Genomic information can allow investigators to devise precision therapies that target molecular lesions specific to a patient's cancer. One of the molecular lesions present in many malignant tumors is loss of the NF1 tumor suppressor, which is a driver in neurofibromatosis type 1 (NF1), one of the most frequently inherited genetic disorders. NF1 exhibits a broad clinical spectrum including benign nervous system tumors called neurofibromas, low-grade astrocytomas, pheochromocytoma, and juvenile myelomonocytic leukemia. Plexiform neurofibromas (PN) occur in deep nerves and can degenerate into malignant peripheral nerve sheath tumors (MPNSTs), chemo- and radiation-resistant sarcomas with a dismal 20% five-year survival rate. In NF1 carriers, the lifetime risk is 30% for PN and 8-15% for MPNST. NF1 mutations are also found in sporadic tumors including glioblastoma multiforme (GBM), melanoma, pheochromocytoma, ovarian, uterine, and lung cancers. Furthermore, in animal models and human tumor lines NF1 loss has been shown to drive GBM. A targeted molecular therapy designed to inhibit tumor-initiating and -promoting cells would substantially advance our ability to treat tumors that develop as a result of NF1 loss. Using complementary screening platforms, we identified small molecules that selectively killed or stopped the growth of cells carrying a mutation in NF1 as well as their molecular targets. In this application we focus on our top small molecule leads and the power of our model systems to test the hypothesis that aggressive neurological cancers that are known to be driven by NF1 loss and for which no cure exists, including PN, MPNST and GBM, will respond to molecules that we identified as synthetic lethal with NF1 loss. For this, we will determine pharmacokinetic properties of our top small molecule leads and, where needed, conduct structure-activity relationship studies to improve kinetics and/or reduce toxicity in order to test their efficacy in pre-clinical models of PN and GBM. We will leverage our model systems to define the mechanisms of action of our small molecule leads and add to our pipeline drugs in clinical trials that share the same targets. In addition to mutational inactivation, some GBM tumors exhibit down-regulation of the NF1 protein. Our lead compounds also stopped the growth of human GBM and neuroblastoma cells with low NF1 protein in vitro, supporting the broad application of the small molecule leads that we have identified. To capture tumors that have lost NF1 by any mechanism, we constructed an RNA-based classifier, which is capable of identifying downstream transcriptomic effects that indicate NF1 inactivation in GBM, using machine learning. We will apply the RNA-based classifier, along with targeted sequencing of NF1, to identify additional patient derived xenograft (PDX) GBM tumors that have an inactivating mutation of NF1 or molecular signatures of NF1 loss and test their response to our lead compounds in pre-clinical models. We expect that this work will provide new targets and therapeutic leads for aggressive neurological cancers driven by NF1 loss.

Public Health Relevance

The work described in this proposal aims to find the Achilles heel of aggressive neurological cancers driven by NF1 loss for which no cure exists. These cancers include neurofibromas, malignant peripheral nerve sheath tumors, glioblastoma (GBM), and neuroblastomas, all with dismal 5-year survival rates. Our multidisciplinary approach to develop an innovative conduit for identifying new targets with potential to cure such malignancies addresses an unmet clinical need.

National Institute of Health (NIH)
National Institute of Neurological Disorders and Stroke (NINDS)
Research Project (R01)
Project #
Application #
Study Section
Drug Discovery and Molecular Pharmacology Study Section (DMP)
Program Officer
Morris, Jill A
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Dartmouth College
Schools of Medicine
United States
Zip Code
Allaway, Robert J; Wood, Matthew D; Downey, Sondra L et al. (2018) Exploiting mitochondrial and metabolic homeostasis as a vulnerability in NF1 deficient cells. Oncotarget 9:15860-15875
Way, Gregory P; Sanchez-Vega, Francisco; La, Konnor et al. (2018) Machine Learning Detects Pan-cancer Ras Pathway Activation in The Cancer Genome Atlas. Cell Rep 23:172-180.e3
Way, Gregory P; Allaway, Robert J; Bouley, Stephanie J et al. (2017) A machine learning classifier trained on cancer transcriptomes detects NF1 inactivation signal in glioblastoma. BMC Genomics 18:127
Allaway, Robert J; Fischer, Dawn A; de Abreu, Francine B et al. (2016) Genomic characterization of patient-derived xenograft models established from fine needle aspirate biopsies of a primary pancreatic ductal adenocarcinoma and from patient-matched metastatic sites. Oncotarget 7:17087-102