PD-1 Primarily Targets TCR Signal in the Inhibition of Functional T Cell Activation

Abstract

Cancer-immunotherapy targeting programmed cell death 1 (PD-1) activates tumor-specific T cells and provides clinical benefits in various cancers. However, the molecular basis of PD-1 function is still enigmatic. Especially, it is unclear which signaling pathway PD-1 primarily targets. Besides, the capacity of PD-1 to inhibit the T cell receptor (TCR)-dependent activation of T cells in the presence of co-stimulation is also controversial. Here we used co-culture systems of T cells and antigen-presenting cells with targeted deletion and overexpression of co-receptors and ligands and examined the inhibitory potency of PD-1 against T cell activation upon TCR stimulation with CD28 and ICOS co-stimulation. As an unambiguous criterion of T cell activation, we used the acquisition of cytokine production capacity, which represents one of the most important functions of T cells. PD-1 inhibited functional T cell activation upon TCR stimulation in the absence as well as in the presence of CD28 co-stimulation, indicating that PD-1 can directly inhibit TCR signal. Notably, CD28 co-stimulation rather attenuated the efficiency of PD-1 in inhibiting TCR-dependent functional T cell activation. In addition, PD-1 inhibited TCR-dependent functional T cell activation with ICOS co-stimulation as efficiently as that with CD28 co-stimulation. Furthermore, we found that the maintenance of antigen-induced follicular helper T (T_FH) cells that required ICOS co-stimulation was persistently restrained by PD-1 in vivo. These findings indicate that PD-1 primarily targets TCR signal in the inhibition of functional T cell activation. Thus, PD-1 functions as the rheostat of T cell activation rather than an inhibitor of a specific stimulatory co-receptor.

Keywords: CD28; ICOS; PD-1; T cell receptor; co-receptor; cytokine production; follicular helper T cell.

Figure 1

PD-1 inhibited the antigen-dependent functional activation of DO11.10 T cells less efficiently in the presence of CD28 co-stimulation. (A) Schematic representations of the antigen-dependent activation of DO11.10 T cells with or without CD28 engagement. (B) Expression levels of indicated co-receptors and ligands. (C) Inhibition of antigen-dependent activation of DO11.10 T cells by PD-1 engagement. IL-2 secretion from DO11.10 T cells in the absence (white) or presence (gray) of PD-1 engagement by PD-L1 on APCs. Anti-PD-L1 Ab completely blocked PD-1 effect (black). (D) Titration of anti-PD-L1 blocking Ab. (E) Expression levels of CD86 on IIA1.6 cells expressing CD86 to varying degrees. (F) Antigen-dependent activation of DO11.10 T cells in the absence of CD28 co-stimulation and the enhancement of the activation in a manner dependent on the amount of CD86 on APCs. (G) Correlation between the amount of secreted IL-2 and the expression level of CD86 on APCs. (H–J) Robust PD-1-mediated inhibition of IL-2 production from DO11.10 T cells in the absence CD28 co-stimulation and the partial attenuation of PD-1-mediated inhibitory effect by CD28 co-stimulation. IL-2 secretion from DO11.10 T cells upon stimulation with pOVA_323−339-pulsed APCs lacking (left, black and gray) or expressing (right, red and pink) CD86 in the presence (gray and pink) or absence (black and red) of PD-1 engagement (H). The average percent PD-1-dependent inhibition of IL-2 production upon stimulation with indicated APCs pulsed with 0.3, 1, and 3 μM of pOVA_323−339 (I). The percent PD-1-dependent inhibition is plotted in relation to the amount of IL-2 production in the absence of PD-1 engagement for indicated APCs (J). Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of more than two independent experiments. *p < 0.05 and **p < 0.01 by one-way ANOVA with Tukey HSD test. Cells used in this figure are listed in Tables 2, 3.

