Isoxazole compound ML327 blocks MYC expression and tumor formation in neuroblastoma

Abstract

Neuroblastomas are the most common extracranial solid tumors in children and arise from the embryonic neural crest. MYCN-amplification is a feature of ∼30% of neuroblastoma tumors and portends a poor prognosis. Neural crest precursors undergo epithelial-to-mesenchymal transition (EMT) to gain migratory potential and populate the sympathoadrenal axis. Neuroblastomas are posited to arise due to a blockade of neural crest differentiation. We have recently reported effects of a novel MET inducing compound ML327 (N-(3-(2-hydroxynicotinamido) propyl)-5-phenylisoxazole-3-carboxamide) in colon cancer cells. Herein, we hypothesized that forced epithelial differentiation using ML327 would promote neuroblastoma differentiation. In this study, we demonstrate that ML327 in neuroblastoma cells induces a gene signature consistent with both epithelial and neuronal differentiation features with adaptation of an elongated phenotype. These features accompany induction of cell death and G1 cell cycle arrest with blockage of anchorage-independent growth and neurosphere formation. Furthermore, pretreatment with ML327 results in persistent defects in proliferative potential and tumor-initiating capacity, validating the pro-differentiating effects of our compound. Intriguingly, we have identified destabilization of MYC signaling as an early and consistent feature of ML327 treatment that is observed in both MYCN-amplified and MYCN-single copy neuroblastoma cell lines. Moreover, ML327 blocked MYCN mRNA levels and tumor progression in established MYCN-amplified xenografts. As such, ML327 may have potential efficacy, alone or in conjunction with existing therapeutic strategies against neuroblastoma. Future identification of the specific intracellular target of ML327 may inform future drug discovery efforts and enhance our understanding of MYC regulation.

Keywords: ML327; MYCN; epithelial-to-mesenchymal transition; neural crest; neuroblastoma.

Figure 1. ML327 induces cell death and cell cycle arrest in neuroblastomas

(A) Light microscopy (20×) demonstrates induction of an elongated phenotype in neuroblastoma cells treated with ML327 (10 µM). (B) ML327 (10 µM) inhibits adherent colony formation in MYCN-amplified and MYCN-single copy neuroblastoma cell lines. (C) Representative concentration-dependent cellular viability plot for BE(2)-C cells with an estimated IC50 of 4 µM. (D) Time course plot of cellular viability in BE(2)-C cells using CCK-8 colorimetric assay. (E) Cell cycle analysis with propidium iodide demonstrates enhanced G1 cell cycle arrest and induction of subG0 phase at 48 h following treatment. (F, G) Anchorage-independent growth and sphere forming frequency (SFF) is blocked in BE(2)-C cells by treatment with ML327 (10 µM).

Figure 2. Effects of ML327 on neuroepithelial differentiation

RNA sequencing analysis of BE(2)-C cells treated with ML327 and vehicle for 7d. (A) GSEA for the neuroblastoma differentiation signature gene set. ES, enrichment score; NES, normalized enrichment score. (B) GSEA for gene ontology gene set for epithelial development. (C, D) RT-PCR survey of neuroblastoma cell lines for the expression of epithelial hallmarks, CDH1 and OCLN. (E) GSEA for gene ontology gene set for neuronal development. (F, G, H) RT-PCR survey of neuroblastoma cell lines for the expression of neuronal markers, NTRK1, NTRK2, and NTRK3.

Figure 3. ML327 blocks MYC signaling in neuroblastoma

(A, B) RNA sequencing demonstrates repression of hallmark MYC target gene sets by GSEA (C) Western blots demonstrate that ML327 inhibits C-MYC expression and N-MYC protein expression levels in MYCN-single and MYCN-amplified neuroblastoma cell lines, respectively. β-actin was used for protein loading control. (D) Immunoblot demonstrating time course of N-MYC, p21/Cip1, p53, and E-cadherin protein expression changes in the presence and absence of ML327. (E) Co-treatment with cycloheximide demonstrates that ML327 fails to alter protein half-life of MYC. (F) RT-PCR demonstrates early changes in MYCN mRNA levels in the presence of ML327. (G) Co-treatment with actinomycin D demonstrates that ML327 fails to alter mRNA half-life.

Figure 4. Pretreatment with ML327 blocks neuroblastoma proliferative potential and tumor-initiating capacity

Pretreating BE(2)-C cells with ML327 (10 µM) decreased proliferative potential as measured by (A) cell viability with CCK-8 assay, (B) adherent clonogenesis, (C) soft-agar colony formation, and (D) sphere-forming frequency. (E) Pretreatment (7d) of BE(2)-C cells with ML327 limits their tumor-initiating capacity in a subcutaneous xenograft model. (F) BE(2)-C cells pretreated with vehicle were injected into the left subscapular region, while ML327-pretreated cells with injected on the right side. Representative images were obtained two weeks following injection.

Figure 5. Growth inhibition of neuroblastoma xenografts by ML327

BE(2)-C subcutaneous xenografts were established and randomized to receive intraperitioneal injection with ML327 (50 mg/kg b.i.d.; N = 9) or vehicle control (N = 8). (A) Tumor volumes were measured daily and (B) explant weights were obtained at time of sacrifice following 2 weeks of treatment. (C) Daily weight measurements demonstrated a 12% weight loss in ML327-treated mice. (D) MYCN mRNA levels were measured by RT-PCR using RNA isolated from ML327 and control tumors. (E) Representative H&E sections were obtained from ML327 and vehicle-treated tumors and demonstrated large areas of necrosis within ML327-treated tumors (10× magnification). (F) Residual, viable neuroblastoma cells in ML327-treated remained poorly differentiated (40× magnification).