Introduction

Neurofibromatosis type 2 (NF2) is a rare disorder that is inherited in an autosomal-dominant manner and is attributed to the loss of heterozygosity (LOH) of the NF2 gene, which encodes for the tumor suppressor protein Merlin. Patients affected by the disease develop vestibular schwannomas (VS), meningiomas, and ependymomas, resulting in high morbidity and premature mortality [1, 2]. To date, NF2 has no FDA-approved drug-based treatment and there are a limited number of trials underway to evaluate drugs such as bevacizumab, crizotinib and brigatinib [2, 3]

Merlin plays a central role in mediating cell contact inhibition (CI) of proliferation [4]. Loss of Merlin leads to abnormal activation of multiple signaling pathways, including those regulated by small G-proteins Ras/Rac/cdc42 and the Hippo-YAP pathway. The Hippo-YAP pathway plays a significant role in cell growth and organ size control, and, at its core, is comprised of a kinase cascade that regulates the transcriptional regulator YAP [5, 6]. Previous studies demonstrate that the function of YAP is crucial for VS development [7, 8]. YAP’s function as a transcriptional regulator is mediated through interactions with a host of transcriptional factors and epigenetic effectors [9]. Among the effectors that interact with YAP are the Bromodomain and Extra-Terminal domain (BET) proteins, which bind to lysine-acetylated histones to regulate transcription [10]. The BET family member BRD4 has been shown to play central roles in different types of cancer [11]. We have previously shown that the BET inhibitor JQ1 can selectively reduce the growth of NF2-null schwannoma and Schwann cells in vitro and tumorigenesis in vivo [12].

Independently, there is evidence showing that Merlin loss can lead to increased cell proliferation that is dependent on the activity of Focal Adhesion Kinase 1 (FAK1), which is encoded by the PTK2 gene [13,14,15]. This protein tyrosine kinase is involved in regulating cell adhesion, survival, and cell proliferation and its overexpression has been identified in various cancers and shown to promote tumor progression and metastasis [16]. We have previously demonstrated that FAK1 inhibition via crizotinib has antiproliferative effects in NF2-null Schwann cells, and this FDA approved drug is currently in phase II clinical trials for NF2-related VS [1, 13].

There is growing evidence to support the idea that co-inhibiting the integrin/FAK and BRD4 axes could be a promising approach in solid neoplasia, including breast cancer subtypes such as triple-negative breast cancer (TNBC) [17]. In Non-Small Cell Lung Cancer (NSCLC), tumors exhibiting oncogenic activation of KRAS have limited therapeutic options and poor outcomes. Recent evidence has shown that the combination of FAK and BET inhibition synergistically induced apoptotic cell death, DNA damage response, downregulation of epithelial-mesenchymal transition (EMT) and disruption of NSCLC tumorspheres. This provided evidence in support of the notion that the combination treatment could overcome the KRAS-driven malignancy [18]. However, the molecular interactions between these two pathways remain to be elucidated.

In the current study, wild-type and NF2-null Schwann cells and schwannoma cell lines were used to determine the impact on cell proliferation and death by employing combination treatment with the BET inhibitors JQ1 or OTX-015 and the FAK1 inhibitor crizotinib. Our analysis shows that the combination treatment exerts a synergistic antineoplastic effect at nanomolar concentrations and results in selective inhibition of NF2-null Schwann cell proliferation compared to wild-type Schwann cells in vitro and to either drug alone. Moreover, we propose a model highlighting the mechanism underlying this synergy, by demonstrating that BRD4 binds to the PTK2 promoter and regulates FAK1 expression. Our data suggest that combined targeting of BET and FAK1 may offer a potential therapeutic option for treating NF2-related schwannomas.

Results

Combined BET and FAK inhibition selectively impair NF2-deficient Schwann cell proliferation

To assess the effects of combined inhibition of BET proteins and FAK on the proliferation of Schwann cells wildtype (hSC2λ-WT) or deficient for both NF2 alleles (hSC2λ-NF2−/−), and NF2-deficient schwannoma cells (HEI-193) were treated with vehicle (DMSO), JQ1 (0.5 µM), crizotinib (0.5 µM), or JQ1 (0.5 µM) + crizotinib (0.5 µM) and cell proliferation was evaluated after 24, 48 and 72 h. In agreement with previously reported findings, treatment with JQ1 (0.5 µM) or crizotinib (0.5 µM) alone resulted in preferential inhibition of proliferation of the NF2 deficient cells (Fig. 1). The combination treatment inhibited the growth of cells in a time and dose-dependent manner. Similar results were obtained with OTX015 (0.5uM), a clinical-stage BET inhibitor (Fig. S1).

