The phenomenon of NMD was first described in human cells and in yeast almost simultaneously in 1979, which already suggested broad phylogenetic conservation and an important biological role of this intriguing mechanism. Significantly, the recognition of translational sense occurs before the mRNA enters its repeated rounds of translation. The basic signals that are required for mammalian NMD to work are (1) the splicing-dependent deposition of a multi-protein complex, referred to as exon junction complex (EJC), close to the exon-exon boundaries of the mRNA and (2) the interaction of the EJC with components of the terminating ribosome. The translation apparatus can thus distinguish a proper translation termination event in the 3’ terminal exon, which is not followed by a downstream EJC from an improper termination event, which is followed by a downstream EJC.
For our analysis of the NMD mechanism, we used mutation analyses of EJC components and RNAi-mediated depletion of NMD- and EJC-proteins in the context of a functional tethering assay and showed that different EJC components specify distinct routes of NMD with differential cofactor requirements (Gehring et al. 2005). This work challenged the then predominant paradigm and proposed a concept that has since gained much additional support through work by us and others. Subsequent work functionally analyzed the isoforms of the key NMD factor UPF3 (UPF3a and UPF3b) and found that these isoforms serve clearly separable functions in NMD and in translation, although they are found in EJCs with an otherwise similar composition (Kunz et al. 2006). In addressing the important question of whether different modes of translation may serve distinguishable functions in NMD, we found that NMD can be activated by both, cap-dependent and IRES-mediated translation (Holbrook et al. 2006).
Furthermore, we provided direct evidence that normal endogenous cellular mRNAs can be regulated by alternative NMD pathways (Gehring et al. 2005). The existence of such alternative NMD pathways conceptually suggested NMD to represent a variable process affecting RNA stability and potentially regulating different classes of mRNAs in different tissues or at different times of development. This new concept significantly widened the previously assumed function of NMD as a constitutive and mechanistically linear “vacuum cleaner” for misprocessed and mutated mRNAs (Neu-Yilik et al. 2004 & 2008).
As a logical next step in further characterizing the mechanism of NMD pathways, we studied the interaction of components of the EJC and the translation termination complex. It had previously been shown by others that the post-termination complex consists of at least 4 specific proteins, the release factors eRF1 and eRF3, the helicase UPF1 and the UPF1-specific kinase SMG1. The phosphorylation of UPF1 by SMG1 is an essential step of triggering mRNA degradation and SMG1 has been shown to be activated by an interaction with the EJC in general and its component UPF2 in particular.
We have recently found evidence that UPF1 delays termination thus enabling the post-termination complex to either interact with the poly(A) binding protein (PABPC1), likely inducing cycling of the mRNA into the next round of translation, or to interact with the EJC to trigger UPF1 phosphorylation and mRNA decay. We also demonstrated that UPF1 phosphorylation can either be supported by UPF2 or by UPF3b as a biochemical correlate for the previously identified alternative NMD pathways. On the basis of these observations, we propose an integrated model of NMD, where UPF1 delays translation termination and is phosphorylated by SMG1, if the termination-promoting interaction of PABPC1 with eRF3 cannot readily occur. The EJC, with UPF2 or UPF3b as a cofactor, interferes with proper termination and leads to the phosphorylation of UPF1 and subsequent decay. This model integrates previously competing models of NMD and suggests a mechanistic basis for alternative NMD pathways (Ivanov et al. 2008). In our most recent work we have shown that the exon junction complex is assembled in a strict stepwise hierarchical fashion that enables the EJC to functionally connect to different downstream pathways (Gehring et al. PLOS Biology 2009). Further, we have identified the protein PYM as a specific cytoplasmic factor that serves to disassemble the EJC and to recycle its components (Gehring et a. Cell, 2009).
The medical perspective of NMD as a key post-transcriptional mechanism limiting abnormal and regulating normal gene expression is a particular focus of our team (Holbrook et al. 2004). Recently, other groups have shown that NMD can vary between cell types and between individuals. Furthermore, the expression of disease-related genes that carry identical nonsense mutations has been reported to differ and to modulate disease severity. These observations led us to the hypothesis that variations of NMD efficiency may contribute to the phenotypic variability of hereditary and possibly of acquired genetic disorders. It has so far been impossible to test this hypothesis because of the lack of an assay system to reliably quantify NMD efficiency in readily accessible clinical material. We have now established an assay system for the quantification of NMD efficiency, which is based on carefully validated cellular NMD target transcripts.
We demonstrated in a HeLa cell model system that NMD efficiency can be remarkably variable and represents a stable characteristic of different subclones of HeLa cells. In one of these lines, low NMD efficiency was shown to be functionally related to the reduced abundance of the exon junction component RNPS1. Cellular concentrations of RNPS1 can thus modify NMD efficiency and cell type specific co-factor availability emerged as a novel principle that controls NMD (Viegas et al. 2007). In a complementary approach, we designed a cell-based chemiluminescence reporter system that enables us to robustly measure the effect of modifications of the NMD pathway with a high degree of sensitivity and in a format that can be used for high throughput small molecule and/or RNAi screening (Boelz et al., 2007). The medical dimension of the variability of NMD has been directly addressed in cystic fibrosis (CF) by a collaborative project with the Kerem group at Hebrew University in Jerusalem. Previous work of the Kerem group had shown that nonsense suppression in CF-patients with nonsense mutations can reconstitute CFTR function.
However, the readthrough efficiency and the subsequent correction of chloride transport in response to the aminoglycoside gentamicin have been highly variable among a group of CF-patients carrying the same nonsense mutation. We have now demonstrated in nasal epithelial cell lines of such patients that the abundance of nonsense mutated CFTR transcripts differs. A functional role of variable NMD efficiency was implicated by the finding that the differences in CFTR transcript levels correlated with differences in the abundance of normal cellular NMD targets and by experimentally depleting and repleting important NMD factors. Interestingly, chloride transport could be elevated more effectively by nonsense suppression in patients with less efficient NMD than in those with highly efficient NMD (Linde et al, 2007a & 2007b). More generally, this study further supported the concept that variable NMD efficiency may play a role as a genetic modifier of different cellular functions and may govern the response to treatments that aim to promote readthrough of nonsense mutated transcripts.