The Diphtheria Toxin-dependent Adenosine Diphosphate Ribosylation of Rat Liver Aminoacyl Transferase II GENERAL CHARACTERISTICS AND MECHANISM OF THE REACTION*

to catalyze a reversible reaction in which the adenosine diphosphate ribose moiety of NAD is transferred to aminoacyl transferase II, resulting in the concomitant of transferase activity in protein synthesis.

loss of transferase activity in protein synthesis. Some general characteristics of the reaction are reported and a mechanism is proposed. The rate of the reaction is proportional to toxin, aminoacyl transferase II, and NAD concentrations, but, in the presence of excess NAD, the extent of the reaction is dependent only on the amount of aminoacyl transferase II present. ADP ribosylation occurs at all temperatures between 040" with a maximum at 20". The adenosine diphosphoribosylaminoacyl transferase II bond is stable to heating at 90" for 15 min in 5 % trichloracetic acid. The following random mechanism for the reaction is proposed on the basis of results from kinetic studies, use of inhibitors of the reaction and isotope exchange reactions, as well as our failure to find a partial reaction.
A ternary intermediate composed of toxin, NAD, and aminoacyl transferase II is formed by a series of different binary reactions. The results support the participation of all combinations of binary reactions in the formation of the ternary complex. Toxin is bound to the adenosyl moiety of NAD, as well as to the transfer enzyme. Aminoacyl transferase II is bound to the NMN moiety of NAD.
The proposed ternary intermediate consisting of toxin, aminoacyl transferase II, and NAD breaks down rapidly to the final products, adenosine diphosphoribosylaminoacyl transferase II, nicotinamide, and toxin.
Diphtheria toxin inhibits the incorporation of amino acids into protein by mammalian cells (1)  from these cells (2). NAD is required specifically for this in hibition in vitro (2, 3). The site of action of toxin and NAD is aminoacyl transferase II (4), a soluble enzyme required together with aminoacyl transferase I and ribosomes for transfer of amino acids from aminoacyl transfer RNA into protein.
Transferase 111 has been shown (5)(6)(7)(8) to function in polypeptide chain extension by catalyzing translocation of the nascent chain from the unreactive acceptor (aminoacyl) site on the ribosome to the donor (peptidyl) site in which the chain can again react with an incoming aminoacyl-tRNA or with puromycin to form a peptide bond (9-11).
Translocase activity of transferase II, independent of transferase I and aminoacyl-tRNA, is measured by a GTP-dependent stimulation of the number of polypeptide chains that can react with puromycin (5). Toxin and NAD inhibit the stimulation of the puromycin reaction by transferase II (5).
Like the analogous G factor in Escherichia co& transferase II aIso exhibits a ribosome-dependent GTPase activity (6,12). Toxin and NAD inhibit this activity (6). Common inhibition by toxin and NAD of transferase II-dependent GTPase activity, translocase activity, and amino acid incorporation has led to the conclusion that these activities are catalyzed by a single protein in partially purified transferase II or that they share a requirement for such a protein.
The inactivation of transferase II can be reversed by toxin and high concentrations of nicotinamide (13)(14)(15) and recent results indicate that toxin exerts its effect by catalyzing the following reversible reaction (14,15).
The ADP ribosylation of transferase II intimately parallels inactivation of transferase II in amino acid-incorporating systems (15) and reversal of ADP ribosylation by nicotinamide results in restoration of transferase II activity (15).
In the present report, we have investigated some general characteristics of the reaction and have pursued and extended our  (6) that GTP, /3, y-GTP-methylene diphosphonate, and ribosomes each inhibit the ADP ribosylation of transferase II by toxin and NAD.
These inhibitors were first tested because they are known to interact with transferase II. Other compounds (adenine, adenosine, AMP, ADP, ATP, NMN,  nicotinamide, and NADH) were also tested for inhibitory activity.
The inhibitors fall into two groups. Group 1 inhibitors (GMP, GDP, GTP, 0, y-GTP-methylene diphosphonate, ITP, NMN, and nicotinamide) are competitive with respect to both transferase II and NAD.
Group 2 inhibitors (NADH, adenine, adenosine, AMP, ADP, and ATP) are competitive with respect to NAD and noncompetitive with respect to transferase II. On the basis of kinetic, inhibitor, and isotope exchange experiments, as well as our inability to detect a partial reaction, we propose that toxin catalyzes a concerted reaction in which the reactive intermediate is a ternary complex consisting of toxin, NAD, and transferase II. Our results indicate that NAD binds to toxin via the adenine moiety and to transferase II via the nicotinamide-ribose moiety.
Evidence is presented that the ternary intermediate is formed by all possible combinations of binary reactions.
GTP, CDP, CTP, and UTP were products of P-L Biochemicals.
p, y-GTP-Methylene diphosphonate and polyribouridylic acid were supplied by Miles Laboratories, Inc., Elkhart, Indiana. Radioactive Compounds-Adenosine 5'-triphosphate-8-aH tetralithium (7.1 Ci per mmole) was a product of Schwarz BioResearch. 3H-NAD labeled in the adenine moiety was synthesized from NMN and aH-ATP with hog liver NAD-pyrophosphorylase (16) provided by Dr. S. Zimmerman of the National Institutes of Health.
Labeled NAD was purified by DEAE-cellulose chromatography (17) and had a final specific activity of 7 Ci per mmole.
aH-NAD labeled in the nicotinamide moiety was prepared as described by Zimmerman et al. (18) and was a gift of Dr. S. Zimmerman.
