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

On the Mechanism of Protein-Synthesis Inhibition by Abrin and Rich Inhibition of the GTP-Hydrolysis Site on the 60-S Ribosomal Subunit

Stephen BENSON, Sjur OLSNES, and Alexander PlHL

Norsk Hydros Institutt for Kreftforskning. Oslo

Jon SKOKVE and Abraham K. ABRAHAM

Biokjemisk Institutt, Cellebiologisk Forskningsgruppe. Universitetet i Bergen

(Received March lO/July 7, 1975)

The mechanism of protein synthesis inhibition by the toxic lectins, abrin and ricin, has been studied in crude and in purified cell-free systems from rabbit reticulocytes and Krebs 11 ascites cells.

In crude systems abrin and rich strongly inhibited protein synthesis from added aminoacyl-tRNA, demonstrating that the toxins act at some point after the charging of tRNA.

Supernatant factors and polysomes washed free of elongation factors were treated separately with the toxins and then neutralizing amounts of anti-toxins were added. Recombination experiments between toxin-treated ribosomes and untreated supernatant factors and vice versa showed that the toxin-treated ribosomes had lost most of their ability to support polyphenylalanine synthesis, whereas treatment of the supernatant factors with the toxins did not inhibit polypeptide synthesis.

Recombination experiments between toxin-treated isolated 40-S subunits and untreated 60-S subunits and vice versa showed that only when the 60-S subunits had been treated with the toxins was protein synthesis inhibited in the reconstituted system. The incorporation of [3H]puromycin into nascent peptide chains was unaffected by the toxins, indicating that the peptidyl transferase is not inhibited.

Both the EF-1-catalyzed and the EF-2-catalyzed ability of the ribosomes to hydrolyze [ Y - ~ ~ P I G T P was inhibited by abrin and ricin.

An 8-S complex released from the 6 0 4 subunit by EDTA treatment possessed both GTPase and ATPase activity, while the particle remaining after the EDTA treatment had lost most of its GTPase activity. Both enzyme activities of the 8-S complex were inhibited by abrin and ricin.

The present data indicate that there is a common site on the 6 0 4 subunits for EF-1- and EF-2- stimulated GTPase activity and they suggest that abrin and ricin inhibit protein synthesis by modifying this site.

In previous papers we have reported that the toxic plant proteins abrin and ricin inhibit protein synthesis in eucaryotic cells by interfering with peptide chain elongation [1,2]. Both abrin and ricin consist of two polypeptide chains, the A-chain or ‘effectomer’ and the B-chain or ‘haptomer’ [3]. The function of the B-chains is to bind the toxins to the cell surface and it is, therefore, necessary for the cellular uptake of the toxin. The ability of the toxins to inhibit protein synthesis in cell-free systems is associated exclusively with the A-chains [4,5]. Experiments in our laboratory [6] as well as by Montanaro et al. [7] indicate that the ribosomes are the targets for abrin and ricin. In contrast to many antibiotics, which inhibit protein

synthesis by being bound to the ribosomes [8], abrin and ricin inactivate the ribosomes by a catalytic mechanism ([6] and unpublished data).

In the present study experiments were performed to identify the step in peptide chain elongation which is inhibited by abrin and ricin and to elucidate in more detail the mechanism of action of these toxins.

MATERIALS AND METHODS

Toxins

Abrin and ricin and their A-chains were prepared as earlier described [4,5].

514 Inhibition of Ribosome-Dependent GTPase

Preparation of a Crude Cell-Free Protein-Synthesizing System from Rabbit Reticulocytes

The system was essentially an unfractionated reticulocyte lysate prepared according to Lingrel [9] as previously described [l].

Preparation of Ribosomes and Their Isolated Subunits from Rabbit Reticulocytes

Salt-washed ribosomes were isolated by layering 4 ml of reticulocyte lysate over 1 ml of 0.5 M sucrose in buffer A [lo mM triethanolamine (pH 7.5), 0.15 M KCl, 1 mM MgC12] and centrifuging in an SW-50.1 rotor for 2 h at 234000 x g . The pellet was suspended in a small volume of buffer A and the ribosomes were washed with high-salt buffer according to Felicetti and Lipmann [lo].

