1 976 Volume 3 no.,5 no.5 May May1976
Volume 3
Nucleic Acids Research
Nucleic Acids Research
Specific release of ribosomal proteins by nucleic acid-intercalating agents.
Juan P. G. Ballesta, Michael J. Waring* and D. Vazquez Centro de Biologia Molecular, C. S. I. C. and U. A. M., Velazquez 144, Madrid 6, Spain
Received 11 March 1976 ABSTRACT
Increasing concentrations of ethidium bromide cause progressive inactivation of ribosomes, apparently by binding to double-stranded regions of the rRNA. At low drug concentrations (10-4M) the partial inhibition detected is due to specific release of proteins L7 and L12; activity can be restored by addition of an excess of these two proteins. At higher concentrations the inactivation is not reversed by supplementation with released proteins. The presence of ethanol affects the extent of ethidium binding and also the release of ribosomal proteins. In all tests the proteins most sensitive to the presence of the drug are L7 and L12, followed by L8/9, Lll, L27, L28, L29 and L30. Despite the fact that L7 and L12 are the first two proteins released by ethidium they are never totally missing from drug-treated ribosomes, though the other proteins can be displaced completely. About 50% of proteins L7 and L12 remain on the ribosomes at the highest drug concentrations tested, possibly indicating heterogeneity in the binding sites for the several copies present in the ribosome. INTRODUCTION Ethidium bromide binds to double stranded regions of nucleic acids by intercalation between adjacent base pairs. This intercalation causes a perturbation of the double helix reflected in local extension and unwinding of the duplex structure1' Ribosomal RNA has a high content of double stranded regions 3,4 Due to this fact ethidium bromide and other intercalating agents have been used as probes for the study of ribosomal structure. It has been shown that binding of a number of intercalating drugs results in a decrease in the stability of the ribosomal structure5'6 and also that ethidium is able to compete for binding to the ribosome with polyamines7 and even ribosomal proteins 8 . However, to the best of our knowledge no attempt has thus far been made to relate the alterations in the ribosomal structure with possible effects of ethidium on ribosomal activity. The present study was therefore initiated primarily to explore this point.
C) Information Retrieval Limited 1 Falconberg Court London Wl V 5FG England
1307
Nucleic Acids Research MATERIALS AND METHODS
Ribosomes were prepared from log phase Escherichia coli D-10 by alumina grinding and washed 5-6 times with 1 M NH4Cl buffer9. ( C)PhetRNA was obtained by charging commercial deacylated tRNA with ( C)phenylalanine (513 Ci/mmol) (The Radiochemical Centre, Amersham).CACCA(3H)Leu-Ac was prepared by RNase T1 treatment of N-acetyl-leucyl-tRNA as described elsewhere 10 . ((y- 32pP)GTP was prepared following the method of Glynn and Chappell 11 . Elongation factor EF G was purified from E. coli B according to Parmeggiani et al. 12 . Ethidium bromide was purchased from Sigma (St. Louis,Mo., USA); other drugs were gifts from Boots Pure Drug Co. and May & Baker Ltd.13. Polyphenylalanine synthesis, EF G-dependent hydrolysis of GTP and peptide bond formation (peptidyl transferase activity) in the fragment reaction assay were carried out as previously described10'14* Unless otherwise stated in the text, buffer and ion concentrations in the reaction mixtures were 80 mM NH4Cl, 10 mM Mg acetate, 20 mM TrisHC1 (pH 7.8) and 1 mM dithiothreitol (DTT). Treatment of the ribosomes (1 mg/ml) with ethidium bromide was carried out in 20 mM Tris-HCl (pH 7.8), containing 100 mM NH4Cl and 1 mM or 10 mM MgCl2 (buffer A). The reaction mixtures were kept at 30°C for 15 min and then either centrifuged at 200,000 x g in a Beckman rotor type 65 or precipitated by addition of 1 vol of ethanol after increasing the Mg concentration to 15 mM. The treated particles were resuspended in buffer A and dialysed against the same buffer for 24 h. The supernatants were dialysed against 10 mM Tris-HCl (pH 7.8), 3 mM DTT for 24 h with four changes of buffer, then lyophilized, taken up in 1/10 of the original volume of water and again dialyzed against the same buffer for 8 h. They were kept in liquid nitrogen. Bound ethidium bromide was estimated by measuring the absorption at 480 nm in a phenol extract of treated ribosomes using solutions of ethidium bromide in phenol as standards. The rRNA after phenol extraction was completely colourless, indicating that most, if not all, of the dye had been removed by the phenol. The ribosome concentration was determined from A260 considering one A260 unit = 24.5 pmol of 70S ribosomes of molecular weight 2.75 x 10 daltons with a content of 66% RNA. The hypothetical number of intercalation sites was estimated according to Wolfe et al.5 by dividing the number of nucleotides in the rRNA (about
1308
Nucleic Acids Research 5220 in 70S and 3720 in 50S subunits15) by 4.2, the minimum ratio of nucleotide to drug observed for intercalation into DNA. Ribosomal proteins were extracted by acetic acid treatment of the ribosomes. The ribosomal proteins released by ethidium bromide were obtained from supernatants dialysed exhaustively (48 h) against 2% acetic acid and then lyophilized. Two dimensional polyacrylamide gel electrophoresis of ribosomal proteins was carried out using the system described by Kaltschmidt and Wittmann16 as modified by Howard and Traut17. The different types of rRNA were separated by one-dimensional polyacrylamide gel electrophoresis (3.1 and 10% acrylamide in tandem) following the conditions of Yu 18 Quantitative analysis of proteins L7 and L12 by monodimensional gel electrophoresis was carried out using the system described by Li and Subramanian 19 . The gels were stained with Coomassie blue and scanned at 600 nm in a Gilford gel scanner adapted to a Unicam spectophotometer and connected to a Radiometer Servograph recorder. The areas under the peaks were cut out and weighed. The relative amount of proteins L7 + L12 in each sample was estimated from the ratio of the weights of the L7 + L12 area to a reference protein area. As reference we have taken the band moving immediately behind L7 and L12 whose intensity is not affected by ethidium bromide treatment (Fig. 3). RESULTS
Effect of ethidium bromide on ribosomal activity. The effect of increasing concentrations of ethidium bromide added directly to the reaction mixture on EF G-dependent GTPase, poly(U)-directed polyphenylalanine synthesis and peptidyl transferase activity is shown in Fig. 1. The EF G-dependent GTPase reaction was sensitive to the drug; it declined steadily as the concentration of ethidium bromide increased. Poly(U)-directed polyphenylalanine synthesis was also sensitive to the drug. Although it was slightly stimulated by low concentrations (3 x 10 5 M ethidium bromide), higher concentrations brought about a marked decline in polymerizing activity. Peptidyl transferase activity was less sensitive to ethidium bromide. Binding of ethidium bromide to ribosomes. In order to verify that the effects observed on the activities are due to Interaction of ethidium with the ribosomes, the particles were pretreated with the 1309
Nucleic Acids Research drug and then isolated either by centrifugation or by ethanol precipitation. The activity of the treated particles is shown in Fig. 2. EF G-dependent GTP hydrolysis (Fig. 2B) is more affected than )-ISO
I) 140
120
o LU-
SC
ETHIDIUM BROMIDE CONCENTRATION
CM)
Fig. 1. Effect of ethidium bromide on the synthesis of polyphenylalanine (O-cX, hydrolysis of GTP dependent on elongation factor EF G (Ar-A) and peptidyl transferase activity (fragment reaction) (C0-n) in ribosomes. The phenylalanine polymerization and EF G-dependent GTPase assays were performed as indicated in Materials and Methods. The ionic conditions for the fragment reaction were 0.27 M KCl, 0.033 M Tris-HCl (pH 7.8), 0.013 M Mg acetate, 1 mg/ml ribosomes, 2 x 10-3 M puromycin, about 3,000 cpm of CACCA-(3H)Leu-Ac and 33% ethanol. The reaction was allowed to proceed for 30 min at 0°C.
