- Email: [email protected]
Probing the assembly of the 3′ major domain of 16 S ribosomal RNA
J. Mol. Biol.
(lY88)
200, 309-319
Probing the Assembly of the 3’ Major Domain of 16 S Ribosomal RNA Quaternary Interactions Involving Ribosomal Proteins S7, S9 and S19 Ted Powers’, L&Ming Changchien 2, Gary R. Craven2 and Harry F. Noller’ 1Thimann Laboratories University of California at Santa Cruz Santa Cruz, CA 95064, U.S.A. 2Laboratory of Molecular Biology University of Wisconsin Madison, WI 53706, U.S.A. (Received 26 May 1987) We have studied the effect of assembly of ribosomal proteins 57, S9 and S19 on the accessibility and conformation of nucleotides in 16 S ribosomal RNA. Complexes formed between 16 S rRNA and S7, S7+S9, S7+Sl9 or S7+S9+S19 were subjected to a combination of chemical and enzymatic probes, whose sites of attack in 16 S rRNA were identified by primer extension. The results of this study show that: (1) Protein S7 affects the reactivity of an extensive region in the lower half of the 3’ major domain. Inclusion of proteins S9 or S19 with S7 has generally little additional effect on S7specific protection of the RNA. Clusters of nucleotides that are protected by protein S7 are localized in the 935-945 region, the 950/1230 stem, the 1250/1285 internal loop, and the 1350/1370 stem. (2) Addition of protein S9 in the presence of S7 causes several additional effects principally in two structurally distal regions. We observe strong SS-dependent protection of positions 1278 to 1283, and of several positions in the 1125/1145 internal 190~. These findings suggest that interaction of protein S9 with 16 S rRNA results in a structure in which the 1125/1145 and 1280 regions are proximal to each other. (3) Most of the strong SlS-dependent effects are clustered in the 950-1050 and 121CL1230 regions, which are joined by base-pairing in the 16 S rRNA secondary structure. The highly conserved 960-975 stemploop, which has been implicated in tRNA binding, appears t)o be destabilized in the presence of S19. (4) Protein S7 causes enhanced reactivity at several sites that become protected upon addition of S9 or S19. This suggests that S7-induced conformational changes in 16 S rRNA play a role in the co-operativity of assembly of the 3’ major domain.
1. Introduction
the 3’ maior domain was excised from naked 16 S rRNA (Zimmermann et al., 1972) or as an RNP containing a subset of the ribosomal proteins from 30 S subunits (Brimacombe et al., 1971; Yuki & Brimacombe, 1975), by partial nuclease digestion. RNPs isolated by two other groups (Schendel et al., 1972; Roth & Nierhaus, 1973) are probably related to the one that was characterized by Brimacombe and co-workers (Morgan & Brimacombe, 1972; Yuki & Brimacombe, 1975); in all cases, proteins S7, S9 and S19 were found, usually in near-stoichiometric amounts, together with varying amounts of other proteins. Assembly experiments in vitro showed that protein S7 binds independently to 16 S rRNA (Schaup et al., 1970; Mizushima & Nomura, 1970;
Early evidence for the domain organization of ribosomal RNA came from partial nuclease digestion studies of 16 S rRNA and 30 S subunits. In retrospect, the results of several such experiments provide rather convincing evidence that the 3’ major domain of 16 S rRNA (as defined by Noller & Woese, 1981), in particular, forms a relatively autonomous ribonucleoprotein (RNPt) structure within the assembled 30 S subunit. A subfragment corresponding almost precisely to t Abbreviations used: RIVP, ribonucleoprotein; dimethylsulfate; u.v., ultraviolet.
DMS, 309
0022%2836/88/06030~11
$03.00/O
0
1988 .Amdemic
Press Limited
310
T. Powers et al
Garrett et al., 1971); in contrast, proteins S9 and S19 require prior binding of S7 for their assembly (Mizushima & Nomura, 1970; Green & Kurland, 1973). In fact, all of the proteins found in significant amounts in the excised subparticle depend on protein S7 for their assembly (Mizushima & Nomura, 1970). These same proteins are all localized in the “head” of the 30 S subunit, according to immunoelectron microscopy and neutron diffraction studies (Lake, 1980; Stiiffler et al., 1980; Moore, 1980). These observations suggest that the 3’ major domain assembles as a relatively autonomous structural unit. the results of crosslinking Apart) from experiments, which further localize S7 with respect to specific features of the 3’ major domain (Zwieb & Brimacombe, 1979; Wower &, Brimacombe, 1983; Ehresmann et al., 1980), litt,le has been learned concerning the sites of interaction of ribosomal proteins with this domain of 16 S rRNA or, for that matter, whether any proteins other than S7 make contact with this region of the RNA. Nor has any light been shed on the molecular basis of the observed co-operativity of assembly between S7, S9 and S19. In the experiments presented here, we probe the effects of assembly of these three proteins on 16 S rRNA, using a recently developed .approach (Stern et al., 1986). Purified proteins are bound, individually or in combination: to 16 S rRIVA. The resulting RNPs are examined with a set of chemical and enzymatic probes that monitor the accessibility and conformation of each position along the RNA chain. Sites of attack are localized precisely by primer extension (Moazed et al., 1986; Stern et al., 1987). We confirm the co-operativity of assembly of these proteins, and show that each protein influences the reactivity of a distinct set of nucleotides, all of which are found within the 3’ major domain. Both diminished and enhanced reactivities of specific nucleotides result from interaction of 16 S rRNA with the proteins. Clusters of protected bases may indicate sites of protein-RNA contact. Protein S7 causes enhancement of the reactivity of several nucleotides; we interpret this to be the result of local conformational changes in 16 S rRNA caused by interaction with the protein. Some of these sites of enhancement are subsequently protected by S9 or S19, suggesting that S7-induced conformational changes in 16 S rRNA play a part in the co-operativity of assembly of this ribosomal domain.
(a) Isolation
2. Materials
and Methods
of 16 S rRNA
and proteins
S7, S9 and S19
Naked 16 S rRNA was obtained by extraction with phenol of Escherichia coli 30 S ribosomal subunits as described by Stern et aE. (1988). Proteins 57, S9 and S19 were purified by high-pressure liquid chromatography and renatured as described in an accompanying paper (Stern et al., 1988).
The procedure used was essentially the same as that described by Stern et al. (1986). Pro&ins S7. SY and S19 were mixed. either singly or in combination. with 16 S rRNA; 10 or 20 pg of RNA and a 3 to 8-fold molar excess of each prot,ein were used. The mixtures. as well as control samples containing 16 R rRNA without protein. were incubated for 1 h at 42°C’ in 50 ~1 of buffer containing 80 mM-K-Hrpes (pH 7.X), 20 m,n-M&I,. 300 miv-K(yl, 6 mM-B-mercapt’oethanol. 0.01 Oc, Nikko1 detergent (Nikko Chemical Co., Ltd,
and w~z?ymaLic prd~iq
(i) CJhemical modijcatiorc For dimethyl sulfate (DAMS) modification, 2 htl of I)MS (Aldrich), diluted 1 : Id in 95”& (v/v) EtOH. ~WP added KE
DMS
Vl
9: jO-
100 IO-
-
Figure 1. Autoradiographs showing thr results of chemical and enzymatic probing as monitored by primer extension using the 1047 primer (Moazed et al.. 1986: Stern et al.. 1987). Complexes probed: 1. naked 16 S rRNA; 2, S7; 3, S7/S9: 4, S7+819; 5, S7+S9+S19. C. I’. A, G are dideoxy sequencing lanes and refer to the nucleotide sequence of 16 S rRNA. K denotes a 16 8 rRNA control lane treated identically to the other samples except for omission of probes. Probes used: dimethylsulfate (DMS) kethoxal (KE), V, nuclease (V,). Nucleotide positions in 16 S rRNA are indicated. Arrows indicate main protein effects.
Interaction
of 6’7, X9 and S19 with 16 S RNA
to each 50-~1 sample, followed by incubation at 4°C for I .5 h. The reaction was stopped by the addition of 12.5 ~1 (pH 75), 1.0 M-pof DMS stop (1.0 M-T&. HCl mercaptoethanol, 0.1 mM-EDTA) followed by incubation on ice for 10 min. Samples were then precipitated by the addition of 300 ~1 of 95% EtOH and 30 fig of carrier RNA. For kethoxal modification, 2 ~1 of kethoxal (37 mgiml in 207; EtOH; Upjohn) were added to each 50+1 sample. followed by incubation at 4°C for 45 min. Samples were then made 25 mM in potassium borate (pH 7.0) and precipitated as described *above. Precipitated RNP complexes were then centrifuged, extracted with phenol, and precipitated with ethanol as described (Stern et al., 1988). RNA was redissolved in water at a concentration of 0.4 mg/ml. (ii) I’, ,nuclease digestion For V, nuclease digestion. 0.2 units of cobra venom nuclease (V, nuclease; Pharmacia) were added to each 5C)-yl sample and gentlv mixed. Samples were digested for 15 min at 4°C. Reactions were stopped by extract’ion with phenol as described (Stern et al., 1988). After precipitation with ethanol, the RNA was redissolved in water at a concentration of 0.4 mg/ml.
