*I. Mol. Rid. (1991) 221. 441-453
Spatial Organization of Template Polynucleotides on the Ribosome Determined by Fluorescence Methods A.V. Bakin’, 0. F. Borisova’, I. N. Shatsky’ and A. A. Bogdanov’ ‘A. N. Belozersky ‘Institute
Laboratory,
of Molecular
(Received
MOSCOWState University,
Biology of Ac,ademy of Sciences Moscow. U.S.S.R.
21 November
1990: accepted 23 April
MOSCOW119899 of U.S.S’.R.
1991)
The spatial
organization of template polynucleotides on the ribosome and the dynamics of t’heir interaction with 30 S subunits have been studied by fluorescence spectroscopy. The topography of the mRNA in the ribosome has been determined using singlet-singlet energy transfer. This method has allowed us to estimate distances between donors and acceptors of energy which have been linked to the terminal residues of template polynucleotides (polyand oligo(l’) and oligo(A)) and 16 8 RNA or to SH-groups of ribosomal proteins 81 and S8. The dynamics of mRNA-ribosome interaction have been investigated by t’he fluorescence st’opped-flow technique. Tt has been shown that the binding to the 30 6 subunit of poly(U) with length much shorter (16 nucleotides) than that’ covered by the ribosome is greatly enhanced by protein Sl. However, the final posiCon of oligo(U),, on the 30 S subunit. which probably includes the ribosomal decoding sit,e. proves to be quite different from that occupied by oligo(‘l’),, on a free protein Sl: Interact’ion of oligo- and poly(l:) with the 30 S subunit occurs in at least t.wo st,eps: the first one is as fast as the interaction of poly( LT) with free Sl, whereas the second step represents a fir&order reaction. Therefore. the second st,ep may reflect, some rearrangement of the t,emplate in the ribosome after its primary binding. It is suggested that protein Sl in some cases may fulfill the role of a transient binding sit,e for mRNA in the course of its interact,ion with the ribosome. The general shape of the t)emplate in the mRNA binding region of the ribosome has been studied using various synthetic ribopolynurleotides and has been shown to be similar. It can be represented by a loop(s) or “Y-turn(s)“. On the hasis of estimation of distances from the ends of poly(U) to some well-localized points on the 30 S ribosomal surface. a tentative model of mRNA path through the ribosome is proposed.
Kryworrls: singlet-singlet ribosomal
energ,y transfer: mRNA topography protein Sl mRN4 binding region
1. Introduction The interaction of the mRNA with the ribosome is a crucial step in protein biosynthesis. The ribosome is known to shield up to 50 nucleotides of the templat’e polynucleotJide (Kang & C:ant’or. 1985). However, apart from Shine-Dalgarno and codonanticodon interactions. which determine primary events in the mRNAPribosome recognition, little is known about the nat,ure of association of t’he t’emplate with the ribosome. In the absence of information on where the mRNA is situated in t,he ribosome and with what components it, interacts. it is impossible to build any hypothesis as to how the t)rmplate moves in the course of polypeptide elongat)ion. Determination of the mRNA topography on a three-dimensional model of the ribosome can be regarded as a first approximation towards this goal. rnt#il now. the study of the mRNA topography has
on the ribosome:
involved t’wo principal experimentma approaches: crosslinking with ribosomal proteins and rRNA and immune electron microscopy (IEM?) (Evstafieva et al.. 1983: Vladimirov et ~1.. 1990: Stiege rt al., 1988). Here, we present data on the topography of t’emplat’e polynucleotides st,udied by fluorescence methods. As a model system. complexes of synthetic polynucleotides with t’he 30 S subunit or the whole 70 S ribosome have been used. LVe have postulated that the mRNA binding sites for synthetic t Abbreviations usrd: rEPor. imm~lnorlt~c,troll microscopy; FTTC’. fluoresc~ein-:5-isothio(.~~t~~t~: FTS(‘. R~tores~ein-Ti-t,hic,semicarhazide: (‘I’M, 7-tlimrthyl~~-(4’-maleirr~itloph~n~l)-l-mrth~l~oum~ri~~: ETW. eosin5_t,hiosrmi~arhazide; EDC’. l-ethyl-:l-(R’-tli~,tt~~larnino propyl) carbodiirnide; (‘DH. carhoh~draei&: B>lE. “-1nrrcaptoethanoI; DMSO. dimeth~lsulfositle: nt. nu&otide(s): SD. ShinePDalgarno.
442
A. V. Nakin
templates and natural mRNAs are the same, at, least for the elongating ribosomr, and the topography of t’he synthetic template faithfully reflects the general path of any message through the ribosome. Earlier, on the basis of IEM data, we have suggested that poly(Y) makes a “l--turn” as it passes through the ribosome. Here, we have checked this hypothesis by another independent physical method, singlet-singlet energy transfer between two fluorophores linked to the 5’ and 3’ ends of various polynucleotide fragments covered by the ribosome. This has made it possible to determine a general shape of the templat,e in t*he rihosome. The results turn out to be in a good agreement with those of IEM. Also. to have an idea where the template is situat’ed in the 30 S subunit, we hare measured the distances from two well-defined points on the 30 S morphological model t’o the 3’ and 5’ ends (“entry” and “exit,” sit,es, respectively) of the poly(U) fragment covered by the ribosome. Much attention has been paid in this paper t’o the topography of poly(U) with respect to the ribosomal protein Sl. This protein is known to greatly stimuribosomes binding to lat)c, the poly( V) (Subramanian. 1984). Sl has also been shown to be important for the interaction of natural ver! of Esrh~wichin COli with ribosomes messages (Subramanian, 1983). Mutations in this protein are let,hal to E. coli and. probably. t,o other Gramnegative bact’eria (Schnirr of al.. 1986). Sl is a typical polynucleot,ide binding prot’ein with a great preference for polypyrimidine templat,es (Draper 8: van Hippel, 1978; Liperky it ad., 1977: Mulsen ef nl.. 1981). The oligo( U) t.racts have recent Iy been found in many natural bacterial mRNAs. upstream from SD sequences. and are suggest,ed t,o play a,n essential role in primarv recognition of the message by the rihosome (Ron’i 8i Isaeva. 198X: Van 1)irijrn it rrl.. 197X). Therefore. it was important to d&ermine which part of the poly(IT) binding site in the 30 S subunit is occupied by Sl and t.o analyze the t,opography of this site. In the first sections of this paper. WV present data on the topography of oligo(lT) on free protein Sl and when the 1at)ter is an integral part of the 30 S subunit). Tn both cases t,hr tjopc)praphy has been determined with respect to welldefinrd points on the Sl molecule (c’ys349 and (‘ys291). The data thus obtained haw a,llowed us to suggest that in some cases the polypyrimidine site of the protein Sl ma)- fulfill the role of a transient site in the course of mRSA binding to the 30 S subunit. This suggest’ion has gained some support from kinetic experiments on oligo( Ll) and pol,v( I-) binding to 30 S subunits pertormed wit’h fluorwwnw stopped-flow method. 2. Materials
the
help
of the
et al.
2 rnn1-F:1)TA. t! rrl~I-d-mercapt~ot~thilrrol (KME). is 20 rnM-Tris. HC”I (pH 7.2). I m~Mg(‘l~. 200 m.n-NH,CI, 2 mM-BME. Buffer IT is 20 mm-Tris.H(‘I MgC’l,.
