Nonsense-mediated mRNA reduction and pre-mRNA processing in drosophila.

From bacteria to mammalian cells, the presence of a nonsense mutation causes a reduction in the level of the mRNA of the corresponding gene. The reduction is not, contrary to initial expectations, due to a passive mechanism by which non translated mRNAs are degraded; rather it is a active process in which active translation, cis-acting sequences and specific trans-acting factors are required. It is generally accepted that this phenomenon is the consequence of an evolutionary conserved mechanism that evolved to protect cells from the potentially deleterious effect of truncated proteins - this is often referred to as the mRNA surveillance system or nonsense mediated mRNA decay (NMD). This phenomenon has been extensively studied in budding yeast and in mammalian systems and to a lesser extent in C. elegans. In yeast the recognition of the nonsense codon appears to occur during cytoplasmic translation and premature translation termination is thought to activate a specific protein complex - called the surveillance complex - which in turn triggers an accelerated decay of the aberrant mRNA.
However, contrary to the expectation that the recognition of the nonsense codon should occur during cytoplasmic translation, several studies in mammalian cells indicate that NMD may take place in the nucleus by a mechanism that is independent of cytoplasmic translation. For example, several reports indicate that this reduction occurs while the mRNA is still associated with the nucleus, and that the stability of the cytoplasmic mRNA is unchanged relative to a wild-type allele.
The common view in the field is that these apparently discordant results between NMD in yeast and in mammalian cells will eventually be accommodated in a single model in which translation in the cytoplasm plays a prominent role. For example, a commonly given explanation is that the recognition of the nonsense codon takes place during nuclear export, and it has been implied that the apparent effects on nuclear RNA are in fact triggered by the premature abortion of translation at the cytoplasmic side of the nuclear envelope.
However not all the data from mammalian systems can be so easily explained by the above model. For example, several reports indicate that nonsense mutations affect the splicing of the corresponding pre-mRNA, which makes it difficult to imagine how premature translation in the cytoplasm could effect such an early event in mRNA biogenesis.
In summary, although it is a well established fact that premature termination codons can cause a drastic reduction of the mRNA level, it is still controversial whether the mechanism which causes this reduction is similar or not in all organisms. Furthermore it is unclear whether such mechanism(s) are exclusively cytoplasmic (and always linked to translation), or also nuclear (and independent of cytoplasmic translation).
This thesis represents the work that I have done to characterise the NMD phenomenon in Drosophila and to try to understand the underlining mechanism(s).
During this work, I found that nonsense mutations in the Alcohol dehydrogenase gene not only reduce the level of mRNA, but also lead to a longer poly(A) tail in both the mature mRNA and in the unspliced transcript. The effects of nonsense mutations on mRNA processing are happening in-cis, since in heterozygous flies only the transcript of the allele that carries the nonsense mutation is affected, while the mRNA from the other allele is not.
Since a longer poly(A) tail is usually associated with stable transcripts, I suggest that the reduction of the level of mRNA rather than reflecting a change in cytoplasmic stability may actually be the secondary consequence of an increase in abortive pre-mRNA processing events; this hypothesis is also supported by the observation that nonsense mutations in Adh are also associated with an increase of unspliced transcripts.
Collectively my data indicate that the reduction of mRNA level occurs in the nucleus and the lower level of mRNA is probably a consequence of an inhibitory effect of premature stop codons on pre-mRNA processing rather than being the consequence of a reduced stability of the mature transcript. These findings, therefore suggest that the integrity of the coding region may be in-vivo required for accurate pre-mRNA processing.
During this work I have also cloned and partially characterized the Drosophila homolog of Upf1 one of the three "master "genes thought to be involved in the reduction of mRNA harboring a premature stop codon in yeast and other organisms. The cloning of Upf1 in Drosophila opens the possibility of testing whether Upf1 is also involved in this process in Drosophila.