Fig. 1: Impact of BET and FAK inhibition on NF2-null Schwann cell growth.
figure 1

HEI-193, hSC2λ-WT and hSC2λ-NF2−/− cells were cultured with DMSO, JQ1 (0.5 µM), crizotinib (0.5 µM) or JQ1+ crizotinib (0.5 µM each) and counted every 12 h, for 4 days. The data shown are the mean of 3 independent experiments. Two-way ANOVA was used to determine the interaction of time and treatment (control vs OTX, control vs crizotinib, and control vs combination). The p values are indicated in the figure: ns = not significant, *p < 0.05, ***p < 0.001, ****p < 0.0001. Error bars = SD.

Full size image

To determine whether the two drugs work in a synergistic manner we carried out dose-response assays to evaluate the drug combinations in a matrix format, and synergy-scoring models were applied. We first treated the hSC2λ-WT and hSC2λ-NF2−/− cell lines to evaluate the synergy of JQ1 and crizotinib, for 72 h with a matrix of 3-fold dilution series of each drug starting from 10 μM for JQ1 and 10 μM for crizotinib. The results were analyzed using the Loewe model [19]. Typically, a score of >10 indicates a synergistic interaction, whereas a score <10 indicates an additive effect. These analyses indicated that the combined treatment with JQ1 and crizotinib shows synergistic effects on growth suppression in the treated cells, with the NF2-deficient Schwann cells displaying a more pronounced synergy effect (synergy score = 14.08) compared to the hSC2λ-WT cells (synergy score = 10.97) (Fig. 2). Similar findings were obtained when assessing OTX-015 (Fig. S2).

Fig. 2: Combined BET and FAK inhibition synergize in inhibition of NF2-null Schwann cell growth.
figure 2

A hSC2λ-WT or (B) and hSC2λ-NF2−/− cells were treated with increasing concentrations of JQ1 and crizotinib (0–10 μM, 3-fold dilutions) in a 10 × 10 matrix format. Cell viability was determined for each pair of combination treatments after 72 h using a luminescent ATP-dependent assay (Cell Titer Glo), in duplicate for each point. the Loewe synergy method was used to calculate synergy. Heatmap of synergy score (mean +/− SEM) are indicated by the color scale (white = neutral, green = low, red = high). The data shown are combined from 3 independent experiments.

Full size image

Combined inhibition of BET and FAK arrests cells in the G1/S and G2/M Phases of the cell cycle

To further investigate the effects of combined treatment on the cell cycle, hSC2λ-NF2−/− cells were treated with the various drug combinations, and their cell cycle profiles were analyzed by flow cytometry. Treatment with crizotinib alone increased the proportion of NF2-deficient Schwann cells at the G2/M phase, whereas JQ1 and OTX-015 treatment resulted in an increase in the fraction of cells in G1 phase and a reduced fraction in S phase, suggesting BET inhibition leads to an inhibition of the G1/S transition. Significantly, the combination treatment with JQ1 or OTX-015 and crizotinib dramatically increased the proportion of cells at the G2/M phase (Fig. 3). These findings are also bolstered by analysis of CDK2 levels, required for both G1S and G2/M transitions and cyclin B1 required for G2/M transition. This analysis shows a significant decrease in CDK2 and cyclin B1 levels in cells treated with the inhibitor combinations compared to treatment with either drug alone (Fig. 3). This is consistent with the requirement for these proteins for transition through the G1/S and G2/M phases.

Fig. 3: Combined BET and FAK inhibition arrest NF2-null Schwann cells in the G1/S and G2/M phases of the cell cycle.
figure 3

A hSC2λ-NF2−/− cells were treated with DMSO, JQ1 (0.5 µM), OTX-015 (0.5 µM), crizotinib (0.5 µM), JQ1+ crizotinib (0.5 µM each) or OTX-015+ crizotinib (0.5 µM each) for 48 h. Cell cycle distribution was assessed using propidium iodide (PI) incorporation measured by FACS analysis. Cells in G1 and G2 are indicated in red. B Summary of cell cycle distributions in the different treatment conditions. C Western blot analysis of cell cycle markers (CDK2, Cyclin B1) in the cells treated as in A at 48 h post treatment. Vinculin was used as a loading control. D hSC2λ-NF2−/− cells were treated with DMSO, OTX-015 (0.5 µM), crizotinib (0.5 µM) or OTX-015+ crizotinib (0.5 µM each) for 48 h. Cells were labeled for binding of FITC-annexin V and staining with propidium iodide (to exclude necrotic cells). Treated cells were analyzed by FACS and the percent of cells are indicated in each quadrant. E Summary of cells undergoing apoptosis in the different treatment conditions. The data shown are the average of 3 independent experiments with each data point done in triplicate. P values: ***p < 0.001, ****p < 0.0001. Error bars = SD.