Buffer
II
(pH 72).
f0 mM-Mg 120 A, respectively). The oligo(U) and poly(A) have been annealed under conditions that allow the formation of duplexes with a molar ratio 1: 1. It can be done by addition of oligo(U) to a large molar excess of poly(A). As expected, the efficiency of energy transfer for the duplex was negligible. ~oly(U),,.poly(A) Experimental data for poly(U),s, 48 to 58 L%have proved to be very close to those calculated on the base of X-ray data (see Table 9). (c) Topography
Table
3
constants of oligo( CT) and oligo(A) 87 protein
Binding
RNA POlY(~),, PolYw,, PolY(N,, PolyLV,,
WY(A),,
9
t Y is the ratio of mol RNA that $ The experimehs were carried mM-Iv@'
can be bound to protein Sl, out in huffer containing
fluorophore position to the values of association parameters. The formation of oligo( U-S1 complexes results in an increase of fluorescence polarization of fluorescein (for the 5’ end from 8 to 2,514: for t)he 3’ end from 11 to 22%), whereas fluorescence intensity drops by 40 y0 and increases by 30 ‘$6 for the 5’. and 3’-terminal labels, respectively. Scatchard plots of the titration data gave the same results for the 5’. and 3’-labeled oligonucleotides. Tt has been found that two molecules of poly( IT),, can be bound to Sl. with association constants of 31 x 106 M--I and 3 x lo6 M-I. The drop of fluorescence intensity has been observed only for the stronger site whose K,,, virtually coincides with that determined earlier by Lipecky ut nl. (1977). These authors determined K,,, by other techniques: inhibition of reaction of N-ethylmaleimide with SHgroups of protein Sl in the presence of polynucleotides or adsorption on nitrocellulose filt’ers. This may explain why they failed to discover the binding of the second molecule of poly(IT) with a weaker affinity to Sl Tn contrast to oligo(U), only one molecule ot oligo(A) can be bound to protein Sl (see Table 3). The position of oligonucleotides on protein Sl has been det’ermined by estimation of distances between each of the ends of Sl-bound oligonucleotide and the Cys residues of protein Sl (see Table 4). The distance between two ends of Sl bound template itself has also been estimated (Table 6). In the former case, CPM and fluorescein were used as a
of polynucleotide binding sites of ribosomal protein 81
To study the topography of RNA binding sites on protein Sl we have used poly( U),,, poly(A),, and whose lengths are comparable to the PO~Y@)I,, number of nucleotides covered by a single RNA binding site of Sl. First of all, it was necessary to find conditions for formation of stoichiometric complexes between oligoribonucleotides and protein Sl for further experiments using energy transfer. The binding of poly(li),, has been studied by changes in fluorescence parameters of dyes linked either to the 5’ or 3’ ends of polynucleotides. This made it possible to estimate the contribution of the
Table 4 Energy transfer and distances between coumarin attached to Cysd92 or (!ys349 S 7 protein and fluorescein-labeled poly( CT)16
Donor site Cys292 Cys349
Acceptor site 5’ 3 5’ 3
P‘d
p* PO)
IB C”)
31 31 22 22
26 28 26 2x
2 0995 0.40 0.54 0.06
E, is efficiency of energy transfer determined fluorescence polarizations of donor and acceptor.
with, an
by quenching
IZ(d/3)/H
R (A)
090-1’12
I19 52*5 4tik.5 74+7
0~91-1~11
of donor
fluorescence.
oj
Pd. Pa arc thts
447
Spatial Organization of Template Polynucleotides
Table 5 Topography of oligo(A)
on the Sl protein
QS292 Oligo length (nt)
cys349
E cff
15 I0
E df
R (A)
5’
.3 ,
079 0.50
@46
‘ 3I
5’ 37+4 48k5
6’ 0.44 057
49&5
R (A) 3 0.40 0.42
5’
3’
51*5 46+5
5J_f5 51*5
Fluorescence polarization of acceptor (fluorescein) attached to 5’ or 3’ terminus of oligo(A) are lli’:, and 180/b. respectively. Energy transfer was estimated from the decrease of the quantum yield of donor fluorescence. An orientation factor KZ w&8 estimated as described by Haas et al. (1978).
Table 6 Distances between 5’ and 3’ terminus of RNA in the complex with Sl protein RNA
E .a
R (4
Pd (%)
pa (%I
Poly(A),,+Sl -Sl
0.207 062
64+5 49+4
16 8
28 17
Poly(U),, +Sl Poly(CJ),,+Sl
@59 0.36t
47+5 57k6
26 26
28 26
NW/R 092-1.08 O-94-1.06
0.89- 1.11 0.89-1.11
t EC,,was determined from the increase in sensitized emission of acceptor and by quenching of donor fluorescence. The donor and the acceptor of energy are fluorescein and eosin. respectively. donor and acceptor of energy, respectively, whereas in the latter case, the donor-acceptor pair was fluorescein and eosin. The coumarin derivative was attached either to Cys292 or Cys349 of protein Sl as described in Materials and Methods. The topography of p01y(U),~ on the protein Sl has been studied only for the stronger site of Sl since we failed to observe any change in energy transfer from SH-groups of protein Sl to the ends of when both sites of the protein were filled POlYWI,, with the oligonucleotide. Similar experiments carried out with poly(A),, and poly(A),, have resulted in the distances given in Table 5. On the basis of these data, positions of oligo(U) and oligo(A) on the putative rod-like part of protein Sl (Odom et al., 1984) can be represented as shown in Figure 2. Here, the poly(U),, is located almost entirely in the domain RI of the protein and oriented in such a way that the 5’ end points towards the C terminus of the protein, whereas the 3’ end is situated closer to the N-terminal domain of Sl. The 5’ end of the oligonucleotide is located in close proximity of Cys292. The orientation of oligo(A) should be regarded as more preliminary. Nevertheless, it seems to be different from that for pol~(U),~, though both oligonucleotides bind within
5’P.2
Rl I N 01
R3 I
3’>
_ys2z2
As seen from the Scatchard plot shown in Figure 3, up to three molecules of p01y(U),~ can be bound by one 30 S subunit, whereas there is only one
I C
0.00
I
3’
oligo(U) \ oligo(A)
(d) Oligo( U) binding sites on the 30 S subunit
R4 1
cys349
I
Figure 2. Scheme
the same region of protein Sl. In agreement with these data, crosslinking experiments with the use of oligo(U) and oligo(A) carrying photoafinity labels at their 5’ or 3’ ends reveal that both oligonucleotides bind within the N-terminal half of protein Sl (our unpublished results).
of the arrangement on the Sl protein.
;
oligo(A)
of oligo(U)
I and
I.00
,
2.00
3.00
Figure 3. Kinetics of poly( LJ) binding to 30 S subunits and protein Sl. The stopped-flow experiments were Performed by rapidly mixing equal volumes of the oligo(U) solution (final concn is @08pM) with 30 S subunits (@8 PM) or protein Sl (@8 P(M).
