Blockade of PD-L1 Enhances Cancer Immunotherapy by Regulating Dendritic Cell Maturation and Macrophage Polarization

The immuno-inhibitory checkpoint PD-L1, regulated by tumor cells and antigen-presenting cells (APCs), dampened the activation of T cells from the PD-1/PD-L1 axis. PD-L1-expressing APCs rather than tumor cells demonstrated the essential anti-tumor effects of anti-PD-L1 monotherapy in preclinical tumor models. Using the murine tumor model, we investigated whether anti-PD-L1 antibody increased the antigen-specific immune response and anti-tumor effects induced by the antigen-specific protein vaccine, as well as the possible mechanisms regarding activation of APCs. Anti-PD-L1 antibody combined with the PEK protein vaccine generated more potent E7-specific immunity (including the number and cytotoxic activity of E7-specific cytotoxic CD8+ T lymphocytes) and anti-tumor effects than protein vaccine alone. Anti-PD-L1 antibody enhanced the maturation of dendritic cells and the proportion of M1-like macrophages in tumor-draining lymph nodes and tumors in tumor-bearing mice treated with combinatorial therapy. PD-L1 blockade overturned the immunosuppressive status of the tumor microenvironment and then enhanced the E7 tumor-specific antigen-specific immunity and anti-tumor effects generated by an E7-specific protein vaccine through modulation of APCs in an E7-expressing small tumor model. Tumor-specific antigen (like HPV E7 antigen)-specific immunotherapy combined with APC-targeting modality by PD-L1 blockade has a high translational potential in E7-specific cancer therapy.

The therapeutic protocols used to investigate whether the blockade of PD-L1 could enhance the anti-tumor effects of an antigen-specific protein vaccine are shown in Figure 1A. Mice bearing established E7-expressing TC-1 small tumors were treated with PEK protein vaccine with or without anti-PD-L1 Ab. Anti-PD-L1 monotherapy (4605.1 ± 114.0 mm 3 ) had no effect on tumor volume compared with PBS (4846.6 ± 342.1 mm 3 , p = 0.86, Student's t-test; Figure 1B). Mice receiving PEK protein vaccine and anti-PD-L1 Ab (55.1 ± 7.1 mm 3 ) had significantly smaller tumor volumes on day 28 than the group receiving PEK protein vaccine alone (348.4 ± 2.8 mm 3 , p < 0.001, Student's t-test; Figure 1C). Moreover, 80% of the mice treated with PEK protein vaccine and anti-PD-L1 Ab were alive after 100 days of TC-1 tumor cell challenge. In contrast, none of the mice in the group receiving PEK protein vaccine with PBS survived more than 56 days after the tumor challenge (p < 0.001, logrank test; Figure 1D). Anti-tumor effects of mice treated with PBS or anti-PD-L1 Ab alone. Anti-tumor effects were not improved in tumor-bearing mice treated with anti-PD-L1 Ab alone (p = 0.86, Student's t-test). (C) Antitumor effects of mice treated with PEK protein vaccine with or without anti-PD-L1 Ab. Tumor volumes were significantly lower in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than in mice treated with the PEK protein vaccine plus PBS (*: p < 0.001, Student's t-test). (D) Overall survival of mice treated with PEK protein vaccine with or without the anti-PD-L1 Ab. Eighty percent of mice that received PEK protein vaccination combined with anti-PD-L1 Ab were still alive 100 days after TC-1 tumor challenge. None of the mice receiving the PEK protein vaccine plus PBS survived more than 60 days of tumor challenge (*: p < 0.001, log-rank test). All experiments were performed independently in triplicate (mean ± SEM). (C) Anti-tumor effects of mice treated with PEK protein vaccine with or without anti-PD-L1 Ab. Tumor volumes were significantly lower in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than in mice treated with the PEK protein vaccine plus PBS (*: p < 0.001, Student's t-test). (D) Overall survival of mice treated with PEK protein vaccine with or without the anti-PD-L1 Ab. Eighty percent of mice that received PEK protein vaccination combined with anti-PD-L1 Ab were still alive 100 days after TC-1 tumor challenge. None of the mice receiving the PEK protein vaccine plus PBS survived more than 60 days of tumor challenge (*: p < 0.001, log-rank test). All experiments were performed independently in triplicate (mean ± SEM).

