Melanoma Exosomes Enable Tumor Tolerance in Lymph Nodes

Summary

Melanoma preferentially spreads via lymph nodes. Melanoma exosomes can induce angiogenesis and immune suppression. However, a role for melanoma exosomes in facilitating tumor tolerance in lymph nodes has not been considered. Herein, the hypothesis that melanoma exosome mediated induction of vascular endothelial cell (VEC) derived TNF-α results in lymphatic endothelial cell (LEC) mediated tumor tolerance is explored. To support this hypothesis, experiments involving ex vivo lymph node associated VECs, LECs, dendritic cells and T lymphocytes are proposed based upon a previously established fluorescent exosome lymph node trafficking model. The implication of the hypothesis in the context of melanoma exosome mediated induction of tumor tolerance in lymph nodes is then discussed.

Introduction

Melanoma preferentially spreads via lymph nodes (1). Melanoma cells induce angiogenesis through production of growth factors collected by tumor draining lymph nodes. Such factors include vascular endothelial growth factor (VEGF) A and C that can increase growth of traditional vascular endothelial cells (VECs) and lymphatic endothelial cells (LECs). VEGF-C increases the lymphatic flow to the lymph node promoting tumor invasion and delivery of tumor antigens (2). Perpetual delivery of tumor antigens to the node may increase peripheral tolerance to tumor cells. LEC's can induce peripheral tolerance through expression of programmed cell death ligand 1 (PD-L1) in the absence of appropriate T lymphocyte co-stimulatory molecules such as CD80 or CD86 (3). This may also implicate LECs in bolstering melanoma mediated immune tolerance. Melanoma cells can also express PD-L1 resulting in suppression of anti-tumor T-cell activation through programmed death receptor 1 (PD-1) signaling (4).

In recent years, evidence has been accumulating that an additional mechanism for facilitating tumor growth and survival is the production and secretion of tumor exosomes (5-10). Exosomes are naturally occurring biological nanovesicles averaging approximately 100 nm in size depending on their cellular source (11). They relay signals between cells and deliver information rich lipid, protein and RNA cargo (12-14). Similar to soluble mediators, tumor exosomes relay pathogenic processes both locally and distally (14, 15). Distal tumor exosome activity is likely correlated to increased metastatic potential as supported by evidence demonstrating an increase in circulating tumor exosome burden with tumor stage (16, 17). Other distal effects include melanoma exosome mediated re-programming of bone marrow progenitor cells toward a pro-vascular phenotype (17).

Previously we and others discovered that melanoma exosomes directly induce angiogenesis (15, 18). They can also contain pro-angiogenic and immunomodulatory factors (18) and prepare lymph nodes for tumor metastasis (5). Given their participation in melanoma pathogenesis, it is reasonable to propose that melanoma exosomes might be used as tumor vaccines given their capacity to relay tumor antigens between tumors and antigen presenting dendritic cells (DCs). Earlier studies demonstrated that tumor exosomes can harbor tumor rejection antigens capable of driving T cell mediated cross protection against syngeneic and allogeneic tumors in vivo (19). However, in recent years, the potential use of tumor exosomes directly as vaccines has been complicated by their inherent propensity to suppress the immune system. Melanoma exosomes have been shown to suppress natural killer and cytotoxic CD8+ lymphocyte anti-tumor responses (20) and induce MDSCs (21). MDSC's can make anti-tumor cytotoxic CD8+ lymphocytes non-responsive (22, 23) or polarize tumor associated macrophages toward an immunosuppressive M2-like phenotype (24). Alternatively, DC derived exosomes expressing functional MHC-1 tumor peptide complexes have proven more efficacious as tumor vaccines (25). Functional antigen presenting DC derived exosomes can persist in lymph node microenvironments well after DC's themselves have exited the nodes (26). This demonstrates that the presence and function of antigen presenting exosomes in lymph nodes is not dependent on the presence of their source antigen presenting cells. It follows that this would be true for melanoma exosomes in melanoma draining lymph nodes as well. It raises the intriguing possibility that an additional mechanism for melanoma exosome mediated preparation of pre-metastatic niches in lymph nodes beyond angiogenic or direct immunosuppressive processes might be facilitating tolerance to tumor antigens.

