Identification of novel MET fusion transcript amplification in glioblastoma
Recommended Citation
Nuga O, DeCarvalho A, Meng Y, Hasselbach L, Nelson K, Irtenkauf S, and Transou A. Identification of novel MET fusion transcript amplification in glioblastoma. Cancer Res 2018; 78(13)
Document Type
Conference Proceeding
Publication Date
2018
Publication Title
Cancer Res
Abstract
Glioblastoma (GBM) is the most common and aggressive primary CNS malignant tumor with a 15.2% 2 year survival rate for tumors diagnosed between 2008 and 2012. Amplification of MET proto-oncogene has been identified in 4% of glioblastomas, leading to high expression and ligand independent activation in some cases. Our goal was to develop MET-amplified GBM models to study signaling and investigate response to MET inhibitors as a therapy for GBMs. We identified 2 GBM patients (HF3035 and HF3077) with MET gene amplifications after low-pass whole genome DNA sequencing of 13 cases. Fluorescent in situ hybridization (FISH) analyses confirmed heterogeneous MET amplification in HF3035 and HF3077 tumors, in 63.5% and 83.0% of nuclei, respectively. In vitro neurosphere cultures derived from these tumors showed drastic depletion of MET amplicons, to 15.5% MET in HF3035 (P7) and 1.5% in HF3077 (P11). FISH in the metaphase neurosphere spreads showed that MET amplification was extrachromosomal. Interestingly, MET-amplified neurospheres were strongly selected for after intracranial (IC) implant in immunocompromised mice. HF3035 and HF3077 PDX presented MET amplified in high frequency: 79.5-86.5% for HF3035 and 47-65% for HF3077. Met expression levels by RNAseq were congruent with the oscillating gene amplification pattern. In depth RNA sequencing analysis using PRADA has revealed genomic rearrangements involving MET, yielding three novel MET-CAPZA2 fusion transcripts. For both cell lines exon 1 of CAPZA2 was fused to exon 2 of MET, resulting in full length MET coding region, with altered 5' cis-regulatory sequences. For HF3035 samples, we observed an additional in frame fusion of exon 1 of CAPZA2 to exon 6 of MET, resulting in a truncated MET transcript with 13 codons from CAPZA2. Co-expression of the wild-type and fusion MET transcripts in the tumors and PDXs were validated using PCR. MET and p-MET levels were high thoughout the parental and PDX tumors. Capmatinib, which is a selective c-MET inhibitor was administered to the PDXs orally 5days/week. The treatment was effective in improving survival of HF3077 IC PDXs (p=0.028) and decreasing subcutaneous tumor size to 30% of the controls after 2 week treatment (t-test, p=0.017). However, treatment of HF3035 IC PDXs did not significantly improve survival (p=0.313). Kaplan-Meier survival curves were compared by log-rank (Mantel-Cox) test, sig. set at p<0.05. MET and p-MET detection by IHC of control and capmatinib treated xenografts show complete inhibition of p-MET, but did not affect MET overexpression in HF3035 PDX. Our results show that highly amplified regions are susceptible to genomic arrangements and the formation of fusion genes. Under investigation, is the basis for the strong selection for MET expressing cells in vivo and potential novel roles for MET in tumor progression.
Volume
78
Issue
13