Figure 2.
Antigen-specific immunoprofiles of mice treated with PEK protein vaccine with or without anti-PD-L1 Ab, and in vitro or ex vivo tumor-specific killing activities of antigen-specific CD8 + cytotoxic T cells or splenocytes from various immunized groups. (A) Representative figures of E7specific IFN-γ-secreting CD4 + helper or CD8 + cytotoxic T cell precursors/3.5 × 10 5 splenocytes on day 28 after tumor challenge in various vaccinated groups, as determined by flow cytometry analysis. (B) Bar figures depict the number of E7-specific IFN-γ-secreting CD4 + helper T cell precursors/3.5 × 10 5 splenocytes on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). Mice treated with the PEK protein vaccine combined with anti-PD-L1 Ab had significantly higher numbers of E7-specific CD4 + T cell precursors than the other groups (*: p < 0.001, Student's t-test). (C) Bar figures depict the number of E7-specific IFN-γ-secreting CD8 + cytotoxic T cell precursors/3.5 × 10 5 splenocytes on day 28 after tumor challenge in various groups, as detected by flow cytometry (n = 5 per group, mean ± SEM). The number of E7-specific CD8 + T cell precursors was highest in mice treated with the PEK protein vaccine combined with anti-PD-L1 Ab (*: p = 0.002, Student's t-test). All experiments were performed independently in triplicate. (D) Representative figures of the in vitro tumor-specific killing activities of E7-specific CD8 + cytotoxic T cells treated with or without anti-PD-L1 Ab. (E) Quantitation of the average luminescence of TC-1-LG cells from mice in various treatment groups. The T cells from mice treated with anti-PD-L1 Ab exhibited lower luminescence than those from mice not treated with anti-PD-L1 Ab (*: p < 0.001, Student's t-test). (F) Representative figures of the ex vivo tumor-specific killing activities of splenocytes from mice on day 28 after tumor challenge in various groups. (G) Quantitation of the average luminescence of TC-1-LG cells co-cultured with splenocytes from mice on day 28 after tumor challenge in various treatment Figure 2. Antigen-specific immunoprofiles of mice treated with PEK protein vaccine with or without anti-PD-L1 Ab, and in vitro or ex vivo tumor-specific killing activities of antigen-specific CD8 + cytotoxic T cells or splenocytes from various immunized groups. (A) Representative figures of E7-specific IFN-γ-secreting CD4 + helper or CD8 + cytotoxic T cell precursors/3.5 × 10 5 splenocytes on day 28 after tumor challenge in various vaccinated groups, as determined by flow cytometry analysis. (B) Bar figures depict the number of E7-specific IFN-γ-secreting CD4 + helper T cell precursors/3.5 × 10 5 splenocytes on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). Mice treated with the PEK protein vaccine combined with anti-PD-L1 Ab had significantly higher numbers of E7-specific CD4 + T cell precursors than the other groups (*: p < 0.001, Student's t-test).
(C) Bar figures depict the number of E7-specific IFN-γ-secreting CD8 + cytotoxic T cell precursors/3.5 × 10 5 splenocytes on day 28 after tumor challenge in various groups, as detected by flow cytometry (n = 5 per group, mean ± SEM). The number of E7-specific CD8 + T cell precursors was highest in mice treated with the PEK protein vaccine combined with anti-PD-L1 Ab (*: p = 0.002, Student's t-test). All experiments were performed independently in triplicate. (D) Representative figures of the in vitro tumor-specific killing activities of E7-specific CD8 + cytotoxic T cells treated with or without anti-PD-L1 Ab. (E) Quantitation of the average luminescence of TC-1-LG cells from mice in various treatment groups. The T cells from mice treated with anti-PD-L1 Ab exhibited lower luminescence than those from mice not treated with anti-PD-L1 Ab (*: p < 0.001, Student's t-test). (F) Representative figures of the ex vivo tumor-specific killing activities of splenocytes from mice on day 28 after tumor challenge in various groups. (G) Quantitation of the average luminescence of TC-1-LG cells co-cultured with splenocytes from mice on day 28 after tumor challenge in various treatment groups (n = 5 per group, mean ± SEM). Splenocytes from mice treated with PEK protein vaccine plus anti-PD-L1 Ab exhibited lower luminescence than splenocytes from mice treated with the PEK protein vaccine plus PBS (*: p = 0.017, Student's t-test). All experiments were performed independently in triplicate. Representative luminescence figures of TC-1/LG tumor cells co-cultured with E7-specific CD8 + cytotoxic T cells with or without anti-PD-L1 Ab are shown in Figure 2D. The E7-specific CD8 + cytotoxic T cells treated with anti-PD-L1 Ab (8.5 ± 0.5 × 10 7 p/s) had significantly less luminescence than those treated with isotype Ab (1.5 × 10 8 ± 2.8 × 10 6 p/s, p < 0.001, Student's t-test; Figure 2E) The ex vivo tumor killing effects of splenocytes were evaluated further. The representative luminescence activities of TC-1/LG cells co-cultured with splenocytes from mice vaccinated with the PEK protein vaccine with or without anti-PD-L1 Ab are shown in Figure 2F. The PEK protein vaccine with anti-PD-L1 Ab (7.7 ± 0.6 × 10 7 p/s) resulted in significantly lower luminescence activity than the PEK protein vaccine alone (9.8 ± 0.2 × 10 7 p/s, p = 0.017, Student's t-test; Figure 2G).  Figure 3D) from splenocytes of mice vaccinated with the PEK protein vaccine are shown. Only low percentages of these cells expressed the PD-L1 molecule. We also investigated whether other cell types, such as BMM-derived DCs and macrophages, expressed PD-L1. As shown in Figure 3E, 11.8% of CD80 − CD11c + DCs and 9.2% of CD86 − CD11c + DCs expressed PD-L1. In contrast, only 4.1% of CD80 + CD11c + DCs and 3.5% of CD86 + CD11c + DCs expressed PD-L1 ( Figure 3E). PD-L1 expression was more common among MHC-II + M1-like macrophages: 87.4% of F4/80 + ( Figure 3F) and 23.5% of MHC-II + M1-like macrophages ( Figure 3G) expressed PD-L1. In addition, IFN-γ induced PD-L1 expression in 83.2% of F4/80 + M0 macrophages ( Figure 3F). Only 7.7% of CD19 + B lymphocytes expressed PD-L1 ( Figure 3H). PD-L1 expression was not limited to immunocytes; the TC-1 tumor cells also expressed PD-L1 ( Figure 3I).