To isolate ribosomal subunits, a crude cell-free protein-synthesizing system was prepared as above and incubated for 20 min at 37 "C. Puromycin-HC1 (CalBioChem) was added to a final concentration of 1 mM and the incubation was continued for an addi- tional 40 min. The mixture was then layered on to 1 ml of 1 M sucrose in buffer A and centrifuged at 234000 x g in an SW-50.1 rotor at 2 "C for 2 h. The pelleted ribosomes were resuspended in 0.3 M KCl, containing 3 mM MgC12, 10 mM triethanolamine (pH 7.9, and 1 mM dithiothreitol and layered on to 13-ml 10-30%, (w/v) sucrose gradients in the same buffer.

The gradients were centrifuged for 4.5 h at 201 000 x g in an SW-40 Ti rotor and fractionated on a Micro Radiograph fractionator coupled to a Gilford 240 spectrophotometer (Fig. 1 A). The fractions containing the separated 40-S and 60-S subunits were pooled and diluted with an equal volume of 0.3 M KCl, containing 3 mM MgC12, 10 mM triethanolamine (pH 7.5) and 1 mM dithiothreitol, and concentrated by centrifugation at 234000 x g for 5 h in an SW-50.1 rotor. The subunits were resuspended in 0.1 M NH4C1 containing 2 mM magnesium acetate, 20 mM Tris-HC1 (pH 7.5) and 1 mM dithiothreitol at a con- centration giving an absorbance at 260 nm of 70 and 45 for the 60-S and the 40-S subunits respectively.

Preparation 018-S Complex from Reticulocyte Ribosomes

A ribonucleoprotein complex consisting of 5-S RNA and one ribosomal protein was released from the 60-S ribosomal subunits, as described by Grummt et al. [ l l ] with some modifications. Poly- somes were isolated from a rabbit reticulocyte lysate by layering the lysate on to 1 ml of 1 M sucrose in buffer A and centrifuging at 234000 x g in an SW-50.1

rotor for 2 h at 2 "C. The incubation with puromycin (1 mM) lasted 20 rnin and the subunits were separated by sucrose gradient centrifugation in rotor SW-27 for 9.5 h at 94000 x g. The fractions containing 60-S subunits were pooled and pelleted by centrifugation at 234000 x g in an SW-50.1 rotor for 5 h at 2 "C. 1.4 pmol EDTA was added/mg of 60-S subunits, and the mixture was centrifuged in a 5-20'j/, linear sucrose gradient in 20 mM Tris-HCI (pH 7.9, 6 mM 2-mercaptoethanol and 10 mM KCI for 24 h at 201 000 x g in an SW-40, Ti rotor. The gradients were fractionated and the peak fraction sedimenting at about 8 S (Fig. 1 B), as calculated from the sedimenta- tion tables of McEwan [12], was collected and assayed for GTPase and ATPase activity.

Preparation of Ribosomes ,from Krebs Ascites Cells

Krebs I1 ascites cells were propagated in BALB/c mice. The cells were washed twice in 35 mM Tris-HC1 (pH 7.6) containing 0.14 M NaCl and then resuspend- ed in buffer B containing 5 % sucrose. The cells were disrupted by nitrogen cavitation as described by Dowben et al. [13] and then centrifuged for 15 min at 30000 x g. The supernatant was used for the isola- tion of tRNA. To prepare ribosomes the pellet was resuspended in buffer B containing 5 sucrose. Then 1/5 volume of 2.5 M NH4Cl was added and the mixture was layered on to 2 ml of 20% sucrose and centrifuged for 90 min at 100000 xg. This washing was repeated once, and the final ribosome pellet was resuspended in buffer B. The washed polysomes were stored at - 20 "C until use.

Preparation of Elongation Factors

EF-1 and EF-2 were prepared from a reticulocyte lysate pH-5 supernatant as described by Felicetti and Lipmann [lo] and from a postribosomal supernatant obtained from Krebs IT ascites cells, essentially as described by Moldave et al. [14] and by Felicetti and Lipmann [lo].

