PD-1 Blockade Unleashes Effector Potential of Both High- and Low-Affinity Tumor-Infiltrating T Cells

Amaia Martínez-Usatorre, Alena Donda, Dietmar Zehn and Pedro Romero
J Immunol July 15, 2018, 201 (2) 792-803; DOI: https://doi.org/10.4049/jimmunol.1701644

Abstract

Antitumor T cell responses involve CD8+ T cells with high affinity for mutated self-antigen and low affinity for nonmutated tumor-associated Ag. Because of the highly individual nature of nonsynonymous somatic mutations in tumors, however, immunotherapy relies often on an effective engagement of low-affinity T cells. In this study, we studied the role of T cell affinity during peripheral priming with single-peptide vaccines and during the effector phase in the tumor. To that end, we compared the antitumor responses after OVA257–264 (N4) peptide vaccination of CD8+ T cells carrying TCRs with high (OT-1) and low (OT-3) avidity for the N4 peptide in B16.N4 tumor-bearing C57BL/6 mice. Additionally, we assessed the response of OT-1 cells to either high-affinity (B16.N4) or low-affinity (B16.T4) Ag-expressing tumors after high-affinity (N4) or low-affinity (T4) peptide vaccination. We noticed that although low-affinity tumor-specific T cells expand less than high-affinity T cells, they express lower levels of inhibitory receptors and produce more cytokines. Interestingly, tumor-infiltrating CD8+ T cells show similar in vivo re-expansion capacity to their counterparts in secondary lymphoid organs when transferred to tumor-free hosts, suggesting that T cells in tumors may be rekindled upon relief of tumor immunosuppression. Moreover, our results show that αPD-1 treatment enhances tumor control of high- and low-affinity ligand-expressing tumors, suggesting that combination of high-affinity peripheral priming by altered peptide ligands and checkpoint blockade may enable tumor control upon low-affinity Ag recognition in the tumor.

Introduction

CD8+ T cells play a critical role in the immune response against intracellular pathogens and cancer cells. Upon Ag recognition, naive CD8+ T cells proliferate and follow an orchestrated differentiation program, giving rise to different CD8+ T cell subsets with different phenotype, function, and anatomical location. The majority of newly activated CD8+ T cells differentiate into primary cytotoxic effector cells, which migrate to clear pathogen-infected or transformed cells (1). A small proportion of the expanded T cells differentiate into quiescent memory CD8+ T cells that may persist for a lifetime. Upon re-encounter with Ag, memory T cells rapidly generate secondary cytotoxic effectors, but some of them are renewed as long-lived memory T cells. Several T cell differentiation models have been proposed to explain the formation of effector and memory T cells during an immune response: the asymmetric division model (2), the decreasing potential model (3), and the signal strength model. According to the last model, costimulatory molecules, proinflammatory cytokines, and the strength of Ag recognition during T cell priming determine the expansion amplitude and CD8+ T cell fate (4).

Three terms may describe the interaction strength between a T cell and APC/target cell: T cell affinity, avidity, and functional avidity. T cell affinity is the physical binding strength between a single TCR and peptide–MHC (pMHC) molecule, whereas T cell avidity is the interaction strength of multimeric TCR–pMHCs. Nonetheless, T cell avidity correlates with T cell affinity (5). Finally, functional avidity is the sensitivity of a T cell to the Ag, which inversely correlates with the Ag dose needed for a T cell response. It normally correlates with T cell avidity, but many factors, such as TCR expression levels, coreceptors, adhesion molecules, and inhibitory and costimulatory molecules, can modulate the sensitivity of a T cell for its Ag.

TCR–pMHC recognition is essential for T cell activation and differentiation. In fact, TCR affinity affects T cell–APC interaction (6, 7) and survival (8) and directly correlates with the amplitude of T cell expansion in bacterial infections (9) without impacting phenotypic differentiation. In contrast, in the lymphocytic choriomeningitis virus (LCMV) clone 13 chronic viral infection model, it was shown that low-affinity stimulation leads to decreased PD-1 expression and increased frequencies of IFN-γ and TNF-α–producing CD8+ T cells (10), suggesting that, in the case of Ag persistence, phenotype and functionality of CD8+ T cells are impacted by the TCR–pMHC interaction strength.

Tumor-specific CD8+ T cells with increased affinity for the ligand show enhanced intracellular signaling, proliferation, and target cell lysis in vitro (11, 12). However, unlike viral and bacterial Ags, most tumor-associated Ags (TAAs) are self-antigens. As a consequence, high-affinity tumor-reactive CD8+ T cells are eliminated during thymic selection, and the remaining tumor-reactive T cells found in cancer patients recognize the Ag with low avidity (13). Recently, it is becoming increasingly clear that nonsynonymous somatic mutations are a source of neoantigens. T cell recognition of these tumor neoantigens is presumably of high affinity because of the absence of central tolerance.

Cancer vaccines using TAAs have been shown to induce TAA-specific T cell responses in cancer patients in combination or not with other therapies (14–16). However, not all TAAs can elicit specific T cell responses (17, 18), and even in cases of specific immune responses, vaccination with TAAs did not confer significant benefit in clinical trials (19). Although vaccination with neoantigens may induce tumor-protective CD8+ T cell responses, the need to identify and formulate neoantigen-based vaccines in a personalized manner imposes a major hurdle and prohibitive cost. Therefore, it is imperative to understand how T cells differentiate and respond to vaccination with nonmutated Ags and to engagement in the tumor microenvironment, where the target cells may express both tumor-associated self-antigens and neoantigens.

In this study, we studied the role of TCR–pMHC interaction strength during peripheral priming with a peptide vaccine and in the effector phase within the tumor. To that end, we first compared the antitumor responses after peptide vaccination of CD8+ T cells carrying TCRs with high or low avidity for the ligand expressed by the tumor. Additionally, we also assessed the response of CD8+ T cells with a defined TCR to either high- or low-affinity Ag-expressing tumors after high- or low-affinity peptide vaccination. We observed that the affinity for the Ag in the vaccine formulation and the affinity for the ligand expressed by the tumor determined the expansion and differentiation of CD8+ T cells, whereby low-affinity stimulation led to decreased tumor control. Nonetheless, low-affinity stimulation in the tumor led to decreased PD-1 expression and increased IFN-γ and TNF-α–producing CD8+ T cell frequencies. Moreover, both high- and low-affinity stimulated tumor-infiltrating CD8+ T cells showed the same re-expansion capacity when challenged with a high-affinity bacterial infection in a tumor-free host. Importantly, αPD-1 treatment enhanced tumor control of either high- or low-affinity ligand-expressing tumors. Thus, TAA-specific low-affinity CD8+ T cells found in cancer patients may also benefit from αPD-1 treatment.

Materials and Methods

Mice

C57BL/6 mice were obtained from Envigo, OT-1 transgenic mice from The Jackson Laboratory, and OT-3 transgenic mice from D. Zehn (20). Mice were all females, at least 7 wk old at the beginning of the experiment, and were maintained in conventional facilities of the University of Lausanne (UNIL). This study was approved by the Veterinary Authority of the Swiss canton Vaud and performed in accordance with Swiss ethical guidelines.

Generation of B16.N4 and B16.T4 cell lines

B16-F10 cell line was transduced with MigR1-N4-GFP or MigR1-T4-GFP retroviral vectors provided by D. Zehn. Transduced cells were cloned, and cell lines with similar GFP expression were selected.

Melanoma tumor models

For C57BL/6 mice, 105 B16.OVA or 2 ×105 B16.N4 and 2 ×105 B16.T4 cells were s.c. engrafted on each flank. After 6 d, CD45.1 105 OT-1 or 106 OT-3 T cells were i.v. transferred. One day later, mice were s.c. vaccinated with 10 μg SIINFEKL (N4) or SIITFEKL (T4) peptide (Protein and Peptide Chemistry Facility, UNIL) and 50 μg CpG (CpG-ODN 1826, U133-L01A) (TriLink BioTechnologies). A second s.c. or intratumoral (i.t.) vaccination was performed on day 14 with 50 μg CpG alone or in combination with 10 μg N4 peptide. Mice were grouped so that tumor size mean and SD were similar among groups before the second vaccination. Tumors were measured manually with a caliper every 2 d from day 6 or 8. Spleens and tumors were harvested 14 and 21 d after tumor engraftment. Spleens were mashed through a 100-μm diameter filter (Falcon), and RBCs were lysed with RBC lysis buffer (Qiagen). Tumors were dissociated with Tumor Dissociation Kit (Miltenyi Biotec) following the manufacturer’s instructions.

