Crosslinking of 4.5S RNA to the Escherichia Coli Ribosome in the Presence Or Absence of the Protein Ffh

Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the Escherichia coli ribosome in the presence or absence of the protein Ffh

JUTTA RINKE-APPEL,1 MONIKA OSSWALD,1 KLAUS VON KNOBLAUCH,1 FLORIAN MUELLER,1 and RICHARD BRIMACOMBE1; and PETR SERGIEV,2 OLGA AVDEEVA,2 ALEXEY BOGDANOV,2 and OLGA DONTSOVA2 1 Max-Planck-Institut für Molekulare Genetik, D-14195 Berlin, Germany 2 Department of Chemistry, Moscow State University, Moscow 119899, Russia

ABSTRACT Radioactively labeled 4.5S RNA containing statistically distributed 4-thiouridine residues in place of normal uridine was prepared by T7 transcription. The ability of this modified 4.5S RNA to form a complex with the protein Ffh was demonstrated by a gel shift assay. The modified 4.5S RNA, with or without Ffh, was added to Escherichia coli ribosomes under various conditions, and crosslinking from the thiouridine residues was induced by irradiation at 350 nm. The crosslinked ribosomal components were analyzed by our standard procedures. Two clearly defined types of crosslinking were observed. The first was a crosslink to 23S rRNA, which was entirely dependent both on the presence of Ffh and a nascent protein chain in the 50S subunit. This crosslink was localized to nt ; 2828–2837 of the 23S rRNA, from position 84 of the 4.5S molecule. The second type of crosslinking, to the 30S ribosomal subunit, was independent of the presence of Ffh, and was found both with vacant 70S ribosomes or isolated 30S subunits. Here the crosslink was localized to the 39-terminal region of the 16S rRNA, from positions 29–50 of the 4.5S RNA. Cross- linking to ribosomal protein S1 was also observed. The known crystal structure of the protein Ffh/4.5S RNA fragment complex was extrapolated by computer modeling so as to include the whole 4.5S molecule, and this was docked onto the ribosome using the crosslinking data. The results are discussed in terms of multiple functions and binding sites of the 4.5S RNA. Keywords: elongation factor G; functions of 4.5S RNA; nascent peptide chain; ribosome structure; signal recognition particle

INTRODUCTION (Gowda et al+, 1997; Jovine et al+, 2000)+ An X-ray crys- The prokaryotic equivalent of the signal recognition par- tallographic structure has recently been reported for ticle (SRP) is the 4+5S RNA, complexed with the 48 the Ffh M domain complexed with a 49-nt RNA frag- +, , kDa protein Ffh (Poritz et al+, 1990; Ribes et al+, 1990)+ ment (Batey et al 2000) and this structure shows a Protein Ffh is organized into three distinct domains, prominent groove in the protein moiety that has been whereby the C-terminal M domain is the one that is proposed to interact with the signal peptide (Keenan +, + involved in interactions both with the 4+5S RNA and et al 1998) , with the signal peptide sequence at the N-terminus of a In contrast to these detailed structural studies very nascent protein chain emerging from the 50S ribo- little is known about the sites of interaction between the somal subunit (Römisch et al+, 1990; Zopf et al+, 1990)+ SRP (either prokaryotic or eukaryotic) and the ribo- + , The 4+5S RNA correspondingly consists of 114 nt, which some Furthermore there is a considerable body of + are folded into a single long irregular hairpin (Lentzen evidence indicating that the 4 5S RNA has more than et al+, 1996; Schmitz et al+, 1996), the binding site for one function and hence probably more than one ribo- + , + the M domain of protein Ffh being contained within a somal binding site First the ratio of 4 5S RNA to pro- : , stretch of about 43 nt at the loop end of the hairpin tein Ffh in the cell is about 4 1 (Jensen & Pedersen 1994), suggesting that the 4+5S molecule may have + : , functions that do not involve Ffh Consistent with this is Reprint requests to Richard Brimacombe Max-Planck-Institut für + Molekulare Genetik, Ihnestrasse 73, D-14195 Berlin, Germany; e-mail: the now well-established fact that 4 5S RNA is able to brimacombe@molgen+mpg+de+ interact with elongation factor G (EF-G; Shibata et al+, 612 Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 613

1996), and the binding site for the factor has been fected their ability to form a complex with Ffh+ The shown to lie within the same 4+5S region that binds to extent of complex formation was analyzed by a gel shift Ffh (Nakamura et al+, 2001)+ In other studies, mutations assay (cf+ Lentzen et al+, 1994; Suzuma et al+, 1999), have been identified in ribosomal RNA that suppress an example being illustrated in Figure 1+ It can be seen the requirement for 4+5S RNA (O’Connor et al+, 1995); that as the proportion of thio-UTP in the 4+5S transcrip- these mutations are distributed over the 16S and 23S tion mixtures is increased, the ability to form a complex rRNA at locations that are widely separated in the three- with Ffh is progressively reduced+ For the subsequent dimensional structures (Ban et al+, 2000; Wimberly et al+, crosslinking experiments, an input ratio of thio-UTP:UTP 2000) of the ribosomal subunits, again suggestive of of 5:1 was chosen, a value between those of lanes 2 multiple binding sites for the 4+5S molecule+ and 3 in Figure 1, leading to approximately 50% com- We are interested in investigating the interactions plex formation+ The smeared nature of the residual 4+5S between 4+5S RNA and the Escherichia coli ribosome, RNA bands in lanes 1 to 4 of Figure 1 suggests, how- with the help of two different approaches+ In the first ever, that the complex is rather unstable in this gel approach, a complex has been isolated consisting of electrophoresis system, and hence that the level of the M domain of protein Ffh, 4+5S RNA, and the 70S 50% complex formation is likely to be an underesti- ribosome carrying a nascent peptide (Avdeeva et al+, mate+ In control experiments with transcripts contain- 2002), with a view to making cryoelectron microscopic ing no thio-UTP, complex formation was complete under and structural probing studies+ This article however is the same conditions (data not shown)+ concerned with the application of the second approach, Thio-UTP is incorporated into RNA less efficiently which is to make use of established crosslinking tech- than normal UTP during T7 transcription (Dontsova et al+, niques (Rinke-Appel et al+, 1993; Dontsova et al+, 1994) 1994), and the ratio of thio-U to U in the transcribed to identify points of contact or close neighborhood be- 4+5S RNA was accordingly assessed by digestion with tween the 4+5S RNA and the components of the ribo- ribonuclease T2 followed by two-dimensional thin-layer some+ To this end, we have prepared a 4+5S RNA analog chromatography (Stade et al+, 1989) in a system that carrying statistically distributed 4-thiouridine (thio-U) res- separates thio-Up from Up+ After digestion with ribonu- idues in place of normal uridine+ Irradiation with UV clease T2, the 32P-label moves to the next nucleotide light at 350 nm generates effectively zero-length cross- 59 with respect to each U residue, but, because there links from the thio-U residues, and previous experience are a number of sites in the 4+5S molecule where two with similar analogs of mRNA (Dontsova et al+, 1992; Rinke-Appel et al+, 1993) or 5S rRNA (Dontsova et al+, 1994; Dokudovskaya et al+, 1996) has indicated that this crosslinking method gives reliable results that are compatible (Sergiev et al+, 2002) with the established X-ray crystallographic structures of the ribosomal subunits (Ban et al+, 2000; Schluenzen et al+, 2000; Wim- berly et al+, 2000)+ Here we present the results of crosslinking experiments with this 4+5S RNA analog, and describe the analysis of a highly specific Ffh- dependent crosslink between the 4+5S RNA and 23S rRNA, as well as Ffh-independent crosslinks from the 4+5S RNA to the rRNA and protein moieties of the 30S subunit+

