHC class I or MHC class II molecules
According to the NetMHC 4.0 server-prediction program, WT1 KP37-45 within WT1 HP34-51 exhibited high affinity for HLA-A*02:01 and A*02:07. Moreover, according to the NetMHCII-2.3 server-prediction program, WT1 HP34-51 and WT1 KP37-45 exhibited high affinity for MHC class II (data not shown).
Characterization of imDCs, mDCs or NAD cells
Frozen and stored PBMCs were used, and the yield of mDCs at the end of the production process was 2.47±1.86% (n=3). As not enough cells were available for patients 4 and 9, the immunophenotype results of the imDCs and mDCs from the 12 cases are presented in Tables SI and SII. Both imDCs and mDCs generated from 12 of the analyzed cancer patients displayed a characteristic phenotype comprising MHC class I (HLA-ABC), MHC class II (HLA-DR), CD14, CD11c, CD80, CD86, CD40 and CD83 expression (Fig. S1, Tables SI and SII). In addition, the percentages of mDCs expressing CD80, CD86, CD40 and CD83 were significantly increased compared with imDCs (P<0.001, P=0.002, P=0.021 and P<0.001, respectively), indicating that mDCs were activated by exposure of imDCs to OK-432 (Tables SI and SII). In contrast, the NAD cell composition (n=5) included lymphocytes (CD8+ cells, 27.14±8.44%; CD4+ cells, 54.49±10.38%; CD19+ cells, 0.78±0.42%; CD56+ cells, 9.80±3.86%) and monocytes (CD14+ cells, 2.24±1.20%; data not shown).
WT1-reactive T cells generated by mDC/WT1 HP<sub>34-51</sub> stimulation
To assess the stimulatory capacity of mDC/WT1 HP34-51 for induction of HLA-A*02:01-restricted CD8+ T cells, NAD cells from patient 4 (HLA-A*02:01/24:02; Table I) were first stimulated 4 times in 7-day intervals with mDC/WT1 HP34-51. As a control, NAD cells were also first cultured with mDC/WT1 KP37-45, or not stimulated but maintained with low doses of IL-2 and IL-7 4 times in 7-day intervals. The cells were then restimulated with imDC/WT1 HP34-51, imDC/WT1 KP37-45 or imDCs alone (Fig. 1A). At 4 days after the final stimulation with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 (Fig. 1A), cell clusters had formed (Fig. 1B). In contrast, there was no cluster formation by cells in the no stimulation group (Fig. 1B). These results indicated that both mDC/WT1 HP34-51 and mDC/WT1 KP37-45 exhibited the ability to stimulate these cells.
Then, it was evaluated as to whether mDC/WT1 HP34-51 induced WT1-reactive T cells with the functional capacity to produce IFN-γ by monitoring the activity of CD4+ and CD8+ T cells in response to imDC/WT1 HP34-51 (Fig. 2A). The percentages of CD4+IFN-γ+ T cells among total CD4+ T cells were 0.58% [0.29/(0.29+49.4)] and 0.58% [0.28/(0.28+48.4)] after restimulation with imDC/WT1 HP34-51 or imDCs alone, respectively (Fig. 2B). In contrast to CD4+ T cells, the percentages of CD8+IFN-γ+ T cells among total CD8+ T cells after restimulation with imDC/WT1 HP34-51 or imDCs alone were 7.26% [3.10/(3.10+39.6)] and 0.05% [0.02/(0.02+43.9)], respectively (Fig. 2B). These results suggested that mDC/WT1 HP34-51 possessed the ability to stimulate CD8+ T cells that can produce IFN-γ.
