HIF1A (Hypoxia-Inducible-Factor 1A) expression in solid tumors is relevant to establish resistance to therapeutic approaches. The use of compounds direct against hypoxia signaling and HIF1A does not show clinical efficiency because of changeable oxygen concentrations in solid tumor areas. The identification of HIF1A targets expressed in both normoxia and hypoxia and of HIF1A/hypoxia signatures might meliorate the prognostic stratification and therapeutic successes in patients with high-risk solid tumors. In this study, we conducted a combined analysis of RNA expression and DNA methylation of neuroblastoma cells silenced or unsilenced for HIF1A expression, grown in normoxia and hypoxia conditions. The analysis of pathways highlights HIF-1 (heterodimeric transcription factor 1) activity in normoxia in metabolic process and HIF-1 activity in hypoxia in neuronal differentiation process. HIF1A driven transcriptional response in hypoxia depends on epigenetic control at DNA methylation status of gene regulatory regions. Furthermore, low oxygen levels generate HIF1A-dependent or HIF1A-independent signatures, able to stratify patients according to risk categories. These findings may help to understand the molecular mechanisms by which low oxygen levels reshape gene signatures and provide new direction for hypoxia targeting in solid tumor.

Novel druggable targets have been discovered in neuroblastoma (NB), paving the way for more effective treatments. However, children with high-risk NB still show high mortality rates prompting for a search of novel therapeutic options. Here, we aimed at repurposing FDA-approved drugs for NB treatment by performing a high-content screening of a 349 anticancer compounds library. In the primary screening, we employed three NB cell lines, grown as three-dimensional (3D) multicellular spheroids, which were treated with 10 μmol/L of the library compounds for 72 hours. The viability of 3D spheroids was evaluated using a high-content imaging approach, resulting in a primary hit list of 193 compounds. We selected 60 FDA-approved molecules and prioritized drugs with multi-target activity, discarding those already in use for NB treatment or enrolled in NB clinical trials. Hence, 20 drugs were further tested for their efficacy in inhibiting NB cell viability, both in two-dimensional and 3D models. Dose-response curves were then supplemented with the data on side effects, therapeutic index, and molecular targets, suggesting two multiple tyrosine kinase inhibitors, ponatinib and axitinib, as promising candidates for repositioning in NB. Indeed, both drugs showed induction of cell-cycle block and apoptosis, as well as inhibition of colony formation. However, only ponatinib consistently affected migration and inhibited invasion of NB cells. Finally, ponatinib also proved effective inhibition of tumor growth in orthotopic NB mice, providing the rationale for its repurposing in NB therapy. Mol Cancer Ther; 17(7); 1405–15. ©2018 AACR.

Neuroblastoma (NBL) is a heterogeneous tumor characterized by a wide range of clinical manifestations. A high tumor cell differentiation grade correlates to a favorable stage and positive outcome. Expression of the hypoxia inducible factors HIF1-α (HIF1A gene) and HIF2-α (EPAS1 gene) and/or hypoxia-regulated pathways has been shown to promote the undifferentiated phenotype of NBL cells. Our hypothesis is that HIF1A and EPAS1 expression represent one of the mechanisms responsible for the lack of responsiveness of NBL to differentiation therapy. Clinically, high levels of HIF1A and EPAS1 expression were associated with inferior survival in two NBL microarray datasets and patient subgroups with lower expression of HIF1A and EPAS1 showed significant enrichment of pathways related to neuronal differentiation. In NBL cell lines, the combination of all-trans retinoic acid (ATRA) with HIF1A or EPAS1 silencing led to an acquired glial-cell phenotype and enhanced expression of glial-cell differentiation markers. Furthermore, HIF1A or EPAS1 silencing might promote cell senescence independent of ATRA treatment. Taken together, our data suggest that HIF inhibition coupled with ATRA treatment promotes differentiation into a more benign phenotype and cell senescence in vitro. These findings open the way for additional lines of attack in the treatment of NBL minimal residue disease.

The contribution of coding mutations to oncogenesis has been largely clarified, whereas little is known about somatic mutations in noncoding DNA and their role in driving tumors remains controversial. Here, we used an alternative approach to interpret the functional significance of noncoding somatic mutations in promoting tumorigenesis. Noncoding somatic mutations of 151 neuroblastomas were integrated with ENCODE data to locate somatic mutations in regulatory elements specifically active in neuroblastoma cells, nonspecifically active in neuroblastoma cells, and nonactive. Within these types of elements, transcription factors (TF) were identified whose binding sites were enriched or depleted in mutations. For these TFs, a gene expression signature was built to assess their implication in neuroblastoma. DNA- and RNA-sequencing data were integrated to assess the effects of those mutations on mRNA levels. The pathogenicity of mutations was significantly higher in transcription factor binding site (TFBS) of regulatory elements specifically active in neuroblastoma cells, as compared with the others. Within these elements, there were 18 over-represented TFs involved mainly in cell-cycle phase transitions and 15 under-represented TFs primarily regulating cell differentiation. A gene expression signature based on over-represented TFs correlated with poor survival and unfavorable prognostic markers. Moreover, recurrent mutations in TFBS of over-represented TFs such as EZH2 affected MCF2L and ADP-ribosylhydrolase like 1 expression, among the others. We propose a novel approach to study the involvement of regulatory variants in neuroblastoma that could be extended to other cancers and provide further evidence that alterations of gene expression may have relevant effects in neuroblastoma development. Significance: These findings propose a novel approach to study regulatory variants in neuroblastoma and suggest that noncoding somatic mutations have relevant implications in neuroblastoma development.

The genetic aetiology and the molecular mechanisms that characterize high‐risk neuroblastoma are still little understood. The majority of high‐risk neuroblastoma patients do not take advantage of current induction therapy. So far, one of the main reasons liable for cancer therapeutic failure is the acquisition of resistance to cytotoxic anticancer drugs, because of the DNA repair system of tumour cells. PARP1 is one of the main DNA damage sensors involved in the DNA repair system and genomic stability. We observed that high PARP1 mRNA level is associated with unfavourable prognosis in 3 public gene expression NB patients’ datasets and in 20 neuroblastomas analysed by qRT‐PCR. Among 4983 SNPs in PARP1 , we selected two potential functional SNPs. We investigated the association of rs907187, in PARP1 promoter, and rs2048426 in non‐coding region with response chemotherapy in 121 Italian patients with high‐risk NB. Results showed that minor G allele of rs907187 associated with induction response of patients ( P = .02) and with decrease PARP1 mRNA levels in NB cell line ( P = .003). Furthermore, rs907187 was predicted to alter the binding site of E2F1 transcription factor. Specifically, allele G had low binding affinity with E2F1 whose expression positively correlates with PARP1 expression and associated with poor prognosis of patients with NB. By contrast, we did not find genetic association for the SNP rs2048426. These data reveal rs907187 as a novel potential risk variant associated with the failure of induction therapy for high‐risk NB.