Biochimie, 69 (1987) 1071 - 1080 1071 © Soci6t6 de Chimie biologique/Elsevier, Paris

Probing ribosome structure using short oligodeoxyribonucleotides: the question of resolution

Walter E. HILL and Anchalee TASSANAKAJOHN

Chemistry Departement, University o f Montana, Missoula, MT, 59812 USA

(Received 16-5-1987, accepted 25-5-1987)

Summary - The structure of ribosomal RNA in situ can be probed using short, complementary DNA oligomers. As these oligomers bind to exposed, single-stranded regions of rRNA, the stability of the hybridiz- ed complex can be assayed. Differences in binding stability between cDNA probes of similar length and composition may be indicative of the presence of competing structure, such as base-paired rRNA regions, tRNA interactions or protein interactions. In this study the degree to which such interactions can be distinguished is studied. It is found that by using suitable controls, interactions between rRNA and tRNA or rRNA can be discriminated to a resolution of one or two bases. This resolution promises to be impor- tant in delineating the higher, order structure of the rRNA.

ribosomal RNA l DNA oligomers I hybridization I structure of rRNA

REsum~ - Sondage de la structure des r ibosomes en ptilisant des oligod~soxynucl~otides courts; la quest ion de r6solution. L a structure de i ' A R N ribosomal in situ peut ~tre sondEe en utilisent de courts oligombres d ' A D N compl~mentaires. Comme ces oligombres s'attachent ~ des rdgions exposEes, monocatdnaires du rARN, on peut tester la stabilitE du complexe hybridd. Des diffdrences de stabilitd d'attachement entre des sondes cADN de longueur et composition similaires peuvent donner une indica- tion sur la prdsence de structures compdtitives, telles que des rdgions apparides d ' A R N ribosomal, des interactions de t A R N ou des interactions de protdines. Nous dtudions dans ce travail le degrd oft l'on peut distinguer de telles interactions. Nous avons trouvd qu'en utilisant des contr61es addquats, les interactions entre r A R N et t A R N ou r A R N peuvent ~tre distinguEes ~t une rdsolution d'une ou deux bases. Cette rEsolu- tion promet d'Etre importante en ddlindant la structure d'ordre supdrieur de I 'ARN ribosomal.

A R N ribosomal I oligom~res A D N I hybridation / structure de I 'ARN ribosomal

Introduction

The finesse with which ribosomes perform their rol~ in protein synthesis is only beginning to I~e understood. Although it has long ~.een postulated that ribosomes are t:lynamic particles, the mechanism by which they f~nctipn refflgins obscure. As tlie structure of the ribosomal RNA

(rRNA) has been unraveled, there has been increas- ing evidence for 'switches' [1], regions of the rRNA which may have altel~, ate base-pairing: In addition, recent studies Using chemical probing techniques have clearly shown varying reactivity of portions of the rRl~?~ ~epencfi'ng upon whether the ribosome was 90mplexgd" with tR.NA, mRNA 9 r W~s in the active or inactive conformatiola [2-5].

1072 W.E. Hill and A . Tassanakajohn

In earlier work we postulated that rRNA should have active regions that were exposed and single- stranded. Using short DNA oligomers which were complementary to various single-stranded regions o f rRNA (eDNA probes), we found sites which were active in subunit association [6] and tRNA binding (Tassanakajohn & Hill, manuscript submitted). In the process of carrying out these studies, it was noted that small changes in the position or length o f the eDNA probe could cause marked differences in binding o f the probe.

The question that arises with eDNA probing is how much detail can be discerned using this techni- que. What is the limit of resolution? Is it possible to discern single or multiple base-pair changes in structure? In an effort to try to evaluate the amount o f resolution offered by this technique, we have analyzed the results obtained f rom probing three different regions o f rRNA.

The 807-809 region o f 23S rRNA has been im- plicated in tRNA binding. Specifically it has been suggested by Barta et al. [7] that this region of 23S rRNA may base pair with the 3"-terminal CCA of tRNA. Bases 807-809, part of a longer single- stranded region (bases 801-811), can readily be pro- bed using eDNA probes complementary to various portions of this longer region to see if the tRNA presence affects the binding of the eDNA probe, specifically at bases 807-809.