Anti-PD-L1 Ab Enhanced the Maturation and M1-Like Polarization of Macrophages of Regional LNs
DCs and macrophages in tumor-draining lymph nodes was used to investigate whether the blockade of PD-L1 could enhance the activation of APCs in mice vaccinated with PEK protein vaccine are shown in Figure 4A. The percentages of CD80 + CD11c + DCs (2.89 ± 0.05%) was significantly higher among mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine alone (2.11 ± 0.05%, p < 0.001, Student's t-test; Figure 4B). The percentage of CD86 + CD11c + DCs (0.91 ± 0.03%) was also significantly higher among mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine alone (0.63 ± 0.01%, p < 0.001, Student's t-test; Figure 4B). In addition, the percentages of MHC-I + CD11c + and MHC-II + CD11c + DCs were significantly higher in mice treated with the PEK protein vaccine plus anti-PD-L1 antibody than among those treated with the PEK protein vaccine plus PBS (2.13 ± 0.04% vs. 1.42 ± 0.05% and 1.47 ± 0.04% vs. 1.08 ± 0.03%, respectively, p < 0.001 for both, Student's t-test; Figure 4B).

Anti-PD-L1 Ab Enhanced the Maturation and M1-Like Polarization of Macrophages of Regional LNs
DCs and macrophages in tumor-draining lymph nodes was used to investigate whether the blockade of PD-L1 could enhance the activation of APCs in mice vaccinated with PEK protein vaccine are shown in Figure 4A. The percentages of CD80 + CD11c + DCs (2.89 ± 0.05%) was significantly higher among mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine alone (2.11 ± 0.05%, p < 0.001, Student's t-test; Figure 4B). The percentage of CD86 + CD11c + DCs (0.91 ± 0.03%) was also significantly higher among mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine alone (0.63 ± 0.01%, p < 0.001, Student's t-test; Figure 4B). In addition, the percentages of MHC-I + CD11c + and MHC-II + CD11c + DCs were significantly higher in mice treated with the PEK protein vaccine plus anti-PD-L1 antibody than among those treated with the PEK protein vaccine plus PBS (2.13 ± 0.04% vs. 1.42 ± 0.05% and 1.47 ± 0.04% vs. 1.08 ± 0.03%, respectively, p < 0.001 for both, Student's t-test; Figure 4B). The M1 markers, including CD80, CD86, and MHC-II molecules, of macrophages in vivo were also evaluated. The percentages of CD80 + F4/80 + and CD86 + F4/80 + macrophages were higher in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine plus PBS (1.76 ± 0.09% vs. 1.55 ± 0.02%, p = 0.03, and 0.89 ± 0.03% vs. 0.78 ± 0.02%, p = 0.041, Student's t-test for both; Figure 4C). In addition, the percentage of MHC-II + F4/80 + macrophages was higher in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine plus PBS (1.55 ± 0.05% vs. 1.06 ± 0.02%, p < 0.001, Student's t-test; Figure 4C).

Anti-PD-L1 Ab Enhanced the Infiltration of CD4 + and CD8 + T Cells into the Tumors of Mice Vaccinated with Antigen-Specific Protein Vaccine
The protocols used to investigate whether the blockade of PD-L1 could enhance the anti-tumor immunities elicited by antigen-specific protein vaccine in tumor-infiltrating lymphocytes (TILs) are shown in Figure 5A. Representative results of the flow cytometric analysis of CD4 + and CD8 + T cells from TILs are shown in Figure 5B. The percentages of both CD4 + and CD8 + T lymphocytes were significantly higher in TILs from mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than The percentage of CD11c + DCs expressing CD80 significantly increased in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab compared with mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of CD11c + DCs expressing CD86 significantly increased in mice treated with the PEK vaccine plus anti-PD-L1 Ab compared with mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of MHC-I-expressing CD11c + DCs was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of MHC-IIexpressing CD11c + cells was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). (C) Percentages of activation markers of F4/80 + macrophages in tumor-draining lymph nodes from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). The percentage of CD80-expressing F4/80 + macrophages was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p = 0.03, Student's t-test). The percentage of CD86-expressing F4/80 + macrophages was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p = 0.041, Student's t-test). The percentage of MHC-IIexpressing F4/80 + macrophages was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). All experiments were performed independently in triplicate.