Immune checkpoint blockade

B16.N4 and B16.T4 tumor-bearing mice received 200 μg of αPD-1 (clone RMP-1-14, rat IgG2a; Bio X Cell) or 2A3 isotype control (ISO) (rat IgG2a; Bio X Cell) on day 10, 13, and 16 after tumor engraftment. Mice were grouped so that tumor size mean and SD were similar among groups before the first treatment.

Listeria monocytogenes N4 infection

A total of 5000 flow cytometry–sorted naive OT-1 cells or 5000 OT-1 cells from spleen, B16.N4, or B16.T4 tumors 14 and 21 d after tumor engraftment were i.v. transferred into naive C57BL/6 mice. Secondary recipients were infected the same day with N4-expressing L. monocytogenes (Lm-N4) provided by D. Zehn. Eight days postinfection, spleens from infected mice were collected and processed for flow cytometry analysis of OT-1 cells.

BrdU administration and staining

Seven days after tumor engraftment, mice received 1.8 mg BrdU (B5002; Sigma-Aldrich) i.p. and, from then on, 0.8 mg/ml BrdU in drinking water until the end of the experiment. Intracellular BrdU staining was performed with allophycocyanin BrdU Flow kit (BD Pharmingen) following the manufacturer’s instructions.

Ex vivo stimulation of OT-1 cells

Mouse splenocytes or processed tumor samples were stimulated with 10 μg/ml SIINFEKL peptide or 10 ng/ml PMA (Sigma-Aldrich) and 500 ng/ml ionomycin as positive controls (Sigma-Aldrich) for 30 min at 37°C in 10% FCS, 5 mg/ml penicillin (Invitrogen), 5 mg/ml streptomycin (Invitrogen), 10 mg/ml neomycin (Invitrogen), 0.05 mM 2-ME (Invitrogen), and 10 mM HEPES (Amimed) in DMEM (complete DMEM). Unstimulated cells were used as negative controls. BD GolgiPlug and GolgiStop (BD Biosciences) were added to the cells and incubated for another 4 h at 37°C.

Surface and intracellular Ab staining for flow cytometry

Surface Ab staining was performed for 20 min at 4°C with different combinations of CD3-A700 (cl.17A2; eBioscience), CD8-PE–Texas Red (cl.MCD0817; Life Technologies), CD8-A700 (cl.536.7; FACS facility UNIL), CD45.1 FITC (cl.A20.1; FACS facility UNIL), CD45.2 allophycocyanin-eFluor 780 (cl.104; eBioscience), KLRG1-PE–Cy7 (cl.2F1/KLRG1; BioLegend), CD127-PE (cl.A7R34; eBioscience), PD-1–allophycocyanin (cl29F.1A12; eBioscience), LAG-3–PE (clC9B7W; BD Biosciences), CD44–Pacific Blue (cl.IM781; FACS facility UNIL), CD62L-PE–Cy5 (cl.MEL-14; eBioscience), CD39-PE–Cy7 (cl.24DMS1; eBioscience), CD122–Pacific Blue (cl.TM-1b; eBioscience), CD25-PerCp–Cy5.5 (cl.PC61.5; eBioscience), and CD4-BV605 (cl.RM4-5; BioLegend) in 2 mM EDTA and 2% FBS in PBS (FACS buffer). After surface staining, cells were washed with PBS and stained with LIVE/DEAD Fixable Aqua Dead (L34957; Thermo Fisher Scientific) for 20 min at 4°C. Then, cells were washed again with FACS buffer. For intracellular Foxp3 staining, cells were fixed and permeabilized with Foxp3/Transcription Factor Staining Buffer Set (00-5523-00; eBioscience) according to the manufacturer’s instructions and stained for 30 min at room temperature with Foxp3–Alexa Fluor 488 (cl.MF14; BioLegend). For intracellular staining of cytokines, cells were fixed and permeabilized with Intracellular Staining Permeabilization Wash Buffer (421002; BioLegend) according to the manufacturer’s instructions and stained for 20 min at 4°C with IFN-γ–PerCp-Cy5.5 (cl.XM61.2; eBioscience), TNF-α–Pacific Blue (cl.MP6-XT22; BioLegend), CTLA-4-PE (cl.UC19-4F10-11; BD Biosciences), and granzyme B–PE–Texas Red (GRB17; Molecular Probes). Finally, cells were washed and resuspended in FACS buffer for flow cytometry analysis with an LSRI II flow cytometer (BD Biosciences).

For flow cytometry cell sorting, cells were sorted with FACSAria (BD Biosciences) and collected in complete DMEM.

Transferred CD45.1 OT-1 or OT-3 cells were identified as CD45.1+ cells on the CD8+ gate in single-cell (side scatter [SSC]-width versus SSC-height) live (Vivid−) lymphocytes (forward scatter–area versus SSC-area). Naive cells were defined as CD44low CD62L+, central memory (CM) as CD44high CD62L+, and effector cells as CD44high CD62L− cells. Memory precursor effector cells (MPECs) and terminal effector cells (TEs) were defined as CD127+KLRG1− and CD127+KLRG1−, respectively, on CD44+ OT-1 or OT-3 cells. Tregs were defined as Foxp3+ cells on CD3+CD4+ cells.

Data analysis and statistics

Flow cytometry data were analyzed with FlowJo (Tree Star). Graphs and statistical analysis were made with Prism (GraphPad Software). Specific statistical analyses are described in the figure legends. Overall, normality of data distribution was analyzed by Shapiro–Wilk normality test. Comparison between two unpaired groups was performed by parametric Student t test or nonparametric Mann–Whitney U test. When the two groups were paired, parametric paired t test or nonparametric Wilcoxon test was used. For multiple comparison, a parametric one-way ANOVA or nonparametric Kruskal–Wallis test was performed, followed by Tukey multiple comparison test or Dunn multiple comparison test, respectively. Finally, simultaneous analysis of two variables among multiple groups was performed by two-way ANOVA or two-way repeated measurements (RM) ANOVA, followed by Tukey multiple comparison test. The p values are indicated as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 in the figures. All results are representative of at least two independent experiments.

Results

Reduced T cell expansion of low-avidity OT-3 cells leads to weak tumor control

To assess the role of TCR avidity in CD8+ T cell–mediated antitumor responses, we compared naive OT-1 and OT-3 cells. TCR-transgenic OT-3 cells are OVA (OVA257–264)–specific CD8+ T cells expressing a TCR that responds to OVA with lower functional avidity than OT-1 cells (20). Because we expected that OT-3 cells would expand less than OT-1 cells, we transferred 106 OT-3 cells or 105 OT-1 cells 6 d after B16.OVA tumor engraftment in C57BL/6 mice followed by s.c. vaccination with SIINFEKL OVA (N4) peptide and CpG-ODN 1 d later (21) (Fig. 1A).

FIGURE 1.
FIGURE 1.

