Unresectable or recurrent hepatocellular carcinoma (HCC) is associated with poor prognosis. With the exception of sunitinib, there are no established systemic antitumor therapies. Recently, immune checkpoint inhibitors have become available for the treatment of HCC and the results of phase I/II trials of nivolumab, an antiprogrammed cell death-1 antibody, showed good response in some patients, but the majority of the patients remain refractory.1,2 In contrast, despite the fact that the recurrence rate is still high even after curative resection of HCC, postoperative adjuvant therapy has not been established.3 Proteomic and immunohistochemical staining experiments in our department demonstrated that heat-shock protein 70 (HSP70) is highly and specifically expressed by HCC cells.4,5 On the basis of these results, we developed an HSP70 mRNA-transfected dendritic cell (DC) therapy for treating unresectable or recurrent HCC, and a phase I trial confirming the safety and efficiency of this therapy has been completed.6 As this cell-based therapy is complex and therefore suitable for only a limited number of facilities, in the present study, we developed a novel peptide vaccine-based immunotherapy by identifying the functional epitope peptide in the HSP70 mRNA DC therapy using a computational algorithm and patient-derived peripheral blood mononuclear cells (PBMCs).7
The HLA-A gene with the highest frequency in Asian populations, especially in the Japanese, is HLA-A*24:02, whereas the frequency of this gene is low in Caucasians.8 In contrast, the gene frequency of HLA-A*02:01 is high among a variety of ethnic groups, including both Asians and Caucasians.9 These findings suggest that HSP70-derived cytotoxic T lymphocyte (CTL) epitopes that bind to both HLA-A*24:02 and A*02:01 would be highly useful in immunotherapies aimed at treating HCC. In the present study, we identified a human HSP70-derived epitope peptide that binds to HLA-A*24:02, A*02:01, and *A02:06.
MATERIALS AND METHODS
Fourteen HLA-A*24:02-positive, *02:01-positive, or *02:06-positive patients who received the DC-based therapy participated in the clinical trial (UMIN000010691).
Selection of Candidate Peptides
To select candidate peptides derived from HSP70, we used a peptide prediction system developed by the NEC Corporation. The system employs a committee-based framework of query learning using hidden Markov models as the component algorithm.7 The system enables comprehensive searching of a large number (209) of peptides in a small number of experiments. Learning was achieved over 7 rounds of feedback loops, in which our computational algorithm was used to determine the next set of peptides to be analyzed based on the results of the previous iterations. After these training cycles, the algorithm provided a real number prediction of major histocompatibility complex binding peptides with high accuracy. Using this prediction system, we selected a total of 17 candidate peptides (Table 1).
HSP70-derived 9-mer peptides that exhibited high binding affinity to HLA-A*24:02, 02:01, and 02:06 were synthesized by Scrum (Tokyo, Japan) and purified by HPLC to purity of >90%. The epitope peptides HIV-*A24:02 (RYLRDQQLL), HIV-A*02:01 (SLYNTVATL), and HIV-*A02:06 (ATLEEMMTA) were synthesized as negative controls.10–12
Peptide Binding Assay
Binding of peptides to HLA-A*24:02 was examined using acid stripping and a reconstitution assay, as previously described by Zeh et al,13 with minor modifications. Briefly, C1R-A24 cells were exposed to citrate-phosphate buffer (pH 3.3) and then reconstituted with graded concentrations of peptide and 0.1 μM human β2-microglobulin (M-4890; Sigma, St Louis, MO) in DMEM containing 0.25% bovine serum albumin. Fluorescein isothiocyanate–labeled monoclonal antibody 17A12 (Tahara and colleagues) was used to detect properly folded and peptide-bound HLA-A*24:02 molecules.14 Fluorescence intensity was measured using a FACScan instrument (Becton-Dickinson Japan, Tokyo, Japan). Both high binding and low binding peptides (HER2-63 TYLPTNASL and Met149 RVWESATPL, respectively) were included in each assay, and their binding was used to normalize variations between experiments. The affinity of each peptide was calculated as previously described (Udaka and colleagues). A similar method was used to assess the binding of peptides to HLA-A*02:01 and HLA-A*02:06, and details will be published elsewhere.15
Separation of Adherent and Nonadherent Cells
PBMCs were harvested from patients enrolled in the clinical trial described previously using a COBE Spectra Apheresis System (COBE BCT Inc., Lakewood, CO). PBMCs from 3000 mL of blood were enriched by density gradient centrifugation with Ficoll-Paque (Amersham Pharmacia Biotech, Uppsala, Sweden). The PBMCs were incubated for 45 minutes in a 5% CO2 atmosphere at 37°C in serum-free AIM-V medium (Gibco, Paisley, Scotland). Plastic-adherent cells were used to generate DCs, whereas nonadherent cells were used to generate CTLs.
