NGS-based targeted RNA sequencing for expression profiling and relative quantitation of specific gene isoforms and fusions in tumor-specific panels.

Abstract

11108 Background: Gene expression signatures have become a useful tool for the identification of tumor subtypes and response to specific therapies. Expression of tumor, metastatic and macrophage specific transcripts utilizing alternative promoters and transcriptional start sites can further characterize these tumors . NGS is a powerful tool for gene expression analysis, however larger sample input requirements (>100ng) and excessive sequencing depth requirements (30-40M tags/sample) to detect the expression of rare isoforms or fusions in tumor samples are prohibitive for clinical assay development. We describe the development of a targeted RNA sequencing assay for the relative quantitation of specific gene expression signatures, known splice variants and gene fusions from less than 100 ng of starting material in a single tube universal amplification format. Methods: Primers for 52 genes, isoforms and gene fusion products were designed using the universal amplification strategy. 10 ng of RNA from 5 matched tumor/adjacent normal breast cancer tumor pairs were assayed. Libraries were prepared for sequencing by emPCR and sequenced on Ion Torrent PGM. Data were aligned via TMAP. Relative expression was determined vs. housekeeping genes or wild type transcripts. Results: All gene targets were detected at significant levels in at least one tumor sample. Robust expression profiling (5 log dynamic range) was obtained from FFPE macrodissected tumor and normal samples with as little as 200K reads/sample. Immune specific transcripts demonstrated differential expression (CCL3, AIF, FCGR3A and CSF1) across patients and matched pairs as well as an upregulation of CXCL12, indicative of tumor associated macrophages. Conclusions: Targeted RNAseq demonstrates detection and quantitation of relative expression levels of not only tumor subclass specific gene expression signatures, but immune cell specific transcripts from 10ng of FFPE derived total RNA derived from macrodissected tumor samples. The lower input requirements, quicker turnaround time and incredible sensitivity of targeted RNAseq make this assay a useful tool for clinical assay development.