Whether mDC/WT1 HP34-51 had the capacity to induce IFN-γ-producing CD8+ T cells similar to that of mDC/WT1 KP37-45 was also evaluated (Fig. 3A). The percentages of CD4+IFN-γ+ T cells among total CD4+ T cells stimulated with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 or left unstimulated, all of which were subsequently restimulated once with imDC/WT1 KP37-45, were 0.68% [0.34/(0.34+49.3)], 1.00% [0.51/(0.51+50.3)], and 0.19% [0.08/(0.08+41.9)], respectively (Fig. 3B). Moreover, imDCs were used as a control for the second stimulator, and the percentages of CD4+IFN-γ+ T cells among total CD4+ T cells stimulated with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 following restimulation with imDCs alone were 0.58% [0.28/(0.28+48.4); Fig. 2B] and 0.68% [0.34/(0.34+49.6); Fig. S2], respectively. These results suggested that mDC/WT1 HP34-51 and mDC/WT1 KP37-45 exhibited a capacity to induce IFN-γ-producing CD4+ T cells after restimulation with imDC/WT1 KP37-45 when compared with no stimulation. In contrast to CD4+ T cells, the percentages of CD8+IFN-γ+ T cells among total CD8+ T cells stimulated 4 times with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 and restimulated once with imDC/WT1 KP37-45 were 7.96% [3.39/(3.39+39.2)] and 5.13% [2.64/(2.64+48.8)], respectively (Fig. 3B). Furthermore, the percentages of CD8+IFN-γ+ T cells among total CD8+ T cells stimulated with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 following restimulation with imDCs alone were 0.05% [0.02/(0.02+43.9); Fig. 2B] and 0.10% [0.05/(0.05+52.2); Fig. S2], respectively. In addition, 0.60% [0.30/(0.30+50.0)] of total CD8+ T cells were CD8+IFN-γ+ following no stimulation during 29 days of culture and then one stimulation with imDC/WT1 KP37-45 (Fig. 3B). Of note, CD8+ T cells stimulated 4 times with mDC/WT1 HP34-51 produced IFN-γ at higher levels than those stimulated 4 times with mDC/WT1 KP37-45 after one restimulation with imDC/WT1 KP37-45 (Fig. 3B). These results suggested that the combination treatment of mDC/WT1 HP34-51 and imDC/WT1 KP37-45 exhibited the capacity to induce IFN-γ-producing CD8+ T cells more frequently than that of mDC/WT1 KP37-45 and imDC/WT1 KP37-45.
WT1-specific memory cells in activated CD4<sup>+</sup> or CD8<sup>+</sup> T cells stimulated with mDC/WT1 HP<sub>34-51</sub>
As mDC/WT1 HP34-51 and mDC/WT1 KP37-45 stimulated WT1 KP37-45-reactive CD4+ T cells, patient 4 was selected for an analysis of the memory cell subsets of CD4+ T cells (Fig. 4A). The percentages of EM cells among total CD4+ T cells stimulated with mDC/WT1 HP34-51, mDC/WT1 KP37-45 or left unstimulated were 78.3, 73.7 and 30.8%, respectively (Fig. 4B and C). The percentages of EMRA cells among total CD4+ T cells stimulated with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 were 8.29 and 6.67%, respectively (Fig. 4B and C). These percentages were higher than those in the no stimulation group (3.54%). The results suggested that mDC/WT1 HP34-51 and mDC/WT1 KP37-45 induced EM and EMRA CD4+ T cells. Additionally, the percentage of naïve T cells among total CD4+ T cells was decreased after stimulation with mDC/WT1 HP34-51 or mDC/WT1 KP37-45 (Fig. 4B and C).
As mDC/WT1 HP34-51 and mDC/WT1 KP37-45 also stimulated WT1 KP37-45-reactive CD8+ T cells, patient 4 was also selected for an analysis of the memory cell subsets of CD8+ T cells (Fig. 5A). In contrast to CD4+ T cells, the percentages of EM cells among total CD8+ T cells stimulated with mDC/WT1 HP34-51, mDC/WT1 KP37-45 or left unstimulated after one restimulation with imDC/WT1 KP37-45 were 51.3, 53.2 and 56.4%, respectively (Fig. 5B and C). The percentages of EMRA cells among total CD8+ T cells stimulated with mDC/WT1 HP34-51, mDC/WT1 KP37-45 or left unstimulated following one restimulation with imDC/WT1 KP37-45 were 45.0, 42.4 and 37.6%, respectively (Fig. 5B and C). These results suggested that the frequencies of CD8+ memory T cell subsets from a patient with PDA showed no notable differences between the mDC/WT1 HP34-51, mDC/WT1 KP37-45 and no stimulation groups.
