Cavity macrophages stop anti-tumor T cells

Abstract

Precisely how macrophages regulate response to immunotherapy remains unclear. In this issue of Cancer Cell, Chow et al. demonstrate a tumor-promoting role of TIM-4⁺ cavity macrophages affecting the response of metastatic malignancies to immunotherapy. Specifically, TIM-4⁺ macrophages engage with phosphatidylserine-expressing cytotoxic T lymphocytes, inhibiting their activity even during PD-1 blockade.

The wealth of clinical data on patient responses to anti-PD1 therapy has opened up new avenues for identifying response correlates beyond the traditional focus on T cell infiltration and tumor mutational burden (Jardim et al., 2021). Chow et al. (this issue) identify the inhibitory role TIM-4⁺ cavity-resident macrophages in the activity of effector T cells and in the responses of pleural or peritoneal metastatic tumors to immunotherapy. The investigators begin by validating previous findings in Non-small cell Lung Cancer (NSCLC) highlighting that the response rate to anti-PD1 is substantially lower in patients with pleural or peritoneal metastasis, with a significant reduction in both progression-free and overall survival (Osorio et al., 2019). These correlations are also present in multi-institutional cohorts of patients with metastatic NSCLC or colorectal cancer. Only metastasis to the bone and liver are similarly predictive of poor survival, hinting that poor patient responses may be associated with a unique immunosuppressive microenvironment within these select tissues.

The immune composition of the pleural and peritoneal cavities is dominated by macrophages – cells with an established role at directly and indirectly suppressing T cell responses in cancer (DeNardo and Ruffell, 2019). Resident macrophages within these tissues express high levels of TIM-4; a receptor for phosphatidylserine (PS) that promotes efferocytosis and can suppresses anti-tumor immunity (Baghdadi et al., 2013). After confirming the expression pattern of TIM-4 in human and murine cavity-resident macrophages, but not in tumor-associated macrophages or monocytes, Chow et al. establish a model of peritoneal carcinomatosis to evaluate the role of TIM-4⁺ macrophages in response to PD-1 blockade. They show that these mice respond fairly well to PD-1 blockade alone, with 50% of animals surviving past 100 days of being challenged with the direct injection of MC38 colon carcinoma cells. However, mice survival increases to over 70% upon the addition of a TIM-4 blocking antibody, or when the studies are conducted with TIM-4-deficient animals. TIM-4 is even more relevant in models of resistant disease. Mice injected with CT26 colon carcinoma cells or the Kras^G12Dp53^–/– lung cancer cell line are non-responsive to anti-PD-1 or anti-TIM-4, but gain a survival advantage when the combination of antibodies is employed.

Reduced tumor burden and improved survival following dual TIM-4/PD-1 blockade is associated with higher numbers of peritoneal CD8⁺ T cells. How does TIM-4 expression by macrophages control the T cell response? Interfering with PS-dependent efferocytosis can lead to expression of type I interferon through activation of the STING pathway in macrophages (Cunha et al., 2018; Zhou et al., 2020), but Chow et al. find no evidence of this after performing RNA sequencing on sorted peritoneal macrophages. Instead, they notice that a high number of T cell-selective transcripts are detected within the TIM-4⁺ population, and that the level of these transcripts is substantially reduced following TIM-4 blockade. These results hint that TIM-4⁺ macrophages are adhering to T cells within the peritoneal cavity, a phenotype that could be recapitulated in vitro by adding activated T cells together with peritoneal macrophages. Importantly, this physical connection is not associated with efferocytosis, as evidenced by a lack of signal when T cells labeled with a pH-sensitive dye are added into the peritoneal cavity.

