Targeting of Toll-like receptors inhibits CD4⁺ regulatory T cell function and activates lymphocytes in human PBMCs

Abstract

Accumulating evidence suggests elements within tumors induce exhaustion of effector T cells and infiltration of immune-suppressive regulatory T cells (Tregs) thus preventing the development of durable anti-tumor immunity. Therefore, the discovery of agents that simultaneously block Treg suppressive function and reinvigorate effector function of lymphocytes is key to the development of effective cancer immunotherapy. Previous studies have shown that Toll-like receptor ligands (TLRL) could modulate the function of these T-cell targets; however, those studies relied on cell-free or accessory cell-based assay systems that do not accurately reflect in vivo responses. In contrast, we employed a human PBMC-based proliferation assay system to simultaneously monitor the effect of TLRLs on T cells (CD4⁺, CD8⁺, Tregs), B cells and NK cells, which gave different and even conflicting results. We found that the TLR7/8L:CL097 could simultaneously activate CD8⁺ T cells, B cells and NK cells plus block Treg suppression of T cells and B cells. The TLRLs TLR1/2L:Pam₃CSK4, TLR5L:flagellin, TLR4L:LPS and TLR8/7L:CL075 also blocked Treg suppression of CD4⁺ or CD8⁺ T cell proliferation but not B cell proliferation. Besides CL097, TLR2L:PGN, CL075 and TLR9L:CpG-(A-C) were strong activators of NK cells. Importantly, we found that Pam₃CSK4 could: 1) activate CD4+ T cells proliferation; 2) inhibit the expansion of IL-10⁺ nTregs and induction of IL-10⁺ CD4⁺ Tregs (Tr1); and 3) block nTreg suppressive function. Our results suggest these agents could serve as adjuvants to enhance the efficacy of current immunotherapeutic strategies in cancer patients.

Introduction

Active immunotherapy is a promising approach for the treatment of cancer; however, the clinical response rates following therapeutic cancer vaccination have been low (1, 2). Many studies have reported that the immune-suppressive elements within a tumor induce exhaustion of effector T cells (Teff), infiltration of immune-suppressive regulatory T cells (Tregs) and secretion of the anti-inflammatory cytokines TGF-β and IL-10 (3-6). These cytokines can induce the generation of regulatory DCs (DCregs) and maintain CD4⁺ natural occurring FOXP3⁺ regulatory T cells (nTregs) or convert CD4⁺ T cells into inducible IL-10⁺/TGF-β⁺Tregs (iTregs) (4-8). All these elements work against the development of effective cancer immunotherapy strategies by suppressing the immune system and providing an environment that favor tumor growth.

Evidence from the literature suggests that these suppresive elements within the tumor microenvironment can be modulated by triggering signals from members of the toll-like receptor (TLR) family (9, 10). TLRs belong to a family of conserved pattern recognition receptors (PRRs) that recognize unique molecular structures of pathogens in order to distinguish “infectious non-self” from “self” antigens (11), allowing them to sense and initiate innate and adaptive immune responses. To date, ten functional TLRs have been identified in humans with nine known agonists (TLRL1-9) (12). These TLRs are expressed by antigen-presenting cells (APCs), tumor cells and both Teff and Treg cells (13-15). Recent studies using TLR agonists have shown that certain types of TLRs, expressed on different cells, display alternate functions. For instance: 1) on T cells, they function as co-stimulatory receptors to enhance TCR-induced Teff cell proliferation, survival and cytokine production (16); 2) on suppressive Tregs, they can function to block Treg function (10, 17); and 3) on APCs, they induce autocrine maturation and secrete pro-inflammatory cytokines leading to the modulation of Teff cell and Treg function (18). Although these studies identified TLRLs that can reinvigorate Teff cells function and block Treg suppressive function, they showed conflicting results, probably because they relied on cell-free (plate-bound or beads conjugated with anti-CD3) or accessory cell-based experimental systems (soluble anti-CD3 plus monocytes, DCs or CD3-depleted PBMCs) that do not necessarily reflect the in vivo response. For instance, by using a DC-based proliferation system, Peng et al., (17) reported that only CpG-A could block Treg suppressive function, while other TLRLs had no effect. In contrast, by using a cell-free proliferation system, Nyirenda and colleagues (10) showed that a TLR2 ligand blocked Treg function. Because responder T cells are likely to interact with different T cell subtypes and with APCs in vivo, we believe that the use of whole PBMCs, which contain most cell types found in vivo (CD4⁺, CD8⁺, γδ⁺ T cells, CD4⁺Tregs, CD8⁺Tregs, Th17 cells, monocytes, mDCs, pDCs, among others), would result in mimicking the in vivo responses following TLRL stimulation.

