Developing multiplexed functional cancer genomic approaches
Genome sequencing has catalogued somatic alterations in human cancers and identified many putative tumor suppressor genes. However, simply identifying genomic alterations does not reveal their functional importance to cancer growth. Genetically engineered mouse models uniquely enable the introduction of defined genetic alterations into normal adult cells, which results in the initiation and growth of tumors entirely within their natural in vivo setting. However, conventional Cre/Lox-based systems are not readily scalable or sufficiently quantitative to investigate the deluge of candidate genes being uncovered by genome sequencing. To increase the scope and precision of in vivo cancer modeling, we have developed methods to integrate conventional genetically engineered mouse models, CRISPR/Cas9-based somatic genome engineering, and quantitative genomics with mathematical approaches. Tumor barcoding coupled with CRISPR/Cas9-mediated gene inactivation and high-throughput barcode sequencing (Tuba-seq) enables the investigation of multiple tumor genotypes in parallel, as well as the quantification of cancer initiation and growth. We have applied these methods to analyze the impact of inactivating many diverse genes in parallel, uncover novel functional tumor suppressors, and map the genetic interactions that drive cancer cell fitness. We are developing and optimizing other multiplexed techniques to generate additional dimensions of information across all aspects of cancer initiation, growth, and therapy responses.
Uncovering the molecular outputs of driver gene alterations
Understanding the functional outputs of driver genes remains a major gap in our comprehension of the drivers of cancer growth and progression. We are interested in using genetic approaches, molecular analyses, and single cell genomics methods to investigate the mechanisms by which tumor suppressors constrain cancer growth in vivo. Our work on the tumor suppressive serine/threonine kinase LKB1 has uncovered the Sik family of kinases as key tumor-suppressive substrates and suggests a broad impact of Lkb1 on chromatin accessibility and lung cancer cell state. We have also employed homology directed repair in somatic cells to induce an array of oncogenic KRAS variants in lung epithelial cells, which revealed an unexpectedly dramatic difference in the oncogenic potential of these variants in vivo. By integrating proteomic and gene expression data, we also identified KRAS-interacting proteins that affect oncogenic KRAS-driven lung tumor growth in vivo. We continue to be interested in unraveling the molecular functions of novel tumor suppressors and uncovering the interplay between different driver gene pathways.
Quantifying genotype by environment interactions
Gene by environment interactions are key drivers of evolution and fitness. Cancer cells do not exist in isolation but interact broadly with components of their microenvironment. Understanding how cancer genotype alters the response of cancer cells to different microenvironments is critical for understanding how different cancers respond to diverse contexts and challenges. We have previously investigated how tumor genotype impacts therapy responses, arguably one the most important acute environmental changes that cancers experience. We coupled Tuba-seq with robust statistical methods to create a pharmacogenomic map of lung cancer treatment responses in vivo. Interestingly, over 20% of possible tumor genotype-specific therapeutic responses had significant resistance or sensitivity, suggesting that tumor suppressor genotype may be an important driver of patient responses. Beyond genotype-driven responses to therapies, we are interested in how physiological states (such as aging or anti-cancer immune responses) interface with cancer genotype. Uncovering how different environments influence the fitness of cancers of diverse genotypes could have important implications for cancer prevention, detection, and treatment.
Dissecting the molecular drivers of metastasis
Metastasis, the spread of cancer from its original site to other sites within the same organ or in different organs, is responsible for the overwhelming majority of cancer deaths. However, the mechanisms underlying metastasis remain poorly understood. Genetically engineered mouse models of metastatic cancer provide a unique opportunity to understand the molecular biography of metastatic cancer. We have used these models to uncover key drivers of some of the most prevalent and metastatic tumor types, including lung adenocarcinoma, small cell lung cancer and pancreatic ductal adenocarcinoma. For instance, by examining chromatin accessibility in cancer cells isolated from primary tumors and metastases in a genetically engineered mouse model of small cell lung cancer, we discovered that the transcription factor Nfib broadly increases chromatin accessibility, leading to increased expression of neuronal genes and driving metastatic ability. By incorporating quantitative methods and novel in vivo models, our current work is focused on uncovering general rules that govern each step of the metastatic cascade.