Toll like receptor agonist effects on human CD8+ T cell activation and expression of T cell checkpoint receptors Authors: Donghwan Jeon, Jena E Moseman, Ethan Hill, Douglas G McNeel Published: 2025 ABSTRACT ABSTRACTWe have previously found that murine CD8+ T cells activated with cognate antigen increase the expression of PD-1 and other checkpoint receptors, and this can be attenuated in the presence of specific Toll-like receptor (TLR) agonists. In the current report, we sought to investigate whether TLR agonists similarly affect the expression of T cell checkpoint receptors on human CD8+ T cells and their function. Cryopreserved human peripheral blood mononuclear cells (PBMCs) containing CD8+ T cells reactive to human cytomegalovirus (CMV) or Epstein-Barr virus (EBV) were stimulated with virus-specific peptide epitopes in the presence of TLR agonists. Activation of human CD8+ T cells in the presence of TLR agonists did not further expand tetramer-positive CD8+ T cells or significantly alter the expression of checkpoint receptors (PD-1, CTLA-4, LAG-3, TIGIT, TIM-3, or CD160). Activation of CD8+ T cells by viral antigen-loaded dendritic cells (DC) in the presence of a TLR3 agonist increased the number of antigen-specific IFNγ-secreting cells and the magnitude of IFNγ expression. DCs were found to secrete IFNβ and IL-12 in response to stimulation with TLR3 agonist and activation of CD8+ T cells in the presence of IFNß and IL-12 led to a slight decrease in expression of CTLA-4 and TIGIT. These findings demonstrate that, among TLR agonists, TLR3 agonists can enhance the activation of human CD8+ T cells and might be preferred TLR agonists for use as adjuvants in human anti-cancer vaccines or other immunotherapies aimed at activating antigen-specific CD8+ T cells.KEYWORDS: TLR agonist, CD8+ T cell, PD-1, T cell checkpoint receptor Introduction Toll like receptors (TLRs) are the largest family of pattern recognition receptors (PRRs) that provide danger signals from pathogens or cellular damage. Diverse TLRs have been identified in both humans (TLR1-10) and mice (TLR1-9 and 11, 12, 13). Each TLR recognizes a different type of conserved motifs, called pathogen-associated molecular patterns (PAMPs) in foreign pathogens, or damage-associated molecular patterns (DAMPs) in internal molecules from damaged cells.1 TLRs are broadly expressed in both immune cells and nonimmune cells, including DCs, T cells, NK cells, epithelial cells, and fibroblasts. Stimulation of TLRs primarily leads to the activation of innate immune responses, and initiates signaling pathways that enhance the ability of cells to phagocytose and eliminate pathogens.2,3 However, each TLR exhibits distinct expression patterns and cytokine production in immune and nonimmune cells, depending on its role in recognizing various molecular patterns.4 Conceptually, each TLR can trigger different innate immune responses, and certain TLRs may be activated to enhance the desired immune response.5 Therefore, TLR agonists have been used as vaccine adjuvants, targeting infectious diseases as well as various types of cancer.6–9 In our previous studies, we found that TLR agonists improved the anti-tumor efficacy of murine CD8+ T cells when TLR agonists were used as vaccine adjuvants for peptide or DNA anti-tumor vaccines.10 This was mediated by secretion of IL-12 from professional antigen presenting cells, resulting in the downregulation of expression of T cell checkpoint receptor PD-1 on CD8+ T cells. Additionally, we found that the activation of multiple TLR at the time of T cell activation by vaccine led to further enhanced CD8+ T cell-mediated anti-tumor responses in murine tumor models.11 Specifically, certain combination of TLR agonists (TLR1/2 + TLR3, TLR1/2 + TLR9, and TLR3 + TLR9) downregulated additional T cell checkpoint receptors, including PD-1, LAG-3, and CD160 on murine CD8+ T cells. IL-12 and IFNβ were identified as key mediators of these effects, and expression of IL-12 and IFNβ by professional antigen presenting cells increased following exposure to these TLR agonists. These findings in murine models suggested that certain TLR agonists or their combinations possess a great potential to augment the T cell mediated anti-tumor response triggered by cancer vaccines or in combination with T cell checkpoint blockade.12 However, the effect of TLR agonists during activation of human CD8+ T cells and their expression of T cell checkpoint receptors have not been well studied to date. In general, the basic biological function and downstream signaling of TLRs are thought to be conserved between human and mouse.13 Each TLR family member from humans and mice recognizes the same, conserved classes of PAMPs from microorganisms. The Toll/IL-1 receptor (TIR) motif in the cytoplasmic domain of TLR is also highly conserved between humans and mice.14 However, many studies have demonstrated that members of the TLR family show divergence in their baseline expression level, ligand affinity, or downstream gene expression among species. For example, murine TLR2 shows low or undetectable expression unless the expression is triggered by pro-inflammatory cytokines, while human TLR2 shows high, constitutive expression in peripheral blood leukocytes.15,16 The ligand-binding domain of TLR4 has been shown to be highly polymorphic between mice and humans,17 and stimulation of TLR4 showed diverse gene regulation among different species.18 The human immune system is also likely more