Antigen-Presenting Cell (APC) Subsets in Ovarian Cancer

Abstract

The major human antigen-presenting cells (APCs) include monocytes/macrophages, myeloid dendritic cells (mDC), plasmacytoid dendritic cells (pDC), and B cells. These APC subsets have been observed in ovarian tumor environments. Their phenotypes and functionalities are subjected to alteration by multiple factors in the tumor environment. In this review, we summarize the nature, cellular interactions, and prognostic significance of the main APC populations in ovarian cancer, and discuss the relevance of manipulating APC subsets for patient treatment.

Keywords :

• ovarian cancer
• plasmacytoid dendritic cells
• myeloid dendritic cells
• B cells
• SDF-1
• macrophages
• tumor-associated macrophages
• B7-H1
• B7-H4

INTRODUCTION

The tumor microenvironment is a complex network of innate and adaptive immune cells, each category of which encompasses both antitumor and suppressive cell subsets. How well these cellular populations work (or do not work) together determines the outcome of tumor immune responses. T cell subsets—often the “stars” of antitumor immunity—are popularly reviewed and frequently manipulated in order to improve patient prognosis. However, it is important also to examine the cell populations that affect and direct the T cells: the antigen-presenting cells (APC). Differences in activating capacity or in antigen presentation can make or break the antitumor response. We have, therefore, reviewed what is currently known about the phenotype, function, immunological, and predictive significance of the main APC populations in ovarian cancer.

MYELOID DENDRITIC CELLS

Myeloid dendritic cells (mDC) may be the most often studied of the antigen-presenting cell subsets. They play a key role in the adaptive arm of the immune system by stimulating the activation of naive T cells [1]. Pulsing of dendritic cells (DC) with killed ovarian tumor cells has been shown to effectively stimulate tumor-specific, blood-derived T cells, and these MHC-I-restricted T cells can produce IFNγ upon encountering autologous tumor cells [2]. Various other studies have been completed that demonstrate the potential of tumor antigen-pulsed DC to stimulate cytotoxic T lymphocytes (CTL) responses in vitro [3, 4]. Although adoptive transfer of appropriately primed mDC to tumor patients can engender powerful antitumor immunity [5, 6], this protection is rarely seen in the unmanipulated development of human tumors [7]. Often, the tumor or tumor environment produces factors that suppress the development and stimulatory function of DC [8, 9], which in turn undermines antitumor immunity and leads to accelerated tumor growth. In 2003, we documented low expression levels of the inhibitory molecule B7-H1 on blood- and lymph node-derived mDC in healthy individuals, but observed a striking upregulation of B7-H1 on mDC from tumor-draining lymph nodes and tumors [10] from ovarian cancer patients. In this study, B7-H1 expression on these cells was upregulated via interleukin (IL)-10 and vascular endothelial growth factor (VEGF). Interestingly, IL-10 had previously been shown to decrease costimulatory molecule expression on DC [11], while VEGF could inhibit DC differentiation from hematopoetic precursors [8]. B7-H1 blockade enhanced mDC-mediated T-cell activation, and was associated with downregulation of IL-10 and upregulation of IL-2 and interferon-γ (IFNγ) production by T cells. Interestingly, this treatment also downregulated IL-10 in mDC and stimulated an increase in IL-12 expression. Finally, T cells conditioned with the B7-H1−blocked mDC were more potent inhibitors of autologous human ovarian carcinoma growth in non-obese diabetic-severe combined immunodeficient (NOD-SCID) mice. Interestingly, in 2008, Huarte et al. demonstrated that CD11c⁺DEC205⁺ DCs coexpressing alpha-smooth muscle actin and VE-cadherin migrate to perivascular areas in ovarian carcinoma and are essential in maintaining intratumoral tumor vasculature [12]. Perhaps not surprisingly, subsequent experiments involving DC depletion in mice bearing various established ovarian cancers delayed tumor growth and enhanced chemotherapeutic effects. Decelerated tumor growth after depletion of mDC was associated with vascular apoptosis. Subsequent tumor necrosis triggered an increase in antitumor immune responses as measured by increases in tumor antigen-specific CTL populations. Our laboratory's more recent studies on the function and interactions of T-helper-17 (Th17) cells in the ovarian cancer microenvironment again focused our attention on mDC [13]. Both mDC and macrophages (but not pDC) from normal donors were capable of inducing Th17 cells from memory but not naive CD4⁺ T cells, and mDC and macrophages in the ovarian tumor microenvironment were similarly capable of inducing Th17 cells. We will discuss the relevance of Th17 induction in the next section. mDC are thought to be the major functional DC subsets in the tumor environment. Although clinical efficacy has not yet been suitably achieved, mDC vaccination has been utilized in clinical trials to treat cancer patients. However, in patients with ovarian cancer, functional mature mDC exist in limited numbers within the tumor. This fact, along with data demonstrating that mDC are phenotypically and functionally altered by tumor environments and are either dysfunctional or mediate immune suppression, supports the heretofore unsatisfying clinical observations.