RESULTS

Complex formation between thio-U-containing 4.5S RNA and protein Ffh Protein Ffh carrying a His-tag was isolated from an expression clone of E. coli, and 32P-labeled 4+5S RNA was transcribed from a PCR-generated DNA fragment using T7 polymerase, as described in Materials and Methods+ The transcription mixtures contained both thio- FIGURE 1. Gel shift assay showing complex formation between Ffh UTP and a-32P-UTP, and in preliminary experiments and 32P-labeled 4+5S RNA containing increasing amounts of thio-U+ In lanes 1 to 4 the ratio of thio-UTP:UTP in the transcription mixtures the molar ratio of thio-UTP to UTP was varied to see was 3+3:1, 4+3:1, 7+3:1, and 9+0:1, respectively+ Lane 5 is a control whether incorporation of thio-U into the transcripts af- sample minus Ffh+ Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

614 J. Rinke-Appel et al. or more adjacent U residues occur (see Fig+ 4A), the in place of the 4+5S–Ffh complex+ In further experi- thin-layer chromatograms show spots corresponding ments, vacant 70S ribosomes or isolated 30S subunits to both radioactive Up and radioactive thio-Up, from were used in place of the translation mixture; in these which the ratio of thio-U to U in the transcript can be cases (also with or without Ffh), the sucrose cushion estimated+ At the input ratio of thio-UTP:UTP of 5:1 centrifugation step was omitted+ chosen for the crosslinking experiments, the ratio of In typical experiments, with an initial input of 20– thio-U to U incorporated into the 4+5S product was found 50 ϫ 106 cpm of 32P-labeled thio-U-containing 4+5S by this method to be approximately 30%, a value sim- RNA, the amount of radioactivity cosedimenting with ilar to that observed in our previous studies on 5S the 70S ribosomes in the first sucrose gradient was of rRNA (Dontsova et al+, 1994)+ the order of 500–1,000 ϫ 103 cpm (i+e+, about 2%)+ This It is also noteworthy in this context that conforma- low level of binding reflects a number of factors, includ- tional variants of the 4+5S RNA have been mentioned ing the small proportion of ribosomes engaged in syn- in the literature (Lentzen et al+, 1994)+ We occasionally thesis of the full-length b-lactamase peptide, a nonideal observed such a variant in our T7 transcripts on gels length or composition of that peptide with respect to similar to that of Figure 1, which appeared as a band SRP binding, or an intrinsically weak binding of the running at a position halfway between that of the 4+5S 4+5S RNA per se, causing the complex to dissociate RNA and the complex+ This conformational variant during the sucrose gradient step+ Nevertheless, the (which could also possibly be a dimer) had the same crosslinking results were extremely reproducible+ The nucleotide sequence as the 4+5S RNA, but was unable second sucrose gradients (at low magnesium concen- to participate in complex formation with Ffh and had tration) showed that 30–90 ϫ 103 cpm of 4+5S RNA the additional property of being remarkably resistant to remained bound to the 50S subunits, and this radio- ribonuclease digestion+ activity was distributed approximately equally between the 23S rRNA and the ribosomal protein fractions in the third sucrose gradient (in SDS)+ In the absence of Ffh, Preparation of ribosomal complexes or if the mRNA was omitted from the translation mix- and crosslinking ture, these levels of radioactivity were reduced by an To generate a nascent peptide on the ribosome carry- order of magnitude+ In contrast, the amount of radio- ing a signal sequence, a T7 transcript was prepared activity associated with the 30S subunit was about three- from a fragment of the b-lactamase gene (see Materi- fold higher (100–300 ϫ 103 cpm), whereby about 20% als and Methods)+ This transcript contained the se- of these counts cosedimented with the 16S rRNA and quence coding for the first 69 amino acids of the protein 80% with the protein fraction in the final sucrose gra- (including the signal peptide) with no termination co- dient+ These levels of radioactivity were not affected by don, so that the nascent peptide would remain bound the omission of Ffh or mRNA from the reaction mix- to the ribosome as a run-off translation product+ Tests tures, and were even slightly increased when vacant with radioactive amino acids showed that a peptide 70S ribosomes or isolated 30S subunits were used in product of the expected length could be isolated from place of the full translation system+ In all cases, when translation mixtures containing this mRNA analog in an 4+5S RNA containing no thio-U was used, no significant E. coli protein-synthesizing system, and that approxi- levels of radioactivity cosedimenting with the 50S or mately 7% of the ribosomes carried a nascent chain+ 30S subunits were observed, indicating that the cross- For crosslinking experiments, translation mixtures linking reaction does indeed involve the thio-U residues+ were prepared with the mRNA analog, and, after incu- The identities of the crosslinked proteins were as- bation, the 70S ribosomes were separated from the sayed with antibodies to the individual ribosomal pro- other translational components by centrifugation through teins bound via a second antibody to agarose (Gulle a sucrose cushion (High et al+, 1991)+ Next, the 32P- et al+, 1988)+ This test showed a clear and reproducible labeled 4+5S RNA–Ffh complex was added, and, after reaction with anti-S1 in the case of the crosslinking to a short incubation, crosslinking was induced from the the 30S subunit, and an example of the antibody assay thio-U residues by irradiation at 350 nm+ The cross- is illustrated in Figure 2+ (The weak positive reaction linked products were separated by our usual procedure with anti-S21 that is visible in Fig+ 2 was also repro- (Stade et al+, 1989) in a series of sucrose gradient ducible+) In contrast, the corresponding assays of the centrifugations+ The first gradient, at high magnesium crosslinked protein fraction from the 50S subunit gave concentration, separates the 70S ribosomes from un- no positive signal+ bound 4+5S RNA and Ffh; the second, at low magne- , sium concentration dissociates the 70S ribosomes into Localization of crosslink sites ; , , 30S and 50S subunits and the third gradient in SDS on 23S and 16S rRNA separates ribosomal proteins from 16S or 23S rRNA+ In control experiments, the mRNA was omitted from The crosslink sites on the rRNA were localized by a the translation mixture, or 4+5S RNA alone was added series of digestions with ribonuclease H in the pres- Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 615