The M1 markers, including CD80, CD86, and MHC-II molecules, of macrophages in vivo were also evaluated. The percentages of CD80 + F4/80 + and CD86 + F4/80 + macrophages were higher in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK The percentage of CD11c + DCs expressing CD80 significantly increased in mice treated with the PEK protein vaccine plus anti-PD-L1 Ab compared with mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of CD11c + DCs expressing CD86 significantly increased in mice treated with the PEK vaccine plus anti-PD-L1 Ab compared with mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of MHC-I-expressing CD11c + DCs was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of MHC-II-expressing CD11c + cells was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). (C) Percentages of activation markers of F4/80 + macrophages in tumor-draining lymph nodes from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). The percentage of CD80-expressing F4/80 + macrophages was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p = 0.03, Student's t-test). The percentage of CD86-expressing F4/80 + macrophages was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p = 0.041, Student's t-test). The percentage of MHC-II-expressing F4/80 + macrophages was significantly higher in mice treated with the PEK vaccine plus anti-PD-L1 Ab than among those treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). All experiments were performed independently in triplicate. shown in Figure 5A. Representative results of the flow cytometric analysis of CD4 + and CD8 + T cells from TILs are shown in Figure 5B. The percentages of both CD4 + and CD8 + T lymphocytes were significantly higher in TILs from mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than among those treated with the PEK protein vaccine plus PBS (15.42 ± 0.65% vs. 3.30 ± 0.05% and 12.14 ± 0.38% vs. 3.71 ± 0.20%, respectively, p = 0.002 for both, Student's t-test; Figure 5C).
There was an increase in numbers of CD4 + and CD8 + T lymphocytes in the tumors of mice that were controlled by antigen-specific protein vaccine combined with blockade of PD-L1.  Representative figures depicting tumor-infiltrating CD4 + and CD8 + T lymphocytes from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry. (C) Percentages of CD4 + and CD8 + T lymphocytes from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). The percentage of CD4 + T lymphocytes was higher in tumors from mice treated with PEK protein vaccine plus anti-PD-L1 Ab than in tumors from mice treated with PEK vaccine plus PBS (*: p = 0.002, Student's t-test). The percentage of CD8 + T lymphocytes was also higher in tumors from mice treated with PEK vaccine plus anti-PD-L1 Ab than in tumors from mice treated with PEK vaccine plus PBS (*: p = 0.002, Student's t-test). (D) Representative figures of tumor-infiltrating CD80 + CD11c + and CD86 + CD11c + DCs from mice on day 28 after tumor challenge in various groups as determined by flow cytometry. (E) Percentages of CD80 + CD11c + and CD86 + CD11c + DCs from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). The CD80 maturation marker of CD11c + DCs significantly increased in tumors from mice treated with the PEK protein vaccine plus anti-PD-L1 Ab compared with tumors from mice treated with the PEK vaccine plus PBS (*: p = 0.002, Student's t-test). The CD86 maturation marker of CD11c + DCs also significantly increased in tumors from mice treated with the PEK protein vaccine plus anti-PD-L1 Ab compared with tumors from mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). (F) Representative figures depicting tumor-infiltrating CD80 + F4/80 + and CD86 + F4/80 + macrophages from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry. (G) Percentages of tumor-infiltrating CD80 + F4/80 + and CD86 + F4/80 + macrophages from mice on day 28 after tumor challenge in various groups, as determined by flow cytometry (n = 5 per group, mean ± SEM). The percentage of CD80-expressing F4/80 + macrophages was significantly higher in tumors from mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than in tumors from mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). The percentage of CD86-expressing F4/80 + macrophages was significantly higher in tumors from mice treated with the PEK protein vaccine plus anti-PD-L1 Ab than in tumors from mice treated with the PEK vaccine plus PBS (*: p < 0.001, Student's t-test). All experiments were performed independently in triplicate. There was an increase in numbers of CD4 + and CD8 + T lymphocytes in the tumors of mice that were controlled by antigen-specific protein vaccine combined with blockade of PD-L1.
The anti-PD-L1 Ab promoted the maturation of DCs, and these antigen-pulsed mature DCs were able to enhance the activation of antigen-specific CD8 + cytotoxic T precursors.
The anti-PD-L1 Ab promoted the maturation of DCs, and these antigen-pulsed mature DCs were able to enhance the activation of antigen-specific CD8 + cytotoxic T precursors.