In this study we used PBMCs that contained all T cell subtypes and APCs as accessory cells for our proliferation/suppression assays (19). We found that five of the nine known TLRL (Pam₃CSK4, LPS, flagellin, CL097 and CL075) were able to completely block nTreg suppression on CD4⁺ or CD8⁺ Teff cell proliferation. Analyzing the complete dataset, we found that the TLR7/8L:CL097 could simultaneously activate CD8⁺ T cells, B cells and NK cells plus block Treg suppression on CD4⁺/CD8⁺ T and B cells proliferation. Furthermore, we found that TLR1/2L:Pam₃CSK4 could work directly to: 1) stimulate CD4⁺T cell proliferation, 2) inhibit the expansion of IL-10⁺ nTregs, 3) block the induction of IL-10⁺ CD4⁺ Tregs (Tr1)from total CD4⁺ T cells; and 4) block nTreg suppressive function. Our results suggest the potential use of these agents as adjuvants to enhance the efficacy of therapeutic vaccines and other immunotherapeutic strategies in cancer patients.

Materials and Methods

Reagents and cell lines

TLR ligands: Pam₃CSK4, PGN, LTA, poly(I:C), CL075, CL097, LPS-SM, flagellin, CpG-A (ODN2216), CpG-B (ODN2006) and CpG-C (ODN M362) were purchased from Invivogen (San Diego, CA, USA). Phorbol myristate acetate (PMA), ionomycin (ION) and brefeldin A (BFA) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Elisa kits for IL-2, IL-10, TNF-α and IL-6 and neutralizing antibodies against IL-2 and IL-6, and normal goat IgG control were purchased from R&D Systems (Minneapolis, MN, USA). Fetal calf serum and human serum were purchased from Gemini (Manhattan, New Jersey, USA). Carboxyfluorescein succinimidyl ester (CFSE) was purchased from Invitrogen (Grand Island, NY, USA).

PBMC-based proliferation/Treg suppression assays/cytokine production

Adult blood buffy coat samples from healthy donors were obtained from the Gulf Coast Regional Blood Center in Houston, Texas (Human Research Protocol LAB-03-0390- MDACC). Human PBMCs were isolated from the adult blood buffy coats by Ficoll-Paque (GE Healthcare, Waukesha, WI, USA) density gradient centrifugation according to manufacturer’s procedures. Briefly, a diluted suspension of the buffy coat was layered over 15 ml of Ficoll-Paque and centrifuged at 1100 x g for 15 minutes at 20°C without the brake. The mononuclear cell layer containing lymphocytes, monocytes, and thrombocytes was then transferred to a new 50 ml conical tube, filled with PBS and centrifuged at 260 x g for 8 min. After the spin, the cell pellet was resuspended, diluted with PBS and centrifuged at 180 x g for 8 min. The cell pellet was washed one more time with PBS by spinning at 625 x g for 7 minutes. Cells were then resuspended with PBS and labeled with 3 μM of CFSE. Next, 2.5 × 10⁵ CFSE-labeled PBMCs were stimulated with soluble anti-CD3 (1 μg/ml) in T cell medium containing 10% human AB serum (Gemini) in RPMI 1640-GlutaMAX plus 1% penicillin-streptomycin. After 3.5 days of culturing, the PBMCs were stained with APC-Cy7-CD4, Pacific Blue-CD8 and PE-CD19 antibodies. B cells were identified by gating on CD4^-CD8^-CD19⁺ cells. To determine the Treg suppressive activity, 2.5 × 10⁵ CFSE-labeled PBMCs were stimulated with soluble anti-CD3 (1 μg/ml) in the presence of autologous 1 × 10⁵ CD4⁺CD25^highCD127^low nTregs in T cell medium. Proliferation of CD4⁺ T, CD8⁺ T, and B cells was monitored by CFSE dilution assessed by flow cytometry. For evaluation of cytokine production, 2.5 × 10⁵ PBMCs were stimulated with soluble anti-CD3 (1 μg/ml) for 3.5 days. Supernatants were collected and analyzed for IL-10, IL-6, IFN-γ and TNF-α production.

Screening of TLR ligands that block nTreg suppression of lymphocytes

TLR ligands were screened using PBMC-based proliferation/suppression assays. The following TLRLs were tested at 4 different concentrations for T cell proliferation according to the manufacturer’s specification and our preliminary data: a) TLR1/2L: Pam₃CSK4 (2, 10, 50, 100 ng/ml); b) TLR2L: PGN (0.5, 1, 5, 10 μg/ml) and LTA (0.1, 0.25, 0.5, 1.0 μg/ml); c) TLR3L: poly (I:C) (0.1, 0.5, 1, 5 μg/ml); d) TLR4L:LPS-SM (0.25, 1.0, 2.5, 5.0 μg/ml); e) TLR5L: flagellin (ultrapure, 2, 20, 50, 100 ng/ml); f) TLR2/6L: FSL (2, 10, 50, 100 ng/ml); g) TLR7/8L: CL075 and CL097 (0.1, 0.5, 1, 5 μg/ml); h) TLR9L: CpG-A (ODN2216), CpG-B (ODN2006) and CpG-C (ODN M362)(0.25, 1.0, 2.5, 5.0 μM). After 3-4 days in culture, the ability of the added TLRL to enhance T cell proliferation or reverse Treg suppression was analyzed based on the proliferation of CD4 or CD8 T effector (T_eff) cells by CFSE dilution assays. The following TLRLs were tested at three concentrations for CD4⁺ and CD8⁺ T cells and B cells proliferation in the presence of nTregs: a) Pam₃CSK4 (10, 50, 100 ng/ml); b) PGN (1, 5, 10 μg/ml); c) poly (I:C) (0.1, 0.5, 1.0 μg/ml); d) LPS-SM (0.1, 0.25, 1.0 μg/ml); e) flagellin (20, 50, 100 ng/ml); f) FSL (10, 50, 100 ng/ml); g) CL097 (0.1, 0.5, 1.0 μg/ml); h) CpG-A (ODN2216), CpG-B (ODN2006) and CpG-C (ODN M362)(0.25, 1.0, 2.5 μM).

