**1. Introduction**

RASopathy Neurofibromatosis 1 (NF1) is an autosomal dominant hereditary cancer predisposition syndrome that a ffects ~1:3000 individuals [1]. It is caused by mutations in the *neurofibromin 1* (*Nf1*) tumor suppressor gene, which encodes the GTPase-activating protein-related domain (GRD) that catalyzes the inactivation of Ras by accelerating guanosine triphosphate (GTP) hydrolysis to guanosine diphosphate (GDP) [2]. In NF1 individuals, loss of *Nf1* results in high levels of activated Ras, leading to the formation of multiple benign and malignant tumors via multiple e ffector pathways, including the Ras–MAPK pathway, with subsequent activation of the RAF–MEK–ERK cascade.

Patients with NF1 have an increased cancer risk and mortality, and lower survival compared with the general population [3,4]. Based on the Finnish NF1 Registry, the estimated lifetime cancer risk in patients with NF1 is 59.6%, with an estimated cumulative cancer risk of ~25% and ~39% by age 30 and 50 years, whereas the respective percentages in the general Finnish population are much lower, at 30.8%, 0.8% and 3.9% [3]. The most common malignancies are of nervous system origin, such as malignant peripheral nerve sheath tumors (MPNSTs) and astrocytomas, which comprise 63% of all malignancies [3]. Other malignancies include breast cancer, rhabdomyosarcomas, pheochromocytoma, gastrointestinal stromal tumor (GIST), malignant fibrous histiocytoma, and thyroid cancer [3].

MPNST is a very aggressive spindle cell sarcoma which accounts for the majority of cancer deaths in all NF1 patients and is a hallmark complication of this condition [3–6]. MPNST may arise from any of the pre-existing plexiform neurofibromas distributed throughout a patients body. Unfortunately, thereis no way of knowing whichindividual and, more specifically, whichlesions within any one individual are likely to behave in a malignant fashion and thus many patients require regular screening with standard radiographic techniques such as MRI and PET/CT. Patients with *Nf1* microdeletion, i.e., a large deletion of the *Nf1* gene and its flanking regions, are especially susceptible to MPNSTs [7,8].

NF1-specific malignancies, including MPNSTs, typically manifest early in life and are responsible for the relative excess in cancer incidence and mortality observed in children and young adults [4]. Those malignancies are typically very di fficult to treat and current therapies have shown little long-term benefit despite extensive research e fforts [9]; however, early chemoprevention to delay cancer occurrence and reduce cancer risk remains largely unexplored. The success of chemoprevention has been impressively demonstrated in epithelial malignancies, particularly breast, prostate and colorectal cancers, with the use of selective estrogen receptor modulators (SERM) (e.g., tamoxifen), 5 α-reductase inhibitors (e.g., finasteride) and cyclooxygenase-2 (COX-2) inhibitors, a type of non-steroidal anti-inflammatory drug (NSAID, e.g., sulindac, aspirin, celecoxib) that inhibited the appearance of colorectal polyps in various familial colorectal cancer predisposing syndromes [10].

The development of new chemical agents for chemoprevention is a long, di fficult and expensive process. A potential strategy to circumvent these challenges is to discover new uses for compounds with an established track record of safe and long-term use in humans, alone or in combination with already known cancer prevention agents, such as widely used cyclooxygenase-2 (COX-2) inhibitors, whose anti-neoplastic e ffects are mediated through the inhibition of angiogenesis via decreasing COX-2-induced vascular endothelial growth factor (VEGF) production [11] and apoptosis via altered caspase signaling [12,13]. Notably, COX-2 overexpression has been found in a variety of sarcomas and has been associated with poor prognosis [14–16], thus suggesting that COX-2 inhibitors could play a role in NF1 cancer prevention.

We previously identified that mebendazole (MBZ), an FDA-approved low molecular weight benzimidazole derivative with a lengthy track record of safe long-term human use, significantly reduced tumor growth and improved survival in the animal models of glioblastoma multiforme (GBM) and medulloblastoma (Sonic Hedgehog (SHH) Group and c-Myc/OTX2 amplified Group 3) and also reduced tumor formation in a Familial Adenomatous Polyposis (FAP) colon cancer model [17–20]. A number of mechanisms for MBZ's anti-neoplastic activity have been proposed by us and others, including microtubule disruption, pro-apoptosis, and the inhibition of growth factor signaling through the blockage of various tyrosine kinases, particularly VEGFR2 [17,18].

The current study evaluates the feasibility of a cancer prevention strategy using non-toxic MBZ alone and in combination with COX-2 inhibitors in a *cis Nf1*+/−*;Tp53*+/− (NPcis) mouse model of NF1 [21]. Like NF1 patients, NPcis mice spontaneously develop predominantly soft tissue sarcomas including MPNSTs (genetically engineered murine (GEM) PNSTs) and malignant Triton tumors, as well as rhabdomyosarcomas and astrocytomas that severely limit their life expectancy to ~5 months [21–24]. The addition of heterozygous *Tp53* knock-out (KO) accelerates the cancer development, which mimics the secondary mutations required for the transformation to malignancies such as MPNST, where the second copy of *Nf1* is also lost due to the loss of heterozygosity (LOH) [21,22].

#### **2. Material and Methods**

#### *2.1. Tissue Culture and Cell Lines*

The human NF1-associated MPNST cell line NF90.8 was provided by Dr. Michael Tainsky (Wayne University, Detroit, MI) and sNF96.2 was purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). Cells were cultured in DMEM (ATCC) supplemented with 10% fetal bovine serum (FBS) (Sigma, St. Louis, MO, USA) and penicillin/streptomycin (Thermo Fisher, Waltham, MA, USA). These cell lines were not authenticated. All cells were tested and found free of mycoplasma contamination.

#### *2.2. Reagents and Antibodies*

Rabbit anti-Nf1 antibody (A300-140A, Lot 3) was purchased from Bethyl Laboratories and anti-βActin horseradish peroxidase (HRP) antibody (C-11, SC-1615HRP, Lot G3015) was purchased from Santa Cruz Biotech. An Active Ras Detection Kit (#8821, antibody Lot 7), including the anti-Ras antibody, was purchased from Cell Signaling Technology.