Discussion
The results of the present study revealed that PD-L1 blockade enhanced the antigen-specific anti-tumor effects and cell-mediated immunities of an antigen-specific protein vaccine beyond those

Discussion
The results of the present study revealed that PD-L1 blockade enhanced the antigen-specific anti-tumor effects and cell-mediated immunities of an antigen-specific protein vaccine beyond those of the protein vaccine alone. The anti-PD-L1 Ab increased the percentages of tumor-infiltrating CD4 + and CD8 + T cells in mice vaccinated with an antigen-specific protein vaccine. Lin et al. showed that combination therapy of anit-PD-L1 Ab and Lm-LLO-E6 vaccine which secreted HPV-16 E6 antigen fused to listeriolysin O (LLO) pore-forming toxin by bacterial vector could exhibit potent anti-tumor efficacy against HPV-16 E6-positive human TL-1 lung cancer in nude mice [22]. In comparison with Lin et al.'s study, our investigation demonstrated that anti-PD-L1 Ab could enhance antigen-specific anti-tumor effects by modulating immune cells in immune-competent mice, compared to different anti-tumor mechanisms through neutralizing E6 protein by Lm-LLO-E6 vaccine to further inhibit PD-L1 expression, and consequently could suppress tumor growth in immunocompromised mice [22].
Recently, increasing evidence has emerged in support of the hypothesis that the anti-tumor mechanisms of anti-PD-L1 Ab in cancer immunotherapy come mainly through its targeting of PD-L1-expressed antigen-presenting cells, such as DCs and macrophages [10,23]. In the current study, the anti-PD-L1 Ab also enhanced the maturation of DCs and M1-like polarization of macrophages in both regional LNs and tumors in mice. In addition, anti-PD-L1 Ab enhanced the maturation of BMM-derived DCs in vitro. The blockade of PD-L1 enhanced the anti-tumor efficacy of an E7 antigen-specific protein vaccine by modulating the maturation of DCs and M1-like polarization in macrophages in an E7-expressing small tumor model.
DCs play a central role in antigen-specific anti-tumor T cell immune responses. The maturation status of DCs determines their potency during antigen presentation. Immature DCs that reside in nonlymphoid tissues are able to engage in antigen uptake and processing, but do not provide the signals needed to prime T cell responses [24]. Mature DCs induced by extraneous signals are able to migrate to secondary lymphoid organs to upregulate immunogenicity to initiate T cell responses and downregulate antigen capture and antigen-processing abilities [24,25]. Additionally, tumor cells are able to reverse these mature, tumor-infiltrating, immune-active anti-tumor DCs into immune-suppressive, pro-tumor DCs by upregulating inhibitory receptors, including PD-1 and T cell immunoglobulin and mucin domain 3 (TIM-3) molecules on DCs [26].
Macrophages tend to be polarized toward an M2-like phenotype associated with repair and immune suppression (pro-tumor), rather than the M1-like phenotype associated with inflammation and immunity (anti-tumor) in established tumors [27,28]. Interleukin (IL)-10 secreted by tumor-associated M2-like macrophages inhibits CD8 + T cell responses to chemotherapy by suppressing the inflammatory cytokine IL-12 from tumor-infiltrating DCs [29]. In addition, arginase-1 induction and PD-L1 expression by tumor-associated M2-like macrophages can suppress T cell immune responses [30,31]. Therefore, identifying ways to promote the maturation of DCs and M1 shifting of macrophages has became an important strategy for cancer immunotherapy.
Our previous study demonstrated that the PEK protein vaccine is able to generate potent antigen-specific immunity, including antigen-specific CD8 + and CD4 + T cell precursors, antigen-specific Abs, and anti-tumor effects to control micro-metastatic tumor lesions in a pulmonary metastatic model [20]. However, the PEK protein vaccine alone is not potent enough to control established tumors in a subcutaneous tumor model ( Figure 1C,D). Therefore, a strategy is required to enhance the anti-tumor effects of the antigen-specific protein vaccine.