Low-avidity OT-3 cells show reduced expansion and tumor control as well as phenotypic differences after peptide–CpG vaccination. (A) Scheme of the experimental design. 105 B16.OVA cells were s.c. engrafted on the flank of C57BL/6J mice. After 6 d, 105 naive OT-1 or 106 naive OT-3 cells were transferred, and mice were s.c. vaccinated with OVA–CpG 1 d later. (B) Tumor growth curve. Dots represent the mean tumor volume in cubic millimeters + SD (n = 10 per group). (C) Total OT-1 or OT-3 cell numbers in spleen and dLNs 14 and 21 d after tumor engraftment. (D) Number of OT-1 or OT-3 cells per cubic millimeter of tumor. (E) Percentage of naive (CD44low CD62L+), CM (CD44high CD62L+), and effector (CD44high CD62L−) cells in OT-1 or OT-3 cells in spleen, dLN, and tumor at day 14. (F) Representative histograms and PD-1 mean fluorescence intensity of endogenous naive CD8+ cells (gray, filled), OT-1 (red, continuous), and OT-3 (blue, dotted) cells from spleen, dLN, and tumor at day 14. (G) Percentage of MPECs (CD127high KLRG1low) and TEs (CD127low KLRG1high) in CD44high OT-1 or OT-3 cells in spleen, dLN, and tumor at day 14. A two-way ANOVA followed by Tukey multiple comparison test was performed in (B), a two-way ANOVA followed by Sidak multiple comparison test in (C), a two-way RM ANOVA followed by Sidak multiple comparison test in (E) and (G), and an unpaired t test in (F). Dots represent individual mice, and the bar represents the mean. OT-1 cells are represented in red and OT-3 cells in blue. Representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Although mice receiving OT-1 cells could control B16.OVA tumor growth, those receiving OT-3 cells showed similar tumor growth to those receiving vaccination alone (Fig. 1B). Despite transferring 10 times more naive OT-3 than OT-1 cells, ∼10 times lower OT-3 cell numbers were detected in spleen, tumor draining lymph nodes (dLN), and tumor 14 d after tumor engraftment. One week later, OT-1 numbers decreased in spleen and dLNs, although they remained in higher numbers in the tumor as compared with OT-3 cells (Fig. 1C, 1D). Thus, curtailed peripheral expansion of low-avidity OT-3 cells may largely explain the reduced tumor-specific T cell numbers in the tumor associated with impaired tumor control.

At the peak of the response, 14 d after tumor engraftment, increased naive (CD44low CD62L+) and decreased effector (CD44high CD62L−) T cell frequencies were found in OT-3 compared with OT-1 cells in dLNs (Fig. 1E). In addition, OT-3 cells displayed reduced PD-1 expression in dLN (Fig. 1F). In contrast, most T cells in the tumor were CD62L- differentiated T cells, and there were no differences in CM (CD44high CD62L+) and effector T cell frequencies (Fig. 1E) or PD-1 expression levels between OT-1 and OT-3 cells (Fig. 1F). Thus, whereas differentiation status differs between high- and low-avidity T cells in the periphery 7 d postvaccination, no differences are found in the tumor, where mainly differentiated PD-1high T cells are present.

Based on CD127 (IL-7Rα) and KLRG1 expression, CD127low KLRG1high TEs and CD127high KLRG1low MPECs can be distinguished (22). TE frequencies were higher among i.t. OT-3 compared with OT-1 cells in contrast with the lower MPEC/TE ratio in low- compared with high-avidity T cells, at this time point, in the lymphoid organs and tumor (Fig. 1G).

T cell affinity during peripheral priming is critical for successful CD8+ T cell response and tumor control

To exclude cell-intrinsic differences between OT-1 and OT-3 T cells, we directly compared OT-1 responses to high- or low-affinity peptide vaccines. In addition, to discriminate the impact of T cell avidity on the CD8+ T cell response occurring in the periphery versus in the tumor, mice were engrafted with high- and low-affinity Ag-expressing tumors. To address the latter, we generated B16 melanoma cell lines expressing SIINFEKL (N4) wild-type OT-1 ligand or its 10 times lower–affinity SIITFEKL (T4) variant (9, 23). As previously published, the functional avidity of OT-1 cells in response to the T4 altered peptide ligand is similar to that of OT-3 T cells responding to the N4 peptide (20). To normalize Ag expression levels between N4- and T4-expressing cell lines, B16 cells were transduced with B16.N4-GFP or B16.T4-GFP retrovirus, and stable clones with similar GFP expression levels were selected (Supplemental Fig. 1A). We also confirmed that the functional avidity of OT-1 cells to B16.T4 was lower than to the B16.N4 cell line by analyzing IFN-γ secretion after in vitro restimulation of OT-1 cells with the cell lines and by an in vitro killing assay (Supplemental Fig. 1B, 1C).

To assess tumor control of low- and high-affinity ligand-expressing tumors after high- or low-affinity peripheral priming, mice were engrafted with B16.N4 and B16.T4 tumors on either flank and vaccinated with N4-CpG or T4-CpG 7 d after tumor engraftment and 1 d after 105 OT-1 T cell transfer (Fig. 2A). Because B16.N4 and B16.T4 cell lines showed different tumor growth kinetics (data not shown), values were expressed as relative tumor size to the volume at day 6, before OT-1 cell transfer.

FIGURE 2.
FIGURE 2.

Low-affinity peripheral priming leads to decreased OT-1 cell activation, effector differentiation, expansion, and tumor control. (A) Scheme of the experimental design. A total of 2 × 105 B16.N4 and B16.T4 cells were s.c. engrafted on each flank of C57BL/6J mice. After 6 d, 105 naive OT-1 cells were transferred and mice were s.c. vaccinated with the OVA wild type (SIINFEKL, single-letter amino acid code) antigenic peptide mixed with CpG-ODN (N4-CpG) or the OVA antigenic peptide with a Thr substitution at position P4 (SIITFEKL) mixed with CpG-ODN (T4-CpG) 1 d later. (B) Tumor growth curves represented as tumor size relative to tumor volume at day 6. Dots represent the mean ± SD (n = 9 in N4 vaccinated [vacc.] group, n = 10 in T4 vacc. group, and n = 5 in PBS group). (C) Number of OT-1 cells in spleen of N4- or T4 vacc. mice 14 and 21 d after tumor engraftment. (D) Number of OT-1 cells per cubic millimeter of tumor in OT-1 + N4 or OT-1 + T4 vacc. mice in N4 (red) or T4 (blue) tumors. (E) Representative histogram and PD-1 MFI in OT-1 cells at day 14 in spleen of mice vaccinated with N4 or T4 peptide. (F) Percentage of naive, CM, and effector cells in OT-1 cells in spleen at day 14. A two-way ANOVA followed by Tukey multiple comparison test was performed in (B), an unpaired t test or Mann–Whitney U test after Shapiro–Wilk normality test in (C) and (E), a two-way ANOVA followed by Sidak multiple comparison test in (D), and a two-way RM ANOVA test followed by Sidak multiple comparison test in (F). Dots represent individual mice, and the bar represents the mean. Representative of two independent experiments. *p < 0.05, **p < 0.01, ****p < 0.0001. MFI, mean fluorescence intensity.

Mice receiving T4 vaccination could control neither N4- nor T4-expressing B16 tumors (Fig. 2B). This was associated with reduced peripheral expansion and OT-1 cell numbers infiltrating the tumor (Fig. 2C, 2D). Again, reduced PD-1 expression and effector cell frequencies were observed in OT-1 cells after T4 vaccination compared with N4 vaccination in the spleen 7 d after vaccination, 14 d after tumor engraftment (Fig. 2E, 2F). To exclude the possibility that the observed phenotypic differences between high- compared with low-affinity stimulated OT-1 cells are due to differences in the response kinetics as described in a L. monocytogenes infection model (9), we followed the response of OT-1 cells after N4 or T4 peptide vaccination in peripheral blood for 21 d (Supplemental Fig. 1D). We observed an earlier peak of expansion upon low- compared with high-affinity vaccination, yet the expansion magnitude was around 10 times smaller (Supplemental Fig. 1E), as described in microbial infections (9). Nonetheless, although complete CD8+ T cell effector differentiation was observed by Zehn et al. upon Lm-T4 infection (9), we observed sustained reduced effector cell frequencies after low-affinity vaccination (Supplemental Fig. 1F) as well as absence of PD-1 upregulation (Supplemental Fig. 1G). Therefore, T cell avidity during peripheral priming with a vaccine plays an important role in CD8+ T cell activation, effector differentiation and expansion, and, consequently, in the magnitude of tumor control.

CD8+ T cell accumulation in the tumor is T cell affinity dependent

Reduced numbers of CD8+ T cells in the tumor may explain the lack of tumor control by OT-3 after N4 vaccination or OT-1 after T4 vaccination. Therefore, we compared OT-1 responses to B16.N4 or B16.T4 tumors after high-affinity N4 vaccination (Fig. 3A). High-affinity primed OT-1 cells could partially control B16.T4 tumors (Fig. 3B). Therefore, high-affinity T cell priming in the periphery still allows significant tumor control even in the case of low-affinity tumor Ag recognition. Nevertheless, Ag affinity recognition in the tumor still impacts on tumor control, as OT-1 cells controlled better N4- than T4-expressing tumors (Fig. 3B).

