Created by: Milos Tomovic

      T-box riboswitches in gram-positive bacteria are responsible for regulating the expression of aminoacyl-tRNA synthetases and other proteins (1). T-box regulation of these synthetases is in response to varying tRNA aminoacylation levels (1). T-box riboswitches are in the 5' untranslated regions of mRNA and are separated into a ligand binding and regulatory domains (2). The regulatory domain, formed by three complex stems (Stems I-III) is responsible for transcription elongation and translation initiation (2). Stem 1 is the necessary complex for specific, high-affinity tRNA binding and contains several regions that are conserved across all T-box riboswitches (1,2). These regions are the GA/k-turn motif, specifier loop, L3/4, AG bulge, and the distal loop (2). This paper identifies the specific amino acids and nucleotides responsible for the structure and function of the T-box riboswitch Stem I as it forms complex with its cognate tRNA (PDB ID: 4LCK). As well, H/ACA ribonucleoprotein particle pseuduridine synthase (PDB ID: 3LWQ), a comparative protein, is analyzed for structural differences between it and the T-box riboswitch in complex with the cognate tRNA. 

     The Stem I region is an approximately 100 nucleotide region of conserved motifs in the regulatory domain of T-box riboswitch (2). The motifs of the stem are responsible for many functions (2). The GA/k-turn motif, explained in the next paragraph, orients the 3' region of the T-box riboswitch for receiving the tRNA (3). The specifier loop recognizes the cognate tRNA (1). The L3/4 motif likely provides flexibility to the stem for tRNA binding, but more concrete evidence is necessary to conclude this (2). The AG bulge and distal loop motifs form an interlocking T-loop which provides rigidity and a well-formed base-stacking surface in the structure for docking against RNA bases (2). 

       At the proximal end, the k-turn introduces a 120^o bend towards the minor groove in the helical path of Stem I (1, 3). This bend allows for a small enhancement of the interaction between tRNA anticodon and the specifier loop domain (3). The motif of the k-turn has an asymmetrical internal loop that has a canonical (C) flanking stem and non-canonical (NC) flanking stem (3). A pair of 5'-GA-3' dinucleotide sequences terminate the NC stem and mark the kink (3). The phosphate backbone kinks between the second and third unpaired nucleotides (3). The kink is stabilized by stacking of these bases on pairs at the C and NC stem (3). The base of the second nucleotide stacks on the G-C pair of the C stem and the base of the third nucleotide stacks on the G-A pair of the NC stem (3). Subsequent binding of the tRNA to the T-box riboswitch stabilizes the kink in the Stem I (3). 

       Research on Stem I has focused on the specifier loop to determine how the T-box riboswitch binds to the cognate tRNA with high specificity and affinity (2). More is known about this motif of Stem I than any of the other motifs (1-5). A conserved purine (A90), 3' to the specifier nucleotides of the tRNA, stacks under the wobble base pair and hydrogen bonds to A16 in order to increase stability of the specifier loop (1). Upon binding of the tRNA, the specifier residues rotate roughly 34^o which causes the moving of A90 so that it stacks under the duplex (1). tRNA A37 stacks on top of the specifier:anticodon duplex just above the stacking site of A90 and allows for entry of the tA37 nucleotide (1). Any changes to the specifier loop or the adjacent A90 purine are very detrimental to the binding ability of Stem I (1). The Specifier Sequence is three nucleotides that are complementary to the anticodon nucleotides of the cognate tRNA and determine the specificity of the riboswitch (2). 

     The L3/4 motif is thought to cause flexibility in the Stem I by imparting a kink between P3 and P4 which alters the directionality of the helix relative to the axis (2). At the distal end of Stem I, the distal loop and AG bulge interdigitate and form a compact structure of six stacked layers (1). The last of these stacks is formed by three coplanar bases: C44, A56, and G63 (1). A 20^o hinge motion must occur proximal to the T-loops in order to avoid steric clash with tRNA (1). This mutual induced fit between tRNA and Stem I organizes the overall T-box riboswitch 5'-UTR so that it may make the aminoacylation-dependent transcriptional termination decision (1). 

      The T-box riboswitch is a homodimer that associates with the specific tRNA to the Specifier Sequence on the specifier loop of Stem I (3, 4). The importance of Stem I is for high affinity and specificity binding to the cognate RNA (1). The T-box RNA binds to the cognate tRNA through induced fit (2). The alpha helix in the Stem I region at the point of binding is slightly bent to allow for binding without steric hindrance (1). Aside from this helix, the riboswitch has three other alpha-helices and an antiparallel beta-sheet (4). For the purposes of this paper, these subunits will not be discussed since the focus of this paper is on the Stem I domain. In addition, there is no evidence from the sources that any drug developments are being made for the complex. Through Expasy analysis, it was determined that the Stem I-tRNA complex has a molecular weight of 138,696.7 Da (6). Due to the association with cognate tRNA, the isoelectric point was not given in the analysis. Both associated ions, magnesium and strontium, do not associate with the protein but with the tRNA to ensure correct folding (2). 

     Through the use of Dali, Expasy, and PSI-BLAST, several comparative proteins were identified. The purpose of the Dali Server was to find similar proteins based on similarities in the tertiary structures between the proteins (7). Dali determines this similarity by comparing intramolecular distances (7). Expasy and PSI-BLAST work in similar methods by analyzing the nucleotide or amino acid sequence of the protein of interest and comparing that sequence with a databank of proteins (6, 8). The returned proteins are those with the fewest and smallest holes of difference in the primary structure between the comparative protein and protein of interest (6, 8). The cut off scores for analysis are Z>2 and E-value<0.05. After Dali analysis, the protein chosen for comparison was H/ACA ribonucleoprotein particle pseudouridine synthase from Pyrococcus furiosus (7). The results from PSI-BLAST and Expasy analysis did not provide back comparative proteins for use (6, 8). The closest promising protein had an E-value of 5e-43, suggesting that this comparative protein is nearly identical to the protein of interest (8). This particle pseudouridine synthase has a Z-score of 14.5, which indicates a close but not perfect alignment between tertiary structures of the proteins. The particle pseudouridine synthase has three different protein subunits and two RNA subunits, whereas the T-box riboswitch Stem I domain has only two protein subunits and two RNA subunits (9). The synthase is a heterotrimer while the Stem I is a homodimer (2, 9). The RNA subunit specifies the substrate RNA via base complementarity (9). The Cbf5 is the key subunit that both organizes the assembly of the RNP and catalyzes the isomerization process (9). In the case of pseudouridine synthase, the cognate tRNA is bound by three structurally different subunits whereas the cognate tRNA is bound by a homodimer pair of proteins in the T-box riboswitch Stem I domain (9). The pseudouridine synthase minimally requires three of the crucial residues for isomerization: D-85, a hydrophobic ring at position 182, and a positively charged residue at position 184 (9).