Poly- U-Directed Incorporation of Phenylalanine into Polypeptides

Poly-U-directed polyphenylalanine synthesis by ribosomes or by isolated and reassociated subunits was assayed in 0.25 ml of reaction medium containing 100 mM NH4Cl, 12 mM magnesium acetate, 20 mM Tris-HC1 (pH 7.5),1 mM dithiothreitol, 0.2 mM GTP, 200 pg [l4C]phenyla1anyl-tRNA, 300 pg poly(U), 60pg EF-1, 200 pg EF-2, 18 pg 40-S ribosomal subunits, and 42 pg 60-S subunits. Incubation was carried out at 30 "C and at various times 2 0 4 aliquots were withdrawn into 0.1 M KOH and heated to 37 "C for 20 min. Samples were then adjusted to 10"; (w/v) trichloroacetic acid and acid-insoluble radioactivity

S. Benson, S. Olsnes, A. Pihl, J. Skorve, and A. K. Abraham 575

0.4

0.3

E 8 *

s 0.2

n a

0.1

n I - 0 bottom

3

L tom top

Fig. 1. Sucrosr gradient centri$gation oj’rrticulocyte ribosomal suhunit.s and 8-S comp1e.r. (A) A cell-frcc systcm was treated with puromycin and the isolated ribosomes concentrated by centrifugation and resuspended in buffer containing high salt concentration as described in Materials and Methods. The ribosomal subunits were then layered on to a 10- 30 ”/, sucrose gradient and centrifuged at 201 000 x g in rotor SW-40, Ti for 4.5 h. The peak fractions containing 40-S and 60-S subunits were pooled as indicated. (B) Isolated 60-S subunits were treated with EDTA and layered on to a 5-20% sucrose gradient and centrifuged at 201000 x g for 24 h. The inserted picture represents polyacryl- amide gel electrophoresis of the protein moiety (left) and the RNA moiety (right) of the 8-S complex

was collected on glass-fiber filters (Whatman 6F/c) and washed with 5 % trichloroacetic acid. The filters were dried and the radioactivity was measured by liquid-scintillation counting as earlier described [15].

Preparation of Specific Antibodies

Rabbits were immunized with toxoids prepared from abrin and ricin by formaldehyde treatment as earlier described [4,5]. The crude antisera were purified by passage through a column of Sepharose particles to which the pure A-chains had been covalently bound. The adsorbed antibodies were subsequently eluted at pH 2.8 as earlier described [16]. The fractions were immediately neutralized with concentrated Tris, and concentrated by precipitation with 60 % ammo- nium sulphate. The precipitated antibodies were dissolved in a small volume of water and dialyzed against 10 mM Tris-HC1 (pH 7.7) containing 0.14 M NaCl.

Polyacrylamide Gel Electrophoresis

The 8-S complex was treated with 1% sodium dodecylsulphate and the protein moiety was analyzed

by polyacrylamide gel electrophoresis in the presence of sodium dodecylsulphate [4]. The RNA moiety was studied by electrophoresis in 5 polyacrylamide gels according to Peacock and Dingman [17].

Assay of GTPase and ATPase Activity

The assay system (1OOp1) contained 200 pg of N-ethylmaleimide-treated ribosomes, 10 pg of EF-1 or EF-2, 50 pmol [Y-~’P]GTP, 5 mM magnesium acetate, 95 mM KC1, 20 mM Tris-HC1 (pH 7.6) and 5 mM 2-mercaptoethanol.

To test ATPase and GTPase activity of the 8-S complex the assay was carried out in a total volume of 100 p1 containing 3 pg 8-S complex, 50 pmol of [p3’P]ATP or [Y-~’P]GTP (spec. act. 50 counts min-’ pmol-I), 3 mM sodium acetate, 7 mM KCI, 20 mM Tris-HCI (pH 7.8) and 6 mM 2-mercaptoethanol. The mixtures were incubated at 37 “C for the indicated periods of time, and the reaction was stopped by the addition of 1 ml of a mixture containing 5 % Norite, 5 % trichloroacetic acid and 5 mM sodium phosphate to reduce the binding of inorganic phosphate to the charcoal. The mixture was vigorously shaken for 30 s

516 Inhibition of Ribosome-Dependent GTPase

on a Rotamix and then centrifuged for 5 min at 10000 xg. Samples (200 pl) were removed from the super- natant and the radioactivity was measured by scintilla- tion counting. In all cases the amounts of 32P liberated increased linearly during the incubation times used.

So Iut ions

Buffer A : 0.15 M KCI, 1 mM MgClz and 10 mM triethanolamine (pH 7.5). Buffer B: 0.1 M KC1, 5 mM magnesium acetate, 0.1 mM EDTA, 6 mM 2-mer- captoethanol and 20 mM Tris-HC1 (pH 7.6).