FIGURE 3.
FIGURE 3.

OT-1 cells control better high-affinity Ag-expressing tumors. (A) Scheme of the experimental design. A total of 2 × 105 B16.N4 and B16.T4 cells were s.c. engrafted on each side of C57BL/6J mice. After 6 d, 105 naive OT-1 cells were transferred, and s.c. vaccination with N4-CpG was performed 1 d later. (B) Tumor growth curves represented as tumor size relative to tumor volume at day 6. Dots represent the mean ± SD (n = 9 in OT-1 + N4 vaccination group and n = 5 in N4 vaccination group). (C) Number of OT-1 cells per cubic millimeter 14 and 21 d after tumor engraftment. (D) Number of total OT-1 cells in B16.N4 and B16.T4 tumor draining inguinal lymph nodes. (E) Representative histograms of intracellular BrdU staining of OT-1 cells from B16.N4 (red) and B16.T4 (blue) tumors. Endogenous CD8+ from a mouse not receiving BrdU is represented as control (gray). (F) Percentage of BrdU+ cells in OT-1 cells in B16.N4 and B16.T4 tumors. (G) BrdU MFI of BrdU+ OT-1 cells in tumors. (H) Percentage of MPECs and TEs in total CD44high OT-1 cells in B16.N4 and B16.T4 tumors. (I) Representative histograms of PD-1, CTLA-4, LAG-3, and CD39 expression in OT-1 cells from spleen (black, dashed), B16.N4 (red, continuous), and B16.T4 (blue, dotted) tumors on day 21. Below is the quantification of each marker’s MFI. Dots represent individual mice, and lines connect the two tumors of each mouse. A two-way ANOVA followed by Tukey multiple comparison test was performed in (B); a paired t test or Wilcoxon matched-pairs signed rank test after Shapiro–Wilk normality test in (C), (D), (F), (G), and (I); and a two-way RM ANOVA followed by Sidak multiple comparison test in (H). Representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. MFI, mean fluorescence intensity.

Increased OT-1 cell numbers were found in B16.N4 compared with B16.T4 tumors 14 and 21 d after tumor engraftment (Fig. 3C), but no differences in N4 compared with T4 B16 tumor dLNs were found (Fig. 3D). Moreover, higher BrdU+ OT-1 cell frequencies, with higher BrdU incorporation per cell, were found in B16.N4 tumors at day 21, suggesting increased OT-1 cell proliferation in B16.N4 tumors. However, at day 14, there was no difference in BrdU incorporation (Fig. 3E–G), although OT-1 numbers were higher in B16.N4 tumors (Fig. 3C), suggesting that T cell affinity may also affect T cell survival in the tumor, as previously shown in acute infections (8). To test that, we measured expression levels of CD127 and CD25 that were shown to be important for CD8+ T cell survival (8, 24). Indeed, we observed increased CD127high MPEC frequencies (Fig. 3H) as well as increased CD25 (IL-2Rα) and CD122 (IL-2Rβ) expression in OT-1 cells from B16.N4 tumors compared with B16.T4 tumors (Supplemental Fig. 2A, 2B). Thus, increased tumor control of B16.N4 tumors was probably associated with increased proliferation and responsiveness to homeostatic cytokines upon high-affinity Ag recognition.

Low-affinity Ag recognition leads to decreased and slower acquisition of activation markers, including inhibitory receptors

Tumor-infiltrating OT-1 cells expressed increased CD44 compared with those found in the spleen. Moreover, OT-1 cells in B16.T4 tumors had lower CD44 expression levels than in B16.N4 tumors both at 14 and 21 d after tumor engraftment (Supplemental Fig. 2C). Therefore, within the tumor, OT-1 cells upregulate CD44 in a T cell affinity–dependent manner.

CD45RB [a CD45 isoform highly expressed in naive cells but downregulated in effector and memory T cells (25)] expression patterns were opposite to those of CD44, also in a T cell affinity–dependent manner (Supplemental Fig. 2D).

Similar to the observations we made when comparing OT-1 and OT-3 cells, decreased MPEC frequencies were observed in OT-1 cells from low-affinity B16.T4 compared with high-affinity B16.N4 tumors (Fig. 3H). Thus, T cell affinity impacts MPEC CD8+ T cell formation both upon peripheral priming and in the tumor.

As previously described for tumor-infiltrating Ag-specific T cells in melanoma patients (26), PD-1 expression was higher in tumor-infiltrating OT-1 cells compared with those from the periphery. However, PD-1 expression was lower in B16.T4 compared with B16.N4 tumors at day 14 (data not shown) and 21 (Fig. 3I). Similarly, expression levels of CTLA-4, LAG-3, and CD39 T cell exhaustion markers were strongly decreased in B16.T4 tumors as compared with N4 tumors (Fig. 3I).

Altogether, T cell affinity affects CD8+ T cell differentiation in the tumor. Low-affinity Ag recognition leads to decreased acquisition of activation and differentiation markers, including inhibitory receptors. Thus, despite numeric differences, low-affinity T cells are interesting for therapeutic purposes, as they may be less susceptible to negative regulators despite showing decreased tumor control.

i.t. low-affinity Ag recognition leads to decreased granzyme B production but preserves cytokine production capacity

Decreased expression of inhibitory receptors by OT-1 cells in B16.T4 versus B16.N4 tumors suggests decreased susceptibility to T cell suppression. Therefore, we assessed at day 14 the ability of OT-1 cells from spleen and B16.N4 or B16.T4 tumors to produce cytokines and express granzyme B after in vitro restimulation with N4 peptide. As expected, IFN-γ+ and TNF-α+ OT-1 cell frequencies were lower in tumors than in the spleen (Fig. 4A, Fig. 4B, Fig. 4D). although the cytokine levels on a per-cell basis appeared similar in all conditions (Fig. 4C, Fig. 4E). Interestingly, when comparing B16.N4 and B16.T4 tumors, higher frequencies of IFN-γ–producing OT-1 T cells were found in T4 tumors (Fig. 4B). Similar results were obtained when measuring TNF-α (Fig. 4D). However, unlike cytokine production, both the frequencies and expression levels of granzyme B in OT-1 cells were instead decreased in T4 tumors as compared with N4 tumors (Fig. 4F, 4G), suggesting that cytotoxic activity might be less affected by immunosuppression via PD-1 signaling, which correlates with higher tumor control of N4 tumors. When restimulating on day 21 OT-1 cells from tumors and spleen, cytokine production was further decreased in tumor-infiltrating OT-1 cells as compared with spleen, yet higher frequencies of IFN-γ+ and TNF-α+ cells were still found in restimulated OT-1 cells from B16.T4 compared with B16.N4 tumors (Fig. 4B, 4D). Interestingly, frequencies of granzyme B+ OT-1 cells remained lower in OT-1 cells from B16.T4 as compared with B16.N4 tumors (Fig. 4F).

FIGURE 4.
FIGURE 4.

Increased cytokine-producing OT-1 cells but reduced granzyme B+ cells in B16.T4 compared with B16.N4 tumors. (A). Representative dot plots of OT-1 cells from spleen, B16.N4, and B16.T4 tumors in vitro restimulated with N4 peptide 14 and 21 d after tumor engraftment. (B, D, and F) Percentage of IFN-γ+, TNF-α+, and granzyme B+ cells in OT-1 cells. (C, E, and G) IFN-γ, TNF-α, and granzyme B MFI in IFN-γ+, TNF-α+, and granzyme B+ OT-1 cells, respectively. Dots represent individual mice, and bars represent the mean + SD (n = 5). A one-way ANOVA or Kruskal–Wallis test followed by Tukey multiple comparison or Dunn multiple comparison test was performed, respectively, after Shapiro–Wilk normality test in (B)–(G). Representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. MFI, mean fluorescence intensity.

Therefore, low-affinity Ag recognition in the tumor may lead to decreased cytotoxicity but preserved ability to produce cytokines.