ETICM
EOMIDE,M
Fig. 2. Activity of ribosomes after treatment with ethidium bromide. (A) Peptidyl transferase activity measured by the "fragment reaction" and (B) EF G-dependent GTP hydrolysis by ribosomes pretreated with ethidium bromide at the indicated concentrations. (A--s4 70S ribosomes isolated after ethidium treatment by ethanol precipitation; (A,-4 as above plus supernatant fraction; (O-O) 70S ribosomes isolated by high speed centrifugation; (W-s) 50S subunits obtained by ethanol precipitation.
1310
Nucleic Acids Research peptidyl transferase (measured by "fragment reaction") (Fig. 2A) by ethidium bound to ribosomes. At low ethidium bromide concentrations (104 M) the elongation factor-dependent activity can be restored to the particles by addition of the supernatant fraction. However, the supernatant has no stimulatory effect on peptidyl-transferase. 1t is also clear from Fig. 2 that 50S subunits are more sensitive to the drug than 70S ribosomes. Three times less drug is required to obtain a similar degree of inhibition in the subunits. Estimates of the amount of drug bound to the ribosomes at different concentrations are given in Table 1. Less drug is found in the 70S ribosomes than in the 50S subunits, in agreement with the activity tests (Fig. 2) which show higher inhibition of the treated subunits. The method of isolation of the ribosomes after treatment also affects the amount of drug bound. It is clear from the results in Table I that ethanol precipitation removes part of the ethidium bound to the ribosomes. Two to three times more drug is found in ribosomes isolated by centrifugation, and again in agreement with these results ribosomes isolated in this way are proportionally less active than those isolated by ethanol precipitation (Fig. 2). TABLE I. Ethidium bromide bound to ribosomes
Ethidium bromide
Isolated by ethanol precipitation
Isolated by centrifugation
(M) 50S subunits
molecules/ ribosome 10 -4
3 x
10-4
lO3
45.3 141.1 381.4
70S ribosomes mo I ecu I es / ribosome
50S subunits
molecules/ ribosome
23.2
166.9
78.1 229.0
925.0
390.8
70S ribosomes mo I ecu I es / ribosome
43.0 166.0 382.1
Release of ribosomal components by ethidium bromide. The stimulation caused by the supernatant fraction of the EF G-dependent GTPase activity of the treated ribosanes (Fig. 2B) indicates that treatment with the drug releases some ribosomal component required for this function. Monodimensional gel electrophoresis did not detect the presence of rRNA in
1311
Nucleic Acids Research the supernatant after ethidium bromide treatment up to 10 3 M. The protein content of these fraction was analyzed by two dimensional. gel electrophoresis. The results are summarized in Table II. In supernatants obtained by ethanol precipitation from either 50S subunits or 70S ribosomes treated up to 3 x 10 4 M ethidium bromide only proteins L7 and L12 are present. At 10 3 M ethidium bromide proteins LI, L3, L6 or S5, L8/9, L25 and S1O are also released from the 70S ribosome while in the supernatant from 50S subunits, in addition to L7 and L12, proteins L29 and L30 give very strong spots and Lii is present but less conspicuous. In the supernatant obtained from treated 70S particles by centrifugation L7 and L12 are again the main spots at low ethidium bromide concentrations, but proteins L6, L3/9, L1O, Lll and traces of L25, L29 and L30 are also present in supernatants from treatment at ilO3 M ethidium bromide. Treatment of 50S subunits releases more proteins into the supernatant fraction. At the lowest ethidium bromide concentration six proteins L8/9, L6, L29, L30, L7 and L12, in this order of intensity, are already present. At 3 x 104 M ethidium bromide proteins LI, LIO, Ll and L25 are also found, and in addition at 10 3 M ethidium bromide proteins L3, L13, LIl, L22, L23, L24, L27 and L32 showed up in the gel plates. TABLE II. Ribosomal proteins found in the supernatant after ethidium bromide treatment
Ethidium bromide concentration (M)
10
Ribosomes isolated after treatment by: Ethanol precipitation
High speed centrifugation
50S subunits
70S ribosomes
50S subunits
L7,L12
L7,Li2
L,L6,L7,L8/9, L7,L12
70S ribosomes
L12,L29,L30
3
x
10
L7,L12
L7,L12
as
above plus
L7,L12
LIO,LlI ,L25
io-3
-L7,L1-,L12, L29,L30
L1,L3,L6,L7, L8/9,L12,L25, S1O
1312
plus L6,L7,L8/9, L3,L13,L18,L22, LO,L11,12, L23,L24,127,132 L25,L29,L30
as above
Nucleic Acids Research When we analyzed the proteins remaining in the treated particles, only at the highest ethidium bromide concentration used (10 3 M) were some proteins totally missing from the 50S subunit. Proteins Lll, L29 and L30 were undetectable in 50S particles isolated by precipitation and in addition proteins L8/9, L27 and L28 were absent from those obtained by centrifugation. In spite of the fact that L7 and L12 are the first to be released by ethidium bromide even at low concentrations these two proteins are never totally missing from the particles even after 10 M ethidium bromide treatment. We have quantified the real loss of these two proteins using a monodimensional gel electrophoresis system 19 . In this system proteins L7 and L12 move ahead of the other proteins (Fig. 3) and after staining their bands can be quantified by scanning. In Table III the relative content of proteins L7 plus L12 in 70S ribosomes treated with ethidium bromide at 1 mM and 10 mM MgCl2 and isolated by ethanol precipitation are shown. It is clear that the ribosomes lose no more than 45% of their L7 and L12 in both cases and that the loss is not specific for either of them, since the relative intensity of the two bands remains constant during treatment (Fig. 3).
TABLE III. Estimation of proteins L7 + L12;ethidium bromide treated 70S ribosomes
Ethidium bromide concentration (M)
Ethidium bromide treatment carried out at: 10 mM Mg ++ M g 1MM Mg area L7 + L12 bands area L7 + L12 bands area reference band
0
lO4 3 x 10
lO3 3 x 10i 3
area reference band
1.96 1.44
2.10
1.25 l .15 1.12
-
1.41
1.27 1.14
The ribosomes were recovered after treatment with the drug by ethanol precipitation. Control ribosomes were treated in the same way but in the absence of drug. The value of the areas ratio for ribosomes that have suffered no manipulation was 2.17.
1313
Nucleic Acids Research
_
L12 ~~L7C
ABC DE Fig. 3. Separation of proteins L7 and L12 by monodimensional gel electrophoresis. A, controlp articles; B, C, D and_E particles treated with lo04 M, 3 x 10-~m, lo-3 M and 3 x 10-3 M ethidium bromide.
It is also interesting to note that the majority of the loss of the two proteins has already occurred at the lowest ethidium bromide concentration used. Effects of proteins L7 and L12 on the activity of the treated particles. In order to determine whether the decrease in activity of ribosomes treated with 104 M ethidium bromide is exclusively related to the specific release of proteins L7 and L12, as the analysis of the supernatants seems to indicate, the effect of exogenous addition of these two purified proteins on the EF Gdependent GTPase of ethidium-treated ribosomes was examined. The results presented in Table IV show that these two proteins are indeed able to restore the ribosomal activity to a level similar to that of control ribosomes, but not with ribosomes treated with ethidium at concentrations higher than about M. It is interesting to note
104
1314
Nucleic Acids Research TABLE IV. Effect of proteins L7 and L12 on EF G-dependent GTP hydrolysis by ethidium bromide treated 50S ribosome subunits Ethidium bromide concentration l(M) 0
1o
5 x
4
10-4 10-3
aThe
Molecules of GTP hydrolyzed per ribosome per minutea
Minus methanol + L7/12 - L7/12 0.77 0.53 0.42 0.05
0.81 0.75 0.55 0.08
Plus methanol -
L7/12
+
1.5 1.08 0.84 0.10
L7/12 1.62 1.54 1.04 0.13
assay was carried out in the absence of 30S subunits.
in the same Table that methanol, although producing a general stimulation of the activity of the particles as previously reported20 is unable to compensate for a deficiency of proteins L7 and L12 as is the case for other types of protein-deficient ribosomes lacking these two proteins9,21,22 Proteins L7 and L12 are easily removed by ethanol washing when the concentration of monovalent ions in the medium is sufficiently high23. Under our conditions such loss does not occur to any significant extent, for control particles treated in the same way but without exposure to ethidium bromide contain not less than 90% of the L7 and L12 content of untreated particles (Table III). Nevertheless we thought it of interest to compare the activities of particles partially depleted of L7 and L12 by treatment with 1.5 x 10 -4 M ethidium bromide or by two ethanolwashings at 300 mM NH4Cl. As shown in Fig. 4 the two types of particles behave differently in response to increasing concentrations of Li plus L12. Ethidium bromide-treated particles require higher amounts of exogenous proteins,approximately ten times more, in order to restore full activity.