((1) Primer
extension
Hybridization, extension and gel electrophoresis were performed as described by Stern et al., (1988). For each experiment. the entire 16 S rRNA chain (except residues 1508 to 1542) was probed using the same set of DNA primers described by Moazed et al. (1986), with one exception. A new oligonucleotide primer with the sequence TGCTCTCGCGAGGTCGC was used to prime at position 1257. Typically, 16 S rRNA chain was scanned several times for each modification experiment. The data presented in Table 1 are the consensus of visual
--CUAGKl
KE
DMS
311
estimation of band intensities tion experiments.
from 5 separate modifica-
3. Results (a) Experimental
strategy
Assembly studies in vitro have shown that protein S7 binds independently and specifically to 16 S rRNA; binding of S9 and S19 is dependent on prior binding of S7 (Mizushima & Nomura, 1970: Green & Kurland, 1973). Accordingly, our strategy was to form RNP complexes consisting of S7, S7+S9, S7+S19 and S7+S9+S19 together with 16 S rRNA, and to compare the susceptibility of each nucleotide in the RNA to kethozal, dimethyl sulfate and V, nuclease in the various complexes and in naked 16 S rRNA. Mixtures of the secondary binders and 16 S RNA, in the absence of 57, were also probed. Effects attributable to interactions between S9 or S19 and the RNA should be S7dependent, but should require additionally the presence of S9 or S19. It is important to note that effects of the latter class may also be due to S7RNA contacts that somehow depend on the presence of the other two proteins. Tdentification of specific nucleotides and estimation of their altered reactivities in the complexes was accomplished by primer extension using a series of synthetic DNA oligomers, as described previously (Moazed et al., 1986; Stern et al., 1988). In each case, the entire 16 S rRNA chain was scanned, except for the 3’terminal 35 nucleotides, which are not accessible to examination by primer extension. The data, in the form of autoradiographs, are presented in Figures 1 to 3, and summarized in Table 1 and Figure 4; KE --CUAGK123451234512345
Vl
23451234512345
DYS
Vl
lllO-
(b) 1150-
DMS CU AGK
KE
Vl
I 2 3 4 5 1 2 3 4 5 I 2 3 4 5
1230-
Figure 2. Autoradiographs showing the results of chemical and enzymatic probing as monitored by primer extension using (a) and (c) t,he 1257 primer and (b) the 1200 primer. Symbols, numbers and abbreviations are as described for Fig. 1.
+I
;
+t
V
ti
3
+ +o+l=oc++l=c +
=;=
+I
=
ii
0 +I
+ i-
+I v +
+I tl t-1
+ A
+ ;+++I + sF :! Fq
P -2 2 5
+ +I +v
+ E
fl fl +I
===
==c
+ 0 =
ii
+j z
0
“3; 15s I +
=0300+1
++
+
+I
+I+I+++VI+++I+++++I++o
=c
=5
+I ” c +I
+I +I vv
++
+ +I +
+ +I
0
+I +I +I c v v v
+I===
+++;
c t’ +-I +I
==
+
+ +
+ :+
+ +I t
t
;
tl +!
<+t
0
+ + <+ + 0 ++ 0 ++ 0 0 <+ k zk
* + + + +
0 III
I!c
0 0 0 0
& k 0 i + 0 + <++ + + 0 + + iY k +
0 0 0 0
+ + + z!T
+ * 0 ii + ++ + ++ + + 0 + +
+ t +
0 0 <* f * *
0
+ + <+ + 0
<+ ++ ++
++
+ t++
++ +
* * iz
0 0
Al319 G1331 Al332 Al333 G1334 G1338 A1339 Al346 A1349 u1351 Cl352 G1353 u1354 Al357 Cl359 Al360 G1361 Al362 Al363 G1365 Cl367 G1371 Al374 Al375 Al377 Cl378 U1381 Cl382 Cl383 Al394 Cl395 A1396 Cl397 A1398 Cl399 + k + 0 0 xi + k + + <+ <+
+ + + k + 0 0 i AZ
+ + 5 f * +
0 <+ + 0
0
+ <+ ++ 5 XL 5
0 0 0 0 ++ k 0 0
+ + <+ * f +
0 f 0 + 0 *
f + <+ k k +
<++
0 f 0 k
+ ++ +++ if
0 0
0 * + +
0 0 0
Data are the consensus of visual estimates of relative band intensities from 5 dillerent probing experiments. Chemical probing data indicate kethoxal modification of G, DMS modification of A, except where noted. Keactivities are summarized qualitatively by the symbols 0, f, +, + + and + + f, where 0 indicates no reactivity and + + + indicates maximal or hyper-reactivity. In the case of Vi nuclease, the nucleotide 5’ to the cut site is indicated. For clarity, data are indicated only where a result differs from the column immediately to the left; i.e. a blank indicates that the result is the same as that found in the column to its left. Reverse transcriptase pauses or stops at positions U950 and l’1126 when naked 16 S rNNA is treated with kethoxal and U1116, lJ1212, U1224 and U1381 when treated with DMS. The chemical basis of these apparently anomalous reactivitirs is not known.
I:989 11991 G993 A4994 A996 1’1107 U1008 81014 A1016 A1036 A1046 G1047 41110 All11 Cl112 Ull16 U1126 G1127 A1130 G1138 All45 Al 146 A1150 G1156 G1182 IJ1212 A1213 Cl214 Cl217 Cl218 A1219 U1224 Al225 Cl226 Cl228 A1229
314
T. Powers et al KE
-CUAGK123451234512345
DMS
VI
d
5 I -
3
1
Figure 3. Autoradiographs showing the results of chemical and enzymatic probing as monitored by primer extension using the 1490 primer. Symbols, numbers and abbreviations are as described for Fig. 1.
below, we present separately the results obtained for each of the RNP complexes. (b) S7-speci$c
effects
The reactivities of nucleotides from a large region of the 3’ major domain are affected by binding protein S7. This is seen as increased or decreased susceptibility to the chemical and enzymatic probes when complexes of S7 and 16 S rRNA are compared with naked 16 S rRNA (Figs 1 and 3, lanes 1 and 2; Table 1). The distribution of nucleotides affected by 57 is shown in Figure 4(a); they are found in the lower half of the 3’ major domain. As for all of the complexes studied here, we
failed to detect any effects outside of the 3’ major domain (nucleotides 920 to 1400). Contacts between S7 and 16 S rRNA are most likely t,o occur in regions where the stronger protections are concentrated. These regions include the 935-945 region, the 950/1230 stem and its flanking sequences on the lower side, the 1250/1285 internal loop, the 1300 and 1330 regions, and the 1350/1370 stem-loop structure. We also observe several instances where S7 causes increased reactivity of a nucleotide toward our probes. Such enhanced reactivity has been observed in previous studies, both for 30 S subunit assembly (Moazed et al., 1986) and in binding individual proteins to the 5’ and central domains (Stern et al., 1986, 1988; Svensson et al.. 1988). Major enhancements occur at positions 978 and 979 (toward DMS), at’ positions 1278, 1279 and 1338 (toward kethoxal) and in the 1310 helix (toward V, nuclease). Significantly, many of these posit,ions become protected as a result of binding S9 or Sl9. as described below. (c) S9-dependent
eflects
Protein S9 by itself has no detectable effect on the modification pattern of 16 S rRNA (data not shown), but when both S9 and S7 are present’, we observe several effects that are distinct from those produced by S7 alone (Figs 1 to 3, lanes 3 and 5). Strong SS-dependent protections are found in t,he compound stem comprising positions 1241 to 1296 (Fig. 3, lanes 3 and 5). These are concent,rated in the 1245/1290 helix (V, protection) and in the 1255 1285 helix and its flanking sequence around position 1280 (V,, kethoxal and DMS protections). Three of these protected positions (1278, 1279 and 1281) are the sites of S7-induced enhancements. These results give the impression of intermingling of proteins S7 and S9 in the 1250/1280 region. A second cluster of SS-dependent protections is detected in the 1125-1145 region, quite distal in the secondary structure from the site of the other S9 effects (Fig. 2). Most evident is protection of Ull26. G1127, All45 and A1156: protection of several other nucleotides in this region is also observed. as enhancements. These are some SS-dependent findings suggest, that protein S9 may mediate thr close approach of the 1130 and 1280 regions (set, Fig. 4(b)). (d) SlS-dependent
effects
Numerous strong Sly-dependent’ effects are observed throughout the 950-1050 and 1210-1230 regions (Figs 1 and 2; lanes 4 and 5). The main protections are concentrated around the 98511220 and 1010/1020 stems and their respective singlestranded flanking sequences (Fig. 4(c)). Interseveral enhancements of chemical estingly, reactivity and diminished attack by V, nuclease suggest disruption of the 960-9631972-975 stem. Other effects of S9 and S19 are presented below.
Interaction
of S7, S9 and S19 with 16 S RNA
(b)
(c)
Figure 4. A diagram of the 3’ major domain of 16 S 16 S rRNA-protein complexes: (a) S7 relative to naked and (d) S7+S9+819 relative to both S7+S9 and 87 toward the chemical probes are indicated by (0) or enhanced reactivities toward V, nuclease are indicated
(d)
rRNA (nucleotides 920 to 1400) summarizing results of probing 16 S rRNA; (b) 87 + S9 relative to S7; (c) S7 +S19 relative to S7; +S19. Nucleotides exhibiting reduced or enhanced reactivities (A), respectively. Phosphodiester bonds exhibiting reduced or by (open arrows) or (closed arrows), respectively.
