Systematic Detection of Alternative Open Reading Frames (altORFs) in Cancer Driver Genes.

The discovery of translated alternative open reading frames (altORFs) in protein-coding regions has expanded the coding potential of viral, prokaryotic and eukaryotic genes. Experimental and computational approaches indicate that overlapping coding regions occur in mammals. In this study, I used a prediction method based on five criteria to detect novel altORFs in the human genes taken from the COSMIC Cancer Gene Census Database. Apart from the well characterized examples of human cancer-specific antigens expressed from altORF, the vast catalogue of nucleotide substitutions across cancer genes (the COSMIC database) is also likely to harbor previously uncharacterized altORFs. Under the five prediction criteria, I found 251 novel altORFs, 41 of which highly conserved in mammals and 60 uniquely resulting from nucleotide substitutions in the primary ORF of cancer genes. I found experimental evidence for 38% of the 251 novel altORFs from mass spectrometry and ribosome profiling databases. In particular, I found three altORFs in the proto-oncogene RET, three expressed altORfs in the isocitrate dehydrogenase-2 gene, and one expressed large altORF (498 nt) in the mutated TP53 gene. This study may offer clinical perspectives, because a potential source of cancer antigens may include antigens derived from translation of currently unannotated open reading frames. The altORFs detected in this study could be candidates for future experimental validation.