Skip to main content

Flow in the tumor microenvironment: Linking mechanobiology with immunology

Final Report Summary - LYMPHIMMUNE (Flow in the tumor microenvironment: Linking mechanobiology with immunology)

Tumor-associated lymphangiogenesis has been correlated with poor prognosis for many cancers, including metastases. This project aimed to elucidate the role of lymphangiogenesis and flow induced tissue property changes in the immunological response of two such tumor types using murine models: breast and skin cancer (melanoma).

Lymphangiogenesis, the new growth of lymphatic vessels, is for one, driven by vascular endothelial growth factor (VEGF)-C, signaling through the VEGF-receptor 3, amongst other factor/receptor combinations. We explored the effect of lymphangiogenesis, either by over-expressing VEGF-C in our tumor models, through the release of VEGF-C from synthetic gels, or blocking lymphangiogenesis by administering a functionally blocking antibody against VEGFR3.

Our project has shown that one of the important roles of expanding lymphatics in these tumor models, and our synthetic VEGF-C releasing gel contexts is the recruitment and accumulation of immune infiltrates. We further obtained support for this concept by analysing the data from The Cancer Genome Atlas (TCGA). From this data, we saw that VEGF-C expression in primary cutaneous melanoma or in the serum of patients are indicative of primary tumor inflammation, and CCL21 and its receptor CCR7, which attracts naïve T cells and dendritic cells correlate with VEGF-C. This was also reflected in our murine models. In breast cancer, VEGF-C levels correlate with TGFb1. It was seen in our murine model, that lymphatic endothelial cells (LECs) also contribute to TGFb1 in the tumor microenvironment.
Tumor stromal changes were not always correlating with metastasis in our mouse models. While tumor progression in our model MMTV-PyMT corresponded with increased stromal stiffening, it was difficult to recapitulate what we observe in the clinic, i.e. tumor metastasis in the draining lymph node in the mouse. On the other hand, the growth kinetics of the B16-F10 orthotopic melanoma model, we believe, does not allow sufficient time for stromal remodeling. We have, however, achieved in the mouse, a new breast cancer model wherein MMTV-PyMT cells transduced to express high levels of VEGF-C metastasized to the tumor draining lymph node.

While lymphatic vessels have traditionally been seen as conduits for metastatic cells to drain to the draining lymph node, our project has clearly shown that LECs scavenge peripheral antigens and can present these antigens to T cells, with the outcome that the CD8+ T cells have a central memory phenotype. We have shown that these central memory-like CD8+ T cells can be reactivated under immunogenic conditions. This also confirms that LECs themselves, or CD8+ T cells educated by LECs as targets for immunotherapy.

We have also made in vitro models that can add insight to the role of LECs as well as fibroblasts and immune cells in tumor metastasis in the lymph node. Specifically, our microfluidic device models the lymph node cortex, and we have demonstrated the proof of principal that by activating CD8+ T cells in the artificial lymph node cortex, the tumor cells are less invasive into this space.

Our work, therefore argues that lymphangiogenesis, while an indicator and possibly a route for metastasis, can be harnessed for anti-tumor immunotherapies.