High resolution ribosome profiling reveals gene‐specific details of UGA re‐coding in selenoproteins

Abstract

SECIS binding protein 2 (SECISBP2) increases the efficiency of re-coding of UGA codons as selenocysteine (Sec) during translation of mRNAs containing a Sec insertion sequence (SECIS) in the 3’ untranslated region. Our previous ribosomal profiling (RiboSeq) studies in mice showed that re-coding of UGA occurred in the absence of Secisbp2 on a higher level than if tRNASec was inactivated. We concluded that UGA recoding was an intrinsic property of UGA codons in selenoproteins which can be massively increased by the SECIS: Secisbp2 interaction. Since we worked on cell-type specific liver and brain tissues, heterogeneity of cells expressing or not wildtype Secisbp2 might have confounded our analyses and conclusions. In order to study a homogenous sample and thus gain a more detailed in-depth view on UGA recoding, we decided to subject the haploid human HAP1 cell line deficient in SECISBP2 by RiboSeq. Expression of selenoproteins was found at a reduced, but still detectable, level in the mutated cells. Metabolic labeling demonstrated incorporation of 75Se into selenoproteins. UGA recoding was assessed individually for all expressed selenoproteins by RiboSeq. Analysis of ribosomes residing on UGA either at the A-site or the P-site revealed transcript-specific responses to SECISBP2 inactivation. We propose a new measure to assess UGA/Sec read-through in all selenoproteins, i.e. the proportion of ribosome protected fragments (RPF) carrying UGA adjacent to the P-site, pUGA. Omission of translation inhibitors like cycloheximide during preparation allowed to asses not only ribosomes with occupied A-sites, but also ribosomes with unoccupied A-sites, which are differentiated by footprint size. Considering UGA positions in the RPF and footprint size, new details of UGA/Sec re-definition in individual selenoproteins can be uncovered.