The Ridky Lab

Research Program: Three-Dimensional Human Cancer Models

Widely used current cancer models lack clinical faithfulness by failing to study normal human tissue undergoing malignant conversion to invasive cancer. Instead, they employ surrogate platforms that often have limited medical relevance, with experimantal results typically translating into low rates of treatment success. For example, the most commonly studied animal model, the mouse, is well known to differ substantially from humans with respect to critical parameters important in cancer, such as stromal architecture, cellular metabolism, response to oncogene signaling, DNA repair and susceptibility to specific spontaneous cancers.   Additionally, other approaches, such as the use of cell culture-adapted human cancer cell lines injected ectopically into mouse tissues where such cancers never normally arise, as in the standard injection of epithelial cells in the subcutaneous space, suffer from an additional set of intrinsic limitations in approximating the human tissue setting. Another major current bottleneck in cancer research is the inability to rapidly assess the functional importance of a given gene in a human disease process. Sequencing of human cancer tissue genomes has unearthed thousands of mutated and rearranged genes. Distinguishing casual "driver" mutations from unimportant "passengers" requires rigorous genetic experiments in tissue, underscoring the importance of developing new approaches to functionally assess the oncogenicity of specific genetic alterations in the human tissue context. It is increasingly evident that traditional assays used to study cancer processes, such as anchorage independent growth in soft agar, immortalization, and invasion assays in vitro (i.e, Matrigel, which lacks an intact basement membrane) are not faithful surrogates for malignancy in vivo. Thus, there has been an explosion of targets to test, but a paucity of rapid and clinically relevant models in which to test them. An ideal complementary approach to current models is a novel, rapid, and physiologically-relevant platform to characterize and test cancer targets.

     A primary focus in the lab is to develop experimental platforms to rapidly study cancer development in real-time in an entirely human tissue setting complete with human epithelium and stroma from multiple tissues in correct 3-D "seed and soil" architecture.   Skin is the prototype organ for many of these studies and we have optimized tissue-engineering methodologies for the generation of both normal skin and invasive epidermal cancers. Skin malignancies, including epidermal squamous cell carcinoma (SCC), account for more cancers than those arising in all other human tissues combined. SCC and other epithelial cancers develop through a process of tumor-stroma co-evolution in which epithelial cells undergo a stepwise progression toward invasive cancer in conjunction with remodeling of the underlying dermal stroma.  Identifying central cellular pathways and elements in both the tumor and stromal cells necessary for oncogenesis is a primary effort in the lab.
University of Pennsylvania | Perelman School of Medicine