In pre-mRNA processing the multi-domain splicing factor U2AF65 recognizes a uridine rich RNA sequence to promote spliceosome assembly. The protein possesses two RNA recognition motifs, RRM1 and RRM2 connected by a flexible linker. PREs data obtained by spin-labelling different residues of either RRM1 or RRM2 in the RRM1–RRM2 construct revealed the presence of a conformational equilibrium between PD-0332991 cost an “open-state”, where both RRM domains are capable of binding the RNA, and a “closed-state”, where only RRM2 binds to the RNA and the RNA binding surface of RRM1 is partially engaged
in electrostatic interactions with RRM2. By analysing the percentage of “open” versus “closed” conformations in the presence of substrate RNAs of different sequence, the authors could correlate the amount of protein in the “open-state” with the efficiency GSK1120212 of the U2AF65–RNA interaction in promoting spliceosome assembly. Furthermore, they could demonstrate
that protein mutations destabilizing the “open-state” are impaired in their ability to bind the RNA. This study demonstrates the usefulness of PRE data for characterizing the relative orientation of protein domains or of distinct components of a complex, including even the detection of multiple conformations. An extensive set of PRE-derived distances can be used to guide molecular docking and determine the conformation of RNP complexes. As mentioned above, site-directed paramagnetic labelling of proteins is only possible in the absence of multiple accessible cysteines. If more than one cysteine is located
on the surface Parvulin of the protein, these residues can be mutated to serine, under the provision that mutagenesis does not alter the protein folding. Alternatively, a different implementation of the PRE effect has been proposed, which does not requires site-directed spin-labelling [44]. A soluble paramagnetic agent Gd(DTPA–BMA) (DTPA: diethylenetriamine pentaacetic acid, BMA: bismethylamide) is added to the solvent, resulting in line broadening of the accessible nuclear spins. This data can be translated into structural information defining the distance of the nuclear spins from the surface, or in other words the solvent accessibility (Fig. 4). Solvent PREs have been used in a combined structure-selection/structure-refinement protocol to calculate the conformation of the Ran-CRM1-PKI NES complex together with sparse NOEs [44]. More recently an empirical function translating solvent accessibility data into structural information has been implemented in Xplor-NIH for structure calculations [45]. A similar approach has been applied to nucleic acids as well.