sensitive to LPS-mediated toxicity compared to mice.19 These findings suggest that studies with human immune cells are needed to determine if TLR agonists can similarly modulate the function of human CD8+ T cells as previously identified in murine models. In this study, we hypothesized that TLR agonists, or specific combinations of them, when used at the time of human T cell activation, may suppress the expression of T cell checkpoint receptors, and affect the function of human CD8+ T cells. Our goal was to identify TLR agonists that might be preferred for use as adjuvants for human anti-tumor vaccines. Using human PBMCs, we assessed the expression of T cell checkpoint receptors on human CD8+ T cells specific for human cytomegalovirus (CMV) or Epstein-Barr virus (EBV), following stimulation with the antigen alone or in combination with TLR agonists. TLR agonists that downregulated the expression of T cell checkpoint receptors during their activation, or increased the expression of IFNγ with antigen-specific activation, were prioritized for further studies. Our findings suggest that TLR3 agonists, in particular, affect CD8+ T cells during their activation and might be preferred TLR agonists to use as adjuvants in vaccines or other immunotherapies aimed at activating CD8+ T cells. Materials and methods Mice OT-I (C57BL/6-Tg(TcraTcrb)1100Mjb/J, Stock No: 003831) and C57BL/6J (Stock No: 000664) mice were purchased from The Jackson Laboratory (Jax, Bar Harbor, MA). Mice were maintained in microisolator cages under aseptic conditions. All experiments were conducted under an approved protocol (M005690) by the University of Wisconsin School of Medicine and Public Health Institutional Animal Care and Use Committee that conforms to the NIH guide for the care and use of laboratory animals. Peptides Peptides for HLA-A2-restricted epitopes in cytomegalovirus (CMV) pp65 (NLVPMVATV), Epstein-Barr virus (EBV) latent membrane protein 2 (LMP2) (FLYALALLL), a pool of 32 MHC class I-restricted peptides derived from EBV, human CMV, and Influenza A (CEF), and H-2b-restricted epitope from chicken ovalbumin (SIINFEKL) were synthesized and the purity and identity were confirmed by mass spectrometry and gas chromatography (LifeTein, LLC., Hillsborough, NJ). Peptides were reconstituted in DMSO (2 mg/mL) and stored at −80°C until use. TLR agonists TLR agonists were purchased from InvivoGen (San Diego, CA) and used at the following concentrations: 300 ng/mL Pam3CSK4, 2 × 108 HKLM/mL HKLM, 10 µg/mL Poly(I:C) HMW, 10 µg/mL Poly(I:C) LMW, 10 µg/mL LPS-EK, 100 ng/mL FLA-ST, 100 ng/mL FSL-1, 5 µg/mL Imiquimod, 5 µg/mL ssRNA/LyoVec, 5 µmol/L ODN 2006, 3 µg/mL Guardiquimod, 5 µmol/L ODN 1826. For comparisons among TLR3 agonists, final molar concentration of agonists were matched to 5.05 nM based on the estimated molecular weight (g/mol) of each TLR3 agonist [Poly(I:C) HMW: 1.98 × 106, Poly(I:C) LMW: 3.3 × 105, Poly(A:U): 7 × 105, NexaVant: 2.75 × 105] In vitro stimulation of OT-I CD8+ T cells that were pre-exposed to ovalbumin Splenocytes were prepared from the spleens of OT-I mice, and 2 × 106 naïve OT-I splenocytes were adoptively transferred into 6-week-old female C57BL/6 mice intraperitoneally (n = 3/group). On the following day, mice were immunized subcutaneously with 100 µg SIINFEKL peptide weekly for two weeks, with or without TLR agonists at the following concentration: TLR3 (Poly(I:C) HMW, 100 μg/mouse), TLR9 (ODN 1826, 50 μg/mouse). 14 d after the final vaccination, splenocytes were isolated from each mouse as previously described.11,12 Splenocytes were then cultured in RPMI 1640 medium, supplemented with 10% fetal calf serum (FCS), and 2 µg/mL SIINFEKL peptide. TLR agonists were added 1 hour before stimulating with peptide as described above. Cells were analyzed after 48 hours via flow cytometry with the following antibodies: SIINFEKL H2Kb tetramer-BV421 (NIH Tetramer Core Facility), CD3-PE-Cy5 (Biolegend, San Diego, CA 100310), CD4-BV711 (Biolegend 100550), CD8-Alexa700 (Biolegend 100730), PD-1-PE-CF594 (BD 562523), 4-1BB-BV605 (BD 106110), and Live/Dead Ghost dye 780 (Tonbo, San Diego, CA 13–0865–T100). Preparation of human peripheral blood mononuclear cells (PBMC) Peripheral blood mononuclear cells (PBMC) were collected from patients with prostate cancer previously treated in clinical trials between 2004 and 2016, conducted in accordance with the Declaration of Helsinki, and for which all patients gave written informed IRB-approved consent to permit use of remaining blood specimens for immunological research. The HLA-A2+ haplotype of patients was determined by flow cytometry with a PE-labeled anti-HLA-A2 antibody (Biolegend 343306). The presence of CMV or EBV antigen-specific CD8+ T cells in PBMC were determined by flow cytometry with the following tetramers and antibodies: NLVPMVATV HLA-A2 tetramer-Alexa Fluor 647 (NIH Tetramer Core Facility), FLYALALLL HLA-A2 tetramer-Alexa Fluor 647 (NIH Tetramer Core Facility), CD3-PE-Cy7 (BD 557851), CD4-PE (BD 555347), CD8-BV786 (BD 563824), Live/Dead Ghost dye 780 (Tonbo, 13–0865-T100). Preparation of human immature dendritic cells (iDCs) PBMCs were re-suspended in warm (37°C) Aim V medium (Gibco, Billings, MT 12055091) at a cellular concentration of 3x106 cells/mL. Cells were plated in flasks to allow cells to adhere for 2 hours at 37°C in 5% CO2. Cells were gently washed with PBS and incubated for 6 d with X-VIVO-15 medium (Lonza, Walkersville, MD 04-418Q), supplemented with 