This is the m ost common outcome of a PTC m utation. The transcript carrying the PTC is rapidly degraded in vivo by a RNA surveillance mechanism, know n as NMD (M aquat et al., 1995). NMD has captured m uch attention as one of the m ost enigmatic processes in RNA metabolism (Hentze and Kulozik., 1999).
The process can be divided into three phases. Firstly, a transcript is proof-read to identify faulty open reading frames. This is know n as the "discrim ination phase", where the norm al stop codon is discrim inated from the prem ature one. Subsequently, the m utant transcript is comm itted to the NMD pathw ay "com m itm ent phase", that eventually results in its degradation "degradation phase" (Culbertson., 1999).
The ability of NMD to limit the effects of truncated dom inant negative proteins has been appreciated in three species: nem atode (C. elegans)', yeast (S. cerevisiae) and hum an (H. sapiens).
1.15.3.1. C. elegans
Seven genes term ed smgl-smg? are responsible for the efficiency of NMD in C. elegans (Pulak and Anderson 1993; H odgkin et al, 1989). Nonsense m utations that are norm ally recessive often confer a dom inant negative phenotype in worm s carrying a smg- m utation (Cali et al., 1993). The best example of this is in the gene UNC-54, which encodes one of two myosin heavy chains expressed in w orm body wall muscles. Heterozygous worm s are phenotypically normal, due
to NMD of the UNC-54 allele. Homozygotes show severe abnormalities of m uscle architecture and function. M utations in the smg genes (inactivating NMD) cause heterozygous UNC-54 worm s to become immobile. Hence, this is a clear example where mRNA surveillance avoids the production of dom inant - negative polypeptides (Pulak and Anderson 1993).
1.15.3.2. S. cerevisiae
In 1979, it w as observed that PTC m utations of yeast Ura3 gene destabilised the mRNA (Losson and Lacroute 1979). It is established that NMD in yeast requires at least three frans-acting factors, term ed UPFlp, UPF2p and UPF3p (He et al, 1995; Cui et al., 1995; Lee et al 1995). Here sm gl in c.elegans and U PFlp in yeast are counterparts. However, the other smg counterparts have not been identified in yeast, suggesting a more complex NMD process in higher eukaryotes.
Analysis of the CYH2 gene in yeast provides a good example of inefficient splicing leading to NMD. The CYH2 gene codes for a ribosomal protein in S.cerevisiae. It contains an intron with a stop codon that is spliced incorrectly. Functional NMD ensures that the unspliced prim ary transcript is diminished. However, m utations in any of the three UPF genes inactivates NMD and causes an accumulation of unspliced CYH2 mRNA (He et al, 1993). Loss of NMD function in S. cerevisiae is not drastic, as near norm al grow th occurs, except in the mitochondria. It has been show n that UPF- m utations affect telomere length (Lew et al, 1998), gene silencing and kinetochore assembly (Dahlseid et al, 1998).
The significance of NMD on gene expression has been investigated comprehensively in S. cerevisiae. Here, deletions of any or all UPF genes results in alteration of 8% of all mRNAs, where 90% of these show an increase in steady state mRNA levels. Preliminary evidence suggests that mRNAs coding for a small subset of transcription factors act as direct targets for the UPF proteins. Therefore it is possible that gene expression occurs via NMD sensitive transcriptional regulation (Culbertson 1999).
The NMD process has been well characterised in S. cerevisiae (see Figure 1.6 for the sequential steps of the NMD pathw ay in S.cerevisiae). There is m uch controversy over the sublocalisation of NMD, but the favoured m odel in yeast proposes that the decay occurs in the cytoplasm and is dependent on the translation process. There are two m ain lines of evidence that suggest that NMD occurs in the cytoplasm:
1. Most UPF proteins are localised in the cytoplasm and associate w ith polysomes (Atkin 1995; 1997).
2. Pharmacological translational inhibitors and expression of suppressor tRNAs allow the "read-through'" of nonsense codons and result in elevated levels of PTC transcripts (Gozalbo et al., 1990).
At translation, a ribosome pauses upon encountering a PTC, then signals the recruitm ent of a "surveillance complex". This is know n as the recruitm ent/com m itm ent phase. This surveillance complex consists of eukaryotic release factors eRFl, eRF3 and UPFlp. This complex consisting of the release factors, binds to the transcript to stim ulate hydrolysis of the
unfinished polypeptide. The eRF3, hydrolyses the GTP to signal translation termination. The UPFlp scans downstream of the PTC for a sequence know n as dow nstream destabilising element (DSE). Here, the position of the DSE determ ines the fate of the PTC transcript. If the DSE is encountered within about 200bp dow nstream of the PTC, then the stop codon is considered ''prem ature" (discrimination phase) and is prone for degradation (degradation phase) (Ruiz-Echevarria et al, 1998). There are several unansw ered questions concerning DSEs as discussed by Frishmeyer and Dietz., (1999):
1. 25% or so genes do not contain this putative czs-element (DSE), does this make them imm une to NMD?
2. W hat percentage of DSEs are functional?
3. Are there other unrecognised sequences that function as DSEs?
4. Are there elements that stabilise PTC transcripts in the vicinity of a PTC instead of destabilising it? Therefore, it could be a lack of these elements that lead to increased decay.
Norm al mRNA turnover requires deadenylation dependent 5' decapping, which is followed by 5 '-^ 3' decay by X rnlp exonuclease (Decker et al, 1994). How ever NMD has the ability to bypass the deadenylation step prior to the decapping and decay (Caponigro et al, 1996).
The overall interaction of the UPF proteins to each other may be described as follows: UPF3p and UPFlp may function in the recruitm ent of transcripts in the decay pathw ay; UPF3p may transport the nonsense transcript to the
surveillance complex (which contains UPFlip) and as UPFlp interacts with both U PFlp and UPF3p, this may bridge the gap betw een both factors (Culbertson 1999; Frischmeyer and Dietz., 1999) (Figure 1.6).
See discussion for evidence generated in this thesis and other reports of NMD in hum an disease genes in proposing a m odel for m am m alian NMD.
UPF2