The height of each letter is proportional to the frequency of the corresponding residue at that position, and the letters are ordered so the most frequent is on the top. The invariant residues are indicated with dots.Figure 2. A representation of the 3D organization of the catalytic site of PARN. The RNA interacting and structurally conserved residues (Asp324, Thr325, Gly70, Gln68, Leu343, Asn288, Lys326) are shown in an electrostatic cloud, whereas the four evolutionary invariant amino acids that conformationally support the catalytic residues are shown in specefill representation (labeled as under-layer, Asp324, Thr325, Gln68, Gly70). The invariant residues that were detected in the PARN protein motifs by our phylogenetic analysis are showing below the 3D structure. evolutionary conserved, which makes them key pharmacological targets. In any case, nucleotides two and three were both less defined, since they were not rigidly fixated in the 3D conformational space of PARN’s active site. This is supported by crystal structure of PARN, where only the first poly(A) nucleotide was stable enough to return a well defined electron density9. The previous observations suggest that having suppressed the potential degrees of movement for the first two nucleotides by interactions with their bases, PARN can precisely coordinate the positioning of the scissile bond towards the catalytic triad (Asp28, Asp292, Asp382). The optimal catalytic positioning of the scissile bond is directed by His377, which hydrogen bonds to the = O oxygen of the phosphodiesteric bond. Consequently, since His377 and Arg99 establish hydrogen bonds with the first nucleotide from different directions, they play a very crucial role in its threedimensional stabilization and positioning in the catalytic site of PARN (Fig. 3B).
This is in agreement with the reported observation that in the crystal structure of PARN, only the first poly(A) nucleotide was stable enough to return a well defined electron density [9]. Previous work has revealed that the b5 strand of one protomer of PARN forms an antiparallel b-sheet with its counterpart fromthe other protomer. This orientation allows the side chains of conserved residues Phe93, Cys108, Phe106, Ile113, Phe123 and Phe127 from one protomer form extensive hydrophobic interactions with the same set of residues from the other one. Among these, Phe123 is an invariant residue in PARN across species, which when mutated to alanine led to loss of activity [9]. Conclusively, Arg99 may represent another important residue as it links the two monomers, it contributes to overall stability, and directs the substrate to an optimal position for the cleavage reaction. The other invariant residue across species that was identified from the phylogenetic analysis, Gln109, is located in the two antiparallel b5 strands of the homodimeric interface region. To understand the role of Gln109 we performed the Gln109Ala and Gln109Trp in silico mutations. The Gln109Ala mutation revealed a significant loss in the packing and association of the two b5 strands (4 fold energetic loss in packing and association), whereas the Gln109Trp mutation constantly failed, since the bulkier Trp residue could not be accommodated in the homodimerization interface region of PARN (Fig. 3A). It is also evident that the bulky side chain of Gln109 defines the shape and the size of the catalytic pocket that is available to accommodate the poly(A) substrate.Figure 3. The role of Arg99 in the catalytic mechanism of PARN. (A) PARN – Poly(A) interactions have been calculated for both active sites. The Arg99 residues have been highlighted while they are H-bonding with the base moiety of the first poly(A) nucleotide. (B) The interaction map of poly(A) and the catalytic site of human PARN, showing the water mediated bridges of the Aspartic residue attacking the first phosphodiesteric bond, and the vital contribution of the invariant, structurally conserved His377 residue. Insights into PARN’s Nuclease Domain and Interdimeric Interface
We performed structural molecular modeling study of the nuclease domain of PARN. The in silico analysis revealed a series of previously unreported amino acids, which are essential for the function of the enzyme. A conformational under-layer in its catalytic site is shaped by Gly70, Gln68, Asp324 and Thr325 residues (Fig. 2); Gly70 and Gln68 provide structural support for Asp28 and Glu30 catalytic residues, while Asp324 and Thr325 support the poly(A)-interacting Lys326 and His287 residues [9]. The functional role of the PARN under-layer catalytic site amino acids was consistent with the phylogenetic analysis of PARN,where Gly70, Gln68, Asp324 and Thr325 residues were found to be invariant across species. Additionally, in silico mutagenesis studies followed by molecular dynamics simulations (MDs), to each one of the Asp324, Thr325, Gly70 and Gln68 amino acids produced structural conformational changes in the relative positioning of the supported catalytic residues. More specifically, in silico mutation of residues Asp324 and Thr325 to either Alanine (Asp324Ala and Thr325Ala) or to the bulkier Phenylalanine residue (Asp324Phe and Thr325Phe) resulted in loss of the hydrogen bonds between the Lys326 and the Asn288 amino acids with the second scissile bond of the poly(A) substrate (Fig. 3B). Substitution of Gly70 to either Alanine or Phenylalanine aminoacid resulted in the shifting of the three Aspartic residues towards the poly(A) substrate, which consequently pushed the latter away from the catalytic site, losing the Arg99 hydrogen bonding. Finally, mutating the Gln68 residue to either Alanine or Phenylalanine resulted in a slight rearrangement of the 3D positioning of the Glu30 residue that led to the tilting of the whole poly(A) substrate and the complete loss of its hydrogen bonding interactions with the catalytic site of PARN. These findings suggest that an evolutionary conserved and highly sophisticated under-layer structure in the catalytic site of PARN is essential for the function of the enzyme (Fig. 2). Furthermore, it was observed that although the catalytic triad was in very close proximity to the scissile bond, it did not seem to directly interact with it [9]. In depth examination of the active site revealed a smaller cavity within the active site, which in the original X-ray structure coordinate file (RCSB entry: 2A1R) accommodates two water molecules. A MD simulation was set in the presence of the crystallographic waters, and concluded that two water molecules had occupied the small pocket in the active site, now linking Asp28 and Asp292 via a H-O-H bridge to the -P?O group of the scissile bond, whereas Asp382 now interacted with via a water mediated bridge with the = O group as His377 amino acid (Fig. 3B).