between the downstream oligodeoxynucleotide and the 39 end of the molecule in each case+ However, in lane 4 the released fragment “flips” to the position between the two oligodeoxynucleotides (a 120-nt fragment), and is further localized in lanes 5 and 6 to the region between oligodeoxynucleotides centered on positions 2822 and 2841 of the 23S rRNA+ Lanes 7–10 show this more dramatically; here, in all four lanes, the upstream oligodeoxynucleotide was kept constant (centered on position 2628), whereas the downstream oligodeoxynucleotide was moved a few bases at a time in the 39 direction, showing clearly that the “flip” occurs between lanes 8 and 9 (with the downstream oligodeoxynucleotide centered on positions 2828 and 2837, respectively)+ It is noteworthy that, in all cases (lanes 1 to 10), the ribonuclease H digestion was complete, with no material remaining at the start of the gel, and in the control samples without Ffh (where, as already noted above, the amount of crosslinked material was reduced by an order of magnitude), no trace could be observed on the gels of the crosslinked bands seen in Figure 3, even after long exposure times; the crosslink is thus entirely dependent on the presence of Ffh+ With the Ffh-independent crosslink to 16S rRNA (on the right in Fig+ 3), the crosslinked fragment released becomes progressively smaller in lanes 1 to 6, indicating that, in each case, the crosslink lies between the downstream oligodeoxynucleotide and the 39 terminus of the 16S molecule+ However, in lanes 5 and 6, the digestion is incomplete, suggesting that the exact crosslink site is either somewhat heterogeneous, or else lies FIGURE 2. Immunological analysis of crosslinked 30S proteins+ The very close to the midpoint of the oligodeoxynucleotide histogram shows the radioactivity remaining bound to antibodies to , + each of the small subunit ribosomal proteins (S1 to S21), immobi- in lane 6 centered on position 1525 The same pattern lized on agarose+ of ribonuclease H digestion was found, regardless of whether vacant 70S ribosomes, 30S subunits, or the full translation system (with or without Ffh) were used as substrate for the crosslinking+ ence of pairs of oligodeoxynucleotides complementary Our usual procedure in crosslink site determinations to selected sequences in the 16S or 23S molecules is to follow up the ribonuclease H digestions with a (Rinke-Appel et al+ 1991; Dontsova et al+, 1994)+ In the primer extension analysis to pinpoint the crosslink sites subsequent gel electrophoresis, the radioactive 4+5S (e+g+, Rinke-Appel et al+, 1991; Dontsova et al+, 1992)+ RNA will appear crosslinked either to the rRNA frag- However, in the case of the 16S crosslink, the site is ment lying between the two oligodeoxyncleotide posi- obviously too close to the 39 end of the molecule to tions (if these have straddled the crosslink site) or to permit such an analysis+ With the 23S crosslink, our the rRNA fragment between one of the oligodeoxy- attempts to apply the primer extension method did not nucleotides and the 59 or 39 end of the molecule+ Fig- give an unequivocal result, perhaps because of the ure 3 shows examples of the ribonuclease H digestions, very small amounts of crosslinked material that we both for the Ffh- and mRNA-dependent crosslink to the obtained+ 23S rRNA, and for the Ffh-independent crosslink to the 16S rRNA+ The sketches at the bottom of the figure Localization of the crosslinked illustrate the positions and sizes (rounded to the near- thio-U residues on 4.5S RNA est 5 nt) of the rRNA fragments that could be released in the various digestions+ Identification of the thio-U residues in the 4+5S RNA With the Ffh-dependent crosslink to 23S rRNA, it can that were involved in the crosslinks to 23S or 16S rRNA be seen from Figure 3 that, in lanes 1 to 3, the released was achieved by a combination of ribonuclease H di- fragment attached to the 4+5S RNA becomes progres- gestion and classical oligonucleotide fingerprinting sively smaller, thus corresponding to the rRNA region (cf+ Dontsova et al+, 1992, 1994); examples are illus- Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

616 J. Rinke-Appel et al.

FIGURE 3. See caption on facing page. Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 617 trated in Figure 4+ (It should be noted that the primer that band b has moved downwards (to band c), be- extension method cannot be used for this purpose, because the 39 terminal sequence has now been re- cause the thio-U residues themselves cause pauses in moved+ It follows that the crosslink site on the 4+5S the patterns of reverse transcription)+ Figure 4A shows RNA must be between the two thio-U residues at po- the 4+5S RNA sequence, and indicates the oligonucle- sitions 75 and 98 (Fig+ 4A)+ otides that would be released by ribonuclease T1 di- This site localization was confirmed and extended by gestion as radioactive entities, visible on the fingerprint, oligonucleotide fingerprinting of the 4+5S–23S rRNA from a transcript labeled with 32P-UTP+ These oligonu- crosslinked complex (Fig+ 4C)+ It can be seen from a cleotides are numbered according to their U content comparison of the control fingerprint (noncrosslinked and chain length (Brimacombe et al+, 1990); for ex- 4+5S RNA) with that of the crosslinked complex that the ample “38” is an octanucleotide containing three UG spot (12) is clearly missing from the latter+ The UG U-residues; the two parameters determine the position sequence (as a released ribonuclease T1 oligonucle- of the oligonucleotide on the fingerprint+ Figure 4A also otide) occurs at two adjacent positions in the 4+5S RNA, shows the positions of the complementary oligodeoxy- namely 82–83 and 84–85 (Fig+ 4A)+ In the radioactive nucleotides (numbered I–IV) that were used for ribo- 4+5S RNA transcript before crosslinking, each U posi- nuclease H digestion; the secondary structure of the tion is a mixture of U (from the 32P-labeled UTP) and 4+5S RNA is exceptionally stable, and accordingly rel- thio-U (from the nonradioactive thio-UTP), whereas at atively long oligodeoxynucleotides had to be applied in a crosslink site, the U position concerned has obvi- these digestions to allow DNA–RNA hybridization to ously been selected so that only thio-U is present+ Thus, occur+ if the crosslinked nucleotide in this case was U-82, Figure 4B shows some typical ribonuclease H diges- then two effects would be expected: (1) the radioactive tion patterns+ Lane 1 is a control of undigested, non- UG spot from positions 82–83 should disappear from crosslinked 4+5S RNA, whereas lane 2 shows the the fingerprint (by virtue of the crosslink itself), but also digestion products obtained from the latter in the pres- (2) the radioactive CG spot at positions 80–81 should ence of oligodeoxynucleotide III (Fig+ 4A)+ The two prin- concomitantly disappear, because U-82 must be exclu- cipal bands of 70 and 40 nt in lane 2 represent the 59 sively thio-U in the crosslinked complex and hence G-81 and 39 regions of the molecule, respectively, and this would not be radioactive+ Because the CG spot was was confirmed by oligonucleotide fingerprinting (not not absent from the fingerprint (although it was some- shown)+ Lanes 3 to 5 show the corresponding patterns times reduced in intensity, as in the example given in obtained from the Ffh-dependent crosslink to 23S rRNA+ Fig+ 4C), it follows by the same argument that the cross- In lane 3, the crosslinked product was excised from the linked residue is thio-U-84, whereby G-83 has lost its 23S rRNA with ribonuclease H as an 80-nt fragment radioactivity+ attached to the complete 4+5S molecule (band a; An example of the ribonuclease H digestion of the cf+ Fig+ 3, lane 7)+ In lane 4, oligodeoxynucleotide III Ffh-independent crosslink to 16S RNA is illustrated in (Fig+ 4A) was also present in the digestion, and it can Figure 4B, lanes 6 and 7+ Here, lane 6 is a control be seen that, whereas the 70-nt fragment (cf+ lane 2) is digest of 4+5S RNA in the presence of oligodeoxynucle- still present, the 40-nt fragment is absent+ This indi- otides I and III (Fig+ 4A), which leads to the appearance cates that the crosslink must lie within the 39 fragment of three products, a short fragment (of about 10 nt) of the 4+5S RNA, and the digested crosslinked species from the 59 terminus of the 4+5S molecule, a 60-nt frag- runs as a new band (b in lane 4) at approximately the ment from the central region, and the same 40-nt frag- position of the undigested 4+5S RNA+ When oligodeoxy- ment from the 39 region as in lane 2 of Figure 4B+ (A nucleotides III and IV were added simultaneously to small amount of undigested 4+5S RNA is also present+) the digestion (lane 5), the situation is similar, except In the crosslinked sample (lane 7), which was excised