A. V. Hakin et, al.
448
Table 7 Associated
constants qf poEy( U)
with
30 I)’subunits
RNA powv,6
Polyu(U),,-Sl
Poly(lV,,
Table 8
Distalbces between CysB92 or Cys349 qf Sl protein and 5’ or poEy(U) in the complex with 30 S subunits
3’ twminus of
Polg(lJ) length W)
binding site for p01y(U),~ (see Table 7). The most plausible explanation of these results is (1) that longer polynucleotides are capable of forming more stable contacts with the ribosome per unit’ length than shorter templates, and (2) the longer polynucleotides interact with the greater part of all binding sites of poly( U),,. If so. the binding of two molecules of poly(U) 16 is not equivalent to that of p~ly(U)~~. The latter may occlude a larger part of the mRNA binding site of the ribosome. It is per& nent that the poly(U),, is nearly equal t,o the minimal size of the poly(U) fragment covered by the ribosome (Kang & Cantor, 1985). The binding of both poly(C),, and poly(lJ),, is strongly promoted by the ribosomal protein Sl Omission of Sl from 30 S subunits leads to a disappearance of t’he oligo(l!)-binding sites character,istic of the initial ribosomal subunits. Tn the absence of Sl, only one type of complex with a low h’,,, can be titrated (see Table 7) in agreement with other data (Katunin et al.. 1980). Addition of Sl restores the binding sites of native 30 S subunits.
out whether the position of poly(U),,. with respect to defined points on protein Sl. changes when the protein becomes an integral part of the ribosome. For this, we have measured the distances between 8H-groups of protein Sl and the 5’ or 3’ ends of bound oligo(li). Both these distances (see Tables 4 and 8) and the separation of the ends of the bound oligonucleotide (Tables 6 and 9) are quite different for the complexes of polp( CT),, with free Sl and 30 S subunits. Whereas in the binary complex. the 5’ end of the templat,r is in close proximity to Cys292 (o that for oligo(U) binding to RI-RI1 domains of free Sl has been found on the 30 S subunit and can be filled on increasing the concentration of poly(U),,. The most plausible explanation of these findings is t’hat the main polypyrimidine binding domain of Sl, in some cases, can fulfill the role of a transient site in the course of mRNA interaction with the 30 S subunit: the template may transiently bind within RI-RI1 domains and t,hen “travels” to a final position on the ribosome. It may still include some part’
Spatial
Organization
of Template
of protein Sl , too, thus accounting for its stronger affinity for oligo(U) as compared to 30 S subunits lacking this protein. A similar function for SD sequence has recently been proposed by Gualerzi and his co-workers (Gualerzi et al., 1988). To gain some support for this hypothesis, we have studied kinetics of poly(U) i6 binding with free protein Sl and 30 S subunits using the fluorescence stoppedflow technique. Indeed, the kinetics have turned out to be complex. The first step of oligo(U) binding to both the 30 S subunit and free Sl have been found to last less than five milliseconds. However, in the case of 30 S subunits, this step is followed by some rearrangement of the message, which is not observed for Sl-oligo(U) complex. We realize that these data are open to other interpretations as well. When bound to the ribosome, the template is not extended into “a straight line” as is depicted in many textbooks. Rather, it forms a loop or “U-turn” as it passes through the ribosome with the separation between the ends about 45 A. It is of interest that this value changes very little for poly(U) lengths in the range of 18 to 35 nucleotide residues, as if the template goes around some convex surface. In contrast, for a linear form of poly(U),,, for example, this separation would be at least 90 A if the bound template has fully stacked bases (3 A x 30) (Bloomfield et al., 1974). In reality, it must be much larger since, with the exception of two mRNA triplets interacting with tRNA anticodons, other nucleotide residues of the template do not’ appear to be in a stacked conformation. This follows circular dichroism of poly(thioU) bound to the 30 S subunit, which does not reveal any stacking of nucleotide bases of the template (A. V. Bakin et al., unpublished results). The “U-turn” represents a general shape of the message. It does not exclude other bends or turns within the segment of mRNA covered by the ribosome. One of these bends most certainly exists between two codons of mRNA in A and P sites of the ribosome when they interact with anticodons of tRNAs (Rich, 1974). However, the general shape of the message does not depend on the presence or absence of tRNAs (see Results, section (g)) and, hence, is determined by the ribosome itself. The distances between the ends of a poly(U) fragment covered by the ribosome and the 5’ end of 16 S rRNA or prot,ein S8 have been estimated in order to gain more information on the location of the mRNA binding site on the 30 S subunit. The 5’ end of the 16 S rRNA and protein S8 are wellmapped points on the morphological model of the 30 S subunit (Lake, 1979; Mochalova et al., 1982). Their positions are indicated in Figure 5. The set of measured distances thus estimated is not enough to determine the exact position of the mRNA binding site. However, some important observations can be drawn from these experiments: (1) the 3’ end of the message is much closer to the 5’ terminus of 16 S rRNA than the 5’ end; (2) as the length of poly(U) increases, the 3’ end tends to approach to the 5’ end of 16 S rRNA; (3) the distance between the 3’ end of
Polynucleotides
30s
451
30s
Figure 5. Schematic presentation of the 30 S subunit in 2 projections. The position of the mRiVA binding site is shown by the hatched area. The black half-circle and arrow indicate the position of the 5’end and 3’end of 16 S RNA, respectively.
poly(U) and protein S8 decreases with increasing of the polymer length, while the distance between the 5’ end of RNA and protein S8 does not change significantly. These data combined with those obtained by IEM (Evstafieva et al., 1983) have allowed us to determine the orientation of the message and propose a tentative model of the mRNA path on the 30 S subunit. As shown in Figure 5, the 5’ end of the template is located near the 3’ end of 16 S rRNA, in agreement with our IEM data on the poly(U) 5’ end location (Evstafieva et al., 1983). Then t,he template descends along the 30 S “cleft” or “groove”, which is thought to be the site of SD (Oakes et al., 1986a) and the codon-anticodon interaction (Prince et al., 1982), thus reaching the “neck” of the subunit (the lower part of the groove). Such an orientation of the message is similar to that suggested by Olson et al. (1988) and seems to contradict that proposed by Lake (Oakes et al., 19866). The further path of mRNA is less evident. To form a loop or U-turn, there are two possibilities: (1) the message comes back along the groove (or near it) to its 5’ end; (2) the message continues its way around the neck of the 30 S subunit and quits the latter somewhere on its cytoplasmic side. In both cases the template may form on its way other bends or turns. Indeed, according to our experiments on affinity labeling of 16 S rRNA with the templates carrying photoactivated probes, the mRNA binding site may also include some parts of 16 S RNA located in the head of the 30 S subunit (A. V. Bakin et al., unpublished results). For many reasons, we prefer the latter model, which is shown in Figure 5 and will be discussed in detail elsewhere (I. N. Shatsky et al., unpublished results). In particular, it is in better agreement with the positions of the 5’ and 3’ ends of poly(U) with respect to protein S8 and the 16 S RNA 5’ end determined in this paper as well as with our IEM data on the position of the 3’ end of the template on the 70 S ribosome (Evstafieva et al., 1983). This model is also compatible with the position of the
452
A. V. Bakin
terminal
nucleotide
to SH-groups Results, latter IEM
section on
the
residues
of the template
of the ribosome-bound (e))
and
30 S subunit
(Stoffler-Meilicke,
with
the
recently
personal
relative
protein
Rl
(see
location
of
the
determined
by
communication).