CD8+ T cells from both high- and low-affinity ligand-expressing tumors show similar re-expansion capacity in recall tumor-free responses

To evaluate whether increased B16.N4 tumor control was due to increased OT-1 cell numbers that could compensate their exhausted profile, we assessed the fitness of B16.N4 or B16.T4 tumor-infiltrating OT-1 cells by analyzing their re-expansion capacity upon Ag reencounter in a tumor-free inflammatory environment. To this end, 5000 OT-1 cells sorted from spleen, B16.N4, or B16.T4 tumors were transferred to naive C57BL/6 mice, which were infected with 2000 CFU of Lm-N4 (Fig. 5A). As control, a group of mice received 5000 naive OT-1 cells and Lm-N4.

FIGURE 5.
FIGURE 5.

Tumor-infiltrating OT-1 cells from B16.N4 and B16.T4 tumors show similar re-expansion capacity in secondary tumor-free responses but decreased cytokine production over time. (A) Scheme of the experimental design. Fourteen or twenty-one days after tumor engraftment, 5000 OT-1 cells from spleen, B16.N4, or B16.T4 tumors were transferred into naive CD57BL/6 mice. The same day, secondary hosts were infected with 2000 CFU L. monocytogenes expressing the wild type OVA antigenic peptide (Lm-N4). As controls, a group of mice received 5000 naive OT-1 cells and Lm-N4 infection. (B) Fold expansion of day 14 or day 21 OT-1 cells in spleen of secondary hosts 8 d postinfection. Each dot represents individual mice, and the bar represents the mean ± SD. (C) Representative histograms of CD44, PD-1, and CTLA-4 expression of naive OT-1 cells (gray, filled) or OT-1 cells from spleen (black, dashed), B16.N4 (red, continuous), and B16.T4 (blue, dotted) tumors at day 14 or day 21 pretransfer (Pre) or 8 d postinfection in secondary hosts (Post). (D). Quantification of CD44, PD-1, and CTLA-4 MFI. (E) Percentage of IFN-γ+, TNF-α+, and granzyme B+ OT-1 cells from spleen of secondary hosts in vitro restimulated with N4 peptide. A two-way ANOVA followed by Sidak multiple comparison test was performed in (B) and (E) and a one-way ANOVA or Kruskal–Wallis test followed by Tukey multiple comparison or Dunn multiple comparison test after Shapiro–Wilk normality test in (D). Representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. MFI, mean fluorescence intensity.

Strikingly, OT-1 cells from day 14 or 21 B16.N4 and B16.T4 tumors re-expanded to a similar extent upon Lm-N4 infection (Fig. 5B). When comparing tumor and spleen, OT-1 cells from day 14 tumors expanded three times less than OT-1 cells from spleen of the same tumor-bearing mice (Fig. 5B). However, such difference may be due to the presence of CM OT-1 cells in the spleen at day 14, as 1 wk later, at day 21, when only effector cells remained in spleen and tumors (Supplemental Fig. 3A), OT-1 cells showed the same re-expansion capacity regardless of their tissue origin (Fig. 5B). Thus, effector OT-1 cells from the periphery or from high- and low-affinity tumors show the same re-expansion capacity in secondary tumor-free responses.

Moreover, 8 d postinfection, the differences in CD44, PD-1, and CTLA-4 expression levels between OT-1 cells from spleen, B16.N4, and B16.T4 tumors in primary hosts were lost upon re-expansion in the secondary host (Fig. 5C, 5D). Remarkably, PD-1 expression levels in secondary hosts receiving tumor-infiltrating OT-1 cells were much lower than within the tumors during the primary response (Fig. 5D). Yet, PD-1 expression was higher in the secondary response of Ag-experienced OT-1 compared with naive OT-1 cell controls. Similarly, there were no longer differences in MPECs and TE OT-1 cell frequencies in secondary hosts. TE frequencies were much higher in secondary hosts’ spleens compared with primary hosts’ day 14 and day 21 tumors (Supplemental Fig. 3B). Thus, OT-1 cell phenotype defined during the primary antitumor response is reset in a high-affinity secondary acute infection.

Whereas day 14 OT-1 cells from N4 and T4 tumors showed reduced cytokine and increased granzyme B production upon in vitro N4 peptide restimulation as compared with those from spleen (Fig. 4), similar frequencies of IFN-γ, TNF-α, and granzyme B–producing OT-1 cells were found upon Lm-N4 infection of secondary hosts from all groups (Fig. 5E). Nonetheless, like in the in vitro OT-1 restimulation from primary hosts (Fig. 4), the frequencies of IFN-γ, TNF-α, and granzyme B–producing OT-1 cells were lower in mice receiving day 21 compared with day 14 tumor-infiltrating OT-1 cells (Fig. 5E). Thus, prolonged exposure to the tumor microenvironment dampens CD8+ T cell capacity to produce IFN-γ, TNF-α, and granzyme B even upon rechallenge in a tumor-free environment.

Altogether, prolonged exposure to the tumor microenvironment, regardless of the affinity for the ligand, decreases the capacity of CD8+ T cells to produce IFN-γ, TNF-α, and granzyme B. Yet, the re-expansion capacity of tumor-infiltrating CD8+ T cells in secondary tumor-free responses is preserved over time and is similar to the expansion amplitude of effector CD8+ T cells from the periphery.

αPD-1 treatment enhances tumor control of both high- and low-affinity Ag-expressing tumors

To address the impact of anti–PD-1 blockade on T cells activated by high- or low-affinity tumor Ag, mice were engrafted with B16.N4 and B16.T4 tumors. On day 6 after tumor graft, 105 naive OT-1 or irrelevant P14 cells were transferred, and 1 d later, N4-CpG vaccine was administered. αPD-1 or ISO Abs were given three times every 3 d from day 10 after tumor engraftment (Fig. 6A). PD-1 blockade enhanced OT-1 cell–mediated tumor control of both B16.N4 and B16.T4 tumors by 2.4 and 1.6 times compared with untreated mice, respectively (Fig. 6B). These tumor responses were correlated with increased OT-1 numbers 21 d after tumor engraftment in the spleen (Fig. 6C), as well as to a lesser extent in tumors (Fig. 6D), of αPD-1 compared with ISO-treated mice. In addition, in situ proliferation of OT-1 cells was enhanced in B16.N4 tumors of αPD-1–treated mice (Fig. 6E, 6F) but not T4 tumors. However, when mice were grafted only with B16.T4 tumors (Supplemental Fig. 4A), αPD-1 treatment further enhanced tumor control (by 2.8) (Supplemental Fig. 4B), and increased OT-1 proliferation was also observed (Supplemental Fig. 4C, 4D), suggesting some preferential OT-1 cell recruitment to N4 tumors in B16.N4 and B16.T4 double tumor–bearing mice.

FIGURE 6.
FIGURE 6.

αPD-1 enhances B16.N4 and B16.T4 tumor control. (A) Scheme of the experimental design. A total of 2 × 105 B16.N4 and B16.T4 cells were s.c. engrafted on each side of C57BL/6J mice. After 6 d, 105 naive OT-1 or P14 cells were transferred, and s.c. vaccination with N4-CpG was performed 1 d later. On days 10, 13, and 16, αPD-1 or ISO Abs were administered i.p. Mice were sacrificed on day 21 for analysis (B). Tumor growth curves. Dots represent the mean tumor volume in cubic millimeters ± SD (n = 8 per group). (C) Number of OT-1 cells in spleen. (D) Number of OT-1 cells per cubic millimeter of tumor in B16.N4 and B16.T4 tumors. (E) Percentage of BrdU+ cells in OT-1 cells from tumors. (F) BrdU MFI of BrdU+ OT-1 cells in tumors (G). Representative histograms and PD-1 MFI of OT-1 cells from spleen and tumors. (H) Percentage of IFN-γ+, TNF-α+, and granzyme B+ cells in OT-1 cells from B16.N4 or B16.T4 tumors in vitro restimulated with N4 peptide. (I) CD8/Treg (defined as CD4+Foxp3+ cells) ratio in B16.N4 or B16.T4 tumors. Bars represent the mean ± SD. Dots represent individual mice, and the bar represents the mean. A two-way RM ANOVA followed by Tukey multiple comparison test was performed in (B) and an unpaired t test or Mann–Whitney U test after Shapiro–Wilk normality test in (C)–(I). Representative of four independent experiments. *p < 0.05, **p < 0.01, ****p < 0.0001. MFI, mean fluorescence intensity.