1% penicillin/streptomycin, 20 ng/mL rhGM-CSF (Biolegend, 572902) and 10 ng/mL rhIL-4 (Biolegend, 574002), as previously described.20 Adherent DCs, characterized phenotypically and by CD11c expression, were flushed with warm PBS, resuspended in RPMI-1640 with 10% Human AB serum (Valley Biomedical, HP1022-HI). DCs were used directly or cryopreserved in cold freezing medium (90% FBS/10% DMSO) for later use. In vitro stimulation of HLA-A2 specific CD8+ T cells in PBMCs Cryopreserved PBMC were thawed, washed, re-suspended in RPMI-1640 with 10% human AB serum (Valley Biomedical, HP1022-HI), and rested at 37°C for one hour before further use. PBMCs were then stimulated with IL-2 (10 IU/mL) and TLR agonists as above for one hour. In other experiments, PBMCs were stimulated with human IL-12 (R&D Systems 10,018-IL-010) or IFNβ (R&D Systems 8499-IF-010/CF) for one hour. HLA-A2 specific viral antigen peptides (CMV: NLVPMVATV, EBV: FLYALALLL) were then added at a concentration of 2 µg/mL, a saturating concentration for stimulating antigen-specific CD8+ T cell when using well-defined epitopes. Cells were analyzed after 72 hours by flow cytometry with the following tetramers and antibodies: NLVPMVATV HLA-A2 tetramer-Alexa Fluor 647(NIH Tetramer Core Facility), FLYALALLL HLA-A2 tetramer-Alexa Fluor 647(NIH Tetramer Core Facility), CD3-PE-Cy5 (BD 555334), CD4-BV711 (BD 563028), CD8-Alexa Fluor 700 (BD 557945), LAG-3- Alexa Fluor 488 (BD 567,610), PD-1-PE-CF594 (BD 566845), CTLA-4-PE-Cy7 (Biolegend 369614), CD160-PE (Biolegend 341206), TIGIT-BV605 (Biolegend 372712), CD244.2-BV510 (Biolegend 329534) VISTA-BV421 (BD 566750), 4-1BB-PerCP-eFluor710 (Invitrogen 46–1379-42), Live/Dead Ghost dye 780 (Tonbo, 13–0865-T100). IFNγ ELISPOT Ninety-six-well ELISPOT plates (Fisher, MAIPS4510) were coated with 1 μg/mL of anti-human IFNγ (Fisher, M700A) in phosphate buffered saline (PBS). Plates were incubated overnight at 4°C, washed, and then blocked with RPMI-1640, with 10% human AB serum, for one hour at room temperature. Immune cell populations were enriched from PBMCs via immunomagnetic negative selection (STEMCELL, 17953 (CD8+ T cells), 17955 (NK cells), 17954 (B cells)). DCs and CD4+ T cells were enriched via immunomagnetic PE-positive selection ((STEMCELL, Vancouver, BC, Canada 17654, CD4-PE (BD 555347), CD11c-PE (BD 560999)). Cryopreserved PBMCs or enriched immune cells were stimulated with TLR agonists as above for an hour, and then incubated with 2 μg/mL of a pool of peptides derived from viral antigens (CEF). After 48 hours, the plates were incubated with 0.5 μg/mL of biotinylated anti-human IFNγ (Fisher, M701B) in PBS/4% bovine serum albumin (BSA) for 90 minutes, followed by one-hour incubation with 1 U/mL of Streptavidin-AP conjugate (Roche, 11089161001). The plates were then developed in buffer (0.1 M Tris pH 9.5, 0.01 M MgCl2, 0.1 M NaCl) supplemented with 300 μg/mL Nitro Blue Tetrazolium (RPI, N66000-1.0) and 150 μg/mL 5-bromo-4-chloro-3-indolyl phosphate (BCIP, Fisher, BP1610-500). IFNγ spots were counted, or the median size of spots were determined, by an automated ELISPOT reader (Cellular Technology Limited, Cleveland, OH). IL-12p70 and IFNβ ELISA Immature dendritic cells prepared from PBMCs from 6 different human blood donors were stimulated with TLR agonists (TLR1 (Pam3CSK4), TLR2 (HKLM) TLR3H (Poly I:C HMW), TLR3L (Poly I:C LMW), TLR4 (LPS-EK), TLR5 (FLA-ST), TLR6 (FSL-1), TLR7 (Imiquimod), TLR8 (ssRNA/LyoVec), TLR9 (ODN2006)) for 24 hours. Cell culture supernatants were then evaluated for the presence of IL-12p70 or IFNβ by ELISA (R&D systems, Minneapolis, MN DY1270-05, DY814-05), according to the manufacturer’s instructions. Statistical analysis Group mean comparisons were performed using GraphPad Prism software, v10. Analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test was used to compare individual group means. For all comparisons, p values < .05 were considered statistically significant. Results Murine CD8+ T cells previously stimulated by the cognate antigen showed decreased PD-1 expression in the presence of TLR agonists at the time of T cell re-activation in vitro We previously demonstrated that TLR agonists or certain combination of TLR agonists administered at the time of T cell activation could improve the anti-tumor activity of CD8+ T cells by downregulation of T cell checkpoint receptor expression.11 However, this was during priming of CD8+ T cells activated at the time of their first exposure to antigen. This situation is not analogous to the human cancer or infectious disease situation in which CD8+ T cells would already have been exposed to target antigens. Consequently, we first wished to determine whether TLR agonists were able to downregulate the expression of T cell checkpoint receptors on murine CD8+ T cells when CD8+ T cells were initially activated by antigen exposure and subsequently re-activated in vitro with the cognate antigen. In this study, we used an adoptive cell transfer model in which OT-1 T cells were transferred to naïve mice, and then immunized, to activate antigen-specific CD8+ T cells present at a more biologically relevant frequency, rather than activating all CD8 T cells by immunizing OT-1 mice directly (Figure 1(A)). Immunization of mice with SIINFEKL alone or with SIINFEKL and TLR agonists led to increased numbers of antigen-specific CD8+ T cells, and in vitro stimulation of splenocytes with peptide and TLR agonists further expanded the antigen-specific CD8+ T cell