Induction of WT1-reactive CD4<sup>+</sup> or CD8<sup>+</sup> T cells derived from patients with different types of cancer by mDC/WT1 HP<sub>34-51</sub>
The ability of mDC/WT1 HP34-51 to induce WT1-reactive CD4+ or CD8+ T cells obtained from patients with different types of cancer and HLA-A2 alleles [02:01 (n=5), 02:06 (n=7), 02:07 (n=2)] was then subsequently assessed. For 10 (patients 1, 3, 5, 6, 7, 10, 11, 12, 13 and 14) of the 14 patients with cancer (71.4%), >1% of the total CD4+ T cell population expressed IFN-γ after four rounds of stimulation with mDC/WT1 HP34-51 followed by one restimulation with imDCs (Table II). Furthermore, for 10 (patients 1, 2, 4, 6, 7, 8, 9, 10, 13 and 14) of the 14 patients examined (71.4%), >10% of the total CD4+ T cell population expressed IFN-γ after four rounds of stimulation with mDC/WT1 HP34-51 and restimulation once with imDC/WT1 KP37-45, which was higher compared with the percentage after restimulation once with imDCs alone (Table II). In contrast to the results for WT1-reactive CD4+ T cells, >1% of the total CD8+ T cell population from 10 (patients 1, 3, 5, 6, 7, 8, 10, 11, 12 and 13) of the 14 patients (71.4%) examined produced IFN-γ after four rounds of stimulation with mDC/WT1 HP34-51 followed by one stimulation with imDCs (Table III). For 6 (patients 2, 3, 4, 7, 9 and 14) of the 14 patients with cancer (42.9%), a >10% increase in the IFN-γ-producing CD8+ T cell population among total CD8+ T cells was detected after restimulation once with imDC/WT1 KP37-45 compared with restimulation once with imDCs alone (Table III). WT1-reactive CD8+ T cells from patient 14 with the HLA-A*02:07 allele exhibited greater IFN-γ production than the cells from the other patient with the HLA-A*02:07 allele. In summary, for 5 of the 12 patients with cancer (41.7%) with the HLA-A*02:01 or HLA-A*02:06 allele, WT1-reactive CD8+ T cells stimulated with mDC/WT1 HP34-51 exhibited enhanced levels of WT1 KP37-45-induced IFN-γ production, with a >10% increase following one restimulation with imDC/WT1 KP37-45 in this experimental setting. These results suggested that mDC/WT1 HP34-51 generated from patients with different types of cancer may induce WT1-specific CD8+ T cells at least in part in an HLA-A*02:01-, HLA-A*02:06- or HLA-A*02:07-restricted manner.
Discussion
The present study indicated that mDC/WT1 HP34-51 exhibited the ability to induce not only WT1-reactive EM CD4+ Th1 cells but also CD8+ CTLs producing IFN-γ in an HLA-A*02:01- or HLA-A*02:06-restricted manner in an in vitro human model. The combination of mDC/WT1 HP34-51 and imDC/WT1 KP37-45 induced higher levels of IFN-γ-producing CD8+ T cells than stimulation with mDC/WT1 KP37-45 and imDC/WT1 KP37-45.
The impact of mDC/WT1 HP34-51 on the activation of WT1-specific T cells was assessed by monitoring IFN-γ production following restimulation with imDC/WT1 HP34-51 or imDC/WT1 KP37-45. As restimulation with mDCs alone induced a degree of nonspecific IFN-γ production in the present study, WT1-loaded or unloaded imDCs were used as restimulators. CD4+ and CD8+ T cells were stimulated with mDC/WT1 HP34-51 together at the same time, resulting in IFN-γ production after restimulation with imDC/WT1 HP34-51 or imDC/WT1 KP37-45, but not imDCs alone. Our previous phase I study with mDCs pulsed with a mixture of three WT1 peptides, including WT1 HP332-347 and WT1 KP, observed that simultaneous induction of WT1-specific CD4+ and CD8+ T cells was associated with improved long-term survival (4). The novel peptide WT1 HP34-51 was subsequently developed to enhance WT1 peptide immunogenicity, and early clinical trials of a cocktail WT1 peptide vaccine containing the novel peptide WT1 HP34-51 and WT1 KP have been conducted in patients with pediatric brain cancer or myelodysplastic syndromes (20,21). The results demonstrated WT1-specific CTL induction in clinical responders, suggesting that WT1 HP34-51 has effective activity. Moreover, vaccination of a mouse model with both WT1 HP34-51 and WT1 KP also induced strong infiltration of WT1-specific CD4+ Th1 cells and CD8+ CTLs into tumor microenvironments (17).
