Investigation of premature termination codon recognition in nonsense-mediated mRNA decay

Nonsense-mediated mRNA decay (NMD) is best known for its role in quality control of mRNAs, where it recognizes premature translation termination codons (PTCs) and rapidly degrades the corresponding mRNA. The basic mechanism of NMD appears to be conserved among eukaryotes: aberrant translation termination triggers NMD. According to the current working model, correct termination requires the interaction of the ribosome with the poly(A)-binding protein (PABPC1) mediated through the eukaryotic release factors 1 (eRF1) and 3 (eRF3). The model predicts that in the absence of this interaction, the NMD core factor UPF1 binds to eRF3 instead and initiates the events ultimately leading to mRNA degradation. However, the exact mechanism of how the decision between proper and aberrant (i.e. NMD-inducing) translation termination occurs is not yet well understood. We address this question using a tethering approach in which proteins of interest are bound to a reporter transcript into the vicinity of a PTC. Subsequently, the ability of the tethered proteins to inhibit NMD and thus stabilize the reporter transcript is assessed. Our results revealed that the C-terminal domain interacting with eRF3 seems not to be necessary for tethered PABPC1 to suppress NMD. In contrast, the N-terminal part of PABPC1, consisting of 4 RNA recognition motifs (RRMs) and interacting with eukaryotic initiation factor 4G (eIF4G), retains the ability to inhibit NMD. We find that eIF4G is able to inhibit NMD in a similar manner as PABPC1 when tethered to the reporter mRNA. This stabilization by eIF4G depends on two key interactions. One of these interactions is to PABPC1, the other is to eukaryotic initiation factor 3 (eIF3). These results confirm the importance of PABPC1 in inhibiting NMD but additionally reveal a role of translation initiation factors in the distinction between bona fide termination codons and PTCs.

Probing interactions of NMD factors in a distance-dependent manner by BioID

The understanding of molecular mechanisms requires the elucidation of protein-protein interaction in vivo. For large multi-factor complexes like those assembling on mRNA, co-immunoprecipitation assays often identify many peripheral interactors that complicate the interpretation of such results and that might conceal other insightful mechanistic connections. Here we address the protein-protein interaction network for key factors in the nonsense-mediated mRNA decay (NMD) pathway in a distant-dependent manner using BioID1,2. In this novel approach, the mutant E. coli biotin-protein ligase BirAR118G is fused to the bait protein and biotinylates proximal proteins promiscuously. Hence, interactors positioned close to the bait in vivo are enriched by streptavidin purification and identified by mass spectrometry or western blotting. We present a validation of the BioID assay and preliminary results for close interactors of UPF1 and other key players in NMD.

SMG6 mediated degradation of nonsense mRNA requires phosphorylation-independent interaction with the helicase domain of UPF1

Eukaryotic mRNAs with premature translation-termination codons (PTCs) are recognized and eliminated by nonsense-mediated mRNA decay (NMD). NMD targeted mRNAs can be degraded by different routes that all involve phosphorylated UPF1 (P-UPF1) as a starting point. The endonuclease SMG6, which cleaves mRNA near the PTC, is one of three known NMD factors thought to be recruited to nonsense mRNAs by interaction with P-UPF1, leading to eventual mRNA degradation. By MS2-mediated tethering of SMG6 and mutants thereof to a reporter RNA combined with knockdowns of various NMD factors, we demonstrate that besides its endonucleolytic activity, SMG6 also requires UPF1 and SMG1 for inducing RNA decay. Our experiments revealed a phosphorylation-independent interaction between SMG6 and UPF1 that is important for SMG6-mediated mRNA decay and using yeast two hybrid assays, we mapped this interaction to the unique stalk region of the UPF1 helicase domain. This region of UPF1 is essential for SMG6-mediated reporter RNA decay and also for NMD. Our results postulate that besides recruiting SMG6 to its RNA substrates, UPF1 is also required to activate its endonuclease activity.

