The Structure of Ribosomal RNA: A Three‐Dimensional Jigsaw Puzzle

Abstract

CONTENTS

Functional implications of ribosomal protein L2 in protein biosynthesis as shown by in vivo replacement studies

“…This domain, together with domain V of 23 S rRNA, is now generally accepted as being an integral and functional important part of the peptidyltransferase centre [42]. Intra-rRNA, inter-rRNA and tRNA-rRNA cross-linking experiments [43] and hydroxyl-radical footprinting experiments [44] indicated a very close proximity of L2 to potential functionally important components of the peptidyl-transfer reaction (e.g. acceptor end of tRNA, decoding region of the 16 S rRNA, growing peptide chain).…”

Structure of 18 S Ribosomal RNA in Native 40 S Ribosomal Subunits

“…Unfortunately, nothing is known about the folding of 18 S rRNA in the small ribosomal subunit. However, Brimacombe (34,35) has suggested a model for the three-dimensional folding of 16 S rRNA in the prokaryotic 30 S subunit. The general structures of the 16 S-like rRNAs and the basic topology of the 30 S and 40 S subunits are similar (36 -38).…”

New features of 23S ribosomal RNA folding: The long helix 41–42 makes a “U-turn” inside the ribosome

“…Locations of cleavage sites and crosslink sites in regions of the secondary structure of 23S rRNA+ A: Crosslinks from psUp at position 384, in Domain I of the 23S rRNA+ B: Crosslinks from psUp at positions 867, 1045, and 1117, in Domain II of the 23S rRNA+ In each case, the cleavage site is ringed, and the crosslink sites are indicated by lines with arrowheads+ Helices are numbered as in Brimacombe (1995), and the crosslinks are numbered (1-4) according to the positions of the cleavage sites in the rRNA (Table 1)+ C-385 and G-411 proposed by Larsen (1992) on the basis of phylogenetic analysis+ In view of the wellestablished tertiary interaction (Leffers et al+, 1987) already mentioned in the Introduction between the loop ends of helix 22 (in Domain I of the 23S rRNA) and helix 88 (in Domain V), these data also imply that helices 21 and 88 are in close proximity+ Crosslink 2, from the psUp residue at C-867 (Fig+ 5), clearly lies within the secondary structure of the 23S rRNA in helix 38+ As noted in the Results, at least two crosslink sites are involved here, separated by several nucleotides, but the sites cannot be specified precisely as a result of the "stuttering" of the reverse transcriptase in the primer extension analysis (Fig+ 4B)+ The cleavage at C-867 generates a potentially flexible singlestranded "arm" comprising residues 863-867 (although it should be noted that additional base pairs could be formed within the internal loop of helix 38, such as 863-864/912-913), and this flexibility would enable the psUp residue to reach a number of sites on the opposite strand of the helix+ The distance between the two principal reverse transcriptase stops in Figure 4B represents almost one full turn of an A-form RNA helix, suggesting that the crosslink sites are concentrated on one side of the helix+ Crosslink 2 is very similar to the psoralen crosslink observed by Turner and Noller (1983), which they localized to positions C867 and U-913+ Crosslink 4, from the psUp residue at C-1117, is-like crosslink 2-also a "secondary structural" crosslink, within helix 42 of the 23S molecule+ Crosslink 3, on the other hand, from the psUp residue at C-1045, defines an important new tertiary structural feature of the 23S rRNA+ This crosslink, from C-1045 to G-993, demonstrates that the internal loop of helix 42 (Fig+ 5) is folded back so as to contact the base of helix 41, thus generating a "U-turn" within the helix 41-42 region+ The established pseudoknot in this area between nt 1137-1138 and 1005-1006 is likely to be important for the formation of this U-turn; it has been demonstrated (Rosendahl et al+, 1995) that mutations that disrupt the pseudoknot base pairing are detrimental to ribosomal activity, but that compensatory mutations cause the activity to be restored+ The U-turn has the effect of bringing the well-known GTPaseassociated area of the ribosome (defined by the conserved region between nt 1050 and 1110 in helices 42-44) close to the "ring" in Domain II enclosed by helices 36-41 and 45 (Fig+ 5)+ It is noteworthy that such a U-turn is implicit from sequence comparison studies on the large subunit rRNA (e+g+, Gutell & Fox, 1988); large deletions occur within helices 41-42 in certain mitochondrial rRNAs and, if the 41-42 stem did not fold back, these deletions would result in the "misplacing" of the GTPase area in the mitochondria ribosomes+ The 1050-1110 region i...…”

