Based on the experiments shown before it seems unlikely that Mtr4 KOW interacts with ssRNA which agrees with literature (Figure 5.5 A)194. Most peaks that show up in presence of ssRNA can also be found in presence of dsRNA with or without the Nop53 peptide.
A possible explanation could be the existence of a low populated conformation of KOW that interacts (maybe unspecifically) with any kind of RNA. Additionally, the negative charge of the RNA could interact unspecifically with parts of the protein.
The NMR data presented in this thesis revealed that structured RNAs and Nop53 bind to the Mtr4 KOW domain via distinct interfaces (Figure 5.5 C, D). The solution data for the peptide complex is in agreement with the crystal structure from our collaborator Sebastian Falk (data not shown). Among the residues affected are leucine, isoleucine and valine which could bind the AIM of Nop53 via hydrophobic interactions and the crystal structure indeed confirms that these residues interact with residues 59-69 of Nop53 (ALFHVDVEGDE) which comprise a small β-strand that packs antiparallel against β-strand 5 of KOW.
The RNA titrations show that KOW interacts with dsRNA and tRNA via the same binding site suggesting recognition of tRNAs (or otherwise structured RNAs) is depending on double-stranded regions within the RNA. This site comprises residues that are good h-bond donors (histidine, serine, asparagine, aspartate and glutamate) which makes them suitable RNA binding residues.
The different binding sites allow simultaneous binding of both Nop53 and RNA substrates, however, based on the NMR titrations the binding sites overlap partially. The residues involved switch to the RNA bound form when both ligands are present (Figure 5.5 E). In presence of 20-fold Nop53 peptide a 7-fold access of RNA was needed to achieve similar shifts as for 5-fold access of RNA only. This indicates that the two ligands influence each other and might modulate affinities of KOW. Individual affinities and ligand concentrations will fine tune complex formation. A simultaneous binding mode seems important for the biological function of Mtr4 and the exosome: The RNA binding site points to the inside of Mtr4 and the ssRNA binding channel of the core while Nop53 binds to the top of the domain. This
arrangement enables Mtr4 to be recruited to its target via Nop53/Utp18 while binding its RNA substrate. Depending on the affinities in the natural complex (means with all cofactors) it could even be possible that upon ribosome/substrate binding Nop53/Utp18 is released from Mtr4.
Overall KOW has a low affinity (based on the NMR titrations where no saturation was reached) for all tested ligands199, which agrees with reports that mutations in the KOW domain reduce RNA binding and unwinding but do not abolish it completely210. Since both the Mtr4 core (Rec domains) as well as Air1/2 bind the RNA substrate, the KOW domain probably assists in RNA binding and assures correct orientation of the substrate to enable proper unwinding and access to the Mtr4 channel. Furthermore, it serves as a platform to mediate interactions to its ribosome targets via Nop53 and Utp18.
Figure 5.5: Interaction modes of Mtr4 KOW domain with different ligands. A) ssRNA might interact with a minor population of KOW (blue, adapted from PDB: 2XGJ) though direct proof is missing. B, C, D, E) Two potential sequential binding schemes for Nop53 and tRNA binding are shown. The binding site for D structured RNAs (grey) differs from the C Nop53 (gold) binding interface. F) The stalk (lilac) interacts with the KOW domain and G) might contribute to RNA binding or modulate it.
The NMR experiments for the extended KOW domain (including parts of the stalk) point at an interaction of the stalk with the KOW domain (Figure 5.5 F) and a possible participation of the stalk in RNA binding or at least modulation of RNA binding (Figure 5.5 G).
A scenario where the stalk modulates accessibility and/or affinity of KOW for RNA seems thus plausible. Due to large shifts of the free extended form compared to the KOW domain a new
backbone assignment would be necessary to allow a proper analysis and extended model.
The influence of the stalk could also be an artefact that could be absent in the context of full length Mtr4 or in presence of cofactors. Here further structural experiments would be necessary, e.g. SAXS/SANS to determine the overall shape.
Based on this work and literature data a possible model for pre-ribosomal RNA processing involves initial binding of Nop53/Utp18 via its AIM to KOW (Figure 5.6 A, B).
Thereby the respective factor recruits the TRAMP-exosome complex to the pre-ribosome and KOW binds the ribosomal RNA precursor (Figure 5.6 C). This binding is accompanied by a conformational change within the AIM of Nop53/Utp18. Here the cofactor might get released from the complex. When the RNA is properly bound by the Mtr4 core and Air1/2 it gets unwound through the central channel of Mtr4 and degraded by the connected exosome (Figure 5.6 D).
Figure 5.6: A model for ribosome precursor-RNA processing via the TRAMP-Exosome complex.
A) Mtr4 (blue, adapted from PDB: 2XGJ) as part of the TRAMP complex (green) binds the exosome (green). B) Nop53//Utp18 bind via their AIM to the KOW domain of Mtr4 and recruit the entire complex to C) the pre-ribosome. Upon binding of ribosomal precursor-RNAs the AIM of Nop53/Utp18 performs a conformational switch and partially releases Mtr4. D) The RNA gets processed and unwound by Mtr4 (TRAMP) and subsequently degraded by the exosome.