Patient samples and laser-capture microdissection
Snap-frozen NPC biopsies were obtained from NPC patients and normal, healthy nasopharyngeal epithelial samples from biopsy-negative cases were used as control. The criteria of clinical staging of NPC samples was based on the 2008 staging system of NPC, which was established in 2008 according to NPC 92 and AJCC staging system [32
]. NPC samples in clinical staging I-IV were used (numbers I to IV, with IV having greater progression). Samples were collected from Xiangya Second Hospital, Central South University. The patients were informed about the sample collection and had signed informed consent forms. Collections and use of tissue samples were approved by the ethical review committees of Xiangya Second Hospital (Additional file 1
). Laser capture microdissection was used to separate the cancer tissues from the normal tissues. Samples were first frozen-sectioned by using a LEICA CM 1900 cryomicrotome. Phase contrast images were acquired using LEICA CTR 6500 microscope.
Total RNA was extracted using Trizol® reagent (Invitrogen) from samples. Two hundred nanograms (200 ng) of total RNA from each sample were used for the follow-up microarray. Poly(A) polymerase (PAP) was used to add a stretch of Poly-A tail to the 3' end of each sequence in total RNA. The Ambion Illumina TotalPrep RNA Amplification Kit was used to synthesize biotinylated cDNA. MicroRNA expression profiling kit contains primers for 1146 human miRNAs. The biotinylated cDNAs were hybridized with microRNA-specific oligonucleotides. The unbinding oligos were washed away and followed by the extension and ligation reaction. Polymerase Chain Reactions (PCR) were performed with fluorescently labelled universal primers, followed by hybridizing of the fluorescently labelled, single-stranded PCR products to capture beads. The fluorescent signals were then detected by Illumina's iScan System. All steps were performed according to Illumina's instructions manual.
1. Multi-Class Dif (RVM-F test)
Raw data from each array were analyzed using Multi-Class Dif (RVM-F test) which is applicable to small sample size analysis for multiple groups. The RVM F-test was applied to screen the dynamic differentially-expressed genes. The detailed methods were applied as previously described [34
2. Gene Ontology (GO)
The TargetScan database was used to predict the target gene of miRNAs. To understand the functions of predicted target genes, we used the ontology classification of genes based on gene annotation and summary information available through DAVID (Database for Annotation, Visualization and Integrated Discovery). The predicted target genes were assigned to functional groups based on molecular function, biological processes and specific pathways.
miRNAs sequence was mapped to the genome in the Sanger database. The Jemboss software was used to examine the alignment of sequences of pre-miRs and putative transcription factor binding sequences. A genome browser database was used to build the relationship of transcription factors and miRNAs network. An adjacency matrix was implemented in Java (programming language) according to the binding of pre-miRNAs and transcription factors. The network's core transcription factor is the most important centre with the biggest degree [37
]. The Pearson correlation analysis [37
] is used to measure the regulatory ability of transcription factors by calculating the correlation between transcription factors and the miRNAs.
The MicroRNA-Gene-Network was based on the interactions of miRNAs and target genes [39
]. Twenty miRNAs of interest were built determined by pathways extracted from KEGG as primary nodes-networks. The significance of relationship of the miRNAs and target genes network was evaluated by the number of nodes in the network with degree greater than 10. In the MicroRNA-Gene-Network, the circle represents gene and the square represents MicroRNA, and their relationship was represented by one edge.
Quantitative reverse transcription-polymerase chain reaction analysis
Total RNA was extracted using Trizol® reagent (Invitrogen) from samples. The primers for RT-PCR to detect miRNA were designed based on the miRNA sequences provided by the Sanger Center miRNA Registry. The primers were synthesized and purified by the Shanghai Gene-Pharma Co. (Shanghai, China). RT reactions were performed by means of the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). Real-time PCR was performed on the BIO-RAD IQTM5 Multicolor Real-Time PCR detection System (Bio-Rad). The qPCR cycle was 98°C for 2 min., 40 cycles of 95°C for 15 sec., 60°C for 30 sec. Final melt-curve analysis (60°-95°C) was included. The standard curve was produced with slopes at approximately -3.32 (~100% efficiency); miRNA PCR quantification used 2ΔΔct method against the U6 for normalization. mRNA PCR quantification used 2ΔΔct method against the GAPDH for normalization. The data are representative of the means of three experiments.