A base-pair ing interact ion between bases 1394-1506 and 1395-1505 has been postdated to oc- cur in the active form of the 30S subunit based on results found from chemical probing by Moazed et al. [5]. The 1394 and 1395 bases are found in a long, single-stranded region (1394-1408) of 16S rRNA that is known to interact closely with tRNA. By probing this two-base interaction site using probes which either cover or miss those bases, evidence for probe specificity good to a resolution of two bases has been obtained.

Resolution to a single base can be checked using C1400 in 16S rRNA. This base occurs in the long 1394-1408 single-stranded region noted above. It has been shown that this base - and only this base - can be cross-linked to the 5'-wobble base in the anti-codon loop of tRNA when the tRNA is bound to either the A- or the P-site [8, 9]. This suggests that this base is uniquely situated with respect to the tRNA in the P-site. Although it is understood that tRNA does not 'bind ' the rRNA here, cross- linking evidence shows that the distance between C 1400 and the 5'-base in the anti-codon loop must be less than 4 ~, [8]. Probing this region using probes that either do - or do not - cover C1400

has shown that probes can be used to discriminate single-base interactions.

In this paper we present the results of such studies and apply the implications derived therefrom to define the resolution of probing studies using short cDNA oligomers.

Materials and methods

Ribosomes and ribosomal subunits Ribosomes were prepared from freshly-grown E. coil strain MRE 600 by the methods of Hill et ai. [10]. Ribosomal subunits were then purified using zonal centrifugation as outlined by Tam and Hill [11] except that the suburtits were pelleted out of the sucrose using centrifugation rather than using ethanol [12]. Ribosomes and ribosomal subunits were stored in small aliquots at -70°C. Tight-couple ribosomes were separated from crude ribosomes by centrifuging crude ribosomes through a 5-20o70 sucrose gradient in 10 mM Tris-HC1, pH 7.2, 100 mm KCI and 6 mM MgCl 2. The fractions were pool- ed and the ribosomes were pelleted by centrifuging in a Beckman Ti70 rotor at 60000 rpm overnight. The ribosomal pellet was then resuspended in a buffer con- taining 10 mM Tris-HCl, pH 7.2, 100 mM NH4CI and 10 mM MgCl 2 and dialyzed against this buffer overnight with three changes of dialysate. The 30S ribosomal subunits were activated by incubating the 30S ribosomal subunits in a buffer containing I0 mM MgCI 2, 100 mM KCI and l0 mM Tris-HCl, pH 7.4 at 42°C for 20 min [13].

tRNA preparation and labeling Transfer RNA Ph~ was purchased from Boehringer (Mannheim) in lyophilized form, dissolved and dialyzed against a buffer containing 10 mM Tris-Ac, pH 7.8, 14 mM Mg(OAc)2, 60 mM K(OAc) and 0.1 mM DTT. The tRNA Ph~ was dephosphorylated and 5' labeled with [~,-32-p]ATP [14].

DNA oligonucleotides The DNA probes were chemically synthesized using phosphoramidite chemistry on a Biosearch 8600 automatic DNA synthesizer. Deblocked synthetic DNA was purified using a Gilson HPLC using a reverse phase C-18 column. Oligomers were stored at -70°C. Purified DNA probe was 5'-end labeled using [?.-32-P]ATP (New England Nuclear) and T4 polynucleotide kinase (Bethesda Research Laboratory) according to the method outlined by Chaconas and Van de Sande [14].

Binding cDNA probes Binding of cDNA probes to ribosomal subunits was car- ried out by incubating 30S, 50S or 70S ribosomes with 5'- [3z.p] end-labeled probe for 5-18 hr at 4°C in 50-200/~1 binding buffer (10 mM Tris-HCl, pH 7.4, 100 mM KCi, 10 mM MgCI2). The reactions were filtered through nitrocellulose filters and washed twice with I ml

1073

Results

aliquots o f binding buffer . The amount o f radioactivity retained by the filters was determined by liquid scintilla- t ion. The specificity o f a probe for a given site or sites is indicated when the amount o f subunit bound saturates to a constant level as the amoun t o f probe in the binding react ion is increased.