References Bakin. A. V. & Shatskv. I. N. (1989). Selective fluorescent labelling of the RkA ends to study its macromolecule structure in the solution. BiopoliwL. Kletka (CJSSR). 5(3), 71-76. Bloomfield, V. A.. Crothers, D. M. CI:Tinoco, I. (1974). Tn Physical Chemistry of Nucleic Acids. p. 128, Harper & Row, New York. Boni, I. V. & Isaeva. D. M. (1988). Localization of RNA sites of Qfi. fr and MS2 bacteriophages which are involved in interactions with ribosomal protein RI during the 30 S*mRNA complex formation. Dokl. Akad. Nauk S&S%. 298(4), 1015-1018. Borisova. 0. F.. Kichevsakaya, A. A., Posvirin, V. V. & Samarina, 0. P. (1979). Nuclear ribonucleoproteins containing pro-mRNA. Investigation of their structure using ethidium-bromide and fluorescamin. Mol. Biol. (USSR), 13, 422-436. Dale, R. F. 8: Eisinger, J. (1975). Polarized excitation Concepts energy transfer. In Biochemical Fluorescence (Chen, R. F. & Edelhoch, H., eds). pp. 115-224. Marcel Dekker Inc., New York. Draper, D. E. & von Hippel, R. H. (1978). Nucleic acid binding properties of Escherichia coli ribosomal protein Sl. J. Mol. Biol. 122, 321-359. Evstafieva, A. G., Shatsky. T. N.. Bogdanov. A. A.. Sedmenkov. Y. P. & Vasiliev, V. D. (1983). Localization of 5’ and 3’ ends of the ribosome-bound segment of template polynucleotides by immune electron microscopy. EMBO J. 2. 799-804. Farclough. R. H. & Cantor, C. R. (1978). Singlet-singlet 48, 347-379. energy transfer. Methods &zymol. Forster. Th. (1966). In Modern Quantum Chemistry (Sinanoglu. 0.. ed.). vol. 3. pp. 93-137, Academic Press. New York. (Zorginis, S. bz Subramanian. A. R. (1980). The major ribosome binding sit.e of Escherichin coli ribosomal protein Sl is located in its N-terminal segment. J. Mol. Biol. 141. 393-408. Gottich, M. B.. Ivanovskaya. M. 0. & Shabarova, %. A. (1983). Synthesis of phosphamidmono and oligodeoxynucleotides in water medium. Bioorgan. (Ihem. (VSSR), 9. 1063-1067. Ciualerzi. C. 0.. C’alogero, R. A.. Canonaco. M. 8.. Brombach, M. &, Pon, C. L. (1988). Selection of mRNA by ribosomr during prokariotic translational init,iation. In Grnetics of Translation (Tuite, M. F., eds) , Picard. M. Pr Bolotin-Fukuhara. M.. pp. 317-330, Springer-Verlag. Berlin. Haas. E.. Katchalski-Katzir. E. 6 Steinberg, T. %. (1978). Effect of the orientation of donor and acceptor on t,he probability of energy t,ransfer involving electronic Biochemistry, 17. transitions of mixed polarization. 5064-5070. Kang, C. & Cantor, (‘. R. (1985). Structure of ribosomrby enzymatic bound messenger RNA as revealed accessibility studies. J. Mol. Biol. 181, 241-251. Katunin. V. I.. Semenkov. Yu. P.. Makhno. V. I. & Kirillov. S. V. (1980). Comparativr study of the int,eraction of polgucidilic acid with 30 S suhunit,s and 70 S rihosomes of E. coli. Nucl. drids i&s. 8. 403.-42 1.
et al.
Lake. ,J. A. (1979). Rihosome structure and functional Function and Genetics sites. Tn Ribosomes, Structure, (Chambliss et al.. eds), pp. 201-236. I’niversity Park Press. Baltimore. MD. Laemmli, LT. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London), 227, 680-683. Lipecky. R.. Kohlschein. .J. & Gassen. H. (1. (1977). Complex formation between rihosomal protein Sl oligo- and polynucleotides: chain length dependence and base specificity. ilrucl. Acids Res. 4. 3627-3642. Melhuish. W. H. (1961). Quantum efficiencies of fluorescence of organic substances: effect of solvent and concentration of the fluorescent solute. .J. Phys. C’hem. 65. 229-235. Mochalova. L. N.. Shatsky* I. N’.. Bogdanov, A. A. & Vasiliev, V. I). (1982). Topography of RNA in t,he ribosome: localization of the 16 S 5’ end by immune electron microscopy. J. Mol. Biol. 52. 637-650. Mulsen. A.. (‘olpan. M., Wollny. E.. Gassen. H. (:. & Riesner. I>. (1981). Mechanism of the interaction between rihosomal protein Sl and olieonucleotides. IYucl. ilcids Res. 9. 2367-2386. Oakes, M.. Clark, M.. Henderson. E. & Lakts. J. A. (1986a). DNA hybridization electron microscop?: ribosomal RNA nucleotides 1392-1407 are exposed m the cleft of the small subunit. Proc. Sat. .-lend. Sri.. ~:.s.d. 83 .I*“75-279. I Oakes. MM., Henderson. E.. Schienman. A.. (‘lark. M. Hr Lake, .I. A. (1986b). Rihosome structure. function and evolution: mapping rihosomal RNA. prot,rins. and functional sites in three dimensions. In Str~ct~rr. Function. a.nd Genetics of Ribosomes (Hardest., K. & Kramer. (i., eds), pp. 49-69. Springer-Verlag. Berlirl. Odom, 0. W.. Deng. H.-Y.. Subramanian, A. R. & Hardesty. B. (1984). Relaxation time. internal distance, mechanism of action of rihosomal protein Sl ilrch. BiochewL. Biophys. 230, 178-I 03. Olson, H. K.. Lasater, I,. S.. Cann, P. 21. & Blitz. I). (:. (1988). Messenger RNA orientation on t.hr rihoxomr. J. Hiol. ('hem. 263, 15196-15204. Prince. .J. B.. Taylor. B. H.. Thrulov. 11. I,.. Ofengand. .I. &. Zimmerman. R. A. (1982). Covalent crosslinking of tRNA\(a’ to 16 S RNA at t,he rihosomal 1’ sit.r: identification of crosslinked residues. Proc. ,Vnt. Acad. Sri.. I’.S.A. 79, 5450-5454. Rich, A. (1974). How tRNA may movt insitir the ribosome. Tn RibosvrnPs (NoAura. vl nl.. eds). pp. X71-884. (~!old Spring Harbor Lahorator,v Press. C’old Spring Harbor, NY. Schnier. ,I.. St,offler. (4. & Xshi. K. (l!%S). Deletion and insertion mutants in the structural genes for rihosomal protein Sl from Escherichia co/i. J. Rio!. (‘haunt.. 261. (“5). 11866~11871. Stiege. W.. Stade. K.. Srhulrr. I). & Brimac*ornhr. R (1988). (‘ovalent cross-linking of poly(A) to rihosomes and loc.alizat,ion of the cross-link site within the 16 S RNA. ili~cl. dcids 12~s. 16, 2369-238X. Suhramanian. A. R.. (1983). Structure and functions ot’ ribosomal prot,rin Sl. I’roy. Xucl. _4fids Hrs. .lfol. Biol. 28. 101- 142. Suhramanian, A. R.. (1984). Struct,ure and functions of rihosomal protein Sl. Trends Hiochrw. ,Sci. 9. 191-4!44. Suhramanian. A. R.. Rienhardt. P.. Kitnura. %I. & Surpanarayana. T. (1981). Fragments of ribosotnal protein Sl and it,s mutant form ml-S1 Rtrr. .I. Biochum. 119. 245-249. Survanaravana. T. & Subramanian. :I. I
Spatial Organization of Template Polynucleotides on the Ribosome Determined by Fluorescence Methods A.V. Bakin’, 0. F. Borisova’, I. N. Shatsky’ and A. A. Bogdanov’ ‘A. N. Belozersky ‘Institute
Laboratory,
of Molecular
(Received
MOSCOWState University,
Biology of Ac,ademy of Sciences Moscow. U.S.S.R.
