Eur. J. Biochem. 161,111-117 (1986) 0 FEBS 1986

The mechanism of the protein-synthesis elongation cycle in eukaryotes Effect of r ich on the ribosomal interaction with elongation factors

Lars NILSSON and Odd NYGARD Department of Cell Biology, The Wenner-Gren Institute, University of Stockholm

(Received June 2/August 4, 1986) - EJB 86 0517

The functional significance of the post-translocation interaction of eukaryotic ribosomes with EF-2 was studied using the translational inhibitor ricin. Ribosomes treated with ricin showed a decreased rate of elongation accompanied by altered proportions of the different ribosomal phases of the elongation cycle. The content of ribosome-bound EF-2 was diminished by approximately 65% while that of EF-1 was unaffected. The markedly reduced content of EF-2 was caused by an inability of the ricin-treated ribosomes to form high-affinity pre- translocation complexes with EF-2. However, the ribosomes were still able to interact with EF-2 in the form of a low-affinity post-translocation complex. Ricin-treated ribosomes showed an altered ability to stimulate the GTP hydrolysis catalysed by either EF-1 or EF-2. The EF-1 -catalysed hydrolysis was reduced by approximately 70%, resulting in a decreased turnover of the quaternary EF-1 . GTP . aminoacyl-tRNA . ribosome complex. In contrast, the EF-2-catalysed hydrolysis was increased by more than 400%, despite the lack of pre-translocation complex formation. The effect was not restricted to empty reconstituted ribosomes since gently salt-washed polysomes also showed an increased rate of GTP hydrolysis. The results indicate that the EF-1- and EF-2- dependent hydrolysis of GTP was activated by a common center on the ribosome that was specifically adapted for promoting the GTP hydrolysis of either EF-1 or EF-2. Furthermore, the results suggest that the GTP hydrolysis catalysed by EF-2 occurred in the low-affinity post-translocation complex.

The translational elongation in eukaryotes is mediated through the two elongation factors EF-1 and EF-2 and in- volves a cyclic series of reactions by which the amino acids are added to the growing polypeptide chain [l, 21. Each cycle can be divided into three basic reaction steps. In the first step cognate aminoacyl-tRNAs are brought to the A-site of the mRNA-programmed ribosome in the form of a ternary complex with EF-1 and GTP [l, 21. During this reaction, GTP is hydrolysed to GDP and inorganic phosphate [3]. This step is followed by the transfer of the nascent peptide in the ribo- somal P-site to the newly attached aminoacyl-tRNA [l , 21. Finally, elongation factor EF-2 promotes the translocation of the elongated peptidyl-tRNA from the A-site to the P-site under the hydrolysis of GTP [4]. A common GTPase center on the ribosome, involving the 5s rRNA . L5 complex, has been suggested to participate in both the EF-1- and EF-2- dependent GTP hydrolysis [5].

The two elongation factors bind to the ribosome at identi- cal or partially overlapping binding sites [6 - 91. Elongation factor EF-1 associates exclusively with post-translocation ribosomes, i. e. ribosomes having their peptidyl-tRNA in the P-site, whereas EF-2 binds only to pre-translocation ribo- somes, i.e. ribosomes carrying the peptidyl-tRNA in the A- site [6, 71.

~

Correspondence to 0. NygBrd, Avdelningen for Cellbiologi, Wenner-Gren Institut, Stockholms Universitet, S-106 91 Stockholm, Sweden

Abbreviutions. EF, elongation factor; GuoPqCHzlP, guanosine 5’-[B,y-methylene]triphosphate; poly (U), poly(uridy1ic acid).

We have previously demonstrated that EF-2 forms two types of ribosomal complexes [lo]. In the presence of non- hydrolysable GTP analogues a high-affinity pre-translocation complex is formed, whereas GDP promotes the formation of a low-affinity post-translocation complex [l 1, 121. In a coupled translation system, EF-2 only exerts its function on pre-trans- location ribosomes. The physiological meaning of the post- translocation complex is therefore unclear.

Ricin has been shown to inhibit the eukaryotic elongation cycle [13 - 181. The effect is due to a catalytic inactivation of the large ribosomal subunit [18], but the mechanism is not known. Attempts to demonstrate a ricin-dependent modifica- tion of ribosomal proteins has been unsuccessful [ 19, 201. Unlike a-sarcin, ricin does not show any ribonucleolytic effect on 18s or 28s rRNA [21] but, in a recent report, rich was shown to degrade isolated 5s and 5.8s rRNA [22]. Both the EF-1- and EF-2-dependent GTPases have been reported to be inhibited after ricin treatment [14, 231. It has also been shown that the binding of EF-2 to ribosomes is reduced after ricin treatment [17]. However, the EF-1 -dependent binding of aminoacyl-tRNA to the ribosomal A-site seems not be affected [24].

