form the hairpin structure. Stability of RNA hairpin depends on the length of loop size. Stability decreases with increasing loop length, and four or five nucleotides in the loop form the most stable hairpin [14]. Closing base pair, which is a base pair located at the boundary between the stem and loop region, also affects the stability. G·C and C·G base pairs are preferred to make stable and robust hairpin structures [15]. Some specific loop sequence, which is typified by UNCG, in which N is any nucleobase, can drastically stabilize the loop region [16]. For example, 5′-UUCG-3′ loop with C·G closing base pair forms one of the most stable hairpin loops, which can be often seen in natural RNAs [17]. NMR structure of 5′-GGACUUCGGUCC-3′ demonstrated that the hairpin loop structure is stabilized by unusual base pairing between uracil and guanine in the loop, in which glycosidic bond angle of guanine is in syn conformation. The structure is also stabilized by a cytosine-phosphate contact and extensive base stacking on the crossing base pair (Figure 2.8) [18].
2.6.2.2 Bulge Loop
Bulge loop is a region in which unpaired nucleotides present on one strand of a continuous duplex. In structures of natural RNAs, the most common bulge loop contains one nucleotide. When single nucleotide bulge loop forms internal stack within the duplex, the helicity of the duplex is distorted, resulting in a kink of the helix axis (Figure 2.9). When the bulged nucleotide is looped out in solvent, the overall duplex geometries can remain close to the canonical A-form (Figure 2.9). The bulged nucleobases create unique recognition sites for proteins both directly, by acting as a molecular handle, and indirectly, by distorting the RNA backbone and allowing the proteins access to base pairs in a groove [19]. The kink of the helix caused by the bulge loop also contributes to the shaping of the overall RNA structure. Thermodynamic parameters for single nucleotide bulges indicate that the bulges destabilize the helix. The destabilization effect little depends on the identity of the bulged nucleobase [20]. On the other hand, destabilization does depend on the adjacent base pairs [20b]. As it can be expected, bulge loop with longer nucleotides more destabilizes the RNA structure [21].
Figure 2.8 Hairpin structure. (a) General secondary structure of hairpin. (b) Sequence and tertiary structure of typical RNA hairpin with stable UUCG tetraloop sequence (PDB ID: 2KOC). Top view of UUCG tetraloop sequence is shown on right side of the structure. Hydrogen bonds formed within the UUCG tetraloop are shown as dashed lines.
Figure 2.9 Bulge structure. (a) General secondary structure of bulge. (b) Sequence and tertiary structure of RNA hairpin containing single nucleotide bulges, which is derived from an RNA element responsible for dynein-mediated localization of Drosophila mRNA (PDB ID: 2KE6). Bulged nucleobases are emphasized dark. Secondary structure and side view of the region forming single nucleotide bulges are shown in right boxes surrounded by solid and dashed lines.
2.6.2.3 Internal Loop
An internal loop is formed when single-stranded loops on two strands are enclosed by adjacent two stems (Figure 2.10). Internal loops can be symmetrical or asymmetrical with respect to the number of loop residues on each strand. In the symmetrical internal loops, there is high possibility of forming the unusual base pairs, which are described above, between the nucleobases present at opposite positions. Fully paired and stacked internal loops consisting of eight unusual base pairs have been structurally observed (Figure 2.10) [22]. In contrast, asymmetric internal loops often become motifs that shape a three-dimensional RNA structures by bringing large changes in the helical structure such as bends and turns. Certain asymmetric internal loop motifs, such as kink-turn (K-turn) and reverse K-turn, have been identified and characterized as resulting in sharp turns important for tertiary structure formation (Figure 2.10) [23]. By considering the possible internal loops diversities, relatively few experimental data are available for stabilities of internal loops.
Figure 2.10 Internal loop structure. (a) General secondary structure of internal loop. (b) Sequence and tertiary structure of RNA hairpin containing internal loop forming consecutive mismatched base pairs (PDB ID: 1MNX). RNA sequence is derived from loop E region of 5S ribosomal RNA. Nucleobases forming mismatched base pairs in the internal loop, in which base pairs are shown with black circles in the secondary structure, are emphasized dark. Sequences and structures of typical kink-turn (c; PDB ID: 5FJ1) and reverse kink-turn (d; PDB ID: 1ZZN) motifs. Mismatched base pairs are shown with black circles in the secondary structures.
2.6.3 Elements in Tertiary Interactions of RNA
RNAs are commonly represented planarly as their secondary structure. However, most RNAs are folded into their defined tertiary structures that correlate with their unique functions including contribution to the gene regulation. To form a stable tertiary structure, three-dimensional interactions between secondary structure motifs including loop–loop, loop–helix, and helix–helix interactions are indispensable (Figure 2.11) [24].
Figure 2.11 Distribution of RNA tertiary motifs except coaxial stacking of helices in the nonredundant data set of 54 high-resolution crystal structures [24a].