At the peak of the response, at day 14, higher PD-1 expression levels were observed in OT-1 cells from spleen of αPD-1–treated mice, indicating an enhanced activation of OT-1 cells during peripheral priming (data not shown). In the tumor, PD-1 levels were at least twenty times higher than in the spleen but were similar between αPD-1–treated and untreated mice (data not shown). However, 1 wk later, lower PD-1 expression levels were found in tumors of αPD-1–treated mice, whereas PD-1 expression levels were similar between groups in the spleen (Fig. 6G). This lower PD-1 expression in tumor-infiltrating OT-1 cells would suggest a decreased susceptibility to PD-L1/L2–mediated immunosuppression and enhanced functionality of T cells. Thus, we checked whether cytokine production capacity of OT-1 cells was increased in αPD-1–treated mice. Interestingly, selectively increased IFN-γ+ frequencies were observed in OT-1 cells from B16.N4 tumors, whereas no differences were observed in B16.T4 tumors (Fig. 6H). Nonetheless, in single B16.T4 tumor-bearing mice, increased IFN-γ+ frequencies were also observed in αPD-1–treated mice (Supplemental Fig. 4E). The overall increased CD8/Treg ratio observed in αPD-1–treated mice (Fig. 6I) also explains the increased OT-1 functionality and antitumor response.

Altogether, αPD-1 treatment enhanced tumor control of both high- and low-affinity Ag-expressing tumors by increasing peripheral expansion of OT-1 cells, resulting in higher OT-1 numbers in the tumor. Moreover, tumor-infiltrating OT-1 cells displayed lower PD-1 expression levels and enhanced cell proliferation and IFN-γ+ production.

High-affinity peptide-CpG boost notably enhances low-affinity antigen-expressing tumor control

In our experimental model, the vaccine was administered s.c. However, alternative routes of administration, such as i.t. injection of immunomodulators, have also shown the capacity to elicit specific T cell responses (27). In fact, it was recently reported that i.t. administration of CpG in combination with anti-OX40 blocking Abs could enhance B16 tumor control (28). Thus, we tested whether i.t. administration of N4-CpG in combination with αPD-1 treatment enhanced tumor control, particularly of B16 tumors expressing the low-affinity T4 peptide. The initial s.c. vaccination with N4-CpG was performed in all B16.N4 and B16.T4 tumor-bearing mice 7 d after tumor engraftment, and on day 14, mice received either a second s.c. vaccination or i.t. administration of N4-CpG. CpG alone was also administered i.t. in a group of mice, and αPD-1 treatment was performed as previously described.

We noticed that a second administration of N4-CpG, either s.c. or i.t. but not CpG alone, led to increased tumor control of mainly B16.T4 tumors (Fig. 7B). Enhanced tumor control upon a second vaccination was also observed among αPD-1–treated mice (Fig. 7B). This improved tumor control was associated with increased OT-1 numbers in tumors (Fig. 7C).

FIGURE 7.
FIGURE 7.

Second N4-CpG s.c. and i.t. vaccination notably enhances low-affinity Ag-expressing tumor control. (A) Scheme of the experimental design. A total of 2 × 105 B16.N4 and B16.T4 cells were s.c. engrafted on each side of C57BL/6J mice. After 6 d, 105 naive OT-1 cells were transferred and s.c. vaccination with N4-CpG was performed 1 d later. On days 10, 13, and 16, αPD-1 or ISO Abs were administered i.p. On day 14, CpG alone or in combination with N4 was given s.c. or i.t. (B) Tumor size on day 21 relative to tumor volume at day 14 before the second vaccination. (C) Percentage of OT-1 cells in total live cells in B16.N4 and B16.T4 tumors. Dots represent individual mice, and the bar represents the mean ± SD (n = 7–13 per group). A one-way ANOVA followed by Tukey multiple comparison test was performed. Representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Thus, a second delivery of the high-affinity N4 peptide with CpG, either s.c. or i.t., can further enhance control of αPD-1–treated low-affinity B16.T4 tumors by increasing infiltration of specific OT-1 cells.

Discussion

The weak immunogenicity of cancer cells is one of the hurdles in achieving clinically effective antitumor T cell responses. Indeed, most TAAs are self-proteins, and therefore, tumor-reactive T cells found in cancer patients generally recognize their Ag with low avidity (13). The advent of affordable whole-exome sequencing has made possible the routine identification of mutant protein Ags (neoantigen) for which specific high-avidity CD8+ T cells may be found in blood and tumors of cancer patients (29, 30). Thus, high- and low-avidity TAA-specific T cells may be found in cancer patients, but the contribution of each type of CD8+ T cell to the overall antitumor response is only partially understood. Moreover, the response of each type of CD8+ T cell, in the same host, to cancer vaccines and immune checkpoint blockade therapies has not been studied.

Comparison of high- and low-affinity T cell responses by varying either the TCR or its ligand showed in this study that the affinity for the ligand in a peptide vaccine greatly impacts the magnitude of CD8+ T cells’ peripheral expansion, as previously shown in bacterial infections (9). Curtailed expansion of low-affinity primed CD8+ T cells was accompanied by reduced early PD-1 expression and effector CD8+ T cell frequencies in dLNs. Moreover, low-affinity priming by systemic vaccination hampered effective tumor control of either high- or low-affinity ligand-expressing tumors. On the contrary, high-affinity priming led to a significant retardation of tumor growth regardless of the Ag recognition strength in the tumor microenvironment. Interestingly, a high-affinity systemic or i.t. boost particularly enhanced control of low-affinity ligand-expressing tumors. These observations support the use of high-affinity altered-peptide ligands to optimize therapeutic cancer vaccines (14–16).

Although high-affinity peripheral priming allowed tumor control of low-affinity ligand-expressing tumors, our results clearly show that the magnitude of tumor control is dependent on the avidity of TAA recognition by T cells recruited to the tumor. This was probably due to decreased in situ OT-1 cell expansion and granzyme B expression in low-affinity tumors. Yet, OT-1 cells infiltrating the low-affinity Ag-expressing tumors exhibited lower inhibitory receptor expression levels and higher cytokine-producing cell frequencies than those thriving in the high-affinity Ag-expressing tumors. These results suggested increased functionality of low-affinity stimulated tumor-infiltrating CD8+ T cells. Notwithstanding, tumor-infiltrating OT-1 cells showed the same re-expansion capacity in secondary tumor-free acute responses regardless of the affinity for the Ag ligand expressed by the tumor. This outcome is similar to recent observations made in the model of specific CD8+ T cell exhaustion in the context of LCMV cl13 chronic infection. Indeed, although low-affinity stimulated CD8+ T cells showed lower PD-1 expression levels in a primary chronic virus infection, high- and low-affinity stimulated CD8+ T cells maintained the same re-expansion capacity in a secondary acute viral infection (10). Thus, chronic high- and low-affinity stimulation, whether in an infection or in the tumor, do not impair re-expansion capacity of CD8+ T cells in secondary acute responses.

Interestingly, the CD8+ T cell exhausted phenotype acquired in the tumor microenvironment appears reversible upon secondary acute infection. This is in contrast to previous observations in the LCMV cl13 chronic infection model, in which high PD-1 levels in chronically stimulated CD8+ T cells were retained in a secondary acute infection (31), probably because of permanent demethylation of the PD-1 locus (32). Therefore, it would be interesting to analyze whether epigenetic regulation of the PD-1 locus may explain the differences observed between the two models. In this regard, a recent report showed that T cells in mouse spontaneous liver tumors may acquire the exhausted functional state by transitioning through two successive discrete chromatin states: one plastic dysfunctional state that is reversible and a fixed one in which cells become resistant to reprogramming (33). Whether this sequence of events occurs in chronically activated tumor-infiltrating lymphocytes on a regular basis in all types of tumors remains to be elucidated.

