TAS0314, a novel multi-epitope long peptide vaccine, showed synergistic antitumor immunity with PD-1/PD-L1 blockade in HLA-A*2402 mice

Cancer peptide vaccines are a promising cancer immunotherapy that can induce cancer-specific cytotoxic T lymphocytes (CTLs) in tumors. However, recent clinical trials of cancer vaccines have revealed that the efficacy of the vaccines is limited. Targeting single antigens and vaccination with short peptides are partly the cause of the poor clinical outcomes. We synthesized a novel multi-epitope long peptide, TAS0314, which induced multiple epitope-specific CTLs in HLA knock-in mice. It also showed superior epitope-specific CTL induction and antitumor activity. We also established a combination treatment model of vaccination with PD-1/PD-L1 blockade in HLA-A*2402 knock-in mice, and it showed a synergistic antitumor effect with TAS0314. Thus, our data indicated that TAS0314 treatment, especially in combination with PD-1/PD-L1 blockade, is a promising therapeutic candidate for cancer immunotherapy.


Scientific Reports
| (2020) 10:17284 | https://doi.org/10.1038/s41598-020-74187-6 www.nature.com/scientificreports/ is altered by epigenetic regulation or mutation, enabling the tumors to evade the immune system 25 . For example, in the phase II clinical trial of the glypican-3 peptide vaccine, tumors that relapsed after vaccination showed the loss of glypican-3 antigen 26 . Immunization with multiple epitopes is one of the ways to overcome the immune evasion of tumors, and would enhance the antitumor effect 27 . Therefore, it is possible to enhance the efficacy of anti-cancer peptide vaccines to overcome these three problems.
To overcome the issues of existing antitumor peptide vaccines, we designed and developed a novel multiepitope long peptide vaccine, TAS0314, which contains four epitopes from SART2 and SART3. The aim of this study was to demonstrate the immunological characteristics of TAS0314 in vitro and in vivo as an antitumor peptide vaccine. We also evaluated the benefit of combination treatment with PD-1/PD-L1 blockade using recently established HLA-A*2402 mice.

Results
TAS0314 induced multiple epitope-specific CTLs. To evaluate the efficacy of a multi-epitope long peptide vaccine, we synthesized a novel peptide, TAS0314, which contains four epitopes from SART2 and SART3 linked with an arginine dimer.
To evaluate whether the CTLs induced by TAS0314 can exert long-term surveillance, we examined the recall response against SART2 93-101 peptide. Mice were immunized with TAS0314 or SART2 93-101 peptide and were boosted with SART2 93-101 148 days after the last immunization (Fig. 2c). Mice immunized with TAS0314 had significantly increased numbers of specific CTLs after boosting when compared to those immunized with SART2 93-101 .
To evaluate the mechanism of antigen presentation for TAS0314, we examined the antigen presentation of TAS0314 by professional antigen-presenting cells (DCs) and non-professional antigen-presenting cells (T cells). As shown in Fig. 2d, TAS0314-pulsed CD11c + DCs efficiently activated CTLs while TAS0314-pulsed T cells exhibited a significantly lower IFN-γ response. In addition, the CTL activation by TAS0314-pulsed DCs was significantly decreased by the proteasome inhibitor epoxomicin 29 .
Antitumor effect of TAS0314 against B16F10.A24/SART2 93-101 tumor cells. We evaluated the antitumor activity against subcutaneous B16F10.A24/SART2 93-101 tumors in a syngeneic mouse model. To compare the efficacy, mice were immunized with SART2 93-101 peptide or TAS0314 at a molar-equivalent dose (Fig. 4a). At 18 days after tumor implantation, TAS0314 significantly reduced the tumor volume by 70.25% when compared to the control. SART2 93-101 peptide only reduced the tumor volume by 47.7%, and did not show any significant difference from the control group (Fig. 4b). These data suggested that TAS0314 induced better antitumor activity than SART2 93-101 .
Scientific Reports | (2020) 10:17284 | https://doi.org/10.1038/s41598-020-74187-6 www.nature.com/scientificreports/ TAS0314 and PD-1/PD-L1 blockade synergistically suppressed tumor growth. We evaluated the synergistic antitumor activity of TAS0314 and anti-PD-1 antibody against B16F10.A24/SART2 93-101 tumors (Fig. 5a). Anti-PD1 antibody treatment partially suppressed tumor growth, while TAS0314 treatment significantly suppressed tumor growth (Fig. 5b). Furthermore, the combined therapy of TAS0314 and anti-PD-1 antibody dramatically suppressed tumor growth when compared to monotherapy. We also evaluated the survival for 57 days. Anti-PD-1 antibody did not prolong the survival when compared to the control group (Fig. 5c). Combined treatment with TAS0314 and anti-PD-1 antibody significantly prolonged the survival when compared to the anti-PD-1 antibody alone (p < 0.001) or TAS0314 alone (p < 0.05). At 57 days after tumor implantation, all mice in the control group and the anti-PD-1 antibody-treated group were dead. In contrast, 13.33% of the mice were tumor-free in the TAS0314-treated group, and 53.33% of the mice were tumor-free in the group treated with the combination of TAS0314 and anti-PD-1 antibody.
We also evaluated the antitumor activity of TAS0314 in combination with anti-PD-L1 antibody (Fig. 5d). The combination of TAS0314 and anti-PD-L1 antibody also synergistically inhibited tumor growth when compared to monotherapy (Fig. 5e). Combination therapy with anti-PD-L1 antibody dramatically prolonged the survival (Fig. 5f). Eighty-five days after tumor implantation, 66.67% of the mice treated with the combination of TAS0314 and anti-PD-L1 antibody were tumor-free. In contrast, no mice were alive in the other groups.
These data strongly suggested that the combination treatment of TAS0314 and PD-1/PD-L1 blockade has synergistic antitumor effects.

