luti-mRNA mediated gene repression
The time and location of gene expression affects how organisms differentiate their cells into distinct lineages. While protein degradation and translational regulation affect the final level and localization of protein output, transcription factors are considered to be the dominant drivers of gene regulation throughout development. It is well established that transcription factor cascades coordinates the timely activation of gene expression clusters during development. In comparison, much less is understood about how gene repression events are coordinated with the transcription factor-driven waves of gene activation and how this repression is achieved at a mechanistic level.
Recently, we uncovered a novel mechanism, where a central meiotic transcription factor induces the expression of an mRNA that serves a purely regulatory function. This mRNA is never translated into a functional protein due to the upstream open reading frames (uORFs) in its 5’ leader region (Chen et al., 2017). Instead, it serves to inactivate a gene through an integrated transcriptional and translational mechanism (Chen et al., 2017; Chia et al., 2017). This new insight challenges the assumption that mRNA molecules must produce the gene product encoded in their open reading frames and provides a fresh perspective on how gene activation and inactivation waves can be temporally coordinated during differentiation by utilizing the same set of transcription factors.
Our previous work on meiotic kinetochore regulation in budding yeast demonstrated that the essential kinetochore protein Ndc80 is downregulated during S-phase and prophase (Miller et al., 2012). This assists in kinetochore remodeling which allows homologous chromosomes to be segregated in meiosis I. A deeper investigation into the mechanism by which the Ndc80 protein level decreases led us to discover an initially counterintuitive mechanism by which cells can downregulate gene expression in a cell.
At the NDC80 locus, a 5’-extended transcript isoform is expressed exclusively during meiosis. It is developmentally regulated by the master meiotic transcription factor Ime1 and its binding partner Ume6. The extended transcript contains 9 uORFs in addition to the entire NDC80 coding sequence. Using single molecule fluorescence in situ hybridization, we have shown that this transcript is exported from the nucleus. Its translation status, determined by ribosome profiling, confirms that the extended transcript is engaged with the ribosome (Brar et al., 2012; Miller et al., 2012). By all accounts, this RNA molecule would be considered an mRNA. Intriguingly, due to the uORFs, it cannot be translated. For this reason, we have termed this class of RNAs, luti-mRNAs, for long undecoded transcript isoform mRNAs and for which NDC80 is the founding member.
In budding yeast, hundreds of transcripts with 5’-extended leaders are expressed in meiosis; the majority of which also contain translated AUG initiated uORFs, as observed by ribosome profiling (Brar et al., 2012). On an individual gene basis, BOI1, which is involved in polarized growth, has an extended meiosis-specific transcript that, similar to NDC80luti, is regulated by Ume6/Ime1 (Liu et al., 2015). In another case, it was demonstrated that the origin of recognition gene ORC1 has an extended transcript that is regulated by the mid-meiotic transcription factor Ndt80 (Xie et al., 2016). We predict that other extended meiotic transcript isoforms are regulated by these two master transcription factors and that upon further investigation many will prove to be translationally repressed luti-mRNAs.
Approximately 50% of mouse, >40% of drosophila, and 30-50% of human genes have alternative start site usage during development (Batut et al., 2013; Kimura et al., 2006). In those same organisms, uORF translation is prevalent with an estimated 50% of human genes harboring translated uORFs (Calvo et al., 2009; Chew et al., 2016; Dunn et al., 2013; Johnstone et al., 2016). Transcription-coupled chromatin modifications are also highly conserved across evolution (Eissenberg and Shilatifard, 2010; Wagner and Carpenter, 2012). Therefore, it does not seem far-fetched to propose that the form of regulation controlling the NDC80 locus in budding yeast meiosis could be responsible for fine-tuning gene expression in other organisms.