Translational inhibitors like puromycin have been used successfully to gain insight into the mechanisms involved in protein synthesis. In this report we have studied the functional effects of ricin to gain a better understanding of the mechanism of the elongation cycle. Ricin was shown to decrease the EF-1-dependent GTPase activity, thereby re- ducing the turnover of the ribosome-bound aminoacyl- tRNA . EF-1 . GTP complex. Although ricin almost abolished the pre-translocation binding of EF-2 to the

112

ribosome, a conspicuous increase in the EF-2-catalysed hy- drolysis of GTP was observed. Since ricin-treated ribosomes were fully active in forming post-translocation ribosomal complexes with EF-2, the data suggest that the EF-2-depend- ent GTP hydrolysis takes place after translocation.

MATERIALS AND METHODS

Chemicals

[14C]Leucine (57 Ci/mol), [35S]methionine (1450 Ci/mol), [ 14C]valine (290 Ci/mol), ['4C]phenylalanine (522 Ci/mol), [y- 32P]GTP (13.8 Ci/mmol) and [14C]NAD+ (226 Ci/mol) were from Amersham International, UK GTP, GuoPP[CH2]P and Met-Val dipeptide were from Sigma Chemical Co. (St Louis, MO, USA). Micrococcus nuclease was from Boehringer Mannheim (Mannheim, FRG). Cronex lo'dose mammo- graphy X-ray film was from DuPont De Nemours & Co. Inc. (Wilmington, DE, USA). Poly(U) was obtained from Miles Laboratories (Elkhardt, IN, USA). Nitrocellulose filters HAWP 0.45 pm were from Millipore (Bedford, MA, USA). Ricinus toxin was isolated from the seeds of Ricinus communis (castor beans) according to Olsnes et al. [25]. Diphtheria toxin was a generous gift of Dr M. Tiru (The National Bacteriologi- cal Laboratories, Stockholm, Sweden).

Preparation of reticulocyte lysates and rat liver ribosomal subunits

Rabbit reticulocyte lysates were prepared as previously described [lo]. The lysates were treated with Micrococcus nuclease as described by Pelham and Jackson [26]. The ribosome content of the lysates was determined by centrifuga- tion for 170 rnin at 257000 x g,, through discontinuous su- crose gradients composed of 1 ml of 1 .O M sucrose in 20 mM Tris/HCl (pH 7.6), 10 mM 2-mercaptoethanol, 100 mM KC1, 5 mM Mg(CH3C00)2 superimposed on 1 ml of 1.8 M su- crose in the same buffer. The pellets were rinsed with 0.25 M sucrose, 70 mM KC1, 30 mM Hepes/KOH (pH 7.6), 2 mM Mg(CH3C00)2, 1 mM dithiothreitol. The particles were re- suspended in the same buffer and the absorbance at 260 nm determined. Rat liver ribosomal subunits were prepared as previously described [27].

Preparation of '251-labelled EF-I and EF-2

Elongation factors EF-1 and EF-2 were prepared from rat liver as described by Moon el al. [28] and Nilsson and Nygird [29], respectively. The purity of the factors were analysed as previously described [lo]. The factors were labelled with '''I according to Bolton and Hunter [30].

Preparation of globin mRNA and [14C]Phe-tRNA

Globin mRNA was prepared from isolated reticulocyte polysomes by phenol extraction [31] and repeated oligo (dT)- cellulose chromatography [32]. ['4C]Phe-tRNA was prepared as described previously [33].

Synthesis of Met- Val dipeptide by the reticylocyte lysate system

Micrococcus-nuclease- trea ted reticulocyte lysates were in- cubated in the presence of 30 pg globin mRNA, 2.3 pM [14C]valine and 450 nM [35C]methionine. After incubation

for 10 rnin at 30 "C, the ribosomes were pelleted by centrifuga- tion at 234000 x g,, for 180 rnin through a discontinuous su- crose gradient composed of 2.3 ml of 0.35 M sucrose in (25 mM KCl, 20 mM Tris/HC1 pH 7.6,2 mM MgCI2, 0.1 mM EDTA, 5 mM 2-mercaptoethanol) superimposed on 1 ml 1.11 M sucrose in the same buffer. The ribosomal pellets were carefully rinsed with distilled water, resuspended in water and the RNA was hydrolysed by incubation for 30 rnin at 30'C in the presence of 0.1 M NaOH. The released peptides were acetylated with 1 % (v/v) acetic acid anhydride in the presence of 1 M triethanolamine pH 8.0. After acidification with HCl, the acetylated peptides were extracted with ethyl acetate and analysed by thin-layer chromatography followed by auto- radiography.