Effects of related drugs. In an effort to elucidate whether the release of ribosomal proteins by ethidium bromide can be correlated with its binding to the rRNA a few experiments were performed with a series of closely related phenanthridinium drugs whose interaction with DNA has 1315
Nucleic Acids Research
100
50~~~~
10
20
molecules L7+LI2/ribosome Fig. 4. Recovery of EF G-dependent GTPase activity by addition of purified proteins L7 and L12. (A-) 70S ribosomes washed with 300 mM NH Cl and 50% ethanol; (0-O) 70S ribosomes treated with 1.5 x 10-4 h ethidium bromide.
been investigated 3. Six compounds were selected for study, chosen for the known differences in their affinity for DNA and ability to induce distortion (unwinding) of the double helix. No qualitative differences from ethidium were detected as regards the ability of the phenanthridines to promote release of particular ribosomal proteins, but their relative potency compared to ethidium varied greatly when assessed in terms of the release of L7 and L12 (Table V). The drugs are listed in order of decreasing affinity for DNA. It can be seen that, with one exception, this order correlates satisfactorily with their ability to displace proteins L7 and L12 from ribosomes. The exception is the drug RD 16101, a dicarbethoxy analogue of ethidium, which promotes quite significant release of the two proteins (though not to the same extent as does ethidium) notwithstanding its considerably lower affinity for DNA. The interaction between this drug and nucleic acids is already known to be anomalous in several respects 3, and its action on ribosomes observed here provides a further instance. Its effect could be explained by preferential interaction with RNA as opposed to DNA, an intriguing possibility, but without further evidence such a suggestion is at best tentative. Leaving aside this single anomaly it can be concluded that
1316
Nucleic Acids Research TABLE V. Estimation of proteins L7 + L12 in ribosomes treated with phenanthridine drugs Drug
Designation
Relative K
Concentration
area L7 + L12 band area
None M&B 2421 M&B 2421 M&B 4594 M&B 3492 M&B 3492 RD 16101
M&B 3427 M&B 3427 M&B 1765
70S control Des-phenyl -d im id i um Des-phenyl-dimidium
8-Acetyl-dimidium p-Carboxy-dimidium p-Carboxy-dimidium
3,8-Dicarbethoxyethidium Des-amino-dimidium Des-amino-dimidium
3,8-Dibromo-des-
-
-
0.96
3 x
0.96 0.5 0.2 0.2 0.07 0.05 0.05 0.03
104 310
M 3 M 10 3 M 3 x 10 M 10-3 M 10o3 M
3 x 10
10o3 3 x 10
reference band
2.03 1 .90 1.32 1.78 2.10 1.75
1.53
M M M
2.15
M
2.10
2.03
1.85
amino-d imid ium M&B 1765
3,8-Dibromo-des-
0.03
10-3
amino-dimidium
Relative DNA-binding constant referred to that of ethidium set equal to 1.013.
the release of 17 and L12 by analogues of ethidium correlates broadly with their capacity to bind to and distort the structure of DNA, consistent with the notion that perturbation of the structure of helical regions in rRNA constitutes the origin of the loss of ribosomal proteins. Two additional experiments were performed with quite unrelated drugs, the steroidal diamines irehdiamine A and dipyrandium 24'25 to see whether the pattern of release of ribosomal proteins might be a common feature of drug binding, irrespective of its molecular characteristics. Again there was evidence of preferential release of L7 and L12 from ribosomes, and in one case (dipyrandium) a peculiar pattern of proteins was found in a supernatant prepared from treated ribosomes after high-speed
1317
Nucleic Acids Research centrifugation. These observations have not been pursued further. DISCUSSION Comparison of the inhibition curves caused by ethidium bromide when added to a polymerization system (polyphenylalanine synthesis) or partial protein synthesis reaction (peptidyl transferase and EF Gdependent GTP hydrolysis) with those obtained from ethidium bromide treated ribosomes (Fig. 2) seems to indicate that the inhibition is probably due to interaction of the drug with the ribosomes and not with other components of the polymerization system. A similar conclusion was reached by Krey and Hahn26 in their studies using other intercalating agents. At low ethidium bromide concentrations ( 3 x 10 4 M) the ribosomeand EF G-dependent hydrolysis of GTP is preferentially affected by the drug due to a specific and partial release of proteins L7 and L12 which appear to be essential for this function 27. The addition of these two proteins to the treated particles restores their activity. The release of proteins L7 and L12 must be mediated by the intercalating effect of ethidium bromide and not just by a non-specific ionic effect as occurs in concentrated monovalent salt solutions which also remove these two proteins. This is made clear by the fact that analogues of ethidium fail to produce the same effect at equivalent molar concentrations (Table V). In fact, the release of these two proteins by treatment with phenanthridines is well correlated with their known capacity to interact with and distort the structure of nucleic acids. The fact that the ethidium-treated particles require ten times more L7 and L12 than do monovalent salt-washed ribosomes in order to regain full activity also supports the notion that the .unwinding caused by the drug alters the binding site for these proteins, lowering its affinity. In this connection it is interesting to note that only a fraction, roughly 4576, of the L7 and L12 complement of the ribosomes is released by ethidium. This may imply that either the ribosome population is heterogeneous only part of it being sensitive to the drug or the binding sites for the various copies of these two proteins present in the ribosome are not equivalent, and that there exist two types having different sensitivity to the changes of conformation caused by the intercalated ethidium. If it is so, this is to the best of our knowledge the first time that a difference in behaviour between the various copies
1318
Nucleic Acids Research of L7 and L12 in the ribosome has been detected. The number of copies of these two proteins in the ribosome has been a matter of discussion. Recently Subramanian28 has convincingly shown that four molecules are present per ribosome. The relative amount of the two species however changes w.ith the growth conditions 29. The ratio L7/L12 seems to be 1:1 in our samples judging from the intensity of the bands in the monodimensional gel electrophoresis. If this is so, our results indicate that approximately two molecules of protein are released by ethidium, one of L7 and one of L12, implying that at least from the point of view of ethidium bromide sensitivity the two species are equivalent. Our results do not, of course, imply that the distribution of the various copies of L7 and L12 is the same in all the ribosomes of a given population. The existence of different ribosome subpopulations containing different proportions of L7 and L12 (as discussed by Subramanian2) is perfectly possible and we feel that the investigation of ethidium interaction with ribosome samples having different proportions of the two protein species might cast some light on its distribution among the various binding sites. Experiments along these lines are in progress. Although methanol is able partially to restore elongation factordependent functions to ribosomes totally depleted of proteins Li and L129'21 it is without effect in partially depleted ethidium-treated ribosomes. It is possible that in order for the alcohol to induce the appropriate conformation in the ribosome for interaction with the factors in the absence of L7 and L12, the proteins have to be totally removed. Another plausible hypothesis is that intercalated ethidium hinders this change of conformation. The higher accessibility of RNA for binding ethidium in 50S subunits as compared with 70S couples can be accounted for by either a looser conformation adopted by the subparticles when free in solution or by the RNA forming part of the interface that is not available in the 70S ribosome. The large difference found in the amount of drug bound by the two types of particles makes it quite possible that both explanations are correct. In fact changes in conformation upon association of ribosomal subunits have been documented using a variety of techniques3 32 Methanol, which also affects the conformation of the ribosome, probably tightening its structure33, strongly decreases the amount of ethidium bound as well.
