Cellular Senescence Contributes to the Outcome of Lymphoma Chemotherapy by Cell-Autonomous and Paracrine Immune Mechanisms

Introduction: Premature senescence reflects an acutely inducible, irreversible growth arrest as a cellular response to stresses such as oncogenic activation and DNA damage, including chemotherapeutic anticancer agents. Senescence complements apoptosis as a tumor suppressive and therapeutic effector principle, but whether a selective disruption of the senescence machinery impairs treatment outcome is unknown. Moreover, function and fate of senescent tumor cells within the tumor site remain unclear. Here, we analyze the impact of defined genetic alterations, i.e. Bcl2 overexpression (blocking apoptosis), deletion of the histone H3 lysine 9 methyltransferase Suv39h1 (controlling senescence), and conditional expression of p53 (mediating both apoptosis and senescence), on therapy-induced senescence (TIS) in the Eµ-myc mouse lymphoma model with specific emphasis on immunological tumor-host and growth-modulating senescent/non-senescent cell interactions as a consequence of TIS in vitro and in vivo. Methods: Lymphoma cells (LCs) of various genetic backgrounds were retrovirally transduced with the bcl2 gene to study TIS in the absence of drug-induced apoptosis. Bcl2-protected LCs were treated with the DNA damaging anticancer agent adriamycin in vitro, or were exposed to the alkylating agent cyclophosphamide upon lymphoma formation in normal immunocompetent mice in vivo. TIS was detected by staining for senescence-associated β-galactosidase activity (SA-β-gal) and other senescence-related markers, including Ki67 and BrdU incorporation. To study tumor-host cell interactions, isolated normal splenocytes were co-incubated with proliferating or senescent LCs in vitro. Immunophenotyping was carried out with antibodies specific for macrophages, granulocytes, natural killer cells and T-lymphocytes. Cytokine production was measured by protein arrays. Results: Senescent LCs engage in cell-cell interactions with different immune cell subsets, in particular macrophages, granulocytes and T-cells in vitro. Fluorescence microscopy reveals that macrophages engulf LCs after they entered TIS. In vivo, TIS correlates with the quantitative attraction of immune cell populations to the tumor site and subsequent clearing of senescent cells. Ongoing mechanistic studies on underlying ligand/receptor interactions will be reported at the meeting. TIS cells exhibit a specific pro-inflammatory secretory profile whose functional impact on tumor and bystander cells is currently being investigated. Importantly, this profile is distinguishable from cytokine profiles of senescence-compromised Suv39h1- or p53-deficient lymphomas, and, thus, reflects a senescence- rather than a DNA damage-associated secretory response. Discussion: The study unveil a functional interaction of senescent LCs with different immune cell subsets in vitro and in vivo. The cytokine arrays show that senescent cells produce a specific secretory profile, which might stimulate immune cell attraction. Therefore, immune cells could be recruited to lymphomas in vivo specifically after TIS with the potential to clear senescent – and possibly non-senescent – cells from the tumor site. The data demonstrate genetically that senescence is a beneficial effector principle of DNA damaging chemotherapy and encourage further exploration of this program to limit cancer expansion in vivo.