Trying to make sense in nonsense-mediated mRNA decay

Despite over 30 years of research, the molecular mechanisms of nonsense-mediated mRNA decay (NMD) are still not well understood. NMD appears to exist in most eukaryotes and is intensively studied in S. cerevisiae, C. elegans, D. melanogaster and in mammalian cells. Current evidence suggests that the core of NMD – involving UPF1, UPF2 and UPF3 – is evolutionarily conserved, but that different species may have evolved slightly different ways to identify target mRNAs for NMD and to degrade them. Our lab has shown that the exon junction complex (EJC) is not absolutely required for NMD in human cells (Bühler et al., NSMB 2006) and that it is neither restricted to CBP80-bound mRNAs as classical models claim (Rufener & Mühlemann, NSMB 2013). Together with the finding that long 3’ UTRs often are an NMD-inducing feature (Eberle et al, PLoS Biol 2008; Yepiskoposyan et al., RNA 2011), our data is consistent with much of the data from other species and hence has led to a “unified” working model for NMD (Stalder & Mühlemann, Trends Cell Biol 2008; Schweingruber et al., Biochim Biophys Acta 2013). Our recent iCLIP experiments with endogenous UPF1 indicate that UPF1 binds mRNAs indiscriminately with respect to being an NMD target or not before they engage with ribosomes (Zünd et al., NSMB 2013). After onset of translation, UPF1 is cleared from the coding region but remains bound to the 3’ UTR of mRNAs. Why this 3’ UTR-associated in some cases induces NMD and in others not is currently being investigated and not yet understood. Following assembly of a phospho-UPF1-containing NMD complex, decay adaptors (SMG5, SMG7, PNRC2) and/or the endonuclease SMG6 are recruited. While the latter cleaves the mRNA in the vicinity of the termination codon, the former proteins induce deadenylation, decapping and exonucleolytic degradation of the mRNA. In my talk, I will give an overview about the latest developments in NMD – with a focus on our own work – and try to integrate the bits and pieces into a somewhat coherent working model.

An mRNA-derived ncRNA targets and regulates the ribosome

Small non-protein-coding RNA (ncRNA) molecules are key players in controlling gene expression at multiple steps in all domains of life. While the list of validated ncRNAs that regulate crucial cellular processes grows steadily (such as micro RNAs and small-interfering RNAs), not a single ncRNA has been identified that directly interacts and regulates the ribosome during protein biosynthesis (with the notable exceptions of 7SL RNA and tmRNA). This is unexpected, given the central position the ribosome plays during gene expression. To investigate whether such a class of regulatory ncRNAs does exist we performed genomic screens for small ribosome-associated RNAs in various model organisms of all three domains [1,2]. Here we show that an mRNA-derived 18 nucleotide long ncRNA is capable of down-regulating translation in Saccharomyces cerevisiae by directly targeting the ribosome [3]. This 18-mer ncRNA binds to polysomes upon salt stress and is crucial for efficient growth under hyperosmotic conditions. Although the 18-mer RNA originates from the TRM10 locus, which encodes a tRNA methyltransferase, genetic analyses revealed the 18-mer RNA nucleotide sequence, rather than the mRNA-encoded enzyme, as the translation regulator under these stress conditions. Our data reveal the ribosome as a target for small regulatory ncRNAs and unveil the existence of a novel mechanism of translation regulation. Analogous genomic screens in organisms spanning all three domains of life demonstrate the existence of thousands of ncRNA candidates putatively regulating the ribosome. We therefore anticipate that ribosome-bound ncRNAs are capable of fine tuning translation and might represent a so far largely unexplored class of regulatory ncRNAs.