Figure 2

PD-1 inhibited the antigen-dependent functional activation of DO11.10 T cells in the absence of CD28 co-stimulation. (A) Expression levels of CD28 on DO11.10 and DO11.10-CD28KO T cells. (B) PD-1-dependent inhibition of the antigen-dependent functional activation of DO11.10 T cells without CD28 co-stimulation. IL-2 secretion from DO11.10-CD28KO T cells upon stimulation with pOVA_323−339-pulsed IIAdL1-CD86KO cells in the presence (gray) or absence (black) of PD-1 engagement. Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of three independent experiments. Cells used in this figure are listed in Tables 2, 3.

Figure 3

PD-1 inhibited the TCR-dependent functional activation of primary T cells less efficiently in the presence of CD28 co-stimulation. (A) Schematic representations of the TCR-dependent activation of primary T cells. Pre-activated CD4⁺ and CD8⁺ T cells were stimulated with anti-CD3 Ab presented on APCs with or without CD86 expression. (B,C) Expression levels of CD28 and PD-1 on CD4⁺ (B) and CD8⁺ (C) T cells with (lower) or without (upper) pre-activation. (D–G) Robust PD-1-mediated inhibition of cytokine production from primary T cells in the absence CD28 co-stimulation and the partial attenuation of PD-1-mediated inhibitory effect by CD28 co-stimulation. IL-2 and IFNγ secretions from pre-activated primary CD4⁺ (D) and CD8⁺ (E) T cells, respectively upon TCR stimulation with no (left, black and gray), weak (middle, green and olive), and strong (right, red and pink) CD28 co-stimulation. PD-1-dependent inhibitory effect was evaluated by comparing amounts of cytokines in the presence (gray, olive, and pink) or absence (black, green, and red) of PD-1 engagement. The average percent PD-1-dependent inhibition of cytokine production upon stimulation with indicated APCs pulsed with 0.08, 0.4, 2, and 10 μg/ml of anti-CD3 Ab is shown for pre-activated primary CD4⁺ (F) and CD8⁺ (G) T cells. Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of more than two independent experiments. **p < 0.01 by one-way ANOVA with Tukey HSD test. Cells used in this figure are listed in Tables 2, 3.

Figure 4

Stimulation of primary T cells using CH27 B lymphoma cells as APCs. (A–C) Expression of CD80 (A), CD86 (B), and PD-L1 (C) on CH27 cells. CH27 cells endogenously express CD80, CD86, and PD-L1. Genes for CD80, CD86, and PD-L1 were knocked-out in CH27TKO cells. (D–G) Robust PD-1-mediated inhibition of cytokine production from primary T cells in the absence of CD28 co-stimulation and the partial attenuation of PD-1-mediated inhibitory effect by CD28 co-stimulation. IL-2 and IFNγ secretion from pre-activated primary CD4⁺ (D) and CD8⁺ (E) T cells, respectively upon TCR-stimulation in the absence (left, black and gray) or presence (right, red and pink) of CD28 engagement. PD-1-dependent inhibitory effect was evaluated by comparing amounts of cytokines in the presence (gray and pink) or absence (black and red) of PD-1 engagement. The average percent PD-1-dependent inhibition of cytokine production upon stimulation with indicated APCs pulsed with 2 and 10 μg/ml of anti-CD3 Ab is shown for pre-activated primary CD4⁺ (F) and CD8⁺ (G) T cells. (H,I) PD-1-dependent inhibition of the antigen-dependent functional activation of 2B4.11 T cells without CD28 co-stimulation. CH27TKO cells with (right, red and pink) or without (left, black and gray) CD86 were pulsed with indicated amounts of pMCC and used to stimulate 2B4.11 T cells. IL-2 secretion in the presence (gray and pink) or absence (black and red) PD-1 engagement is shown (H). The percent PD-1-dependent inhibition denotes the average inhibitory effects with 0.3 and 1 μM of pMCC (I). Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of two independent experiments. *p < 0.05 and **p < 0.01 by two-tailed unpaired Student's t-test. Cells used in this figure are listed in Tables 2, 3.