Determination of ribosome-bound EF-I and EF-2 Samples of reticulocyte lysate were supplemented with

increasing amounts of '251-labelled EF-1 and incubated in the presence of unlabelled leucine as described above. The ribosomes were pelleted by centrifugation through discontinu- ous sucrose gradients and the ribosome-bound radioactivity was determined [lo]. For the determination o l ribosome- bound EF-2, purified reticulocyte polysomes were washed with 0.5 M KC1 and the EF-2 content of the salt-wash fraction was analysed by incubation with diphtheria toxin and [14C]NAD+ as described [34].

Formation of the EF-2 ' ribosome complex '251-labelled EF-2 was incubated with isolated 40s and

60s ribosomal subunits and GuoPP[CH2]P as previously de- scribed [l ll.

Determination of EF-1 and ribosome-dependent GTPase activity

The ribosome- and EF-1-dependent GTPase activity was determined using duplicate 100-p1 samples of 100 mM KCl, 20 mM Tris/HCl (pH 7.6), 5 mM MgCI2, 7 mM 2-mercap- toethanol, 40 pM [y-32P]GTP, 20 pg poly(U), 25 pmol 40s subunits, 25 pmol 60s subunits, 25 pmol Phe-tRNA and 100pmol EF-1. The ribosomal subunits had been pre- incubated for 5 min at 37°C with or without ricin. The re- leased [32P]phosphate was extracted as phosphomolybdate with benzene/isobutanol (1 : 1) [35] and the radioactivity deter- mined.

Determination of EF-2- and ribosome-dependent GTPase activity

The ribosome- and EF-2-dependent GTPase activity was determined using duplicate 5O-pl samples of 100 mM KCI, 20 mM Tris/HCI (pH 7.6), 5 mM MgCI2, 7 mM 2-mercapto- ethanol, 25 pmol 40s subunits, 25 pmol 60s subunits, 40 pM [Y-~~P]GTP and indicated amounts of EF-2 or ADP- ribosylated EF-2. The ribosomal subunits were preincubated for 5 rnin at 30°C with or without ricin. The inhibitor/ ribosome ratio was 0.12. The released [32P]phosphate was determined as for EF-1.

Determination of EF-2- and polysome-dependent GTPase activity

Reticulocyte lysates were layered onto a discontinuous sucrose gradient containing 1 ml 0.5 M sucrose in 25 mM

113

KC1, 20 mM Tris/HCl (pH 7.6), 2 mM Mg(CH3C00)2, 5 mM 2-mercaptoethanol superimposed on 1 ml of 1 .O M su- crose in the same buffer. The lysates were centrifuged for 180 min at 257000 x gav. The resulting polysomal pellets were carefully rinsed with 0.25 M sucrose in 70 mM KCI, 30 mM Hepes/KOH (pH 7.6), 2 mM Mg(CH3C00)2, 1 mM di- thiothreitol and resuspended in the same buffer to a final concentration of 500 pmol ribosomes/ml. Salt-washed poly- somes were isolated in the same way except that the lysates were treated with 0.5 M NH4C1 for 10 min at 0°C prior to centrifugation. In this case the discontinuous sucrose gradients contained 0.5 M NH4C1 in the 0.5 M sucrose solu- tion and 0.35 M KC1 in the 1.0 M sucrose solution. The polysome- and EF-2-dependent GTPase activity was deter- mined as described above with 20 pmol EF-2, but the reaction mixtures contained 30 pmol polysomes or salt-washed polysomes instead of isolated subunits.

RESULTS Distribution of the dgjerent ribosomal phases ojthe elongation cycle

Ricin has been reported to inhibit the translocation of peptidyl-tRNA from the ribosomal A-site to the P-site by a catalytic inactivation of the large ribosomal subunit [18]. Due to this effect, ricin was expected to alter the steady-state dis- tribution of the different ribosomal phases in the elongation cycle. We have earlier demonstrated that these phases can be characterised by the ribosomal content of the elongation factors EF-1 and EF-2 [lo].