Purification of T-cell subsets and NK cells

CD4⁺ T cells were enriched from peripheral blood buffy coat samples using a CD4 T cell isolation kit (StemCell Techologies, Vancouver, BC, Canada) according to manufacturer’s procedures. CD4⁺ T cells were stained with APC-Cy7-CD4, PE-Cy7-CD25, PE-CD127, perCP-Cy5.5-CD45RA, APC-CD45RO antibodies and FITC-labeled lineage cocktail antibodies against CD11c, CD14, CD16, CD19, CD56, CD303 and γδ-TCR (Becton Dickinson Biosciences, San Jose, California, USA and Mitenyi Biotec, Auburn, CA, USA; FITC-CD303). Stained cells were sorted on a FACSAria cell sorter (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA). CD4⁺FITC lineage negative cells were sorted into three fractions: 1) naïve T (CD4⁺CD25^lowCD45RA⁺); 2) memory T (CD4⁺CD25^lowCD45RA^-CD45RO⁺); and 3) nTregs {CD4⁺CD25^highCD127^low (top 4-7%)}. NK cells were enriched using EasySep human NK cell enrichment kit (StemCell Techologies) according to manufacturer’s protocols and stained with APC-CD56, PE-CD16, Pacific Blue-CD3 antibodies and FITC-labeled lineage cocktail antibodies against CD14, CD19, CD11c and γδ-TCR. Pure CD14^-CD19^-CD11c^-γδ^-TCR^-CD3^-CD56⁺CD16⁺ NK cells were obtained by sorting on a FACSAria.

Screening of TLR ligands that directly enhance T cell proliferation

Freshly sorted 1×10⁵ T cells (CD4 naive, CD4 memory, CD4 nTregs, CD3⁺CD45RA⁺CD27⁺ CD8) were stimulated with plate-bound anti-CD3 {CD4 (2 μg/ml), CD8 (1 μg/ml)} in T cell medium in the absence or presence of single TLRLs. The following TLRLs were tested at five concentrations: Pam₃CSK4, FSL and flagellin (400, 200, 100, 50, 25 ng/ml); poly (I:C) (5, 2.5, 1.25, 0.62, 0.315 μg/ml) and CpG-(A-C) (5, 2.5, 1.25, 0.62, 0.315 μM); LPS-SM, CL097 and CL075 (1, 0.5, 0.25, 0.125, 0.063 μg/ml). ³H-thymidine was added on the third day of culture and cells were harvested after another 15 hours of incubation. Proliferation of T cells was evaluated by thymidine incorporation.

FOXP3⁺nTregs treatment with TLRLs

Freshly sorted CD4⁺CD25^highCD127^low nTregs were cultured in T cell medium containing 5% human AB serum in the presence of each separate TLRL tested: Pam₃CSK4, FSL and flagellin (400 ng/ml); poly (I:C) (5 μg/ml); CpG-(A-C) (2.5 μM); LPS-SM, CL097 and CL075 (1 μg/ml) for 24 h. nTregs were then washed three times and co-cultured with autologous CFSE (4 μM)-labeled PBMCs (1×10⁵ nTregs:2.5×10⁵ PBMCs) in the presence of 1.5 μg/ml of anti-CD3 (OKT3). Proliferation of CD8⁺ T cells after four days of stimulation was monitored by CFSE dilution assessed by flow cytometry.

Stimulation of nTregs in the presence of Pam₃CSK4 and flagellin

Freshly sorted 4 × 10⁵ nTregs were stimulated with plate-bound anti-CD3 (2 μg/ml) in T cell medium containing 10% FCS plus IL-2 (300 IU/ml) and soluble anti-CD28 (0.25 μg/ml) in a 24-well tissue culture plate in the presence or absence of 50 ng/ml of Pam₃CSK4 or flagellin for seven days. Expanded nTregs were either stained with a FoxP3 antibody or restimulated with PMA (50 ng/mL) and ION (2 μg/mL) for 6 h. During the last 4 h, BFA-protein trafficking blocker (10 μg/mL) was added. The cells were stained with IL-2, IFN-γ, TNF-α or IL-10 antibodies (eBioscience) using Caltag FIX and PERM kit (Invitrogen).

