Aberrant RNAs are intercepted in higher organisms and degraded in the nucleus by the nuclear exosome to avoid inappropriate cellular effects. The exosome is a large RNA degrading machine known to be composed of 10-12 exonucleases that is responsible for most 3'-5' RNA degradation both in the nucleus and cytoplasm, but the regulation of this huge complex remains a mystery. Extended 3' polyadenylation is known to function as a stabilising element in eukaryotic mRNAs, however, recent results indicate that moderate polyadenylation of RNAs can also function as a signal for nuclear exosome-mediated degradation in the nucleus in analogy with what is seen in bacteria . This type of polyadenylation in yeast is carried out by a newly discovered nuclear polymerase called Trf4p [1, 2]. In contrast to the classic poly(A) polymerases, Trf4p does not contain an RNA-binding domain and it is believed that the detection of aberrant RNAs is performed by two other proteins, called Air1p and Air2p, that act in concert with Trf4p. The exosomemediated degradation of RNAs also seems to be dependent of a third protein called Mtr4p, which contains a putative helicase activity important for the unwinding of secondary structures in the RNA substrates during degradation. These three proteins make up the so-called Trf4p/Air1p/2p/Mtr4p polyadenylation (TRAMP) complex, which is able to change the activity of the exosome from its normal slow, distributive activity into a processive degradative activity and hereby stimulate RNA decay. One of the key questions is how the recognition of aberrant RNAs as substrates for TRAMP/exosome degradation is achieved. In other words, what makes an RNA defect? Recent results have shown that the recognition of defect tRNAs is determined by missing modifications or deletions at specific positions in the tRNA molecules . We believe that in such cases, the tRNAs are recognized by a common structural alteration that causes part of the molecule to unfold and thereby mark the tRNA as defect. We therefore want to isolate and crystallize the RNA binding proteins Air1p or Air2p in complex with aberrant tRNA analogous to gain knowledge of how the cell determines which RNAs are good and bad and use this as a model system for how the cell in general recognizes aberrant RNAs. So far, Air1p and Air2p have been cloned from different yeast organisms and the first expression tests of the proteins have been carried out. However, it seems that the proteins, which each contain five zinc knuckle motifs, are insoluble in the absence their normal complex partners. We are therefore in the process of making bicistronic constructs to try to express Air1p/2p in the context of the poly(A) polymerase, Trf4p.  Tollervey et al. Cell. (2005), 5, 713-24.  Keller et al. PLoS Biol. (2005), 6, 0001-0012.