In order to determine the distribution of the ribosomal phases after ricin treatment the content of ribosome-bound EF-1 was analysed by an isotope dilution assay [lo]. Rabbit reticulocyte lysates were incubated in the presence of in- creasing amounts of radioactively labelled EF-1 and the ribosome-bound radioactivity was measured after isolation of the ribosomes [lo]. The labelled factor competed effectively with the endogenous EF-1 in binding to ribosome both in control and ricin-treated lysates. The binding data were linearised by plotting the reciprocal of the added amount of labelled EF-1 against the reciprocal of the ribosome-bound radioactive EF-1 (Fig. 1). The intercept on the Y-axis repre- sents the reciprocal of the total number of ribosome-bound EF-1 [lo]. By regression analysis of the data, 1 mol of control ribosomes were calculated to contain 0.36 mol EF-1 whereas the ricin-treated ribosomes contained 0.30 mol. Thus, the ri- bosomal content of EF-1 was not significantly changed after ricin treatment. Furthermorc, the results indicate that the ability of the ricin-treated ribosomes to interact with EF-3 was unimpaired. This is in agreement with the lack of effect of ricin on the EF-1-catalysed binding of aminoacyl-tRNA

The ribosomal content of EF-2 was determined by the specific ADP-ribosylation of the factor in the presence of [l4C]NAD' and diphtheria toxin after detachment of EF-2 from the isolated ribosomes by high-salt treatment [lo]. As seen in Table 1, the content of EF-2 was decreased by more than 60% after ricin treatment, indicating a diminished ability of the ribosomes to interact with the factor.

~ 4 1 .

Ejfeect of ricin on the formation of'EF-2 * ribosome complcxes

Elongation factor EF-2 associates with high affinity to pre-translocation ribosomes, i. e. ribosomes having peptidyl-

2 4 6 lO-'O/EF-l added (mol-')

Fig. 1. Ribosomal content oJ'EF-1 ufier ricin treatment. Rcticulocyte lysates, 200 pl, were incubated in the presence of increasing amounts of '251-labelled EF-1, as described in Materials and Methods. The ribosomes were pelleted by centrifugation and the amount of ribosome-bound radioactivity determined [lo]. The binding data were linearised by plotting the reciprocal of the added amount of EF-1 against the reciprocal of the recovered ribosome-bound EF-1. The total ribosomal conlent of EF-1 was calculated from the reciprocal of the intercept on the Y-axis [lo]. Ribosomal contcnt of EF-1 in control lysates (0) and after ricin-treatment (M)

Table 1. Effect of Ricinus toxin on the riho.somu1 content of EF-2 in reticulocyte lysutcs Reticulocyte lysates were incubated with or without ricin (46 mmol/ mol ribosomes) as indicated in Materials and Methods. After 10 min at 37"C, the ribosomes were isolated and the content of EF-2 mea- sured as described in Materials and Methods

Treatment EF-2 bound to ribosomes

mo1/100 mol (% control)

Control 20.2 (100) Ridn 7.5 (37)

tRNA located in the A-site [6, 71. The diminished ribosomal content of EF-2 in ricin-treated ribosomes must therefore result either from a decreased ribosomal affinity for EF-2 or from a lack of available pre-translocation binding sites. We have previously demonstrated that the binding of EF-2 to ribosomes requires guanosine nucleotides and that the factor forms two different types of ribosomal complexes depending on the type of guanosine nucleotide present ill]. Non- cleavable GTP analogues stimulate the formation of a high- affinity pre-translocation complex that is stable during cen- trifugation [ll, 121, whereas the formation of a low-affinity post-translocation complex is stimulated by GDP [ l l , 12].The latter complex can be demonstrated by gradient centrifugation after fixation with glutaraldehyde. As seen in Table 2, rich treatment of the ribosomes drastically reduced the formation of the pre-translocation complex. This is in agreement with a previous observation [17]. However, the ribosomes were still fully active in the formation of the low-affinity post-transloca- tion complex (Table 2).

114

Fig. 2. Peptidisation reaction in the presence of ricin. Micrococcus- nuclease-treated reticulocyte lysates [26] supplemented with globin mRNA, [35S]methionine and ['4C]valine, were incubated for 10 min at 3 0 T , in the presence of 15 nM ricin. The ribosome-bound amino acids and oligopeptides were acetylated, extracted and analysed by thin-layer chromatography as described in Materials and Methods. (A) A~etyl-['~C]valine; (B) extract from control lysate; (C) acetyl- [35S]methionine and a~etyl-['~C]valine; (D) ricin-treated lysate; (E) a~etyl-[~'S]rnethionine. Ac-Met-Val was used as marker and detected by staining with ninhydrin