Generation of DCs
DCs were generated as described previously.6 Briefly, plastic-adherent cells were cultured in AIM-V medium containing 800 U/mL of granulocyte-macrophage colony-stimulating factor (Osteogenetics GmbH, Wurzburg, Germany) and 500 U/mL of interleukin (IL)-4 (Osteogenetics GmbH). On day 6, immature DCs were cultured in AIM-V medium containing 300 U/mL of tumor necrosis factor-α (R&D Systems, Minneapolis, MN). Cultures were checked for endotoxins and contamination with Mycoplasma and other bacteria before administration. On day 10, floating and loosely adherent cells were collected as mature DCs.
Preparation of CD8+ T Lymphocytes
Nonadherent cells were cultured in AIM-V medium. On days 3, 5, and 7, recombinant human IL-2 (Shionogi Pharmaceutical Co., Tokyo, Japan) was added to the cultures to a final concentration of 10 IU/mL. The plates were incubated in a 5% CO2 atmosphere at 37°C. On day 10, cultured cells were collected and washed 3 times with phosphate buffered saline (PBS). T cells were then isolated by MACS depletion (Miltenyi Biotec, Bergisch Gladbach, Germany). Cultured cells were labeled with microbeads using a CD8 T-cell isolation kit and separated on magnetic columns in a VarioMACS separator, according to the manufacturer’s recommendations (Miltenyi Biotec). CD8+ T cells were suspended at 5×106 cells/mL.
Coculture of Peptide-pulsed DCs and CD8+ T Lymphocytes
On day 10, mature DCs were cultured with HSP-derived positive or negative peptides at a final concentration of 20 μg/mL for 2 hours in a 5% CO2 atmosphere at 37°C. The peptide-pulsed DCs were collected and washed 3 times and suspended at 2×106 cells/mL. CD8+ T cells at 5×105 cells/well and peptide-pulsed DCs at 2×105 cells/well were cocultured in 48-well plates in a 5% CO2 atmosphere at 37°C. On day 12, IL-2 was added to the cultures at a final concentration of 20 U/mL.
Interferon (IFN)-γ ELISPOT Assay of Stimulated CD8+ T Cells
On day 17, stimulated CD8+ T cells were plated in triplicate wells at a density of 5×104 cells/well in the medium for IFN-γ using an ELISPOT assay, according to the manufacturer’s instructions (Mabtech, Cincinnati, OH). CD8+ T cells were incubated for 18–20 hours in an incubator with stimulator cells. The number of spots on each plate was determined using an Eliphoto Scan (Minerva Tech, Tokyo, Japan). Similar results were obtained in 5–26 independent experiments. This experiment has been performed at least 5 times for each peptide. Because the amount of sample was limited, the peptides that are not likely to be candidates have stopped the experiment any time. In contrast, to improve the accuracy for potential candidate peptides, additional experiments were repeated up to 26 times.
Lymphocyte Preparation for Cytotoxicity Detection Assay
PBMCs were obtained from HLA-A*24:02-positive or *02:01-positive healthy volunteers. Peripheral blood was obtained via venipuncture, collected in EDTA tubes, and transferred to the central laboratory at room temperature. Within 24 hours of blood collection, PBMCs were isolated using Ficoll-Paque (Amersham Pharmacia Biotech) density gradient solution. For in vitro culture, the PBMCs were thawed simultaneously, and 1×106 cells/well were incubated in medium with peptide stimulation (20 μg/mL) performed twice on days 1 and 8 in combination with the HIV-A*24:02 (RYLRDQQLL) and HIV-A*02:01 (SLYNTVATL) epitope peptides as negative controls. A total of 20 IU/μL of recombinant IL-2 (Novartis) was added on days 2, 5, 9, and 13. On day 15, the cultured lymphocytes were subjected to an xCELLigence cytotoxicity detection assay after negative selection of CD8+ T cells using magnetic beads (Miltenyi Biotec). xCELLigence cytotoxicity detection assays were performed as described.
