Next-generation RNA sequencing reveals differential expression of MYCN target genes and suggests the mTOR pathway as a promising therapy target in MYCN-amplified neuroblastoma.

In many cancer types, MYC proteins are known to be master regulators of the RNA-producing machinery. Neuroblastoma is a tumor of early childhood characterized by heterogeneous clinical courses. Amplification of the MYCN oncogene is a marker of poor patient outcome in this disease. Here, we investigated the MYCN-driven transcriptome of 20 primary neuroblastomas with and without MYCN amplification using next-generation RNA sequencing and compared the results to those from an in vitro cell model for inducible MYCN (SH-EP MYCN-ER). Transcriptome sequencing produced 30-90 million mappable reads for each dataset. The most abundant RNA species was mRNA, but snoRNAs, pseudogenes and processed transcripts were also recovered. A total of 223 genes were significantly differentially expressed between MYCN-amplified and single-copy tumors. Of those genes associated with MYCN both in vitro and in vivo, 32% of MYCN upregulated and 37% of MYCN downregulated genes were verified either as previously identified MYCN targets or as having MYCN-binding motifs. Pathway analyses suggested transcriptomal upregulation of mTOR-related genes by MYCN. MYCN-driven neuroblastomas in mice displayed activation of the mTOR pathway on the protein level and activation of MYCN in SH-EP MYCN-ER cells resulted in high sensitivity toward mTOR inhibition in vitro. We conclude that next-generation RNA sequencing allows for the identification of MYCN regulated transcripts in neuroblastoma. As our results suggest MYCN involvement in mTOR pathway activation on the transcriptional level, mTOR inhibitors should be further evaluated for the treatment of MYCN-amplified neuroblastoma.

Characterization of pancreatic glucagon-producing tumors and pituitary gland tumors in transgenic mice overexpressing MYCN in hGFAP-positive cells.

Amplification or overexpression of MYCN is involved in development and maintenance of multiple malignancies. A subset of these tumors originates from neural precursors, including the most aggressive forms of the childhood tumors, neuroblastoma and medulloblastoma. In order to model the spectrum of MYCN-driven neoplasms in mice, we transgenically overexpressed MYCN under the control of the human GFAP-promoter that, among other targets, drives expression in neural progenitor cells. However, LSL-MYCN;hGFAP-Cre double transgenic mice did neither develop neural crest tumors nor tumors of the central nervous system, but presented with neuroendocrine tumors of the pancreas and, less frequently, the pituitary gland. Pituitary tumors expressed chromogranin A and closely resembled human pituitary adenomas. Pancreatic tumors strongly produced and secreted glucagon, suggesting that they derived from glucagon- and GFAP-positive islet cells. Interestingly, 3 out of 9 human pancreatic neuroendocrine tumors expressed MYCN, supporting the similarity of the mouse tumors to the human system. Serial transplantations of mouse tumor cells into immunocompromised mice confirmed their fully transformed phenotype. MYCN-directed treatment by AuroraA- or Brd4-inhibitors resulted in significantly decreased cell proliferation in vitro and reduced tumor growth in vivo. In summary, we provide a novel mouse model for neuroendocrine tumors of the pancreas and pituitary gland that is dependent on MYCN expression and that may help to evaluate MYCN-directed therapies.

Robust selection of cancer survival signatures from high-throughput genomic data using two-fold subsampling.

Identifying relevant signatures for clinical patient outcome is a fundamental task in high-throughput studies. Signatures, composed of features such as mRNAs, miRNAs, SNPs or other molecular variables, are often non-overlapping, even though they have been identified from similar experiments considering samples with the same type of disease. The lack of a consensus is mostly due to the fact that sample sizes are far smaller than the numbers of candidate features to be considered, and therefore signature selection suffers from large variation. We propose a robust signature selection method that enhances the selection stability of penalized regression algorithms for predicting survival risk. Our method is based on an aggregation of multiple, possibly unstable, signatures obtained with the preconditioned lasso algorithm applied to random (internal) subsamples of a given cohort data, where the aggregated signature is shrunken by a simple thresholding strategy. The resulting method, RS-PL, is conceptually simple and easy to apply, relying on parameters automatically tuned by cross validation. Robust signature selection using RS-PL operates within an (external) subsampling framework to estimate the selection probabilities of features in multiple trials of RS-PL. These probabilities are used for identifying reliable features to be included in a signature. Our method was evaluated on microarray data sets from neuroblastoma, lung adenocarcinoma, and breast cancer patients, extracting robust and relevant signatures for predicting survival risk. Signatures obtained by our method achieved high prediction performance and robustness, consistently over the three data sets. Genes with high selection probability in our robust signatures have been reported as cancer-relevant. The ordering of predictor coefficients associated with signatures was well-preserved across multiple trials of RS-PL, demonstrating the capability of our method for identifying a transferable consensus signature. The software is available as an R package rsig at CRAN (http://cran.r-project.org).