316
T. Powers et al. (e) Complex effects
In addition to the strong effects described above, which appear to be highly specific for the presence of a given protein, some regions of the RNA show a less well-defined pattern of protein-dependent perturbations. These include the 1245-1290, 13101320, 135&1365 and 1394-1399 regions (Figs 3 and 4). Most complex among these is the 1245-1290 region. G1268 is not perturbed unless all three proteins are present, in which case it has diminished reactivity. Reactivity of G1279 is enhanced by 57, further enhanced by Sl9, but protected by S9. Several other nucleotides in the 126&1280 region effects that depend on the presence of all three proteins. Nucleotides in this region are also affected by 57, S7+S9 or 57 +Sl9. The 1260-1280 region is clearly the site of extensive co-operative protein-RNA interactions. The 13lCl320 and 135&1365 stem-loop regions are similar, in that they both seem to be affected by each of the three proteins to some extent. Moreover, protection from chemical probes tends to be incomplete in many cases. There are two examples where the same end result is achieved in the presence of either S7+S9 or S7+Sl9. The 1350 stem is fully protected from V, nuclease by either combination (Fig. 3). Similarly, a series of protections and enhancements at positions 1394 to 1399 is induced by either S7+S9 or S7+519. although the S7+Sl9 combination appears to give the stronger effect (Fig. 3).
4. Discussion The experiments presented here demonstrate that specific nucleotides in the 3’ major domain of 16 S rRNA show decreased (or increased) susceptibility to chemical and enzymatic probes when various combinations of the ribosomal proteins X7, S9 and Sl9 are bound (summarized in Fig. 4). Although we believe that many of these effects, and clusters of strongly protected nucleotides in particular, are indicative of specific RNA-protein contacts, any rigorous interpretation demands caution in two respects. First, protection of a nucleotide need not be the result of direct contact with protein, but could as well be attributed to conformational changes caused by interaction of the protein with a more remote region of the RNA. A second consideration arises from the dependence of S9 and Sl9 on X7 for their assembly. Here, we cannot exclude the possibility that protections dependent on the presence of S9 or Sl9 are the result of new contacts between 57 and 16 S rRNA, rather than to direct interaction between S9 or Sl9 with the RNA. Protein S7 protects ma’ny nucleotides in the lower half of the 3’ major domain (Fig. 4(a)), and also causes several enhancements there. Major clusters of protected nucleotides are found in four main regions: the 935-945 region, the 950/1230 stem, the 1250-1285 internal loop, and the 1250/1370 stemloop (Fig. 4(a)). Assignment of S7 contacts to these
three regions is in close agreement witch the results of chemical crosslinking studies reported by Brimacombe and his colleagues. Tn one study. protein S7 was crosslinked to Ul240 by ultraviolet (u.v.) irradiation of 30 S subunits (Zwieb 8 Brimacombe, 1979). Tn another st,udy, using U.Y. irradiation of 2-iminothiolane-derivated subunits, S7 was crosslinked to the same region (within residues 1234 to 1241) and to a second site, within the dinucleotide AC! 1377-1378 (Wower & Brimacombe, 1983). The former crosslink lies precisely between the first two clusters of Hiprotected residues. and is proximal to both of them; the latter crosslink is closely adjacent to the third cluster in the 1350/1370 stem. Ehresmann et al. (1980) have reported a crosslink between protein 87 and C1265, in the same general region as the observed protections. However, t,he placement of the latter crosslink has been questioned by Brimacombe and co-workers (Brimacombc et trl., 1983), who reinterpret the relevant RNA sequentring data to infer that the crosslink is in fact identical with that reported by Zwieb & Brimacombe (1979), i.e. to position 1240. Finally, all of the nucleotides affected by protein S7 art’ contained well within the limits of its binding region, as defined by RNA subfragments obtained from either S7-16 S rRNA complexes (Zimmermann et (~1.. 1972) or from subparticles obtained by nuclease digestion of 30 S subunits (Yuki $ Krimacornbr, 1975). Proteins S9 and 819 have no effect on 16 S rRNA, either singly or in combination with each ot’her. When present in combination with Si, however, both proteins cause protection or enhanct%ment of nucleotides that is distinct from that observed with S7 alone. These findings support, t,hc results of earlier assembly studies in &ro: binding to S9 or Sl9 requires S7 (Mizushima tl: Xomura, 1970; Green & Kurland, 1973). Protein SS-dependent protections arc clustered in two regions. One is the cluster of protected nucleotides in the 1241.-1296 compound stem (Fig. 4(b)). Here. the effects of 57 and S9 are closely intertwined. and even cxo-operative interactiotl between all three proteins is evident in this region. A second region affected by protein S9 is centered on the 1125/l 145 internal loop, at the top of’ the domain (Fig. 4(b)). Simultaneous contact of prot,eill S9 with the 1125 and 1280 regions would imply that these two extremities of the domain are brought into mutual proximity by the three-dimensional folding of 16 S rRNA. This conclusion is in agreement’ with the results of’ Atmadja &, Brimacombe (1985) and Stiege it al. ( 1986), who reported a, crosslink between positions 1125-I 127 and 1280-1281 from U.V. irradiation of 30 S subunits, both in vitro and in viva. Our results suggest that protein S9 may stabilize the folding together of these two extremities of the 3’ domain. Protein Sl9 protects many sites in the 950-1050 and 1210-1230 regions (Fig. 4(c)). The effects seem to center around the 985/1220 stem. and include
Interaction
of S7, S9 and S19 with 16 S RNA
many of its llanking single-stranded nucleotides. Extensive enhancements of reactivity toward single-strand-specific probes and loss of attack by the double-strand-specific V, nuclease in the 966 975 region provide good evidence that the 960-9631 972-975 stern is disrupted in the presence of protein S19. The reactivity of this stem in 30 S subunits, however, is quite different, and suggests that it becomes double-stranded in a later step during assembly (Moazed et al., 1986). Thus, the S19dependent disruption of structure in the 960-9X region may provide a site for interaction with another ribosomal component. This may be of special significance, because of the functional importance of this region, as discussed below. Our observations bear on the mechanism of cooperative assembly. One can envision two extreme models to account for the observed assembly interdependencies of these three proteins. At one ext,reme. only the primary binding protein (57) would make direct contact with 16 S rRNA. The secondary binders (S9 and S19) would then bind through protein-protein contacts with S7 (and possibly with each other), somehow strengthening its contacts with the RNA as a result. At the opposite ext,reme. all three proteins would bind directly and solely to the RNA, potentiating each other’s binding interactions by including conformational alterat,ions in the RNA structure. Our results favor a model that incorporates elements of both kinds of mechanism. Proteins S9 and S19 each cause protection of regions of I6 S rRNA that overlap or are closely juxtaposed with sites that are protected by S7 (Fig. 4). This suggests t]hat these three proteins bind to neighboring feat,ures of the RNA structure. I)rpendenct> of S9 and S19 upon S7 for their assembly could t,hus be due to a requirement for protein-protein contact, in addition to proteinRNA contact. for assembly of these secondary binding prot,eins. We note that certain sites that are protected in the presence of S9 or S19 are relatively unreaot,ive, and hence presumably unavailable for pro,tein interaction, in the naked RNA. A particularly clear example is the series of 87. enhanced bases at positions 977 to 980, which is protected in the presence of S19. A more complex example is G1279, which is enhanced by S7, furthei enhanced by S19, and then strongly protected upon addition of S9. It seems likely that protein-induced conformational changes play a significant role in the co-operat,ive assembly of these three proteins. Overlap of the RNA binding sites for 57 and S9 is consistent with the substantial evidence for close proximity of t,hese two proteins in the 30 S subunit. There are seven independent reports of chemical crosslinking between 57 and S9 (for reviews, see Garrett, 1979; Traut. 1980), and one study reported t
317
very close to one another (Moore, 1980: Huang et al., 1975), and immunoelectron microscopy studies are also in support of these placements (Stiiffler et al., 1980: Lake, 1980). Although overlap between S7 and S19 effects, particularly in the 960 region, also suggests proximity of these two proteins, the placement of protein S19 relative to 57 is a subject of some controversy (for a review, see Prince et al., 1983). Tmmunoelectron microscopy studies have located 819 in the “head” of the 30 S subunit very close to S7 (Stiiffler et al., 1980; Lake, 1980) but at least one group also reports a determinant for S19 at the opposite side of the head (Lake, 1985), in agreement with the placement’ from neutron diffraction measurements (Moore, 1980). On the other hand, energy transfer and crosslinking studies place S13 and S19 in close proximity, while results from nearly all approaches agree that 57 and S13 are close, placing 57 and S19 reasonably near one another by extrapolation. Clearly, the issue of the location of protein S19 is yet to be settled: our results would be most consistent with its placement proximal to S7. It has been argued that ribosome function is likely to be defined by its RNA components (Noller & Woese, 1981). What, then, of the ribosomal proteins? One of their roles is in assembly, to facilitate the folding of ribosomal RNA into a compact form (Nomura et al., 1969). Our present findings also suggest that certain ribosomal proteins may regulate the conformation of functional sites in the 16 S rRNA. G1338 is highly reactJive toward kethoxal in 30 S subunits (Noller, 1974; Moazed et uZ., 1986). lt is protected in ribosomes occupied by t RNA (Brow & Noller, 1983; Moazed & Noller, 1986), suggesting that it participates in tRNA binding. G1338 is unreactive in naked 16 S rRNA (Fig. 1, lane 1: Moazed et al.. 1986), but becomes exposed to kethoxal upon binding of protein S7 (Fig. 1, lanes 2 to 5). Thus, another role for S7 appears to be the adjustment of RNA conformation in this functional site. An analogous observation has been made with another primary binding protein. S4, which appears to make specific adjustments in the structure of the functionally important 530 loop (Stern et al., 1986). Protein S19 itself has been implicated in the tRNA binding function (Rummel & Noller, 1973: Shimizu & Craven, 1976a,b: Lin et al., 1984). Here. we provide evidence that S19 affects the structure of the 960 region. m’G966 is another nucleotide t’hat is protected from kethoxal by t,RNA (Brow & Noller, 1983; Moazed & Noller, 1986). raising the possibility that the functional importance of S19 may be related to its ability to perturb t’he conformation of this region of 16 S rRNA. Of particular functional relevance is the set of perturbations in the 1394-1399 region that are dependent on the presence of S9 or S19. The observed enhancement of Al394 and C1399, and protection of C1395, A1396, Cl397 and Al398 (Fig. 3; summarized in Fig. 4) bring the reactivities in this region more closely in line with those
318
T. Powers
observed for the fully assembled 30 8 subunit (Moazed et al., 1986). These are the only assembly effects that we have observed so far with primary and secondary binding proteins, that affect the 1400 region, suggesting that its final conformation is linked in some way to assembly events in the 3’ major domain. This provides another possible connection between these proteins and a functionally important site in 16 S rRNA in light, of the observations that the wobble base of certain tRNAs can be crosslinked photochemically to Cl400 (Prince et al., 1982), and that s4U8 of tRNA can be crosslinked via an 8 w (1 A = 0.1 nm) linker to protein S19 (Lin et al., 1984). Protein S9 has been implicated in the elongation factor G-related GTPase function (Marsh & Parmeggiani, 1973; Cohlberg, 1974). Although essential for this function, as deduced from some reconstitution experiments in vitro, S9 seems to be less crucial when 30 S particles are reconstituted from core particles and split proteins (Traub et al., 1967). Here, it may be relevant that the site of a mut,ation conferring spectinomycin resistance is found at position Cl192 (Sigmund et al., 1984); this antibiotic is thought to block the translocation factor step, which is triggered by elongation G-dependent GTP hydrolysis (for a review, see Cundliffe, 1981). Cl192 is found in that part of the 3’ major domain spanned by the two distal S9protected sites (Fig. 4(b)). Thus: it’ is conceivable that one role of S9 is to help to arrange this region of the RNA into its proper functional conformation. accounting for its effect when omitted from a total reconstitution. The observation that it can be omitted from a split-core reconstitution may simply imply that this region of 16 S rRNA retains its functional conformation during removal of split proteins. These studies were support,ed by grant no. (iM17129 from the U.S. National
Instit,utes of Health (to H.F.N.).