Labeled Substances

tRNA charged with 14 different 14C-labeled amino acids was prepared essentially as described by Har- desty et al. [18]. The specific activity was 192 nCi/mg. [U-'4C]leucyl-tRNA (spec. act. 355 nCi/mg) and [U-14C]phenylalanyl-tRNA (spec. act. 192 nCi/mg) were obtained from New England Nuclear (Boston). [8-3H]puromycin dihydrochloride (spec. act. 3.7 Ci/ mmol), [p3'P]GTP (spec. act. 1.56 Ci/mmol) and [y-32P]ATP (spec. act. 2.1 Ci/mmol) were obtained from Amersham (England).

RESULTS

Since we have earlier shown that the ability of abrin and ricin to inhibit protein synthesis in a cell- free system is associated exclusively with their A-chains [4,5], we have used here the purified A-chains rather than the complete toxins.

lnhihition by Abrin and Ricin of Polypeptide Synthesis from Aminoacyl-tRNA

When A-chains from abrin and ricin were added to a cell-free protein-synthesizing system prepared from a rabbit reticulocyte lysate, the incorporation of ['4C]leucine from ['4C]leucyl-tRNA was strongly inhibited (Fig. 2). Since this system contained saturat- ing concentrations of the other 19 amino acids in the form of unlabeled aminoacyl-tRNA, the data demon- strate that the inhibition of polypeptide synthesis by abrin and ricin takes place at some step after the charging of tRNA.

Selective Pretreatment of Ribosomes arid Soluble Factors with Abrin and Rich

Previous experiments, where ribosomes and post- ribosomal supernatant from a rabbit reticulocyte lysate were treated separately with abrin and ricin, indicated that the toxins inhibit protein synthesis by inactivating the ribosomes rather than some factor in the supernatant [6]. However, since high-

I I I I

;ib I

R i c h A-chain

5 10 15 2U Time (min)

R i c h A-chain

5 10 15 2U Time (min)

5

Fig. 2. Ej'ecr of abrin and ricin A-chains on protein synthesis from aminoacyl-tRNA. A cell-free system from rabbit reticulocytes was prepared as described in Materials and Methods. The incubation mixture contained in a final volume of 0.5 ml: 0.2 ml lysate, 10 mM Tris-HCI (pH 7.4), 100 mM ammonium acetate, 2 mM magnesium acetate, 1 mM ATP, 0.2 mM GTP, 15 mM creatine phosphate, 50 pg/ml creatine phosphokinase and unlabelled aminoacyl-tRNA (300 pg) and 0.02 pCi of ['4C]leucyl-tRNA. One sample (0) was used as control, whereas to the other ones were added (A) 100 ng of abrin A-chain and ( x ) 100 ng of rich A-chain

salt wash of the ribosomes had not been carried out in these experiments, the possibility could not be ruled out that the toxins acted on ribosome-bound elonga- tion factors. In the present work reticulocyte ribo- somes were, therefore, isolated and washed several times with buffer containing 0.5 M NH4C1. That this procedure removed all elongation factors follows from the finding that when the washed ribosomes were incubated in a protein-synthesizing system containing GTP, poly(U) and ['4C]phenylalanyl-tRNA no in- corporation of radioactivity into polypeptides was observed in the absence of added EF-1 and EF-2 (data not shown). Experiments were then carried out where ribosomes and supernatant factors were sepa- rately preincubated in the presence and absence of A-chains from abrin and ricin. Neutralizing amounts of anti-A-chain antibodies were subsequently added to each preparation and treated ribosomes were mixed with untreated supernatant factors, and vice versa. Complete protein-synthesizing systems were recon- stituted by adding poly(U) and the incorporation of [14C]phenylalanine was measured. It is evident (Fig. 3) that the rate of polyphenylalanine synthesis was almost the same whether or not the supernatant factors had

S. Benson, S. Olsnes. A. Pihl, J. Skorve, and A. K. Abraham

15

517

I I 1 I A

Sup.untr. Rib.untr.

X H X

- X-

I I I I

B Sup.untr. Rib. untr.

X

/

Sup.untr. Rib.ric.

Rib.ric. Sup.ric.