population (Figure 1(B)). As shown in Figure 1(C), 4-1BB expression on tetramer positive CD8+ T cells did not change upon in vitro stimulation with SIINFEKL alone or with TLR3 agonist, but the expression was lower with TLR9 or TLR3+TLR9 combination when mice were previously immunized with SIINFEKL. Stimulation of splenocytes from the mice immunized with SIINFEKL and TLR agonists increased 4-1BB expression on tetramer positive CD8+ T cells, and the expression was further upregulated with TLR9 agonist. Interestingly, baseline PD-1 expression was elevated in tetramer positive CD8+ T cells when the mice were immunized with SIINFEKL peptide along with TLR agonists (Figure 1(D)). However, the presence of TLR agonists at the time of in vitro re-activation of CD8+ T cells still downregulated the expression of PD-1 on both CD8+ T cells from mice previously immunized with SIINFEKL alone or SIINFEKL with TLR agonists, demonstrating that the presence of TLR agonists at the time of re-activation of murine T cells can lower the PD-1 expression on antigen-specific CD8+ T cells. Murine CD8+ T cells previously stimulated by cognate antigen showed decreased PD-1 expression in the presence of TLR agonists at the time of T cell re-activation in vitro. OT-I splenocytes were adoptively transferred to C57BL/6 mice, and the mice were immunized with PBS control (second group), SIINFEKL peptide alone (third group), or SIINFEKL peptide in combination with TLR3 and TLR9 agonists (fourth group). The first group are OT-1 splenocytes alone, without adoptive transfer. Splenocytes obtained from these groups following treatment were then stimulated in vitro with SIINFEKL peptide (or PBS control) in the presence or absence of TLR3 and/or TLR9 agonists, and the median fluorescence intensity (MFI) of 4-1BB and PD-1 on SIINFEKL-tetramer positive CD8+ T cells were determined by flow cytometry. Shown are the schematic of the experimental procedures (A), the number of tetramer positive CD8+ T cells among CD45+ cells (B), and median fluorescence intensity (MFI) of 4-1BB (C) or PD-1 (D) on tetramer positive CD8+ T cells. Statistical comparisons in each splenocyte group were performed separately, due to the variability between groups. Asterisks indicate p < .05 assessed by one-way ANOVA with Tukey’s multiple comparisons test (Prism software, v10). Error bars represent SEM. The results shown are representative of two independent studies. Presence of TLR agonists at the time of activation of CMV- or EBV-reactive human CD8+ T cells had modest effect on the expression of T cell checkpoint receptors in multiple human blood donors To determine the effect of TLR agonists on human CD8+ T cells, PBMCs were stimulated with TLR agonists, and HLA-A2-specific CMV peptide or EBV peptide, and analyzed by flow cytometry for expression of 4-1BB and T cell checkpoint receptors on tetramer+ antigen-specific CD8+ T cells. In most cases, no significant change was detected in MFI of 4-1BB or PD-1 (Figure 2(A,B), Supplemental Figures S1 and S2) with TLR agonists. However, certain TLR agonists, alone or in combination, showed trends in lowering several T cell checkpoint receptors. In particular, the combination of TLR7 and TLR9 agonists led to lower expression of PD-1, LAG-3 and VISTA on CMV-specific CD8+ T cells (Figure 2(A,G), Supplemental Figure S1(D)). TLR8 agonist and TLR6 + TLR8 combination lowered the expression of TIGIT on CMV-specific CD8+ T cells (Figure 2(E)). In EBV-reactive PBMCs, TLR7 and TLR8 agonists and their combination with TLR9 agonist also lowered the expression of TIGIT (Figure 2(F)). Notably, several combinations with TLR8 or TLR9 agonists showed a trend in lowering the expression of CTLA-4 in both CMV- and EBV- specific CD8+ T cells (Figure 2(C,D)). Additionally, TLR3 agonists also showed a trend in lowering CTLA-4 expression on EBV-specific CD8+ T cells (Figure 2(D)). These data indicate that the presence of TLR agonists at the time of activating human T cells can modulate the expression of T cell checkpoint receptors on antigen-specific CD8+ T cells. Presence of TLR agonists at the time of activation of CMV- or EBV-reactive human CD8+ T cells had modest effect on the expression of T cell checkpoint receptors in multiple human blood donors. Cryopreserved HLA-A2+ human PBMCs containing CMV- or EBV-reactive CD8+ T cells were stimulated with HLA-A2 specific CMV peptide or EBV peptides, and TLR agonists. After 72 hours, the median fluorescence intensity (MFI) of T cell checkpoint receptor expression on CD8+ T cells was determined by flow cytometry. Shown are the average MFI fold-change of PD-1 (A, B), CTLA-4 (C, D), TIGIT (E, F), and LAG-3 (G, H) on tetramer+ CD8+ T cells from each human blood donor. Asterisks indicate p < .05 assessed by one-way ANOVA with Tukey’s multiple comparisons test. Error bars represent SEM. The red line in each panel is the MFI following peptide stimulation without TLR agonist stimulation, for comparison. The results shown are the representative of two independent studies. Stimulation of PBMCs with viral peptide antigens in the presence of TLR3 or TLR8 agonists increased the number of IFNγ-secreting cells. We next tested whether the presence of TLR agonists at the time of T cell activation can improve the functionality of human CD8+ T cells as measured by IFNγ secretion upon the stimulation with a pool of peptides from viral antigens (CEF) and TLR agonists. As indicated in Figure 3(A,B), several TLR agonists slightly