Wr thank Drs B. Nag and R. R. Traut, for providing purified ribosomal proteins for preliminary experiments.
et al
Expert-Besaqon. A., Barritault. I>.. &&let. M.. Guitrin. M.-F. & Hayes, 1). H. (1977). J. Mol. Hlol. 112. 603.. 629. Garrett, R. A. (1979). fnt. tlvu. Hiochem. 25, l21--177. Garrett, It. A.. Rak. K. H.. Daya, I,. dz Stiiffler. G. (1971). &fol. Gen. Cenet. 144, 112-124. Green, M. & Kurland. c. C. (1973). Mol. Rlol. RPJI. 1. 105&111. Huang, K. H.? Fairclough, R. H. &. Ilantor.. (‘. R. (1975). J. Mol. iSol. 97. 433-470.
Lake. ,J. A. (1980). Tn R~JOSOVWS. Structure. k’unction nrtd Genetics (Chambliss, G.. Craven. (+. R.. Davies. J.. I)avis, K., Kahan, L. & Nomura, M.. rds). pp. 207 236, University Park Press. Baltimore. Lake, ,I. L (1985). Ann/u. Rw. Biochem. 54. 307~-530. Lin. F.-L.. Kahan. L. & Ofengand, .J. (1984). J. Mol. Jkiol. 172, 77786. Marsh. R,. 11. 8r Parmeggiani. A. (1973). I’roc. ,Vut. rlccrd. Ax.. 1:ds.d. 70, 151~155. Mizushima, S. & Xomura, M. (1970). rVcltlrrt> ~L~JM~J~L), 266, 1214-1218. Monzed. I). & Noller. H. F. (1986). (Jell. 47. 9X5-994. Moazed. I). I St,ern. S. $ Noller. H. F. (1986). J. ,WoZ.lZ&/. 187, 399--416. Moore. I’. R. (1980). In R~~~J.SOVUM. Structure, FuM&~~JI/, and Genetics (Chambliss. G.. Craven. G. R.. Davies. J., Davis, K.. Kahan, L. & Nomura, M.. eds). pp. 111 133, lTniversit,y Park Press. Baltimore. Morgan. J. h Brimacombe. R. (1972). Eur. J. Hiorh.un,. 29, 542- 562. Noller. H. F. (1974). f~io~h~emistry, 13, 4694 4703. Boiler. H. 17. & Woese. C. RI. (1981). Scielrcr. 212. 403 111. Nomura. kl.. Mieushimat. S.. Ozaki. ,M.. Traub. I’. & lAowry. (‘. v. (1969). (‘old Spring f~arbor t~yntp. d_)uant.Bid. 34, l9-Al. Prince. ,I. K.. Taylor, B. H., Thurlow. I). l,.. C)fengand. .I. Cyr.Zimmrrmann. R. A. (1982). Proc. .Va/. .Acnd. Sri.. I:.K.A
79. 54X-
Trends
Rio&em
Atmadja, J. & Brimacombe, R. (1985). AV~cl. Acids Res. 13. 6919-6936. Brimacombe? R., Morgan, J.. Oakley. D. G. bt (lox, R. A. (1971). Nature New Biol. 231, 209-212. Brimacombe. R., Maly, P. & Zwieb, C. (1983). Proqr. Nucl.
Acid. Res. Mol. Biol.
28, l--48.
Brow, I). A. &, Noller, H. F. (1983). .I. Mol. Biol. 163. 27-46. (‘hangchien, L.-M. & Craven. G. R. (1977). ,J. Ho/. Biol. 113. 103-122. Cohlberg, J. A. (1974). RiQchenr. Biophys. Res. Commun. 57, 225-231. Cundliffe. E. (1981). In The Molecular Basis 01 Antibiotic Action,
pp. 402-547, John Wiley & Sons, London.
Ehresmann, B., Backendorf, C., Ehresmann, C.. Millon. R. & Ebel, J-P. (1980). Eur. .I. Biochem. 104, 255262.
R. It. 6i Garrett.
It. A. (l!W).
Sci. 8. 359-363.
Roth. H. E. & Nierhaus. K. (1973). F’ENS I,rttpr.~. 31. R5-~38. Rummel. I). P. & Noller. H. F. (IY74). lVa/lr,re A:rcrsflioi. 245, 72.-75. Schaup. H. &:.. (ireen. M. & Kurland. (‘. (; (1970). .Wvl. Grrt, Canrt. 109. 193-205. Schendrl. P.. Marba. I’. & (“raven. G. It. (iY72). I’~w. Nat. drad.
References
5454.
Prince. J. B.. Gutell.
Sci.. I‘.$.A.
69, 544 54X.
Shimizu, 31. & (?avrrr. (2. R. (1976n). :Ilol. /2iol. fir,//. 3. IOG- 11I. Shimizu. M. & (‘raveu. (:. K. (19766). h’ur. ,J. fjiochrnc. 61. 307 -315. Sigmund. (‘. I).. Ettayebi, M. & Morgan. E. A. (1984). .Vucl. ilcids Rrs. 12. 4653-4663. Stern, S.. Wilson, R. (‘. & Nollrr. H. F. (1986). ,J. :\lol. Bid. 192, 101 I lo. Stern. S.. C’hangchirn. L-M.. Ciraven. (:. II. & Sollrr~. H. F. (19%). .J. Mol. Biol. 200, 291~~29’3. Stern, S.. hlloazed. I). & Noller. H. F. (1987). .Wethods fhzymo/. in the press. Stiege, W.. Atmadja. J.. Zobawa. M. & f)rimacombr. lb. ,I. Mol. Hiol. 191, 1X-138. I:.. Htiiffler, G.. Bald. lt.; Kastner. IS.. r,utll~f~~arlrr. Sttiffler-Meilicke, -MM..Tischendorf, (:. 6 Trsctrr, 1s. (1980).
In
~?ihoso?ms.
&+?LCtZLM,
~umtion
and &netic’s
((‘hamhliss. G.. Craven. G. R.. Davies. .J.. Davis. K., Kahau. 1,. & Nomura. M.. eds). lq). 171 ~205. llniversit; l’ark Press. Balt,imoreh.
Interaction
of X7, X9 and S19 with 16 S RNA
Svensson. P., Changchien, L.-M., Craven, G. R. & Noller, H. F. (1988). J. Mol. Riol. 200, 301-308. Traub, I’.. Hosokawa, Ii., Craven, G. R. & Nomura, M. (1967). Proc. Nat. Acad. Sci., U.S.A. 58, 2430-2436. Traut, R. R., Lambert, J. M., Boileau, G. & Kenny, Structure, Function and J. vi’. (1980). In Ribosmes. Genetics. (Chambliss. G., Craven, G. R., Davies, J., Davis, K., Kahan, L. & Nomura, M., eds), pp. 89110, 1Jniversity Park Press, Baltimore.
319
Wower, I. & Brimacombe, R. (1983). Nucl. Acids Res. 11, 1419-1437. Yuki, A. & Brimacombe, R. (1975). Eur. J. Biochem. 56, 23-34. Zimmermann, R. A., Muto, A., Fellner, I’., Ehresmann, C. & Branlant. C. (1972). Proc. Nat. Acad. Sci., U.S.A. 69, 1282-1286. Zwieb, C. & Brimacombe, R. (1979). Nucl. rlcids. Res. 6, 1775-1790.