1319
Nucleic Acids Research In addition to proteins L7 and L12, there are a number of other proteins that are also released by ethidium bromide from the ribosome. Perhaps the most striking fact is that most of them come from the 50S subunit (except for protein S10) suggesting that in the small subunit when it forms part of the 70S couple the protein binding sites are less dependent on double stranded regions of RNA or in general on the 16S RNA tertiary structure. This may be related to the facility with which reconstitution of this subunit has been achieved 34 Proteins L29 and L30 are the most sensitive to the drug, especially in free 50S subunits. They are totally missing from particles treated with 10 3 M ethidium bromide. Lll is occasionally missing and always strongly diminished, as also are proteins L8/9, LlO and L25 although only in the case of ribosomes isolated by high speed centrifugation, indicating that they require higher amounts of drug bound for their release. The results are surprising in the case of protein L25. This protein binds directly to 5S RNA and has been shown not to affect the fluorescence of ethidium bound to the RNA molecule35. However, these studies were carried out using purified RNA and proteins and the possibility cannot be excluded that in the ribosomal structure the proteins forming the complex with 5S RNA also bind to 23S RNA. Data in the literature indicate that while the 5S RNA molecule requires the presence of its binding proteins L5, L18 and L25 in order to be incorporated into the ribosome36 once it becomes part of the ribosomal structure the three proteins can be removed without effect on the 5S RNA binding22 indicating that the conformation of the 5S RNA-protein complex may be different depending whether it is free or in the ribosome. It is difficult to tell to what extent the release of these proteins by ethidium implicates a binding site related to double stranded RNA regions. This is however an attractive suggestion, especially in the case of proteins L7 and L12. These two proteins are involved in elongation factor-dependent activities and therefore possibly play a role in translocation as has been discussed by Mo"ller37. This leads to the hypothesis that they may be implicated somehow in the transmission of the energy liberated by GTP hydrolysis. If so, the conformation of the double stranded regions of rRNA associated with these proteins might be altered upon absorption of the GTP energy through a contraction or relaxation of the protein structure, inducing the changes in ribosomal structure supposedly required for the translocation step in protein synthesis. .
1320
Nucleic Acids Research ACKNOWLEDGEMENTS We are grateful to Miss Pilar Ochoa for expert technical assistance. This work has been supported by personal Grants to one of us (D.V.) from "Fondo Nacional para el Desarrollo de la Investigacion Cientifica" and "Lilly Indiana of Spain" and by Institutional Grants from "Comision Administradora del Descuento Complementario (Instituto Nacional de Previsio6n)" and "Direccion General de Sanidad " (Spain). We are also grateful to the British Council and the Medical Research Council for personal support to one of us (M.J.W.).
*Department of Pharmacology, School of Medicine, University of Cambridge, UK. REFERENCES
1. Crawford, L.V. and Waring, M.J. (1967) J.Mol.Biol. 25, 23-30. 2. Gale, E.F., Cundliffe, E., Reynolds, P.E., Richmond, M.H. and Waring, M.J. (1972) in The Molecular Basis of Antibiotic Action, pp. 173-277, John Wiley & Sons, London. 3. Ehresmann, C., Stiegler, P., Mackie, G.A., Zimmermann, R.R., Ebel, J.P.
and Fellner, P. (1975) Nucleic Acids Res. 2, 265-278. 4. Gatti, C., Houssier, C. and Fredericq, E. (1975) Biochim.Biophys.Acta 407, 308-319. 5. Wolfe, A.D., Allison, R.G. and Hahn, F.E. (1972) Biochemistry 11, 1569-1572. 6. Survanarayana, J. and Burma, D.P. (1975) Biochem.Biophys.Res.Commun.
65, 708-713. 7. Stevens, L. and Pascoe, G. (1972) Biochem.J. 128, 279-289. 8. Bollen, A., Herzog, A., Favre, A., Thibault, J. and Gros, F. (1970) FEBS Letters 11, 49-54.
9. Ballesta, J.P.G. and V3zquez, D. (1972) FEBS Letters 28, 337-342. 10. Monro, R.E. (1971) Methods Enzymol. 20, 472-481. 11. Glynn, I.M. and Chappell, J.B. (1964) Biochem.J. 90, 147-149. 12. Parmeggiani, A., Singer, C. and Gottschalk, E.M. (1971) Methods Enzymol. 20, 291-302. 13. Wakel in, L.P.G. and Waring, M.J. (1974) Mol.Pharmacol. 10, 567-586. 14. Ballesta, J.P.G., Montejo, V., Hern3ndez, F. and V3zquez, D. (1974) Eur.J.Biochem. 42, 167-175.
1321
Nucleic Acids Research 15. Fellner, P. (1971) Biochemie 53, 573-583. 16. Kaltschmidt, E. and Wittmann, G.H. (1970) Anal.Biochem. 36, 401-412. 17. Howard, G.A. and Traut, R.R. (1974) Methods Enzymol. 30, 526-539. 18. Yu, R.S.T. (1973) Hoppe-Seyler's Z.Physiol.Chem. 354, 125-130. 19. Li, K. and Subramanian, A.R. (1975) Mol.Biochem. 64, 121-129. 20. Ballesta, J.P.G., Montejo, V. and Vazquez, D. (1971) FEBS Letters
19, 75-78. 21. Hamel, E. and Nakamoto, T. (1972) J.Biol.Chem. 247, 6810-6817. 22. Sander, G., Marsh, R.C. and Parmeggiani, A. (1975) Biochemistry 14,
1805-1814. 23. Hamel, E., Koka, M. and Nakamoto, T. (1972) J.Biol.Chem. 247, 805-314. 24. Waring, M.J. and Henley, S.M. (1975) Nucleic Acids Res. 2, 567-586. 25. Silver, S., Wendt, L. and Bhattacharyya, P. (1975) in Antibiotics. III. Mechanism of Action of Antimicrobial and Antitumor Agents (J.Corcoran and F.E.Hahn, editors), pp. 614-622, Springer-Verlag,
Heidelberg.
26. Krey, A.K. and Hahn, F.E. (1975) Naturwissenchaften 62, 99-100. 27. Kisha, K., Mol ler, W. and Stoffler, G. (1971) Nature 233, 62-63. 28. Subramanian, A.R. (1975) J.Mol.Biol. 95, 1-8. 29. Ramagopal, S. and Subramanian, A.R. (1974) Proc.Nat.Acad.Sci. 71,
2136-2140. 30. Ball, L.A., Johnson, P.M. and Walker, J.0. (1973) Eur.J.Biochem. 37,
12-20. 31. Sherman, M.I. and Simpson, M.V. (1969) Cold Spring Harbor Symp.Quant.
Biol. 34, 220-222. 32. Zitomer, R.S. and Flaks, J.G. (1972) J.Mol.Biol. 71, 263-279.
33. Ballesta, J.P.G. and Vazquez, D. (1973) Biochemistry 12, 5063-5067. 34. Traub, P. and Nomura, M. (1968) Proc.Nat.Acad.Sci. 59, 777-784. 35. Feunteun, J., Monier, R., Garrett, R., Le Bret, M. and Le Pecq, J.B.
(1975) J.Mol.Biol. 93, 535-541. 36. Yu, S.T. and Wittmann, H.G. (1973) Biochim.Biophys.Acta 324, 375-385. 37. Moller, W. (1974) in Ribosome, M.Nomura, A.Tissieres and P.Lengyel, editors), pp. 711-731, Cold.Spring Harbor Laboratory.