Presentation of the Hypothesis

Melanoma exosome dependent tumor tolerance might be mediated through traditional immunological mechanisms. In a recent study, plasma derived MHC II+ exosomes, presumably derived from antigen presenting cells, were able to mediate tumor tolerance (27). The authors suggested that the mechanism of action may be tumor antigen presentation by MCH II+ exosomes to regulatory T cells. Additionally, given their ability to directly interact with endothelial cells to induce angiogenesis (15, 18) and traffic to lymph nodes (5, 28), melanoma exosomes may also mediate tumor tolerance in lymph nodes through endothelial cells.

It is well established that TNF-α is expressed by macrophages and lymphocytes (29). TNF-α is an inflammatory cytokine that upregulates expression of leukocyte ligands such as intercellular adhesion molecule 1 (ICAM-1) on endothelial cells allowing for transmigration of leukocytes into tissues. However, endothelial cells can also produce TNF-α (30). Endothelial production of TNF-α may have autocrine, juxtacrine or paracrine effects within the context of the lymph node environment. One conceivable juxtacrine or paracrine effect might be VEC mediated TNF-α stimulation of neighboring or local LECs. Given the role of LECs in facilitating tumor spread, it is reasonable to hypothesize that LECs may also participate in tumor tolerance (31).

Previous studies demonstrate that TNF-α enables LEC participation in normal peripheral immune tolerance (32). TNF-α stimulated LECs influence the maturation of dendritic cells (DCs) by suppressing expression of dendritic cell CD86 co-receptors through interactions between ICAM-1 and macrophages 1 antigen (MAC-1). This inhibits the ability of DCs to induce T-cell proliferation in response to self-antigens to prevent autoimmunity. Unfortunately melanoma cells and exosomes are also composed of self-antigens. As a result, melanoma cells may potentially exploit this normal process to enable their survival. In other studies, melanoma exosomes were able to increase VEC expression of TNF-α in 3D cell culture assays (15) and lymph nodes in vivo (5). Combined these investigations support the hypothesis that melanoma exosome mediated induction of vascular endothelial cell (VEC) derived TNF-α results in lymphatic endothelial cell (LEC) mediated tumor tolerance (Fig. 1).

Figure 1.

Melanoma exosomes mediate tumor immune tolerance through endothelial cells in lymph nodes. Melanoma exosomes deliver tumor antigens to dendritic cells (DCs)(19) and induce vascular endothelial cell (VEC) production of tumor necrosis factor alpha (TNF-α) (5, 15). TNF-α stimulates interactions between lymphatic endothelial cells (LECs) and DCs through intercellular adhesion molecule 1 (ICAM-1) and macrophage 1 antigen (MAC-1) (32). This results in decreased DC surface expression of the CD86 co-receptor. Interactions between the DC MHC-1 receptor and the T-lymphocyte T cell receptor (TCR) in the absence of CD86 interactions with the T-cell CD28 co-receptor generate unresponsive cytotoxic CD8+ anti-tumor T-cells resulting in tolerance to tumor antigens.