Table 2. Ability of EF-2 to form ribosomal complexes with rich-treated empty reconstituted ribosomes For complex formation, 50 pmol of '2s1-labelled EF-2 was incubated with 31 pmol of 40s subunits and 31 pmol of 60s subunits in 50 pl of 100 mM KCI, 4.5 mM magnesium acetate, 30 mM Hepes/KOH @H 7.6), 7.5 mM 2-mercaptoethanol, 100 mM sucrose and 0.5 mM GuoPfiCH,]P. After incubation for 10 min at 37"C, the incubation mixtures were anlysed by sucrose gradient centrifugation [I I]. Values in parentheses represent percentage of control. n. d. = no determined

Inhibitor/ EF-Z/ribosomes ribosomes

without fixation with fixation

mol/mol mo1/100 mol (YO)

0 (control) 90.4 (100) 91.2 (100) 5 x 61.6 (68) n.d. 5 x 10-2 11.6 (13) n.d. 5 x 10-1 0 (0) 76.0 (83)

Effect of ricin on the format ion of pre-translocation ribosomes

The inability of EF-2 to bind to pre-translocation ribosomes in a coupled translation system would prevent the translocation of peptidyl-tRNA from the ribosomal A-site to the P-site, thus leading to a decreased proportion of ribosomes in the post-translocation phase of the elongation cycle and an accumulation of pre-translocation ribosomes, i. e. ribosomes

Table 3. Eflect of ricin on the ribosome- and EF-1-dependent GTPase Isolated 40s and 60s ribosomal subunits were pre-incubated with or without ricin as described in Materials and Methods. The inhibitor/ ribosome ratio was 0.12. The GTPase activity was determined using duplicate samples as described in Materials and Methods. The GTP hydrolysis observed in the complete system lacking Phe-tRNA was considered as background. Values in parentheses represent percentagc of control

Additions [32P]P04 released

EF-1 Phe-tRNA ribosomes actual above background

- + + + + + +

-

- control ricin control control control ricin ricin

pmol (YO)

0 44 39 15 18

3

14 141 67 (1 00)

74 96 22 (33)

having peptidyl-tRNA in the A-site. Since EF-1 has been shown to bind exclusively to ribosomes containing peptidyl- tRNA in the P-site [6, 71, a shortage of post-translocation ribosomes should be expected to give a decreased ribosomal content of EF-1. As this was not the case (cf. Fig. l), the treatment must, in addition to the reduced ribosomal affinity for EF-2, interfere with the formation of pre-translocation ribosomes, either by an inhibition of the peptidyltransferase reaction or by a reduced turnover of the ribosome-bound EF- 1 . aminoacyl-tRNA . GTP complex.

In order to distinguish between these two possibilities, the peptidyltransferase activity of the ricin-treated ribosomes was studied using a Micrococcus-nuclease-treated reticulocyte lysate programmed with globin mRNA. The two initial codons in both a- and P-globin mRNA code for, consecutively, methionine and valine [36]. Thus, peptide bond formation in the lysates can be assayed by the synthesis of the dipeptide Met-Val. As seen in Fig. 2, both lysates formed dipeptides. The increase in the amount of Met-Val after ricin treatment was a result of the decreased synthesis of non-extractable oligopeptides and polypeptides (not shown). The data suggest that the peptide bond formation was unimpaired by the treat- ment. This is in agreement with previous results [37, 381.

EF-1 and EF-2 bind to identical or partially overlapping binding sites on the ribosomes [6, 71. Because of the mutually exclusive binding, EF-1 has to be detached from the ribosome in order to allow for the binding of EF-2. The detachment of the ribosome-bound ternary complexes requires GTP hy- drolysis. As seen in Table 3, the activity of the EF-1- and ribosomes-dependent GTPase was markedly reduced after ricin treatment. The data indicate a decreased formation of pre-translocation ribosomes as a result of an inhibited turnover of the EF-1-containing ribosomal complex, thus counterbalancing the reduced formation of post-translocation ribosomes.

115

A

20 40 60 80 EF-2 added(pmo1)

Fig. 3. EF-2- and ribosome-dependent hydrolysis of GTP. The EF-2- and ribosome-dependent GTPase activity was determined, as de- scribed in Materials and Methods, using increasing amounts of EF-2. The ribosomes were preincubated with (A) or without (0 ) rich as described in Materials and Methods

Table 4. EF-2- and ribosome-dependent GTPase in he presence ofricin and diphtheria toxin EF-2 was pre-incubated in the presence of NAD' and diphtheria toxin as described [34]. For control experiments, the factor was in- cubated in the presence of NAD'. The excess of NAD' was removed by microdialysis. The ribosomes were pre-incubated with or without ricin as described in Materials and Methods. EF-2- and ribosome- dependent GTP hydrolysis was determined using duplicate samples of 18 pmol EF-2 as described in Materials and Methods. Values within parentheses represent percentage stimulation