In general, simultaneous activation of WT1-specific CD4+ Th1 cells and CD8+ CTLs may be essential when treating patients with advanced-stage cancer. Notably, the novel WT1 HP34-51 amino acid sequence includes WT137-45, which has binding affinity for HLA-A*02:01 molecules (18). Therefore, the novel peptide WT1 HP34-51 may stimulate not only WT1-specific CD4+ T cells but also CD8+ T cells in a manner, at least in part, restricted by HLA-A*02:01 molecules. To evaluate this hypothesis, a patient with PDA possessing HLA-A*02:01/24:02 alleles was initially selected for further analysis. Notably, mDC/WT1 HP34-51 activated not only CD4+ T cells but also CD8+ T cells to produce IFN-γ after restimulation with imDC/WT1 KP37-45, which according to the NetMHC 4.0 server-prediction program has a high affinity for HLA-A*02:01 but not for HLA-A*24:02. Thus, WT1-specific CD8+ T cells from patient 4 induced by mDC/WT1 HP34-51 may be HLA-A*02:01-restricted.
Stimulation of T cells by 4 incubations with mDC/WT1 HP34-51, followed by one restimulation with imDC/WT1 HP34-51 or imDC/WT1 KP37-45 resulted in IFN-γ production by CD8+ T cells. When imDCs are pulsed with WT1 HP34-51, WT1 HP34-51 may be taken up by imDCs, which can be activated by OK-432. Both imDCs and mDCs may be able to process WT1 HP34-51 and load WT1 KP37-45 onto HLA-A*02:01 molecules in the endoplasmic reticulum, resulting in their surface expression of WT1 KP37-45/HLA-A*02:01 complexes for presentation to CD8+ T cells. Endogenously processed WT1 in DCs has relatively good access to MHC class I and II molecules, resulting in the efficient induction of antitumor immunity (22). Thus, mDC/WT1 HP34-51, imDC/WT1 HP34-51 or imDC/WT1 KP37-45 may also stimulate CD8+ T cells through WT1 KP37-45/HLA-A*02:01 complexes, resulting in the production of IFN-γ.
Indeed, a >10% increase in IFN-γ-producing cells among total CD8+ T cells was detected in 3 of 5 patients with the HLA-A*02:01 allele and 2 of 7 patients with the HLA-A*02:06 allele when their NAD cells were stimulated 4 times with mDC/WT1 HP34-51 followed by restimulation with imDC/WT1 KP37-45 compared with restimulation with imDCs alone. Therefore, WT1 HP34-51 may have the ability to induce WT1-specific CD8+ CTLs in patients with cancer carrying the HLA-A*02:01 or HLA-A*02:06 allele. Nonetheless, 1 of 2 cancer patients with the HLA-A*02:07 allele exhibited IFN-γ production by CD8+ T cells. According to the NetMHC 4.0 server-prediction program, WT1 KP37-45 within WT1 HP34-51 has a relatively high affinity for HLA-A*02:07. However, as HLA-A*02:07+ patients are not a major population according to the national marrow donor program (23), only two patients with HLA-A*02:07 were enrolled in the present study. Accordingly, further studies are required to assess whether mDC/WT1 HP34-51 consistently induces HLA-A*02:07-restricted WT1-specific CD8+ T cells in patients with cancer.