xCELLigence Cytotoxicity Detection Assay
All experiments were performed using the respective target cell culture medium. A 50-μl aliquot of the medium was added to E-Plates 16 (ACEA Biosciences, San Diego, CA) for measurement of background values. Target cells were seeded in the medium at a density of 10,000 cells/well. Suitable cell densities were determined by prior titration experiments. Cell attachment was monitored using an RTCA SP instrument (Roche) with RTCA software, version 1.1 (Roche), until the plateau phase was reached, which usually required ∼22–26 hours T cells were added at different effector to target (E:T) ratios, ranging from 20:1 to 5:1. Upon addition of effector cells, impedance measurements were performed every 15 minutes for up to 81 hours. All experiments were performed in duplicate. Changes in electrical impedance were expressed as a dimensionless cell index (CI) value, which was derived from relative impedance changes corresponding to cellular coverage of the electrode sensors, normalized to baseline impedance values with medium only. To analyze the acquired data, CI values were exported and the percentage of lysis was calculated in relation to control cells lacking any effector T cells. Percent lysis was determined using the following formula at various time points: (CI tumor only−[CI tumor+T cells])/(CI tumor only)×100.16
Induction of HSP70-derived Peptide-specific CTLs in Vivo
HLA-A 24 transgenic (Tg) mice were kindly provided by Dr Nishimura and Dr Irie (Lemonnier, unpublished data). Mice were maintained at the Yamaguchi University Animal Facility (Ube, Japan) under specific pathogen-free conditions. All experiments used mice that were 8–12 weeks old at the time of the first procedure. All mice were used according to the guidelines of the institutional animal care and use committee of Yamaguchi University, which approved this study (ID number: 33038).
HLA-A 24 Tg mice were subcutaneously immunized with PBS, adjuvant of 50 μg poly(I:C) plus 1 μg LAG-3-Ig, or 50 μg of HSP70-derived peptide with adjuvant of 50 μg poly(I:C) plus 1 μg LAG-3-Ig17 once a week for 8 weeks. At 7 days after the last immunization, lymph nodes were extracted and homogenized. For the in vitro culture, 5×105 lymphocytes/well were maintained in complete medium consisting of RPMI 1640 (Mediatech, Herndon, VA) supplemented with 10% fetal bovine serum (HyClone, Logan, UT), 25 mM HEPES (N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid), 2 mM glutamine, 100 U/mL penicillin G, and 100 g/mL streptomycin sulfate with HSP70-derived peptide (10 μg/mL) after lysing red blood cells in ACK lysis buffer. Five days later, the lymphocytes were harvested and the mouse IFN-γ ELISPOT assay (Mabtech, Nacka Strand, Sweden) was performed according to the manufacturer’s protocol. Because of the small number of cells, lymphocytes from each group of mice were mixed and used in the ELISPOT assay. Briefly, lymphocytes (3×106 cells/80 μL) were added to a 96-well ELISPOT plate pre-coated with anti-mouse monoclonal IFN-γ antibodies. Experiments were carried out in triplicate. The cells were incubated with 10 μg/mL of HSP70-derived peptide for 24–48 hours. After stimulation, the wells were washed and incubated with biotinylated anti-mouse IFN-γ antibody for 2 hours at room temperature. Subsequently, the wells were washed and incubated with streptavidin-horseradish peroxidase for 1 hour. Spot-forming cells were developed with TMB substrate and counted using an Eliphoto Scan (Minerva Tech).
Data are presented as mean and SD. Statistical differences between 2 groups were evaluated using the unpaired Student t test with JMP software, version 12.0 (SAS Institute Japan, Tokyo, Japan). P-values <0.05 were considered significant.
Analysis of Binding of HSP70-derived Peptides to HLA-A*24:02, *02:01, and *02:06
HSP70-derived 9-mer peptides that exhibited high binding affinity to HLA-A*24:02, 02:01, and 02:06 were selected as candidate peptides using a binding prediction system (NEC Corporation) that utilizes a data-mining technique and a query learning algorithm based on hidden Markov models.7
For this study, we examined binding to HLA-A*24:02, *02:01, and *02:06. HLA-A*24:02 is the most frequent allele among Asians (eg, 33% in Japanese18 versus ∼10% in western populations. Seventeen peptides exhibiting an affinity of 4.0 or higher in terms of log Ka for binding to at least 1 of the 3 HLA-A types were selected as targets for the experiment (Table 1). To achieve therapeutic effects in a wide range of cases, peptides with high affinity for all 3 HLA-A molecules would be more preferable, and these are indicated in bold font in Table 1.