1322
Volume 3
Nucleic Acids Research
Nucleic Acids Research
Specific release of ribosomal proteins by nucleic acid-intercalating agents.
Juan P. G. Ballesta, Michael J. Waring* and D. Vazquez Centro de Biologia Molecular, C. S. I. C. and U. A. M., Velazquez 144, Madrid 6, Spain
Received 11 March 1976 ABSTRACT
Increasing concentrations of ethidium bromide cause progressive inactivation of ribosomes, apparently by binding to double-stranded regions of the rRNA. At low drug concentrations (10-4M) the partial inhibition detected is due to specific release of proteins L7 and L12; activity can be restored by addition of an excess of these two proteins. At higher concentrations the inactivation is not reversed by supplementation with released proteins. The presence of ethanol affects the extent of ethidium binding and also the release of ribosomal proteins. In all tests the proteins most sensitive to the presence of the drug are L7 and L12, followed by L8/9, Lll, L27, L28, L29 and L30. Despite the fact that L7 and L12 are the first two proteins released by ethidium they are never totally missing from drug-treated ribosomes, though the other proteins can be displaced completely. About 50% of proteins L7 and L12 remain on the ribosomes at the highest drug concentrations tested, possibly indicating heterogeneity in the binding sites for the several copies present in the ribosome. INTRODUCTION Ethidium bromide binds to double stranded regions of nucleic acids by intercalation between adjacent base pairs. This intercalation causes a perturbation of the double helix reflected in local extension and unwinding of the duplex structure1' Ribosomal RNA has a high content of double stranded regions 3,4 Due to this fact ethidium bromide and other intercalating agents have been used as probes for the study of ribosomal structure. It has been shown that binding of a number of intercalating drugs results in a decrease in the stability of the ribosomal structure5'6 and also that ethidium is able to compete for binding to the ribosome with polyamines7 and even ribosomal proteins 8 . However, to the best of our knowledge no attempt has thus far been made to relate the alterations in the ribosomal structure with possible effects of ethidium on ribosomal activity. The present study was therefore initiated primarily to explore this point.
C) Information Retrieval Limited 1 Falconberg Court London Wl V 5FG England
1307
Nucleic Acids Research MATERIALS AND METHODS
Ribosomes were prepared from log phase Escherichia coli D-10 by alumina grinding and washed 5-6 times with 1 M NH4Cl buffer9. ( C)PhetRNA was obtained by charging commercial deacylated tRNA with ( C)phenylalanine (513 Ci/mmol) (The Radiochemical Centre, Amersham).CACCA(3H)Leu-Ac was prepared by RNase T1 treatment of N-acetyl-leucyl-tRNA as described elsewhere 10 . ((y- 32pP)GTP was prepared following the method of Glynn and Chappell 11 . Elongation factor EF G was purified from E. coli B according to Parmeggiani et al. 12 . Ethidium bromide was purchased from Sigma (St. Louis,Mo., USA); other drugs were gifts from Boots Pure Drug Co. and May & Baker Ltd.13. Polyphenylalanine synthesis, EF G-dependent hydrolysis of GTP and peptide bond formation (peptidyl transferase activity) in the fragment reaction assay were carried out as previously described10'14* Unless otherwise stated in the text, buffer and ion concentrations in the reaction mixtures were 80 mM NH4Cl, 10 mM Mg acetate, 20 mM TrisHC1 (pH 7.8) and 1 mM dithiothreitol (DTT). Treatment of the ribosomes (1 mg/ml) with ethidium bromide was carried out in 20 mM Tris-HCl (pH 7.8), containing 100 mM NH4Cl and 1 mM or 10 mM MgCl2 (buffer A). The reaction mixtures were kept at 30°C for 15 min and then either centrifuged at 200,000 x g in a Beckman rotor type 65 or precipitated by addition of 1 vol of ethanol after increasing the Mg concentration to 15 mM. The treated particles were resuspended in buffer A and dialysed against the same buffer for 24 h. The supernatants were dialysed against 10 mM Tris-HCl (pH 7.8), 3 mM DTT for 24 h with four changes of buffer, then lyophilized, taken up in 1/10 of the original volume of water and again dialyzed against the same buffer for 8 h. They were kept in liquid nitrogen. Bound ethidium bromide was estimated by measuring the absorption at 480 nm in a phenol extract of treated ribosomes using solutions of ethidium bromide in phenol as standards. The rRNA after phenol extraction was completely colourless, indicating that most, if not all, of the dye had been removed by the phenol. The ribosome concentration was determined from A260 considering one A260 unit = 24.5 pmol of 70S ribosomes of molecular weight 2.75 x 10 daltons with a content of 66% RNA. The hypothetical number of intercalation sites was estimated according to Wolfe et al.5 by dividing the number of nucleotides in the rRNA (about
1308
Nucleic Acids Research 5220 in 70S and 3720 in 50S subunits15) by 4.2, the minimum ratio of nucleotide to drug observed for intercalation into DNA. Ribosomal proteins were extracted by acetic acid treatment of the ribosomes. The ribosomal proteins released by ethidium bromide were obtained from supernatants dialysed exhaustively (48 h) against 2% acetic acid and then lyophilized. Two dimensional polyacrylamide gel electrophoresis of ribosomal proteins was carried out using the system described by Kaltschmidt and Wittmann16 as modified by Howard and Traut17. The different types of rRNA were separated by one-dimensional polyacrylamide gel electrophoresis (3.1 and 10% acrylamide in tandem) following the conditions of Yu 18 Quantitative analysis of proteins L7 and L12 by monodimensional gel electrophoresis was carried out using the system described by Li and Subramanian 19 . The gels were stained with Coomassie blue and scanned at 600 nm in a Gilford gel scanner adapted to a Unicam spectophotometer and connected to a Radiometer Servograph recorder. The areas under the peaks were cut out and weighed. The relative amount of proteins L7 + L12 in each sample was estimated from the ratio of the weights of the L7 + L12 area to a reference protein area. As reference we have taken the band moving immediately behind L7 and L12 whose intensity is not affected by ethidium bromide treatment (Fig. 3). RESULTS
Effect of ethidium bromide on ribosomal activity. The effect of increasing concentrations of ethidium bromide added directly to the reaction mixture on EF G-dependent GTPase, poly(U)-directed polyphenylalanine synthesis and peptidyl transferase activity is shown in Fig. 1. The EF G-dependent GTPase reaction was sensitive to the drug; it declined steadily as the concentration of ethidium bromide increased. Poly(U)-directed polyphenylalanine synthesis was also sensitive to the drug. Although it was slightly stimulated by low concentrations (3 x 10 5 M ethidium bromide), higher concentrations brought about a marked decline in polymerizing activity. Peptidyl transferase activity was less sensitive to ethidium bromide. Binding of ethidium bromide to ribosomes. In order to verify that the effects observed on the activities are due to Interaction of ethidium with the ribosomes, the particles were pretreated with the 1309
Nucleic Acids Research drug and then isolated either by centrifugation or by ethanol precipitation. The activity of the treated particles is shown in Fig. 2. EF G-dependent GTP hydrolysis (Fig. 2B) is more affected than )-ISO
I) 140
120
o LU-
SC
ETHIDIUM BROMIDE CONCENTRATION
CM)
Fig. 1. Effect of ethidium bromide on the synthesis of polyphenylalanine (O-cX, hydrolysis of GTP dependent on elongation factor EF G (Ar-A) and peptidyl transferase activity (fragment reaction) (C0-n) in ribosomes. The phenylalanine polymerization and EF G-dependent GTPase assays were performed as indicated in Materials and Methods. The ionic conditions for the fragment reaction were 0.27 M KCl, 0.033 M Tris-HCl (pH 7.8), 0.013 M Mg acetate, 1 mg/ml ribosomes, 2 x 10-3 M puromycin, about 3,000 cpm of CACCA-(3H)Leu-Ac and 33% ethanol. The reaction was allowed to proceed for 30 min at 0°C.
