Bioinformatic analysis
Data processing and analysis of gene expression profile
The public microarray dataset GSE90121, which was obtained based on the Affymetrix GPL570 platform Affymetrix Human Genome U133 Plus 2.0 Array), was downloaded from the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/), [14]. This dataset was deposited by Kaplan D et al., containing information from human NB SK-N-AS metastatic subpopulations isolated after in vivo selection, aimed to identify genomic signatures that regulate metastasis and candidate therapeutics for NB patients. A total of 16 samples were included in the current dataset, including 12 metastatic samples and 4 primary samples. Robust Multi-array Average (RMA) affy package of Bioconductor was used to adjust the raw data. The processed gene expression data was then filtered to include those probe sets with annotations which reference the new version annotation files. To identify DEGs, we used the Linear Models (Microarray Data package in Bioconductor) to compare the expression levels of genes between the metastatic group and the localized tumor group[15].
An adjusted p-value of <0.05 and a |log2FC (fold change) | of ≥2 was used as the threshold.
Functional enrichment analysis of DEGs
Database for annotation, visualization and integrated discovery (DAVID) integrates a set of functional enrichment tools to distinguish functional genes underlying diseases processing (http://david.abcc.ncifcrf.gov/)[16]. GO (Gene Ontology) function and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analyses of DEGs were performed based on DAVID. P value < 0.05 and count ≥ 2 was regarded as statistically significant differences.
Protein–protein interaction network construction by STRING
We used the Search Tool for the Retrieval of Interacting Genes (STRING), an online tool and biological database for prediction of interactions between proteins, to construct the PPI network [17]. According to our analysis based on STRING, score（median confidence）> 0.15 was the standard of PPIs of DEGs selections. The Cytoscape software was used to visualize the PPI network.[18]. The proteins that have many interaction partners constitute the extremely important nodes in the PPI network. These proteins were defined as the hub proteins in this study. To identify such hub proteins in the PPI network, we utilized six bioinformatic tools, namely Closeness, Degree, EPC, MNC, Radiality, and Stress centrality. Sub-network analysis was then conducted to help us discover the outstanding genes.
NB patients, tissue samples and cell lines
Primary tumor tissues were obtained from 9 NB patients with bone marrow metastasis and 10 NB patients without bone marrow metastasis who had undergone tumor resection surgery at the Affiliated Hospital of Qingdao University. None of the included patients was treated with chemotherapy, hormonal therapy or radiotherapy before the tumor resection surgery. Written informed consent was obtained from all the participants. The current research was conducted with the permission of the Medical Ethical Committee of Affiliated Hospital of Qingdao University (Qingdao, China).
Human NB cell line SH-SY5Y was kindly provided by Professor Xiao from the Guizhou Medical University. The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) contained 10% fetal bovine serum (FBS, Hyclone, USA) under the conditions of 37°C, 5% CO2.
NB serum samples and Quantitative real-time PCR
The expression of TAC1, PTGS2 and FGF1 was examined by quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR). The collected tissues were immediately frozen at -80°C after surgery. Total RNA was extracted from cancer tissues by TRIzol reagent (Invitrogen Life Technologies). Then the extracted RNA was transcribed into cDNA using random primers and analyzed with an ABI 7000 Real-Time PCR System (Applied Biosystems). PCR primers were as follows: TAC1 primers: (forward) 5′-TGA TCT GAA TTA CTG GTC CGA CT-3′ and (reverse) 5′-TCC GGC AGT TCC TCC TTG A-3′; PTGS2 primers: (forward) 5′-TAA GTG CGA TTG TAC CCG GAC-3′ and (reverse) 5′-TTT GTA GCC ATA GTC AGC ATT GT-3′; FGF1 primers: (forward) 5′-CTC CCG AAG GAT TAA ACG ACG-3′ and (reverse) 5′-GTC AGT GCT GCC TGA ATG CT-3′; GAPDH primers (forward) 5′-CAG CGA CAC CCA CTC CTC-3′ and (reverse) 5′-TGA GGT CCA CCA CCC TGT-3′. Reactions were performed in triplicate using SYBR Green master mix (TaKaRa, Japan) and normalized to GAPDH mRNA level using the ΔΔCt method.
ELISA
To quantify levels of S100A9 in plasma, ELISA was performed as previously described [19] By using human S100A9 (JYM0539Hu, JYM, China) ELISA kits, S100A9 in plasma of the NB patients were detected according to the manufacturer’s recommended procedure.