: /.' I=&'

I I I I

10 20 30 40 Time (rnin)

Fig. 3. Polyphenylalanine synthesis after separate treatment of ribosomes and supernatant factors from rabbit reticulocytes with A-chains from abrin ( A ) and ricin ( B i . Salt-washed reticulocyte ribosomes (0.4 mg) in 60 p1 of 50 m M Tris-HC1 (pH 7.5) containing 2 mM MgC1, and 1 mM dithiothreitol were incubated in the presence and absence of 60 ng of abrin and ricin A-chains at 37 'C for 10 min. Similarly a mixture of 300 pg of poly(U), 50 pg of EF-1, 150 pg of EF-2,200 pg of [14C]phenylalanyl-tRNA and 0.5 pmol G T P were incubated in a volume of 120 pI with and without abrin and ricin A-chains (60 ng) at 37 "C for 10 min. After the incubation 7 pg of specific antibodies was added to all tubes. Toxin- treated ribosomes were then recombined with untreated soluble factors and vice versa, and the rate of polyphenylalanine synthesis was measured in the reconstituted system. ( x ) Ribosomes and soluble factors not treated with toxins; (0) ribosomes untreated, soluble factors toxin-treated ; (A) ribosomes toxin-treated, soluble factors untreated; (0) ribosomcs and soluble factors toxin-treated

been preincubated with toxins. On the other hand, when the ribosomes had been preincubated with toxins, the incorporation of phenylalanine was strongly inhibited. Control experiments demonstrated that saturating amounts of anti-A-chain antibodies im- mediately neutralized even high amounts of A-chains. The results confirm our previous conclusion that the toxins act on the ribosomes rather than on soluble factors. The results also show that the effect of the A-chains is irreversible. Thus, neutralization of the A-chain activity with antibodies did not reactivate the toxin-treated ribosomes. This is consistent with the view [3,6,19] that the A-chains possess enzymic activity.

Effect on Ribosomal Subunits

To elucidate which of the ribosomal subunits is attacked by the toxins, we exposed separately 40-S and 60-S subunits to toxin A-chains, added neutraliz- ing amounts of antibodies and then tested the ability of the reconstituted ribosomes to polymerize phenyl- alanine in the reconstructed system containing poly(U). The results in Fig. 4 show that toxin-treated 40-S

particles, when combined with untreated 60-S sub- units, supported incorporation of [14C]phenylalanine into acid-precipitable material to nearly the same extent as when untreated 40-S subunits were used. On the other hand, 60-S particles pretreated with toxins had lost most of their activity when tested together with untreated 40-S particles. In other experiments we incubated intact lysate with toxins under conditions as in Fig. 2 and then prepared 40-S and 60-S subunits from the toxin-treated ribosomes. These subunits were then incubated with each other or with subunits from untreated ribosomes under conditions as in Fig. 4. The results (not shown) were similar to those in Fig. 4. The data therefore indicate that the site of action of abrin and ricin is the 6 0 4 subunits, confirming the conclusion of Sperti et al. [20] that the 60-S subunit is the site of action of ricin.

Effect o j Abrin and Ricin on Peptidyl Transferase

The central reaction taking place on the 60-S sub- unit is the formation of the peptide bond [21]. To see whether abrin and ricin inhibit the peptidyl transfer- ase, we studied the incorporation of [3H]puromycin

578 Inhibition of Ribosome-Dependent GTPase

~~

I I

A

/ 40 S - Abrin

I I

B

A 40 s 60 S - Ricin

40 S - Ricin 60 S -Rich

I I I I I I

Time ( r n i n ) Fig. 4. E//ect of A-chains,from abrin ( A ) and ricin ( B ) on isolated ribosomal subunirs/rom rabbit ruticulocytes. 404 and 60-S subunits were incubated separately with and without 60 ng of abrin and ricin A-chains and then 7 pg of specific antibodies was added. Subsequently the 40-S and 60-S subunits were recombined as described in Materials and Methods. Poly(U), ['4C]phenylalanyl-tRNA, EF-I, EF-2 and GTP were added, and the rate of polyphenylalanine synthesis was measured. ( x ) Both subunits untreated; (0) 40-S subunits pretreated with toxin A-chain. 6 0 3 subunits untreated: (A) 6 0 3 subunits Dretreated with toxin A-chain, 404 subunits untreated; (0) both subunits toxin-treated; . (m) untreated 404 subunits alone; (0) untreated 60-S subunits alone

