Dissecting forces in tumor evolution using zebrafish models
The laboratory currently has multiple projects that aim to understand how heterogeneity is generated in the tumor cell population, and how the tumor microenvironment influences those heterogeneous populations. Some specific areas of focus include:
Melanoma is amongst the most plastic of tumors, with each individual cell capable of existing on a spectrum between highly undifferentiated to deeply pigmented and differentiated. Using our zebrafish models, we discovered that success at metastatic sites is associated with a “phenotype switch” in which the microenvironment produces factors which promote this plasticity. These plasticity factors include molecules such as the endothelin EDN3b, which normally acts to promote melanocyte specification during embryonic development. Using CRISPRs, we found that deletion of EDN3b from the microenvironment strongly inhibits tumor progression and growth, identifying this as a relevant target in halting metastatic progression. These data imply that cellular identity of melanoma is not fixed, but instead is strongly dominated by the surrounding microenvironment in which the cell finds itself.
Building upon this work, we are interested in identifying novel factors that control cell identity in melanoma. These include embryonic lineage programs that are important in normal melanocyte specification, and whether such embryonic programs are subverted for melanoma initiation, progression and metastasis. In some cases, these embryonic lineage programs are controlled cell-autonomously, and in other cases they are strongly influenced by the surrounding microenvironment. The mechanisms by which these embryonic lineage programs are maintained in melanoma, or not, is a significant open question for us.
We have identified new and unexpected cell types in the microenvironment which promote melanoma progression. These include adipocytes, or fat cells, as mediators of melanoma growth and invasion. We are using both focused and unbiased screening approaches to identify the factors present in these fat cells which promote these behaviors in the tumor cell itself. We also find that there are reciprocal interactions between the melanoma cells and the adipocytes, such that the tumor microenvironment itself is not a fixed entity, but highly influenced by the presence of nascent tumor cells.
Despite the absence of ultraviolet light exposure, our zebrafish melanomas exhibit a surprisingly high degree of genetic diversity. The origins and consequences of this diversity are unclear, and we are undertaking projects to identify how cell stress may alter mutational patterns in melanoma. The identification of such mechanisms will shed light on strategies that tumors may use to survive in new microenvironments. Some of these mechanisms may be amenable to pharmacologic inhibition.
It is increasingly recognized that there is a close link between physical fitness, obesity and risk of cancer progression. Based on these observations, we have recently developed a system in which fish are exercised in a swim tunnel and the effects on melanoma growth can be precisely quantified. The genetic flexibility of this system is now allowing us to dissect both microenvironmental and somatic changes that mediate the effect of exercise on cancer.