Polyamines Contribute to the Stability of the tRNA Tertiary Structure
We demonstrate via iodine cleavage of phosphorothioated tRNAs that, in solution, tRNAs adopt different conformations under polyamine and conventional buffer conditions. Under conventional buffer conditions a Mg2+ shift from 10 to 20 mm had no significant influence on tRNA conformation. In contrast to conventional buffers, strong protection using polyamine buffer was seen, for example, at the phosphate groups of the anticodon stem (24–26 and 39–43) and around phosphate 10 (residues 7, 8, and 11–13; see TableII). This suggests that, despite the low Mg2+concentration, the tRNA tertiary structure in the polyamine buffer is more compact and stable than in the conventional buffer system with 10–20 mm Mg2+. This is surprising, as the general wisdom is that RNA structure becomes more stable at higher Mg2+ concentrations. This indicates that polyamines not only compensate for the lower magnesium, but also rather induce a more compact structure, a feature that might be related to the near in vivo performance of protein synthesis under polyamine conditions, in contrast with the conventional buffer systems (for review, see Ref.23).
Interestingly, the regions of increased protection in polyamine buffer locate in the vicinity of two polyamine binding sites. In a crystal structure of yeast tRNAPhe, besides four distinct Mg2+ ions, two spermine molecules are found. The first is located in the major groove at one end of the anticodon stem (phosphates 25, 26, and 41–43), and the second is near the variable loop and curls around phosphate 10 in a region where the polynucleotide chain takes a sharp turn (26). Because the differences in protection between the polyamine system and the conventional systems accumulate at the polyamine binding sites, we conclude that many of these additional protection sites can be attributed to the bound polyamines, thus underlining not only the impact of polyamines on RNA stabilization, but also the sensitivity of the phosphorothioate technique.
Refining the Ribosomal Components Responsible for Fixation of a tRNA at the P Site
The reduction in the accessibility of the phosphate groups of an AcPhe-tRNA located at the P site (16, 19) can be convincingly correlated with the points of contact between the ribosomal matrix and a P site fMet-tRNA as revealed by cryo-electron microscopy (27).
Here we analyze deacylated tRNA, and find that the protection patterns obtained with deacylated tRNAs bound to individual ribosomal subunits complement each other so as to equate with the pattern observed when a tRNA is bound to the P site of a 70 S ribosome. This observation allows a sharp delineation of the tRNA contact regions within the 70 S ribosome that are contributed by each of the component subunits. The small subunit contacts include the anticodon loop and the first two base pairs of the adjacent anticodon stem (positions 30 ± 1 to 43 ± 1), whereas the remaining 85% of the tRNA is in contact with the large subunit (Fig. 4C). The contribution of each subunit to the overall distribution of contacts made with the P site tRNA is in remarkable concordance with the 5.5-Å map of the tRNA·70 S complex (see Table III in Ref. 10) and also with the mimic of an anticodon-stem loop structure in the crystal of 30 S subunit (28). Fig. 6 illustrates the excellent agreement of our data with the 5.5-Å map of a 70 S complex; the 30 S protections of the anticodon stem-loop structure (positions 30 ± 1 to 41 ± 1 in yellow) are located exclusively in the neighborhood of 30 S components (gold and red) of the 70 S map, whereas the remaining portion of the tRNA (dark blue) lies within the domain of the 50 S subunits (cyan andgreen).
tRNA neighborhoods at the ribosomal P site of the 70 S ribosome according to Ref. 10 (see also Table IV).A, view from the A site; B, view from the E site. tRNA: yellow, contacts with the small subunit;blue, contacts with the large subunit; pink, strongly protected sites in both tRNAPhe and tRNA at the P site (see Fig.2C). Regions of the ribosomal components that are within a 10-Å radius of the P site tRNA are indicated with spheres. 30 S components: gold, 16 S rRNA; red, S-proteins. 50 S components: cyan, 23 S rRNA;green, L-proteins.
Furthermore, we have determined a number of strong protections that are common between two different species of elongator tRNA, namely tRNAPhe and tRNAMet. We believe that these phosphates may represent strategic fixation points on a deacylated tRNA at the P site. If this were so, then one could expect conservation of the tRNA bases adjacent to these phosphates and conservation of the neighboring ribosomal components. Indeed, this is the case. Eight of 10 strong protections are adjacent to conserved bases of the tRNA (TableIII), and a detailed inspection of the 5.5-Å map of the 70 S complex revealed that the ribosomal components neighboring the 10 tRNA bases are remarkable conserved (Table IV). The rRNA bases are conserved in >95% bacteria and 80–100% across all three phylogenetic domains. Furthermore, we identify a number of conserved positions of ribosomal proteins that neighbor these phosphates, for example, position 120 (E. coli numbering) of S13 is always a Lys or an Arg residue and lies next to tRNA base G30, a highly conserved Arg-128 neighbors tRNA base Y32 and positions Arg/Lys-56 and Arg/Lys-64 of L5 are in close proximity to C56. This further corroborates the suggestion that the identified tRNA nucleotides are of strategic importance for tRNA fixation at the P site.