An mRNA-derived ncRNA targets and regulates the ribosome

Small non-protein-coding RNAs (ncRNAs) are key players in controlling gene expression. The advantage of ncRNA regulators is their almost immediate availability since they act on the RNA level. The list of validated ncRNAs regulating translation, such as micro RNAs, is growing steadily, however, they almost exclusively target the mRNA rather than the ribosome. This is unexpected given the central position the ribosome plays. Here we show that an mRNA-derived 18 nucleotide long ncRNA is capable of down-regulating translation in Saccharomyces cerevisiae by targeting the ribosome. This 18-mer ncRNA binds to polysomes upon salt stress and is crucial for efficient growth. Although the 18-mer RNA originates from the TRM10 locus, which encodes a tRNA methyltransferase, genetic analyses revealed the 18-mer RNA nucleotide sequence as the translation regulator. Our data reveal the ribosome as a target for a small regulatory ncRNA and demonstrate the existence of a yet unknown mechanism of translation regulation.

Atomic mutagenesis of the ribosomal Sarcin-Ricin-Loop to study EF-G GTPase activation

Translocation factor EF-G, possesses a low basal GTPase activity, which is stimulated by the ribosome. One potential region of the ribosome that triggers GTPase activity of EF-G is the Sarcin-Ricin-Loop (SRL) (helix 95) in domain VI of the 23S rRNA. Structural data showed that the tip of the SRL closely approaches GTP in the active center of EF-G, structural probing data confirmed that EF-G interacts with nucleotides G2655, A2660, G2661 and A2662.1-3 The exocyclic group of adenine at A2660 is required for stimulation of EF-G GTPase activity by the ribosome as demonstrated using atomic mutagenesis.4 Recent crystal structures of EF-G on the ribosome, gave more insights into the molecular mechanism of EF-G GTPase activity.5 Based on the structure of EF-Tu on the ribosome1, the following mechanism of GTPase activation was proposed: upon binding of EF-G to the ribosome, the conserved His92 (E.coli) changes its position, pointing to the γ-phosphate of GTP. In this activated state, the phosphate of residue A2662 of the SRL positions the catalytic His in its active conformation. It was further proposed that the phosphate oxygen of A2662 is involved in a charge-relay system, enabling GTP hydrolysis. In order to test this mechanism, we use the atomic mutagenesis approach, which allows introducing non-natural modifications in the SRL, in the context of the complete 70S ribosome. Therefore, we replaced one of the non-bridging oxygens of A2662 by a methyl group. A methylphosphonat is not able to position or activate a histidine, as it has no free electrons and therefore no proton acceptor function. These modified ribosomes were then tested for stimulation of EF-G GTPase activity. First experiments show that one of the two stereoisomers incorporated into ribosomes does not stimulate GTPase activity of EF-G, whereas the other is active. From this we conclude that indeed the non-bridging phosphate oxygen of A2662 is involved in EF-G GTPase activation by the ribosome. Ongoing experiments aim at revealing the contribution of this non-bridging oxygen at A2662 to the mechanism of EF-G GTPase activation at the atomic level.

Atomic mutagenesis of the ribosomal Sarcin-Ricin-Loop to study EF-G GTPase activation

Translocation factor EF-G, possesses a low basal GTPase activity, which is stimulated by the ribosome. One potential region of the ribosome that triggers GTPase activity of EF-G is the Sarcin-Ricin-Loop (SRL) (helix 95) in domain VI of the 23S rRNA. Structural data showed that the tip of the SRL closely approaches GTP in the active center of EF-G, structural probing data confirmed that EF-G interacts with nucleotides G2655, A2660, G2661 and A2662.1-3 The exocyclic group of adenine at A2660 is required for stimulation of EF-G GTPase activity by the ribosome as demonstrated using atomic mutagenesis.4 Recent crystal structures of EF-G on the ribosome, gave more insights into the molecular mechanism of EF-G GTPase activity.5 Based on the structure of EF-Tu on the ribosome1, the following mechanism of GTPase activation was proposed: upon binding of EF-G to the ribosome, the conserved His92 (E.coli) changes its position, pointing to the γ-phosphate of GTP. In this activated state, the phosphate of residue A2662 of the SRL positions the catalytic His in its active conformation. It was further proposed that the phosphate oxygen of A2662 is involved in a charge-relay system, enabling GTP hydrolysis. In order to test this mechanism, we use the atomic mutagenesis approach, which allows introducing non-natural modifications in the SRL, in the context of the complete 70S ribosome. Therefore, we replaced one of the non-bridging oxygens of A2662 by a methyl group. A methylphosphonat is not able to position or activate a histidine, as it has no free electrons and therefore no proton acceptor function. These modified ribosomes were then tested for stimulation of EF-G GTPase activity. First experiments show that one of the two stereoisomers incorporated into ribosomes does not stimulate GTPase activity of EF-G, whereas the other is active. From this we conclude that indeed the non-bridging phosphate oxygen of A2662 is involved in EF-G GTPase activation by the ribosome. Ongoing experiments aim at revealing the contribution of this non-bridging oxygen at A2662 to the mechanism of EF-G GTPase activation at the atomic level.