Among the total CD8+ T cell populations from patients with cancer, WT1-specific IFN-γ-producing CD8+ T cells were more frequently induced by four rounds of stimulation with mDC/WT1 HP34-51 followed by one stimulation with imDC/WT1 KP37-45 compared with four rounds of stimulation with mDC/WT1 KP37-45 followed by one stimulation with imDC/WT1 KP37-45. Previous reports indicate that CD4+ and CD8+ T cell stimulation occurs when peptide vaccination with combined WT1 HP and WT1 KP is performed; however, vaccination with WT1 HP alone does not strongly activate T cells (4,17). It has also been reported that activation of CD4+ T cells is essential for CD8+ CTL induction, potentiating CD8+ CTL proliferation and maintaining effector functions in the tumor microenvironment (5,24). Therefore, effective cancer vaccines must include a mechanism to activate CD4+ T cells. Simultaneous activation of CD4+ and CD8+ T cells by DCs pulsed with both WT1 HP34-51 and WT1 KP37-45 may improve the efficient induction of WT1-specific CD8+ T cells that produce IFN-γ. In fact, for 10 of the 14 patients in the present study, a >10% increase in the IFN-γ-producing CD4+ T cell population among total CD4+ T cells was induced by four rounds of stimulation with mDC/WT1 HP34-51 followed by one stimulation with imDC/WT1 KP37-45. According to the NetMHCII-2.3 server-prediction program, WT1 KP37-45, which is a part of WT1 HP34-51, has a relatively high affinity for MHC class II. Thus, WT1 KP37-45 may also be present in MHC class II with imDC/WT1 KP37-45, resulting in stimulation of CD4+ T cells producing IFN-γ. As one of the major subtypes of HLA-A molecule is HLA-A*24:02 (23), mDCs may be pulsed with a cocktail of WT1 HP34-51 and WT1 KP235-243 for HLA-A*24:02 (4) to maximize the induction and maintenance of CD4+ Th1 cells and CD8+ CTLs, at least in part, in an HLA-A*02:01-, HLA-A*02:06- or HLA-A*24:02-restricted manner for application in a wider cohort of patients with cancer.
WT1-reactive CD4+ T cells from patient 4 induced by mDC/WT1 HP34-51 or mDC/WT1 KP37-45 stimulation were almost all EM T cells, whereas those maintained with low doses of IL-2 and IL-7 were not. In contrast, cancer patient CD4+ T cells in the no stimulation group were almost all naïve. Naïve T cells differentiated into EM T cells via the effector T cell route (25) by undergoing several rounds of stimulation with mDC/WT1 HP34-51 or mDC/WT1 KP37-45. It has been reported that peptide-specific memory T cell frequencies are increased in vaccine responders with a positive delayed-type hypersensitivity (DTH) test (4,26-28). Our previous study also demonstrated that in super-responders with WT1-specific DTH induced by vaccination with mDCs pulsed with WT1 HP and WT1 KP, continuous elicitation of WT1-specific long-lived EM and CM T cells occurred and could be boosted; these long-lived WT1-specific EM T cells patrol and recognize tumor cells and are therefore associated with a clinical benefit (4). Although EM CD4+ T cells were induced by mDC/WT1 HP34-51 stimulation, compared with EM CD4+ T cells, the EM CD8+ T cell frequency was not strongly increased. As some WT1 overexpression occurs in various types of solid tumor, WT1-specific CD8+ T cells are spontaneously induced (11-13), suggesting that some memory CD8+ T cells may be detected without any stimulation in vitro. As memory CD4+ T cells particularly support the priming and maintaining of antigen-specific CD8+ T cells (29,30), the increase in memory CD4+ T cell frequency induced by mDC/WT1 HP34-51 stimulation may support an increase in WT1-reactive CD8+ T cell frequency. In summary, T cells stimulated with mDC/WT1 HP34-51 may have the capacity to elicit WT1-specific EM CD4+ T cells and CD8+ CTLs, which was at least in part restricted by HLA-A*02:01 or HLA-A*02:06, both of which produce IFN-γ. However, the memory cell subsets were only analyzed in a single case, a patient with PDA, as a sufficient number of cells was not available. The induction and maintenance of the durable memory type of CD4+ Th1 cells by mDC/WT1 HP34-51 may produce promising clinical activity to prolong overall survival in advanced-stage cancer patients.
As a sufficient number of PBMCs and autologous tumor cells were not available, the killing activity of WT1-ractive T cells induced by mDC/WT1 HP34-51 could not be assessed in this preclinical study. A phase I clinical trial involving patients with advanced pancreatic cancer who express HLA-A*02:01, HLA-A*02:06 or HLA-A*24:02 is being performed using mDCs pulsed with a novel peptide cock-tail containing WT1 HP34-51 and HLA-A*24:02-restricted WT1 KP235-243 combined with standard chemotherapy; the results of this clinical trial will be available soon. Additionally, future studies should optimize cancer vaccines incorporating mDCs loaded with an immunogenic novel WT1 peptide cocktail in combination with an immune checkpoint-blocking antibody.