Generation of IL-10-producing Tr1 cells from CD4⁺ T cells

A total of 2 × 10⁵ freshly isolated CD4⁺ T cells were cultured with irradiated (60 Gy) ICOSL-expressing CD32-L cells (8 × 10⁴), which were plated for 2 hours and pre-coated with anti-CD3 (0.2 μg/ml) for another hour, in the presence of dexamethasone (5 × 10^-8 M) and 1α,25-dihydroxyvitamin D3 (1 × 10^-7 M) (Life Technologies, Carlsbad, CA, USA) in T cell medium containing 10% FCS plus IL-2 (50 IU/ml) and soluble anti-CD28 (0.2 μg/ml) for 7 days in a 48-well tissue culture plate. Expanded T cells were restimulated with PMA/ION as above described. Intracellular staining was performed on cells with AlexaFluor⁶⁴⁷-IL-10 antibody (clone JES3-9D7, eBioscience) using Caltag FIX and PERM kit (Invitrogen).

Statistical analysis

Statistical difference between experimental groups was determined by paired or unpaired t test or two-way ANOVA test using Prism software (GraphPad Software, Inc).

Results

Identification of TLRLs that promote T cell proliferation, block nTreg function and activate NK cells in a PBMC-based proliferation system

Responder T cells and CD4⁺FOXP3⁺Treg cells are likely to interact with different T cell subtypes and with antigen-presenting cells (APCs) in vivo. Therefore, to evaluate the ability of TLRLs to enhance T cells proliferation and block Treg suppression function, we employed a PBMC-based proliferation assay system (19), which contains most cell types found in vivo. First, we evaluated lymphocyte proliferation using CFSE-labeled PBMCs polyclonally stimulated with anti-CD3 mAb in the presence or absence of the detailed TLRL, each tested at four different concentrations. Then, to evaluate the ability of the TLRLs to CD4⁺CD127^low CD25^high nTregs were added at a 1:1 (lymphocyte : nTreg) ratio under the same culture conditions. After 3.5 days of culture, the proliferation of CD4⁺ T, CD8⁺ T and B cells was assessed by CFSE dilution, as described in Methods, and found to have undergone 5-6 rounds of cell division (Supplementary Fig. 1). In the absence of nTregs, TLR1/2L:Pam₃CSK4, TLR4L:LPS and TLR5L:flagellin were the most potent ligands to enhance TCR-mediated CD4⁺ T cell proliferation, while the remaining ligands tested were ineffective (Fig. 1A, upper panel and Table I). We also found that TLR7/8L:CL097 potently stimulated CD8⁺ T cell proliferation in addition to Pam₃CSK4, LPS and flagellin, (Fig. 1A, middle panel and Table I). In contrast, only TLR7/8L:CL097 and TLR9Ls:CpG-(B and C) were able to potently stimulate B cell proliferation (Fig. 1A, lower panel). In the presence of autologous nTregs, however, we found that TLR2/6L:FSL and TLR7/8L:CL097 that did not enhance CD4⁺ T cell proliferation,, were able to block nTreg suppression of CD4⁺ T cell proliferation (Fig. 1B, upper panel and Table I). When CD8⁺ T cell proliferation was evaluated, Pam₃CSK4, LPS and CL097 simultaneously exerted both functions of activation of proliferation and inhibition of suppression mediated by Tregs (Fig. 1B, middle panels, Table 1). We then expanded our studies to multiple donor PBMCs and found that TLR7/8L:CL097 could consistently block nTreg suppression of all CD4⁺ T and CD8⁺ T and B cell proliferation (Fig. 2A-C). Other TLRLs such as Pam₃CSK4, flagellin, LPS and TLR8/7L:CL075 only blocked Treg suppression of CD4⁺and CD8⁺ T cell proliferation (Fig. 2A&B) but did not affect B cell proliferation (Fig. 2C). Of note, we found that FSL could block nTreg suppression of CD4⁺ but not CD8⁺ T cell proliferation (Fig. 2A&B). Besides CL097, CpG-B and CpG-C were the most potent in blocking Treg suppression of B cell proliferation. Because NK cells were shown to kill tumor targets in vitro and in vivo (20), we sought to determine the effect of TLRLs in activating NK cells. CFSE-labeled, freshly sorted NK cells were added to autologous PBMCs and stimulated with a TLRL for one day. Afterward, NK cell activation was determined by the expression of CD69 on CFSE⁺ cells. Figure 2D shows that PGN, CL097, CL075 and CpG-(A-C) were strong activators of NK cells. Because a small percentage of lymphocytes are already activated in vivo, we sought to determine if TLRLs could enhance their proliferation in the absence of stimulation by anti-CD3 mAb. We found that Pam₃CSK4, LPS and Flagellin do not require polyclonal stimulation with anti-CD3 to enhance CD4+ T cell proliferation but they need this stimulus to activate proliferation on CD8+ T cells (Fig. 3A&B). On the contrary, 6 out of 11 TLRLs increased B cell proliferation in the absence of anti-CD3 (Fig. 3B) compared to only three active TLRLs when PBMCs were polyclonal activated. We next sought to evaluate the pro- and anti-inflammatory cytokines that are induced by TLRLs in the polyclonal activated TCR PBMC-based system. We found that the majority of the 11 TLRLs could induce anti-inflammatory IL-10 secretion except Pam₃CSK4, FSL, Ploy(IC) and flagellin (Fig. 4A). Of the three pro-inflammatory cytokines tested, only IL-6 and TNF-α responded to TLRL stimulation (Fig. 4B-D). We found that CpG-A, -B, and –C, and poly(IC) did not induce IL-6 and TNF-α Fig. 4C&D). In addition, Flagellin and CL097 also failed to induce TNF-α Taken together, these results suggest that Pam₃CSK4, FSL and Flagellin are better agonists that only induce pro-inflammatory cytokines.