Control - Ricin Abrin

-

Control - Ricin Abrin

-

-

1 1 I I I 1 0 2 4 6 a 10 12

Time (min) Fig. 5. Elyecr q/' ahrin and ricin A-chains on the incorporution oJ purornycin inro nascent peptide chains. Rabbit reticulocyte ribosomes (0.15 mg) were incubated in a 0.5-ml reaction mixture containing 50 mM Tris-HC1 (pH 7.8), 10 m M MgCI,, 2 nmol of [3H]puro- mycin (5x lo5 counts/min) and either 60 mM KC1 (open symbols) or 0.6 M KCI (filled symbols). The samples were incubated at 37 "C, 10O-pl aliquots were removed at the times indicated, and the acid-precipitable radioactivity was measured. (0, 0) No toxin added; (0, m) 100 ng abrin A-chain added; (A, A) 100 ng ricin A- chain added

into acid-precipitable material [22] in the presence and absence of abrin and ricin. The data in Fig. 5 show that the toxin A-chains had no effect on this reaction, either at low (60 mM) or at high (600 mM) concentra- tion of KCl. It may, therefore, be concluded that abrin and ricin do not inhibit the formation of the peptide bond.

Effect of Abrin and Ricin on Ribosome-Dependent GTPase Activity

It is well known that the ribosome-dependent GTPase activity requiring the elongation factors EF-1 or EF-2 is localized on the 60-S ribosomal sub- unit [23]. The data in Table 1 show the effect of abrin and ricin A-chains on the EF-1 -catalyzed ribosome- dependent GTPase activity. It is clear that both toxins exerted a marked inhibition of the GTP hydrolysis in this system. Also the EF-2-stimulated ribosome-dependent GTPase activity was inhibited by abrin and rich (Table 2). The data thus show that the toxins inhibit both the EF-1 and the EF-Zstimulat- ed ribosome-dependent GTPase activity.

Effect of Abrin and Rich on the 8-S Complex Isolated from the Large Ribosomal Subunit

Recently an 8-S complex containing 5-S RNA and a single protein (Fig. 1B) has been isolated from the

S. Bcnson. S. Olsnes. A. Pihl, J. Skorve, and A. K. Abraham 579

Table 1. EJyeci of ahrin and ricin A-chains on EF-I-catalyzed ribo- .same-dcpendeni GTPase activity Ribosomes and EF-1 were isolated from Krcbs TI ascites cells as described in Materials and Methods. The complete system contained in a total volume of 100 pl: 125 pg of N-ethylmaleimide-treated ribosomes or 6 0 3 subunits. 10 pg EF-1, 25 pg unlabcllcd amino- acyl-tRNA and 5 nmol [y-3zP]GTP (50 counts min-' pmol-') in 20 mM Tris-HCI (pH 7.6) containing 5 mM magnesium acetate, 95 mM KCI and 5 mM 2-mercaptoethanol. Abrin and ricin A- chains were prcscnt in final concentrations of 0.2 pg. The mixture was incubated at 37 "C for 15 min and the reaction was stopped by the addition of 1 ml of a mixture containing 5 % Norite, 5% tri- chloroacetic acid and 5 mM sodium phosphate to reduce the binding of inorganic phosphate to the charcoal. The mixture was vigorously shaken for 30 s on a Rotamix and then centrifuged for 5 min at 1OOOOxg. Samples (200 pl) were removed from the supernatant and the radioactivity was measured by scintillation counting

Assay system Conditions GTP hydrolyzed - -. - ____ no toxin abrin ricin

pmol . ...

Krebs 11 ascites complete system 594 335 338 cells -ribosomes 115 105 117 (ribosomes) - EF-1 30 30 30

- tRNA 236 242 23 1

Krebs 11 ascites complete system 1068 701 -

cells - tRNA 507 504 ~

(60-subunits)

Reticulocytes complete system 754 612 679 (ribosomes) -ribosomes 190 198 -

- EF-1 95 103 -

- tRNA 285 343 327

60-S ribosomal subunit by treatment with EDTA [11,24]. This complex has been shown to have some GTPase activity [l l] . Our data indicate that this activity is not stimulated by either EF-1 or EF-2 (data not shown). The 50-S particle remaining after treatment of the 60-S subunit with EDTA does not hydrolyze GTP, either in the presence or absence of elongation factors. Since it thus appears that the 8-S complex carries the GTPase site of the 60-S subunits, we studied the effect of A-chains on the GTPase activity of this complex. The data in Table 3 confirm that the 8-S complex indeed contains a low GTPase activity and they demonstrate that abrin and ricin A- chains inhibit the GTPase activity of the complex. Interestingly, the 8-S complex also catalyzes the hydrolysis of ATP, an activity which is inhibited by abrin and ricin A-chains to an even greater extent than the GTPase activity. Similar results were obtained by Grummt et al. [l 11 with fusidic acid. It is interesting that the extent of the inhibition by abrin and ricin A- chains is similar to that found with fusidic acid. The data thus suggest that the toxins inactivate the 60-S. ribosomal subunits by interfering specifically with that part of the subunit which may be released as an 8-S complex upon treatment with EDTA.