ETICM
EOMIDE,M
Fig. 2. Activity of ribosomes after treatment with ethidium bromide. (A) Peptidyl transferase activity measured by the "fragment reaction" and (B) EF G-dependent GTP hydrolysis by ribosomes pretreated with ethidium bromide at the indicated concentrations. (A--s4 70S ribosomes isolated after ethidium treatment by ethanol precipitation; (A,-4 as above plus supernatant fraction; (O-O) 70S ribosomes isolated by high speed centrifugation; (W-s) 50S subunits obtained by ethanol precipitation.
1310
Nucleic Acids Research peptidyl transferase (measured by "fragment reaction") (Fig. 2A) by ethidium bound to ribosomes. At low ethidium bromide concentrations (104 M) the elongation factor-dependent activity can be restored to the particles by addition of the supernatant fraction. However, the supernatant has no stimulatory effect on peptidyl-transferase. 1t is also clear from Fig. 2 that 50S subunits are more sensitive to the drug than 70S ribosomes. Three times less drug is required to obtain a similar degree of inhibition in the subunits. Estimates of the amount of drug bound to the ribosomes at different concentrations are given in Table 1. Less drug is found in the 70S ribosomes than in the 50S subunits, in agreement with the activity tests (Fig. 2) which show higher inhibition of the treated subunits. The method of isolation of the ribosomes after treatment also affects the amount of drug bound. It is clear from the results in Table I that ethanol precipitation removes part of the ethidium bound to the ribosomes. Two to three times more drug is found in ribosomes isolated by centrifugation, and again in agreement with these results ribosomes isolated in this way are proportionally less active than those isolated by ethanol precipitation (Fig. 2). TABLE I. Ethidium bromide bound to ribosomes
Ethidium bromide
Isolated by ethanol precipitation
Isolated by centrifugation
(M) 50S subunits
molecules/ ribosome 10 -4
3 x
10-4
lO3
45.3 141.1 381.4
70S ribosomes mo I ecu I es / ribosome
50S subunits
molecules/ ribosome
23.2
166.9
78.1 229.0
925.0
390.8
70S ribosomes mo I ecu I es / ribosome
43.0 166.0 382.1
Release of ribosomal components by ethidium bromide. The stimulation caused by the supernatant fraction of the EF G-dependent GTPase activity of the treated ribosanes (Fig. 2B) indicates that treatment with the drug releases some ribosomal component required for this function. Monodimensional gel electrophoresis did not detect the presence of rRNA in
1311
Nucleic Acids Research the supernatant after ethidium bromide treatment up to 10 3 M. The protein content of these fraction was analyzed by two dimensional. gel electrophoresis. The results are summarized in Table II. In supernatants obtained by ethanol precipitation from either 50S subunits or 70S ribosomes treated up to 3 x 10 4 M ethidium bromide only proteins L7 and L12 are present. At 10 3 M ethidium bromide proteins LI, L3, L6 or S5, L8/9, L25 and S1O are also released from the 70S ribosome while in the supernatant from 50S subunits, in addition to L7 and L12, proteins L29 and L30 give very strong spots and Lii is present but less conspicuous. In the supernatant obtained from treated 70S particles by centrifugation L7 and L12 are again the main spots at low ethidium bromide concentrations, but proteins L6, L3/9, L1O, Lll and traces of L25, L29 and L30 are also present in supernatants from treatment at ilO3 M ethidium bromide. Treatment of 50S subunits releases more proteins into the supernatant fraction. At the lowest ethidium bromide concentration six proteins L8/9, L6, L29, L30, L7 and L12, in this order of intensity, are already present. At 3 x 104 M ethidium bromide proteins LI, LIO, Ll and L25 are also found, and in addition at 10 3 M ethidium bromide proteins L3, L13, LIl, L22, L23, L24, L27 and L32 showed up in the gel plates. TABLE II. Ribosomal proteins found in the supernatant after ethidium bromide treatment
Ethidium bromide concentration (M)
10
Ribosomes isolated after treatment by: Ethanol precipitation
High speed centrifugation
50S subunits
70S ribosomes
50S subunits
L7,L12
L7,Li2
L,L6,L7,L8/9, L7,L12
70S ribosomes
L12,L29,L30
3
x
10
L7,L12
L7,L12
as
above plus
L7,L12
LIO,LlI ,L25
io-3
-L7,L1-,L12, L29,L30
L1,L3,L6,L7, L8/9,L12,L25, S1O
1312
plus L6,L7,L8/9, L3,L13,L18,L22, LO,L11,12, L23,L24,127,132 L25,L29,L30
as above
Nucleic Acids Research When we analyzed the proteins remaining in the treated particles, only at the highest ethidium bromide concentration used (10 3 M) were some proteins totally missing from the 50S subunit. Proteins Lll, L29 and L30 were undetectable in 50S particles isolated by precipitation and in addition proteins L8/9, L27 and L28 were absent from those obtained by centrifugation. In spite of the fact that L7 and L12 are the first to be released by ethidium bromide even at low concentrations these two proteins are never totally missing from the particles even after 10 M ethidium bromide treatment. We have quantified the real loss of these two proteins using a monodimensional gel electrophoresis system 19 . In this system proteins L7 and L12 move ahead of the other proteins (Fig. 3) and after staining their bands can be quantified by scanning. In Table III the relative content of proteins L7 plus L12 in 70S ribosomes treated with ethidium bromide at 1 mM and 10 mM MgCl2 and isolated by ethanol precipitation are shown. It is clear that the ribosomes lose no more than 45% of their L7 and L12 in both cases and that the loss is not specific for either of them, since the relative intensity of the two bands remains constant during treatment (Fig. 3).
TABLE III. Estimation of proteins L7 + L12;ethidium bromide treated 70S ribosomes
Ethidium bromide concentration (M)
Ethidium bromide treatment carried out at: 10 mM Mg ++ M g 1MM Mg area L7 + L12 bands area L7 + L12 bands area reference band
0
lO4 3 x 10
lO3 3 x 10i 3
area reference band
1.96 1.44
2.10
1.25 l .15 1.12
-
1.41
1.27 1.14
The ribosomes were recovered after treatment with the drug by ethanol precipitation. Control ribosomes were treated in the same way but in the absence of drug. The value of the areas ratio for ribosomes that have suffered no manipulation was 2.17.
1313
Nucleic Acids Research
_
L12 ~~L7C
ABC DE Fig. 3. Separation of proteins L7 and L12 by monodimensional gel electrophoresis. A, controlp articles; B, C, D and_E particles treated with lo04 M, 3 x 10-~m, lo-3 M and 3 x 10-3 M ethidium bromide.
It is also interesting to note that the majority of the loss of the two proteins has already occurred at the lowest ethidium bromide concentration used. Effects of proteins L7 and L12 on the activity of the treated particles. In order to determine whether the decrease in activity of ribosomes treated with 104 M ethidium bromide is exclusively related to the specific release of proteins L7 and L12, as the analysis of the supernatants seems to indicate, the effect of exogenous addition of these two purified proteins on the EF Gdependent GTPase of ethidium-treated ribosomes was examined. The results presented in Table IV show that these two proteins are indeed able to restore the ribosomal activity to a level similar to that of control ribosomes, but not with ribosomes treated with ethidium at concentrations higher than about M. It is interesting to note
104
1314
Nucleic Acids Research TABLE IV. Effect of proteins L7 and L12 on EF G-dependent GTP hydrolysis by ethidium bromide treated 50S ribosome subunits Ethidium bromide concentration l(M) 0
1o
5 x
4
10-4 10-3
aThe
Molecules of GTP hydrolyzed per ribosome per minutea
Minus methanol + L7/12 - L7/12 0.77 0.53 0.42 0.05
0.81 0.75 0.55 0.08
Plus methanol -
L7/12
+
1.5 1.08 0.84 0.10
L7/12 1.62 1.54 1.04 0.13
assay was carried out in the absence of 30S subunits.
in the same Table that methanol, although producing a general stimulation of the activity of the particles as previously reported20 is unable to compensate for a deficiency of proteins L7 and L12 as is the case for other types of protein-deficient ribosomes lacking these two proteins9,21,22 Proteins L7 and L12 are easily removed by ethanol washing when the concentration of monovalent ions in the medium is sufficiently high23. Under our conditions such loss does not occur to any significant extent, for control particles treated in the same way but without exposure to ethidium bromide contain not less than 90% of the L7 and L12 content of untreated particles (Table III). Nevertheless we thought it of interest to compare the activities of particles partially depleted of L7 and L12 by treatment with 1.5 x 10 -4 M ethidium bromide or by two ethanolwashings at 300 mM NH4Cl. As shown in Fig. 4 the two types of particles behave differently in response to increasing concentrations of Li plus L12. Ethidium bromide-treated particles require higher amounts of exogenous proteins,approximately ten times more, in order to restore full activity.