FIGURE 3. Autoradiograms of ribonuclease H digests on 6% polyacrylamide gels of 32P-labeled 4+5S RNA crosslinked to 23S or 16S rRNA+ In each gel lane, the ribonuclease H digest was performed in the presence of one or two oligodeoxynucleotides (10–17 nt long) complementary to selected regions of the rRNA+ The central complementary positions of each of these oligodeoxynucleotides within the 23S or 16S rRNA sequence, together with the approximate lengths of the rRNA fragments that could be released in each case, are indicated in the diagrams at the bottom of the figure+ The fragment lengths are rounded to the nearest 5 nt; solid lines indicate the fragment that was actually observed in the corresponding gel lane, whereas dotted lines with fragment lengths in parentheses indicate the alternative possibility that was not observed+ 23S rRNA: Lane 1: oligodeoxynucleotides centered on 23S rRNA positions 2281 and 2442; lane 2: positions 2442 and 2628; lane 3: 2572 and 2822; lane 4: 2739 and 2857; lane 5: 2739 and 2841; lane 6: 2822 and 2841; lanes 7 to 10: 2628 and 2822, 2828, 2837, and 2841, respectively+ (Position 2281 from lane 1 is not included in the lower diagram+) 16S rRNA: Lane 1: oligodeoxynucleotides centered on 16S rRNA positions 846 and 1055; lane 2: positions 972 and 1139; lane 3: 1139 and 1399; lane 4: 1399 and 1499; lane 5: position 1513; lane 6: position 1525+ (Positions 846 and 972, from lanes 1 and 2, respectively, are not included in the lower diagram+) Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

618 J. Rinke-Appel et al.

FIGURE 4. Localization of thio-U residue(s) within the 4+5S RNA involved in crosslinking to 23S or 16S rRNA+ A: The primary sequence of the 4+5S RNA, divided into ribonuclease T1 oligonucleotide digestion products+ The asterisks denote residues carrying a radioactive 39 phosphate group, and the corresponding radioactive oligonucleotides are numbered according to their U content and chain length (see text)+ The bars marked with roman numerals (I to IV) indicate the positions of the complementary oligodeoxynucleotides used for ribonuclease H digestion+ B: Ribonuclease H digestions (cf+ Fig+ 3)+ Lane 1: undigested, noncrosslinked 4+5S RNA; lane 2: the same, digested with oligodeoxynucleotide III (see A. The principal digestion products, 70 and 40 nt, are marked); lane 3: 4+5S RNA crosslinked to an 80-nt fragment of 23S rRNA (a, see text); lane 4: the same, digested with oligodeoxynucleotide III; lane 5: the same, digested with oligodeoxynucleotides III and IV (b and c represent the residual fragments of 4+5S RNA crosslinked to the 80-nt 23S fragment); lane 6: noncrosslinked 4+5S RNA control, digested with oligodeoxynucleotides I and III (see A. The principal digestion products, 60, 40, and 10 nt, are marked); lane 7: 4+5S RNA crosslinked to an 85-nt fragment of 16S rRNA digested with oligodeoxynucleotides I and III (d represents an undigested remnant of the crosslinked 4+5S/85-nt material, and e the corresponding product after digestion)+ C: Ribonuclease T1 fingerprints of free 32P-labeled 4+5S RNA (Control) and of 4+5S RNA crosslinked to 23S rRNA (23S X-Link)+ Oligonucleotide spots are numbered as in A. Direction of the first chromatographic dimension is from right to left and that of the second from bottom to top+ Small arrows mark the sample application point, and the large arrow in the crosslinked sample shows the position of the missing 12 spot+ Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 619 as an 85-nt fragment from the 16S RNA (the residual 23S rRNA (Fig+ 5)+ These two helices are located low undigested band d), the 40-nt and 10-nt fragments are down on the L7/L12 side of the 50S subunit (Fig+ 6A; still present, but the 60-nt central fragment is clearly Ban et al+, 2000), and the 4+5S RNA can be rotated absent, the digested crosslinked complex now being about the U-84 crosslink site so as to bring the Ffh represented by band e+ This locates the crosslink site moiety into the direct proximity of the peptide tunnel to the region approximately between positions 10 and exit site at the base of the subunit (Fig+ 6B)+ In this 70 of the 4+5S sequence+ A similar digestion with oligo- orientation, the groove in the M domain of Ffh that has deoxynucleotide II in place of I (not shown) enabled been proposed to accommodate the signal peptide se- this region to be narrowed down to positions 20 and 70, quence (cf+ Keenan et al+, 1998; Batey et al+, 2000) lies and hence the crosslink must lie between the two thio-U in a plausible position at the tunnel exit, where it could residues at positions 29 and 50 (Fig+ 4A)+ A further interact with the signal sequence as soon as the latter localization by fingerprinting was not possible in this emerges from the 50S subunit+ The most direct evi- case, as a result of the degeneracy of the oligonucle- dence that the nascent peptide does indeed pass otide pattern (two 01 and two 38 products) in this re- through the tunnel was provided by crosslinking stud- gion of the 4+5S sequence+ A localization of the site or ies using a diazirine photoreagent attached to the sites on the 4+5S RNA involved in the Ffh-independent N-terminus of in situ synthesized peptides of different crosslink to protein S1 in the 30S subunit (Fig+ 2) was lengths (Stade et al+, 1995; Choi & Brimacombe, 1998); not undertaken+ when the peptide was ;30 amino acids in length, crosslinking was observed in those experiments to E. coli 23S rRNA nt 91+ The equivalent nucleotide in H. maris- DISCUSSION mortui 23S rRNA is also located at the tunnel exit (Ban et al+, 2000), just opposite the modeled position of the The two inter-RNA crosslinks identified in this study are Ffh moiety in Figure 6B+ The Ffh-dependent crosslink in many respects distinct+ One is entirely Ffh depen- from U-84 to the 23S rRNA is thus fully consistent with dent and the other totally Ffh independent, one is to the the SRP function of the 4+5S/Ffh complex in signal 23S rRNA and the other to the 16S rRNA, and the two peptide sequence recognition+ crosslinks are to opposite strands of the 4+5S RNA In the absence of any data beyond the U-84 crosslink molecule+ The crosslink sites are shown in the appro- (in the direction of the 39 and 59 termini of the 4+5S priate regions of the secondary structures of the three RNA), the extrapolated region of the molecule (shown RNA molecules in Figure 5+ In the case of the 23S green in Fig+ 6) has been modeled as an approximately rRNA, the corresponding structure for Haloarcula maris- A-form helix, extending the crystal structure (Batey et al+, mortui is also illustrated, as this is the species for which 2000)+ It should be noted, however, that there are sev- the atomic structure of the 50S subunit has been de- eral looped-out sequences in the secondary structure termined (Ban et al+, 2000); the secondary structures (Fig+ 5), which could allow this extrapolated region to of H. marismortui and E. coli are somewhat different in bend so as to wrap more closely around the 50S sub- the region (helices h99–h101) encompassing the cross- unit (Fig+ 6A,B)+ link site+ The same extrapolated structure for the 4+5S RNA The X-ray crystallographic structure reported for the was used to model the Ffh-independent crosslink site 4+5S RNA/Ffh complex (Batey et al+, 2000) comprises to helix 45 of the 16S rRNA (Fig+ 6C)+ In this case, the the M domain of the protein and nt 33–74 of the 4+5S orientation of the molecule relative to the crosslink site molecule, stabilized by an extra 3 bp+ Whereas the is arbitrary (but see below), there being no second Ffh-independent crosslink (nt 29–50) lies largely within “anchor point” corresponding to the peptide tunnel exit this RNA region, the Ffh-dependent crosslink (nt 84) in Figure 6B+ Because the crosslink to 16S rRNA is lies a few nucleotides outside it (Fig+ 5)+ Accordingly, to formed even with isolated 30S subunits, we have no investigate possible orientations of the 4+5S RNA on “functional control” for it and we therefore cannot ex- the ribosome in the light of our crosslinking data, the clude that it represents an experimental artefact+ (This crystal structure (Batey et al+, 2000) was extrapolated is in direct contrast to the situation regarding the cross- to include the rest of the 4+5S molecule using the pro- link to 23S rRNA, where omission of Ffh or mRNA gram ERNA-3D (Mueller & Brimacombe, 1997), and causes total abolition of the crosslink+) There is, for was docked onto the atomic structure of the 50S (Ban example, a strong Shine–Dalgarno-type sequence at et al+, 2000) or 30S (Wimberly et al+, 2000) ribosomal the loop end of the 4+5S molecule (Fig+ 5) that could subunit+ The results of these modeling studies are il- give rise to a possibly fortuitous interaction with the 39 lustrated in Figure 6+ end of the 16S rRNA+ On the other hand, it is interest- Figure 6A,B shows the modeled structure for the Ffh/ ing that three of the mutations reported by O’Connor 4+5S RNA complex docked onto the 50S subunit via et al+ (1995) that suppress the requirement for 4+5S the crosslink from U-84 of the 4+5S molecule to the RNA lie in helices 23a and 34 of the 16S rRNA, all three junction of helices h100 and h101 in domain VI of the of these sites being located quite close to our 4+5S Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