RNase H digestion To determine the exact site(s) of the hybridization of the cDNA probes, the heteroduplex probe-subunit complex was incubated with RNase H and the digestion products analyzed. The conditions were similar to those outlined by Donis-Keller [15] and Mankin et ai. [16]. Digestion reactions containing 20 pg ribosomal subunits, 20/zg cDNA probe and 2-3 units of RNase H (Promega) in 20 pl of 40 mM Tris-HCl, pH 7.9, 10 mM MgC12, 60 mM KCI and 1 mM dithiothreitol were incubated at 4°C for 5-18 hr. Ribosomal RNA was purified by extracting the sample three consecutive times with buffer- equilibrated phenol, precipitating with 2.5 volumes 95% ethanol at - 70°C for 1 hr and pelleting by centrifuga- tion. The rRNA pellets were resuspended in 20pl of 8M urea, 0.025070 bromphenol blue, 0.025070 xylene cyanol and the digestion products analyzed by eleztrophoresis in a 5 070 polyacrylamide gel containing 7M urea, 89 mM Tris-horate, pH 8.3, and 1 mM EDTA.

I I I I q II 12 16 2O

PIe~E]SIJI~ITS ~TIO [1~./1~1.1

The 801-811 region o f 23S rRNA

Two probes complementary to the 803 region were made, one being complementary to bases 801-806 and the other complementary to bases 803-811. These probes were chosen to specifically answer the question as to whether the presenc¢~ of tRNA in the P-site would compete with the probe across the bases 807-809, postulated to be base paired with the 3'-terminal CCA in tRNA [7l. The probes initially were labeled at the 5 -end using [32p] in the man- ner outlined above. Filter-binding assays confirm- ed (Fig. 1) that both probes bound well. The difference in the amount of binding is probably a result of additional sites having partial homology to the longer probe, as was shown by a computer search of the 23S rRNA structure.

Ribonuclease H assays confirmed that the probes were binding the target region, but the amount of clipping was very low, no matter how long the sam- ple was incubated with RNase H. Comparing this with the filter-binding assay results, it appears that the lack of clipping may be entirely due to steric hindrance of the RNase H. Other rRNA fragments were seen as well, confirming the fact that partial- ly homologous regions elsewhere in the rRNA do have the ability to bind the probe sufficiently well to allow RNase H activity.

8 0

a3 .,7"J i,-,.

,...., t,o

ii,I

70

6 0

SO

II0

30

2O

I 0

0

0

Probing rRNA structure

o - - a ~ 1 ~ 1 6 OF ~ 80J-2O6 x - - x I I J l ~ l ~ aF I ' 1 ~ ( 803-~ J

Fig. 1. Filter binding assay of probes complementary to regions 801-806 and 803-811 of 23S rRNA. Increasing amounts of [3:P]-iabeled DNA were incubated in 50/~1 volumes containing 50 pmol of 50S subunits in 10 mM Tris-HCI, pH 7.4, 10 mM MgCI: and 100 mM KCI (binding buffer) at 4~C for 18 hr, Reactions were filtered under vacuum through nitrocellulose filters and washed with 2 × l ml portions of the above buffer. After the filters were dried, the radioactivity retained was monitored with liquid scintillation.

The binding of labeled, deacylated t R N A phe was assayed in the presence and absence of unlabeled probes. In the case of the probe complementary to bases 801-806, tRNA bound to the ribosomes to the same extent, in the presence or absence of the probe (Fig. 2a). However, when the probe complemen- tary to bases 803-811 was present, tRNA binding decreased approximately 50% (Fig. 2b).

The converse experiment was also carried out, wherein the binding of labeled probes was measured in the presence and absence of unlabeled, deacy- lated tRNA. The results show that the binding of the probe complementary to bases 801-806 was not reduced when tRNA was previously bound to the ribosome (Fig. 3a), but in the case of the probe complementary to bases 803-811, a significant decrease in probe binding occurred at low probe concentrations, when tRNA was previously bound to the ribosome (Fig. 3b). As the probe concentra- tion increased, the probe appeared to compete more favorably for the site.