increased the number of cells secreting IFNγ, including TLR3H and TLR3L. Notably, TLR8 and the TLR6 + 8 combination showed the greatest increase of IFNγ-secreting cells. We next investigated whether TLR agonists or their combinations affected CD8+ T cells in an antigen-dependent manner. Since TLR3H and TLR8 agonist showed the greatest increase in the number of IFNγ secreting cells, these agonists were prioritized for these studies. Interestingly, TLR8 agonist increased the number of IFNγ-secreting cells from PBMCs regardless of peptide-specific activation (Figure 3(C)). However, no significant increase of the number of IFNγ-secreting cells was detected from isolated CD8+ T cells, CD4+ T cells, NK cells, B cells or DCs in the presence of TLR8 agonist (Figure 3(D), Supplemental Figure S3). Co-incubation of CD8+ T cells with B cells or DCs in the presence of cognate peptide also did not show increased number of IFNγ-secreting cells (Figure 3(E), Supplemental Figure S4). However, the numbers of IFNγ-secreting cells was increased when CD8+ T cells were incubated with DCs in the presence of TLR3 agonist and cognate antigens (Figure 3(H)), whereas no significant increase in IFNγ-secreting cells was detected in CD8+ T cells, CD4+ T cells, NK cells, B cells, or co-incubation of CD8+ T cell with B cells, in the presence of TLR3 agonist without the cognate peptides (Figure 3(G), and Supplemental Figures S3 and S4). These data indicate that TLR8 agonist stimulation can increase the IFNγ secretion from immune cells in an antigen-independent manner, but this is not related to specific activation of CD8+ T cells alone or following activation by antigen-presenting cells. In contrast, activation of TLR3 during CD8+ T cell activation increased the number of IFNγ-secreting cells in an antigen-dependent manner, and this was dependent on activation of CD8+ T cells by antigen-loaded dendritic cells. Stimulation of PBMCs with viral peptide antigens in the presence of TLR3 or TLR8 agonists increased the number of IFNγ-secreting cells. Cryopreserved PBMCs were stimulated in vitro with a pool of peptides from viral antigens (CEF) in the presence of TLR agonists. After 48 hours, the detection of cells secreting IFNγ was evaluated by ELISPOT. Shown are (A) IFNγ spot-forming unit (SFU) per 106 PBMCs, and (B) representative ELISPOT well images from each group. Dendritic cells and CD8+ T cells were isolated from PBMCs and stimulated in vitro with a pool of peptides from viral antigens (CEF) in the presence of TLR3H or TLR8 agonists. After 48 hours, the number of IFNγ-secreting cells was evaluated by ELISPOT. Shown are IFNγ SFU per 106 PBMCs (C,F), isolated CD8+ T cells (D,G), or combined DCs and CD8+ T cells (E,H). Asterisks indicate p < .05 assessed by one-way ANOVA with Tukey’s multiple comparisons test. Error bars represent SEM. The results shown are the representative of two independent studies. TLR3 agonists increased the secretion of IL-12 and IFNβ in human dendritic cells We have previously shown in mice that the effect of TLR agonists and their combinations on CD8+ T cells is due, at least in part, to the secretion of IL-12 and IFNβ from antigen-presenting cells at the time of CD8+ T cell activation.11 Given this, we tested whether TLR agonists affected the production of IL-12 and IFNβ from human dendritic cells. As shown in Figure 4(A), treatment of human DCs with TLR3H, TLR3L showed increases of IL-12 secretion, while TLR8 agonist treatment did not. IFNβ secretion was increased following treatment with TLR3H or TLR9 agonists (Figure 4(B)). This indicates that, similar to murine cells, TLR agonist stimulation can cause human antigen-presenting cells to release IL-12 or IFNβ. TLR agonists increased the secretion of IL-12 and IFNβ from human dendritic cells. Dendritic cells were stimulated with TLR agonists as above for 24 hours, and culture supernatant was evaluated for the presence of (A) IL-12 or (B) IFNβ by ELISA. Asterisks indicate p < .05 assessed by one-way ANOVA with Tukey’s multiple comparisons test. Error bars represent SEM. IL-12 and IFNβ present at the time of antigen-specific CD8+ T cell activation can lower the expression of specific T cell checkpoint receptors We have previously shown that stimulation of murine splenocytes with cognate peptide in the presence of recombinant murine IL-12 and/or IFNβ led to decreased expression of PD-1 on CD8+ T cells.11 Consequently, we tested whether the presence of human IL-12 or IFNβ at the time of T cell activation affected the expression of T cell checkpoint receptors on antigen-specific human CD8+ T cells. In most cases, no significant change was detected in the number of antigen-specific CD8+ T cells or expression of 4-1BB or PD-1 on antigen-specific CD8+ T cells (Figure 5(A,B), Supplemental Figures S5 and S6). IFNβ, alone and/or in combination with IL-12, showed a trend in lowering the expression of CTLA-4, LAG-3 and TIGIT on both CMV- and EBV-specific CD8+ T cells (Figure 5(C–H)). These data indicate that IL-12 and IFNβ released following TLR activation can, at least in part, serve as mediators to modulate T cell checkpoint receptor expression on human CD8+ T cells at the time of their activation. IL-12 and IFNβ present at the time of antigen-specific CD8+ T-cell activation can lower the expression of specific T cell checkpoint receptors. Cryopreserved human PBMCs were stimulated with HLA-A2-specific CMV peptide (NLVPMVATV, left panels) or EBV peptide (FLYALALLL, right