Edited by P. uon Hippel
(lY88)
200, 309-319
Probing the Assembly of the 3’ Major Domain of 16 S Ribosomal RNA Quaternary Interactions Involving Ribosomal Proteins S7, S9 and S19 Ted Powers’, L&Ming Changchien 2, Gary R. Craven2 and Harry F. Noller’ 1Thimann Laboratories University of California at Santa Cruz Santa Cruz, CA 95064, U.S.A. 2Laboratory of Molecular Biology University of Wisconsin Madison, WI 53706, U.S.A. (Received 26 May 1987) We have studied the effect of assembly of ribosomal proteins 57, S9 and S19 on the accessibility and conformation of nucleotides in 16 S ribosomal RNA. Complexes formed between 16 S rRNA and S7, S7+S9, S7+Sl9 or S7+S9+S19 were subjected to a combination of chemical and enzymatic probes, whose sites of attack in 16 S rRNA were identified by primer extension. The results of this study show that: (1) Protein S7 affects the reactivity of an extensive region in the lower half of the 3’ major domain. Inclusion of proteins S9 or S19 with S7 has generally little additional effect on S7specific protection of the RNA. Clusters of nucleotides that are protected by protein S7 are localized in the 935-945 region, the 950/1230 stem, the 1250/1285 internal loop, and the 1350/1370 stem. (2) Addition of protein S9 in the presence of S7 causes several additional effects principally in two structurally distal regions. We observe strong SS-dependent protection of positions 1278 to 1283, and of several positions in the 1125/1145 internal 190~. These findings suggest that interaction of protein S9 with 16 S rRNA results in a structure in which the 1125/1145 and 1280 regions are proximal to each other. (3) Most of the strong SlS-dependent effects are clustered in the 950-1050 and 121CL1230 regions, which are joined by base-pairing in the 16 S rRNA secondary structure. The highly conserved 960-975 stemploop, which has been implicated in tRNA binding, appears t)o be destabilized in the presence of S19. (4) Protein S7 causes enhanced reactivity at several sites that become protected upon addition of S9 or S19. This suggests that S7-induced conformational changes in 16 S rRNA play a role in the co-operativity of assembly of the 3’ major domain.
1. Introduction
the 3’ maior domain was excised from naked 16 S rRNA (Zimmermann et al., 1972) or as an RNP containing a subset of the ribosomal proteins from 30 S subunits (Brimacombe et al., 1971; Yuki & Brimacombe, 1975), by partial nuclease digestion. RNPs isolated by two other groups (Schendel et al., 1972; Roth & Nierhaus, 1973) are probably related to the one that was characterized by Brimacombe and co-workers (Morgan & Brimacombe, 1972; Yuki & Brimacombe, 1975); in all cases, proteins S7, S9 and S19 were found, usually in near-stoichiometric amounts, together with varying amounts of other proteins. Assembly experiments in vitro showed that protein S7 binds independently to 16 S rRNA (Schaup et al., 1970; Mizushima & Nomura, 1970;
Early evidence for the domain organization of ribosomal RNA came from partial nuclease digestion studies of 16 S rRNA and 30 S subunits. In retrospect, the results of several such experiments provide rather convincing evidence that the 3’ major domain of 16 S rRNA (as defined by Noller & Woese, 1981), in particular, forms a relatively autonomous ribonucleoprotein (RNPt) structure within the assembled 30 S subunit. A subfragment corresponding almost precisely to t Abbreviations used: RIVP, ribonucleoprotein; dimethylsulfate; u.v., ultraviolet.
DMS, 309
0022%2836/88/06030~11
$03.00/O
0
1988 .Amdemic
Press Limited
310
T. Powers et al
Garrett et al., 1971); in contrast, proteins S9 and S19 require prior binding of S7 for their assembly (Mizushima & Nomura, 1970; Green & Kurland, 1973). In fact, all of the proteins found in significant amounts in the excised subparticle depend on protein S7 for their assembly (Mizushima & Nomura, 1970). These same proteins are all localized in the “head” of the 30 S subunit, according to immunoelectron microscopy and neutron diffraction studies (Lake, 1980; Stiiffler et al., 1980; Moore, 1980). These observations suggest that the 3’ major domain assembles as a relatively autonomous structural unit. the results of crosslinking Apart) from experiments, which further localize S7 with respect to specific features of the 3’ major domain (Zwieb & Brimacombe, 1979; Wower &, Brimacombe, 1983; Ehresmann et al., 1980), litt,le has been learned concerning the sites of interaction of ribosomal proteins with this domain of 16 S rRNA or, for that matter, whether any proteins other than S7 make contact with this region of the RNA. Nor has any light been shed on the molecular basis of the observed co-operativity of assembly between S7, S9 and S19. In the experiments presented here, we probe the effects of assembly of these three proteins on 16 S rRNA, using a recently developed .approach (Stern et al., 1986). Purified proteins are bound, individually or in combination: to 16 S rRIVA. The resulting RNPs are examined with a set of chemical and enzymatic probes that monitor the accessibility and conformation of each position along the RNA chain. Sites of attack are localized precisely by primer extension (Moazed et al., 1986; Stern et al., 1987). We confirm the co-operativity of assembly of these proteins, and show that each protein influences the reactivity of a distinct set of nucleotides, all of which are found within the 3’ major domain. Both diminished and enhanced reactivities of specific nucleotides result from interaction of 16 S rRNA with the proteins. Clusters of protected bases may indicate sites of protein-RNA contact. Protein S7 causes enhancement of the reactivity of several nucleotides; we interpret this to be the result of local conformational changes in 16 S rRNA caused by interaction with the protein. Some of these sites of enhancement are subsequently protected by S9 or S19, suggesting that S7-induced conformational changes in 16 S rRNA play a part in the co-operativity of assembly of this ribosomal domain.
(a) Isolation
2. Materials
and Methods
of 16 S rRNA
and proteins
S7, S9 and S19
Naked 16 S rRNA was obtained by extraction with phenol of Escherichia coli 30 S ribosomal subunits as described by Stern et aE. (1988). Proteins 57, S9 and S19 were purified by high-pressure liquid chromatography and renatured as described in an accompanying paper (Stern et al., 1988).
The procedure used was essentially the same as that described by Stern et al. (1986). Pro&ins S7. SY and S19 were mixed. either singly or in combination. with 16 S rRNA; 10 or 20 pg of RNA and a 3 to 8-fold molar excess of each prot,ein were used. The mixtures. as well as control samples containing 16 R rRNA without protein. were incubated for 1 h at 42°C’ in 50 ~1 of buffer containing 80 mM-K-Hrpes (pH 7.X), 20 m,n-M&I,. 300 miv-K(yl, 6 mM-B-mercapt’oethanol. 0.01 Oc, Nikko1 detergent (Nikko Chemical Co., Ltd,
and w~z?ymaLic prd~iq
(i) CJhemical modijcatiorc For dimethyl sulfate (DAMS) modification, 2 htl of I)MS (Aldrich), diluted 1 : Id in 95”& (v/v) EtOH. ~WP added KE
DMS
Vl
9: jO-
100 IO-
-
Figure 1. Autoradiographs showing thr results of chemical and enzymatic probing as monitored by primer extension using the 1047 primer (Moazed et al.. 1986: Stern et al.. 1987). Complexes probed: 1. naked 16 S rRNA; 2, S7; 3, S7/S9: 4, S7+819; 5, S7+S9+S19. C. I’. A, G are dideoxy sequencing lanes and refer to the nucleotide sequence of 16 S rRNA. K denotes a 16 8 rRNA control lane treated identically to the other samples except for omission of probes. Probes used: dimethylsulfate (DMS) kethoxal (KE), V, nuclease (V,). Nucleotide positions in 16 S rRNA are indicated. Arrows indicate main protein effects.
Interaction
of 6’7, X9 and S19 with 16 S RNA
to each 50-~1 sample, followed by incubation at 4°C for I .5 h. The reaction was stopped by the addition of 12.5 ~1 (pH 75), 1.0 M-pof DMS stop (1.0 M-T&. HCl mercaptoethanol, 0.1 mM-EDTA) followed by incubation on ice for 10 min. Samples were then precipitated by the addition of 300 ~1 of 95% EtOH and 30 fig of carrier RNA. For kethoxal modification, 2 ~1 of kethoxal (37 mgiml in 207; EtOH; Upjohn) were added to each 50+1 sample. followed by incubation at 4°C for 45 min. Samples were then made 25 mM in potassium borate (pH 7.0) and precipitated as described *above. Precipitated RNP complexes were then centrifuged, extracted with phenol, and precipitated with ethanol as described (Stern et al., 1988). RNA was redissolved in water at a concentration of 0.4 mg/ml. (ii) I’, ,nuclease digestion For V, nuclease digestion. 0.2 units of cobra venom nuclease (V, nuclease; Pharmacia) were added to each 5C)-yl sample and gentlv mixed. Samples were digested for 15 min at 4°C. Reactions were stopped by extract’ion with phenol as described (Stern et al., 1988). After precipitation with ethanol, the RNA was redissolved in water at a concentration of 0.4 mg/ml.