MACROPHAGES

Tumor-associated macrophages (TAMs) are numerically the major APC subset in solid epithelial cancer in humans. Several years ago, our group examined regulatory T cells in the ovarian cancer microenvironment and discovered that both tumor cells and microenvironmental macrophages expressed CCL22, a chemokine instrumental in attracting regulatory T cells to the tumor environment [14]. Interestingly, because regulatory T cells predict poorer survival and are associated with a high death hazard in patients with ovarian carcinoma, tumor-associated macrophages may contribute indirectly to their prognoses. Indeed, we subsequently demonstrated that although ovarian cancer cells express high levels of the inhibitory molecule B7-H4, they do not directly mediate suppression of antitumor T cells. However, macrophages from the human ovarian tumor microenvironment express B7-H4 and are powerful suppressors of tumor-associated antigen-specific T cell immunity [15]. Blockade of B7-H4 restored the macrophages’ stimulatory capacity and mediated tumor regression in vivo in NOD/SCID mice. IL-10 and IL-6, cytokines which are often found in high concentrations in the tumor environment, induce macrophage expression of B7-H4. Contrastingly, two cytokines typically expressed at low levels in the same environment—granulocyte/macrophage colony-stimulating factor and IL-4—inhibit B7-H4 expression. Finally, we found that forced expression of B7-H4 in macrophages from healthy donors rendered them suppressive. A subsequent study investigated the prognostic significance of B7-H4⁺ macrophages in ovarian cancer, and demonstrated an inverse relationship between the intensity of B7-H4 expression on macrophages and patient survival. Importantly, Treg cells, often found to be predictors of poor prognoses in cancer patients [16], induced B7-H4 expression on myeloid APCs (including macrophages), and were associated with B7-H4⁺ macrophages in ovarian tumors [17]. This study is in line with the later observation of Wan and colleagues that the mean density of TAMs is significantly higher in ovarian cancer than in benign ovarian lesions, and that the average 5-year survival rate in patients with low densities of TAMs was significantly higher than in patients with increased TAM populations. Multivariate analysis demonstrated that TAM infiltration status is an independent negative predictor for overall survival of ovarian cancer patients [18].

The debate continues on the functionality and prognostic significance of Th17 cells in human cancer [19, 20]. Although few human studies have been completed, it seems that Th17 in established epithelial cancers, such as ovarian, act to recruit other effector T cell subsets and, in fact, support antitumor immunity [13]. As we mentioned above, both mDC and macrophages from the ovarian cancer environment are capable of Th17 induction. TAMs are more potent inducers of Th17 cells than both tumor-derived mDC and blood macrophages from healthy volunteers, and this induction was dose-dependent. Additionally, Th17 induction was dependent upon TAM expression of IL-1β and IL-23; blockade of either molecule significantly decreased resultant Th17 populations, and abrogation of both further diminished IL-17⁺ expression in CD4⁺ T cells. However, Th17 induction is suppressed by Treg cells in the same microenvironment [13]. In summary, macrophages are the largest APC subset in ovarian cancer. Ovarian cancer-associated macrophages may suppress antitumor immunity through multiple modes of action, including attraction of Treg cells and expression of inhibitory B7 family members.

PLASMACYTOID DENDRITIC CELLS (pDC)

We completed some of the first studies of plasmacytoid dendritic cells (pDC) in the tumor environment. A decade ago, we found that malignant human ovarian epithelial cells express very high levels of stromal-derived factor-1 (SDF-1), which via signaling through CXC chemokine receptor-4 (CXCR4) induced pDC tumor trafficking [9, 21]. SDF-1 also increased expression of very late antigen-5 (VLA-5) on pDCs, which interacted with VCAM-1 to mediate adhesion and cell migration through vascular endothelium. SDF-1 also served to protect pDC from apoptosis affected by TAM-derived IL-10. Tumor-associated pDCs were capable of inducing IL-10 production from bystander T cells, which contributed to poor T cell activation by local mDC. This was some of the first data showing that pDCs could undermine antitumor immunity and take part in the creation of a suppressive milieu within the tumor environment. In 2004, our laboratory demonstrated a role for pDC in promoting ovarian tumor angiogenesis [22]. SDF-1 attracted pDCs into the tumor environment, where they induced angiogenesis through production of tumor necrosis factor alpha (TNFα) and IL-8. Functional mDCs, although limited in the tumor microenvironment, could suppress angiogenesis in vivo through production of IL-12. Thus, malignant cells may attract pDCs through expression of SDF-1, to augment vessel formation, while excluding angiogenesis-inhibiting mDCs. In a study from the subsequent year, we observed that pDCs from malignant ascites were capable of inducing CD8⁺ regulatory T cell populations [23]. While tumor ascites macrophage-derived DCs induced tumor-associated antigen-specific CD8⁺ T cells with effector functions, pDCs induced suppressor cells. These cells were IL-10⁺ CCR7⁺ CD45RO⁺, and could suppress mDC-mediated tumor-associated antigen-specific T cell effector functions via their production of IL-10. CCR7 on these cells was functional, and they migrated efficiently under the influence of the lymphoid homing chemokine MIP-3β. In ovarian cancer patients, suppressive populations of CCR7⁺ CD45RO⁺ CD8⁺ T cells are found in the tumor environment, suggesting the in vivo functionality of tumor-associated pDCs. Altogether, ovarian cancer-associated pDCs induce CD8⁺ Treg cells and promote tumor angiogenesis. Thus, pDCs are detrimental to antitumor immunity in ovarian cancer.