panels) in the presence of recombinant human IL-12 or IFNβ. After 72 hours, the MFI of T cell checkpoint receptor expression on tetramer+ antigen-specific CD8+ T cells was determined by flow cytometry. Shown are the average MFI fold-change of PD-1 (A,B), CTLA-4 (C,D), TIGIT (E,F), or LAG-3 (G,H) from each human blood donor. Error bars represent SEM. The results shown are representative of two independent studies. Stimulation of PBMCs with viral peptide antigens in the presence of TLR3H agonist increased the number of IFNγ-secreting cells and magnitude of IFNγ expression. Since the number of IFNγ-secreting cells from PBMCs increased in the presence of TLR3 agonists in an antigen-dependent manner, we next examined the effects of different TLR3 agonists. As shown in Figure 6(A), distinct IFNγ secretion was detected when PBMCs were stimulated with cognate peptides and different TLR3 agonists. Notably, Poly I:C HMW showed the greatest increase of IFNγ, both in terms of the number of cells expressing IFNγ (SFU, Figure 6(A,B)) and the amount of secretion per cell (mean spot size), from PBMCs (Figure 6(B,C)). Stimulation of PBMCs with viral peptide antigens in the presence of TLR3H agonist increased the number of IFNγ-secreting cells and magnitude of IFNγ expression. (A) Cryopreserved PBMCs from a single donor were stimulated in vitro with a pool of peptides from viral antigens (CEF) in the presence of the following TLR3 agonists: Poly I:C HMW, Poly I:C LMW, Poly A:U, or NexaVant. After 48 hours, the number of IFNγ-secreting cells was evaluated by ELISPOT. Shown are (A) IFNγ SFU per 106 PBMCs, (B) ELISPOT well images from each group, and (C) mean spot size of IFNγ spots from the same ELISPOT wells. Asterisks indicate p < .05 assessed by one-way ANOVA with Tukey’s multiple comparisons test. Error bars represent SEM. The results shown are the representative of two independent studies with another blood donor. Discussion In previous reports, we found that murine T cell activation in the presence of TLR agonists resulted in greater anti-tumor immunity with altered expression of several T cell checkpoint receptors.10,11 Here we investigated whether TLR agonists similarly impact human CD8+ T cells by lowering the expression of T cell checkpoint receptors during their activation. We first tested whether TLR agonists can suppress the expression of PD-1 when re-stimulating murine CD8+ T cells that were previously exposed to the antigen, a situation more analogous to human CD8+ T cells specific for viral or tumor antigens. We found that antigen-experienced CD8+ T cells also exhibited lower PD-1 expression when T cells were re-activated in vitro with antigen and TLR agonists, although the magnitude of expression was markedly reduced in antigen-experienced CD8+ T cells activated without prior TLR agonist stimulation. Consequently, we investigated whether the presence of TLR agonists at the time of antigen-specific CD8+ T cell activation can lower the expression of T cell checkpoint receptors, using human PBMCs containing known antigen-specific CD8+ T cells. Specifically, we used PBMC containing CD8+ T cells reactive to either CMV pp65 (NLVPMVATV) or EBV LMP2 (FLYALALLL). In contrast to expectations, TLR agonists, and combination of TLR agonists such as TLR3+TLR9, showed no significant alteration of PD-1 expression on antigen-specific human CD8+ T cells. Certain TLR agonists showed a modest decrease in the expression of other T cell checkpoint receptors on either CMV- or EBV-specific CD8+ T cells, but little consistency between CMV- and EBV- specific CD8+ T cells. However, the presence of TLR3 or TLR8 agonists at the time of T cell activation led to increased numbers of IFNγ-secreting cells, suggesting these TLR agonists can affect the functionality of T cells. Unlike TLR3 agonist, stimulation of PBMC with TLR8 led to increased IFNγ secretion even in the absence of antigen-specific activation. Stimulation of human dendritic cells with a TLR3 agonist (Poly I:C HMW) significantly increased IL-12 and IFNβ secretion, and the presence of recombinant IL-12 or IFNβ at the time of T cell activation modestly decreased the expression of some immune checkpoint receptors on CD8+ T cells. Taken together, these results suggest: 1) T cell checkpoint receptor expression appears to be less affected by TLR agonists during activation of human CD8+ T cells compared to mice; 2) TLR3 agonists had the greatest effect on the functionality of human CD8+ T cells when used during T cell activation, at least in terms of IFNγ expression; and 3) there are likely differences in TLR3 agonists for their effects on CD8+ T cell function. Our findings indicate that T cell checkpoint receptors appear to be less affected by TLR agonists employed at the time of CD8+ T cell activation on human CD8+ T cells compared to mouse CD8+ T cells. We have previously demonstrated that the presence of several TLR agonists used at the time of murine CD8+ T cell activation lowered the expression of PD-1 and other T cell checkpoint receptors.10,11 Specifically, combinations of TLR1 and TLR2, or TLR3 and TLR9, agonists significantly downregulated the expression of PD-1, LAG-3, and CD160. In the current studies, no TLR agonists demonstrated significant decreases in any of the T cell checkpoint receptors evaluated on both EBV- and CMV-specific CD8+ T cells. This difference likely stems, at least in part, from the antigen-experienced state of the CD8+ T cells. Specifically, PD-1 expression