Full size image

To investigate the potential effects of the combined therapy on cell viability NF2-deficient Schwann cells were treated with crizotinib and OTX-015 inhibitors, and cell death was analyzed by assessing Annexin V levels using flow cytometry. While either drug alone did induce a significant increase in annexin V (approximately 3-4 fold increase) the combined drug treatment resulted in a slight increase in annexin V levels to an average of approximately 5-fold) (Fig. 3D, E). Our combined analysis suggests that the combination of BET and FAKi in suppressing NF2-deficient Schwann cell proliferation is mediated through both blocking the cell cycle and an increase in apoptosis.

Combined inhibition of BET and FAK inhibition suppresses tumorigenic capacity of NF2-deficient Schwann in vivo

To assess the impact of combined BET and FAK inhibition in an in vivo setting, we employed an orthotopic model based on the Nf2-deficient SC4-luciferase cells (SC4-Luc). This model has been previously employed extensively to assess the antineoplastic activity of various drugs and small molecules [7, 12, 20]. As outlined in Fig. 4A, SC4-Luc cells were orthotopically implanted into the sciatic nerves of NOD/SCID mice and tumor growth was monitored via Bioluminescent Imaging. Ten days post-surgery similar flux readings for all animals were validated and animals were enrolled into 4 cohorts: control (vehicle only), OTX-015-treated cohorts (5 mg/kg, IP, every other day), crizotinib-treated cohorts (0.5 mg/kg, IP, every other day) and combined treatment for 14 days. Upon completion of treatment, the mice were imaged, euthanized, and tumors removed and weighed. A comparison of tumor weights demonstrated a significantly lower average tumor weight in the OTX-015 and crizotinib-treated groups compared to the control (Fig. 4B, C). Notably, the combined treatment of OTX-015 and crizotinib resulted in strong inhibition of tumor growth when compared to both control treatment and compared with either inhibitor alone. Taken together, these data demonstrate that combined BETi/FAK inhibition has significant antitumor activity in vivo.

Fig. 4: Combined BET and FAK inhibition synergize in inhibition of NF2-null Schwann tumor growth.
figure 4

A Diagram of 4 arm study design (B) Representative BLI images of animals from the different cohorts at 14 days post initiation of treatment. C Distribution of tumor weight from vehicle or JQ1-treated cohorts. The results of t-test with equal variances show the OTX-015/crizotinib combination treated cohort has significant lower average tumor weight compared to control group. N = 10 in each cohort. Error bars = SD, *P < 0.05, **P < 0.01, ****P < 0.001.

Full size image

BET inhibition selectively reduces FAK1 levels in NF2-deficient Schwann cells

BRD4 is upregulated in VS and, as discussed in the introduction, there is evidence suggesting the involvement of BRD4 in regulation of FAK expression [12, 18]. To directly assess the potential involvement of BRD4, we assessed the presence of this protein in the proximity of the PTK2 (the gene coding for FAK1) promoter in human Schwann cells (hSC2λ-NF2−/−) using ChIP-seq analysis [12]. A number of peaks, correlating with enrichment for BRD4, were identified in the promoter region of the PTK2 locus. Moreover, a portion of these peaks were diminished in cells treated with JQ1 (Fig. 5A). The binding of BRD4 to the promoter of PTK2 was validated using ChIP-qPCR and, as expected, was found to be reduced in cells treated with JQ1 or OTX-015 (Fig. 5B). In further validation of the role BRD4 has in regulating the expression of FAK1, we find that treatment of the hSC2λ-NF2−/− cells with JQ1 resulted in decreased levels of FAK1 mRNA and protein levels. Moreover, these effects were significantly diminished in cells that are wildtype for the NF2 alleles (Fig. 5C, D).