620 J. Rinke-Appel et al.

FIGURE 5. Locations of crosslink sites in the secondary structure of 4+5S RNA and in appropriate regions of the secondary structures of 16S and 23S rRNA+ For the 16S rRNA, the 39-terminal helix (h45) is shown, with the approximate region of the crosslink site localized by ribonuclease H digestion (Fig+ 3) in yellow+ Similarly, for the 23S rRNA, helices h99–h101 in domain VI of the molecule are shown, with the approximate region of the crosslink site in magenta; the equivalent region of the H. marismortui 23S rRNA is shown for comparison (see text)+ The corresponding crosslinked regions localized on the 4+5S RNA (Fig+ 4) are indicated in yellow and magenta, respectively+ The part of the 4+5S molecule encircled in green is that contained in the crystallized 4+5S/Ffh complex (Batey et al+, 2000)+

crosslink site at the base of helix 45 in the three- link+ Protein S1, which we identified as the crosslinked dimensional structure of the 30S subunit (Schluenzen protein target from the 4+5S RNA (Fig+ 2), also belongs et al+, 2000; Wimberly et al+, 2000)+ This proximity argues to the same topographical neighborhood (Sengupta in favor of some functional significance for the cross- et al+, 2001); S1 was not included in the X-ray structure Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 621

FIGURE 6. Computer modeling of interactions between 4+5S RNA and the ribosome, as evidenced by the crosslinking data+ A,B: The 4+5S RNA/Ffh complex docked onto the atomic structure of the H. marismortui 50S subunit+ The 23S and 5S rRNAs of the latter are rendered as gray backbone tubes, with the proteins as thinner blue tubes+ The 4+5S RNA is shown by the green and magenta backbone tube, the magenta representing the part of the 4+5S molecule in the crystallized Ffh/4+5S fragment complex, and the green the extrapolated region+ Protein Ffh in the crystallized complex is shown as the orange backbone tube+ The crosslink site at position U-84 of the 4+5S RNA (Fig+ 5) is highlighted as a yellow CPK nucleotide, and the corresponding crosslinked region of the 23S rRNA as a red backbone tube+ The red CPK nucleotide shows the equivalent of position 91 in the 23S rRNA of E. coli+ In A, the view is from the L7/L12 side of the 50S subunit, in B from the “underside” of the subunit with the red arrow indicating the peptide tunnel exit site+ C: The 4+5S RNA docked onto the atomic structure of the T. thermophilus 30S subunit+ Proteins and rRNA in the latter are rendered as in A and B, and the magenta and green regions of the 4+5S molecule are also marked as in A and B+ The yellow backbone tube represents the crosslinked nt 29–50 of the 4+5S RNA (Fig+ 5) with the corresponding crosslinked region in 16S rRNA as a red backbone tube+ D: The two positions of 4+5S RNA (with and without Ffh) from A, B, and C shown relative to a semitransparent cryo-EM contour of the 50S subunit, with the location of EF-G in the presence of fusidic acid indicated by the orange backbone tube+ The red backbone tube (upper right) gives the position of E. coli 23S rRNA nt 1068–1077; this region of the 23S rRNA was disordered in the crystal structure of the 50S subunit, and its position is accordingly taken from the modeled structure of Mueller et al+ (2000)+ See text for further references and explanation+ Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