Combination of TAS0314 and anti-PD-1 antibody increased the number of epitope-specific CTLs in tumors.
To analyze the mechanism of the synergistic effect, we measured SART2 93-101 epitope-specific T cells in tumors using the SART2 93-101 -HLA-A24 tetramer (Fig. 6). A number of tetramer-positive T cells was detected after immunization with TAS0314, whereas no SART2-specific CTL infiltration was detected in the vehicle-immunized groups. Surprisingly, the number increased approximately three-fold in the group treated with the combination treatment of TAS0314 and anti-PD-1 antibody. Increases in the number of epitope-specific CTLs in tumors would contribute to the synergistic antitumor activity of TAS0314 and anti-PD-1 antibody.

Discussion
Numerous clinical trials have been conducted, but the efficacy of cancer vaccines has been limited 30,31 . To develop useful cancer peptide vaccines, their efficacy must be improved.   www.nature.com/scientificreports/ We developed a novel multivalent long peptide vaccine, TAS0314, and evaluated its CTL induction ability, antitumor effect in monotherapy, and its antitumor effect in combination with anti-PD-1 antibody. We previously reported that the induction of epitope-specific CTLs is affected by the position of the epitope in a long peptide 32 . We evaluated and screened numerous peptides to obtain suitable long peptides with efficient CTL induction ability by changing the linkage order of four epitopes. Finally, we designed TAS0314, which could efficiently induce epitope-specific CTL induction. Indeed, TAS0314 showed superior CTL induction and antitumor effects when compared to short epitope peptide vaccination.
Synthetic long peptide vaccines that were efficiently processed by DCs and induced robust CD8 + T cell responses have been reported 33 . In our study, we confirmed that TAS0314 epitopes were presented by DCs in a proteasome-dependent pathway. In contrast, non-professional antigen-presenting cells, such as T cells, showed little antigen presentation of the epitopes in TAS0314. Antigen presentation by non-professional antigenpresenting cells induces the exhaustion of CTLs, and exhausted T cells show loss of multifunctional cytokine production 34 . Indeed, we obtained similar data showing that T cells induced by short epitope immunization had a smaller population of multifunctional (IFN-γ + /TNF-α + /IL-2 + , IFN-γ + /TNF-α + ) CTLs when compared to those induced by TAS0314. We considered that TAS0314 antigen presentation via DCs may contribute to the enhanced CTL induction and antitumor effects of TAS0314.
According to the current study, immunization with TAS0314 and boosting 148 days after the last immunization significantly induced a large number of CTLs, suggesting that TAS0314 vaccination resulted in memory CTL induction. Among peripheral blood mononuclear cells, central memory T cells produce IL-2 35 , and autocrine IL-2 production is required for efficient expansion after a second challenge 36 . Therefore, multifunctional CTLs that are induced by TAS0314 might differentiate into memory T cells, and contribute to long-term surveillance. It has been reported that epitope-specific memory T cells remained for a long time in long-surviving patients treated with cancer vaccines 37 . Thus, TAS0314 might also contribute to the long-term control of cancer progression by inducing memory CTLs.
Cancer antigen-specific T cells upregulate immunosuppressive molecules (e.g., PD-1) after activation, and cause dysfunction by interacting with their ligands (e.g., PD-L1) in tumors 17 . Indeed, it has been reported that epitope-specific CTLs induced by peptide vaccines highly expressed PD-1 in cancer patients 38 . It is believed that the combination therapy of a cancer vaccine and PD-1/PD-L1 blockade synergistically enhances the antitumor activity. However, few HLA-A2-restricted epitope peptides from human cancer antigens have been evaluated for their efficacy in combination with anti-PD-1 antibody due to the HLA restriction in mouse models 39 . To evaluate the combination of other HLA allele-restricted epitopes with anti-PD-1 antibody, mice expressing the HLA allele are required. In the current study, we focused on the HLA-A24-restricted epitope, which is the most common HLA allele in Japanese. The synergistic antitumor effect of HLA-A24-restricted epitopes and anti-PD-1 antibody was evaluated using our HLA-A*2402 KI mice.