Fig. 5: BET inhibition negatively impacts the expression of FAK1.
figure 5

A Landscape of BRD4 binding to the PTK2 promoter region in the absence (DMSO) or presence of JQ1 treatment. Arrow highlights the peak examined in subsequent experiments. B Validation of BRD4 binding to the PTK2 promoter by ChIP-qPCR. Cells were treated with DMSO, JQ1 (1 μM) or OTX-015 (1 μM) for 12 h, crosslinked and immunoprecipitated using an anti-BRD4 antibody. C Analysis of FAK1 mRNA expression by RT-qPCR. Cells were treated as in B. D Analysis of FAK1 protein levels by western blotting in hSC2λ-WT cells or hSC2λ-NF2−/− cells treated with JQ1 or OTX-015, as indicated. GAPDH was used as a loading control. The data shown are representative of 3 independent experiments.

Full size image

Discussion

Approximately 1 in 25,000 people are diagnosed with NF2-related schwannomatosis, formerly known as Neurofibromatosis type 2 (NF2), with limited treatment options available and substantial morbidity. Clinical management of vestibular schwannoma is still a major concern as the available options include close observation and off-label use of bevacizumab for rapidly growing vestibular schwannomas [21]. This leaves invasive approaches such as surgery and stereotactic radiosurgery as the only options when life-threatening risks develop. Unfortunately, to date, there is no FDA-approved pharmacotherapy for this type of tumor. Therefore, there is an urgent need to find new treatment options to improve the quality of life for these patients [22]. Studies have shown that YAP is involved in NF2-associated vestibular schwannoma tumorigenesis, which has led to assessment of TEAD-YAP inhibitors as potential therapeutic strategies in this disease [7, 8, 23]. It has also been demonstrated that small molecule inhibitors of BET proteins and drugs like JQ1 and OTX-015 are effective in reducing the size of vestibular schwannomas [12]. Additionally, another drug, crizotinib, has demonstrated promising tumor size reduction of this type of tumor by targeting the Focal Adhesion Kinase 1 (FAK1) [13] and is currently in phase 2 clinical trials. Considering the potential use of these therapeutic targets, we hypothesized that co-targeting BET and FAK1 proteins could exert a more effective and less toxic antineoplastic effect in NF2-null Schwann cells compared to a single agent.

BET inhibitors, such as JQ1 and OTX-015, have shown promise for their anticancer effects when used in combination with other treatments. These combined treatments have been effectively used to treat triple-negative breast cancer, in which the BET inhibitor JQ1 and the FAK inhibitor VS-6063 were used in vitro and in vivo [17]. In non-small cell lung cancer, specifically the EGFR/KRAS adenocarcinoma subtype, a screening analysis identified a synergistic effect with the use of JQ1 and VS-6063 [18]. In the context of NF2-null Schwann cells, our findings indicate that combining the inhibitors can effectively inhibit cell proliferation, arrest the cell cycle, and induce apoptosis. In order to better understand the mechanism behind this effect, we examined the landscape of BRD4 binding to chromatin in human Schwann cells that were deficient for NF2 and were treated with JQ1. We focused on BRD4 since we have previously shown it to be the main mediator of the inhibitory effects of BETi in NF2 deficient Schwann cells [12]. We identified a number of BRD4 peaks localized in proximity of the promoter of PTK2, the gene that codes for FAK1. Notably, we observed a significant reduction in BRD4 binding in the proximity of the PTK2 transcriptional start site (TSS) in cells treated with JQ1. A reduction in FAK1 expression levels in NF2-deficient Schwann cells treated with BETi was confirmed by RT-qPCR and western blotting. Interestingly, the reduction occurred to a greater extent in the hSC2λ-NF2−/− cells in comparison to WT.

The combined inhibition had a significant impact on cellular phenotypes in the treated cells compared to single agent alone, most notably related to the cell cycle. While treatment with JQ1 or OTX-015 resulted in a moderate increase in cells in the G1 phase, treatment with crizotinib resulted in a moderate accumulation of cells in G2. However, when the treatments were combined, this manifested as a significant increase in numbers of cells arrested in G2/M phase, as demonstrated by FACS analysis and reduced expression levels of CDK2 and cyclin B1. It has been previously shown that in human alveolar rhabdomyosarcoma, crizotinib caused a dose-dependent G2/M cell growth arrest [24]. In addition to the effects on the cell cycle, we find that the combined treatment results in an increase in apoptosis, when compared to single agent alone. Finally, the in vivo studies using an orthotopic mouse model of NF2-associated schwannomatosis demonstrated that a combining crizotinib and OTX-015 was more effective in reducing tumor size, when compared to the either agent alone, thus confirming the synergistic effect of OTX-015 and crizotinib. Whether the cellular phenotypes of the combined treatment are similar in vitro and in vivo remains to be determined. It should be noted that the observed synergy between BET and FAK inhibition was observed in the NF2-deficient Schwann cells and, to a lesser extent, in the NF2 wildtype cells. Importantly, the synergy in the NF2-deficient cells was evident across a wide range of drug treatment concentrations, indicating the existence of a therapeutic window that will need to be defined in future pre-clinical studies.