622 J. Rinke-Appel et al. of Wimberly et al+ (2000), and is accordingly not visible most favorable orientation (Fig+ 6D), the loop end of the in Figure 6C+ 4+5S RNA is still far from the homologous decanucleo- As noted in the Introduction, 4+5S RNA is well known tide sequence in the 23S rRNA in h43 mentioned above to interact with EF-G, and the factor binding site over- (the red element on the upper right in Fig+ 6D)+ laps the binding site for Ffh at the loop end of the 4+5S In conclusion, we have identified two highly specific molecule (Jovine et al+, 2000; Nakamura et al+, 2001)+ crosslinks between the 4+5S RNA and the ribosomal Furthermore, the loop-end region contains a 10-nt se- RNA+ The Ffh-dependent crosslink to the 23S rRNA is quence that is identical to that of nt 1068–1077 in the compatible with the SRP function of the 4+5S RNA, and E. coli 23S rRNA (Brown, 1989); the latter sequence is gives a first approximation as to how the 4+5S molecule located in the loop end of helix h43, which contains is oriented on the ribosome in this mode+ On the other footprint sites for EF-G (Moazed et al+, 1988)+ EF-G in hand, the Ffh-independent crosslink to the 16S rRNA various different functional conditions has been located represents—if any—a hitherto unknown function of the on the E. coli 70S ribosome by cryoelectron micros- 4+5S RNA+ Neither of the two crosslinks appears to be copy (cryo-EM; Agrawal et al+, 1998; Stark et al+, 2000)+ compatible with an interaction of 4+5S RNA with EF-G All of the locations found lie in the same general area on the ribosome, and further crosslinking experiments of the intersubunit space, the most well-defined state are in progress to see whether we can observe such being that observed in the presence of fusidic acid an interaction in the presence of antibiotics, and if so, (Agrawal et al+, 1998; Stark et al+, 2000)+ In this state, where the 4+5S molecule is located relative to the ribo- the orientation of the factor is almost identical to that some in this situation+ seen in a similar cryo-EM reconstruction made with the EF-Tu/tRNA complex (Stark et al+, 1997), and, accordingly, we modeled the EF-G molecule onto the latter MATERIALS AND METHODS reconstruction at 13 Å resolution (Brimacombe et al+, 2000), superimposing it as closely as possible onto the Preparation of 4.5S RNA containing thio-U / + electron density of the EF-Tu tRNA This location was DNA coding for 4+5S RNA was amplified from total E. coli then transferred to the 9 Å reconstruction of the 50S DNA by PCR using the primers ATGGATCCTAATACGACT subunit, as in Mueller et al+ (2000)+ The result is shown CACTATAGGG and TATGCATGGTGGGGGCCCTGCCAG+ together with the two positions for the 4+5S RNA dis- The PCR product was cleaved with BamHI and NsiI restric- cussed above in Figure 6D+ tion endonucleases and ligated to the vector pSL1180 (Phar- The view in Figure 6D is from the interface side of macia), cleaved by the same endonucleases+ In the resulting the 50S subunit, and the first and most obvious con- construct, the 4+5S RNA is under the control of the T7 pro- clusion to be drawn is that the position of the 4+5S moter, so that a 4+5S molecule with a mature 39 end could be RNA/Ffh complex in its SRP functional state (as in obtained by in vitro run-off transcription of the plasmid after + Fig+ 6A,B) is nowhere near the location of EF-G+ This cleavage with Nsi restriction endonuclease The PCR products were verified by automated sequencing+ observation is of course inherent in the fact that, Aliquots of the purified 4+5S PCR-DNA (7 pmol) were tran- whereas the binding sites for Ffh and EF-G overlap at scribed directly with T7 polymerase in the presence of ATP, + , the loop end of the 4 5S molecule the binding site for CTP, and GTP, together with a mixture of a-32P-UTP (Amer- EF-G on the ribosome is on the interface side of the sham-Buechler), unlabeled UTP, and thio-UTP to give a 50S subunit and the peptide tunnel exit is on the sol- UTP:thio-UTP molar ratio of 1:5 (cf+ Dontsova et al+, 1994)+ vent side+ The second binding site for the 4+5S RNA, as The yield of 4+5S RNA obtained was typically ;300 pmol, modeled onto the 30S subunit, is shown in the same with a specific activity of 1+3 ϫ 106 cpm/pmol+ The ratio of position as in Figure 6C (but now relative to the 50S Up:thio-Up in the transcripts was determined by digestion subunit) on the upper left side of Figure 6D+ As already with ribonuclease T2 followed by two-dimensional thin-layer +, + noted, the modeled orientation of the 4+5S molecule chromatography (Stade et al 1989) relative to the crosslink site on the 16S rRNA is arbi- , , , trary but in such an orientation there could in principle Isolation of protein Ffh be some interaction between the loop end of the 4+5S and domain 4 of the EF-G (the domain extending out to A gene coding for the full-length Ffh protein was amplified the left side of the latter)+ This does not seem very from total E. coli DNA by PCR using the primers TTTGAT plausible, however, because the 4+5S RNA bound (or AATTTAACCGATCG and ATCTCGAGGCGACCAGGGAA + crosslinked) to the base of h45 of the 16S rRNA GCC The PCR product was cleaved with XhoI endonuclease and ligated to the vector pET33b (Novagen) cleaved by NcoI, would—in any orientation—interfere with P site tRNA + , treated with T4 DNA polymerase and cleaved with XhoI In binding and the P site is always occupied during the the resulting construct, the sequence coding for the full- , elongation phase of protein synthesis both in the pre- length Ffh protein together with a C-terminal tag of six histi- and posttranslocational states+ Thus, the crosslink to dine residues is located downstream of the T7 promoter+ 16S rRNA is unlikely to be relevant to the interaction Expression of the protein was induced in BL21 DE3 cells by between 4+5S RNA and EF-G+ Moreover, even in this 1 mM IPTG+ The overexpressed protein was isolated from Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 623 the induced cells using the Ni-NTA spin kit (Qiagen) accord- 5 min at 37 8C the elongation and initiation reactions were ing to the manufacturer’s instructions+ The final product was mixed to give a total volume of 150 mL, and incubated for a dialyzed against 1,000 vol of 20 mM Tris-HCl, pH 7+6, 10 mM further 15 min at 37 8C+ MgCl2, 50 mM NaCl, 1 mM dithiothreitol, 0+1% Tween 20, and At the end of the translation incubation, the mixtures were 50% glycerol+ The product was analyzed for purity and pro- kept on ice for 5 min and then layered over a 0+5 M sucrose tein content by polyacrylamide gel electrophoresis, and was cushion (2 mL) containing 10 mM Tris-HCl, pH 7+8, 100 mM stored in aliquots at Ϫ20 8+ NH4Cl, 10 mM magnesium acetate, and 6 mM 2-mercaptoethanol (cf+ High et al+, 1991), followed by centrifugation at 36,000rpmfor4hat48C in a Beckman Ti 50 rotor+ The Ffh/4.5S RNA complex formation ribosomal pellets were resuspended in 100 mL of translation buffer (20 mM HEPES-KOH, pH 7+5), 5 mM phosphoenol 32 + Equimolar amounts of P-labeled thio-U-containing 4 5S RNA pyruvate, and 0+6 mg/mL pyruvate kinase)+ Aliquots of the and Ffh protein were incubated together in a buffer consisting purified nascent chain/ribosome complexes were subjected , + , , , of 20 mM Tris-HCl pH 7 5 10 mM MgCl2 200 mM NaCl to protein analysis on 15% polyacrylamide gels in SDS fol- , , + 10 mM dithiothreitol 30 mM 2-mercaptoethanol 0 01% Igepal lowed by autoradiography, to examine the products of pep- , / 8 , CA 630 (Sigma) and1mgmL BSA for 10 min at 37 C tide synthesis+ This showed that approximately 7% of the 8 + +, + followed by 20 min at 4 C (cf Batey et al 2000) Complex ribosomes had synthesized the nascent chain+ formation was monitored by electrophoresis on 5% nonde- naturing polyacrylamide gels containing 5 mM magnesium , , , + + acetate 20 mM NH4Cl and 75 mM Tris-HCl ph 7 8 (cf Lentzen Crosslinking of the Ffh/4.5S RNA complex et al+, 1994; Suzuma et al+, 1999)+ The gels were run with low to the nascent peptide/ribosome complex voltage at 4 8C+ One hundred microliters of purified nascent peptide/ribosome complex (containing ;120 pmol of 70S ribosomes) was ad- Preparation of mRNA justed to the buffer concentration of the preformed Ffh/4+5S RNA complex (see above; 20 mM Tris-HCl, pH 7+5, 10 mM b The -lactamase (bla) gene was excised from the vector MgCl , 200 mM NaCl, 10 mM dithiothreitol, 30 mM 2-mercapto- + + , + 2 pTYB-ZZ (A V Petrov unpubl data) by cleavage with SspI ethanol, 0+01% Igepal, and1mg/mL BSA) and then mixed , endonuclease then cloned back into pTYB-ZZ cleaved by with the Ffh/4+5S complex containing 30–40 pmol of 32P- + XbaI and treated with T4 DNA polymerase In the resulting labeled 4+5S RNA+ This amount of complex, although sub- , construct the full-length bla gene is under the control of the molar with respect to the total amount of 70S ribosomes (120 + T7 promoter DNA fragments of suitable length for run-off pmol), represents an excess over the fraction of ribosomes transcription were prepared by PCR amplification with appro- carrying a nascent chain (see previous paragraph)+ This ratio + , priate primers For this study the DNA construct used (blaR1) of components was chosen so as to increase the absolute contained the T7 promoter together with the first 69 amino levels of crosslinked products formed+ The mixtures were b + acids of the -lactamase Transcription of the PCR DNA incubated for 5 min at 25 8C then 5 min on ice, followed by fragment with T7 polymerase was carried out as previously irradiation at 350 nm for 15 min as described by Dontsova , described (Choi & Brimacombe 1998) and the transcribed et al+ (1994)+ mRNA was purified by gel electrophoresis (Stade et al+, 1989; Dontsova et al+, 1992)+ Analysis of crosslinked products