Effects of related drugs. In an effort to elucidate whether the release of ribosomal proteins by ethidium bromide can be correlated with its binding to the rRNA a few experiments were performed with a series of closely related phenanthridinium drugs whose interaction with DNA has 1315
Nucleic Acids Research
100
50~~~~
10
20
molecules L7+LI2/ribosome Fig. 4. Recovery of EF G-dependent GTPase activity by addition of purified proteins L7 and L12. (A-) 70S ribosomes washed with 300 mM NH Cl and 50% ethanol; (0-O) 70S ribosomes treated with 1.5 x 10-4 h ethidium bromide.
been investigated 3. Six compounds were selected for study, chosen for the known differences in their affinity for DNA and ability to induce distortion (unwinding) of the double helix. No qualitative differences from ethidium were detected as regards the ability of the phenanthridines to promote release of particular ribosomal proteins, but their relative potency compared to ethidium varied greatly when assessed in terms of the release of L7 and L12 (Table V). The drugs are listed in order of decreasing affinity for DNA. It can be seen that, with one exception, this order correlates satisfactorily with their ability to displace proteins L7 and L12 from ribosomes. The exception is the drug RD 16101, a dicarbethoxy analogue of ethidium, which promotes quite significant release of the two proteins (though not to the same extent as does ethidium) notwithstanding its considerably lower affinity for DNA. The interaction between this drug and nucleic acids is already known to be anomalous in several respects 3, and its action on ribosomes observed here provides a further instance. Its effect could be explained by preferential interaction with RNA as opposed to DNA, an intriguing possibility, but without further evidence such a suggestion is at best tentative. Leaving aside this single anomaly it can be concluded that
1316
Nucleic Acids Research TABLE V. Estimation of proteins L7 + L12 in ribosomes treated with phenanthridine drugs Drug
Designation
Relative K
Concentration
area L7 + L12 band area
None M&B 2421 M&B 2421 M&B 4594 M&B 3492 M&B 3492 RD 16101
M&B 3427 M&B 3427 M&B 1765
70S control Des-phenyl -d im id i um Des-phenyl-dimidium
8-Acetyl-dimidium p-Carboxy-dimidium p-Carboxy-dimidium
3,8-Dicarbethoxyethidium Des-amino-dimidium Des-amino-dimidium
3,8-Dibromo-des-
-
-
0.96
3 x
0.96 0.5 0.2 0.2 0.07 0.05 0.05 0.03
104 310
M 3 M 10 3 M 3 x 10 M 10-3 M 10o3 M
3 x 10
10o3 3 x 10
reference band
2.03 1 .90 1.32 1.78 2.10 1.75
1.53
M M M
2.15
M
2.10
2.03
1.85
amino-d imid ium M&B 1765
3,8-Dibromo-des-
0.03
10-3
amino-dimidium
Relative DNA-binding constant referred to that of ethidium set equal to 1.013.
the release of 17 and L12 by analogues of ethidium correlates broadly with their capacity to bind to and distort the structure of DNA, consistent with the notion that perturbation of the structure of helical regions in rRNA constitutes the origin of the loss of ribosomal proteins. Two additional experiments were performed with quite unrelated drugs, the steroidal diamines irehdiamine A and dipyrandium 24'25 to see whether the pattern of release of ribosomal proteins might be a common feature of drug binding, irrespective of its molecular characteristics. Again there was evidence of preferential release of L7 and L12 from ribosomes, and in one case (dipyrandium) a peculiar pattern of proteins was found in a supernatant prepared from treated ribosomes after high-speed
1317
Nucleic Acids Research centrifugation. These observations have not been pursued further. DISCUSSION Comparison of the inhibition curves caused by ethidium bromide when added to a polymerization system (polyphenylalanine synthesis) or partial protein synthesis reaction (peptidyl transferase and EF Gdependent GTP hydrolysis) with those obtained from ethidium bromide treated ribosomes (Fig. 2) seems to indicate that the inhibition is probably due to interaction of the drug with the ribosomes and not with other components of the polymerization system. A similar conclusion was reached by Krey and Hahn26 in their studies using other intercalating agents. At low ethidium bromide concentrations ( 3 x 10 4 M) the ribosomeand EF G-dependent hydrolysis of GTP is preferentially affected by the drug due to a specific and partial release of proteins L7 and L12 which appear to be essential for this function 27. The addition of these two proteins to the treated particles restores their activity. The release of proteins L7 and L12 must be mediated by the intercalating effect of ethidium bromide and not just by a non-specific ionic effect as occurs in concentrated monovalent salt solutions which also remove these two proteins. This is made clear by the fact that analogues of ethidium fail to produce the same effect at equivalent molar concentrations (Table V). In fact, the release of these two proteins by treatment with phenanthridines is well correlated with their known capacity to interact with and distort the structure of nucleic acids. The fact that the ethidium-treated particles require ten times more L7 and L12 than do monovalent salt-washed ribosomes in order to regain full activity also supports the notion that the .unwinding caused by the drug alters the binding site for these proteins, lowering its affinity. In this connection it is interesting to note that only a fraction, roughly 4576, of the L7 and L12 complement of the ribosomes is released by ethidium. This may imply that either the ribosome population is heterogeneous only part of it being sensitive to the drug or the binding sites for the various copies of these two proteins present in the ribosome are not equivalent, and that there exist two types having different sensitivity to the changes of conformation caused by the intercalated ethidium. If it is so, this is to the best of our knowledge the first time that a difference in behaviour between the various copies
1318
Nucleic Acids Research of L7 and L12 in the ribosome has been detected. The number of copies of these two proteins in the ribosome has been a matter of discussion. Recently Subramanian28 has convincingly shown that four molecules are present per ribosome. The relative amount of the two species however changes w.ith the growth conditions 29. The ratio L7/L12 seems to be 1:1 in our samples judging from the intensity of the bands in the monodimensional gel electrophoresis. If this is so, our results indicate that approximately two molecules of protein are released by ethidium, one of L7 and one of L12, implying that at least from the point of view of ethidium bromide sensitivity the two species are equivalent. Our results do not, of course, imply that the distribution of the various copies of L7 and L12 is the same in all the ribosomes of a given population. The existence of different ribosome subpopulations containing different proportions of L7 and L12 (as discussed by Subramanian2) is perfectly possible and we feel that the investigation of ethidium interaction with ribosome samples having different proportions of the two protein species might cast some light on its distribution among the various binding sites. Experiments along these lines are in progress. Although methanol is able partially to restore elongation factordependent functions to ribosomes totally depleted of proteins Li and L129'21 it is without effect in partially depleted ethidium-treated ribosomes. It is possible that in order for the alcohol to induce the appropriate conformation in the ribosome for interaction with the factors in the absence of L7 and L12, the proteins have to be totally removed. Another plausible hypothesis is that intercalated ethidium hinders this change of conformation. The higher accessibility of RNA for binding ethidium in 50S subunits as compared with 70S couples can be accounted for by either a looser conformation adopted by the subparticles when free in solution or by the RNA forming part of the interface that is not available in the 70S ribosome. The large difference found in the amount of drug bound by the two types of particles makes it quite possible that both explanations are correct. In fact changes in conformation upon association of ribosomal subunits have been documented using a variety of techniques3 32 Methanol, which also affects the conformation of the ribosome, probably tightening its structure33, strongly decreases the amount of ethidium bound as well.