IHC staining
In brief, the formalin fixed, paraffin-embedded tissues sections were deparaffinized, rehydrated and boiled in 0.01 M citrate buffer for 10 min, then incubated with 0.3% H2O2 in methanol to block endogenous peroxidase activity. Then the sections were incubated with the anti-S100A9 antibody (Cell Signaling Technology, USA), followed by incubation with secondary antibody tagged with the peroxidase enzyme for 30min at room temperature, finally visualized with 0.05% DAB (3,3'-diaminobenzidine) until the desired brown reaction product was obtained. All slides were observed under a OLYMPUS BX41 Microscope, and representative photographs were taken.
Construction of plasmids and establishment of stably transfected cells
SBI-piggyBac vector, GST-S100A9, were kindly provided by Professor T.C. He from the University of Chicago. To construct an S100A9 overexpression plasmid, the complete coding sequence of human S100A9 gene was subcloned into the SBI-piggyBac vector. For S100A9 silencing, siRNAs targeting human S100A9 with the sequences of 5’- GCAAGACGAUGACUUGCAA -3’ and 5’- UUGCAAGUCAUCGUCUUGC -3’were synthesized and assembled into the SBI-piggyBac vector, resulting in SBI-siS100A9. After transfecting SH-SY-5Y NB cells with the constructed plasmids, the stably transfected cells were selected by incubation with puromycin for one week. The stably transfected cell lines, namely Control (SH-SY5Y transfected with SBI empty vector), S100A9(SH-SY5Y transfected with SBI-S100A9) and siS100A9(SH-SY5Y transfected with SBI-siS100A9) briefly.
Cell viability assay
The viability of SH-SY5Y cells was assessed by 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphe-nyltrazolium bromide (MTT) assay. Briefly, stably transfected SH-SY5Y cells were seeded in 96-well plates (1000 cells/well). The cells were incubated in DMEM supplemented with 1% FBS for 24, 48, 72, 96 and 120 h, then incubated with MTT reagent (Progema, Madison, WI, USA, 20 µL /well) for another 4 h at 37°C to allow the formation of formazan. After that, 100 µL of dimethyl sulfoxide was added into the cell culture medium for another 10-min incubation at room temperature. At last, in every day of the next five days, a microplate reader (Bio-rad, iMark) was used to measure the absorbance at 492 nm of each well. The experiment included three independent replicates for each sample.
Colony formation assay
Exponentially growing stably transfected SH-SY5Y cells were seeded at a low density (100 cells/well) in cell culture medium containing 1% FBS in 6-well plates. The cells were allowed to grow for about 10 days to form colonies. The culture medium was refreshed every 3-4 days. Crystal violet was used to stained the colonies. Colony numbers from 3 wells were used to calculated the average colony number.
Scratch wound healing assay
The scratch wound healing assay was performed as described previously [20,21]. Briefly, stably transfected SH-SY5Y cells were seeded in 6-well plates and grown to ~90% confluency. Then, sterile micro-pipette tips were used to scratch the monolayer formed by SH-SY5Y cells to create a wound. After that, the medium (DMEM containing 1% FBS) was refreshed every day to remove the floating cells. Bright field microscopy was used to monitor the wound healing status at 24 h, 48 h, and 72 h after the wound was created. Each assay was repeated three times. imageJ software was used to calculate the wound healing ratio.
Transwell invasion assay
A chamber coated with non-type I-collagen (Millipore, USA) was used for the Transwell assay. The upper chamber coated with ECM gel (Sigma, USA) was filled with 400 µL of serum-free DMEM and seeded with exponentially growing stably transfected SH-SY5Y cells (1×104 cells). The lower chamber was filled with 500 µL of DMEM supplemented with 20% FBS, which served as a chemoattractant. After 24 h of incubation, the cells migrated across the Transwell membrane were dried, fixed with methanol, and then stained with hematoxylin-eosin (H&E). Cotton swabs were used to remove the cells on the upper surface of the Transwell membrane. At last, an inverted microscope (×100 magnification) was used to count the number of cells migrated across the Transwell membrane. Five randomly-selected fields were examined to obtain the mean value of the number of cells migrated across the Transwell membrane. The experiment was repeated three times.
Statistical analysis
All data are presented as means ± standard deviations. T test was performed in the GraphPad Prism software to determine the statistical significance of differences between groups. A P value of less than 0.05 was considered statistically significant.