was demonstrably lower on previously antigen-exposed CD8+ T cells following peptide activation in mice (Figure 1). It is conceivable that activation of human CD8+ T cells with completely foreign (not previously encountered) antigens would have led to a different outcome. That situation may be relevant to the use of TLR agonists as vaccine adjuvants for antigens to which the host has never been exposed, such as for infectious disease prophylaxis; however, it is not relevant to cancer treatment vaccines in which the host would have been previously exposed to the antigen. Our results indicate that the presence of TLR3 agonist stimulation at the time of human CD8+ T cell activation can improve the functionality of human CD8+ T cells, at least in terms of IFNγ expression. It is well known that activation of TLR3 results in secretion of type-I IFNs, such as IFNβ. Type-I IFNs elicit effective antiviral immunity and T-cell activation against viral antigens, however their role in anti-tumor immunity is more controversial; type-I IFNs are indeed essential for immune surveillance and promoting anti-tumor responses, but excessive or prolonged type-I IFN signaling has been shown to be associated with the induction of immunosuppressive mechanisms within the tumor microenvironment.21,22 Therefore, several studies have explored combining TLR3 agonists with other TLR agonists, such as TLR7, 8, or 9, since these TLRs have been shown to induce robust Th1 responses and increase the secretion of Th1-associated cytokines such as IL-12.23,24 We previously demonstrated that several TLR agonists, alone or in combination, used at the time of T cell activation, significantly increased CD8+ T cell secretion of IFNγ and their anti-tumor efficacy in murine models.10–12 In particular, we demonstrated that the combination of TLR3 and TLR9 agonists elicited the greatest anti-tumor immune response, along with significant increases in IL-12 and IFNβ secretion from DCs.11 In the current study, however, we found that only TLR3 agonist stimulation led to increased secretion of IFNγ by antigen-specific human CD8+ T cells, which was dependent on the co-incubation of antigen-loaded DCs with CD8+ T cells. Moreover, we found that TLR3 activation led to secretion of both IL-12 and IFNβ by human DC, while TLR3 activation primarily led to secretion of IFNβ by murine DCs. Our results suggest that TLR3 is thus a preferred receptor to target during the activation of human antigen-specific CD8+ T cells, such as would be used during anti-tumor immunization. Our results suggest that different agonists intended to activate the same TLR3 receptor may have different effects on CD8+ T activation. For example, the expression of T cell checkpoint receptors on human CD8+ T cell was slightly different when activated with TLR3H (Poly I:C HMW) stimulation versus TLR3L (Poly I:C LMW) stimulation. TLR3H stimulation showed a trend toward decrease of CD160 expression, while TLR3L stimulation showed decreased VISTA expression on EBV-specific CD8+ T cells. Differences in IFNγ secretion were also detected when CD8+ T cells were activated in the presence of different TLR3 agonists, favoring the use of TLR3H (PolyI:C HMW, Figure 6). We used equivalent molar concentration of each TLR3 agonist, however it is possible that different TLR3 agonists have different optimal concentration to most effectively activate TLR3. Different TLR3 agonists have shown distinct binding kinetics25 and intensity of downstream signaling and cytokine production,26,27 either of which could lead to differences in modulating the expression of T cell checkpoint receptors on CD8+ T cells and their functionality. Besides TLR3, several reports have demonstrated that different agonists targeting other TLR can have distinct immunological functions. For example, TLR4 ligands, LPS and MPLA, have been shown to have different effects on CD8+ T cells; LPS can impact the generation of memory CD8+ T cell populations, while MPLA, a low-toxicity derivative of LPS, led to enhanced clonal expansion of effector CD8+ T cell populations.28 Two different TLR7 agonists, imiquimod and gardiquimod, differentially affected the proliferation of splenocytes, the expression of costimulatory molecules on dendritic cells, and displayed different anti-tumor efficacy in a B16 murine melanoma model.29 Consequently, future studies should evaluate the distinct immunological functions of different agonists targeting the same TLR to identify optimal agents for use in specific applications. Overall, our findings demonstrate that TLR agonists, and notably TLR3 agonists, can affect the function of human CD8+ T cells during their activation. These findings suggest that TLR3 agonists are ideally poised to use as adjuvants for human anti-tumor vaccines. However, our study has several limitations. First, our evaluations used PBMCs from a small number of human blood donors and most of our studies used CD8+ T cells specific for only two epitopes, both restricted by HLA-A2. It is conceivable that there may be differences in other MHC backgrounds. Moreover, it is possible that results obtained using PBMCs from cancer patients may have differences compared to those from healthy donors. However, our study specifically used blood obtained from cancer patients, as this was most relevant to our long-term goal of identifying TLR to serve as adjuvants for anti-tumor vaccines. Finally, viral antigens were used as model antigens in the current study