PD-1⁺ T_reg cells: a foe in cancer immunotherapy?

Abstract

The mechanisms that drive responses to PD-1-blocking immunotherapy in some but not all patients have been puzzling. A new study suggests that the balance of PD-1 expression levels between CD8⁺ T cells and T_reg cells might provide an answer.

PD-1 is an inhibitory receptor expressed by activated T cells. The discovery that PD-1 ligands are expressed in the tumor microenvironment (TME) has led to the use of antibodies that block PD-1 or its ligands (for example, PD-L1) to induce antitumor T cell responses and has revolutionized cancer immunotherapy. Despite considerable progress, durable clinical responses are observed in only a small fraction of patients, but the reasons remain poorly understood. In this issue of Nature Immunology, Kumagai et al. demonstrate for the first time that the balance of PD-1 expression between CD8⁺ effector T (T_eff) and regulatory T (T_reg) cells in the TME might be utilized as a clinically meaningful biomarker to predict the efficacy of PD-1-blocking immunotherapy in various cancers¹.

Checkpoint immunotherapy is increasingly used for the treatment of cancers that do not respond to traditional therapies. Blocking inhibitory PD-1 signaling enables antigen-specific CD8⁺ T cells to respond to tumor antigens and become activated, ultimately reducing the tumor size. The predictability of clinical responses on the basis of reliable biomarkers remains a major question and a central subject of research efforts. Kumagai et al. examined determinants of responses in patients with non-small cell lung cancer (NSCLC), gastric cancer (GC) and malignant melanoma (MM) treated with PD-1 blockade monotherapy using anti-PD-1 (nivolumab or pembrolizumab) or anti-PD-1 ligand (atezolizumab). Despite a favorable trend in therapeutic responses in patients with PD-L1-positive tumors, several patients with PD-L1-positive tumors did not develop any responses, while other patients with PD-L1-negative tumors responded to PD-1 blockade therapies. These findings suggest that PD-L1 positivity alone is not sufficient to predict clinical responses to PD-1 blockade, and more precise markers are needed. To achieve this, Kumagai et. al. used tumor-infiltrating lymphocytes (TILs) to perform a combined analysis involving flow cytometry and next-generation sequencing. This method allowed them to identify the significance of the balance of PD-1 expression between CD8⁺ T cells and T_reg cells.

In humans, FoxP3⁺CD4⁺ T_reg cells represent a heterogeneous population that can be subdivided into three functionally and phenotypically distinct subsets²; thymus-derived naive T_reg cells (tT_reg, CD45RA⁺FoxP3^lo) and activated effector T_reg cells (eT_reg, CD45RA⁻FoxP3^hi) are highly suppressive, while CD45RA⁻FoxP3^lo T_reg cells represent a non-T_reg cell population with the potential for inflammatory cytokine production. Among these T_reg cell subsets, eT_reg cells are the predominant FoxP3⁺ T_reg cells infiltrating the TME in most cancers³. Kumagai et al. found no significant differences in the ratio of tumor-infiltrating CD8⁺ T cells to eT_reg cells between patients with cancer who responded to PD-1 blocking immunotherapy and those who did not. However, when they looked at the expression of PD-1 on these T cell populations, they found that tumor samples from non-responders exhibited a higher frequency of PD-1⁺ eT_reg cells. They examined the functional outcome of PD-1 expression by eT_reg cells and the specific impact of blocking PD-1 signals by in vitro suppression assays using TILs from NSCLC samples. When PD-1^lo T_reg cells were cultured with PD-1^hi CD8⁺ T cells, PD-1 blockade enhanced the proliferation of CD8⁺ T cells. Conversely, when PD-1^hi T_reg cells were cultured with PD-1^lo CD8⁺ T cells, blocking the PD-1 signals increased the suppressive activity of T_reg cells, preventing the proliferation of PD-1^lo CD8⁺ T cells. PD-1-blocking immunotherapy induced activated eT_reg cells in TILs, as determined by upregulation of the receptors CTLA4, GITR and ICOS. Additional studies showed that PD-1 blockade enhanced T cell receptor (TCR) and CD28-mediated signaling in mouse CD8⁺ T cells and T_reg cells to similar extents, as determined by activation of the kinases ZAP-70 and Akt.