21 November
1990: accepted 23 April
MOSCOW119899 of U.S.S’.R.
1991)
The spatial
organization of template polynucleotides on the ribosome and the dynamics of t’heir interaction with 30 S subunits have been studied by fluorescence spectroscopy. The topography of the mRNA in the ribosome has been determined using singlet-singlet energy transfer. This method has allowed us to estimate distances between donors and acceptors of energy which have been linked to the terminal residues of template polynucleotides (polyand oligo(l’) and oligo(A)) and 16 8 RNA or to SH-groups of ribosomal proteins 81 and S8. The dynamics of mRNA-ribosome interaction have been investigated by t’he fluorescence st’opped-flow technique. Tt has been shown that the binding to the 30 6 subunit of poly(U) with length much shorter (16 nucleotides) than that’ covered by the ribosome is greatly enhanced by protein Sl. However, the final posiCon of oligo(U),, on the 30 S subunit. which probably includes the ribosomal decoding sit,e. proves to be quite different from that occupied by oligo(‘l’),, on a free protein Sl: Interact’ion of oligo- and poly(l:) with the 30 S subunit occurs in at least t.wo st,eps: the first one is as fast as the interaction of poly( LT) with free Sl, whereas the second step represents a fir&order reaction. Therefore. the second st,ep may reflect, some rearrangement of the t,emplate in the ribosome after its primary binding. It is suggested that protein Sl in some cases may fulfill the role of a transient binding sit,e for mRNA in the course of its interact,ion with the ribosome. The general shape of the t)emplate in the mRNA binding region of the ribosome has been studied using various synthetic ribopolynurleotides and has been shown to be similar. It can be represented by a loop(s) or “Y-turn(s)“. On the hasis of estimation of distances from the ends of poly(U) to some well-localized points on the 30 S ribosomal surface. a tentative model of mRNA path through the ribosome is proposed.
Kryworrls: singlet-singlet ribosomal
energ,y transfer: mRNA topography protein Sl mRN4 binding region
1. Introduction The interaction of the mRNA with the ribosome is a crucial step in protein biosynthesis. The ribosome is known to shield up to 50 nucleotides of the templat’e polynucleotJide (Kang & C:ant’or. 1985). However, apart from Shine-Dalgarno and codonanticodon interactions. which determine primary events in the mRNAPribosome recognition, little is known about the nat,ure of association of t’he t’emplate with the ribosome. In the absence of information on where the mRNA is situated in t,he ribosome and with what components it, interacts. it is impossible to build any hypothesis as to how the t)rmplate moves in the course of polypeptide elongat)ion. Determination of the mRNA topography on a three-dimensional model of the ribosome can be regarded as a first approximation towards this goal. rnt#il now. the study of the mRNA topography has
on the ribosome:
involved t’wo principal experimentma approaches: crosslinking with ribosomal proteins and rRNA and immune electron microscopy (IEM?) (Evstafieva et al.. 1983: Vladimirov et ~1.. 1990: Stiege rt al., 1988). Here, we present data on the topography of t’emplat’e polynucleotides st,udied by fluorescence methods. As a model system. complexes of synthetic polynucleotides with t’he 30 S subunit or the whole 70 S ribosome have been used. LVe have postulated that the mRNA binding sites for synthetic t Abbreviations usrd: rEPor. imm~lnorlt~c,troll microscopy; FTTC’. fluoresc~ein-:5-isothio(.~~t~~t~: FTS(‘. R~tores~ein-Ti-t,hic,semicarhazide: (‘I’M, 7-tlimrthyl~~-(4’-maleirr~itloph~n~l)-l-mrth~l~oum~ri~~: ETW. eosin5_t,hiosrmi~arhazide; EDC’. l-ethyl-:l-(R’-tli~,tt~~larnino propyl) carbodiirnide; (‘DH. carhoh~draei&: B>lE. “-1nrrcaptoethanoI; DMSO. dimeth~lsulfositle: nt. nu&otide(s): SD. ShinePDalgarno.
442
A. V. Nakin
templates and natural mRNAs are the same, at, least for the elongating ribosomr, and the topography of t’he synthetic template faithfully reflects the general path of any message through the ribosome. Earlier, on the basis of IEM data, we have suggested that poly(Y) makes a “l--turn” as it passes through the ribosome. Here, we have checked this hypothesis by another independent physical method, singlet-singlet energy transfer between two fluorophores linked to the 5’ and 3’ ends of various polynucleotide fragments covered by the ribosome. This has made it possible to determine a general shape of the templat,e in t*he rihosome. The results turn out to be in a good agreement with those of IEM. Also. to have an idea where the template is situat’ed in the 30 S subunit, we hare measured the distances from two well-defined points on the 30 S morphological model t’o the 3’ and 5’ ends (“entry” and “exit,” sit,es, respectively) of the poly(U) fragment covered by the ribosome. Much attention has been paid in this paper t’o the topography of poly(U) with respect to the ribosomal protein Sl. This protein is known to greatly stimuribosomes binding to lat)c, the poly( V) (Subramanian. 1984). Sl has also been shown to be important for the interaction of natural ver! of Esrh~wichin COli with ribosomes messages (Subramanian, 1983). Mutations in this protein are let,hal to E. coli and. probably. t,o other Gramnegative bact’eria (Schnirr of al.. 1986). Sl is a typical polynucleot,ide binding prot’ein with a great preference for polypyrimidine templat,es (Draper 8: van Hippel, 1978; Liperky it ad., 1977: Mulsen ef nl.. 1981). The oligo( U) t.racts have recent Iy been found in many natural bacterial mRNAs. upstream from SD sequences. and are suggest,ed t,o play a,n essential role in primarv recognition of the message by the rihosome (Ron’i 8i Isaeva. 198X: Van 1)irijrn it rrl.. 197X). Therefore. it was important to d&ermine which part of the poly(IT) binding site in the 30 S subunit is occupied by Sl and t.o analyze the t,opography of this site. In the first sections of this paper. WV present data on the topography of oligo(lT) on free protein Sl and when the 1at)ter is an integral part of the 30 S subunit). Tn both cases t,hr tjopc)praphy has been determined with respect to welldefinrd points on the Sl molecule (c’ys349 and (‘ys291). The data thus obtained haw a,llowed us to suggest that in some cases the polypyrimidine site of the protein Sl ma)- fulfill the role of a transient site in the course of mRSA binding to the 30 S subunit. This suggest’ion has gained some support from kinetic experiments on oligo( Ll) and pol,v( I-) binding to 30 S subunits pertormed wit’h fluorwwnw stopped-flow method. 2. Materials
the
help
of the
et al.
2 rnn1-F:1)TA. t! rrl~I-d-mercapt~ot~thilrrol (KME). is 20 rnM-Tris. HC”I (pH 7.2). I m~Mg(‘l~. 200 m.n-NH,CI, 2 mM-BME. Buffer IT is 20 mm-Tris.H(‘I MgC’l,.