B CELLS

As is widely known, higher numbers of tumor-infiltrating CD8⁺ T lymphocytes typically correlate with improved patient survival. B cells often colocalize with T cells and are known to provide various supportive and stimulatory functions; however, their association with patient prognoses has not been well-studied [24]. Milne et al. recently demonstrated that CD20⁺ tumor-infiltrating B cells could be found in more than 40% of high-grade serous ovarian cancers [25]. In these tissues, B cell presence was strongly associated with both CD4⁺ and CD8⁺ T cells, the T cell activation markers CD25 and CD45RO, and T cell markers indicating effector functionality including TIA-1 and granzyme B. Interestingly, B cells were also associated with T cell expression of FoxP3, a marker that indicates either activated or regulatory T cells [26, 27]. Intraepithelial presence of B cells was positively correlated with improved patient disease-specific survival (DSS), while fascinatingly, the combination of CD8⁺ and CD20⁺ tumor-infiltrating lymphocytes (TILs) in the same tumor indicated significantly increased DSS over tumors that contained only CD8⁺ or only CD20⁺ TILs. It is possible that CD20⁺ cells can support the actions of tumor-associated effector T cells through multiple mechanisms. In mice, B cells have been shown to produce autoantibodies directed against tumor targets [28]; tumor-infiltrating B cells may raise the concentration of autoantibodies in the tumor microenvironment to physiologically effective levels. Tumor-infiltrating B cells can also secrete granzyme B [29] and induce tumor cell death through TRAIL signaling [30]. New evidence of the “killer” potential of B cells is coming to light and will no doubt inform future studies of these cells in tumor microenvironments [31].

Interestingly, an earlier study examined CD19⁺ cell presence in post-chemotherapy effusions from advanced metastatic ovarian cancer and found that it was predictive of poorer survival [32]. This begs the question of how B cells can be prognostically good in one investigation of ovarian carcinoma, while detrimental in another. First, it is important to note that patients in the second study were subjected to chemotherapy (which can profoundly affect the numbers and functionality of immune cell subsets), while those in the first study were not. Second, while CD20 is present on the surface of all mature B cells [33–35], and CD19 is predominantly expressed on B cells [36], these surface markers have slightly different expression profiles. Third, the activation state of B cells has been observed to contribute to their effector or suppressor functions in various pathologies; resting B cells inhibit antitumor immunity [37], while activated B cells can facilitate T cell responses [38]. B cells are capable of enacting regulatory functions. They can be polarized by T-helper (Th) subsets into subpopulations that produce IFNγ, IL-12, and TNFα and accordingly promote Th1 development, or producers of IL-2, IL-4, TNFα, and IL-6 that support Th2 differentiation. B cell production of these groups of cytokines feeds back into maintenance and expansion of the Th populations that stimulated their cytokine expression in the first place, therefore maintaining and propagating the initial cytokine milieu [39, 40]. B cells are capable of influencing T cell memory, survival, and proliferation [41, 42], as well as presenting antigen to both CD4⁺ and CD8⁺ T cells [38, 43]. In advanced tumors, where DC may have become suppressive or rare within the tumor, B cells might take on a larger antigen (Ag)-presentation role [24]. However, this may act as a double-edged sword: presentation to helper or cytotoxic T cells may aid in antitumor immunity, while Ag presentation to regulatory T cells could lead to stronger immunosuppression of the antitumor response. B cells can also enact immunosuppressive functions through their cytokine products. IL-10 and TGFβ, both powerful immunoregulatory proteins, can be produced by B cells [43] and contribute to down-regulation of Ag presentation, suppression of T cell activation, and maintenance of suppressor function of regulatory T cells [44–46]. However, our laboratory has recently demonstrated several new roles for IL-10 in support of antitumor immunity (unpublished data), including support of the development of effector T cells, and restraint of suppressive cellular networks, including regulatory T cells and myeloid-derived suppressor cells in the tumor microenvironment. Thus, further research is required to determine whether B cell-derived IL-10 may act to suppress or support antitumor immunity. The B cells themselves may be beneficial or detrimental to antitumor immunity, depending on their phenotype within the tumor.

CONCLUSIONS

In this review, we have discussed the major APC subsets in patients with ovarian cancer. It is evident that the phenotype and function of APCs can be altered by mediators within the tumor microenvironment; in many cases, APC subsets become either dysfunctional or immunosuppressive. Therefore, targeted manipulation of tumor-associated APC populations may result in improved tumor immunity. Notably, mDC vaccination has been tested in patients with various cancers. Although we have learned important scientific rationale regarding DC vaccination, the therapeutic efficacy of these treatments has not been satisfactory. Therefore, we suggest that a combinatorial strategy targeting multiple suppressive elements within the tumor environment may be warranted for successful immune therapy.

Declaration of Interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.