Strongly protected bases in both tRNAPhe and tRNA
Contact sites with tRNA phosphates at the P site that were strongly protected in two different elongator tRNAs, namely tRNAPheand tRNA (see also Fig. 6)
Identification of a Mobile Ribosomal Domain Associated with tRNA Transport
Although the protection experiments present a “static” picture of the ribosome, which is exemplified by the identification of fixation of the P site tRNA as described in the previous section, a comparison of protection patterns under different conditions and between different states enables an interpretation of two dynamic features of the ribosome.
First, the protection patterns afforded by isolated 30 S and the 50 S subunits could be combined to reconstruct the P site pattern of 70 S ribosomes. Isolated 30 S subunits have a single binding site, the prospective P site after association with the large subunit, as demonstrated with binding experiments (18) and via the toeprinting method (29). In contrast, isolated E. coli 50 S subunits bind exclusively deacylated tRNA to the E site and have no available P or A site (18, 30). The fact that the 50 S E site pattern is practically the same as the 50 S part of the 70 S P site pattern seems to be reminiscent of a P/E hybrid site of the hybrid site model for elongation, where the tRNAs are thought to creep through intermediary hybrid sites (A/P and P/E) before arriving, after translocation, at the classical P and E sites (Ref. 31; see also Ref. 32 for review). However, the similarity with a hybrid site does not hold here, because the 70 S P site pattern was obtained under polyamine-buffer conditions, where the tRNA is found in a canonical P site (22).
We note that a similar conservation in protection patterns was observed between pre- and post-translocation complexes (PRE and POST, respectively), with a deacylated tRNA at the P site in the PRE state and at the E site in the POST state. This led to the α-ε model for the ribosomal elongation cycle (reviewed in Ref. 23), which proposes the existence of a movable domain that binds and guides tRNAs during translocation. The movable domain contains two binding regions, α and ε, each of which bind a tRNA with a characteristic protection pattern. During translocation, the α-region carries a tRNA from the A to the P site and the ε-region a tRNA from the P to the E site.
Our finding that the protection pattern of a tRNA bound to isolated 50 S subunits (E site) is the same as that of the 50 S part of the 70 S P site can be interpreted in the frame of the α-ε model to suggest that the ε-part of the mobile tRNA carrier is at the E site in isolated 50 S subunits, but “swings” into the P site upon association with the 30 S subunit forming 70 S ribosomes.
Second, the accessibility pattern of a deacylated thioated tRNA in the P site of programmed ribosomes is almost identical under polyamine and conventional buffer conditions. However, it is known from cryo-electron microscopy that the locations of the tRNA are strikingly different,i.e. a deacylated tRNA is found at a classical P site under polyamine conditions and at a P/E hybrid site under conventional conditions (12). This suggests that the ribosomal components that hold the 50 S portion of the tRNA are located in a classical P site under polyamine conditions but slip into the E site position under conventional buffer conditions. The physiological relevance of the latter finding is immediately compromised by the buffer conditions themselves, i.e. their non-physiological nature (see Ref. 23for discussion), but may nevertheless provide some insight into the mechanism of translocation.
In the frame of the α-ε model, the conservation of protection patterns between the P and P/E sites suggests that under conventional buffer conditions the ε module of the movable domain has slipped into the E site on the 50 S subunit. This situation induced by non-physiological buffer conditions may seem again reminiscent to a hybrid site. However, the hybrid site model does not propose a movable domain and thus would predict alternative patterns for P/P and P/E sites.
The α-ε model suggests that movement of the tRNAs occurs simultaneously on both large and small subunit in a co-coordinated fashion. Our observation of similar contacts between P/P and P/E indicates that the mutual arrangement of the α and ε regions of the movable domain may differ in parts substantially, when moving between PRE and POST states. Although tRNAs in the PRE and POST states display a similar mutual arrangement relative to each other (the angles between the tRNAs are 39° and 35° in the PRE and POST state, respectively (Ref. 12)), the positions of the CCA ends differ dramatically. Prior to translocation the CCA ends of the two tRNAs present at A and P sites are directly adjacent at the peptidyl-transferase center, an obvious requirement for peptide-bond formation. Following translocation the CCA ends are separated by over 50 Å (10, 12); after formation of the peptide bond, there is no requirement that the CCA ends remain together. This finding indicates that the postulated α region and ε region do not move strictly side-by-side during translocation.