1074 W.E. Hill and A. Tassanakajohn

U

10

e.~ Z

r O

e n

N

(a)

|

Q

go i n

me

O~ i •

I / I I

O . q 0 . 8 l . ~ 1 . 6 ~l.O

T - R N I : I / R I B O S O t ~ R I I T I O (HOL / t IOL1

SD

I10

20

(b)

10

0 0 I Ol.O 0 , 0 1.6

s

s s

s J

/ / /

I ! ! O.q 0 . 8 i . 2

T - R N , q / R J B O S O ~ ( R R T I O ( ~ O L / ~ O L }

a - - o T.-MI~ MITH ~ BOi,...B06 v - - v T-ibqR k lT~ FIMef. 803-811 z - - .~ l- l l~lR 14I'Tit~lJT ~ 801-806 x - - ~ T.-I~I~ tlITtRIIJT f R ~ BO3-BII

Fig. 2. Binding of deacylated tRNA TM tO tight-couple ribosomes in the presence and absence of probe coml~!ementary to regions 801-806 (a) and 803-8110b) of 23S rRNA. Fifty pmol tight-couple ribosomes were incubated with increasing concentration of [32p]_ labeled tRNA ehe in the presence ( ) and absence ( . . . . ) of 800 pmol unlabeled probe, The 50-/zl reaction mixtures contained 50 mM Tris-HCl, pH 7.2, 10 mM MgCI2, 100 mM KCI, I mM DTT, and 5 mg/ml poly(U) (tRNA binding buffer) and were !.ncubated for 2 hr at 4°C. The binding of the tRNA was assayed by filtering over nitrocellulose and counting.

(a) (b) SO DO

S

~ . - ' j SO

3O

~O

H

e.1, z ,,.,j

Ts en

,.w 30

I I I I 0 q R 12 AS

P R O O ( / R I O O $ O N E R A T I O { H O L / M O L I

IS

Of[ N

JlF .

vl /

Y t e ! e tl 8 12 16

P R O g E / R I O O S f l H E R A T I O (H I ' I L / t lOL1

O - o, ~ w l . - e w ; HITH T - I ~ 6 - A ~ 003-011 MIIN T-RNR + - + ~ m ( eas-eo6 , , m ~ T-J~,~n . - 0. ~ e e ( e 0 3 - s ~ ,, 'w.e.~'r 'r-~,,m

Fig. 3. Binding of cDNA probes complementary to regions 801-806(a) and 803-81 lob) of 23S rRNA to tight-couple ribosomes in the presence and absence of deacylated tRNA Phe. Fifty pmol tight-couple ribosomes were incubated with increasing concentrations of P~P]-labeled probe in the presence ( ) and absence ( . . . . ) of 100 pmol unlabeled tRNA. The 50-/zl reaction mixture con- taining tRNA binding buffer (see legend to Fig. 2) was allowed to incubate for 18 hr at 4°C. The binding of eDNA tc the ribosomes were assayed using nitrocellulose filters and counting.

These results show that probe binding can be markedly attenuated when three or four bases in the probe binding site are not available for hybridization. Further studies o f tRNA binding to this region are presently underway.

Bases 1394 and 1395

The 1400 region o f 16S rRNA

It has been shown by Moazed et al. [5] that upon activation of the 30S ribosomal subunit, bases 1394 and 1395 in 16S rRNA become much less reactive to chemical modification compared to their reac- tivity in the active form. The same is true for bases 1505 and 1506. It was suggested that these two sets o f bases become base paired when the 30S subunit is activated.

To see if our probes were able to document such a change, two probes having nine bases were made, one complementary to bases 1393-1401 and the other complementary to 1396-1404. These probes were both labeled and then bound to the inactive ribosomal subunits. The results are shown in Fig. 4a. It is seen that both probes bind equivalently to the inactive form of 30S ribosomal subunit. However, upon subunit activation, the probe corn-

plementary to the bases 1396-1409 showed no reduction in binding, but there is a marked reduc- tion in the binding o f the probe complementary to the bases 1393-1401 region (Fig. 4b). These results suggest that upon activation, bases 1394 and 1395 are no longer available as a hybridization site.

