P-13 lncRNA co-expression network in Neuroblastoma
by 陳輔卿, 2015-09-14 19:34, 人氣(794)

lncRNA co-expression network in Neuroblastoma


Divya Sahu1,2,3, Chia-Lang Hsu5, Hsueh-Fen Juan4,5,6, Hsuan-Cheng Huang3*


1Institute of Bioinformatics and Systems Biology, National Chiao Tung University, Hsinchu, Taiwan , 2Bioinformatics, Institute of Information Science, Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan, 3Institute of Biomedical Informatics, Center for Systems and Synthetic Biology, National Yang Ming University, Taipei, Taiwan, 4Institute of Molecular and Cellular Biology, 5Department of Life Science, 6Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei , Taiwan



Neuroblastoma (NB) is one of the most common childhood tumor of sympathetic nervous system. Amplification of MYCN proto-oncogene is a strong prognostic marker of this disease indicating poor survival rate. It encodes a transcription factor N-Myc, which either induce or repress numerous target genes involved in cell cycle, neuron differentiation, cell proliferation, metabolism and apoptosis in NB. Currently, several long non-coding RNAs (lncRNAs) observed to be aberrantly expressed in cancer which modulate the fundamental biological process. Co-expression analysis has been widely used to find the degree of association between genes, which might share similar functions. In this study, we aim to find differentially co-expressed lncRNA and mRNA regulated by N-Myc in high risk MYCN - amplified NB.


Materials and Methods:

GEO microarray dataset was used with R program and Bioconductor package (limma) for differential expression analysis. Gene Ontology for differentially expressed genes was performed with Cluego, TopGene and DAVID. Association between lncRNA and coding gene was determined by spearman correlation coefficient (SCC) and Fischer’s Z transformation. Positively correlated z-score with cut off ≥ 3.0 was used to build amplified and non-amplified network in Cytoscape. To confirm the results, 3 other microarray and RNA-seq dataset were taken. SCC and mutual rank were calculated for the whole lncRNA and mRNA. Moreover, Kaplan-Meier survival analysis was performed in two cohort of neuroblastoma patients (n=86 and n= 493, respectively).



Of the neuroblastoma cohort, 47 patients (14 were MYCN amplified and 33 non-amplified) we found 592 mRNA and 13 lncRNA to be differentially expressed with fold change ≥ 2 and p-value ≤ 0.05. Gene Ontology were found to be significantly enriched. The co-expression network analysis predicts a denser amplified network compared to the non-amplified. Moreover, with SCC cut off ≥ 0.8, we found 39 co-expressed pairs, of which SNHG1 (lncRNA) and TAF1D (mRNA) were significantly up-regulated and highly correlated. In another cohort of 86 patients, the pair shown to be significantly up-regulated in stage 4 of amplified compared to non-amplified. In concordance, the high expression of SNHG1 and TAF1D (n = 247) is also associated with poor patient survival. The 18 years survival probability for both of the genes in stage 4 of amplified (n= 65) reduced to 20-30%, and interestingly of the males (n=31) is lower than the females (n= 31). Furthermore, ChIP-seq data also reveals the binding of MYCN on the promoter of SNHG1 and TAF1D.


Taken together, our finding indicates that SNHG1 and TAF1D are significantly up-regulated and highly correlated pair which are regulated by N-Myc in the amplified NB. High expression of the both genes linked to poor patient survival. Thus, the pair can be served as novel biomarkers for neuroblastoma intervention