FIGURE 1.

Identification of TLRLs that enhance proliferation and/or block nTreg suppression of lymphocytes within PBMCs. (A) Proliferation of CD4⁺ and CD8⁺ T cells and B cells following 3.5 day culture of CFSE labeled PBMCs stimulated with soluble anti-CD3 in the absence or presence of a TLRL, 4 concentrations were tested as described in Methods. White color bars and dark color bars indicate lowest and highest concentration, respectively. T and B cell proliferation was determined by CFSE dilution assessed by flow cytometry as described in Methods. (B) Proliferation of CD4⁺ and CD8⁺ T cells and B cells under culture conditions describe in (A) with autologous CD4⁺CD25^highCD127^- nTregs added to the PBMCs at a 1:1 ratio of nTregs to effector cells (assuming that lymphocytes constitute about 40% of total PBMCs). Data are representative of independent experiments using PBMCs derived from n=3 (2A), n=4(2B), and n=5(2C).

Table I. Summary of TLRLs that enhance T and B cell proliferation, block nTreg function and activate NK cells in two different proliferation systems.

TLRL/Cells	CD4+Proliferation	CD8+Proliferation	B Proliferation	Blocking T reg/CD4+ suppression		Blocking T reg/CD8+ suppression	Blocking Treg/B suppression	NK proliferation
				PBMC system	Three components system
Pam	+	+	-	++	++	+	-	-
LPS	+	+	-	++	NA	+	-	-
Flagellin	+	+	-	+	++	+	-	-
CL097	-	+	+	++	-	+	+	+
CL075	NA	NA	NA	++	-	+	-	+
CpG-A	-	-	-	-	++	-	-	+
CpG-B	-	-	+	-	NA	-	+	+
CpG-C	-	-	+	-	NA	-	+	+
PGN	-	-	-	-	+	-	-	+
FSL	-	-	-	+	-	-	-	-
Poly IC	-	-	-	-	NA	-	-	-
LTA	-	-	NA	NA	NA	NA	NA	NA

FIGURE 2.

Identification of TLRLs that consistently block nTreg suppression of lymphocytes and activate NK cells. (A-C) Relative proliferation of CD4⁺ and CD8⁺ T cells and B cells under conditions described in Figure 1 using three concentrations of TLRL described in Methods. Lymphocyte proliferation was assessed by flow cytometry for CFSE dilution. Data points for each TLRL were taken from one of the three TLRLs concentration tested that yielded the highest lymphocyte proliferation. Error bars represent means ± SEM. Statistical significance between treatment groups (P<0.05 compared to responder cells plus nTregs) is indicated by a left square bracket ([). P values were calculated by paired t test. Data are representative of independent experiments using PBMCs derived from 4-5 donors. (D) Relative percentage of 69⁺ NK cells following 1 day culture of PBMCs (0.25×10⁶) and autologous CFSE-labeled NK cells (1×10⁵/well) in the presence or absence of a TLRL at different concentrations achieved by using four successive one-half dilutions: Pam₃CSK4, FSL and flagellin (400-50 ng/ml); poly(I:C) (5-0.63 μg/ml); and CpG-(A-C) (5-0.63 μM); LPS, CL097 and CL075 (1-0.125 μg/ml). NK cell activation was evaluated by the expression of CD69 activation markers by flow cytometry analyses. Data are representative of 2 donors.

FIGURE 3.

Identification of TLRLs that enhance proliferation of CD4⁺T, CD8⁺ T and B cells without activation through TCR receptor. Proliferation of CFSE labeled non-stimulated PBMCs in the absence or presence of single TLRLs, at a fix concentration. T and B cell proliferation was determined by CFSE dilution assessed by flow cytometry as described in Methods. Data are representative of experiments using PBMCs derived from 6 donors plotted individually. left square brackets ([) indicate significance compared to the control with no TLRL added.

FIGURE 4.

