Generation of 3D melanoma models using an assembloid-based approach

Abstract

While 3D tumor models have greatly evolved over the past years, there is still a strong requirement for more biosimilar models which are capable of recapitulating cellular crosstalk within the tumor microenvironment while equally displaying representative levels of tumor aggressiveness and invasion. Herein, we disclose an assembloid melanoma model based on the fusion of individual stromal multicellular spheroids (MCSs). In contrast to more traditional tumor models, we show that it is possible to develop self-organizing, heterotypic melanoma models where tumor cells present stem-cell like features like up-regulated pluripotency master regulators SOX2, POU5F1 and NANOG. Additionally, these assembloids display high levels of invasiveness while embedded in 3D matrices as evidenced by stromal cell promotion of melanoma cell invasion via metalloproteinase production. Furthermore, sensitivity to anticancer drug doxorubicin was demonstrated for the melanoma assembloid model. These findings suggest that melanoma assembloids may play a significant role in the field of 3D cancer models as they more closely mimic the tumor microenvironment when compared to more traditional MCSs, opening the doors to a better understanding of the role of tumor microenvironment in supporting tumor progression. Statement of significanceThe development of complex 3D tumor models that better recapitulate the tumor microenvironment is crucial for both an improved comprehension of intercellular crosstalk and for more efficient drug screening. We have herein developed a self-organizing heterotypic assembloid-based melanoma model capable of closely mimicking the tumor microenvironment. Key features recapitulated were the preservation of cancer cell stemness, sensitivity to anti-cancer agents and tumor cell invasion promoted by stromal cells. The approach of pre-establishing distinct stromal domains for subsequent combination into more complex tumor constructs provides a route for developing superior tumor models with a higher degree of similarity to native cancer tissues.