Figure 5

ICOS engagement augmented the antigen-dependent functional activation of DO11.10 T cells in the presence of PKCθ. (A) Expression levels of ICOS and ICOSL. (B) Increased antigen-dependent secretion of IL-2 from DO11.10-ICOS T cells by ICOS engagement in the presence of PKCθ. ICOS engagement increased the IL-2 secretion from DO11.10-ICOS T cells overexpressing PKCθ (right) but not from DO11.10-ICOS T cells lacking PKCθ expression (left). (C) Lack of Prkcq (PKCθ) expression in DO11.10 T cells. The expression levels of Prkcq, Zap70, and Lat in primary CD4⁺ and CD8⁺ T cells and DO11.10 T cells with or without PKCθ overexpression were quantified by qPCR. Fold changes in expression relative to primary CD4⁺ T cells are shown. Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of two independent experiments. Cells used in this figure are listed in Tables 2, 3.

Figure 6

PD-1 efficiently inhibited the functional activation of DO11.10 and primary T cells upon TCR stimulation with ICOS co-stimulation. (A,B) PD-1-dependent inhibition of the functional activation of DO11.10 T cells upon antigen stimulation with ICOS co-stimulation. IL-2 secretion from DO11.10 T cells upon stimulation with pOVA_323−339-pulsed APCs lacking (left, black and gray) or expressing (right, red and pink) ICOSL in the presence (gray and pink) or absence (black and red) of PD-1 engagement (A). The average percent PD-1-dependent inhibition of IL-2 production upon stimulation with indicated APCs pulsed with 0.3, 1, and 3 μM of pOVA_323−339 (B). (C,D) Expression levels of ICOS and PD-1 on CD4⁺ (C) and CD8⁺ (D) T cells with (lower) or without (upper) pre-activation. (E–H) PD-1-dependent inhibition of the functional activation of primary T cells upon TCR stimulation with ICOS co-stimulation. IL-2 and IFNγ secretion from pre-activated primary CD4⁺ (E) and CD8⁺ (F) T cells, respectively upon TCR stimulation in the absence (left, black and gray) or presence (right, red and pink) of ICOS co-stimulation. PD-1-dependent inhibitory effect was evaluated by comparing amounts of cytokines in the presence (gray and pink) or absence (black and red) of PD-1 engagement. The average percent PD-1-dependent inhibition of cytokine production upon stimulation with indicated APCs pulsed with 2 and 10 μg/ml of anti-CD3 Ab is shown for pre-activated primary CD4⁺ (G) and CD8⁺ (H) T cells. Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of more than two independent experiments. Cells used in this figure are listed in Tables 2, 3.

Figure 7

PD-1 restricted the maintenance of antigen-induced T_FH cells that required ICOS co-stimulation. (A) Schematic representation of the experimental system to analyze PD-1 effects on the maintenance of antigen-induced T_FH cells. Anti-ICOSL and anti-PD-L1 Abs were administrated independently or in combination on days 6 and 8 to analyze their effects on the maintenance of antigen-induced T_FH cells. CD4⁺ T cells in draining lymph nodes were analyzed on day 9. (B–E) Increase of T_FH cells by PD-1 blockade during their maintenance phase. Representative flowcytometric profiles (B,C) and statistics (D,E) of T_FH cells identified by the expression of CXCR5 and PD-1 (B,D) and CXCR5 and Bcl6 (C,E) are shown for mice with the indicated treatment. (F–H) PD-1 blockade facilitated the down-regulation of Klf2 (F) and up-regulation of Il4 (G) and Il21 (H) upon TCR stimulation with ICOS co-stimulation. Pre-activated primary CD4⁺ T cells were stimulated with anti-CD3 Ab (2C11, 0.1 μg/ml) in the indicated combination of ICOS and PD-1 engagements for 6 h. T cells were purified and the expression of Klf2, Il4, and Il21 was quantified by qPCR. Data are the mean ± SEM of technical triplicates in one experiment. Data are representative of more than two independent experiments. *p < 0.05 and **p < 0.01 by one-way ANOVA with Tukey HSD test. Cells used in this figure are listed in Tables 2, 3.