In vitro translation and purification of The crosslinked ribosomal complexes were subjected to three + nascent chain/ribosome complexes rounds of sucrose gradient centrifugation First the crosslinked 70S ribosomes were separated from unbound 4+5S Initiator tRNAfMet was charged with 35S-methionine using RNA in a gradient at 10 mM magnesium, and then dissoci- tRNA-free S100 enzymes (Rheinberger et al+, 1988) and sub- ated into 50S and 30S subunits in a gradient at 0+3mM sequently formylated+ Initiation reaction mixtures (75 mL) were magnesium (Stade et al+, 1989)+ In a final gradient in the prepared containing 120 pmol E. coli 70S tight couple ribo- presence of SDS, the crosslinked 4+5S RNA/subunit com- somes (Rheinberger et al+, 1988), 188 pmol blaR1 mRNA plexes were further separated into rRNA and protein frac- (see above), 96 pmol formylated Met-tRNAfMet, and 75 pmol tions+ Ribosomal proteins crosslinked to 4+5S RNA were each of the initiation factors IF-1, IF-2, and IF-3 in 20 mM identified by immunological analysis using antibodies immo- HEPES-KOH, pH 7+5, 60 mM NH4Cl, 7+5 mM magnesium bilized on agarose as before (Gulle et al+, 1988; Rinke-Appel acetate, 1 mM dithiothreitol, and 1 mM GTP, pH 7+5 (Stade et al+, 1991)+ Crosslink sites on 23S or 16S rRNA were local- et al+, 1995)+ The mixtures were incubated for 10 min at 37 8C+ ized by digestion with ribonuclease H in the presence of pairs Similarly, elongation reaction mixtures (75 mL) were pre- of oligodeoxynucleotides (decamers or 17-mers) according pared containing bulk tRNA (Boehringer, 4+5 nmol), amino to the procedure of Rinke-Appel et al+ (1991) and analyzed acid mixture minus methionine (Promega, 45 nmol) and S100 by gel electrophoresis (Dontsova et al+, 1992)+ Identification enzymes (Stade et al+, 1995), together with 1+15 nmol of of crosslinked thio-U residues within the 4+5S RNA was made 35S-methionine at low specific activity (65,000 cpm/pmol) in by a combination of ribonuclease H digestion (using longer 20 mM HEPES-KOH, pH 7+5, 60 mM NH4Cl, 13 mM MgCl2, oligodeoxynucleotides, 15- to 23-mers) and ribonuclease T1 1 mM dithiothreitol, 4mMATP,pH 7+5, 10 mM phosphoenol fingerprinting of the isolated complexes, exactly as described pyruvate, and 1+2 mg/mL pyruvate kinase+ After incubation for by Dontsova et al+ (1992)+ Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

