Christian Metallo, Associate Professor at UC San Diego, in collaboration with Prashant Mali. Cellular transformation and neoplastic growth are often driven by aberrations in the epigenome. At the same time, there is an increasing appreciation for the role of metabolic enzymes in controlling the epigenome, as critical substrates controlling epigenetic modifications are produced and consumed by metabolic pathways. In this project, we will apply combinatorial CRISPR/Cas9 screening to identify key points of interactions between cellular metabolism and epigenetics. We will focus on enzymes involved in methylation, demethylation and the production/consumption of aKG and SAM. Specifically, we will conduct in vitro and in vivo screens in MCF7 and MDA-MB-231 breast cancer cells that exhibit different invasiveness and drug sensitivity. This approach will allow us to identify functional interactions between metabolic enzymes, methyltransferase and demethylases that are involved in anchorage-independent or in vivo growth. Subsequently, we will functionally validate specific metabolic changes using quantitative metabolomics and flux analysis.
Hani Goodarzi, Assistant Professor at UCSF, in collaboration with Nevan Krogan. A major obstacle to developing predictive models of regulatory networks is the multi-functionality and context-dependence of posttranscriptional regulators. We are developing an integrated approach to decipher the combinatorial code between RNA-binding proteins and their regulons. We have built a probabilistic approach to derive overlapping regulatory modules. We are relying on multiple independent sources of information to identify these modules: genome-wide protein-RNA interaction maps (CLIP datasets), Bio-ID based proximity tagging and single-cell protein knockdown followed by RNA-seq (Perturb-seq). Together, these datasets cover physical and functional interactions and provide a suitable framework for our analysis. Upon completion, we will interrogate these modules and dissect their roles in RNA life cycle and their contribution to disease.
Keith Yamamoto, Professor at UCSF, in collaboration with Nevan Krogan. This project seeks to develop a new technology, denoted CasCut&Run, which will enable purification of specific in vivo-assembled transcriptional regulatory complexes (TRCs), which consist of the glucocorticoid receptor, ligand, and numerous possible coregulators, followed by the identification of the protein components. In CasCut&Run, a specific glucocorticoid response element (GRE) with a TRC bound will be solubilized from permeabilized cells using a pair of wtCas9 nucleases targeted to flank the GRE. Liberated chromatin fragment will then be purified by targeting a nuclease-defective strep-tagged-dCas9 to a region close to the GRE. Strep-tagged-dCas9 complexes will be purified and analyzed by liquid chromatography tandem mass spectrometry. We have begun carrying out pilot experiments to test individual steps in this process, assessing efficiency, specificity and yield.
Susan Taylor and Pablo Tamayo, Professors at UC San Diego, in collaboration with Silvio Gutkind. We are in the process of generating the PPI and transcriptional data that will provide us with the landscape of oncogenic signaling in GNAS-PKA cancers. In the meantime, we are investigating the nature of the oncogenic state generated by GNAS mutations in existing cancer cell lines from the Cancer Cell Line Encyclopedia. We have been able to identify the GNAS-driven state (state S4) as a one of several RAS-associated oncogenic substates (Kim et al., Cell Syst 2017). Interestingly, this state appears to be dependent on both GNAS and KRAS as we have been able to assess by analysis data from shRNA synthetic lethal screens (Cowley et al., Sci Data 2014). In addition, we have use our REVEALER algorithm (Kim et al., Nat Biotechnol 2016) to identify a number of complementary mutations besides GNAS that are putative drivers: KRAS, PKA and PDE4D.