Cytokine production by PBMCs activated through TCR stimulus. PBMCs were stimulated with soluble anti-CD3 mAb. Supernatants were collected after 3.5 days and analyzed for (A) IL-10, (B) IFN-γ, (C) IL-6 and (D) TNF-α production with R&D ELISA kits for each cytokine. Error bars represent means ± SEM. Statistical significance between treatment groups (P<0.05 compared to non-treated group, medium alone) is indicated by a left square bracket ([). P values were calculated by unpaired t test. NS indicates non-significant. Data are from PBMCs derived from 9-10 donors plotted individually.

Differential response to TLRLs by lymphocytes in PBMCs versus a three components proliferation system

Because of conflicting results reported in the literature that used non-PBMCs-based proliferation assay system (10, 17), we sought to determine whether the responses to TLRLs by lymphocytes in total PBMCs proliferation system, differed from lymphocytes in a three components system, in which CD3-depleted PBMCs, used as antigen-presenting cells, CFSE-labeled CD4⁺ memory effector T cells (T_eff), used as responders, and nTregs were co-cultured together. Experiments were performed as previously described and all three cell components were isolated from the same donor. Interestingly, we found that Pam₃CSK4, PGN, flagellin and CpG-A potently inhibited nTreg suppression of CD4⁺ T cell proliferation, while CL097 and CL075 were ineffective (Fig. 5). These conditions gave different results from our PBMC-based proliferation assay system where CL097 and CL075 were effective but PGN and CpG-A were ineffective, suggesting that some missing component/s of the total PBMC system might differentially modulate the lymphocyte response to TLRL.

FIGURE 5.

Identification of TLRLs that block nTreg suppression of CD4⁺ T cell proliferation using a CD3-depleted PBMC-based proliferation assay system. Experiments were performed as in Figure 1 except that CD3-depleted PBMCs were used as antigen-presenting cells and CD4⁺ effector T cells were used as responder cells. All cells were derived from the same donor. Anti-CD3 was used at a 0.3 μg/ml concentration. (A) Representative data showing the effects of TLRLs on nTreg suppression of CD4⁺ effector T cell proliferation. Each TLRL was tested at 4 different concentrations as described in Methods. White color bars and dark color bars indicate lowest and highest concentration, respectively. (B) Experiments were performed as described in Methods using three concentrations of TLRL. The ratio of nTregs to effector cells was 1:2. T cell proliferation was assessed by flow cytometry for CFSE dilution. Error bars represent means ± SEM. Statistical significance between treatment groups (P<0.05 compared to responder cells plus nTregs) is indicated by a left square bracket ([). P values were calculated by paired t test.

Pam₃CSK4 and flagellin act directly on T cells to promote proliferation

Because the majority of TLRs, with the exception of TLR9, are expressed on T cells (14, 21, 22), we asked whether these ligands could act directly on T cell subtypes to enhance TCR-mediated proliferation. Freshly sorted, naive CD4⁺ and CD8⁺ T cells were stimulated with anti-CD3 in the absence or presence of single TLRLs. After 4 days of culture, the proliferation of T cells was monitored by CFSE dilution. From the three donors of T cells tested, and in the absence of IL-2, the TLRLs Pam₃CSK4, FSL, LPS and flagellin were able to enhance anti-CD3 stimulated proliferation of naïve CD4⁺ T cells but not CD8⁺ T cells (Fig. 6A & B). Taking into account that Pam₃CSK4 and flagellin lack the undesirable septic shock side effects associated with LPS in vivo (23) and have the ability to consistently block nTreg function (Table 1), we next extended our studies to naïve, memory and regulatory CD4⁺ T cell subsets. Figure 6C & D show that, in the absence of IL-2, Pam₃CSK4 and flagellin could enhance TCR-mediated proliferation of naïve and memory CD4⁺ T cells in a dose-dependent manner, but did not enhance nTreg proliferation (data not shown), meanwhile CpG-A and CpG-B had no effect on the tested cells. However, in the presence of a high IL-2, Pam₃CSK4 and flagellin were able to enhance nTreg proliferation, whereas CpG-B decreased it (Fig. 6E). These results suggest that anti-CD3 stimulation is sufficient for Pam₃CSK4 and flagellin to enhance CD4⁺ T cell proliferation directly without the presence of APCs. In contrast, in concert with TCR stimulation, the TLR9 ligands CpG-A and CpG-B inhibited CD4⁺ T cell and nTreg proliferation in a dose-dependent manner.

FIGURE 6.