EF-2 [32P]P04 released/ribosomes

control ricin

mol/mol (%)

Native 8.9 39.2 (440) ADP-ribosylated 6.0 30.0 (500)

EF-2- and ribosome dependent GTPase In the presence of empty reconstituted ribosomes, EF-2

acts as an uncoupled GTPase [l, 21. The rate of hydrolysis was linearly dependent on EF-2 up to an approximately equimolar ratio of EF-2 to ribosomes (Fig. 3). Since ricin-treated ribosomes were unable to form a high-affinity pre-transloca- tion complex with EF-2, a markedly reduced GTP hydrolysis was expected with these ribosomes. This presumption was supported by the observation that the reduced affinity of ADP-ribosylated EF-2 for ribosomes [I21 was accompanied by a decreased GTPase activity (Table 4). However, the ricin treatment increased the EF-Zdependent GTP hydrolysis conspicuously (Fig. 3, Table 4). At subsaturating levels of EF-2 the stimulation was more than fourfold. Thus, ricin-treated ribosomes formed a low-affinity EF-2 . GTP.ribosome complex which was very active in GTP hy- drolysis. As seen in Table 4, ADP-ribosylation of EF-2 did not inhibit the ricin stimulation of the GTPase. Thus, the two modifications had separate effects on the EF-2/ribosome interaction.

100

a 0

g 8 0 a P 6 60

E 0

40 B

- 20

0

L

0

P E

A B c D

n I

Fig. 4. The influence of translation inhibitors on the distribution of the different ribosomal phases during elongation. The ribosomal content of elongation factors EF-1 (open blocks) and EF-2 (hatched blocks) was determined as described in Material and Methods. Distribution of the ribosomal phases at steady state (A) and in the presence of the translational inhibitors GuoPP[CH2]P (B), inhibitor extracted from barley seeds (unpublished observation) (C) and ricin (D)

Table 5 . EF-2- andpolysome-dependent hydrolysis of GTP The polysome- and EF-2-dependent GTPase was determined as de- scribed in Materials and Methods. The ribosomes were preincubated with or without ricin as described in Materials and Methods. The inhibitor/ribosome ratio was 0.30. Values within parentheses rep- resent stimulation by EF-2

EF-2 Polysomes [32P]P04 released/polysomes Ricin- induced

control ricin stimulation

mol/mol (-fold) -fold

- control 46.2 (1.00) 47.0 (1.00) 1.02 + control 57.5 (1.24) 49.1 (1.04) 0.85

- salt-washed 13.4 (1.00) 11.9 (1.00) 0.89 + salt-washed 27.2 (2.03) 38.4 (3.27) 1.41

To learn whether ricin also simulated the EF-2-dependent GTPase using coupled ribosomes, analogous experiments were carried out with polysomes. As seen in Table 5, control polysomes showed a substantial endogenous hydrolysis of GTP. This activity was not significantly altered after ricin treatment. Addition of EF-2 to the control polysomes in- creased the rate of GTP cleavage by approximately 25% cor- responding to a hydrolysis of approximately 10 GTP molecules/ribosome. However, EF-2 failed to stimulate the GTPase activity in the presence of ricin-treated polysomes (Table 5 ) .

After gently washing the polysomes with KC1, without disturbing the distribution of the translational phases in the ribosome population [lo], the rate of endogenous GTP hy- drolysis was markedly reduced (Table 5). However, EF-2 stimulated the GTPase activity to approximately the same extent as with non-salt-washed polysomes. In the presence of ricin-treated salt-washed ribosomes, EF-2 strikingly stimulated the GTP hydrolysis (Table 5). Thus, the lack of stimulation seen with non-salt-washed polysomes was due to an additional component removed by the KCl treatment rather than to the coupling of the ribosomes to defined phases of the elongation cycle.