In addition to tumor cells, PD-L1 is upregulated on APCs, including DCs and macrophages, in tumor microenvironments due to inflammatory stimuli [32]. Tumor cells or APCs expressing PD-L1 inhibitory ligand can downregulate the amplitude of T cell activation and suppress anti-tumor immunity through the PD-1/PD-L1 axis [10,33]. Moreover, in addition to PD-L1-expressing tumor cells, PD-L1-expressing tumor-infiltrating myeloid cells appeared to provide a compensatory source of the inhibitory ligand to downregulate the anti-tumor immune responses mediated by T cells in mice [34]. Lin et al. further demonstrated that PD-L1 expressed on host immune cells, including DCs and macrophages, rather than PD-L1 intrinsically expressed on tumor cells could account for the anti-tumor efficacy of anti-PD-L1 Ab monotherapy [23]. Lin et al. first demonstrated that E7-specific Sig/E7/LAMP-1 vaccinia could generate in vivo protection against TC-1 tumor [35]. They also demonstrated that treatment with the Sig/E7/LAMP-1 vaccinia vaccine could only cure mice with small TC-1 tumors [35]. Our study revealed that PD-L1 blockade could enhance the anti-tumor effects of E7-sepcific PEK protein vaccine against established TC-1 subcutaneous tumors. PD-L1 was also highly expressed on APCs, including immature DCs, M1 macrophages, IFN-γ-induced M0 macrophages, and B cells in our study ( Figure 3E-H), which implies that the blockade of the PD-1/PD-L1 axis in these cells might be a potential strategy to enhance anti-tumor immunity by blocking the inhibitory signaling pathway.
Antibody-mediated blockade of PD-L1 has shown an induced unprecedented durable response in patients with advanced cancers [36]. Additionally, PD-L1 expressed in APCs rather than on tumor cells could play an essential role in PD-L1 blockade therapy [10]. However, the majority of cancers present resistance to PD-L1 blockade monotherapy due to a lack of pre-existing tumor-infiltrating anti-tumor CD8 + T cells [37,38]. Cancer vaccines could allow the priming and intratumoral recruitment of CD8 + T cells and transform a "non-inflamed" non-permissive tumor resistant to checkpoint blockade into a sensitive "inflamed" tumor [39]. Our survey revealed that the combination of an antigen-specific protein vaccine with anti-PD-L1 therapy increased the number of tumor-infiltrating CD4 + and CD8 + T cells within tumors ( Figure 5B,C). Compared with our anti-PD-L1 primary resistance tumor model, innate immune agonists such as TLR agonists, the Sting pathway, and oncolytic viruses might enhance the anti-tumor effect treated with anti-PD-L1 Ab in PD-L1 sensitive tumor models.
PD-L1 expression on DCs can attenuate the anti-tumor effects of effector T cells by interacting with PD-1 on the surface of T lymphocytes to induce T-cell apoptosis, anergy, and exhaustion [40]. Blockade of PD-L1 has been shown to promote the maturation and function of DCs in the peripheral blood of patients with colorectal cancer in vitro [41]. We determined that anti-PD-L1 Ab was able to enhance DC maturation ( Figure 6A,B) and activate antigen-specific cytotoxic CD8 + T cells ( Figure 6C,D). The results of the in vitro experiments were demonstrated further in the tumor microenvironment in vivo ( Figure 5D,E). Recently, Lin et al. also showed that the PD-L1 expression of intratumoral macrophages rather than that of tumor cells could account for the potential anti-tumor efficacy of PD-L1 blockade therapy [23]. We also demonstrated that anti-PD-L1 Ab improved the proportion of M1 in macrophages of draining LNs ( Figure 4C) and tumor microenvironments ( Figure 5F,G).