Production of IFN-γ by CTL Clones Induced by Candidate Peptides
HLA-A*24:02-positive, *02:01-positive, and *02:06-positive CD8+ T cells were pulsed with respective peptides, and then IFN-γ production was analyzed by ELISPOT assay. Data for positive and negative controls are significantly different. Four peptides induced peptide-specific CTLs producing IFN-γ against HLA-A*24:02-positive CD8+ T cells, and 2 peptides induced peptide-specific CTLs producing IFN-γ targeting HLA-A*02:01-positive and *02:06-positive CD8+ T cells (Fig. 1). Seven of the HLA-A*24:02-positive patients (represented by black bars) had HLA-A*02:01 or HLA-A*02:06. Compared with the negative control, we obtained both HLA-A*24:02-positive CTLs in addition to HLA-A*02:01-positive and *02:06-positive CTLs that specifically produced significant amounts of IFN-γ after CTL expansion when pulsed with peptide Y50. Although HLA-A*24:02-positive Y75 peptide-specific CTLs produced a significant amount IFN-γ as compared with the negative control, HLA-A*02:01-positive and *02:06-positive Y75 peptide-specific CTLs did not. Compared with the negative control, CTLs induced with other peptides did not exhibit significant IFN-γ production, regardless of HLA-A alleles.
Cytotoxic Activity of CTLs
We also examined the cytotoxic activity of CTL clones to evaluate the possibility of developing a cancer vaccine for use in tumor therapy involving candidate peptides. We wished to identify a peptide that could induce all HLA-A*24:02-positive CTLs and HLA-A*02:01-positive and *02:06-positive CTLs promiscuously. Therefore, we used the Y50 peptide for this experiment. Other peptides could not induce CTLs to a similar degree as the Y50 peptide. Therefore, we did not expect cytotoxicity to other peptides and did not conduct experiments using these peptides. A peptide Y50-specific CTL clone with HLA-A*24:02 and *02:01 demonstrated cytotoxic activity against HLA-A*24:02-positive and *02:01-positive HepG2 cells, which express high levels of HSP70. In contrast, HIV peptide–specific CTL clones did not demonstrate cytotoxicity against HepG2 cells (Fig. 2).
Induction of Specific CTLs by Vaccination with HSP70-derived Peptides in Mice
We next examined whether peptide Y50 induces HSP70-specific CTLs in HLA-A 24 Tg mice. Following 8 subcutaneous vaccinations with PBS, adjuvant of (I:C) plus LAG-3-Ig, or peptide Y50 with adjuvant of poly(I:C) plus LAG-3-Ig, lymphocytes were cultured in vitro and then evaluated for IFN-γ production. Lymphocytes from 3 mice immunized with peptide Y50 exhibited significantly higher IFN-γ production (Fig. 3).
We also examined the tissues of the 3 immunized mice that exhibited HSP70 peptide–specific CTL responses. In all 3 mice, no pathologic changes caused by immune responses, such as lymphocyte infiltration or tissue destruction and repair, were noted in the brain or other internal organs except for the intestines, and the tissues exhibited normal structures and cellularity. Minimal to mild cellular infiltration was observed in the intestines of mice in the adjuvant group and adjuvant plus peptide group (Table 2).
In this study, we performed binding assays to identify candidate HSP70-derived peptides and identified a multi-HLA-class I–binding epitope peptide that binds to HLA-A*02:01, *02:06, and *24:02 by immune induction using an in vitro IFN-γ ELISPOT assay. We confirmed cytotoxicity in vitro and induction of an immune response and safety in vivo using HLA-A 24 Tg mice. In this study, we targeted HLA-class I peptides capable of binding to multiple HLAs. We used a peptide prediction system (developed by NEC Corporation) appropriate for screening promiscuous peptides identified in previous reports as capable of binding to HLA-class I molecules.