Ribonuclease H assays o f the binding of both probes to both active and inactive forms of 30S subunits were seen. The results showed that the amount o f 145-base fragment produced was the same in both the active and inactive form when the probe complementary to bases 1396-1404 was used. However, similar experiments using the probe com- plementary to bases 1393-1401 showed much less 145-base fragment in the active complex than in the inactive complex. These results corroborate the f'dter-binding assay results above. It would appear that the creation o f the putative base-pairs between 1394-1506 and 1395-1505 hinders the binding of the probe complementary to that region from binding.

(a)

In order to assess the ability o f cDNA probes to test the uniqueness of a single-base interaction at

20

Probing rRNA structure 1075

IO

N 3O

10

l . i I I l I~ 16

PR~BEIINRC'r |~E 30,5 RR~IO [ N O L / ~ . ~

q o 12 16 gO

P~OBF, JRCT|YF" 305 ~RT|O (NOI./HOl.~

liO

$0

qO

a

30

u) INI 2O

I0

Co) JL

0 - 0 INf~c'Yrd[ 30S.-fq~m£ l~ . ,~ iqo i 4 . - 4. f~"Y|V[ 30~#5~ff1~ l ~ 3 - 1qo l

Fig. 4. (a) Filter binding assay of probes complementary to bases 1393-1401 and 1396-1404 of 16S rRNA to inactive 30S subunits. The conditions are identical to those described in the legend to Fig. 1, except that 30S subunits were used, and the reaction mixtures were incubated at 4°C for 5 hr. (1o) Filter binding assay of probes complementary to bases 1393-1401 and 1394-1406 of 16S rRNA to active 30S subunits. The conditions are as described in the legend to Fig. 1, except that 30S subunits were used and activated by heating at 42oc for 20 rain in a buffer containing 10 mM Tris-HCl, pH 7.4, 10 mM MgCI2, and 100 mM KCI, after which increas- ing amounts of lab¢leA probes were added, and the reactions were incubat~ at 4°C for 5 hr.

1076 W.E. Hill and

C1400 of 16S rRNA, probes were made which were complementary to bases 1393-1399 and 1401-1406, respectively Both of these probes were labeled and then increasing amounts of probe were incubated with 30S ribosomal subunits for 5 hr at 4°C. The results of the filter-binding assays are shown in Fig. 5. Both probes are shown to bind well using this assay.

To certify that the probes were binding to the target site, RNase H assays were made when the 30S subunit was incubated with each probe (Fig. 6). The probe complementary to bases 1401-1406 (lane D) gave a single fragment of about 135 bases, but the probe complementary to region 1393-1399 (lane B) gave two bands, one from a fragment with about 145 bases (very faint) and one from a fragment with about 130 bases. Upon analyzing the structure of 16S rRNA, it was noted that the CACAC sequence found at bases 1395-1399 was identical to bases 1407-1411. This five-base homology only occurs in two other places, bases 312-316 and 1226-1230. Both of these regions are highly structured, con-

A. Tassanakajohn

A B C D

-145

-130

I l l

SO

q8

30 z

I 0

0 I I ! a 0 Ii El 12 16 20

PROBE/~J~LqI'i' ~1"1"10 [HOLIHOI.I

o -- o B,IIIOING mr ~ 13a~-13g9

Fig. 5. Filter binding assay of probes complementary to region 1393-1399 and 1401-1406 of 16S rRNA. Increasing amounts of p2p]-labeled DNA were incubated in 50-~1 volumes containing 50 pmol of 30S subunits, 10 mM Tris, pH 7.4, 10 mM MgCI 2 and 100 mM KCI (binding buffer) at 4°C for 5 hr. Reactions were filtered under vacuum through nitrocellulose filters and washed with 2 x 1 ml portions of the above buffer. After the filters were dried, the radioactivity retained was monitored with liquid scintillation.

Fig. 6. Ribonuclease H assay of probes complementary to regions 1393-1399, 1401-1406, and 1393-1401 of 16S rRNA. Twentytzg of DNA probe were incubated in 20-/zl volumes containing 20/~g of 30S subunits, 2-4 units of RNase H, 40 mM Tris, pH 7.9, 10 mM MgCl 2, 60 mM KC1, and I mM dithiothreitol at 4°C for 5 hr. Proteins were removed by buffer-equilibrated phenol extraction and the resulting RNA and RNA fragments were precipitated with 3 volumes of 95°7o EtOH, pelleted and resuspended in 10 M urea, 0.025070 bromophenol blue, and 0.0250/0 xylene cyanol. The RNA was run on a 5070 poly- acrylamide/7 M urea/89 mM Tris-borate (pH 8.3)/2 mM EDTA gel for 3 hr at 12.5 mA and stained with methylene blue.