Combination treatment with TAS0314 and PD-1/PD-L1 blockade also showed synergistic antitumor activity. The combination therapy dramatically inhibited tumor growth, and prolonged survival. The number of www.nature.com/scientificreports/ epitope-specific CTLs in tumors were increased by approximately three-fold, and this increase was considered to be a mechanism of the synergistic antitumor effect of both PD-1 and PD-L1 antibody treatment with TAS0314. Although a short peptide vaccine containing the SART2 93-101 epitope did not induce a significant clinical response in a recent Phase III clinical trial, our data suggest that the combination of PD-1/PD-L1 blockade might improve its clinical efficacy.
It has been reported that anti-PD-1 antibody enhanced chemokine expression in tumors and T cell infiltration 24,40 . Others have also reported that anti-PD-1 antibody inhibited T cell apoptosis, and enhanced the proliferation of T cells 41 . According to these reports and our current data, we believe that anti-PD-1 antibody treatment induces the infiltration and accumulation of TAS0314-specific CTLs in the tumor microenvironment, which contributes to the synergistic antitumor effects. However, the detailed mechanism should be elucidated in future studies.
There are some limitations in this study. First, we evaluated the antitumor activity only against tumor cells expressing SART2 93-101 as a representative epitope peptide. Other epitopes are also expected to show superior CTL induction and antitumor activity, because they likely behave similarly.
We also developed different multivalent long peptides, TAS0315 and TAS0316, which contain 8 epitopes from EGFR, Lck, MRP3, PTHrP, TMEM189, and WHSC2. Similar to TAS0314, these two long peptides also showed CTL induction ability in HLA-KI mice (Supplementary Fig. 1). According to our results, TAS0314, TAS0315, and TAS0316 are expected to have an antitumor effect and synergistic effect with anti-PD-1 antibody; however, further investigations are needed with the establishment of new HLA-A2-or A24-dependent syngeneic antitumor models.
Second, the antitumor effect of TAS0314 was examined only by an artificial model using HLA-A*2402 and SART2 93-101 stably overexpressing cells. Although we confirmed that TAS0314-induced antigen-specific CTLs recognized and killed HLA-A24 human cancer cells in vitro ( Supplementary Fig. 2), the anti-tumor effect of TAS0314 against human cancers remains to be clarified through clinical trials.
In summary, a novel multi-epitope long peptide vaccine, TAS0314, induced HLA-A2, HLA-A24, and HLA-A3 superfamily-restricted multiple epitope-specific CTLs. TAS0314 showed superior antitumor activity when compared to short epitope vaccines. Furthermore, we developed a combination model of vaccine and PD-1/ PD-L1 blockade in HLA-A*2402 KI mice; combination therapy with anti-PD-1 antibody or anti-PD-L1 antibody increased epitope-specific CTLs in tumors and synergistically enhanced the antitumor activity. We have been developing a peptide mixture, TAS0313, that contains TAS0314, TAS0315, and TAS0316 for clinical use. Our preliminary data obtained during this development suggest that TAS0313 treatment, especially the combination therapy of TAS0313 and PD-1/PD-L1 blockade, represents a novel treatment option for patients suffering from various cancers. Moreover, the induced multiple antigen-specific CTLs that are expected to protect from immune escape due to antigen loss would contribute to long-term antitumor activity.