1319
Nucleic Acids Research In addition to proteins L7 and L12, there are a number of other proteins that are also released by ethidium bromide from the ribosome. Perhaps the most striking fact is that most of them come from the 50S subunit (except for protein S10) suggesting that in the small subunit when it forms part of the 70S couple the protein binding sites are less dependent on double stranded regions of RNA or in general on the 16S RNA tertiary structure. This may be related to the facility with which reconstitution of this subunit has been achieved 34 Proteins L29 and L30 are the most sensitive to the drug, especially in free 50S subunits. They are totally missing from particles treated with 10 3 M ethidium bromide. Lll is occasionally missing and always strongly diminished, as also are proteins L8/9, LlO and L25 although only in the case of ribosomes isolated by high speed centrifugation, indicating that they require higher amounts of drug bound for their release. The results are surprising in the case of protein L25. This protein binds directly to 5S RNA and has been shown not to affect the fluorescence of ethidium bound to the RNA molecule35. However, these studies were carried out using purified RNA and proteins and the possibility cannot be excluded that in the ribosomal structure the proteins forming the complex with 5S RNA also bind to 23S RNA. Data in the literature indicate that while the 5S RNA molecule requires the presence of its binding proteins L5, L18 and L25 in order to be incorporated into the ribosome36 once it becomes part of the ribosomal structure the three proteins can be removed without effect on the 5S RNA binding22 indicating that the conformation of the 5S RNA-protein complex may be different depending whether it is free or in the ribosome. It is difficult to tell to what extent the release of these proteins by ethidium implicates a binding site related to double stranded RNA regions. This is however an attractive suggestion, especially in the case of proteins L7 and L12. These two proteins are involved in elongation factor-dependent activities and therefore possibly play a role in translocation as has been discussed by Mo"ller37. This leads to the hypothesis that they may be implicated somehow in the transmission of the energy liberated by GTP hydrolysis. If so, the conformation of the double stranded regions of rRNA associated with these proteins might be altered upon absorption of the GTP energy through a contraction or relaxation of the protein structure, inducing the changes in ribosomal structure supposedly required for the translocation step in protein synthesis. .
1320
Nucleic Acids Research ACKNOWLEDGEMENTS We are grateful to Miss Pilar Ochoa for expert technical assistance. This work has been supported by personal Grants to one of us (D.V.) from "Fondo Nacional para el Desarrollo de la Investigacion Cientifica" and "Lilly Indiana of Spain" and by Institutional Grants from "Comision Administradora del Descuento Complementario (Instituto Nacional de Previsio6n)" and "Direccion General de Sanidad " (Spain). We are also grateful to the British Council and the Medical Research Council for personal support to one of us (M.J.W.).
*Department of Pharmacology, School of Medicine, University of Cambridge, UK. REFERENCES
1. Crawford, L.V. and Waring, M.J. (1967) J.Mol.Biol. 25, 23-30. 2. Gale, E.F., Cundliffe, E., Reynolds, P.E., Richmond, M.H. and Waring, M.J. (1972) in The Molecular Basis of Antibiotic Action, pp. 173-277, John Wiley & Sons, London. 3. Ehresmann, C., Stiegler, P., Mackie, G.A., Zimmermann, R.R., Ebel, J.P.
and Fellner, P. (1975) Nucleic Acids Res. 2, 265-278. 4. Gatti, C., Houssier, C. and Fredericq, E. (1975) Biochim.Biophys.Acta 407, 308-319. 5. Wolfe, A.D., Allison, R.G. and Hahn, F.E. (1972) Biochemistry 11, 1569-1572. 6. Survanarayana, J. and Burma, D.P. (1975) Biochem.Biophys.Res.Commun.
65, 708-713. 7. Stevens, L. and Pascoe, G. (1972) Biochem.J. 128, 279-289. 8. Bollen, A., Herzog, A., Favre, A., Thibault, J. and Gros, F. (1970) FEBS Letters 11, 49-54.
9. Ballesta, J.P.G. and V3zquez, D. (1972) FEBS Letters 28, 337-342. 10. Monro, R.E. (1971) Methods Enzymol. 20, 472-481. 11. Glynn, I.M. and Chappell, J.B. (1964) Biochem.J. 90, 147-149. 12. Parmeggiani, A., Singer, C. and Gottschalk, E.M. (1971) Methods Enzymol. 20, 291-302. 13. Wakel in, L.P.G. and Waring, M.J. (1974) Mol.Pharmacol. 10, 567-586. 14. Ballesta, J.P.G., Montejo, V., Hern3ndez, F. and V3zquez, D. (1974) Eur.J.Biochem. 42, 167-175.
1321
Nucleic Acids Research 15. Fellner, P. (1971) Biochemie 53, 573-583. 16. Kaltschmidt, E. and Wittmann, G.H. (1970) Anal.Biochem. 36, 401-412. 17. Howard, G.A. and Traut, R.R. (1974) Methods Enzymol. 30, 526-539. 18. Yu, R.S.T. (1973) Hoppe-Seyler's Z.Physiol.Chem. 354, 125-130. 19. Li, K. and Subramanian, A.R. (1975) Mol.Biochem. 64, 121-129. 20. Ballesta, J.P.G., Montejo, V. and Vazquez, D. (1971) FEBS Letters
19, 75-78. 21. Hamel, E. and Nakamoto, T. (1972) J.Biol.Chem. 247, 6810-6817. 22. Sander, G., Marsh, R.C. and Parmeggiani, A. (1975) Biochemistry 14,
1805-1814. 23. Hamel, E., Koka, M. and Nakamoto, T. (1972) J.Biol.Chem. 247, 805-314. 24. Waring, M.J. and Henley, S.M. (1975) Nucleic Acids Res. 2, 567-586. 25. Silver, S., Wendt, L. and Bhattacharyya, P. (1975) in Antibiotics. III. Mechanism of Action of Antimicrobial and Antitumor Agents (J.Corcoran and F.E.Hahn, editors), pp. 614-622, Springer-Verlag,
Heidelberg.
26. Krey, A.K. and Hahn, F.E. (1975) Naturwissenchaften 62, 99-100. 27. Kisha, K., Mol ler, W. and Stoffler, G. (1971) Nature 233, 62-63. 28. Subramanian, A.R. (1975) J.Mol.Biol. 95, 1-8. 29. Ramagopal, S. and Subramanian, A.R. (1974) Proc.Nat.Acad.Sci. 71,
2136-2140. 30. Ball, L.A., Johnson, P.M. and Walker, J.0. (1973) Eur.J.Biochem. 37,
12-20. 31. Sherman, M.I. and Simpson, M.V. (1969) Cold Spring Harbor Symp.Quant.
Biol. 34, 220-222. 32. Zitomer, R.S. and Flaks, J.G. (1972) J.Mol.Biol. 71, 263-279.
33. Ballesta, J.P.G. and Vazquez, D. (1973) Biochemistry 12, 5063-5067. 34. Traub, P. and Nomura, M. (1968) Proc.Nat.Acad.Sci. 59, 777-784. 35. Feunteun, J., Monier, R., Garrett, R., Le Bret, M. and Le Pecq, J.B.
(1975) J.Mol.Biol. 93, 535-541. 36. Yu, S.T. and Wittmann, H.G. (1973) Biochim.Biophys.Acta 324, 375-385. 37. Moller, W. (1974) in Ribosome, M.Nomura, A.Tissieres and P.Lengyel, editors), pp. 711-731, Cold.Spring Harbor Laboratory.
1322