Buffer
II
(pH 72).
f0 mM-Mg 120 A, respectively). The oligo(U) and poly(A) have been annealed under conditions that allow the formation of duplexes with a molar ratio 1: 1. It can be done by addition of oligo(U) to a large molar excess of poly(A). As expected, the efficiency of energy transfer for the duplex was negligible. ~oly(U),,.poly(A) Experimental data for poly(U),s, 48 to 58 L%have proved to be very close to those calculated on the base of X-ray data (see Table 9). (c) Topography
Table
3
constants of oligo( CT) and oligo(A) 87 protein
Binding
RNA POlY(~),, PolYw,, PolY(N,, PolyLV,,
WY(A),,
9
t Y is the ratio of mol RNA that $ The experimehs were carried mM-Iv@'
can be bound to protein Sl, out in huffer containing
fluorophore position to the values of association parameters. The formation of oligo( U-S1 complexes results in an increase of fluorescence polarization of fluorescein (for the 5’ end from 8 to 2,514: for t)he 3’ end from 11 to 22%), whereas fluorescence intensity drops by 40 y0 and increases by 30 ‘$6 for the 5’. and 3’-terminal labels, respectively. Scatchard plots of the titration data gave the same results for the 5’. and 3’-labeled oligonucleotides. Tt has been found that two molecules of poly( IT),, can be bound to Sl. with association constants of 31 x 106 M--I and 3 x lo6 M-I. The drop of fluorescence intensity has been observed only for the stronger site whose K,,, virtually coincides with that determined earlier by Lipecky ut nl. (1977). These authors determined K,,, by other techniques: inhibition of reaction of N-ethylmaleimide with SHgroups of protein Sl in the presence of polynucleotides or adsorption on nitrocellulose filt’ers. This may explain why they failed to discover the binding of the second molecule of poly(IT) with a weaker affinity to Sl Tn contrast to oligo(U), only one molecule ot oligo(A) can be bound to protein Sl (see Table 3). The position of oligonucleotides on protein Sl has been det’ermined by estimation of distances between each of the ends of Sl-bound oligonucleotide and the Cys residues of protein Sl (see Table 4). The distance between two ends of Sl bound template itself has also been estimated (Table 6). In the former case, CPM and fluorescein were used as a
of polynucleotide binding sites of ribosomal protein 81
To study the topography of RNA binding sites on protein Sl we have used poly( U),,, poly(A),, and whose lengths are comparable to the PO~Y@)I,, number of nucleotides covered by a single RNA binding site of Sl. First of all, it was necessary to find conditions for formation of stoichiometric complexes between oligoribonucleotides and protein Sl for further experiments using energy transfer. The binding of poly(li),, has been studied by changes in fluorescence parameters of dyes linked either to the 5’ or 3’ ends of polynucleotides. This made it possible to estimate the contribution of the
Table 4 Energy transfer and distances between coumarin attached to Cysd92 or (!ys349 S 7 protein and fluorescein-labeled poly( CT)16
Donor site Cys292 Cys349
Acceptor site 5’ 3 5’ 3
P‘d
p* PO)
IB C”)
31 31 22 22
26 28 26 2x
2 0995 0.40 0.54 0.06
E, is efficiency of energy transfer determined fluorescence polarizations of donor and acceptor.
with, an
by quenching
IZ(d/3)/H
R (A)
090-1’12
I19 52*5 4tik.5 74+7
0~91-1~11
of donor
fluorescence.
oj
Pd. Pa arc thts
447
Spatial Organization of Template Polynucleotides
Table 5 Topography of oligo(A)
on the Sl protein
QS292 Oligo length (nt)
cys349
E cff
15 I0
E df
R (A)
5’
.3 ,
079 0.50
@46
‘ 3I
5’ 37+4 48k5
6’ 0.44 057
49&5
R (A) 3 0.40 0.42
5’
3’
51*5 46+5
5J_f5 51*5
Fluorescence polarization of acceptor (fluorescein) attached to 5’ or 3’ terminus of oligo(A) are lli’:, and 180/b. respectively. Energy transfer was estimated from the decrease of the quantum yield of donor fluorescence. An orientation factor KZ w&8 estimated as described by Haas et al. (1978).
Table 6 Distances between 5’ and 3’ terminus of RNA in the complex with Sl protein RNA
E .a
R (4
Pd (%)
pa (%I
Poly(A),,+Sl -Sl
0.207 062
64+5 49+4
16 8
28 17
Poly(U),, +Sl Poly(CJ),,+Sl
@59 0.36t
47+5 57k6
26 26
28 26
NW/R 092-1.08 O-94-1.06
0.89- 1.11 0.89-1.11
t EC,,was determined from the increase in sensitized emission of acceptor and by quenching of donor fluorescence. The donor and the acceptor of energy are fluorescein and eosin. respectively. donor and acceptor of energy, respectively, whereas in the latter case, the donor-acceptor pair was fluorescein and eosin. The coumarin derivative was attached either to Cys292 or Cys349 of protein Sl as described in Materials and Methods. The topography of p01y(U),~ on the protein Sl has been studied only for the stronger site of Sl since we failed to observe any change in energy transfer from SH-groups of protein Sl to the ends of when both sites of the protein were filled POlYWI,, with the oligonucleotide. Similar experiments carried out with poly(A),, and poly(A),, have resulted in the distances given in Table 5. On the basis of these data, positions of oligo(U) and oligo(A) on the putative rod-like part of protein Sl (Odom et al., 1984) can be represented as shown in Figure 2. Here, the poly(U),, is located almost entirely in the domain RI of the protein and oriented in such a way that the 5’ end points towards the C terminus of the protein, whereas the 3’ end is situated closer to the N-terminal domain of Sl. The 5’ end of the oligonucleotide is located in close proximity of Cys292. The orientation of oligo(A) should be regarded as more preliminary. Nevertheless, it seems to be different from that for pol~(U),~, though both oligonucleotides bind within
5’P.2
Rl I N 01
R3 I
3’>
_ys2z2
As seen from the Scatchard plot shown in Figure 3, up to three molecules of p01y(U),~ can be bound by one 30 S subunit, whereas there is only one
I C
0.00
I
3’
oligo(U) \ oligo(A)
(d) Oligo( U) binding sites on the 30 S subunit
R4 1
cys349
I
Figure 2. Scheme
the same region of protein Sl. In agreement with these data, crosslinking experiments with the use of oligo(U) and oligo(A) carrying photoafinity labels at their 5’ or 3’ ends reveal that both oligonucleotides bind within the N-terminal half of protein Sl (our unpublished results).
of the arrangement on the Sl protein.
;
oligo(A)
of oligo(U)
I and
I.00
,
2.00
3.00
Figure 3. Kinetics of poly( LJ) binding to 30 S subunits and protein Sl. The stopped-flow experiments were Performed by rapidly mixing equal volumes of the oligo(U) solution (final concn is @08pM) with 30 S subunits (@8 PM) or protein Sl (@8 P(M).
