Browsing by Author "Wood, Kris C"

Enhancers possess both structural elements mediating promoter looping and functional elements mediating gene expression. Traditional models of enhancer-mediated gene regulation imply genomic overlap or immediate adjacency of these elements. We test this model by combining densely-tiled CRISPRa screening with nucleosome-resolution Region Capture Micro-C topology analysis. Using this integrated approach, we comprehensively define the cis-regulatory landscape for the tumor suppressor PTEN, identifying and validating 10 distinct enhancers and defining their 3D spatial organization. Unexpectedly, we identify several long-range functional enhancers whose promoter proximity is facilitated by chromatin loop anchors several kilobases away, and demonstrate that accounting for this spatial separation improves the computational prediction of validated enhancers. Thus, we propose a new model of enhancer organization incorporating spatial separation of essential functional and structural components.

Cancer is the second leading cause of death worldwide. Although the era of targeted therapies has produced encouraging clinical responses across a wide range of genetically defined cancer subtypes—from BRAF-mutant melanomas to EGFR-mutant non-small cell lung cancers—these benefits are, unfortunately, usually short-lived. The fact is that patients typically develop resistance to first-line targeted therapies within a matter of just months. This problem is complicated by two interrelated considerations: (1) cancer cells can develop resistance to their cognate targeted therapies through wide ranges of distinct mechanisms; and, (2) many, if not all, of these mechanisms co-evolve within the same individual patient or tumor. This presents a difficult quandary: on the one hand, treatments which use one drug to target an individual resistance mechanism are insufficient and unlikely to prove curative; on the other hand, treatments which use multiple drugs to target many different mechanisms at the same time are demanding, and are unlikely to be feasible. Thus, new approaches to the problem of multifocal drug resistance are needed. One strategy is to identify, target, and exploit new vulnerabilities which emerge as a consequence of drug resistance itself. Although these “collateral” sensitivities have long been documented in the microbial literature, details of their therapeutic relevance in cancer had, until recently, been limited to isolated reports. A previous, unpublished study of ours found that collateral sensitivities are a stable, predictable, and widespread feature of drug resistance in cancer. Furthermore, drug-resistant, EGFR-mutant NSCLC cells in particular are especially prone to acquisition of targetable collateral sensitivities. The dissertation below builds off of our previous work by identifying the bis-chalcone MCB-613 as a collateral sensitivity to drug resistance in EGFR-mutant NSCLC. Extensive mechanistic dissection reveals that MCB-613 enacts this program through the inhibition of KEAP1, which selectively targets resistant cells through hyperinduction of the integrated stress response. Furthermore, this pattern of activity makes bis-chalcones unique among KEAP1 inhibitors, and relies upon their ability to act as a molecular glue which tethers together KEAP1 monomers by specific lysine residues within their BTB dimerization domains. As a result, this work demonstrates the value of collateral effects not just as a paradigm for approaching multifocal drug resistance in cancer, but also as a starting point for the elucidation of novel underlying biology.