Lane A contains RNA extracted from 30S subunits incubated with RNase H. Lanes B, C and D contain RNA extracted from 30S subunits incubated with RNase H and probes complemen- tary to bases 1393-1399, 1393-1401, and 1401-1406, respectively.

taining five base pairs, and neither would give the approximately !30 base fragment obtained with RNase H digestion. We assume that the shorter fragment may be due to the probe complementary to bases 1393-1399 binding to bases 1407-1411 as well. A RNase H assay would then give two fragments, one of about 145 bases and one having about 130 bases, which is what we observe.

Another probe was made complementary to bases 1393-1401 in order to completely encompass

Probing rRNA structure 1077

the homologaus CACAC region and to cross over C1400. This probe was shown to bind well using filter-binding assays and to give two fragments similar to those fragments obtained when the probe complementary to bases 1393-1399 was incubated with RNase H (Fig. 6, lane C).

The main object of this portion of the study was to look in detail at the C1400 interaction with tRNA to determine if that base was the only base in this region proximal to tRNA was bound to the P-site. In order to test this, we performed two experiments. In the first experiment (Fig. 7) we incubated ribosomes with increasing amounts of the labeled probes which were complementary to regions 1393-1399 (Fig. 7a), 1401-1406 (Fig. 7b) and 1393-1401 (Fig. 7c) in the presence and absence of bound tRNA. As can be seen, in the case of the two probes which are not complementary to C1400, binding was essentially the same whether tRNA was present or not. However, in the case of the probe complementary to bases 1393-1401, a definite decrease in binding was observed when tRNA was present.

In the second experiment, the labeled probes were complexed with the ribosome and increasing amounts of unlabeled tRNA were added to the sam- ple. The results of this experiment are shown in Fig. 8. In this competition experiment, it is seen that the addition of tRNA does not substantially reduce the binding of either the probe complementary to 1393-1399 or the probe complementary to bases 1401-1406. However, there was definite decrease in the binding of the probe complementary to bases 1393-1401. The results of both of these experiments show that only when a probe crosses base C1400 is there competition with tRNA binding.

Discussion

This study clearly shows that probing the structure of rRNA in situ using small DNA oligomers com- plementary to single-stranded regions can be a tool of high resolution. The results show that the in- teractions between single bases can be observed us- ing this technique. A two-base region involved in base pairing can be readily discriminated as well.

Probing with short eDNA oligomers, as it is be- ing refined in our laboratory, provides excellent corroborative data to the results emanating from various laboratories using chemical modification [4, 17-20], nuclease digestion [21-23] and site- directed mutagenesis [24]. These methods all pro- vide high resolution results, allowing single-base in- teractions to be discriminated as well. Probing using

cDNA oligomers does provide the advantage of modifying the structure irt one region of the rRNA alone (or at least a limited number of regions where partially homologous sequences occur).

Along with this single-base resolution, there is also considerable evidence that the stability of the interactions being probed can be assessed to a degree. In the case of the tRNA interaction at the 807-809 region, our results show that a eDNA probe having nine bases can readily melt the putative base- pairing between tRNA and that site. Conversely, the base-pai~ bonds found in the active form of 16S rRNA at bases 1394 and 1395 are sufficiently strong to withstand melting by a probe having nine bases. This suggests that these bonds may be stabilized elsewhere, since under normal circumstances we should be able to melt a dimer bond readily using a probe of nine bases.

It is also noteworthy that the single-base prox- imity of C1400 to the tRNA cannot be disrupted by a probe containing nine bases. This suggests that tRNA in the P-site is stabilized by binding to other portions of the ribosome, which gives it sufficient stability for the anticodon loop to displace a nine- base eDNA probe. We have found, however, that the tRNA cannot displace a probe having thirteen bases complementary to bases 1393-1406 (Tassa- nakajohn & Hill, unpublished observations).