Table 2. Euect of abrin and ricin A-chains on LF-2-catalyzed GTPase activity of salt-washedpolysomes The complete system contained 200 pg of N-cthylmaleimide-treated polysomes and 10 pg EF-2, 5 nmol [y3'P]GTP (50 counts min-' pmol-I) in 100 pl of 20 mM Tris-HCI (pH 7.6) containing 50 mM magnesium acetate, 95 mM KCI, and 5 mM 2-mercaptoethanol. Abrin and ricin A-chains were present in final concentrations of 100 ng/ml. In the Krebs I1 ascites system GTP hydrolysis was measured as in Table 1. In the reticulocyte system the incubation lasted for 90 min

Assay system Conditions GTP hydrolyzed

no toxin abrin ricin

pmol

Krebs I1 ascites complete system 1142 611 814

- EF-2 33 32 28

Reticulocytes complete system 3798 2052 2610

- EF-2 564 462 3 30

cells - polysomes 105 275 215

- polysomes 330 318 282

Table 3. Efleect of abrin and ricin A-chain on the GTPase and ATPase artivity of ihe 8-S complex The system contained 3 pg 8-S complex from reticulocytes, 5 nmol of either [7-32P]GTP or ['J-~'P]ATP (50counts min-' pmol-') in 100 p1 of 20 mM Tris-HCI (pH 7.8), containing 3 mM MgCI,, 7 mM KCI and 6 mM 2-mercaptoethanol. Abrin and ricin A-chains were present in final concentrations of 100 ng/ml. GTP and ATP hydrolyses were measured as in Table 2, except that the incubation time was 3 h instead of 15 min. The blank value obtained in the absence of 8-S complex was subtracted from all values

Expen- Conditions Hydrolysis of ment no GTP ATP

~ ~

pmol %of control

1 control 504 100 abrin 192 38

2 control 1264 100 abrin 460 36

3 control 1428 100 ricin 440 31

pmol y c , of

348 100 44 13

control

DISCUSSION

In earlier papers we have presented evidence that the A-chains of abrin and ricin inactivate eucaryotic ribosomes by an enzymatic mechanism [6,19]. Re- cently, Montanaro et al. [7] have found that ricin inhibits the EF-2-stimulated ribosome-dependent GTPase activity. Moreover, ricin was found to inhibit the 60-S subunits of the ribosomes [20]. The present results confirm and extend these findings. We have shown that the same situation exists with abrin. Moreover, our data demonstrate that also the EF-1-

580 S. Benson, S. Olsnes, A. Pihl, J. Skorve. and A. K . Abraham: Inhibition of Ribosome-Dependent GTPase

stimulated ribosome-dependent GTPase activity is inhibited by abrin and ricin A-chains. It is particularly interesting that abrin and ricin A-chains inhibit the GTPase activity of the 8-S complex, which can be isolated from the 60-S subunit by treatment with EDTA. Since no GTPase activity remained in the large ribosomal subunit after release of the 8-S com- plex, even in the presence of elongation factors, it is likely that there is a common site on the 60-S subunit for EF-1 and EF-2-stimulated GTPase activity and that upon treatment with EDTA this site is released in the 8-S complex. The 5-S RNA appears to be somehow involved in the enzymatic function of the complex, since treatment with ribonuclease strongly reduces its GTPase activity (data not shown). The fact that the GTPase activity of the complex can be sup- pressed by abrin and ricin indicates that either the 5-S RNA or the single ribosomal protein in the 8-S complex is the target for the toxins. By polyacrylamide gel electrophoresis we have not been able to detect any reduction in the size of the RNA or the protein moiety, and it is possible that the toxins induce only a minor modification of the 8-S complex.

The finding that the EF-1-dependent GTPase activity of the ribosomes is inhibited by abrin and ricin is in accordance with the findings of Carrasco et al. [25] that the toxins inhibit the enzymic binding of aminoacyl-tRNA to ribosomes.