((1) Primer
extension
Hybridization, extension and gel electrophoresis were performed as described by Stern et al., (1988). For each experiment. the entire 16 S rRNA chain (except residues 1508 to 1542) was probed using the same set of DNA primers described by Moazed et al. (1986), with one exception. A new oligonucleotide primer with the sequence TGCTCTCGCGAGGTCGC was used to prime at position 1257. Typically, 16 S rRNA chain was scanned several times for each modification experiment. The data presented in Table 1 are the consensus of visual
--CUAGKl
KE
DMS
311
estimation of band intensities tion experiments.
from 5 separate modifica-
3. Results (a) Experimental
strategy
Assembly studies in vitro have shown that protein S7 binds independently and specifically to 16 S rRNA; binding of S9 and S19 is dependent on prior binding of S7 (Mizushima & Nomura, 1970: Green & Kurland, 1973). Accordingly, our strategy was to form RNP complexes consisting of S7, S7+S9, S7+S19 and S7+S9+S19 together with 16 S rRNA, and to compare the susceptibility of each nucleotide in the RNA to kethozal, dimethyl sulfate and V, nuclease in the various complexes and in naked 16 S rRNA. Mixtures of the secondary binders and 16 S RNA, in the absence of 57, were also probed. Effects attributable to interactions between S9 or S19 and the RNA should be S7dependent, but should require additionally the presence of S9 or S19. It is important to note that effects of the latter class may also be due to S7RNA contacts that somehow depend on the presence of the other two proteins. Tdentification of specific nucleotides and estimation of their altered reactivities in the complexes was accomplished by primer extension using a series of synthetic DNA oligomers, as described previously (Moazed et al., 1986; Stern et al., 1988). In each case, the entire 16 S rRNA chain was scanned, except for the 3’terminal 35 nucleotides, which are not accessible to examination by primer extension. The data, in the form of autoradiographs, are presented in Figures 1 to 3, and summarized in Table 1 and Figure 4; KE --CUAGK123451234512345
Vl
23451234512345
DYS
Vl
lllO-
(b) 1150-
DMS CU AGK
KE
Vl
I 2 3 4 5 1 2 3 4 5 I 2 3 4 5
1230-
Figure 2. Autoradiographs showing the results of chemical and enzymatic probing as monitored by primer extension using (a) and (c) t,he 1257 primer and (b) the 1200 primer. Symbols, numbers and abbreviations are as described for Fig. 1.
+I
;
+t
V
ti
3
+ +o+l=oc++l=c +
=;=
+I
=
ii
0 +I
+ i-
+I v +
+I tl t-1
+ A
+ ;+++I + sF :! Fq
P -2 2 5
+ +I +v
+ E
fl fl +I
===
==c
+ 0 =
ii
+j z
0
“3; 15s I +
=0300+1
++
+
+I
+I+I+++VI+++I+++++I++o
=c
=5
+I ” c +I
+I +I vv
++
+ +I +
+ +I
0
+I +I +I c v v v
+I===
+++;
c t’ +-I +I
==
+
+ +
+ :+
+ +I t
t
;
tl +!
<+t
0
+ + <+ + 0 ++ 0 ++ 0 0 <+ k zk
* + + + +
0 III
I!c
0 0 0 0
& k 0 i + 0 + <++ + + 0 + + iY k +
0 0 0 0
+ + + z!T
+ * 0 ii + ++ + ++ + + 0 + +
+ t +
0 0 <* f * *
0
+ + <+ + 0
<+ ++ ++
++
+ t++
++ +
* * iz
0 0
Al319 G1331 Al332 Al333 G1334 G1338 A1339 Al346 A1349 u1351 Cl352 G1353 u1354 Al357 Cl359 Al360 G1361 Al362 Al363 G1365 Cl367 G1371 Al374 Al375 Al377 Cl378 U1381 Cl382 Cl383 Al394 Cl395 A1396 Cl397 A1398 Cl399 + k + 0 0 xi + k + + <+ <+
+ + + k + 0 0 i AZ
+ + 5 f * +
0 <+ + 0
0
+ <+ ++ 5 XL 5
0 0 0 0 ++ k 0 0
+ + <+ * f +
0 f 0 + 0 *
f + <+ k k +
<++
0 f 0 k
+ ++ +++ if
0 0
0 * + +
0 0 0
Data are the consensus of visual estimates of relative band intensities from 5 dillerent probing experiments. Chemical probing data indicate kethoxal modification of G, DMS modification of A, except where noted. Keactivities are summarized qualitatively by the symbols 0, f, +, + + and + + f, where 0 indicates no reactivity and + + + indicates maximal or hyper-reactivity. In the case of Vi nuclease, the nucleotide 5’ to the cut site is indicated. For clarity, data are indicated only where a result differs from the column immediately to the left; i.e. a blank indicates that the result is the same as that found in the column to its left. Reverse transcriptase pauses or stops at positions U950 and l’1126 when naked 16 S rNNA is treated with kethoxal and U1116, lJ1212, U1224 and U1381 when treated with DMS. The chemical basis of these apparently anomalous reactivitirs is not known.
I:989 11991 G993 A4994 A996 1’1107 U1008 81014 A1016 A1036 A1046 G1047 41110 All11 Cl112 Ull16 U1126 G1127 A1130 G1138 All45 Al 146 A1150 G1156 G1182 IJ1212 A1213 Cl214 Cl217 Cl218 A1219 U1224 Al225 Cl226 Cl228 A1229
314
T. Powers et al KE
-CUAGK123451234512345
DMS
VI
d
5 I -
3
1
Figure 3. Autoradiographs showing the results of chemical and enzymatic probing as monitored by primer extension using the 1490 primer. Symbols, numbers and abbreviations are as described for Fig. 1.
below, we present separately the results obtained for each of the RNP complexes. (b) S7-speci$c
effects
The reactivities of nucleotides from a large region of the 3’ major domain are affected by binding protein S7. This is seen as increased or decreased susceptibility to the chemical and enzymatic probes when complexes of S7 and 16 S rRNA are compared with naked 16 S rRNA (Figs 1 and 3, lanes 1 and 2; Table 1). The distribution of nucleotides affected by 57 is shown in Figure 4(a); they are found in the lower half of the 3’ major domain. As for all of the complexes studied here, we
failed to detect any effects outside of the 3’ major domain (nucleotides 920 to 1400). Contacts between S7 and 16 S rRNA are most likely t,o occur in regions where the stronger protections are concentrated. These regions include the 935-945 region, the 950/1230 stem and its flanking sequences on the lower side, the 1250/1285 internal loop, the 1300 and 1330 regions, and the 1350/1370 stem-loop structure. We also observe several instances where S7 causes increased reactivity of a nucleotide toward our probes. Such enhanced reactivity has been observed in previous studies, both for 30 S subunit assembly (Moazed et al., 1986) and in binding individual proteins to the 5’ and central domains (Stern et al., 1986, 1988; Svensson et al.. 1988). Major enhancements occur at positions 978 and 979 (toward DMS), at’ positions 1278, 1279 and 1338 (toward kethoxal) and in the 1310 helix (toward V, nuclease). Significantly, many of these posit,ions become protected as a result of binding S9 or Sl9. as described below. (c) S9-dependent
eflects
Protein S9 by itself has no detectable effect on the modification pattern of 16 S rRNA (data not shown), but when both S9 and S7 are present’, we observe several effects that are distinct from those produced by S7 alone (Figs 1 to 3, lanes 3 and 5). Strong SS-dependent protections are found in t,he compound stem comprising positions 1241 to 1296 (Fig. 3, lanes 3 and 5). These are concent,rated in the 1245/1290 helix (V, protection) and in the 1255 1285 helix and its flanking sequence around position 1280 (V,, kethoxal and DMS protections). Three of these protected positions (1278, 1279 and 1281) are the sites of S7-induced enhancements. These results give the impression of intermingling of proteins S7 and S9 in the 1250/1280 region. A second cluster of SS-dependent protections is detected in the 1125-1145 region, quite distal in the secondary structure from the site of the other S9 effects (Fig. 2). Most evident is protection of Ull26. G1127, All45 and A1156: protection of several other nucleotides in this region is also observed. as enhancements. These are some SS-dependent findings suggest, that protein S9 may mediate thr close approach of the 1130 and 1280 regions (set, Fig. 4(b)). (d) SlS-dependent
effects
Numerous strong Sly-dependent’ effects are observed throughout the 950-1050 and 1210-1230 regions (Figs 1 and 2; lanes 4 and 5). The main protections are concentrated around the 98511220 and 1010/1020 stems and their respective singlestranded flanking sequences (Fig. 4(c)). Interseveral enhancements of chemical estingly, reactivity and diminished attack by V, nuclease suggest disruption of the 960-9631972-975 stem. Other effects of S9 and S19 are presented below.
Interaction
of S7, S9 and S19 with 16 S RNA
(b)
(c)
Figure 4. A diagram of the 3’ major domain of 16 S 16 S rRNA-protein complexes: (a) S7 relative to naked and (d) S7+S9+819 relative to both S7+S9 and 87 toward the chemical probes are indicated by (0) or enhanced reactivities toward V, nuclease are indicated
(d)
rRNA (nucleotides 920 to 1400) summarizing results of probing 16 S rRNA; (b) 87 + S9 relative to S7; (c) S7 +S19 relative to S7; +S19. Nucleotides exhibiting reduced or enhanced reactivities (A), respectively. Phosphodiester bonds exhibiting reduced or by (open arrows) or (closed arrows), respectively.