Because the protection patterns encompass the entire tRNA, from the anticodon loop to the acceptor stem, a contiguous structure spanning the intersubunit space, from the decoding center to the peptidyl-transferase center, should exist. A potential structure has been identified, termed bridge B2a, in 70 S ribosomes (27). A major component of bridge B2a is the universally conserved stem-loop of H69 of 23 S rRNA, which has been proposed to undergo conformational change upon subunit association, enabling it to bridge the intersubunit space and to make contacts with both A and P site tRNAs (4). Other candidates for the movable domain include the upper region of the h44 of the 16 S rRNA (33) and parts of the ribosomal protein L2 (34, 35).
Codon-Anticodon Interaction at the P Site Is a Prerequisite for 30 S-tRNA Contacts within the 70 S Ribosome
Although representing only one of many findings presented in this paper, we believe the implications of this section are worthy of emphasis. Specifically, when the protection pattern of a P site tRNA bound to a non-programmed 70 S ribosome (i.e. no mRNA) was assessed, to our surprise the 30 S subunit did not contribute to the protection pattern at all (Fig. 5), whereas the 50 S pattern was similar to that observed with programmed 70 S ribosomes (Fig. 4). This result suggests that codon-anticodon interaction at the ribosomal P site of 70 S ribosomes is essential for 30 S contacts and that the additional 30 S contacts, those outside of the anticodon, are not available in the absence of codon-anticodon interaction. This finding agrees with and extends a previous observation that 30 S subunits in the absence of mRNA do not bind any tRNA at Mg2+ concentrations that are well suited for protein synthesis (18). Furthermore, this result implies that the 30 S subunit undergoes a conformation change upon codon-anticodon interaction, resulting in additional contacts that further stabilize the P site tRNA.
The conformation of a tRNA in solution depends on the buffer conditions. Under in vivo near conditions (polyamine buffer), the conformation differs from that observed under conventional buffer systems regardless of whether the [Mg2+] is 10 or 20 mm. However, the buffer systems have only little influence on the accessibility of the tRNA phosphates if the tRNA is bound to the P site. An analysis of our findings led to the following conclusions. 1) A comparison of the contact patterns of two different elongator tRNAs at the P site of programmed 70 S ribosome identified 10 common and highly protected sites that might be of strategic importance for the fixation of a tRNA at the P site. 2) The accessibility or contact patterns of the tRNAs with the isolated subunits in the presence of mRNA can be combined to produce the pattern seen at the P site of 70 S ribosomes, thus allowing a sharp delineation of the regions of a tRNA in contact with the 30 and 50 S subunits within the programmed 70 S ribosome. 3) The contact pattern of non-programmed 70 S ribosomes is almost identical to that of isolated 50 S subunits, indicating that codon-anticodon interaction at the P site is required for 30 S contacts with the tRNA. (4) On the basis of our results, we propose the following scheme for conformational rearrangements within 70 S ribosomes upon subunit association and P site tRNA binding. (i) Upon association of the ribosomal subunits forming the 70 S ribosome, the tRNA carrier of the 50 S subunit shifts from the E site to the P site. (ii) Binding a tRNA to the P site in the presence of mRNA establishes codon-anticodon interaction. This in turn induces a conformational change of the 30 S subunit that allows further stabilizing interaction with this subunit, in addition to those already existing with the 50 S subunit, both of which may be important for subsequent translocation. The presence of movable domains supports the α-ε model for the ribosomal elongation cycle.
Messenger RNA is a single stranded nucleic acid containing the sugar ribose and the bases adenine, guanine, cytosine and uracil. It is capable of copying the DNA inside the nucleus in a process called transcription and bringing that message to a ribosome in the cytoplasm. Once the mRNA attaches to a ribosome, a different type of RNA called transfer RNA will bring appropriate amino acids to the ribosome to construct a polypeptide chain. Transfer RNA are folded RNA molecules with an area called an anticodon. It can recognize a specific codon on the mRNA and bring the appropriate amino acid that the codon designates. For example, the first triplet or codon reads AUG. This represents START or the amino acid methionine. The transfer RNA will carry this amino acid to the ribosome and translation begins. Each triplet will be translated, adding another amino acid to the growing polypeptide chain in a process called elongation. Finally, when a stop codon is reached, the process of translation ends and the polypeptide detaches from the ribosome and folds into a functional protein.