624 J. Rinke-Appel et al.

ACKNOWLEDGMENTS proteins L2, L4, L24, and L27 by treatment with 2-iminothiolane+ Nucleic Acids Res 16:815–832+ The vector pTYB-ZZ was a kind gift from Dr+ A+V+ Petrov High S, Flint N, Dobberstein B+ 1991+ Requirements for the mem- (Moscow), initiation factors were generously provided by Dr+ brane insertion of signal-anchor type proteins+ J Cell Biol 113: + , , + + + 25–34+ C Gualerzi (Camerino Italy) and S100 enzymes by Dr K H , + + + + Jensen CG Pedersen S 1994 Concentrations of 4 5S RNA and Ffh Nierhaus (Berlin) We gratefully acknowledge financial sup- protein in E. coli: The stability of Ffh protein is dependent on the port from the Deutsche Forschungsgemeinschaft (Br 632/ concentration of 4+5S RNA+ J Bact 176:7148–7154+ 4-1), the Volkswagen Foundation (I/74 598), the Howard Jovine L, Hainzl T, Oubridge C, Scott WG, Li J, Sixma TK, Wonacott Hughes Medical Institute (55000303), and the Russian Foun- A, Skarzynski T, Nagai K+ 2000+ Crystal structure of the Ffh and + EF-G binding sites in the conserved domain IV of E. coli 4+5S dation for Basic Research (01-04-48565) Alexey Bogdanov RNA+ Structure 8:527–540+ is the recipient of a research prize from the Alexander von Keenan RJ, Freymann DM, Walter P, Stroud RM+ 1998+ Crystal struc- Humboldt Foundation+ ture of the signal sequence binding subunit of the signal recognition particle+ Cell 94:181–191+ Lentzen G, Dobberstein B, Wintermeyer W+ 1994+ Formation of SRP- Received January 18, 2002; returned for revision like particle induces a conformation change in E. coli 4+5S RNA+ February 5, 2002; revised manuscript received FEBS Lett 348:233–238+ February 19, 2002 Lentzen G, Moine H, Ehresmann C, Ehresmann B, Wintermeyer W+ 1996+ Structure of 4+5S RNA in the signal recognition particle of E. coli as studied by enzymatic and chemical probing+ RNA 2:244–253+ , , + + REFERENCES Moazed D Robertson JM Noller HF 1988 Interaction of elongation factors EF-G and EF-Tu with a conserved loop in 23S RNA+ : + Agrawal RK, Penczek P, Grassucci RA, Frank J+ 1998+ Visualization Nature 334 362–364 , + + of elongation factor G on the E. coli 70S ribosome: The mecha- Mueller F Brimacombe R 1997 A new model for the three-dimensional : + nism of translocation. Proc Natl Acad Sci USA 95:6134–6138+ folding of E. coli 16S ribosomal RNA I Fitting the RNA to a 3D + : Avdeeva ON, Myasnikov AG, Sergiev PV, Bogdanov AA, Brima- electron microscopic map at 20 Å resolution J Mol Biol 271 524– + combe R, Dontsova OA+ 2002+ Construction of the minimal SRP 544 interacting with the translating ribosome but not with the specific Mueller F, Sommer I, Baranov P, Matadeen R, Stoldt M, Wöhnert J, membrane receptor in E. coli+ FEBS Lett 514:70–73+ Görlach M, van Heel M, Brimacombe R+ 2000+ The 3D arrange- Ban N, Nissen P, Hansen J, Moore PB, Steitz TA+ 2000+ The com- ment of the 23S and 5S rRNA in the E. coli 50S ribosomal subunit plete atomic structure of the large ribosomal subunit at 2+4Å based on a cryo-electron microscopic reconstruction at 7+5Å resolution+ Science 289:905–920+ resolution+ J Mol Biol 298:35–39+ Batey RT, Rambo RP, Lucast L, Rha B, Doudna JA+ 2000+ Crystal Nakamura K, Miyamoto H, Suzuma S, Sakamoto T, Kawai G, Ya- structure of the ribonucleoprotein core of the signal recognition mane K+ 2001+ Minimal functional structure of E. coli 4+5S RNA particle+ Science 287:1232–1239+ required for binding to elongation factor G+ J Biol Chem 276:22844– Brimacombe R, Greuer B, Gulle H, Kosack M, Mitchell P, Osswald M, 22849+ Stade K, Stiege W+ 1990+ New techniques for the analysis of O’Connor M, Brunelli CA, Firpo MA, Gregory ST, Lieberman KR, intra-RNA and RNA–protein cross-linking data from ribosomes+ Lodmell JS, Moine H, Van Ryk DI, Dahlberg AE+ 1995+ Genetic In: Spedding G, ed+ Ribosomes and protein synthesis; a practical probes of ribosomal RNA function+ Biochem Cell Biol 73:859–868+ approach+ Oxford, England: IRL Press+ pp 131–159+ Poritz MA, Bernstein HD, Strub K, Zopf D, Wilhelm H, Walter P+ 1990+ Brimacombe R, Greuer B, Mueller F, Osswald M, Rinke-Appel J, An E. coli ribonucleoprotein containing 4+5S RNA resembles mam- Sommer I+ 2000+ Three-dimensional organization of the bacterial malian signal recognition particle+ Science 250:1111–1117+ ribosome and its subunits: The transition from low-resolution mod- Rheinberger HJ, Geigenmüller U, Wedde M, Nierhaus KH+ 1988+ els to high-resolution structures+ In: Garrett RA, Douthwaite SR, Parameters for the preparation of E. coli ribosomes and ribo- Liljas A, Matheson AT, Moore PB, Noller HF, eds+ The ribosome: somal subunits active in tRNA binding+ Methods Enzymol 164:658– Structure, function, antibiotics and cellular interactions+ Washing- 670+ ton, DC: ASM Press+ pp 151–163+ Ribes V, Römisch K, Giner A, Dobberstein B, Tollervey D+ 1990+ E. Brown S+ 1989+ Time of action of 4+5S RNA in E. coli translation+ coli 4+5S RNA is part of a ribonucleoprotein particle that has J Mol Biol 209:79–90+ properties related to signal recognition particle+ Cell 63:591–600+ Choi KM, Brimacombe R+ 1998+ The path of the growing peptide Rinke-Appel J, Jünke N, Stade K, Brimacombe R+ 1991+ The path of chain through the 23S rRNA in the 50S ribosomal subunit; a mRNA through the ribosome; site-directed cross-linking of mRNA comparative cross-linking study with three different peptide fam- analogues carrying a photo-reactive label at various points 39 to ilies+ Nucleic Acids Res 26:887–895+ the decoding site+ EMBO J 10:2195–2202+ Dokudovskaya S, Dontsova O, Shpanchenko O, Bogdanov A, Bri- Rinke-Appel J, Jünke N, Brimacombe R, Dokudovskaya S, Dontsova macombe R+ 1996+ Loop IV of 5S ribosomal RNA has contacts O, Bogdanov A+ 1993+ Site-directed cross-linking of mRNA ana- both to domain II and to domain V of the 23S RNA+ RNA 2:146– logues to 16S ribosomal RNA; a complete scan of cross-links 152+ from all positions between “ϩ1” and “ϩ16” on the mRNA, down- Dontsova O, Dokudovskaya S, Kopylov A, Bogdanov A, Rinke-Appel stream from the decoding site+ Nucleic Acids Res 21:2853–2859+ J, Jünke N, Brimacombe R+ 1992+ Three widely separated posi- Römisch K, Webb J, Lingelbach K, Gausepohl H, Dobberstein B+ tions in the 16S RNA lie in or close to the ribosomal decoding 1990+ The 54-kD protein of signal recognition particle contains a region; a site-directed cross-linking study with mRNA analogues+ methionine-rich RNA binding domain+ J Cell Biol 111:1793–1802+ EMBO J 11:3105–3116+ Schluenzen F, Tocilj A, Zarivach R, Harms J, Gluemann M, Janell D, Dontsova O, Tishkov V, Dokudovskaya S, Bogdanov A, Döring T, Bashan A, Bartels H, Agmon I, Franceschi F, Yonath A+ 2000+ Rinke-Appel J, Thamm S, Greuer B, Brimacombe R+ 1994+ Stem- Structure of functionally activated small ribosomal subunit at 3+3Å loop IV of 5S ribosomal RNA lies close to the peptidyl transferase resolution+ Cell 102:615–623+ center+ Proc Natl Acad Sci USA 91:4125–4129+ Schmitz U, Freymann DM, James TL, Keenan RJ, Vinayak R, Walter Gowda K, Chittenden K, Zwieb C+ 1997+ Binding site of the M-domain P+ 1996+ NMR studies of the most conserved RNA domain of the of human protein SRP54 determined by systematic site-directed mammalian signal recognition particle+ RNA 2:1213–1227+ mutagenesis of signal recognition particle RNA+ Nucleic Acids Sengupta J, Agrawal RK, Frank J+ 2001+ Visualization of protein S1 Res 25:388–394+ within the 30S ribosomal subunit and its interaction with messen- Gulle H, Hoppe E, Osswald M, Greuer B, Brimacombe R, Stöffler G+ ger RNA+ Proc Natl Acad Sci USA 98:11991–11996+ 1988+ RNA–protein cross-linking in E. coli 50S ribosomal sub- Sergiev P, Leonov A, Dokudovskaya S, Shpanchenko O, Dontsova units; determination of sites on 23S RNA that are cross-linked to O, Bogdanov A, Rinke-Appel J, Mueller F, Osswald M, von Kno- Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the E. coli ribosome 625

blauch K, Brimacombe R+ 2002+ Correlating the X-ray structures van Heel M+ 1997+ Visualization of elongation factor Tu on the for halo- and thermophilic ribosomal subunits with biochemical E. coli ribosome+ Nature 389:403–406+ data for the E. coli ribosome+ Cold Spring Harbor Symp Quant Stark H, Rodnina MV, Wieden HJ, van Heel M, Wintermeyer W+ 2000+ Biol 56+ In press+ Large-scale movement of elongation factor G and extensive con- Shibata T, Fujii Y, Nakamura Y, Nakamura K, Yamane K+ 1996+ Iden- formational change of the ribosome during translocation+ Cell tification of protein synthesis elongation factor G as a 4+5S RNA- 100:301–309+ binding protein in E. coli+ J Biol Chem 271:13162–13168+ Suzuma S, Hayashi K, Nakamura K, Yamane K+ 1999+ Analysis of E. Stade K, Jünke N, Brimacombe R+ 1995+ Mapping the path of the coli 4+5S RNA binding affinity to Ffh and EF-G+ FEMS Microbiol nascent peptide chain through the 23S RNA in the 50S ribosomal Lett 180:271–277+ subunit+ Nucleic Acids Res 23:2371–2380+ Wimberly BT, Brodersen DE, Clemons WM, Morgan-Warren RJ, Carter Stade K, Rinke-Appel J, Brimacombe R+ 1989+ Site-directed cross- AP, Vonrhein C, Hartsch T, Ramakrishnan V+ 2000+ Structure of linking of mRNA analogues to the E. coli ribosome; identification the 30S ribosomal subunit+ Nature 407:327–339+ of 30S ribosomal components that can be cross-linked to the Zopf D, Bernstein HD, Johnson AE, Walter P+ 1990+ The methionine- mRNA at various points 59 with respect to the decoding site+ rich domain of the 54 kd protein subunit of the signal recognition Nucleic Acids Res 17:9889–9908+ particle contains an RNA binding site and can be cross-linked to Stark H, Rodnina M, Rinke-Appel J, Brimacombe R, Wintermeyer W, a signal sequence+ EMBO J 9:4511–4517+ Downloaded from rnajournal.cshlp.org on September 23, 2021 - Published by Cold Spring Harbor Laboratory Press

Crosslinking of 4.5S RNA to the Escherichia coli ribosome in the presence or absence of the protein Ffh.

Jutta Rinke-Appel, Monika Osswald, Klaus von Knoblauch, et al.

RNA 2002 8: 612-625

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the Service top right corner of the article or click here.

To subscribe to RNA go to: http://rnajournal.cshlp.org/subscriptions