316
T. Powers et al. (e) Complex effects
In addition to the strong effects described above, which appear to be highly specific for the presence of a given protein, some regions of the RNA show a less well-defined pattern of protein-dependent perturbations. These include the 1245-1290, 13101320, 135&1365 and 1394-1399 regions (Figs 3 and 4). Most complex among these is the 1245-1290 region. G1268 is not perturbed unless all three proteins are present, in which case it has diminished reactivity. Reactivity of G1279 is enhanced by 57, further enhanced by Sl9, but protected by S9. Several other nucleotides in the 126&1280 region effects that depend on the presence of all three proteins. Nucleotides in this region are also affected by 57, S7+S9 or 57 +Sl9. The 1260-1280 region is clearly the site of extensive co-operative protein-RNA interactions. The 13lCl320 and 135&1365 stem-loop regions are similar, in that they both seem to be affected by each of the three proteins to some extent. Moreover, protection from chemical probes tends to be incomplete in many cases. There are two examples where the same end result is achieved in the presence of either S7+S9 or S7+Sl9. The 1350 stem is fully protected from V, nuclease by either combination (Fig. 3). Similarly, a series of protections and enhancements at positions 1394 to 1399 is induced by either S7+S9 or S7+519. although the S7+Sl9 combination appears to give the stronger effect (Fig. 3).
4. Discussion The experiments presented here demonstrate that specific nucleotides in the 3’ major domain of 16 S rRNA show decreased (or increased) susceptibility to chemical and enzymatic probes when various combinations of the ribosomal proteins X7, S9 and Sl9 are bound (summarized in Fig. 4). Although we believe that many of these effects, and clusters of strongly protected nucleotides in particular, are indicative of specific RNA-protein contacts, any rigorous interpretation demands caution in two respects. First, protection of a nucleotide need not be the result of direct contact with protein, but could as well be attributed to conformational changes caused by interaction of the protein with a more remote region of the RNA. A second consideration arises from the dependence of S9 and Sl9 on X7 for their assembly. Here, we cannot exclude the possibility that protections dependent on the presence of S9 or Sl9 are the result of new contacts between 57 and 16 S rRNA, rather than to direct interaction between S9 or Sl9 with the RNA. Protein S7 protects ma’ny nucleotides in the lower half of the 3’ major domain (Fig. 4(a)), and also causes several enhancements there. Major clusters of protected nucleotides are found in four main regions: the 935-945 region, the 950/1230 stem, the 1250-1285 internal loop, and the 1250/1370 stemloop (Fig. 4(a)). Assignment of S7 contacts to these
three regions is in close agreement witch the results of chemical crosslinking studies reported by Brimacombe and his colleagues. Tn one study. protein S7 was crosslinked to Ul240 by ultraviolet (u.v.) irradiation of 30 S subunits (Zwieb 8 Brimacombe, 1979). Tn another st,udy, using U.Y. irradiation of 2-iminothiolane-derivated subunits, S7 was crosslinked to the same region (within residues 1234 to 1241) and to a second site, within the dinucleotide AC! 1377-1378 (Wower & Brimacombe, 1983). The former crosslink lies precisely between the first two clusters of Hiprotected residues. and is proximal to both of them; the latter crosslink is closely adjacent to the third cluster in the 1350/1370 stem. Ehresmann et al. (1980) have reported a crosslink between protein 87 and C1265, in the same general region as the observed protections. However, t,he placement of the latter crosslink has been questioned by Brimacombe and co-workers (Brimacombc et trl., 1983), who reinterpret the relevant RNA sequentring data to infer that the crosslink is in fact identical with that reported by Zwieb & Brimacombe (1979), i.e. to position 1240. Finally, all of the nucleotides affected by protein S7 art’ contained well within the limits of its binding region, as defined by RNA subfragments obtained from either S7-16 S rRNA complexes (Zimmermann et (~1.. 1972) or from subparticles obtained by nuclease digestion of 30 S subunits (Yuki $ Krimacornbr, 1975). Proteins S9 and 819 have no effect on 16 S rRNA, either singly or in combination with each ot’her. When present in combination with Si, however, both proteins cause protection or enhanct%ment of nucleotides that is distinct from that observed with S7 alone. These findings support, t,hc results of earlier assembly studies in &ro: binding to S9 or Sl9 requires S7 (Mizushima tl: Xomura, 1970; Green & Kurland, 1973). Protein SS-dependent protections arc clustered in two regions. One is the cluster of protected nucleotides in the 1241.-1296 compound stem (Fig. 4(b)). Here. the effects of 57 and S9 are closely intertwined. and even cxo-operative interactiotl between all three proteins is evident in this region. A second region affected by protein S9 is centered on the 1125/l 145 internal loop, at the top of’ the domain (Fig. 4(b)). Simultaneous contact of prot,eill S9 with the 1125 and 1280 regions would imply that these two extremities of the domain are brought into mutual proximity by the three-dimensional folding of 16 S rRNA. This conclusion is in agreement’ with the results of’ Atmadja &, Brimacombe (1985) and Stiege it al. ( 1986), who reported a, crosslink between positions 1125-I 127 and 1280-1281 from U.V. irradiation of 30 S subunits, both in vitro and in viva. Our results suggest that protein S9 may stabilize the folding together of these two extremities of the 3’ domain. Protein Sl9 protects many sites in the 950-1050 and 1210-1230 regions (Fig. 4(c)). The effects seem to center around the 985/1220 stem. and include
Interaction
of S7, S9 and S19 with 16 S RNA
many of its llanking single-stranded nucleotides. Extensive enhancements of reactivity toward single-strand-specific probes and loss of attack by the double-strand-specific V, nuclease in the 966 975 region provide good evidence that the 960-9631 972-975 stern is disrupted in the presence of protein S19. The reactivity of this stem in 30 S subunits, however, is quite different, and suggests that it becomes double-stranded in a later step during assembly (Moazed et al., 1986). Thus, the S19dependent disruption of structure in the 960-9X region may provide a site for interaction with another ribosomal component. This may be of special significance, because of the functional importance of this region, as discussed below. Our observations bear on the mechanism of cooperative assembly. One can envision two extreme models to account for the observed assembly interdependencies of these three proteins. At one ext,reme. only the primary binding protein (57) would make direct contact with 16 S rRNA. The secondary binders (S9 and S19) would then bind through protein-protein contacts with S7 (and possibly with each other), somehow strengthening its contacts with the RNA as a result. At the opposite ext,reme. all three proteins would bind directly and solely to the RNA, potentiating each other’s binding interactions by including conformational alterat,ions in the RNA structure. Our results favor a model that incorporates elements of both kinds of mechanism. Proteins S9 and S19 each cause protection of regions of I6 S rRNA that overlap or are closely juxtaposed with sites that are protected by S7 (Fig. 4). This suggests t]hat these three proteins bind to neighboring feat,ures of the RNA structure. I)rpendenct> of S9 and S19 upon S7 for their assembly could t,hus be due to a requirement for protein-protein contact, in addition to proteinRNA contact. for assembly of these secondary binding prot,eins. We note that certain sites that are protected in the presence of S9 or S19 are relatively unreaot,ive, and hence presumably unavailable for pro,tein interaction, in the naked RNA. A particularly clear example is the series of 87. enhanced bases at positions 977 to 980, which is protected in the presence of S19. A more complex example is G1279, which is enhanced by S7, furthei enhanced by S19, and then strongly protected upon addition of S9. It seems likely that protein-induced conformational changes play a significant role in the co-operat,ive assembly of these three proteins. Overlap of the RNA binding sites for 57 and S9 is consistent with the substantial evidence for close proximity of t,hese two proteins in the 30 S subunit. There are seven independent reports of chemical crosslinking between 57 and S9 (for reviews, see Garrett, 1979; Traut. 1980), and one study reported t
317
very close to one another (Moore, 1980: Huang et al., 1975), and immunoelectron microscopy studies are also in support of these placements (Stiiffler et al., 1980: Lake, 1980). Although overlap between S7 and S19 effects, particularly in the 960 region, also suggests proximity of these two proteins, the placement of protein S19 relative to 57 is a subject of some controversy (for a review, see Prince et al., 1983). Tmmunoelectron microscopy studies have located 819 in the “head” of the 30 S subunit very close to S7 (Stiiffler et al., 1980; Lake, 1980) but at least one group also reports a determinant for S19 at the opposite side of the head (Lake, 1985), in agreement with the placement’ from neutron diffraction measurements (Moore, 1980). On the other hand, energy transfer and crosslinking studies place S13 and S19 in close proximity, while results from nearly all approaches agree that 57 and S13 are close, placing 57 and S19 reasonably near one another by extrapolation. Clearly, the issue of the location of protein S19 is yet to be settled: our results would be most consistent with its placement proximal to S7. It has been argued that ribosome function is likely to be defined by its RNA components (Noller & Woese, 1981). What, then, of the ribosomal proteins? One of their roles is in assembly, to facilitate the folding of ribosomal RNA into a compact form (Nomura et al., 1969). Our present findings also suggest that certain ribosomal proteins may regulate the conformation of functional sites in the 16 S rRNA. G1338 is highly reactJive toward kethoxal in 30 S subunits (Noller, 1974; Moazed et uZ., 1986). lt is protected in ribosomes occupied by t RNA (Brow & Noller, 1983; Moazed & Noller, 1986), suggesting that it participates in tRNA binding. G1338 is unreactive in naked 16 S rRNA (Fig. 1, lane 1: Moazed et al.. 1986), but becomes exposed to kethoxal upon binding of protein S7 (Fig. 1, lanes 2 to 5). Thus, another role for S7 appears to be the adjustment of RNA conformation in this functional site. An analogous observation has been made with another primary binding protein. S4, which appears to make specific adjustments in the structure of the functionally important 530 loop (Stern et al., 1986). Protein S19 itself has been implicated in the tRNA binding function (Rummel & Noller, 1973: Shimizu & Craven, 1976a,b: Lin et al., 1984). Here. we provide evidence that S19 affects the structure of the 960 region. m’G966 is another nucleotide t’hat is protected from kethoxal by t,RNA (Brow & Noller, 1983; Moazed & Noller, 1986). raising the possibility that the functional importance of S19 may be related to its ability to perturb t’he conformation of this region of 16 S rRNA. Of particular functional relevance is the set of perturbations in the 1394-1399 region that are dependent on the presence of S9 or S19. The observed enhancement of Al394 and C1399, and protection of C1395, A1396, Cl397 and Al398 (Fig. 3; summarized in Fig. 4) bring the reactivities in this region more closely in line with those
318
T. Powers
observed for the fully assembled 30 8 subunit (Moazed et al., 1986). These are the only assembly effects that we have observed so far with primary and secondary binding proteins, that affect the 1400 region, suggesting that its final conformation is linked in some way to assembly events in the 3’ major domain. This provides another possible connection between these proteins and a functionally important site in 16 S rRNA in light, of the observations that the wobble base of certain tRNAs can be crosslinked photochemically to Cl400 (Prince et al., 1982), and that s4U8 of tRNA can be crosslinked via an 8 w (1 A = 0.1 nm) linker to protein S19 (Lin et al., 1984). Protein S9 has been implicated in the elongation factor G-related GTPase function (Marsh & Parmeggiani, 1973; Cohlberg, 1974). Although essential for this function, as deduced from some reconstitution experiments in vitro, S9 seems to be less crucial when 30 S particles are reconstituted from core particles and split proteins (Traub et al., 1967). Here, it may be relevant that the site of a mut,ation conferring spectinomycin resistance is found at position Cl192 (Sigmund et al., 1984); this antibiotic is thought to block the translocation factor step, which is triggered by elongation G-dependent GTP hydrolysis (for a review, see Cundliffe, 1981). Cl192 is found in that part of the 3’ major domain spanned by the two distal S9protected sites (Fig. 4(b)). Thus: it’ is conceivable that one role of S9 is to help to arrange this region of the RNA into its proper functional conformation. accounting for its effect when omitted from a total reconstitution. The observation that it can be omitted from a split-core reconstitution may simply imply that this region of 16 S rRNA retains its functional conformation during removal of split proteins. These studies were support,ed by grant no. (iM17129 from the U.S. National
Instit,utes of Health (to H.F.N.).