Pam₃CSK4, FSL, LPS and flagellin act directly on CD4⁺ T cells but not on CD8⁺T cells to enhance cell proliferation. (A-B) CD3-stimulated proliferation of freshly sorted, naïve CD4⁺and CD8⁺ T cells in the absence or presence of a TLRL, five concentrations tested as described in Methods. White color bars and dark color bars indicate lowest and highest concentration, respectively. (C-E) Dose - dependent enhancement of CD3-stimulated CD4⁺ naïve, memory and regulatory T cell proliferation by Pam₃CSK4 and flagellin. The TLR9Ls CpG-A and CpG-B served as negative controls. The nTregs were stimulated in the presence of a high amount of IL-2 (300 IU/ml). TLRLs were tested at five concentrations: Pam₃CSK4 and flagellin (0, 1, 10, 50, 100 ng/ml) and CpG-(A-B) (0, 0.5, 1, 2.5, 5 mM). Numbers 1 and 5 on x-axis indicate lowest and highest concentration, respectively. (A-E) ³H-thymidine was added on the third day of culture and cells were harvested after another 15 hours of incubation. Proliferation of T cells was evaluated by thymidine incorporation.Each of the data points was performed in duplicates. Data are representative from 2 donors that yielded similar results.

Pam₃CSK4 acts directly on nTregs to block suppressive activity

To evaluate whether TLRLs act directly on nTregs to inhibit their suppressive activity, we treated nTregs with each TLRL for 24h, washed out and then cultured them with autologous CFSE-labeled PBMCs in the presence of anti-CD3. After four days of culture, CD8⁺ T cell proliferation was monitored by CFSE dilution assessed by flow cytometry. We found that only Pam₃CSK4 could consistently inhibit nTreg suppressive activity on CD8⁺ T cell proliferation, whereas the eight other TLRLs tested were ineffective (Fig.7A). To investigate the mechanisms mediated by Pam₃CSK4 in reversing nTreg function, we stimulated nTregs with anti-CD3 and anti-CD28 with high amount of IL-2 and in the presence or absence of the TLRL for 7 days. The nTregs were either stained for the Treg-specific marker FOXP3 or re-stimulated with PMA/ION to detect IL-2 and IL-10 cytokine expression. We found that treatment of nTregs with either Pam₃CSK4 or control flagellin did not result in a significant change in the percentage FOXP3⁺ cells (Fig.7B). However, in the presence of Pam₃CSK4, the percentage of IL-2^-IL-10⁺nTregs was significantly reduced 3 fold compared to no treatment (Fig.7C). The decrease in the expansion of IL-10⁺ Tregs was specific to Pam₃CSK4 because flagellin was ineffective. We note that the high percentage of IL-10⁺ nTregs was the result of our cell culture system that was not observed in freshly sorted nTregs (Supplementary Fig. 2). We next determined whether Pam₃CSK4 could shut down the induction of IL-10-producing Tr1 cells from total CD4⁺ T cells. We stimulated CD4⁺ T cells with the immunosuppressive reagents vitamin D3 and dexamethasone in the absence or presence of increasing concentrations of Pam₃CSK4, flagellin or CpG-A. We found that only Pam₃CSK4 was able to shut down the induction of IL-10-producing Tr1 cells from CD4⁺ T cells in a dose-dependent manner, while CpG-A and flagellin did not show such effect (Fig.7D). These results that Pam₃CSK4 can specifically shut down IL-10-producing regulatory T cells, providing a new function of TLR1/2 stimulation pathway.

FIGURE 7.

Pam₃CSK4 directly inhibits nTreg function and shuts down IL-10-producing regulatory T cells. (A) Loss of nTreg function after Pam₃CSK4 treatment. Freshly sorted nTregs were cultured in the presence of a TLRL, 9 tested, for 24h, washed and then co-cultured with autologous CFSE-labeled PBMCs (1: 2.5, nTregs:PBMCs) in the presence of 1.5 μg/ml OKT3. CD8⁺ T cell proliferation was determined by CFSE dilution assessed by flow cytometry as described in Methods. Error bars represent means ± SEM. Statistical significance between treatment groups (P<0.05 compared to no treatment) is indicated by a left square bracket ([). P values were calculated by paired t test. (B-C) Freshly sorted nTregs were stimulated with anti-CD3 and anti-CD28 plus a high concentration of IL-2 in the presence or absence of Pam₃CSK4 or flagellin for seven days as described in Methods. Intracellular staining was then performed to evaluate the expression of FoxP3 (B) or IL-2 and IL-10 (C). Representative flow cytometry analyses are shown on upper panels. Data were taken from 4 donors. (D) Total CD4⁺ T cells were stimulated with anti-CD3 and anti-CD28 in the presence of vitamin D3, dexamethasone and ICOSL/CD32L cells. Each TLRL was tested at give increasing concentrations: Pam₃CSK4 and flagellin (0, 50, 100, 500, 1000 ng/ml); and CpG-A (0, 0.5, 1, 2.5, 5μM). Numbers 1 and 5 on x-axis indicate lowest and highest concentration, respectively. After seven days, T cells were re-stimulated with PMA/ION and cells were subjected to intracellular staining for IL-10 expression. Results are from two independent experiments, with standard deviation of the mean shown as error bars. Statistical significance between treatment groups (P<0.05 compared to Flagellin) is indicated by a left square bracket ([). P values were calculated by two-way ANOVA test.

