Analysis Of MicroRNA Expression Signatures And Related Factors In Gastrointestinal Stromal Tumors

BackgroundGastrointestinal stromal tumors (GISTs), the most common mesenchymal neoplasm of the digestive tract, usually originate in the wall of the stomach or small intestine and those arising from extragastrointestinal sites are termed extragastrointestinal stromal tumors (EGISTs), the origin of which still remains controversial. Someone hypothesizes that EGISTs are in actual fact mural GISTs with extensive extramural growth resulting in eventual loss of their connection with the gut wall. Presently, numerous studies have compared GISTs with EGISTs in multiaspects including clinicopathologic feature, biological behavior and genetic molecular character.MicroRNAs, recently identified non-protein coding single-stranded-22-nucleotide RNAs, negatively regulate gene expression either by translational inhibition or by mRNA degradation and play a crucial role in pathophysiological processes including tumorigenesis and progression. They have attracted more and more attention for an enormous potential to be biomarkers for diagnosis and prognosis or therapeutic targets. So far, only a few studies have paid attention to miRNA expression signatures of GISTs and comparison between those of EGISTs with GISTs has not yet been reported.ObjectiveTo evaluate the relationships between GISTs and EGISTs concerning miRNA expression and filter out the relevant miRNAs with an aim to help the molecular classifications of GISTs.Methods1. Genomic DNA was extracted from formalin-fixed paraffin-embedded (FFPE) tissues and then KIT and PDGFRA mutations of 38 samples were analyzed by PCR amplification and direct sequencing.2. FFPE tissues from 13 GISTs and 7 EGISTs were evaluated for miRNA expression profile using Agilent microarrays representing 866 human miRNAs.3. Unsupervised hierarchical clustering analysis was used to investigate the differential miRNA expression signatures and possible related factors. Student's t-test and significance analysis of microarrays (SAM) were applied to filter out the differentially expressed miRNAs.4. The putative targeted genes of selected miRNAs were predicted using bioinformatics websites.Results1. KIT mutations were identified in 34 samples, involving 32 mutations in exon 11 and 2 tandemduplications in exon 9. The mutational format of exon 11 included deletions, insertions and point mutations. PDGFRA mutation was detected in only one sample.2. Unsupervised hierarchical clustering analysis revealed that both GISTs and EGISTs could be divided into 3 clusters (Cluster A, B and C) majorly related to loss of 14q and anatomic sites (stomach and small intestine).3. The expression of 32 miRNAs clustered in 14q32 was found to be up-regulated in samples of cluster A and might serve as tumor suppressor genes.4.17 miRNAs were found to be down-regulated in samples of cluster B and most of them were functionally mapped to KIT/PDGFRA signalling and G1/S-phase transition of the cell cycle and might involve in tumorigenesis and progression.5. The expression levels of 5 miRNAs were correlated with tumor risk and 2 of them were found to be down-regulated in high-risk group. Conclusions1. GISTs and EGISTs share similar miRNA expression signatures majorly related to loss of 14q and anatomic location. This may provide further evidence to support "extensive extramural growth" hypothesis and help the molecular classifications of GISTs.2. FFPE derived miRNAs are acceptable for profiling using a microarray platform.3. KIT and PDGFRA mutations are the most common genetic events in GISTs.