Wr thank Drs B. Nag and R. R. Traut, for providing purified ribosomal proteins for preliminary experiments.
et al
Expert-Besaqon. A., Barritault. I>.. &&let. M.. Guitrin. M.-F. & Hayes, 1). H. (1977). J. Mol. Hlol. 112. 603.. 629. Garrett, R. A. (1979). fnt. tlvu. Hiochem. 25, l21--177. Garrett, It. A.. Rak. K. H.. Daya, I,. dz Stiiffler. G. (1971). &fol. Gen. Cenet. 144, 112-124. Green, M. & Kurland. c. C. (1973). Mol. Rlol. RPJI. 1. 105&111. Huang, K. H.? Fairclough, R. H. &. Ilantor.. (‘. R. (1975). J. Mol. iSol. 97. 433-470.
Lake. ,J. A. (1980). Tn R~JOSOVWS. Structure. k’unction nrtd Genetics (Chambliss, G.. Craven. (+. R.. Davies. J.. I)avis, K., Kahan, L. & Nomura, M.. rds). pp. 207 236, University Park Press. Baltimore. Lake, ,I. L (1985). Ann/u. Rw. Biochem. 54. 307~-530. Lin. F.-L.. Kahan. L. & Ofengand, .J. (1984). J. Mol. Jkiol. 172, 77786. Marsh. R,. 11. 8r Parmeggiani. A. (1973). I’roc. ,Vut. rlccrd. Ax.. 1:ds.d. 70, 151~155. Mizushima, S. & Xomura, M. (1970). rVcltlrrt> ~L~JM~J~L), 266, 1214-1218. Monzed. I). & Noller. H. F. (1986). (Jell. 47. 9X5-994. Moazed. I). I St,ern. S. $ Noller. H. F. (1986). J. ,WoZ.lZ&/. 187, 399--416. Moore. I’. R. (1980). In R~~~J.SOVUM. Structure, FuM&~~JI/, and Genetics (Chambliss. G.. Craven. G. R.. Davies. J., Davis, K.. Kahan, L. & Nomura, M.. eds). pp. 111 133, lTniversit,y Park Press. Baltimore. Morgan. J. h Brimacombe. R. (1972). Eur. J. Hiorh.un,. 29, 542- 562. Noller. H. F. (1974). f~io~h~emistry, 13, 4694 4703. Boiler. H. 17. & Woese. C. RI. (1981). Scielrcr. 212. 403 111. Nomura. kl.. Mieushimat. S.. Ozaki. ,M.. Traub. I’. & lAowry. (‘. v. (1969). (‘old Spring f~arbor t~yntp. d_)uant.Bid. 34, l9-Al. Prince. ,I. K.. Taylor, B. H., Thurlow. I). l,.. C)fengand. .I. Cyr.Zimmrrmann. R. A. (1982). Proc. .Va/. .Acnd. Sri.. I:.K.A
79. 54X-
Trends
Rio&em
Atmadja, J. & Brimacombe, R. (1985). AV~cl. Acids Res. 13. 6919-6936. Brimacombe? R., Morgan, J.. Oakley. D. G. bt (lox, R. A. (1971). Nature New Biol. 231, 209-212. Brimacombe. R., Maly, P. & Zwieb, C. (1983). Proqr. Nucl.
Acid. Res. Mol. Biol.
28, l--48.
Brow, I). A. &, Noller, H. F. (1983). .I. Mol. Biol. 163. 27-46. (‘hangchien, L.-M. & Craven. G. R. (1977). ,J. Ho/. Biol. 113. 103-122. Cohlberg, J. A. (1974). RiQchenr. Biophys. Res. Commun. 57, 225-231. Cundliffe. E. (1981). In The Molecular Basis 01 Antibiotic Action,
pp. 402-547, John Wiley & Sons, London.
Ehresmann, B., Backendorf, C., Ehresmann, C.. Millon. R. & Ebel, J-P. (1980). Eur. .I. Biochem. 104, 255262.
R. It. 6i Garrett.
It. A. (l!W).
Sci. 8. 359-363.
Roth. H. E. & Nierhaus. K. (1973). F’ENS I,rttpr.~. 31. R5-~38. Rummel. I). P. & Noller. H. F. (IY74). lVa/lr,re A:rcrsflioi. 245, 72.-75. Schaup. H. &:.. (ireen. M. & Kurland. (‘. (; (1970). .Wvl. Grrt, Canrt. 109. 193-205. Schendrl. P.. Marba. I’. & (“raven. G. It. (iY72). I’~w. Nat. drad.
References
5454.
Prince. J. B.. Gutell.
Sci.. I‘.$.A.
69, 544 54X.
Shimizu, 31. & (?avrrr. (2. R. (1976n). :Ilol. /2iol. fir,//. 3. IOG- 11I. Shimizu. M. & (‘raveu. (:. K. (19766). h’ur. ,J. fjiochrnc. 61. 307 -315. Sigmund. (‘. I).. Ettayebi, M. & Morgan. E. A. (1984). .Vucl. ilcids Rrs. 12. 4653-4663. Stern, S.. Wilson, R. (‘. & Nollrr. H. F. (1986). ,J. :\lol. Bid. 192, 101 I lo. Stern. S.. C’hangchirn. L-M.. Ciraven. (:. II. & Sollrr~. H. F. (19%). .J. Mol. Biol. 200, 291~~29’3. Stern, S.. hlloazed. I). & Noller. H. F. (1987). .Wethods fhzymo/. in the press. Stiege, W.. Atmadja. J.. Zobawa. M. & f)rimacombr. lb. ,I. Mol. Hiol. 191, 1X-138. I:.. Htiiffler, G.. Bald. lt.; Kastner. IS.. r,utll~f~~arlrr. Sttiffler-Meilicke, -MM..Tischendorf, (:. 6 Trsctrr, 1s. (1980).
In
~?ihoso?ms.
&+?LCtZLM,
~umtion
and &netic’s
((‘hamhliss. G.. Craven. G. R.. Davies. .J.. Davis. K., Kahau. 1,. & Nomura. M.. eds). lq). 171 ~205. llniversit; l’ark Press. Balt,imoreh.
Interaction
of X7, X9 and S19 with 16 S RNA
Svensson. P., Changchien, L.-M., Craven, G. R. & Noller, H. F. (1988). J. Mol. Riol. 200, 301-308. Traub, I’.. Hosokawa, Ii., Craven, G. R. & Nomura, M. (1967). Proc. Nat. Acad. Sci., U.S.A. 58, 2430-2436. Traut, R. R., Lambert, J. M., Boileau, G. & Kenny, Structure, Function and J. vi’. (1980). In Ribosmes. Genetics. (Chambliss. G., Craven, G. R., Davies, J., Davis, K., Kahan, L. & Nomura, M., eds), pp. 89110, 1Jniversity Park Press, Baltimore.
319
Wower, I. & Brimacombe, R. (1983). Nucl. Acids Res. 11, 1419-1437. Yuki, A. & Brimacombe, R. (1975). Eur. J. Biochem. 56, 23-34. Zimmermann, R. A., Muto, A., Fellner, I’., Ehresmann, C. & Branlant. C. (1972). Proc. Nat. Acad. Sci., U.S.A. 69, 1282-1286. Zwieb, C. & Brimacombe, R. (1979). Nucl. rlcids. Res. 6, 1775-1790.
Edited by P. uon Hippel