As a tumor grows, it interacts constantly with objects in the surrounding environment, such as blood vessels, signaling molecules and immune cells. Communication between these entities is a two-way street; tumor cells can influence the behavior of other cells in the tumor microenvironment and vice versa. While in some cases the immune system recognizes cancer cells as foreign and works to eliminate them, in other contexts immune cells can actually promote tumor survival. For example, tumor-associated macrophages have been shown to accelerate cancer progression and metastasis, despite their canonical role in protecting the body from harmful agents. Understanding the interactions between macrophages and tumor cells will thus be important in developing strategies to inhibit the spread of cancer.
Interactions between tumor cells and other cells in the tumor microenvironment have historically been studied using in vitro co-culture assays or fixed tumor sections. However, “cancer is a dynamic process,” says Dr. Minna Roh-Johnson, a former postdoctoral fellow in the Moens Laboratory (Basic Sciences Division) who now runs her own lab at the University of Utah. With the advent of high-resolution imaging techniques, it is now possible to observe cell-cell interactions as they happen in live organisms. In work recently published in Developmental Cell and led by Dr. Roh-Johnson, the Moens lab and their collaborators used high-resolution fluorescence microscopy to investigate how macrophages influence the migratory behavior of tumor cells. They focused on metastatic melanoma, a particularly deadly cancer type, hoping to shed light on why melanomas often develop resistance to immune-based therapies.
To perform their experiments, the researchers transplanted human melanoma cells into larval zebrafish via injection into the hindbrain ventricle, a component of the central nervous system. Larval zebrafish are transparent and thus amenable to imaging, and introduction into the hindbrain ventricle allows tumor cells to spread into skin, the organ in which melanoma naturally occurs. Dr. Roh-Johnson and her colleagues began by testing six different metastatic human melanoma cell lines for their ability to disseminate into distal sites within transplanted zebrafish. Dissemination efficiency varied from 15-70% among the cell lines tested, but the metastatic cell lines all migrated better than non-metastatic cell lines, confirming that dissemination is specific to metastatic cells.
Next, the researchers sought to characterize the interactions between human melanoma cells and resident zebrafish macrophages labeled with green or red fluorescent proteins, respectively. The macrophages were observed extending long, thin protrusions that contacted the melanoma cells for prolonged periods of time (Figure 1). To confirm that this macrophage behavior was specific to tumor cells and not due to activation of the innate immune system by transplantation, the authors showed that macrophages acted similarly toward endogenous zebrafish skin cells engineered to express a cancer-driving protein. Further, the authors demonstrated that host macrophages promote tumor cell dissemination by showing that reduced macrophage abundance via knockdown of a macrophage specification gene concomitantly reduced migration of melanoma cells.
Communication between macrophages and tumor cells could theoretically be achieved either by diffusion of signaling molecules or through direct exchange of cytoplasmic contents. Because it was recently shown that migratory melanoma cells can induce metastatic behaviors in other melanoma cells via cytoplasmic transfer, the authors hypothesized that macrophages might similarly promote dissemination of melanoma cells by transferring cytoplasm. To test this idea, they set up a Cre recombinase-based reporter assay in which a zebrafish line engineered to express the Cre recombinase only in macrophages was transplanted with human melanoma cells expressing a specially designed fluorescence cassette (Figure 2A). The cassette contains two genes in tandem that encode red and green fluorescent proteins, respectively, but only the red gene is expressed. However, the sequences flanking the red gene are targets of Cre recombinase, leading to removal of the red gene and expression of the green fluorescent protein upon exposure to Cre. Thus, if a zebrafish macrophage exchanges cytoplasm with a human melanoma cell in this system, the cancer cell will change from red to green (Figure 2B).
Thirteen percent of tumor cells became green within four days after transplantation into zebrafish with Cre-expressing macrophages, compared to only 2% in control zebrafish. This result reveals that macrophages do in fact transfer their cytoplasmic contents to tumor cells. In addition, cytoplasmic transfer by macrophages promotes metastatic behavior because green cells had a greater tendency to disseminate than red cells.
There are two known mechanisms for cells to exchange cytoplasm: secretion of extracellular vesicles or cell-to-cell contact. Due to the prolonged physical interactions observed between macrophages and melanoma cells, the authors hypothesized that cytoplasmic transfer occurs via direct contact. To test this idea, they assessed the ability of Cre recombinase-expressing macrophages to induce a red-to-green color switch in melanoma cells in vitro when the two cell types were either co-cultured or separated by a membrane that prevents cell-to-cell contact but allows diffusion of extracellular vesicles. Consistent with their hypothesis, the researchers found that more tumor cells turned green when co-cultured with Cre-expressing macrophages compared to when the cells were separated.
Because their previous results did not determine whether cytoplasmic transfer occurred before or after dissemination, the authors sought to examine this process over time. They observed that most animals with green cells on day 4 (post-dissemination) also had green cells on day 1 (pre-dissemination), suggesting that cytoplasmic transfer may precede, and possibly trigger, migration. In agreement with this idea, tumor cells that received macrophage cytoplasm had a higher tendency to travel in a consistent direction during migration, although they did not move faster than cells that did not experience cytoplasmic transfer.
The researchers confirmed that the cytoplasmic transfer phenomenon also occurs in mammals by performing analogous experiments in mice; 91% of cells that metastasized to mouse lungs had turned from red to green, compared to 8% in control mice. Together, these results suggest that cytoplasmic exchange between macrophages and tumor cells is broadly relevant in cancer. Says Dr. Roh-Johnson, “we think that these findings suggest a novel method of how cells in the microenvironment can communicate with tumor cells to drive invasive behavior – a process that we could potentially inhibit in terms of therapeutics or perhaps even take advantage of to deliver therapeutics.”
Roh-Johnson M, Shah AN, Stonick JA, Poudel KR, Kargl J, Yang GH, di Martino J, Hernandez RE, Gast CE, Zarour LR, Antoku S, McGarry H, Bravo-Cordero JJ, Wong MH, Condeelis J, and Moens CB. Macrophage-Dependent Cytoplasmic Transfer during Melanoma Invasion In Vivo. Developmental Cell. 2017 Dec 4;43:549-562. doi: 10.1016/j.devcel.2017.11.003
This research was supported by the National Institutes of Health, Fondation ARC pour la Recherche sur le Cancer, European Commission, OHSU Center for Women’s Health, Crohn’s and Colitis Foundation, and Prospect Creek Foundation.