The Unique Molecular and Cellular Microenvironment of Ovarian Cancer

Abstract

The reciprocal interplay of cancer cells and host cells is an indispensable prerequisite for tumor growth and progression. Cells of both the innate and adaptive immune system, in particular tumor-associated macrophages (TAMs) and T cells, as well as cancer-associated fibroblasts enter into a malicious liaison with tumor cells to create a tumor-promoting and immunosuppressive tumor microenvironment (TME). Ovarian cancer, the most lethal of all gynecological malignancies, is characterized by a unique TME that enables specific and efficient metastatic routes, impairs immune surveillance, and mediates therapy resistance. A characteristic feature of the ovarian cancer TME is the role of resident host cells, in particular activated mesothelial cells, which line the peritoneal cavity in huge numbers, as well as adipocytes of the omentum, the preferred site of metastatic lesions. Another crucial factor is the peritoneal fluid, which enables the transcoelomic spread of tumor cells to other pelvic and peritoneal organs, and occurs at more advanced stages as a malignancy-associated effusion. This ascites is rich in tumor-promoting soluble factors, extracellular vesicles and detached cancer cells as well as large numbers of T cells, TAMs, and other host cells, which cooperate with resident host cells to support tumor progression and immune evasion. In this review, we summarize and discuss our current knowledge of the cellular and molecular interactions that govern this interplay with a focus on signaling networks formed by cytokines, lipids, and extracellular vesicles; the pathophysiologial roles of TAMs and T cells; the mechanism of transcoelomic metastasis; and the cell type selective processing of signals from the TME.

Keywords: NFκB; STAT; T cell; T cell checkpoints; adhesion; ascites; extracellular vesicles; invasion; mesothelial cell; metastasis; tumor microenvironment; tumor-associated macrophage.

Figure 1

Signaling network between tumor cells and tumor-associated macrophages (TAMs) in ovarian cancer ascites. Prominent examples of signaling pathways operating between tumor cells (red) and TAMs (blue). Genes for receptors and their cognate ligands are ordered in adjacent blocks on the left and right sides, respectively, of each box. The sizes of the filled squares indicate the level of expression determined by RNA-Seq (large: median TPM > 50; intermediate: TPM > 10; and small: TPM > 0.3) (52). Open squares indicate cases, where substantial expression (TPM > 3) was observed only in a fraction of samples (>10%).

Figure 1

Signaling network between tumor cells and tumor-associated macrophages (TAMs) in ovarian cancer ascites. Prominent examples of signaling pathways operating between tumor cells (red) and TAMs (blue). Genes for receptors and their cognate ligands are ordered in adjacent blocks on the left and right sides, respectively, of each box. The sizes of the filled squares indicate the level of expression determined by RNA-Seq (large: median TPM > 50; intermediate: TPM > 10; and small: TPM > 0.3) (52). Open squares indicate cases, where substantial expression (TPM > 3) was observed only in a fraction of samples (>10%).

Figure 2

Functions of extracellular microvesicles (EVs) in ovarian cancer ascites. EVs are released by virtually all cell types of the tumor microenvironment and shape cellular functions of both tumor cells and host cells via different pathways. Depicted examples affect major hallmarks of ovarian cancer to promote tumor growth and metastasis. A, adipocyte; F, fibroblast; M, mesothelial cell; Mph, macrophage; NK, natural killer cell; T, T cell; Tu, tumor cell.

Figure 3

Functions/dysfunctions of tumor-associated macrophages (TAMs) in the ovarian cancer tumor microenvironment (TME) and examples of major mediators of these functions. Multiple mediators (gray box) in the TME as well as the lack of interferon (IFN)γ determine the activation state and function of TAMs. In response to these triggers, TAMs produce a plethora of soluble factors impinging on tumor cells and other host cells and block the expression of essential antitumor mediators such as interleukin (IL)-12.

Figure 4

Ovarian cancer environment and the CD8⁺ T cell versus T regulatory cells (Tregs) balance. Signals provided by cytokines present in ovarian carcinoma ascites [interleukin (IL)-6, IL-10] as well as by dendritic cells induce exhaustion of CD8⁺ T cells. Cells coexpressing the inhibitory receptors, programmed cells death protein 1 (PD-1) and lymphocyte activation gene 3 (LAG3), display the strongest impairment in interferon (IFN)γ and TNFα production. Presumably, high antigen concentrations in the ascites cause CD8⁺ T-cell overstimulation leading to CD8⁺ T cell exhaustion. Moreover, the migration of CD8⁺ T cells into the tumor environment is diminished due to impaired expression of the T helper 1-associated chemokines CXCL9 and CXCL10. Conversely, the preferential migration of immunosuppressive FOXP3⁺CTLA4⁺, GITR⁺CCR4⁺ Tregs is supported by the upregulated chemokine CCL22.

Figure 5

Ovarian cancer cell adhesion and invasion. Ovarian cancer cells in the ascites adhere to the mesothelium via integrins and CD44. This adhesion is further enhanced by matrix metalloproteinase (MMP)-mediated cleavage of fibronectin on mesothelial cells. Ovarian cancer cells then break mesothelial junctions to invade through the mesothelial cell layer into the submesothelial extracellular matrix (ECM), where they proliferate and form metastases.

Figure 6

Molecular signaling network in the ascites. In addition to cancer cells, the ascites contains several other cell types including tumor-associated macrophages (TAMs), cancer-associated adipocytes (CAAs), cancer-associated fibroblasts (CAFs), and T lymphocytes. These cells communicate via multiple signaling molecules that promote the metastasis of ovarian cancer cells.

Figure 7

A simplified view of the cell type-specific impact of cytokine signaling in the ovarian cancer tumor microenvironment. The simultaneous presence of pro- and anti-inflammatory mediators suggests that biological outcome depends on both prevalence and interpretation of signals. Left: tumor cells respond with persistent activation of the transcription factors RELA (p65) and STAT3, leading to promotion of growth, progression, and poor clinical outcome. Right: cells of the immune system do not receive appropriate combinations of stimuli for pro-inflammatory activation and therefore do not activate REL and STAT1. Integration of the signaling molecules present in ascites leads to an anti-inflammatory state, at least in part through activation of STAT3.

Figure 8

Transcriptional circuitry governing the immunesuppressive state of macrophages and T cells in the ovarian cancer tumor microenvironment (TME). In tumor-associated macrophages (TAMs), interleukin (IL)-10 activates STAT3, which in turn stimulates IL10 transcription. Both cooperate to repress IL12B transcription and REL nuclear translocation by multiple mechanisms, reinforced by GPCR ligands and nuclear receptor (NR) agonists such as glucocorticoids and polyunsaturated fatty acids. In T cells, REL plays an ambivalent role by fostering both IL2 and FOXP3 transcription. The latter depends on the presence of TGFβ and results in conversion into a T regulatory cell phenotype. In the ovarian TME or ascites, TNFR, toll-like receptor (TLR), T cell receptor (TCR), and CD28-activating stimuli are overruled; without long-term pro-inflammatory stimulation, upregulation of IFNG transcription is prevented. As a consequence, pro-inflammatory signals fail to activate STAT1, induction of IL12B, and subsequent sustained production of interferonγ. Dot-shaped nodes denote a requirement for both signals (logical AND).