Mice
Female C57BL/6 mice, 6 to 8 weeks old, were purchased from the National Laboratory Animal Center (Taipei, Taiwan) and bred in the animal facility of the School of Medicine of National Taiwan University. All of the animal procedures were conducted according to approved protocols and in accordance with recommendations for the proper use and care of laboratory animals. (Animal handling and procedures were approved by the animal ethic committee of College of Medicine, National Taiwan University. Ethic Code: "20140053" and "20190074").

Administration of Anti-PD-L1 Ab
Rat anti-mouse PD-L1 (clone 10F.9G2) Ab and rat IgG2b isotype Ab (clone LTF-2) were purchased from BioXCell (West Lebanon, NH, USA). The anti-PD-L1 Ab was diluted with PBS and 200 µg was intraperitoneally injected into each mouse at the indicated intervals.

In Vivo Tumor Treatment Experiments with Protein Vaccine and/or Anti-PD-L1 Ab
The treatment protocols for PEK protein vaccine and/or anti-PD-L1 Ab are presented in Figure 1A. Tumor diameter was measured using calipers weekly starting 7 days after tumor challenge; the tumor volume was defined by 4πR 3 /3, where R is the radius of the tumor. Mice were euthanized when their tumor reached 2 cm in diameter or when they appeared sick; this was recorded as death for the survival curve.

In Vitro and Ex Vivo Tumor-Killing Activity
E7-specific CD8 + T cells were cultured with irradiated TC-1/LG tumor cells (1:4 ratio) in a 96-well plate (2 × 10 4 cells/well) for 24 h in the presence of anti-PD-L1 Ab or isotype Ab (10 µg/mL). Luciferin (Promega) was added and the total flux (p/s) from each well was measured using IVIS Imaging Systems (Caliper Life Sciences, Alameda, CA, USA).
Splenocytes were harvested from various groups (=on day 28 after tumor challenge) as described earlier, and then co-cultured with irradiated TC-1/LG cells (10:1 ratio) in a 96-well plate (2 × 10 4 cells/well) for 24 h, and luciferin was added to detect the total flux (p/s) from each well, as described earlier.

