Tumor-induced senescent T cells promote the secretion of pro-inflammatory cytokines and angiogenic factors by human monocytes/macrophages through a mechanism that involves Tim-3 and CD40L

Solid tumors are infiltrated by immune cells where macrophages and senescent T cells are highly represented. Within the tumor microenvironment, a cross-talk between the infiltrating cells may occur conditioning the characteristic of the in situ immune response. Our previous work showed that tumors induce senescence of T cells, which are powerful suppressors of lympho-proliferation. In this study, we report that Tumor-Induced Senescent (TIS)-T cells may also modulate monocyte activation. To gain insight into this interaction, CD4+ or CD8+TIS-T or control-T cells were co-incubated with autologous monocytes under inflammatory conditions. After co-culture with CD4+ or CD8+TIS-T cells, CD14+ monocytes/macrophages (Mo/Ma) exhibit a higher expression of CD16+ cells and a reduced expression of CD206. These Mo/Ma produce nitric oxide and reactive oxygen species; however, TIS-T cells do not modify phagocyte capacity of Mo/Ma. TIS-T modulated-Mo/Ma show a higher production of pro-inflammatory cytokines (TNF, IL-1β and IL-6) and angiogenic factors (MMP-9, VEGF-A and IL-8) and a lower IL-10 and IP-10 secretion than monocytes co-cultured with controls. The mediator(s) present in the supernatant of TIS-T cell/monocyte-macrophage co-cultures promote(s) tubulogenesis and tumor-cell survival. Monocyte-modulation induced by TIS-T cells requires cell-to-cell contact. Although CD4+ shows different behavior from CD8+TIS-T cells, blocking mAbs against T-cell immunoglobulin and mucin protein 3 and CD40 ligand reduce pro-inflammatory cytokines and angiogenic factors production, indicating that these molecules are involved in monocyte/macrophage modulation by TIS-T cells. Our results revealed a novel role for TIS-T cells in human monocyte/macrophage modulation, which may have deleterious consequences for tumor progression. This modulation should be considered to best tailor the immunotherapy against cancer.Clinical and experimental studies have established that several types of solid tumors are characterized by infiltration of both innate and adaptive immune cells. Indeed, it has been reported that tumors can be infiltrated by different cell populations such as B cells, NK cells, Th1 and Th2 cells, regulatory T cells (Tregs), senescent T cells and macrophages, among others.^{1, 2, 3} Investigating the nature and effector function of these tumor-infiltrating subsets is highly relevant as accumulating evidence indicates that a dynamic cross-talk between tumors and immune system cells can regulate tumor growth and metastasis.^{4, 5}Macrophages constitute a major component of the leukocytes that infiltrate tumors. Tumor-associated macrophages (TAMs) derive almost entirely from circulating monocytes, which acquire distinct phenotypic characteristics and diverse functions according to the tumor microenvironment. Prototypically, two different types of activated macrophages have been recognized: the classically activated (M1) or pro-inflammatory macrophages and the alternatively activated (M2) macrophages. Thus, in response to diverse signals like cytokines or membrane receptor ligation, macrophages undergo M1 or M2 polarization states characterized by particular profiles of cytokine and chemokine production. M1 macrophages express high levels of pro-inflammatory cytokines, major histocompatibility complex (MHC) molecules and inducible nitric oxide (NO) sintetase. By contrast, M2 macrophages downregulate MHC class II and show increased expression of the anti-inflammatory cytokine IL-10 and mannose receptor. In addition, macrophages can also be polarized into a M2-like state, which shares some but not all the signature features of M2 cells.¹ Macrophages with intermediate or overlapping phenotypes have been observed in many pathological conditions in vivo, probably as a result of the effect of diverse signals that occur along different times of the immune response. In fact, many studies emphasize the heterogeneity and plasticity of macrophages and indicate that typical M1 and M2 phenotypes are extremes of a wide spectrum of functional states.^{6, 7, 8}Within the tumor microenvironment, the macrophages interact with or receive signals from different tumor-infiltrating immune cells such as Tregs, myeloid-derived suppressor cells, Th1 and Th2 cells, among others.^{1, 9} This interaction may regulate the profile of macrophage activation and consequently impact on tumor progression. It has been described that the excessive activity of either M1 or M2 subsets may be detrimental to the host by preventing the development of an efficient anti-tumor immune response.¹⁰ Understanding the cellular interactions that lead to the control of monocyte/macrophage (Mo/Ma) activation is, therefore, of fundamental importance to the field of tumor immunology.Senescent T cells are reported to be increased during chronic infections and some tumor processes.^{11, 12} In fact, senescent T cells circulate in the peripheral blood of most cancer patients and infiltrate tumors.² Although the hallmark of human senescent T cells is the loss of CD27 and CD28 expression, other features of these cells include shortened telomeres, reduced proliferative capacity and cytokine production as well as suppressor activity.^{13, 14} We previously reported that, after a brief contact with tumor cells, CD4+ as well as CD8+ T lymphocytes from healthy donors acquire a senescent phenotype. These CD4+ and CD8+ tumor-induced senescent (TIS)-T cells are characterized by the loss of CD27 and CD28 expression, lack of proliferative capacity, telomere shortening and increment in the expression of senescence-associated molecules such as p53, p21 and p16. Remarkably, these CD4+ and CD8+ TIS-T populations also show a potent suppressive ability.¹⁵ We also demonstrated that tumor-induced senescence of T cells is triggered by soluble factors secreted by tumor cells and that this process can be prevented by IL-7.¹⁶ These data support the hypothesis that, within the tumor microenvironment, tumor-infiltrating T lymphocytes encounter tumor cells that promote their senescence and dysfunction. These TIS-T cells would be able to suppress the lympho-proliferative response and potentially modulate other immune cells. Thus, they may serve as an intercellular cross-talk in the tumor microenvironment and impact on tumor progression.Although macrophages and TIS-T lymphocytes are both highly represented in tumors, the biological consequences of a TIS-T cell and macrophage interaction have not been studied so far. Here, we demonstrate that monocytes co-cultured with TIS-T cells in inflammatory conditions increase their production of inflammatory cytokines and angiogenic factors. In addition, we determined that Mo/Ma modulation mediated by TIS-T lymphocytes requires cell-to-cell contact and identified T-cell immunoglobulin and mucin protein 3 (Tim-3) and CD40 ligand (CD40L) molecules as mediators of this previously uncharacterized modulatory function of senescent T cells.