Expression of the vault RNA protects cells from undergoing apoptosis

Non-protein-coding RNAs are a functionally versatile class of transcripts found in all domains of life exerting their biological role at the RNA level. Recently, we demonstrated that the vault-associated RNAs (vtRNAs) were significantly up-regulated in human B cells upon Epstein-Barr virus (EBV) infection [1,2]. vtRNAs are an integral part of the vault complex, a huge and evolutionarily conserved cytoplasmic ribonucleoprotein complex. The major vault protein (MVP) is the main structural component of the complex while vtRNA accounts for only 5% of its mass. Very little is known about the function(s) of the vtRNAs or the vault complex. In particular the role and significance of the previously observed vtRNA up-regulation upon EBV infection remained unclear. We individually expressed EBV-encoded genes in B cells and found the latent membrane protein 1 (LMP1) as trigger for vtRNA up-regulation. To unravel a putative functional interconnection between vtRNA expression and EBV infection, we ectopically expressed vtRNA1-1 in human B cells and observed an improved viral establishment. Furthermore, expression of vtRNA1-1 but not of the other vtRNA paralogs protected cells from undergoing apoptosis. Knock-down of MVP had no effect on these phenotypes thus revealing the vtRNA and not the vault complex to contribute to the enhanced EBV establishment and apoptosis resistance. Mutational analysis highlighted the central domain of the vtRNA to be involved in the anti-apoptotic effect. Ongoing research aims at characterizing the target of vtRNA1-1 in the apoptotic pathway. In summary, our data reveal a crucial cellular function for the so far elusive RNA biology of the vtRNAs.

Functional proteomics and biochemsitry: working together to unravel mitochondrial phenomena of African trypansomes

Eukaryotic cells are compartmentalized into membrane-bound organelles in order to provide sheltered reaction rooms for various specific processes. Organelles are not randomly distributed in a cell or operate isolated from each other. At the contrary — some organelles are closely linked and their functions are tightly orchestrated. The most well-known example of two such organelles acting in concert are the ER and the mitochondrion that work together in order to coordinate cellular lipid biosynthesis, maintain Ca2+-homeostasis, regulate mitochondrial division and control mitochondrial/ER shape as well as to synchronize the movement of these organelles within a cell. To study the mitochondrion and its interface to the ER requires a simplified mitochondrial system. African trypanosomes represent such a system. The unicellular parasite that causes devastating diseases in humans and animals has only one large mitochondrion that does not undergo fission/fusion events except for the context of cell division. Moreover, mitochondrial functions and morphology are highly regulated throughout the life cycle of the protozoan. Central to the understanding of how mitochondria control their morphology, communicate with their surroundings and manage exchange of metabolites and transport of biopolymers (proteins, RNAs) is the mitochondrial outer membrane (MOM), as the MOM defines the boundary of the organelle. Recently, we have purified the MOM of T. brucei and characterized its proteome using label-free quantitative mass spectrometry for protein abundance profiling in combination with statistical analysis. Our results show that the trypanosomal MOM proteome consists of 82 proteins, two thirds of which have never been associated with mitochondria before. Among these, we identified novel factors required to regulate mitochondrial morphology and the long-elusive protein import machinery of T. brucei. A comparison with the MOM proteome of yeast defines a set of 17 common proteins that are likely present in the mitochondrial outer membrane of all eukaryotes. One of these is the Miro-GTPase Gem1. In yeast, this Ca2+-EF-Hand containing polypeptide is thought to be involved in a protein complex that physically tethers the mitochondrion to the ER. Interestingly, a putative tethering complex in mammalian cells was linked to the mitochondrial fusion/fission machinery. Thus, the concept of a protein complex-mediated connection seems to be a general and conserved feature. We are currently investigating, if such a protein complex exists in T. brucei and if the trypanosomal Gem1 protein is involved. This ER-subdomain associated with mitochondria has been termed mitochondria-associated ER-membranes or MAM. The MAM has recently been implicated to play a key role in Alzheimer’s disease. It is therefore of broad and general interest to establish other eukaryotic model systems in order to investigate the MAM-MOM connection in more detail.

