Cancer occurs when mutations in multiple genes make otherwise normal cells cancerous, and these mutations often accumulate over time, creating phenotypic diversity in a patient's tumor. Specific genetic alterations in specific cancer types are associated with prognosis, response or resistance to therapy, especially targeted therapy, and the propensity of tumors to acquire further mutations, among other phenotypes. However, inferring genotype-phenotype links in patients is challenging because any two tumors are too genetically different to isolate the effect of one or a few mutations. The ability to systematically link cancer-associated mutations or combinations thereof to their phenotypes will advance scientists' understanding of cancer pathogenesis and genetically related disease characteristics.
In a new study, researchers from institutions including the Broad Institute, Dana-Farber Cancer Institute, Brigham and Women's Hospital, and Harvard Medical School argue that genome editing and adaptation of cancer-associated mutations can be harnessed degree advantage to establish a cellular model of human tumorigenesis. Such genome editing models would reproduce precise genetic properties, lineage relationships, and step-by-step development of cancer, allowing them to establish genotype-to-phenotype links in a controlled experimental design.
Although similar models exist for tumors derived from self-renewing stem cells, particularly colorectal cancer, there are no comparable models for tumor types derived from nonstem differentiated cells. The researchers came up with a way to expand the cancer by starting with the non-stem cell origin of melanoma -- healthy human melanocytes -- and then generating a series of cells with precise genome editing for key oncogene mutations possible cell models generated.
Researchers generate a series of progressive genome-edited human melanoma models. They started with healthy human melanocytes and introduced, in a step-by-step fashion, mutations in up to five genes spanning six pathways that are frequently dysregulated in melanoma: CDKN2A (part of the RB pathway), BRAF (MAPK) , TERT (telomerase), PTEN (PI3K/AKT), TP53 (P53), and APC (Wnt), a total of nine genetically distinct cell models have been established.
Utilizing physiological assessment, histopathology, and single-cell RNA sequencing (scRNA-Seq), as well as computational methods and machine learning algorithms, the researchers characterized these models during in vitro growth and after intradermal injection via mouse xenografts. Through these models, they linked the genotype of melanocytes to phenotypes such as gene expression programs, persistent replication capacity, malignancy, rapid tumor growth, tumor pigmentation, metastasis, and histopathological features.
In vitro, successive mutations produce ordered progression through an ordered expression continuum. In vivo, mutations in malignant tumor cells also affected cell type composition and expression status of tumor-infiltrating microenvironment cells. These melanoma models share genotype-related expression programs with patient melanomas and partially reproduce the genotype-related histopathological features of patient melanomas.
In brief, the genotype-phenotype link identified by the researchers highlights that the effects of mutations often depend on genetic background. This genetic epistasis makes understanding the phenotypic consequences of the mutational landscape in human cancer a combinatorial problem, and research into this problem requires modeling strategies that can scale to multiple mutations. Genome-edited human cancer models enable causal relationships between a defined set of genetic alterations and disease-related phenotypes, further understanding how cancer mutations contribute to the many different phenotypes of human malignancies.