A. V. Hakin et, al.
448
Table 7 Associated
constants qf poEy( U)
with
30 I)’subunits
RNA powv,6
Polyu(U),,-Sl
Poly(lV,,
Table 8
Distalbces between CysB92 or Cys349 qf Sl protein and 5’ or poEy(U) in the complex with 30 S subunits
3’ twminus of
Polg(lJ) length W)
binding site for p01y(U),~ (see Table 7). The most plausible explanation of these results is (1) that longer polynucleotides are capable of forming more stable contacts with the ribosome per unit’ length than shorter templates, and (2) the longer polynucleotides interact with the greater part of all binding sites of poly( U),,. If so. the binding of two molecules of poly(U) 16 is not equivalent to that of p~ly(U)~~. The latter may occlude a larger part of the mRNA binding site of the ribosome. It is per& nent that the poly(U),, is nearly equal t,o the minimal size of the poly(U) fragment covered by the ribosome (Kang & Cantor, 1985). The binding of both poly(C),, and poly(lJ),, is strongly promoted by the ribosomal protein Sl Omission of Sl from 30 S subunits leads to a disappearance of t’he oligo(l!)-binding sites character,istic of the initial ribosomal subunits. Tn the absence of Sl, only one type of complex with a low h’,,, can be titrated (see Table 7) in agreement with other data (Katunin et al.. 1980). Addition of Sl restores the binding sites of native 30 S subunits.
out whether the position of poly(U),,. with respect to defined points on protein Sl. changes when the protein becomes an integral part of the ribosome. For this, we have measured the distances between 8H-groups of protein Sl and the 5’ or 3’ ends of bound oligo(li). Both these distances (see Tables 4 and 8) and the separation of the ends of the bound oligonucleotide (Tables 6 and 9) are quite different for the complexes of polp( CT),, with free Sl and 30 S subunits. Whereas in the binary complex. the 5’ end of the templat,r is in close proximity to Cys292 (o that for oligo(U) binding to RI-RI1 domains of free Sl has been found on the 30 S subunit and can be filled on increasing the concentration of poly(U),,. The most plausible explanation of these findings is t’hat the main polypyrimidine binding domain of Sl, in some cases, can fulfill the role of a transient site in the course of mRNA interaction with the 30 S subunit: the template may transiently bind within RI-RI1 domains and t,hen “travels” to a final position on the ribosome. It may still include some part’
Spatial
Organization
of Template
of protein Sl , too, thus accounting for its stronger affinity for oligo(U) as compared to 30 S subunits lacking this protein. A similar function for SD sequence has recently been proposed by Gualerzi and his co-workers (Gualerzi et al., 1988). To gain some support for this hypothesis, we have studied kinetics of poly(U) i6 binding with free protein Sl and 30 S subunits using the fluorescence stoppedflow technique. Indeed, the kinetics have turned out to be complex. The first step of oligo(U) binding to both the 30 S subunit and free Sl have been found to last less than five milliseconds. However, in the case of 30 S subunits, this step is followed by some rearrangement of the message, which is not observed for Sl-oligo(U) complex. We realize that these data are open to other interpretations as well. When bound to the ribosome, the template is not extended into “a straight line” as is depicted in many textbooks. Rather, it forms a loop or “U-turn” as it passes through the ribosome with the separation between the ends about 45 A. It is of interest that this value changes very little for poly(U) lengths in the range of 18 to 35 nucleotide residues, as if the template goes around some convex surface. In contrast, for a linear form of poly(U),,, for example, this separation would be at least 90 A if the bound template has fully stacked bases (3 A x 30) (Bloomfield et al., 1974). In reality, it must be much larger since, with the exception of two mRNA triplets interacting with tRNA anticodons, other nucleotide residues of the template do not’ appear to be in a stacked conformation. This follows circular dichroism of poly(thioU) bound to the 30 S subunit, which does not reveal any stacking of nucleotide bases of the template (A. V. Bakin et al., unpublished results). The “U-turn” represents a general shape of the message. It does not exclude other bends or turns within the segment of mRNA covered by the ribosome. One of these bends most certainly exists between two codons of mRNA in A and P sites of the ribosome when they interact with anticodons of tRNAs (Rich, 1974). However, the general shape of the message does not depend on the presence or absence of tRNAs (see Results, section (g)) and, hence, is determined by the ribosome itself. The distances between the ends of a poly(U) fragment covered by the ribosome and the 5’ end of 16 S rRNA or prot,ein S8 have been estimated in order to gain more information on the location of the mRNA binding site on the 30 S subunit. The 5’ end of the 16 S rRNA and protein S8 are wellmapped points on the morphological model of the 30 S subunit (Lake, 1979; Mochalova et al., 1982). Their positions are indicated in Figure 5. The set of measured distances thus estimated is not enough to determine the exact position of the mRNA binding site. However, some important observations can be drawn from these experiments: (1) the 3’ end of the message is much closer to the 5’ terminus of 16 S rRNA than the 5’ end; (2) as the length of poly(U) increases, the 3’ end tends to approach to the 5’ end of 16 S rRNA; (3) the distance between the 3’ end of
Polynucleotides
30s
451
30s
Figure 5. Schematic presentation of the 30 S subunit in 2 projections. The position of the mRiVA binding site is shown by the hatched area. The black half-circle and arrow indicate the position of the 5’end and 3’end of 16 S RNA, respectively.
poly(U) and protein S8 decreases with increasing of the polymer length, while the distance between the 5’ end of RNA and protein S8 does not change significantly. These data combined with those obtained by IEM (Evstafieva et al., 1983) have allowed us to determine the orientation of the message and propose a tentative model of the mRNA path on the 30 S subunit. As shown in Figure 5, the 5’ end of the template is located near the 3’ end of 16 S rRNA, in agreement with our IEM data on the poly(U) 5’ end location (Evstafieva et al., 1983). Then t,he template descends along the 30 S “cleft” or “groove”, which is thought to be the site of SD (Oakes et al., 1986a) and the codon-anticodon interaction (Prince et al., 1982), thus reaching the “neck” of the subunit (the lower part of the groove). Such an orientation of the message is similar to that suggested by Olson et al. (1988) and seems to contradict that proposed by Lake (Oakes et al., 19866). The further path of mRNA is less evident. To form a loop or U-turn, there are two possibilities: (1) the message comes back along the groove (or near it) to its 5’ end; (2) the message continues its way around the neck of the 30 S subunit and quits the latter somewhere on its cytoplasmic side. In both cases the template may form on its way other bends or turns. Indeed, according to our experiments on affinity labeling of 16 S rRNA with the templates carrying photoactivated probes, the mRNA binding site may also include some parts of 16 S RNA located in the head of the 30 S subunit (A. V. Bakin et al., unpublished results). For many reasons, we prefer the latter model, which is shown in Figure 5 and will be discussed in detail elsewhere (I. N. Shatsky et al., unpublished results). In particular, it is in better agreement with the positions of the 5’ and 3’ ends of poly(U) with respect to protein S8 and the 16 S RNA 5’ end determined in this paper as well as with our IEM data on the position of the 3’ end of the template on the 70 S ribosome (Evstafieva et al., 1983). This model is also compatible with the position of the
452
A. V. Bakin
terminal
nucleotide
to SH-groups Results, latter IEM
section on
the
residues
of the template
of the ribosome-bound (e))
and
30 S subunit
(Stoffler-Meilicke,
with
the
recently
personal
relative
protein
Rl
(see
location
of
the
determined
by
communication).