Ribonuelease H clipping of the regions base- paired by the eDNA probes has been used to cer- tify that the eDNA probes are indeed binding to their target sites. Inasmuch as the RNase H does not clip at a specific base, but rather clips the RNA base-paired with the DNA probe, such clipping leaves fragments with variable ends [6,15]. Nonetheless, RNase H is a powerful technique to certify specific binding of eDNA probes to the putative target sites.

It has been shown in this study that RNase H may be hindered from clipping the rRNA. When the RNase H assay is used on a region showing good eDNA binding, and the RNase H gives poor clipp- ing, it is probable that RNAase H is being sterical- ly hindered from activity. In our work we have noted that when RNase H is used on sites consisting of loops on stems, good clipping often results [6]. However, when the region is such as the 803-811 region of 23S rRNA, there appears to be signifi- cant hindrance to RNase H activity.

It ~ important to note that the RNase H assay is also capable of showing considerable resolution. As noted above, when the probe complementary to bases 1393-1399 was used, two fragments were produced. One had about 145 and the other about 130 bases. Finding a partially-homologous CACAC

1078 W.E. Hill and A. Tassanakajohn

M

| ql

S je

m

m

I0

(a)

, . e " . . . . .

|0 iO

SO

qO

m

I t )

tm

10

01 I I I I 0 I O q 8 12 I§ 20 0

PRO~c./RIDOSOfl( RR'rIO (NOl./NOl.1

(b)

x - - x PJ tmE i l a . - i ~ MITH T - m R (> - - o ~ I N e , - i ' m e Idl'rllaUt T- I t l~

| I I I q 8 12 16 20

I ~ O f l E J R I B f l S O H E R R T I O ( I ' G J t l O L I

a - - a ~ l q O l - i q D S MITH T4J4~ O - - O P l tmE ll~ll-tq~ tll'rlllP.IT l-IINII

(c) 70

60

H

Q

m

w

3o m

114 20

I0

n v 0

s /

_ s

s

I . _ _ I I I tJ It 19 16 2'O

PRWE./I~ISMOHE I ~ T I O (HBL/MOLI

4- ~ 4- ~ | N 3 - 1 q b l i41'n4 5,4NR 0 ~ ~ ~ i ~ [ ~ l q O l 1415llOUT T-IlflR

Fig. 7. Binding of eDNA probes complementary to regions 1393-1399(a), 1401-1406(b), and 1393-1401(e) of 16S rRNA to tight-couple ribosomes in the presence and absence of deacylated tRNA phe. Fifty pmol tight-couple ribosomes were incubated with increasing con- centrations of p2p]-Iabeled probe in the presence ( ) and absence ( . . . . ) of 100 pmol unlabeled tRNA. The 50-~1 reaction mixture containing tRNA binding buffer (see legend to Fig. 2) was allowed to incubate for 5 hr at 4°C. The binding of eDNA to the ribosomes were assayed by filtering through nitrocellulose and counting.

Probing rRNA structure 1079

J

SD

qO

i" N

I0

0 I I f |

0.0 o.q 0.8 1.2 h i s 2.0

"r-RNI~/[~J~SOME I~?I0 (HOI./HOL]

Conclusions

We can conclude that the use of oligomeric probes in probing the structure of rRNA in situ should prove to be most rewarding. Not only can this technique be used to discriminate interactions taking place between single bases, but the stability of such interactions can be assessed as well. Coupl- ed with the judicious use of RNase H assays, oligomer probing should glee considerable insight into the structure and function o f the ribosome.

Acknowledgments

The insightful work of David V~quez has been an in- spiration to these authors. This work was supported in part by NSF Grant No. DMB 8417292 and NIH Grant No. GM 35717.