Discussion

A variety of in vitro proliferation assay systems have been used to study the effect of TLRLs in blocking human CD4⁺ Treg suppression of effector T cell proliferation. Some systems involve the stimulation of effector T cells and CD4⁺ Tregs with soluble anti-CD3 in the presence of accessory cells such as DCs (17) or CD3-depleted PBMCs (24), while other systems include the stimulation of T cells without accessory cells but in the presence of anti-CD3/CD28-coupled beads or plate-bound anti-CD3 plus soluble anti-CD28 (21). Conflicting results have often been reported, probably due to the use of these different in vitro proliferation assays. Therefore, to clarify these ambiguous results, we believe that the use of a human total PBMCs culture system would better resemble the lymphocyte environment in vivo and yield findings relevant in the clinic. In fact, by using this PBMC-based proliferation system, we found that six TLR-specific ligands, TLR1/2L:Pam₃CSK4, TLR4L:LPS, TLR5L:flagellin, TLR7/8LCL097, TLR8/7L:CL075 and TLR9Ls:CpG-(B or C), potently inhibit Treg suppression of either CD4⁺T or CD8⁺ T cell, or B cell proliferation, and, in the case of flagellin and CL097, the three cell types simultaneously. The most remarkable difference was that LPS, CL097 and CL075 were able to block nTreg suppression of CD4⁺ T cell proliferation in the PBMC-based assay system, but not in the CD3-depleted PBMC-based (three components) assay system. Conversely, PGN and CpG-A were able to block nTreg suppression of CD4⁺ T cell proliferation only in the CD3-depleted PBMC assay system. Noteworthy, although CL097 failed to enhance CD4⁺ T cell proliferation in the PBMC-assay system, in the presence of nTregs, it was able to completely abrogate nTreg suppression of CD4⁺ and CD8⁺ T cells and B cells. We made attempts to evaluate the ability of TLRLs to enhance T-cell proliferation in the presence of CMV-specific peptides or attenuated FluA-virus. Our results suggest that the effects varied wildly among donors (Supplementary Fig. 3&4). We reasoned that the unpredictable effect could be due to the presence of a significant percentage of activated T cells in PBMCs that responded to TLRL stimulation in the absence of CMV peptides or Flu antigens. Nonetheless, taken together, our results suggest that the cell environment where T cells and nTregs become activated, determines whether a certain TLR ligand is able to block nTreg function.

Our results suggest that Pam₃CSK4 could inhibit human regulatory T-cell function by multiple mechanisms. We found that Pam₃CSK4 could: 1) block induction of IL-10⁺Tr1 cells from CD4⁺ T cells; 2) directly inhibit expansion of IL-10⁺nTregs; and 3) directly inhibit nTreg suppression of CD8⁺ T-cell proliferation. These findings are consistent with our inability to induce IL-10 cytokine secretion from PBMCs, and a recent report from Nyirenda and colleagues showing that stimulation by Pam₃CSK4 drives human naïve and effector nTregs into a Th17-like phenotype with reduced suppressive activity on CD4⁺CD25^-FOXP3^-CD45RA⁺ responder T cell proliferation (10). However, these results are in stark contrast to a recent murine study in which Pam₃CSK4 potently promoted expansion of IL-10-producing T cells (25) suggesting species specific responses to Pam₃CSK4. Interestingly, we further showed that Pam₃CSK4 could directly enhance proliferation of naïve and memory CD4⁺ T cells but not nTregs in a dose-dependent manner in the absence of IL-2, a likely scenario in the tumor microenvironment. Furthermore, both Pam₃CSK4 and flagellin could simultaneously stimulate the release of IL-6 and/or TNF-α cytokines from PBMCs. Together, these results suggest that Pam₃CSK4 or its derivatives might be useful to reinvigorate effector T cell function in humans.

Collectively, we have identified six TLR-specific ligands, Pam₃CSK4, LPS, flagellin, CL097, CL075 and CpG-(B or C) that inhibit Treg suppression of either CD4⁺ T or CD8⁺ T cell, or B cell proliferation. Our results suggest that these agents could serve as adjuvants to enhance the efficacy of current immunotherapeutic strategies in cancer patients. However, due to the different cell types present in the tumor microenvironment, one TLR agonist may trigger either positive or negative signals from different cells, making difficult to predict what the outcome might be in patients receiving TLR ligand adjuvant treatment. Therefore, further studies are needed to determine which TLR agonists, whether used alone or in combination with other reagents, are best suitable for stimulating long-term beneficial immune responses in the tumor environment, which could potentially lead to improve the cancer treatment.

Abbreviations

iTreg

induced regulatory T cell

nTreg

natural regulatory T cell

Treg

CD4⁺ regulatory T cell

mAb

monoclonal antibody

Tr1

IL-10-producing type one regulatory T cell

FOXP3

forkhead box P3

PBMC

peripheral blood mononuclear cell

T_eff

effector T cell

APC

antigen-presenting cells

TLR

Toll-like receptor