instead of tumor antigens. This was necessary due to the paucity of known shared human tumor antigens and the rarity of tumor antigen-specific CD8+ T cells in peripheral blood specimens. We reasoned that viral antigen-specific CD8+ T cells would serve as a reasonable model given the fact that the detection of viral antigen-specific CD8+ T cells was due to prior viral exposure rather than vaccination, similar to cancer antigens, and they are in greater abundance in peripheral blood specimens. However, viral antigens are still non-“self” antigens, and the regulation and activity of virus-specific CD8+ T cells may be different from tumor antigen-specific CD8+ T cells. For example, central and peripheral T cell tolerance mechanisms likely result in lower frequencies of tumor antigen-specific CD8+ T cells compared to viral antigens. Therefore, future studies will aim to explore the use of TLR agonists in modulating the function of tumor antigen-specific CD8+ T cells, perhaps following anti-tumor vaccination, with highly sensitive detection methods. Specifically, methods such as single cell sequencing and TCR sequencing may be necessary to identify rarer tumor antigen-specific CD8+ T cell clones and enable analysis of their functional phenotypes. Our studies were performed with the intent of identifying TLR agonists that might serve as preferred vaccine adjuvants for human anti-tumor vaccines to specifically improve the function of vaccine-activated tumor antigen-specific CD8+ T cells. Our studies suggest that TLR3 agonists may be preferred for this purpose. TLR3 agonists have been previously investigated for other cancer immunotherapy applications, including poly-ICLC and rintatolimod.3 In particular, several trials have used poly-ICLC, administered intratumorally or systemically as an immune modulating agent, in patients with glioblastoma, melanoma, and breast cancer.30–33 It has also been used as a cancer vaccine adjuvant, notably in trials with peptide vaccines for patients with melanoma, and using protein vaccines targeting NY-ESO-1.3,34–36 Our findings suggest it could be used in other contexts as well, for example using bispecific T cell engager agents to more effectively activate tumor-reactive CD8+ T cells, in combination with T cell checkpoint blockade therapy,12 or perhaps in the production of more effective tumor-specific CD8+ CAR-T cells or TCR adoptive therapy approaches. In all of these approaches, TLR3 agonists might be useful to augment the anti-tumor efficacy of tumor-specific CD8+ T cells. Supplementary Material Supplemental Figures and Legends.docx Acknowledgments All studies were conducted and analyzed between July 2022 and September 2025. Some of the data in this publication were previously published as part of Dr. Jeon’s doctoral thesis,37 and were presented at the 2024 national meeting of the Society of ImmunoTherapy of Cancer.38 HJ wrote the manuscript, designed the experiments, and carried out data analysis. JEM and EH participated in data collection and analysis. DGM oversaw the experimental design and is responsible for the overall content as guarantor. All authors contributed to the writing and approval of the final manuscript. Biography Douglas G. McNeel, MD PhD is a tenured Professor of Medicine at the University of Wisconsin, Madison. He is a genitourinary medical oncologist with a laboratory and clinical research program that has focused on prostate cancer immunology since 2001. The long-term goal of these efforts is to develop effective anti-tumor immunotherapies as treatments for prostate cancer. Shorter term goals of this research have been to identify immunologically recognized proteins of the prostate, characterize these as tumor target antigens, evaluate anti-tumor genetic vaccines in pre-clinical models to understand their mechanisms of action and resistance, and translate these findings to early phase human clinical trials. Funding Statement This work was supported by National Institutes of Health Clinical Center [P30 CA014520, R01 CA129154 and P50 CA269011]. Disclosure statement No potential conflict of interest was reported by the author(s). Data availability statement The data generated and/or analyzed during this study are available from the corresponding author on reasonable request. Ethics approval and consent to participate This study was conducted in accordance with the Declaration of Helsinki, using blood samples previously collected from human participants who had participated in clinical trials between 2004 and 2016, and who gave written, informed IRB-approved consent permitting blood specimens to be used for future cancer research. The use was approved by the University of Wisconsin Health Sciences Human Subjects Review Board (IRB) Protocol 2013–0126. The use of laboratory mice was conducted under a protocol (M005690) approved by the University of Wisconsin School of Medicine and Public Health Institutional Animal Care and Use Committee that conforms to the NIH guide for the care and use of laboratory animals. Euthanasia of animals in this study was performed by CO2 asphyxiation, as approved by the American Veterinary Medical Association. Supplementary material Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2025.2598694 References Associated Data Supplementary Materials Supplemental Figures and Legends.docx Data Availability Statement The data generated and/or analyzed during this study are available from the corresponding author on reasonable request.