Methods
Mice. Heterozygous HLA-A*2402, heterozygous HLA-A*0201, and heterozygous HLA-A*3101 KI mice were generated as previously described 28 . All animal procedures were performed in compliance with the National Institutes of Health Guidelines, and were approved by the Taiho Institutional Animal Care and Use Committee.  Immunization of mice. Mice were immunized three times at 7-day intervals. A mixture of peptides emulsified in Montanide ISA-51VG (Seppic, Paris, France) was subcutaneously injected at the base of the tail (100 µL/ mouse). For the experiment to evaluate memory function, mice were immunized with TAS0314 or SART2 93-101 [TAS0314: 100 µg/mouse, SART2 93-101 : 21 µg/mouse (equivalent molar mass to 100 µg of TAS0314/mouse)], and 148 days after the last immunization, the mice were immunized with SART2 93-101 as a booster treatment. Two weeks after the booster treatment, CTL induction in the lymph node was evaluated.
Culture of epitope-specific CTLs. Immunized mice were sacrificed 7 days after the last immunization, and lymphocytes from the lymph node were prepared. The lymphocytes were cultured for 8 days in culture medium supplemented with target peptide (10 µg/mL), rIL-15 (100 ng/mL) and mouse rIL-21 (100 ng/mL). Living cells were collected by density gradient centrifugation using Lympholyte-M (Cedarlane, Hornby, ON, Canada).

Antigen presentation assay. CD11c + dendritic cells (DCs) and CD3 + T cells were purified with CD11c
MicroBeads UltraPure, mouse, and the Pan T Cell Isolation Kit II, mouse (Miltenyi Biotec) from the splenocytes of HLA-A*0201 KI mice. The antigen-presenting cells were incubated with peptides for 3 h (37 °C, 5% CO 2 ) and washed three times. Epoxomicin was added 30 min before the addition of TAS0314. Cultured SART3 302-310 -specific CTLs from immunized HLA-A*0201 KI mice with SART3 302-310 (100 μg/mice) were used as effector cells. The CTLs were co-cultivated with antigen-presenting cells on an anti-mouse IFN-γ-coated plate. Epitope-specific IFN-γ production was detected using an ELISPOT assay.
Evaluation of antitumor activity against SART2 93-101 -expressing syngeneic tumors. For the prophylactic treatment, HLA-A*2402 KI mice (n = 10) were immunized three times with TAS0314 (100 µg/ mouse), SART2 93-101 epitope (21 µg/mouse), or vehicle control. Seven days after the last immunization, immunized mice were subcutaneously inoculated with B16F10.A24/SART2 93-101 tumor cells (5 × 10 5 ) into the right flank. Seven days after tumor inoculation, the tumor size was measured with a caliper. The tumor volume was calculated using the following formula: For the therapeutic treatment, HLA-A*2402 KI mice (n = 15) were intravenously injected with B16F10.A24/ SART2 93-101 tumor cells (1 × 10 6 ). The mice were immunized with TAS0314 (300 µg/mouse) three times with intervals of 7 days starting from the day after the tumor inoculation. On day 28 after tumor injection, mice were euthanized, and the number of lung metastases was counted.