In the transplantable tumor model analyzed here, prolonged exposure to the tumor microenvironment led to decreased cytokine and granzyme B–producing OT-1 cell frequencies in both high- and low-affinity ligand-expressing tumors. Unfortunately, whereas early (day 14) tumor-infiltrating OT-1 cell transfer to a tumor-free environment restored cytokine and granzyme B production capacity, late (day 21) OT-1 cell transfer diminished OT-1 cell functionality in the secondary response. In the chronic LCMV cl13 infection, similar observations were made. Despite similar re-expansion, chronically stimulated CD8+ T cells maintained reduced cytokine production in secondary acute viral responses compared with CD8+ T cells from acute infections (31). Altogether, chronic Ag stimulation, regardless of the affinity to the ligand, leads to decreased cytokine production but preserved re-expansion upon recalling with Ag conveyed by acute infectious agents.

Signaling through the inhibitory receptor PD-1 is one of the immunosuppressive signals that CD8+ T cells encounter in the tumor. Preclinical and clinical studies have shown that αPD-1 blockade greatly ameliorates the outcome of cancer patients, especially in the case of immunogenic tumors (34–38). Our data showed that αPD-1 treatment improves outcomes in both high- and low-affinity Ag-expressing melanoma tumors by enhancing CD8+ T cell numbers and functionality in the tumor. Hence, low-affinity TAA-specific CD8+ T cells may be involved in the antitumor activity observed in αPD-1–treated metastatic melanoma patients (38).

Altogether, our results suggest that early vaccination and boost with high-affinity altered peptide ligands or neoantigens in combination with αPD-1 may be a highly clinically effective strategy to boost endogenous low-affinity TAA-specific CD8+ T cell responses, as well as neoantigen-specific CD8+ T cell responses, in cancer patients. In support of this notion, it is noteworthy that a recent study showed that melanoma patients with tumor recurrences after vaccination with neoantigens had complete tumor responses accompanied by broadening of the neoantigen-specific T cell repertoires following treatment with αPD-1 (39).

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Nina Dumauthioz, Alejandra Gomez-Cadena, Sara Labiano, and Benjamin Tschumi for technical support and insightful discussions. We thank Leyder Lozano, Candice Stoudmann, and Silvia Ferreira for technical help as well as Romain Bedel for technical assistance in the FACS. We also thank the caretakers of the animal facility for excellent assistance.

Footnotes

  • This work was supported in part by grants from the Swiss National Science Foundation to P.R. (Sinergia CRSII3_141879, Sinergia CRSII3_160708, and 31003A_156469) and from the ERC (PROTECTC) to D.Z.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    CM
    central memory
    dLN
    draining lymph node
    ISO
    isotype control
    i.t.
    intratumoral(ly)
    LCMV
    lymphocytic choriomeningitis virus
    Lm-N4
    N4-expressing L. monocytogenes
    MPEC
    memory precursor effector cell
    pMHC
    peptide–MHC
    RM
    repeated measurement
    SSC
    side scatter
    TAA
    tumor-associated Ag
    TE
    terminal effector cell
    UNIL
    University of Lausanne.

  • Received November 30, 2017.
  • Accepted May 9, 2018.

