The PD-1 Interactome

Abstract

T cell activation is a fine-tuned process that involves T cell receptor and costimulation signals. To prevent undue activation of T cells, inhibitory molecules including PD-1 (programmed death 1) are induced and function as brakes for T cell signaling. In a steady state, the interaction of PD-1 with its ligands PD-L1 (B7-H1, CD274) and PD-L2 (B7-DC, CD273) maintains peripheral immune tolerance. However, the expression of PD-L1 on tumor cells and interaction with PD-1 on T cells dampen anti-tumor immunity. Therapeutic inhibitors of the PD-1 pathway have revolutionized tumor immunotherapy. Unfortunately, the majority of patients do not develop sustained anti-tumor responses. However, the knowledge about unique PD-1 interactions and their role in mediating PD-1 inhibitory signals is currently limited. Advances in the mechanistic understanding of the molecular and signaling integration of the PD-1 pathway could unleash the great potential in tumor immunotherapy by allowing the development of combinatorial approaches that target not only PD-1 and its ligands but also its unique downstream signal mediators. In this review, the current advances in understanding the mechanisms of extracellular and intracellular PD-1 interactions and their significance in potential future therapeutic approaches are discussed.

Keywords: SHP-1; SHP-2; T cell activation; programmed death-1 interactome; programmed death-1 pathway.

Figure 1.

Interaction network of PD-1 with its ligands. A) The classic trans interaction of PD-1 on T cells and PD-L1 on APC reduces T cells activation via inhibition of TCR and CD28 signaling (left panel). PD-1 is also expressed on innate immune cells, in which the cis interaction of PD-1/PD-L1 might result in less free PD-L1 to bind PD-1 on T cells, thus prevent PD-1 inhibitory signaling on T cells (middle panel). As PD-L1 is also expressed on T cells and is further upregulated after T cell activation, the cis interaction of PD-1/PD-L1 might take place on T cells after T cell activation (right panel). This cis interaction of PD-1/PD-L1 and self-inhibition on T cells might prevent a continuous activation state. B) PD-L1 is highly expressed on T cells, thus the bidirectional signal of PD-L1 may change the functional fate of T cells in the tumor microenvironment. In this setting, PD-1 might act as a ligand for PD-L1 and transmit an inhibitory signal to T cells and macrophages in the tumor tissue. The backward singling of PD-L1 might shift CD4⁺ T cells from Th1 to Th17 and limit cytotoxic effects of CD8⁺ T cells in the tumor. The forward signaling of PD-L1 changes inflammatory macrophages (M1) to pro-tumor (M2) macrophages. In addition, forward signaling of PD-L1 via PD-1 can reduce production of inflammatory cytokines and effector cytotoxic T lymphocytes (CTL). Overall, the bidirectionality of PD-L1 signaling dampens immune responses against tumor. GzmB, Granzyme B. C) T cell activation requires a first signal from the TCR and second co-stimulatory signal from CD28 (CD28/B7–1). To counteract this process, inhibitory signals from CTLA-4 (CTLA-4/B7–1) and PD-1 (PD-1/PD-L1) act as brakes for further activation of T cells. CTLA-4 mediates its inhibitory function via endocytosis after binding with its ligand B7–1. The trans interaction of PD-L1 and B7–1 was previously reported to inhibit T cell activation. D) The cis interaction of PD-L1 and B7–1 on both APC and T cells. In co-culture of T cell and APC, the cis interaction of PD-L1 and B7–1 blocks the trans interaction of PD-L1/PD-1, thus making PD-L1 less available on APC for binding PD-1 on T cells, resulting in improved T cell activation. However, this relies on more expression of B7–1 than PD-L1 on APC cells. Interestingly, the cis interaction of PD-L1/B7–1 has no effects on the trans B7–1 binding to CTLA-4 and CD28. E) When PD-L1 outnumbers B7–1 on APC cells, the cis interaction of PD-L1 and B7–1 might have limited effects on trans interaction of PD-L1/PD-1. The excess of PD-L1 on APC could still bind to PD-1 and transmit inhibitory signals to T cells. F) When PD-L1 outnumbers B7–1 on APC cells, blocking cis interaction of PD-L1/B7–1 and trans interaction of PD-L1/PD-1 by anti-PD-L1 can further activate T cells. However, in this case, the released B7–1 on APC might bind to CTLA-4 on T cells and cause their inhibition. Thus, combination of anti-PD-L1 and anti-CTLA-4 under these conditions could maximize T cell activation and may be an efficient approach to improve tumor immunotherapy.

Figure 2.

PD-1 signaling and interactors. A) T cell activation requires first signal from the TCR by engagement with peptide-loaded MHC. The recruited Src-family kinases, Lck and Fyn by TCR signaling rapidly phosphorylate ITIM and ITSM tyrosines of the cytoplasmic tail of PD-1 (pPD-1). Both SHP-2 and SHP-1 could be recruited to pPD-1, and promote downstream inhibitory signaling. These phosphatases can counteract positive signaling from TCR and CD28, which includes inhibition of ZAP70, SLP-76, protein kinase C (PKC-θ), phosphatidylinositol-3-kinase (PI3K), and the Ras signaling pathway. B) The self-inhibitory structure of SHP-2 binds to pITIM and pTISM of PD-1. SHP-2 contains two SH2 domains, N-terminal (N-SH2) and C-terminal (C-SH2), followed by a PTP domain. After C-SH2 binds to pITSM domain of PD-1, N-SH2 is released from the autoinhibitory conformation and binds to pITIM tyrosine of PD-1. By using a synthetic peptide (GST-PD-1 cytoplasmic tail) as bait to affinity purify candidate proteins in activated human T cells, signaling lymphocytic activation molecule-associated protein (SAP) was discovered in the PD-1 interactome. SAP blocked PD-1 inhibitory functions in T cells through indirect inhibition of SHP-2 activity. SAP did not show direct association with SHP-2 and might interact through a still unknown adaptor protein to inhibit PD-1 signaling. C) SHP-2 outcompetes SHP-1 for binding to phosphorylated PD-1 under physiologic conditions and inhibits both TCR and CD28 signaling. In the absence of SHP-1, SHP-2 could still inhibit T cell activation by downregulation of pSLP76, pLAT, pCD28, and expression of CD69. Interestingly, SHP-1 could mediate PD-1 inhibitory signaling in the absence of SHP-2. Double deletion of SHP-1 and SHP-2 on T cells showed complete loss of PD-1-mediated inhibition of CD28 and TCR signaling. D) Since pITSM of PD-1 could bind to N-SH2 domain of SHP-2, SHP-2 could bridge two pITSM motifs of PD-1 molecules through its N-SH2 and C-SH2 domains, which leads to dimerization of PD-1. In this model, the increased PTP activity of SHP-2 may inhibit TCR or/and CD28 signaling.