There are several ongoing trials with BETi’s as single agents and as combinations including the assessment of EP31670 (NEO2734) [25], in patients with castration-resistant prostate cancer, NUT midline carcinoma and other advanced solid tumors (NCT05488548). Another trial underway is a phase 1 study of the BET inhibitors BMS-986158 and BMS-986378 in pediatric cancers (NCT03936465). In the case of combination trials, there are several examples including the comBET trial testing the BET inhibitor ZEN-3694 in combination with talazoparib, an orally available poly ADP ribose polymerase (PARP) inhibitor, against advanced solid tumors (NCT05327010). The combination of ZEN-3694 in combination with entinostat, a histone deacetylase inhibitor, for advanced solid tumors and lymphomas (NCT05053971), in combination with binimetinib, a selective MEK inhibitor, against advanced tumors with RAS mutations or triple negative breast cancer (NCT05111561) and several others.

Conclusions

NF2-deficient Schwann and Schwannoma cells are sensitive to BET and FAK inhibition. Our results suggest that the combination of JQ1 and crizotinib exhibits synergy and that this is enhanced in NF2-null Schwann cells. This study supports the idea of testing BET inhibitors in combination with FAK inhibition as a therapeutic option for NF2-deficient schwannoma, which is highly feasible given the current landscape of clinical trial assessing BET inhibitors in combination with other drugs.

Methods

Cell lines

Human-derived Schwann cells were a generous gift from Dr. Margaret Wallace (U. of Florida, USA). The generation of hSC2λ-NF2−/− was carried out by targeting both NF2 alleles using CRISPR/Cas9 mediated genome editing. hSC2λ cells were transfected with 2 independent pX459 plasmids containing guide sequences targeting exon 1 of the NF2 gene. gRNA-1 = GAACTCCATCTCGGCGTCCA, gRNA-2 = GATCCTCACGGTGAACGTCT. Cells were selected with puromycin (0.25 µg/mL) for 48 h and single cell clones picked. Multiple clones were expanded and analyzed by genomic sequencing and western blotting for loss of NF2 protein expression. HEI-193 cells are immortalized human schwannoma cells from with a mutation that inactivates NF2 as previously described [26]. SC4-Luc Nf2-null mouse Schwann cells were used for generating the orthotopic mouse model. All cell lines were cultured in Dulbecco’s Modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin.

Chemicals, reagents, and antibodies

Crizotinib (PF-02341066) and OTX-015 (birabresib), were purchased from MedChem Express (MCE, Monmouth Junction, USA). JQ1 was purchased from Selleck Chemicals. Antibodies used in western blotting assays were purchased from Cell Signaling Technology Inc. (Danvers, USA): FAK (#3285), CDK2 (#2546) and the Cell Cycle Phase Determination Antibody Sampler Kit (#17498). All antibodies were diluted to the optimum concentration, according to the manufacturer’s instructions (1:1000 dilution).

Western Blot analysis

Cells were seeded in 10 cm dishes (10,000 cells/plate) and treated 24 h later with JQ1, OTX015, and/or crizotinib for 48 h, at the concentrations indicated in the Figures. Cells treated with DMSO were used as controls. Cells were harvested by trypsinization and washed with PBS before lysis with RIPA lysis buffer containing protease inhibitor [27]. Blots were developed using ECL Prime chemiluminescent developers and exposed using the Bio-Rad ChemiDoc Imaging System.

Quantitative real-time PCR (qPCR)

RNA was extracted using the RNeasy Mini Kit (Qiagen, Germany. Cat 74104) following the manufacturer’s instructions. 1 ug of mRNA was retrotranscribed into cDNA from hSC2λ-WT, hSC2λ-NF2−/−, and HEI-193 cell lines. Reverse transcription was done using SuperScript II Reverse Transcriptase (Thermo Fisher Scientific) following the manufacturer’s protocol. PCR was performed using SYBER-green. Primers used for RT-qPCR were designed and validated in our laboratory. Results are shown as n-fold differences in the target gene expression relative to the housekeeping gene (actin).