“…Reverse transcriptase analysis of the sites of scission by RNase H in the 23S rRNA+ In each panel, the dideoxy sequencing lanes are indicated by A, C, G, U, and the slots marked S or K are from the cleaved and intact (control) 23S rRNA, respectively+ A: Cleavage site produced by RNase H in the presence of chimeric oligonucleotide No+ 1 (Table 1) Fig+ 3A) shows that the crosslink site lies in the 110-nt stretch between positions ca+ 880 and 990, whereas slot 3 localizes it to between positions 868 and ca+ 940, from which it can be deduced that the site must lie in the 880-940 area+ Slot 2, using an oligodeoxynucleotide complementary to positions 900-920, shows only a faint band (ca+ 40-nt long) in the gel, which suggests that this oligonucleotide may have encompassed the actual crosslink site, and that, as a result, the RNase H digestion was inhibited+ This suggestion was borne out by the primer extension analysis (Fig+ 4B), which shows a strong series of stop signals in the crosslinked sample between nt 909 and 921+ The signals divide into two groups, comprising nt 909-914 and 917-921+ We cannot distinguish whether the stop signals within these groups represent genuine multiple crosslink sites, or whether this is a case of the "stuttering" that is often seen in reverse transcriptase analyses (see Denman et al+, 1988;Brimacombe, 1995, for discussion)+ Nevertheless, the gel of Figure 4B does indicate that there are at least two distinct crosslink sites, separated by 8-10 nt, and this result was observed reproducibly+…”

“…The 50S subunits reconstituted from the four fragmented 23S rRNA species carrying 32 P-labeled psUp at the cleavage sites were separated by sucrose gradient centrifugation and subjected to mild UV irradiation at 350 nm (Rinke-Appel et al+, 1991; Dontsova et al+, 1992)+ The rRNA was then isolated from the irradiated subunits (or from nonirradiated control subunits) by phenol extraction, and the crosslinked regions within the rRNA were localized using the RNase H (Dontsova et al+, 1991;Rinke-Appel et al+, 1991)+ For this purpose, either two or three oligodeoxynucleotides (10-20 nt long) complementary to the 23S rRNA were necessary in the RNase H digestions+ The first oligodeoxynucleotide is complementary to a region upstream of the psUp site, so as to release a short rRNA fragment carrying the 32 P-labeled psUp residue+ This oligodeoxynucleotide is kept constant in the RNase H digestions, whereas the second one (and the third one, if present) is varied so as to excise different sections of the 23S rRNA encompassing the crosslinked target site (or sites)+ In contrast to the digestions with the chimeric oligonucleotides used to generate the initial specific cuts in the rRNA, these RNAse H digestions are performed at 558 with an excess of enzyme (cf+ RinkeAppel et al+, 1991)+ Candidates for the "constant" oligodeoxynucleotide were carefully selected so as to obtain a clean and quantitative digestion in every case+ Thus, in the subsequent gel electrophoresis of the rRNA fragments released by the RNase H, the lanes from the control (nonirradiated) 50S subunits should only show the short fragment carrying the labeled psUp in each case, whereas the lanes from the irradiated subunits should, in addition, show products corresponding to this short fragment crosslinked to the fragments of variable length excised from the target region+ Examples of these gels are illustrated in Figure 3, and will be described below for the individual crosslinks+ When the locations of the crosslink sites had been narrowed down in this way to rRNA regions of less than ca+ 50 nt, the sites of crosslinking within the regions concerned were analyzed by primer extension (Moazed et al+, 1986)+ The substrates for the primer extension reactions were isolated from RNase H digestions on a preparative scale+ For this purpose, the digestions were chosen so as to release the crosslinked material as a relatively large rRNA complex of 100-200 nt (cf+ Fig+ 3A, lane 5, or Fig+ 3B, lane 1), and a control fragment was extracted from the same gel lane in each case+ These control fragments represent the noncrosslinked rRNA from the target region and do not contain the region with the radioactive psUp residue; they could accordingly be located by UV shadowing as bands moving just below the corresponding crosslinked complexes in the gels+ Examples of the primer extension gels obtained from the crosslinked and control fragments in the various cases are shown in Figure 4, and are described below for the individual crosslinks+ As a result of the artifacts that are often observed in primer extension analyses of crosslink sites (see Brimacombe, 1995, for discussion), we only consider a stop signal in the reverse transcriptase gels as being indicative of a crosslink site if it is reproducibly observed within the short rRNA region already defined by the RNase H digestions (cf+ Fig+ 3)+…”