Abstract
The recent advent of whole exon (exome)-capture technology, coupled with second-generation sequencers, has made it possible to readily detect genomic alterations that affect encoded proteins in cancer cells. Such target resequencing of the cancer genome, however, fails to detect most clinically-relevant gene fusions, given that such oncogenic fusion genes are often generated through intron-to-intron ligation. To develop a resequencing platform that simultaneously captures point mutations, insertions-deletions (indels), and gene fusions in the cancer genome, we chose cDNA as the input for target capture and extensive resequencing, and we describe the versatility of such a cDNA-capture system. As a test case, we constructed a custom target-capture system for 913 cancer-related genes, and we purified cDNA fragments for the target gene set from five cell lines of CML. Our target gene set included Abelson murine leukemia viral oncogene homolog 1 (ABL1), but it did not include breakpoint cluster region (BCR); however, the sequence output faithfully detected reads spanning the fusion points of these two genes in all cell lines, confirming the ability of cDNA capture to detect gene fusions. Furthermore, computational analysis of the sequence dataset successfully identified non-synonymous mutations and indels, including those of tumor protein p53 (TP53). Our data might thus support the feasibility of a cDNA-capture system coupled with massively parallel sequencing as a simple platform for the detection of a variety of anomalies in protein-coding genes among hundreds of cancer specimens.