Chromatin Immunoprecipitation and RT-qPCR analysis

After extraction of genomic DNA, the chromatin was sheared using the truCHIP Chromatin Shearing Kit with Formaldehyde (Covaris, USA. Cat PN520154) following the manufacturer’s adherent cells protocol. Next, samples were sonicated in a Covaris M220 according to the High cell Chromatin Shearing Protocol. Sonication was assessed using non-sonicated samples as controls. Next, sonicated chromatin was diluted in Covaris 2X IP Dilution buffer and the BRD4 CHIP antibody and Dynabeads protein G were added. Following overnight incubation at 4 °C, the beads were washed with the indicated buffer and incubated at 65 °C overnight to reverse crosslink. The next day RNase ONE was added and samples were incubated for 1 h at 37 °C followed by the addition of proteinase K and a 5-h incubation at 65 °C. DNA was purified using Qiagen PCR Purification Kit (Qiagen, Germany: Cat 28104). For the qPCR, input samples were used to make 3 stock tubes of 10 ng/ul, 1 ng/ul, and 0.1 ng/ul DNA as controls, and IP samples were run in triplicate as previously described.

Cell proliferation assays

10,000 cells/well were plated in 24-well dishes in triplicate and 24 h later, media was removed and replaced with media containing the corresponding compound was replaced: DMSO, JQ1 (0.5 µM), crizotinib (0.5 µM), OTX-015 (0.5 µM) or combinations as indicated in the figures. The plate was placed into the Incucyte® S3 (Sartorius, USA.) and incubated at 37 °C at 5% CO2. Each well was analyzed every 12 h, taking 16 images at a 10x objective for 4 days. To analyze cell proliferation, The Incucyte® analysis software (Adherent Cell-by-Cell) was used to determine the phase object count per image. The data shown are the mean of 3 independent experiments. Error bars = SD.

Apoptosis assays

10,000 cells/well were plated in 24-well dishes in triplicate and 24 h later, media was removed and replaced with media containing the corresponding compound: DMSO, JQ1 (0.5 µM), crizotinib (0.5 µM), OTX-015 (0.5 µM) or combinations as indicated in the Figures. The cells were then harvested for analysis by FACS using an FITC-Annexin V apoptosis detection kit (BioLegends 640914), according to the manufacturer’s instructions. The data shown are the mean of 3 independent experiments. Error bars = SD.

Drug interaction analysis

500 hSC2λ-WT and hSC2λ-NF2−/− cells were seeded in a 384 well-plate (Greiner Bio One. Cat 781080), white plate in duplicate. 24 h later cells were treated with JQ1, crizotinib or their combination using the Echo 650 Series Acoustic Liquid Handlers (Beckman Coulter, Inc.). Seventy two hours after treatment, cell viability was measured by luminescent ATP-dependent assay (CellTiter-Glo, Progmega, Inc. Madison, WI, USA). Synergism or antagonism was determined with the SynergyFinder R package [28]. A CI > 10 was considered synergistic, CI < 1 was antagonistic, and CI = 1–10 was additive. All data represent the combined data from 3 independent experiments.

Animal experiments

All animal experiments complied with NIH guidelines and were approved by the Moffitt Cancer Center IACUC under protocol IS00008425. SC4-luc cells were resuspended in PBS at a concentration of 1 × 105 cells/uL. 6-week-old NSG mice (Jackson Labs) were anesthetized, and 5 × 105 cells were injected into the sciatic nerve sheath by intraneural injection. Tumor growth was monitored using IVIS 200 imager, drug treatment was started when tumors reached a threshold value of 1 × 106 photons/sec. Mice were randomized into 4 groups: vehicle (10% DMSO, 40% PEG300, 5% Tween-80, and 45% saline), OTX-015 (5 mg/kg every other day), crizotinib (0.5 mg/kg every other day) or combination of OTX-015 /crizotinib (5/0.5 mg/kg, every other day). Mice were sacrificed after 17 days of treatment, tumors were isolated and weighed at the conclusion, as described.

Statistical analysis

Statistical analysis of data was performed using GraphPad Prism (version 6). Unpaired student’s t-test was calculated to determine the significance of the results from the tumor weight studies. For the cell growth analysis, we used a two way ANOVA to determine the interaction of time and treatment. Individual statistical methods are described in the respective figure legends.