Recent work has highlighted glutaminase (GLS) as a key player in cancer cell metabolism, providing glutamine-derived carbon and nitrogen to pathways that support proliferation. There is significant interest in targeting GLS for cancer therapy, although the gene is not known to be mutated or amplified in tumors. As a result, identification of tractable markers that predict GLS dependence is needed for translation of GLS inhibitors to the clinic. Herein we validate a small molecule inhibitor of GLS and show that non-small cell lung cancer cells marked by low E-cadherin and high vimentin expression, hallmarks of a mesenchymal phenotype, are particularly sensitive to inhibition of the enzyme. Furthermore, lung cancer cells induced to undergo epithelial to mesenchymal transition (EMT) acquire sensitivity to the GLS inhibitor. Metabolic studies suggest that the mesenchymal cells have a reduced capacity for oxidative phosphorylation and increased susceptibility to oxidative stress, rendering them unable to cope with the perturbations induced by GLS inhibition. These findings elucidate selective metabolic dependencies of mesenchymal lung cancer cells and suggest novel pathways as potential targets in this aggressive cancer type.

No abstract

No abstract

The pancreas is highly innervated with sensory, autonomic and enteric neurons, and recent studies have highlighted the crosstalk between neurons and pancreatic cells and their potential role in driving tumorigenesis and promoting metastasis. Published data has shown that ablation of sensory neurons in models of pancreatic ductal adenocarcinoma (PDAC) resulted in delayed tumor initiation and progression. In addition, perineural invasion (PNI), the invasion of tumor cells along the nerve, is a pathological characteristic frequently observed in PDAC. PNI has been detected in early stages of pancreatic cancer and associated with a poor outcome. We hypothesized that PDAC tumor cells express neuronal genes which may be playing a role in enabling tumor-nerve interactions and promoting PDAC tumorigenesis. Here we used pooled sgRNA CRISPR screening to probe a library of 5000 sgRNAs targeting 800 neuronal genes. We selected three established (MiaPaca-2, Panc-1 and BxPC3) and two patient derived human PDAC cell lines (PAXF1657 and PAXF1997) and performed in vitro and in vivo CRISPR screens. sgRNAs targeting the CYFIP1 gene were negatively selected across multiple cell line models in vitro and in vivo. Cytoplasmic FMR1 interacting protein 1 (CYFIP1) is a protein that has been described to regulate actin polymerization and protein translation through two distinct protein complexes. CYFIP1 is a member of the heteropentameric WAVE regulatory complex (WRC) which includes CYFIP1, WAVE, NAP1, ABI, and HSPC300. Rac1 binding to CYFIP1 results in activation of WRC, allowing WAVE to interact with Apr2/3 and enabling actin polymerization. In addition, CYFIP1 can interact with fragile X mental retardation protein (FMRP) and eIF4E, acting as a translation repressor. CYFIP1 is highly expressed at excitatory synapses of neurons and is required for proper dendritic spine morphology. CYFIP1 has been linked to neurological and neuropsychiatric conditions such as autism, epilepsy and schizophrenia due to its location in a chromosomal region (15q11.2) with frequent copy number variations. In order to validate the pooled CRISPR screen results we individually knocked out CYFIP1 in PDAC cancer cell lines. We were able to show that knockout of CYFIP1 reduced proliferation of PDAC cells in vitro and dramatically reduced tumor growth in vivo. Additionally, knockout of CYFIP1 resulted in reduced protein levels of WRC members WAVE1 and NCKAP1, suggesting destabilization of the complex. The mechanism by which CYFIP1 promotes tumor growth has yet to be elucidated, but our data shows that CYFIP1 is playing a critical role in PDAC tumorigenesis and is a potential therapeutic target. Citation Format: Kiley Couto, Matthew Strickland, Tiffany Liao, Mortada Najem, Aaron Fulgham, Ameya Apte, Pearl Huang, Jonathan Hurov, Alexandra Lantermann. The autism and schizophrenia-associated CYFIP1 protein is required for pancreatic tumor growth and presents a potential therapeutic target [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr LB-C05. doi:10.1158/1535-7163.TARG-19-LB-C05

The Drosophila extracellular matrix protein Dumpy (Dpy) is one of the largest proteins encoded by any animal. One class of dpy mutations produces a characteristic shortening of the wing blade known as oblique (dpyo), due to altered tension in the developing wing. We describe here the characterization of docked (doc), a gene originally named because of an allele producing a truncated wing. We show that doc corresponds to the gene model CG5484, which encodes a homolog of the yeast protein Yif1 and plays a key role in ER to Golgi vesicle transport. Genetic analysis is consistent with a similar role for Doc in vesicle trafficking: docked alleles interact not only with genes encoding the COPII core proteins sec23 and sec13, but also with the SNARE proteins synaptobrevin and syntaxin. Further, we demonstrate that the strong similarity between the doc1 and dpyo wing phenotypes reflects a functional connection between the two genes; we found that various dpy alleles are sensitive to changes in dosage of genes encoding other vesicle transport components such as sec13 and sar1. Doc’s effects on trafficking are not limited to Dpy; for example, reduced doc dosage disturbed Notch pathway signaling during wing blade and vein development. These results suggest a model in which the oblique wing phenotype in doc1 results from reduced transport of wild-type Dumpy protein; by extension, an additional implication is that the dpyo alleles can themselves be explained as hypomorphs.