The GTPase activity appeared to be less sensitive to high concentrations of abrin and ricin than protein synthesis. However, in control experiments it was found that the toxin concentration needed to give half-maximal inhibition was about the same in the two systcms ( 5 - 10 ng/ml), suggesting that the con- centration dependence of the inhibitory effects was the same. With the amounts of A-chain used in the present experiments (0.1 - 1 pg/ml) close to maximum inhibi- tion was obtained for both activities and an additional increase in the A-chain concentration did not further reduce either of the activities. The fact that the maximal inhibition of protein synthesis, expressed as a percent- age of control, was greater than that of ribosomal GTPase is perhaps not surprising, as the GTP hydro- lysis here measured is uncoupled from polypeptide

synthesis. It is conceivable that the toxin-induced modification of the GTPase site interferes with the coupling and thus contributes also in this way to the inhibition of polypeptide synthesis. Studies along this line are now in progress.

This work was supported by the Norwegian Cancer Society.

REFERENCES 1. Olsnes, S. & Pihl, A. (1972) FEBS Lett. 20, 327-329. 2. Olsnes, S. & Pihl, A. (1972) Nature (Lorid.) 238, 459-461. 3. Olsnes, S. , Refsnes, K. & Pihl, A. (1974) Nafure (Lond.) 249,

4. Olsnes, S. & Pihl. A. (1973) Eur. J . Biochem. 35. 179-185. 5. Olsnes, S. & Pihl, A. (1973) Biochemisrry, 12, 3121 - 3126. 6. Olsnes, S., Heiberg, R. & Pihl, A. (1973) Mol. Biol. Rep. I ,

627- 631.

I.

8. 9.

10.

11.

12. 13.

14.

15. 16.

11.

18.

19. 20.

21.

77 I-.

23.

24. 25.

15-20. Montanaro, L.. Sperti, S. & Stirpe. F. (1973) Biachem. J. 136,

Pestka, S. (1971) Annu. Re) , . Microhiol. 25, 481 -562. Lingrel, J . B. (1972) in Methods in Protein Biosynthesis (Laskin,

A. E. & Last, J. A,, eds) vol. 2, p. 231, Marcel Dekker. Ncw York.

Felicetti, L. & Lipmann. F. (1968) Arch. Biochem. Biophys.

Grummt, F., Grummt, 1. & Erdmann. V. (1974) Eur. J. Bio-

McEwan, C . R. (1967) Anal. Biochem. 20, 114- 149. Dowben, R. M., Lynch, P. M., Nadler, H. L. & Hsia. D. Y.

Moldavc, K., Galasinski, W. & Rao, P. (1971) Merhod,r En:?-

Olsnes, S. (1970) B ~ o c ~ I . Biophys. Aria, 213. 149- 158. Pappenheimer, A. M.. Jr, Olsnes. S. & Harper, A. A. (1974)

Peacock, A. C. & Dingman, C. W. (1968) Biochcmistry, 7.

Hardesty, B., McKeehan, W. & Culp, W. (1971) Muthody

Olsnes, S. & Pihl, A. (1972) FERS Lett. 28, 48 - 50. Sperti, S.. Montanaro, L., Mattioli, A. & Stirpe. F. (1973)

Biochem. J. 136, 813-815. Vazquez, D., Battaner, E., Neth, R., Heller, G . & Munro,

R. E. (1969) Cold Spring Harbor S y p . Quanr. Biol. 34.

Baligd, B. S. , Schechtman, M. G.. Nolan, R. D. & Munro.

McKeehan, W. (1972) Biochem. Bi0phy.v. Rex. Co/nrnun. 48,

Blobel, G. (1971) Proc. Nut / Acad. Sci. U .S .A . 68, 1881 - 1885. Carrasco, L., Fernandez-Puentes, C. & Vazqucz, D. (1975)

671-683.

125,548-557.

chem. 43.343 - 348.

(1969) EXP. Cell. R ~ s . 58, 167- 169.

mol. 20, 337- 348.

J. Im~nunol. 113, 835 - 841.

668 - 674.

Enzymol. 20, 316- 330.

369-375.

H. N. (1973) Biochim. Biophys. . 4 C l a , 312, 349-357.

11 17 - 1122.

Eur. J . Biochem. 54,499- 503.

S. Bcnson, Department of Biological Sciences, California State University, Hayward, California, U.S.A. 94542

S. Olsnes and A. Pihl, Norsk Hydros Institutt for Kreftforskning. Det Norske Radiumhospitalet. Montebello,

J . Skorve and A. K. Abraham, Biokjemisk Institutt. Universitetet i Bergen, Arstadveien 19, N-5000 Bergen. Norway

Oslo 3. Norway