O, - - O. B n ~ I I G 01 r ~ i193-11~1 A _ ,t BINDING 0f ~ 1 ~ 3 - l q 0 l k - i t PIt*PING OF ~aPf. lqOi-)tt06 References

Fig. 8. Competitive displacement of bound probes complemen- tary to regions 1393-1401, 1393-1399, and 1401-1406 of 16S rRNA by deacylated tRNA Phe. Fifty pmol tight-couple 70S ribosomes were pre-incubated with 1000 pmol of p2F]-Iabeled probes. The reaction was carried out in 50-~1 volumes contain- ing 50 mM Tris-HCI, pH 7.2, l0 mM MgCI 2. 100 mM KC1, l mM DTT and 5 mg/ml poly (U) at 4°C for 5 hr. After the pre-incubation, increasing amounts of unlabeled, deacy- lated tRNA TM were added and the incubation was allowed to proceed for 2 hr at 4°C. Following the incubation, the reaction mixture was filtered through nitrocellulose, the filters dried and the radioactivity retained mori~tored with liquid scin- tillation.

region at 1405-1410 suggests that we may be clip- ping this region as well as the 1393-1399 region. This shows that the RNase H assay can discriminate between two regions which are less than 10 bases removed f rom each other. These two fragments were clearly separable under the conditions in which the gel was run.

I t seems very promising that the RNase H assay could distinguish between fragments with only a few base differences in length. In this way probes to adjacent regions of rRNA could be distinguish- ed by small changes in their RNase H fragment pat- terns. Careful analysis o f such fragments will lead to considerable evidence for the amount o f ex- posure o f certain regions o f the rRNA.

1 Bdmacombe R., Maly P. & Zwieb C. (1983) in: Pro- gress in Nucleic Acid Research and Molecular Biology (Cohn W., ed.), Academic Press, New York, pp. 2-49

2 Brow D.A. & Noller H.F. (1983) .I Mo!. Biol. 163, 27-4.6

3 Van Stolk B.J. & NoUer H.F. (1984) J. Mol. BioL 180, 151-177

4 Noller H.F. (1984) Annu. Rev. Biochem. 53, 119-162

5 Moazed D., Van Stolk B., Douthwaite S. & Noiier H. (1986) J. MoL Biol. 191,483-493

6 Tappr~ch W.E. & Hill W.E. (1986) Proc. Nat. Acad. Sci. USA 83, 556-560

7 Barta A., Steiner G., Brosius J., Noller H.F. & Kuechler E. (1984) Proc. Natl. Acad. Sci. USA 81, 3607-3611

8 Prince J.B., Taylor B.H., Thurlow D.L , Ofengand J. & Zimmermann R. (1982) Proc. Natl. Acad. Sci. USA 79, 5450-5454

9 Gornicki P., Ciesiolka J. & Ofengand J. (1985) Biochemistry 24, 4924-4930

10 Hill W.E., Rossetti G.P. & Van Hoide K.E. (1969) J. Mol. Biol. 44, 263-277

I 1 Tam M.F. & Hill W.E. (1981) Biochem. Interna- tional 3, 655-662

12 NolIM.&NolIH. (1976)J. Mol. Biol. 105, 111-120 13 Zamir A., Miskin R. & Eison D. (1971) J. Mol. Biol.

60, 347-364 14 Chaconas G. & Van de Sande J.H. (1980) Methods

Enzymoi. 65, 75-88 15 Donis-Keller H. (1979) Nucleic Acids Res. 7, 179-192 16 Mankin A.S., Skripkin E.A., Chichkova N.V.,

Koplylov A.M. & Bogdanov A.A. (1981) FEBS Lett. 131,253-256

1080 W.E. Hill and A. Tassanakajohn

17 Peattie D.A. & Gilbert W. (1980)Proc. Natl. Acad. Sci. USA 77, 4679-4682

18 Douthwaite S., Christensen J. & Garett R.A. (1983) J. Mol. Biol. 169, 249-279

19 Meier N. & Wagner R. (1985) Eur. J. Biochem. 146, 83-87

20 Peattie D., Douthwaite S., Garrett R. & Noller H. (1981) Proc. Natl. Acad. Sci. USA 78, 7331-7335

21 Santer M. & Shane S. (1977) J. Bacteriol. 130, 900-910 22 Pace B., Matthews E.A., Johnson K.D.o Cantor

C.R. & Pace N.R. (1982) Proc. Natl. Acad. Sci. USA 79, 36-40

23 Burma D.P. (1982) Current Science 51, 723-726 24 Dahlberg A.E. (1986) in: Structure, Function and

Genetics o f Ribosomes (Hardesty B. & Kramer G., eds.), Springer-Verlag, New York, pp. 686-698