116

DISCUSSION

The cyclic character of the elongation implies that the ability of the ribosome to participate in a specific reaction step is dependent on the previous reaction by which the ribosome is converted into the appropriate phase. We have earlier demon- strated that at steady-state translation, approximately 75% of the ribosomes are in the post-translocation phase of the elongation cycle, i.e. carry peptidyl-tRNA in the P-site [lo]. In this phase the ribosomes are capable of interacting with the ternary complex, EF-1 . GTP . aminoacyl-tRNA, for se- lecting the cognate aminoacyl-tRNA [6, 7, 101. The rest of the ribosomes are in the pre-translocation phase, capable of binding the binary EF-2. GTP complex [6, 7, 101. Dis- turbances in the rate of elongation by translational inhibitors or non-cleavable GTP analogues result in the accumulation of ribosomes in the phase preceding that in which the inhibitor exerts its effect. Thus, inhibition of the EF-1-dependent GTPase results in the accumulation of EF-1 on the ribo- somes due to a reduced turnover of the quaternary EF- 1 . GTP . aminoacyl-tRNA . ribosome complex (Fig. 4). In the case of ricin-treated lysates no accumulation of EF-1 was detected despite the decreased EF-1-dependent GTPase activi- ty (Fig. 4). Obviously, the treated lysate was unable to convert available pre-translocation ribosomes to EF-1-binding post- translocation ribosomes because of the low affinity of the ribosomes for EF-2 (Fig. 4). Thus, the two alterations result in an inability of the system to show an accumulation of an individual intermediate of the elongation cycle.

Both EF-1 and EF-2 catalyse the hydrolysis of GTP to GDP in the presence of ribosomes [I]. The two reactions have been suggested to involve the same ribosomal site [5] since both factors associate with the ribosome at identical or partially overlapping ribosomal domains [6,7]. In prokaryotes the catalytical center of the GTP hydrolysis seems to be located on the elongation factors and not on the ribosome [39]. Less is known about the situation in eukaryotes but affinity labelling studies using reactive GTP analogues suggest that the catalytical center is located on the factor [40]. Ricin altered the activity of both elongation-factor-dependent GTPases. The EF-I-dependent hydrolysis was markedly re- duced while the EF-2-dependent activity was conspicuously increased. The effect on both factor-dependent GTPases in- dicates that a common center on the ribosome was responsible for activating the GTPase function of both factors. The remarkable increase in EF-Zdependent GTPase activity was surprising since the ricin-treated ribosomes were unable to take part in the formation of the stable EF-Zcontaining pre- translocation complex, assumed to be a prerequisite for the translocation-coupled hydrolysis. Since EF-2 was only able to form low-affinity post-translocation complexes with ricin- treated ribosomes, the results indicate that the EF-2-catalysed GTP hydrolysis occurs on ribosomes in the post-translocation phase of the elongation cycle. The ricin-treated ribosomes were still able to bind EF-1 [24] but were not adapted for EF-I-dependent GTP hydrolysis. The results suggest that the individual elongation factors associate with ribosomes adapted for GTP hydrolysis catalysed by the complementary elongation factor. Binding of the factor or the ribosome- associated reaction promoted by the factor would readapt the ribosomal domain responsible for inducing the GTP hydroly- sis and thereby activate the GTPase function of the bound factor.

Non-salt-washed control polysomes showed a consider- able endogenous hydrolysis of GTP. Since this activity was

not influenced by rich treatment the data suggest that this activity was not attributed to endogenous elongation-factor- dependent hydrolysis of GTP. The enzyme responsible for the GTPase activity was, however, released from the polysomes by a gentle salt-wash, indicating that the enzyme was loosely attached to the polysomes. The GTP hydrolysis with both native and salt-washed control polysomes was stimulated by added EF-2. Since the polysomes lacked non-ribosome-bound aminoacyl-tRNA, the ribosomes would be expected to undergo at most one translocation per ribosome [I, 21. The EF-2-induced GTP cleavage should therefore be approx- imately stoichiometric to the number of available pre-translo- cation ribosomes. Surprisingly, both native and salt-washed control polysomes showed a more than stoichiometric hy- drolysis of GTP. This indicates that the polysomes were able to participate in considerably more than one round of EF-2- catalysed GTP hydrolysis.

We are indebted to Mrs Birgit Lundberg for skilful technical assistance. This work was supported by a grant (B-BU 8463-100) from the Swedish Natural Science Research Council.

REFERENCES 1.

2. 3.

4.

5.

6.

7.

8.

9. 10.

11. 12.

13. 14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Brot, N. (1980) in Protein biosynthesis in eukaryotes (Pkrez-

Moldave, K . (1985) Annu. Rev. Biochem. 54, 1109-1149. Drews, J., Bednarik, K. & Grasmuk, H. (1974) Eur. J . Biochem.

Tanaka, M., Iwasaki, K. & Kaziro, Y. (1977) J . Biochem. (Tokyo)

Grummt, F., Grummt, I. & Erdmann, V. A. (1974) Eur. J . Bio-

Nolan, R. D., Grasmuk, H. & Drews, J. (1975) Eur. J . Biochem.