Peptide binding assays were performed for candidates selected from the binding prediction system results, and the ability of these peptides to induce an immune response was verified using the IFN-γ ELISPOT assay. Peptide Y50 did not exhibit the highest binding affinity for each HLA, but it did induce immune responses against multiple HLAs. As HSP70 is a relatively ubiquitous molecule, peptides derived from HSP70 could be considered cryptic (ie, immune tolerance could result if the binding affinity of the peptide was too high).19 Regarding HLA-A*02:06, CTL induction was confirmed in the patients with HLA-A*02:06 (Supplemental Fig., Supplemental Digital Content 1, http://links.lww.com/JIT/A530). However, the frequency of HLA-A*02:06 in the Japanese population is reportedly ∼9%, and only 2 patients in our study had HLA-A*02:06. In addition, cases with HLA-A*02:06 always had HLA-A*24:02 or *02:01. Thus, unfortunately, we could not determine whether this CTL induction was against HLA-A*02:06 or another allele. However, for the A2 supertype, some reports have indicated that the peptide for HLA-A*02:01 exhibits cross-reactivity to *02:06 when the binding affinity is high and therefore induces CTLs.20,21 In this study, the Y50 peptide exhibited high binding affinity to HLA-A*02:01, and CTL induction was also observed, suggesting cross-reactivity to HLA-A*02:06. We hope to collect samples and verify this in future experiments.
With respect to cytotoxicity, a peptide Y50-specific CTL clone with HLA-A*24:02 and *02:01 exhibited cytotoxicity against HSP70-expressing cells (HepG2), and this result was consistent with results obtained from the binding and immune response induction assays. Cytotoxicity was evaluated based on electrical impedance measurements using the xCELLigence System. It is reported that a small number of specific T cells and a low effector to target cell ratio commonly leads to a delay in maximal lysis well beyond 24 hours. Using conventional techniques to determine the cytotoxic potential of antigen-specific T cells, such as the 51Cr-release assay, it is not possible to draw conclusions regarding the lysis kinetics of T cells because these are typically endpoint assays. In contrast, xCELLigence is a convenient label-free method that is particularly applicable to the detection of antigen-specific T-cell–mediated cytotoxicity by monitoring lysis over time.22 In this study, we confirmed that cytotoxicity increased with time.
In vivo experiments were performed by administering poly(I:C) plus LAG-3-Ig in combination with peptide to HLA-A 24 Tg mice. Although the efficacy of peptide vaccine anticancer therapies administered alone has generally been limited, recent research indicated that the therapeutic effect of peptide drugs can be enhanced by combining their administration with adjuvants such as immune checkpoint inhibitors.2,23–25 The combination of poly(I:C) and LAG-3-Ig were shown to improve the antitumor effects of cancer vaccines by preventing T-cell exhaustion, and these agents are currently available in clinical practice.17 For this reason, we used combined adjuvants in our in vivo experiments and noted sufficient immune response induction, with no severe adverse events observed. However, we could not verify immune checkpoint changes in the present study due to the lack of an adequate number of samples. In addition, in vivo verification of HLA-A 02 was not carried out because we could not obtain HLA-A 02 Tg mice. These issues will be addressed in future studies.
The present results demonstrate that irrespective of HLA-A*24:02, *02:01, or *02:06 status, DCs pulsed with a specific HSP70-derived peptide induced specific CTLs to exhibit antitumor activity. The immunogenicity of the HSP70-derived peptide should be examined in patients with HSP70-expressing cancers, and we are currently conducting appropriate clinical trials.
The authors thank Akiko Sano and Kaori Kaneyasu for their excellent technical assistance with this work. They also thank Dr F.A. Lemonnier (Department SIDA Retrovirus, Unite d’Immunite Cellulaire Antivirale, Institut Pasteur, France) for providing HLA-A 24 Tg mice.
Conflicts of Interest/Financial Disclosures
This study was funded by NEC Corporation.
H.S., T.K., U.K., and N.H. received research funding from NEC Corporation. M.T. and D.S. are employees of NEC Corporation. The remaining authors have declared that there are no financial conflicts of interest with regard to this work.
1. Nishida N, Kudo M. Immune checkpoint blockade for the treatment of human hepatocellular carcinoma. Hepatol Res. 2018;48:622–634.
2. Hazama S, Tamada K, Yamaguchi Y, et al. Current status of immunotherapy against gastrointestinal cancers and its biomarkers: perspective for precision immunotherapy. Ann Gastroenterol Surg. 2018;2:289–303.
3. Ban D, Ogura T, Akahoshi K, et al. Current topics in the surgical treatments for hepatocellular carcinoma. Ann Gastroenterol Surg. 2018;2:137–146.
4. Takashima M, Kuramitsu Y, Yokoyama Y, et al. Proteomic profiling of heat shock protein 70 family members as biomarkers for hepatitis C virus-related hepatocellular carcinoma. Proteomics. 2003;3:2487–2493.
5. Yoshida S, Hazama S, Tokuno K, et al. Concomitant overexpression of heat-shock protein 70 and HLA class-I in hepatitis C virus-related hepatocellular carcinoma. Anticancer Res. 2009;29:539–544.