Functional proteomics and biochemsitry: working together to unravel mitochondrial phenomena of African trypansomes

Central to the understanding of how mitochondria control their morphology, communicate with their surroundings and manage exchange of metabolites and transport of biopolymers (proteins, RNAs) is the mitochondrial outer membrane (MOM), as the MOM defines the boundary of the organelle. Recently, we have purified the MOM of the mitochondrial model organism T. brucei and characterized its proteome. Our results show that the trypanosomal MOM proteome consists of 82 proteins. Among these, we identified novel factors required to regulate mitochondrial morphology and the long-elusive protein import machinery of T. brucei. A comparison with the MOM proteome of yeast defines a set of 17 common proteins that are likely present in the mitochondrial outer membrane of all eukaryotes. One of these is the Miro-GTPase Gem1. In yeast, this Ca2+-EF-Hand containing polypeptide is thought to be involved in a protein complex that physically tethers the mitochondrion to the ER. In mammalian cells, a putative tethering complex was linked to the mitochondrial fusion/fission machinery. Thus, the concept of a protein complex-mediated connection seems to be a general and conserved feature. We are currently investigating if such a protein complex exists in T. brucei and if the trypanosomal Gem1 protein is involved.

tRNA-derived fragments target the small ribosomal subunit to fine-tune translation

Post-transcriptional cleavage of RNA molecules to generate smaller fragments is a widespread mechanism that enlarges the structural and functional complexity of cellular RNomes. In particular, fragments deriving from both precursor and mature tRNAs represent one of the rapidly growing classes of post-transcriptional RNA pieces. Importantly, these tRNA-derived fragments (tRFs) possess distinct expression patterns, abundance, cellular localizations, or biological roles compared with their parental tRNA molecules (1). Here we present evidence that tRFs from the archaeon Haloferax volcanii directly bind to ribosomes. In a previous genomic screen for ribosome-associated small RNAs we have identified a 26 residue long fragment originating from the 5’ part of valine tRNA (Val-tRF) to be by far the most abundant tRF in H. volcanii (2). The Val-tRF is processed in a stress- dependent manner and was found to primarily target the small ribosomal subunit in vitro and in vivo. Translational activity was markedly reduced in the presence of Val-tRF, while control RNA fragments of similar length did not show inhibition of protein biosynthesis. Crosslinking experiments and subsequent primer extension analyses revealed the Val-tRF interaction site to surround the mRNA path in the 30S subunit. In support of this, binding experiments demonstrated that Val-tRF does compete with mRNAs for ribosome binding. Therefore this tRF represents a ribosome-bound non-protein-coding RNA (ncRNA) capable of regulating gene expression in H. volcanii under environmental stress conditions probably by fine-tuning the rate of protein production (1). (1) Gebetsberger J. and Polacek N. (2013), RNA Biol. 10:1798-1808 (2) Gebetsberger J. et. al. (2012), Archaea, Article ID 260909