Nombela, C. & Ochoa, S. (1913) Proc. Natl Acud. Sci. USA 70,

Collins, J. F., Moon, H. M. & Maxwell, E. S. (1972) Biochemistry

Richter, D. (1973) J. Biol. Chem. 248,2853-2857. Nygtd, 0. & Nilsson, L. (1984) Eur. J . Biochem. 145, 345-

Nygird, 0. & Nilsson, L. (1984) Eur. J . Biorhem. 140, 93-96. Nyglrd, 0. & Nilsson, L. (1985) Biochim. Biophys. Acta 824,

Olsnes, S. & Pihl, A. (1972) FEBS Lett. 20, 327 - 329. Sperti, S., Montanaro, L., Mattioli, A. & Testoni, G. (1975)

Carrasco, L., Fernandez-Puentes, C. & Vasquez, D. (1975) Eur.

Cawley, D. B., Hedblom, M. L. &Houston, L. L. (1979) Biochem-

Nolan, R. D., Grasmuk, H. & Drews, J. (1976) Eur. J . Biochem.

Olsnes, S . & Pihl, A. (1982) in Molecular action of toxins and viruses (Cohen, L. & van Heyningen, H., eds) pp. 51 - 105, Elsevier Biomedical Press, Amsterdam.

Lugnier, A. A. J. , Dirheimer, G., Madjar, J. - J . , Reboud, J.-P., Gordon, J. & Howard, G. A. (1976) FEBS Lett. 67,343 - 341.

Houston, L. L. (1978) Biochem. Biophys. Res. Commun. 85,131 - 139.

Mitchell, S. J., Hedblom, M., Cawley, D. & Houston, L. L. (1976) Biochem. Biophys. Res. Commun. 68, 163 - 169.

Obrig, T. G., Moran, T. P., Colinas, R. J. & Brown, J. E. (1985) Biochem. Biophys. Res. Commun. 130,879-884.

Benson, S., Olsnes, S. & Pihl, A. (1975) Eur. J . Biochem. 59,573 - 580.

Sperti, S., Montanaro, L., Mattioli, A,, Testoni, G., & Stirpe, F. (1976) Biochem. J . 156, 7-13.

Bercoff, R., ed.) pp. 253-273, Plenum Press, Ncw York.

41,211-227.

82,1035-1043.

chem. 43,343 -349.

50,391 -402.

3556-3560.

11,4187-4194,

350.

152 - 162.

Biochem. J , 148,447-451.

J . Biochem. 54,499 - 503.

istry 18, 2648 - 2654.

64,69 - 75.

117

25. Olsnes, S. & Pihl, A. (1973) Biochemistry 12, 3121 -3126. 26. Pelham, H. R. B. & Jackson, R. J. (1976) Eur. J . Biochem. 67,

27. Nyglrd, 0. & Nika, H. (1982) EMBO J . 1,357-362. 28. Moon, H.-M., Refield, B., Millard, S., Vane, F. C . & Weissbach,

29. Nilsson, L. 8c Nygird, 0. (1985) Eur. J. Biochem. 148,299-304. 30. Bolton, A. E. & Hunter, W. M. (1973) Biochem. J . 133, 529-

31. Holmes, D. S. & Bonner, J. (1973) Biochemistry 12,2330-2338. 32. Nygird, 0. & Hultin, T. (1979) Cancer Res. 39, 3349-3352. 33. Nyglrd, 0. & Hultin, T. (1976) Prep. Biochem. 6, 339-346. 34. Honjo, T., Nishizuka, Y., Kato, I . & Hayaishi, 0. (1971) J . Biol.

247 - 256.

H. (1973) Proc. Natl Acad. Sci. USA 70, 3282-8386.

538.

Chem. 246,4251 -4260.

35. Nyglrd, 0. & Hultin, T. (1978) Chem.-Bid. Interactions 20,149- 161.

36. Fasman, G. D. (ed.) (1976) Handbook of biochemistry and tnolw ular biology: Proteins, vol. 111, pp. 441-455, CRC Press, Cleveland, Ohio.

37. Montanaro, L., Sperti, S. & Stirpe, F. (1973) Biochem. J . 136,

38. Benson, S . , Olsnes, S., Pihl, A. Skorve, J. & Abraham, K. A.

39. Girshovich, A. S., Pozdnyakov, V. A. & Ovchinnikov, Y. A.

40. Nilsson, L. &Nygird, 0. (1984) Biochim. Biophys. Acta 782.4Y -

677 - 683.

(1975) Eur. J . Biochem. 59, 573 - 580.

(1976) Eur. J . Biochem. 69, 321 -328.

54.