Solution structure of the fourth FF domain of yeast Prp40 splicing factor

Abstract

Prp40 protein was originally identified as a suppressor of 50 end U1 RNA point mutations.1 Prp40 is a U1 snRNP-associated protein that participates in the early steps of yeast premessenger RNA splicing. Prp40 associates with the branch-binding point protein to bring the 50 splicing site and the intron branch point into spatial proximity.2 Additionally, Prp40 has been implicated in the binding to the phosphorylated C-terminal domain (herein referred as phospho-CTD) of RNA polymerase II through regions involving the WW and FF domains.3 However, a subsequent study on the structure of the Prp40 WW domain pair also showed that, in the absence of additional FF domains, the WW domains do not interact with the phospho-CTD repeats.4 The solution structure of the first FF domain of Prp40 has been determined.5 That study also examined the binding of Prp40FF1 to the splicing factor Clf1 and to a pair of bisphospho-CTD repeats. The binding site for the association with the first TPR motif of Clf1 involves the helices a2, the 310, and the N-terminal half of a3. In contrast, no interaction was detected for the Prp40FF1 domain with the phospho-CTD repeats and for the Prp40FF4 domain with the TPR motif of Clf1.5 Recently, other ligand partners have been identified for Prp40 FF domains, namely Snu71, a component of U1 snRNP, and Luc7, a splicing factor associated with U1 snRNP involved in the 50 splicing site recognition.6–8 Furthermore, it was shown that only the region comprising the first two FF domains of Prp40 is critical for yeast viability, whereas the deletion of the region including FF3 and FF4 results in a slow-growth phenotype.6 These results are in agreement with a previous report that showed that a deletion of the FF1 domain and a fragment of FF2 of Prp40 caused yeast lethality.9 To further study the remaining FF domains present in Prp40, we cloned and produced constructs corresponding to the FF2, FF3, and FF4 domains. Of these, only the constructs corresponding to FF4 gave a folded and stable sample. Indeed, several constructs spanning FF2 and FF3 and even that of the FF1-2 pair were either partially structured or unstable after refolding from inclusion bodies. Here we report the solution structure of the Prp40FF4 domain. Furthermore, prompted by the observation that the charge distribution of the FBP11FF1 region involved in the interaction with the bisphospho-CTD repeats10 is partially conserved in Prp40FF4, we also examined whether this domain also interacted with the phosphoCTD repeats, but no binding was detected under our experimental conditions.