Evaluation of the Hypothesis

Testing this hypothesis will be challenging. As mentioned previously a variety of lymph node cell types can express TNF-α. Moreover, melanoma exosomes can mediate immune suppression through a variety of mechanisms that may confound LEC mediated tumor tolerance experiments. Nevertheless, it is possible to begin to determine whether melanoma exosome mediated tumor tolerance through LECs is possible by focusing on VEC, LEC, DC and T-cell subsets. Based on the hypothetical model (Fig. 1), a good starting point will be to determine whether melanoma exosomes interact with VECs in lymph nodes. This can be accomplished using our previously established exosome lymph node trafficking model (5, 28). Fluorescent melanoma exosomes can be tracked to lymph node endothelial cells in mice, nodes extracted, digested with collagenase and exosome positive cells sorted by fluorescent activated cell sorting (FACS) using antibodies to surface receptors such as VCAM-1 and LYVE-1 to distinguish VECs from LECs. Once sorted, subsequent investigations will be necessary to determine whether exosome targeted VECs express TNF-α. Given the short half-life of cytokines, this might best be done using RT-RT pcr to quantify TNF-α mRNA expression in VECs. Western blot analysis can be used to further validate whether lymph node LECs previously exposed to melanoma exosomes in vivo upregulate surface expression of ICAM-1.

Co-culture experiments containing ex vivo lymph node derived VECs and LECs can be used to determine whether melanoma exosomes increase TNF-α expression in the culture supernatant as well as LEC surface expression of ICAM-1. ICAM-1 expression levels can be determined using Western blot, RT-RT pcr, flow cytometry or a combination. Additional, sequential co-culture experiments might be used to determine whether LEC's derived from VEC/LEC co-cultures containing melanoma exosomes reduce CD86 co-receptor expression on MAC-1+ DCs obtained from mouse collagenase digested lymph nodes. Confirmatory flow cytometry experiments using ex vivo lymph nodes previously treated with melanoma exosomes will validate decreased expression of CD86 co-receptors on DCs.

Additional experiments might be performed using tumor exosomes derived from B16-F10 melanoma cells expressing the highly immunogenic ovalbumin (OVA) antigen (33). B16/OVA exosomes have been used in other tumor antigen based immunosuppressive experiments (34). DC populations obtained from B16/OVA melanoma exosome treated lymph nodes might be used in subsequent antigen cross presentation investigations (35). Such assays can be used to determine whether lymph node DC cell populations obtained from B16/OVA exosome treated C57BL/6 non-transgenic control mice expand OT-1 T lymphocytes derived from OT-1 mice expressing the transgenic T cell receptor for ovalbumin. T-cell proliferation assays (CFSE and flow cytometry) and supernatant IL-2 levels might be used as readouts for OT-1 T cell expansion.

Further, induction of endothelium by melanoma exosomes is not likely limited to VECs but includes LECs as well. Experimental variations for the above proposed experiments will be required to determine whether melanoma exosomes directly induce LEC derived TNF-α. Alternatively, similar experiments can be used to determine the extent to which other cell types such as lymph node associated macrophages and lymphocytes contribute to the pool of melanoma exosome induced TNF- α in lymph nodes.

Consequences of the Hypothesis and Discussion

During tumor pathogenesis, a melanoma exosome concentration gradient would presumably move from high to low with increasing distance from the tumor microenvironment. Distally in tumor draining sentinel lymph nodes where the concentration of melanoma exosomes is initially less, there might exist a fragile homeostasis where pro-tumor immunogenic processes have not yet overcome anti-tumor processes and an immunosuppressive MDSC presence is still lacking or limited. Within this homeostatic immune microenvironment, some melanoma exosomes and their associated antigens might be directly sampled by DCs and macrophages for antigen presentation in an effort to activate cytotoxic anti-tumor immunity. However, while this process is occurring, a second more insidious process is also occurring in the background mediated through melanoma exosome interactions with LECs. As the premetastatic lymph node microenvironment is continuously bombarded with an increasing concentration of tumor draining VEGF-C and melanoma antigens carried on or in melanoma exosomes, LECs are simultaneously proliferating and responding to TNF-α derived from melanoma exosome induced VECs. This results in slow induction of tumor tolerance by LECs via interactions with DCs. Importantly, melanoma exosome mediated tumor tolerance in lymph nodes might be comprised of both MDSC (22, 23) and LEC mediated tumor tolerance mechanisms. Effectively, in combination with direct suppression of anti-tumor immune cell functions, this contributes to a loss of tumor immune surveillance in lymph nodes. This process may be applicable to other tumor types as well.