References Bakin. A. V. & Shatskv. I. N. (1989). Selective fluorescent labelling of the RkA ends to study its macromolecule structure in the solution. BiopoliwL. Kletka (CJSSR). 5(3), 71-76. Bloomfield, V. A.. Crothers, D. M. CI:Tinoco, I. (1974). Tn Physical Chemistry of Nucleic Acids. p. 128, Harper & Row, New York. Boni, I. V. & Isaeva. D. M. (1988). Localization of RNA sites of Qfi. fr and MS2 bacteriophages which are involved in interactions with ribosomal protein RI during the 30 S*mRNA complex formation. Dokl. Akad. Nauk S&S%. 298(4), 1015-1018. Borisova. 0. F.. Kichevsakaya, A. A., Posvirin, V. V. & Samarina, 0. P. (1979). Nuclear ribonucleoproteins containing pro-mRNA. Investigation of their structure using ethidium-bromide and fluorescamin. Mol. Biol. (USSR), 13, 422-436. Dale, R. F. 8: Eisinger, J. (1975). Polarized excitation Concepts energy transfer. In Biochemical Fluorescence (Chen, R. F. & Edelhoch, H., eds). pp. 115-224. Marcel Dekker Inc., New York. Draper, D. E. & von Hippel, R. H. (1978). Nucleic acid binding properties of Escherichia coli ribosomal protein Sl. J. Mol. Biol. 122, 321-359. Evstafieva, A. G., Shatsky. T. N.. Bogdanov. A. A.. Sedmenkov. Y. P. & Vasiliev, V. D. (1983). Localization of 5’ and 3’ ends of the ribosome-bound segment of template polynucleotides by immune electron microscopy. EMBO J. 2. 799-804. Farclough. R. H. & Cantor, C. R. (1978). Singlet-singlet 48, 347-379. energy transfer. Methods &zymol. Forster. Th. (1966). In Modern Quantum Chemistry (Sinanoglu. 0.. ed.). vol. 3. pp. 93-137, Academic Press. New York. (Zorginis, S. bz Subramanian. A. R. (1980). The major ribosome binding sit.e of Escherichin coli ribosomal protein Sl is located in its N-terminal segment. J. Mol. Biol. 141. 393-408. Gottich, M. B.. Ivanovskaya. M. 0. & Shabarova, %. A. (1983). Synthesis of phosphamidmono and oligodeoxynucleotides in water medium. Bioorgan. (Ihem. (VSSR), 9. 1063-1067. Ciualerzi. C. 0.. C’alogero, R. A.. Canonaco. M. 8.. Brombach, M. &, Pon, C. L. (1988). Selection of mRNA by ribosomr during prokariotic translational init,iation. In Grnetics of Translation (Tuite, M. F., eds) , Picard. M. Pr Bolotin-Fukuhara. M.. pp. 317-330, Springer-Verlag. Berlin. Haas. E.. Katchalski-Katzir. E. 6 Steinberg, T. %. (1978). Effect of the orientation of donor and acceptor on t,he probability of energy t,ransfer involving electronic Biochemistry, 17. transitions of mixed polarization. 5064-5070. Kang, C. & Cantor, (‘. R. (1985). Structure of ribosomrby enzymatic bound messenger RNA as revealed accessibility studies. J. Mol. Biol. 181, 241-251. Katunin. V. I.. Semenkov. Yu. P.. Makhno. V. I. & Kirillov. S. V. (1980). Comparativr study of the int,eraction of polgucidilic acid with 30 S suhunit,s and 70 S rihosomes of E. coli. Nucl. drids i&s. 8. 403.-42 1.
et al.
Lake. ,J. A. (1979). Rihosome structure and functional Function and Genetics sites. Tn Ribosomes, Structure, (Chambliss et al.. eds), pp. 201-236. I’niversity Park Press. Baltimore. MD. Laemmli, LT. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London), 227, 680-683. Lipecky. R.. Kohlschein. .J. & Gassen. H. (1. (1977). Complex formation between rihosomal protein Sl oligo- and polynucleotides: chain length dependence and base specificity. ilrucl. Acids Res. 4. 3627-3642. Melhuish. W. H. (1961). Quantum efficiencies of fluorescence of organic substances: effect of solvent and concentration of the fluorescent solute. .J. Phys. C’hem. 65. 229-235. Mochalova. L. N.. Shatsky* I. N’.. Bogdanov, A. A. & Vasiliev, V. I). (1982). Topography of RNA in t,he ribosome: localization of the 16 S 5’ end by immune electron microscopy. J. Mol. Biol. 52. 637-650. Mulsen. A.. (‘olpan. M., Wollny. E.. Gassen. H. (:. & Riesner. I>. (1981). Mechanism of the interaction between rihosomal protein Sl and olieonucleotides. IYucl. ilcids Res. 9. 2367-2386. Oakes, M.. Clark, M.. Henderson. E. & Lakts. J. A. (1986a). DNA hybridization electron microscop?: ribosomal RNA nucleotides 1392-1407 are exposed m the cleft of the small subunit. Proc. Sat. .-lend. Sri.. ~:.s.d. 83 .I*“75-279. I Oakes. MM., Henderson. E.. Schienman. A.. (‘lark. M. Hr Lake, .I. A. (1986b). Rihosome structure. function and evolution: mapping rihosomal RNA. prot,rins. and functional sites in three dimensions. In Str~ct~rr. Function. a.nd Genetics of Ribosomes (Hardest., K. & Kramer. (i., eds), pp. 49-69. Springer-Verlag. Berlirl. Odom, 0. W.. Deng. H.-Y.. Subramanian, A. R. & Hardesty. B. (1984). Relaxation time. internal distance, mechanism of action of rihosomal protein Sl ilrch. BiochewL. Biophys. 230, 178-I 03. Olson, H. K.. Lasater, I,. S.. Cann, P. 21. & Blitz. I). (:. (1988). Messenger RNA orientation on t.hr rihoxomr. J. Hiol. ('hem. 263, 15196-15204. Prince. .J. B.. Taylor. B. H.. Thrulov. 11. I,.. Ofengand. .I. &. Zimmerman. R. A. (1982). Covalent crosslinking of tRNA\(a’ to 16 S RNA at t,he rihosomal 1’ sit.r: identification of crosslinked residues. Proc. ,Vnt. Acad. Sri.. I’.S.A. 79, 5450-5454. Rich, A. (1974). How tRNA may movt insitir the ribosome. Tn RibosvrnPs (NoAura. vl nl.. eds). pp. X71-884. (~!old Spring Harbor Lahorator,v Press. C’old Spring Harbor, NY. Schnier. ,I.. St,offler. (4. & Xshi. K. (l!%S). Deletion and insertion mutants in the structural genes for rihosomal protein Sl from Escherichia co/i. J. Rio!. (‘haunt.. 261. (“5). 11866~11871. Stiege. W.. Stade. K.. Srhulrr. I). & Brimac*ornhr. R (1988). (‘ovalent cross-linking of poly(A) to rihosomes and loc.alizat,ion of the cross-link site within the 16 S RNA. ili~cl. dcids 12~s. 16, 2369-238X. Suhramanian. A. R.. (1983). Structure and functions ot’ ribosomal prot,rin Sl. I’roy. Xucl. _4fids Hrs. .lfol. Biol. 28. 101- 142. Suhramanian, A. R.. (1984). Struct,ure and functions of rihosomal protein Sl. Trends Hiochrw. ,Sci. 9. 191-4!44. Suhramanian. A. R.. Rienhardt. P.. Kitnura. %I. & Surpanarayana. T. (1981). Fragments of ribosotnal protein Sl and it,s mutant form ml-S1 Rtrr. .I. Biochum. 119. 245-249. Survanaravana. T. & Subramanian. :I. I