6. Maeda Y, Yoshimura K, Matsui H, et al. Dendritic cells transfected with heat-shock protein 70 messenger RNA for patients with hepatitis C virus-related hepatocellular carcinoma: a phase 1 dose escalation clinical trial. Cancer Immunol Immunother. 2015;64:1047–1056.
7. Udaka K, Mamitsuka H, Nakaseko Y, et al. Empirical evaluation of a dynamic experiment design method for prediction of MHC class I-binding peptides. J Immunol. 2002;169:5744–5753.
8. Imanishi T, Akaza T, Kimura A, et al. Allele and haplotype frequencies for HLA and complement loci in various ethnic groups. HLA. 1991;1:1065–1220.
9. Browning M, Krausa P. Genetic diversity of HLA-A2: evolutionary and functional significance. Immunol Today. 1996;17:165–170.
10. Ikeda-Moore Y, Tomiyama H, Miwa K, et al. Identification and characterization of multiple HLA-A24-restricted HIV-1 CTL epitopes: strong epitopes are derived from V regions of HIV-1. J Immunol. 1997;159:6242–6252.
11. Tsomides TJ, Aldovini A, Johnson RP, et al. Naturally processed viral peptides recognized by cytotoxic T lymphocytes on cells chronically infected by human immunodeficiency virus type 1. J Exp Med. 1994;180:1283–1293.
12. Kantakamalakul W, de Souza M, Bejrachandra S, et al. Identification of a novel HIV type 1 CRF01_AE cytotoxic T lymphocyte (CTL) epitope restricted by an HLA-Cw0602 allele and a novel HLA-A0206/peptide restriction. AIDS Res Hum Retroviruses. 2006;22:1271–1282.
13. Zeh HJ III, Leder GH, Lotze MT, et al. Flow-cytometric determination of peptide-class I complex formation. Identification of p53 peptides that bind to HLA-A2. Hum Immunol. 1994;39:79–86.
14. Tahara T, Yang SY, Khan R, et al. HLA antibody responses in HLA class I transgenic mice. Immunogenetics. 1990;32:351–360.
15. Udaka K, Wiesmuller KH, Kienle S, et al. An automated prediction of MHC class I-binding peptides based on positional scanning with peptide libraries. Immunogenetics. 2000;51:816–828.
16. Erskine CL, Henle AM, Knutson KL. Determining optimal cytotoxic activity of human Her2neu specific CD8 T cells by comparing the Cr51 release assay to the xCELLigence System. J Vis Exp. 2012;66:e3683.
17. Kano Y, Iguchi T, Matsui H, et al. Combined adjuvants of poly(I:C) plus LAG-3-Ig improve antitumor effects of tumor-specific T cells, preventing their exhaustion. Cancer Sci. 2016;107:398–406.
18. Tokunaga K, Ishikawa Y, Ogawa A, et al. Sequence-based association analysis of HLA class I and II alleles in Japanese supports conservation of common haplotypes. Immunogenetics. 1997;46:199–205.
19. Menez-Jamet J, Gallou C, Rougeot A, et al. Optimized tumor cryptic peptides: the basis for universal neo-antigen-like tumor vaccines. Ann Transl Med. 2016;4:266.
20. Sidney J, Southwood S, Mann DL, et al. Majority of peptides binding HLA-A∗ 0201 with high affinity crossreact with other A2-supertype molecules. Hum Immunol. 2001;62:1200–1216.
21. Zhang HG, Pang XW, Shang XY, et al. Functional supertype of HLA-A2 in the presentation of Flu matrix p58–66 to induce CD8+ T-cell response in a Northern Chinese population. Tissue Antigens. 2003;62:285–295.
22. Cerignoli F, Abassi YA, Lamarche BJ, et al. In vitro immunotherapy potency assays using real-time cell analysis. PLoS ONE. 2018;13:e0193498.
23. Banday AH, Jeelani S, Hruby VJ. Cancer vaccine
adjuvants—recent clinical progress and future perspectives. Immunopharmacol Immunotoxicol. 2015;37:1–11.
24. Kissick HT, Sanda MG. The role of active vaccination in cancer immunotherapy: lessons from clinical trials. Curr Opin Immunol. 2015;35:15–22.
25. Rosenberg SA, Yang JC, Restifo NP. Cancer immunotherapy: moving beyond current vaccines. Nat Med. 2004;10:909–915.