These results highlight a previously unappreciated role of PD-1 signaling in T_reg cells and suggest that PD-1 blockade enhances the suppressive activity of T_reg cells that express high levels of PD-1. Thus, when T cells express high levels of PD-1, PD-1 blockade leads to their activation and enhances their physiological function. While blocking PD-1 on PD-1⁺CD8⁺ T cells converts them to CD8⁺ T_eff cells that have potent effector function, leading to tumor regression, blocking PD-1 on PD-1⁺ T_reg cells converts them to activated eT_reg cells that have potent suppressor functions, leading to tumor progression (Fig. 1). To validate the role of PD-1 expression in T_reg cells on the outcome of cancer immunotherapy that was determined using patients’ samples, Kumagai et al. examined how the ratio of PD-1 expression in T_reg cells versus CD8⁺ T cells shapes antitumor responses in the MC38 mouse colon tumor model. When wild-type CD8⁺ T cells and PD-1-deficient Pdcd1^–/– T_reg cells were adoptively transferred into tumor-bearing mice, PD-1 blockade improved antitumor responses. By contrast, PD-1 blockade increased the tumor burden when Pdcd1^–/– CD8⁺ T cells and wild-type T_reg cells were adoptively transferred, suggesting that inhibition of PD-1 signaling resulted in an increase in the suppressive activity of PD-1⁺ T_reg cells. On the basis of these findings, the authors developed and validated a mathematical formula that incorporates the balance of PD-1 expression between CD8⁺ T_eff cells and eT_reg cells in the TME to predict the clinical efficacy of PD-1-blocking immunotherapy. This is a potentially important advancement in the field of cancer immunotherapy and a step forward in identifying patients that might benefit from PD-1-based immunotherapy. However, several questions arise regarding the effects of PD-1 on the function of T_reg cells in the context of cancer and other conditions in which T_reg cells have important roles.

Fig. 1 |. Levels of PD-1 expression by T_reg and CD8⁺ T cells determine the efficacy of PD-1 immunotherapy.

When T cells express high PD-1 levels, PD-1 blockade leads to their activation and enhances their physiological function. a, When PD-1 is predominantly expressed in CD8⁺ T cells in the Tme, blocking PD-1 converts PD-1⁺CD8⁺ T cells into CD8⁺ T_eff cells with potent effector function, leading to tumor regression. b, Conversely, when PD-1 is predominantly expressed by T_reg cells in the Tme, blocking PD-1 converts them into activated eT_reg cells with potent suppressor function, leading to tumor progression. Credit: Debbie maizels/Springer Nature

It is well established that tT_reg cells are activated after exposure to self-antigens expressed in peripheral tissues, as they interact with MHC class II–bound peptides from peripheral self-antigens substantially more efficiently than CD25^– T cells⁴. As self-antigens are abundant in the TME, tumor-associated T_reg cells may receive continuous antigenic stimulation, leading to high PD-1 expression. Under these conditions, PD-1-blocking immunotherapy may activate T_reg cells in addition to activating exhausted T cells that reside in tumors, as reported by Kumagai et al. Previous studies have assessed the role of T_reg cells in the tolerance mechanisms mediated by the PD-1 pathway in autoimmune conditions. In non-obese diabetic mice, PD-1–PD-L1 blockade reversed tolerance and caused autoimmunity by promoting proliferation and inflammatory cytokine production by antigen-specific T cells infiltrating target tissues without affecting the activity of T_reg cells⁵. Moreover, in a model of autoimmune pancreatitis that is induced by partial FoxP3 insufficiency and PD-1 deficiency, the transfer of T_reg cells rescues the autoimmune phenotype regardless of the presence or absence of PD-1 in T_reg cells⁶. Although these studies suggest that PD-1 does not have a key role in T_reg cell function, it should be noted that PD-1 ligation synergizes with the immunomodulatory cytokine TGF-β to induce FoxP3⁺ iT_reg cells with potent suppressive capacity⁷. Moreover, PD-1 promotes fatty acid oxidation⁸, a metabolic program that supports the differentiation, survival and function of T_reg cells⁹. On the basis of these previous studies, one would anticipate that PD-1 signaling would promote T_reg cell generation and suppressor function, whereas PD-1 blockade would have the opposite effect.

An additional unexpected finding by Kumagai et. al. is related to the signaling changes that PD-1 blockade induced in T_reg cells concurrently with enhancing activation and suppressor capacity. Addition of a blocking PD-1 antibody during T_reg cell activation via CD3 and CD28 resulted in enhanced phosphorylation of ZAP-70 and Akt, similar to what is observed in non-T_reg cells. It has been previously established that inhibition of Akt and its downstream target mTOR is required to induce T_reg cell differentiation and sustain T_reg cell suppressor function^10,11. Thus, Akt activation by PD-1 blockade would be expected to destabilize T_reg cells and diminish their suppressive function. However, in the context of cancer, inhibition of the kinase PI3K might decrease the number and impair the immunosuppressive function of T_reg cells¹².

Together, these studies raise the intriguing hypothesis that PD-1-mediated effects in T_reg cells may be altered by unique cues from distinct microenvironments, such as the presence of distinct soluble factors or types of antigen-presenting cells. Clearly, additional work is required to determine the conditions under which PD-1-mediated signaling preferentially promotes T_reg cell generation and suppression capacity or inhibits activation of T_reg cells. Identifying how PD-1 imprints such distinct T_reg cell fates may have important consequences for cancer immunotherapy, as this may guide appropriate patient selection and predict clinical responses, as indicated by the study of Kumagai et al. This knowledge may also have implications for transplantation tolerance and autoimmune diseases, wherein T_reg cells have key physiological roles for maintaining immune homeostasis.