Preparation of BMM-Derived DCs and Macrophages
Bone marrow monocyte (BMM)-derived dendritic cells (DCs) [45] or macrophages [46] from mice were prepared as described previously, with some modifications. Bone marrow cells were collected from the femurs and tibias by flushing, and the cells were filtered through a 70 µm cell strainer (BD Falcon, San Jose, CA, USA). Red blood cells were removed using RBC lysis buffer (eBioscience, San Diego, CA, USA).

PD-L1 Expression of Immune Cells and Tumor Cells Analyzed by Flow Cytometric Analysis
To determine the expression of PD-L1 by various immune cells and mature/immature BMM-derived DCs or M0/M1 BMM-derived macrophages, these cells were obtained, cultured, and harvested as described previously. To evaluate the effect of IFN-γ on PD-L1 expression of macrophages, BMM-derived M0 macrophages were treated with 500 U/mL recombinant murine IFN-γ (PeproTech) in a 5% CO 2

In Vivo Maturation Status of DCs and M1-Like Polarization of Macrophages from Regional LNs
Tumor-draining LNs were harvested from mice treated with E7-specific protein vaccine and/or anti-PD-L1 Ab 7 days after the last protein vaccination (on day 28 after tumor challenge) ( Figure 4A). Single cell suspensions were prepared as described previously [47]. To detect the maturation status of DCs and the M1-like polarization of macrophages, the cells were stained with FITC-conjugated anti-CD11c Ab, FITC-conjugated anti-F4/80 Ab, PE-conjugated anti-CD80 Ab, and PE-Cy5-conjugated anti-CD86 Ab (BioLegend) or PE-conjugated anti-MHC class I and PE-Cy5-conjugated anti-MHC class II Abs (BioLegend). The cells were analyzed by flow cytometry, as described above.

Isolation of Tumor-Infiltrating Lymphocytes (TILs)
TILs were prepared as described previously, with some modifications [47]. Briefly, mice were challenged with TC-1 tumor cells and immunized with PEK protein vaccine and/or anti-PD-L1 Ab. The mice were sacrificed and tumors excised 7 days after the last protein immunization (on day 28 after tumor challenge) ( Figure 5A). The tumors were dissected into small fragments and digested in 0.1 mg/mL collagenase in CTL medium at 37 • C overnight. After filtering through a 40 µm cell strainer (BD Falcon), the cell suspension was incubated for 30 min at 37 • C. After washing with CTL medium, mixing cell suspensions of CTL medium and balanced salt medium were layered on Ficoll-Paque medium (GE Healthcare, Pittsburgh, PA, USA) before centrifugation. TILs from the white interface layer were collected and washed with PBS.

Maturation Status of BMM-Derived DCs Treated with Anti-PD-L1 Ab
To determine whether anti-PD-L1 Ab enhanced DC maturation, BMM-derived DCs treated with 50 ng/mL LPS were used as a positive control. DCs were treated with or without LPS and anti-PD-L1 Ab (50 µg/mL) in a 5% CO 2 atmosphere at 37 • C overnight. The cells were stained with FITC-conjugated anti-CD11c Ab (BioLegend), PE-conjugated anti-CD80 Ab (BioLegend), anti-MHC class I Ab (Biolegend), PE-Cy5-conjugated anti-CD86 Ab (BioLegend) or anti-MHC class II Ab (Biolegend). The cells were analyzed by flow cytometry, as described above.

Statistical Analysis
All of the data were expressed as mean ± SEM (standard error), which represented at least three different experiments. Student's t-test was used to detect statistical significance. The log-rank test was used to evaluate data from survival experiments. All p values < 0.05 were considered statistically significant.

Conclusions
The strategy of PD-L1 blockade had the effect of increasing the potency of the anti-tumor effects of the E7 antigen-specific protein vaccine, which could directly increase the susceptibility of tumor cells to killing and indirectly enhance the antigen-specific CD8 + T-cell responses through promoting DC maturation and M1-like polarization of macrophages to overturn the immuno-suppressive tumor microenvironment in an E7-expressing tumor model. Therefore, tumor specific antigen (like HPV E7 antigen)-specific immunotherapy combined with APCs targeting modality by PD-L1 blockade has high translational potential in E7-specific cancer therapy.