References

    1. Zhang, N.,
    2. M. J. Bevan
    . 2011. CD8(+) T cells: foot soldiers of the immune system. Immunity 35: 161–168.
    1. Chang, J. T.,
    2. V. R. Palanivel,
    3. I. Kinjyo,
    4. F. Schambach,
    5. A. M. Intlekofer,
    6. A. Banerjee,
    7. S. A. Longworth,
    8. K. E. Vinup,
    9. P. Mrass,
    10. J. Oliaro, et al
    . 2007. Asymmetric T lymphocyte division in the initiation of adaptive immune responses. Science 315: 1687–1691.
    1. Ahmed, R.,
    2. D. Gray
    . 1996. Immunological memory and protective immunity: understanding their relation. Science 272: 54–60.
    1. Lanzavecchia, A.,
    2. F. Sallusto
    . 2002. Progressive differentiation and selection of the fittest in the immune response. Nat. Rev. Immunol. 2: 982–987.
    1. Crawford, F.,
    2. H. Kozono,
    3. J. White,
    4. P. Marrack,
    5. J. Kappler
    . 1998. Detection of antigen-specific T cells with multivalent soluble class II MHC covalent peptide complexes. Immunity 8: 675–682.
    1. Moreau, H. D.,
    2. F. Lemaître,
    3. E. Terriac,
    4. G. Azar,
    5. M. Piel,
    6. A. M. Lennon-Dumenil,
    7. P. Bousso
    . 2012. Dynamic in situ cytometry uncovers T cell receptor signaling during immunological synapses and kinapses in vivo. Immunity 37: 351–363.
    1. Ozga, A. J.,
    2. F. Moalli,
    3. J. Abe,
    4. J. Swoger,
    5. J. Sharpe,
    6. D. Zehn,
    7. M. Kreutzfeldt,
    8. D. Merkler,
    9. J. Ripoll,
    10. J. V. Stein
    . 2016. pMHC affinity controls duration of CD8+ T cell-DC interactions and imprints timing of effector differentiation versus expansion. J. Exp. Med. 213: 2811–2829.
    1. Wensveen, F. M.,
    2. K. P. van Gisbergen,
    3. I. A. Derks,
    4. C. Gerlach,
    5. T. N. Schumacher,
    6. R. A. van Lier,
    7. E. Eldering
    . 2010. Apoptosis threshold set by Noxa and Mcl-1 after T cell activation regulates competitive selection of high-affinity clones. Immunity 32: 754–765.
    1. Zehn, D.,
    2. S. Y. Lee,
    3. M. J. Bevan
    . 2009. Complete but curtailed T-cell response to very low-affinity antigen. Nature 458: 211–214.
    1. Utzschneider, D. T.,
    2. F. Alfei,
    3. P. Roelli,
    4. D. Barras,
    5. V. Chennupati,
    6. S. Darbre,
    7. M. Delorenzi,
    8. D. D. Pinschewer,
    9. D. Zehn
    . 2016. High antigen levels induce an exhausted phenotype in a chronic infection without impairing T cell expansion and survival. J. Exp. Med. 213: 1819–1834.
    1. Schmid, D. A.,
    2. M. B. Irving,
    3. V. Posevitz,
    4. M. Hebeisen,
    5. A. Posevitz-Fejfar,
    6. J. C. Sarria,
    7. R. Gomez-Eerland,
    8. M. Thome,
    9. T. N. Schumacher,
    10. P. Romero, et al
    . 2010. Evidence for a TCR affinity threshold delimiting maximal CD8 T cell function. J. Immunol. 184: 4936–4946.
    1. Dutoit, V.,
    2. V. Rubio-Godoy,
    3. P. Y. Dietrich,
    4. A. L. Quiqueres,
    5. V. Schnuriger,
    6. D. Rimoldi,
    7. D. Liénard,
    8. D. Speiser,
    9. P. Guillaume,
    10. P. Batard, et al
    . 2001. Heterogeneous T-cell response to MAGE-A10(254-262): high avidity-specific cytolytic T lymphocytes show superior antitumor activity. Cancer Res. 61: 5850–5856.
    1. McMahan, R. H.,
    2. J. E. Slansky
    . 2007. Mobilizing the low-avidity T cell repertoire to kill tumors. Semin. Cancer Biol. 17: 317–329.
    1. Murahashi, M.,
    2. Y. Hijikata,
    3. K. Yamada,
    4. Y. Tanaka,
    5. J. Kishimoto,
    6. H. Inoue,
    7. T. Marumoto,
    8. A. Takahashi,
    9. T. Okazaki,
    10. K. Takeda, et al
    . 2016. Phase I clinical trial of a five-peptide cancer vaccine combined with cyclophosphamide in advanced solid tumors. Clin. Immunol. 166–167: 48–58.
    1. Jandus, C.,
    2. D. Speiser,
    3. P. Romero
    . 2009. Recent advances and hurdles in melanoma immunotherapy. Pigment Cell Melanoma Res. 22: 711–723.
    1. Reed, C. M.,
    2. N. D. Cresce,
    3. I. S. Mauldin,
    4. C. L. Slingluff Jr..,
    5. W. C. Olson
    . 2015. Vaccination with melanoma helper peptides induces antibody responses associated with improved overall survival. Clin. Cancer Res. 21: 3879–3887.
    1. Legat, A.,
    2. H. Maby-El Hajjami,
    3. P. Baumgaertner,
    4. L. Cagnon,
    5. S. Abed Maillard,
    6. C. Geldhof,
    7. E. M. Iancu,
    8. L. Lebon,
    9. P. Guillaume,
    10. D. Dojcinovic, et al
    . 2016. Vaccination with LAG-3Ig (IMP321) and peptides induces specific CD4 and CD8 T-cell responses in metastatic melanoma patients--report of a phase I/IIa clinical trial. Clin. Cancer Res. 22: 1330–1340.
    1. Brinckerhoff, L. H.,
    2. L. W. Thompson,
    3. C. L. Slingluff Jr.
    . 2000. Melanoma vaccines. Curr. Opin. Oncol. 12: 163–173.
    1. Liénard, D.,
    2. D. Rimoldi,
    3. M. Marchand,
    4. P. Y. Dietrich,
    5. N. van Baren,
    6. C. Geldhof,
    7. P. Batard,
    8. P. Guillaume,
    9. M. Ayyoub,
    10. M. J. Pittet, et al
    . 2004. Ex vivo detectable activation of Melan-A-specific T cells correlating with inflammatory skin reactions in melanoma patients vaccinated with peptides in IFA. Cancer Immun. 4: 4.
    1. Enouz, S.,
    2. L. Carrié,
    3. D. Merkler,
    4. M. J. Bevan,
    5. D. Zehn
    . 2012. Autoreactive T cells bypass negative selection and respond to self-antigen stimulation during infection. J. Exp. Med. 209: 1769–1779.
    1. Perret, R.,
    2. S. R. Sierro,
    3. N. K. Botelho,
    4. S. Corgnac,
    5. A. Donda,
    6. P. Romero
    . 2013. Adjuvants that improve the ratio of antigen-specific effector to regulatory T cells enhance tumor immunity. Cancer Res. 73: 6597–6608.
    1. Joshi, N. S.,
    2. W. Cui,
    3. A. Chandele,
    4. H. K. Lee,
    5. D. R. Urso,
    6. J. Hagman,
    7. L. Gapin,
    8. S. M. Kaech
    . 2007. Inflammation directs memory precursor and short-lived effector CD8(+) T cell fates via the graded expression of T-bet transcription factor. Immunity 27: 281–295.
    1. Krummey, S. M.,
    2. R. J. Martinez,
    3. R. Andargachew,
    4. D. Liu,
    5. M. Wagener,
    6. J. E. Kohlmeier,
    7. B. D. Evavold,
    8. C. P. Larsen,
    9. M. L. Ford
    . 2016. Low-affinity memory CD8+ T cells mediate robust heterologous immunity. J. Immunol. 196: 2838–2846.
    1. Schluns, K. S.,
    2. W. C. Kieper,
    3. S. C. Jameson,
    4. L. Lefrançois
    . 2000. Interleukin-7 mediates the homeostasis of naïve and memory CD8 T cells in vivo. Nat. Immunol. 1: 426–432.
    1. Hermiston, M. L.,
    2. Z. Xu,
    3. A. Weiss
    . 2003. CD45: a critical regulator of signaling thresholds in immune cells. Annu. Rev. Immunol. 21: 107–137.
    1. Ahmadzadeh, M.,
    2. L. A. Johnson,
    3. B. Heemskerk,
    4. J. R. Wunderlich,
    5. M. E. Dudley,
    6. D. E. White,
    7. S. A. Rosenberg
    . 2009. Tumor antigen-specific CD8 T cells infiltrating the tumor express high levels of PD-1 and are functionally impaired. Blood 114: 1537–1544.
    1. Hammerich, L.,
    2. A. Binder,
    3. J. D. Brody
    . 2015. In situ vaccination: cancer immunotherapy both personalized and off-the-shelf. Mol. Oncol. 9: 1966–1981.
    1. Sagiv-Barfi, I.,
    2. D. K. Czerwinski,
    3. S. Levy,
    4. I. S. Alam,
    5. A. T. Mayer,
    6. S. S. Gambhir,
    7. R. Levy
    . 2018. Eradication of spontaneous malignancy by local immunotherapy. Sci. Transl. Med. 10: eaan4488.
    1. Gros, A.,
    2. M. R. Parkhurst,
    3. E. Tran,
    4. A. Pasetto,
    5. P. F. Robbins,
    6. S. Ilyas,
    7. T. D. Prickett,
    8. J. J. Gartner,
    9. J. S. Crystal,
    10. I. M. Roberts, et al
    . 2016. Prospective identification of neoantigen-specific lymphocytes in the peripheral blood of melanoma patients. Nat. Med. 22: 433–438.
    1. Yarchoan, M.,
    2. B. A. Johnson III.,
    3. E. R. Lutz,
    4. D. A. Laheru,
    5. E. M. Jaffee
    . 2017. Targeting neoantigens to augment antitumour immunity. Nat. Rev. Cancer 17: 209–222.
    1. Utzschneider, D. T.,
    2. A. Legat,
    3. S. A. Fuertes Marraco,
    4. L. Carrié,
    5. I. Luescher,
    6. D. E. Speiser,
    7. D. Zehn
    . 2013. T cells maintain an exhausted phenotype after antigen withdrawal and population reexpansion. Nat. Immunol. 14: 603–610.
    1. Youngblood, B.,
    2. K. J. Oestreich,
    3. S. J. Ha,
    4. J. Duraiswamy,
    5. R. S. Akondy,
    6. E. E. West,
    7. Z. Wei,
    8. P. Lu,
    9. J. W. Austin,
    10. J. L. Riley, et al
    . 2011. Chronic virus infection enforces demethylation of the locus that encodes PD-1 in antigen-specific CD8(+) T cells. Immunity 35: 400–412.
    1. Philip, M.,
    2. L. Fairchild,
    3. L. Sun,
    4. E. L. Horste,
    5. S. Camara,
    6. M. Shakiba,
    7. A. C. Scott,
    8. A. Viale,
    9. P. Lauer,
    10. T. Merghoub, et al
    . 2017. Chromatin states define tumour-specific T cell dysfunction and reprogramming. Nature 545: 452–456.
    1. Leach, D. R.,
    2. M. F. Krummel,
    3. J. P. Allison
    . 1996. Enhancement of antitumor immunity by CTLA-4 blockade. Science 271: 1734–1736.
    1. van Elsas, A.,
    2. A. A. Hurwitz,
    3. J. P. Allison
    . 1999. Combination immunotherapy of B16 melanoma using anti-cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) and granulocyte/macrophage colony-stimulating factor (GM-CSF)-producing vaccines induces rejection of subcutaneous and metastatic tumors accompanied by autoimmune depigmentation. J. Exp. Med. 190: 355–366.
    1. Iwai, Y.,
    2. S. Terawaki,
    3. T. Honjo
    . 2005. PD-1 blockade inhibits hematogenous spread of poorly immunogenic tumor cells by enhanced recruitment of effector T cells. Int. Immunol. 17: 133–144.
    1. Hirano, F.,
    2. K. Kaneko,
    3. H. Tamura,
    4. H. Dong,
    5. S. Wang,
    6. M. Ichikawa,
    7. C. Rietz,
    8. D. B. Flies,
    9. J. S. Lau,
    10. G. Zhu, et al
    . 2005. Blockade of B7-H1 and PD-1 by monoclonal antibodies potentiates cancer therapeutic immunity. Cancer Res. 65: 1089–1096.
    1. Topalian, S. L.,
    2. F. S. Hodi,
    3. J. R. Brahmer,
    4. S. N. Gettinger,
    5. D. C. Smith,
    6. D. F. McDermott,
    7. J. D. Powderly,
    8. R. D. Carvajal,
    9. J. A. Sosman,
    10. M. B. Atkins, et al
    . 2012. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N. Engl. J. Med. 366: 2443–2454.
    1. Ott, P. A.,
    2. Z. Hu,
    3. D. B. Keskin,
    4. S. A. Shukla,
    5. J. Sun,
    6. D. J. Bozym,
    7. W. Zhang,
    8. A. Luoma,
    9. A. Giobbie-Hurder,
    10. L. Peter, et al
    . 2017. An immunogenic personal neoantigen vaccine for patients with melanoma. Nature 547: 217–221.