Although PS exposure is normally associated with the early stages of apoptosis, cytotoxic lymphocytes have also been reported to expose PS on the cell surface as a protective measure to inactivate perforin and protect against self-killing (Rudd-Schmidt et al., 2019). Chow et al. perform bulk and single cell RNA sequencing on sorted CD8⁺ T cells with high or low PS exposure and observe that viable PS^high cells are enriched for genes associated with cytotoxicity, activation, and proliferation. The connection between PS exposure and effector function is maintained even after focusing on T cells expressing TCR sequences associated with an antigen-specific response to MC38. PS^highCD8⁺ T cells are also more prevalent in the peritoneal cavity during dual treatment with anti-PD-1/TIM-4, or when anti-PD-1 is administered to TIM-4 knockout mice. In vitro, PS^highCD8⁺ T cells are found to bind to peritoneal macrophages in a TIM-4-dependent manner, while the same is not true for PS^lowCD8⁺ T cells. These results support a model where TIM-4⁺ macrophages sequester viable PS^high effector T cells in the peritoneal cavity and limit their ability to sustain a state of activation, proliferation, and cytotoxicity.

Macrophages utilize a number of different mechanisms to suppress T cell activation and effector function, and are potent suppressors in T cell proliferation assays. Chow et al. demonstrate that human peritoneal and pleural macrophages are capable of preventing viable CD8⁺ T cells from proliferating in response to anti-CD3/CD28 stimulation, with suppression also observed for murine peritoneal macrophages. It remains an open question how these cavity-resident macrophages are able to mediate this suppression. Chow et al. do observe that human and mouse peritoneal macrophages express CD39, an ectonucleoside that converts ATP to AMP and leads to the production of adenosine. However, inhibition of CD39 with sodium metatungstate only partially reverse macrophage-dependent suppression, depending on the patient sample. It also remains possible that the in vivo impact of TIM-4 blockade relates to functional changes in macrophages. TIM-4 activation in macrophages restricts the presentation of phagocytosed tumor cell antigens and promotes AMP-activated protein kinase activation (Baghdadi et al., 2013; Cunha et al., 2018). In potential agreement, the authors show that anti-TIM-4 reduced the expression of several immunosuppressive genes in peritoneal macrophages (e.g. Ptgs2, Tgfb2, and Tdo2).

Therapeutic targeting of macrophages has traditionally focused on reducing tumor infiltration through inhibition of recruitment or survival pathways (DeNardo and Ruffell, 2019). However, these approaches have less impact on resident macrophage populations and may fail to take advantage of potential anti-tumor functions for recruited cells. A new generation of agents is now focused on altering macrophage polarization or blocking their regulatory activity within tumors. For example, antibodies against TREM2 have shown promising activity in preclinical models (Molgora et al., 2020) and dose escalation trials in combination with Pembrolizumab are ongoing (NCT04691375).

In summary, Chow et al. identify a role for TIM-4⁺ macrophages in the suppression of T cell responses in the peritoneal cavity. Antibodies against TIM-4 do not deplete macrophages, but impair their ability to suppress T cells, resulting in combinatorial efficacy in peritoneal carcinomatosis models. Based on these data, it will be important to determine if anti-TIM-4 can augment immunotherapy against primary malignancies in the serous body cavities, such as ovarian cancer, where TIM-4⁺ tissue-resident macrophages are implicated in cancer progression (Casanova-Acebes et al., 2020; Etzerodt et al., 2020). Additionally, the role TIM-4⁺ macrophages in other anatomic compartments prone to metastasis, such as the bone and liver, merits evaluation. This is especially true given the presence of TIM-4⁺ macrophages in these organs, as well as the association of metastasis to these sites with lower response rates and reduced survival of NSCLC patients during anti-PD-1 immunotherapy.

Figure 1: Cavity-resident macrophages inhibit CD8⁺ cytotoxic T cells via TIM-4.

Macrophages within serous body cavities express high levels of TIM-4, a receptor for phosphatidylserine (PS). T cell activation and cytotoxic effector function is associated with exposure of PS on the cell surface, resulting in macrophage binding and limited CD8⁺ T cell proliferation in mice with peritoneal carcinomatosis, even during treatment with anti-PD-1. Genetic loss or blockade of TIM-4 reduces binding of macrophages to PS-expressing cytotoxic T cells, resulting in T cell expansion and extended survival during anti-PD-1 immunotherapy in preclinical models of peritoneal metastasis.