The characterization of peptides expressed from short open reading frames (sORFs)

Abstract

Potentially translatable short open reading frames (sORFs) of less than 100 codons are present on both mRNAs and non-coding RNAs (ncRNAs) and 50% of mammalian mRNAs contain at least one sORF. We hypothesize that a subset of sORFs encode for functional short peptides (sPEPs) that are expressed and contribute to proteome complexity. Our recent bioinformatic studies showed that nearly 2% of sORFs are conserved between several species indicating that these sORFs may have critical functions. Recent proteomic studies have identified over 1,000 sPEPs in human cell lines showing that some sORFs are indeed translated. Surprisingly, a number of peptides have been identified that are encoded by sORFs present in ncRNAs. In order to extend and validate these studies, I extracted low molecular weight proteins from HeLa and HEK293 cell lysates by either SDS-PAGE or ERLIC fractionation. These extracts were digested with trypsin or LysC and analysed by nano LC-MS/MS. The resulting MS/MS data was searched against the UniProKB/Swiss-Prot using MASCOT version 2.4 to filter out known proteins, and all unmatched spectra were searched against the human RefSeq database. ProteinPilot TM was also used to identify sORF-encoded peptides by searching against an in-house sORF and the Human AlternativeOpen Reading Frame (HaltORF) databases. To date, I have identified several sPEPs including three that are novel. These sPEPs have a mass of less than 20 kDa. One of those is expressed from an ncRNA transcript and is expected to be secreted. The other two sPEPs are encoded by upstream open reading frames (uORFs) on mRNA transcripts; one of these is predicted to localise to the cytoplasm while the other is expected to be secreted. The role, if any, of these peptides has yet to be determined but their identification has provided a pool of candidates for further molecular characterization in order to determine their function.3