The preparation was rechromatographed on DEAE-cellulose (17) before use and 3H-nicotinamide as well as 3H-NAD was collected and saved.
n-PhenylalanylJ4C-tRNA (0.3 /.&i per mg, 350 mCi per mmole of phenylalanine) was purchased from New England Nuclear. It was dissolved in water at 1.7 mg per ml and stored frozen.
Preparations-Partially purified diphtheria toxin was provided by Mr. Leo Levine of the Massachusetts Department of Health.
It was further purified as described by Goor and Pappenheimer (19). The purified toxin contained about 2.6 pg of protein and 60 to 70 minimum lethal dose per flocculation unit. In the presence of 50 mM dithiothreitol, 0.15 pg of this toxin catalyzed 6.3 ppmoles of ADP ribosylation of transferase II in 20 min under the usual assay conditions.
When the toxin was diluted to 30 1.18 per ml or less, a diluent of 0.01 M Tris-HCl buffer containing 100 pg per ml of bovine serum albumin was used. Diphtheria toxoid and horse antitoxin were gifts of the Massachusetts Department of Health.
Highly purified streptococcal NADase (450,000 units per mg) was a gift from Dr. A. Bernheimer of New York University.
Rat liver transferase I and II and purified ribosomes were prepared as previously described (20).
Conditions for Assay of ADP Ribosylation Reaction-The assay of ADP ribosylation of transferase II by toxin and NAD is a modification of the method described by Honjo et al. (15). Unless otherwise specified, each reaction mixture (0.2 ml) contained 1 pmole of Tris-HCl buffer, pH 7.3,0.5 pmole of KCl, 20 mpmoles of dithiothreitol, 0.1 pmole of MgC12, 0.15 pg of diphtheria toxin, 250 pg of transferase II, and 150 to 200 ppmoles of 3H-NAD labeled in the adenine moiety (3.6 x lo5 cpm).
Length and temperature of incubation are indicated in each experiment. The reaction was stopped by adding 1 mg of carrier bovine serum albumin and 2 ml 5% trichloracetic acid. The precipitate was washed twice by centrifugation with 5% trichloracetic acid, collected on Millipore filters, washed again, dried, and counted in a toluene-base solution (Liquifluor, Pilot Chemicals, Inc., Watertown, Massachusetts) in a Nuclear-Chicago Mark I scintillation counter.
In some of the later experiments a method that yielded higher counting efficiency was used. The washed trichloracetic acid precipitates were plated on glass filters (Whatman, GF/C), dried with two aliquots of 5 ml of ether, transferred to vials, and the precipitates dissolved with 0.2 ml of Nuclear-Chicago solubilizer, and counted in 5 ml of Liquifluor; before counting, the scintillation fluid was neutralized with 0.01 ml of 12 N HzS04. This treatment increased the efficiency to 30%.
Amino Acid Incorporation in Cell-free System-Poly rU-directed incorporation of 14C-phenylalanine into trichloracetic acidprecipitable material was carried out in the following system. Each reaction mixture contained 3 pmoles of MgClz, 40 hmoles of mercaptoethanol, 25 pmoles of Tris-HCl buffer, pH 7.3, 0.198 pmole of GTP, 100 pg of poly rU, 100 to 200 pg of ribosomes, 100 I.cg of transferase II, 200 pg of transferase I, and 13.6 pg of 14C-phenylalanyl-tRNA in a final volume of 0.5 ml. Incubation was for 10 min at 37". The reaction was stopped by precipitating with 5 ml of 5% trichloracetic acid. The precipitate was washed and filtered onto Millipore filters and counted in Liquifluor by liquid scintillation counting. Paper Chromatography-Reaction mixtures were either applied directly to paper (Schleicher and Schuell No. 589 Orange ribbon) in several successive aliquots or were first precipitated with 67% ethanol (final concentration), the precipitate removed, and the supernatant evaporated to a volume convenient for spotting onto the paper.
In those experiments, where 3H-NAD labeled in the nicotinamide moiety was used, the paper was developed for 4 hours by descending chromatography with Solvent C of Preiss and Handler (21) modified slightly to contain 7 parts of 95% ethanol to 3 parts of 1 M ammonium acetate buffer, pH 4.8. NAD and nicotinamide standards were cochromatographed with the reaction mixtures.
RF values in this system are NAD, 0.19, and nicotinamide, 0.81. In those experiments where 3H-NAD labeled in the adenine moiety was used, the paper was developed for about 24 hours. NAD, ADP-ribose, and AMP standards were cochromatographed with the reaction mixtures. The appropriate spots were located by quenching of ultraviolet light, cut out, eluted with 1 ml of water, and counted in 10 ml of Bray's (dioxane-base) by liquid scintillation.
Recovery of radioactivity was better than 70%.

Characteristics of ADP Ribosylation Reaction
Honjo et al. (15) reported, and we have confirmed, that toxin is specifically required to catalyze the ADP ribosylation of transferase II. Toxoid, even at a level of 12 1.18, has no activity in the  1. Effect of temperature on the initial rate of ADP ribosylation of transferase II (T2) (A) and on the initial rate of reversal with nicotinamide (B). A, points on the upper curve represent a reaction performed in potassium phosphate buffer, pH 7.6, containing 0.15 rg of toxin and 50 mM dithiothreitol and incubated for 10 min at the specified temperatures.
Points on the lower curve were obtained from an identical reaction performed in 0.01 M Tris-HCl, pH 7.3 (at room temperature), containing 0.15 pg of toxin but no dithiothreitol.
B, in order to measure reversal, 3H-ADP-ribose-transferase II was prepared by incubating transferase II, 3H-NAD labeled in the adenine moiety (0.45 mpmole) and toxin together for 1 hour at 0". The mixture was then treatedwith an excess of streptococcal NADase (4500 units) for 1 hour at 0". Aliquots of this mixture were then treated with 6 pmoles of nicotinamide and incubated for 10 min at the specified temperatures before stopping the reaction with trichloracetic acid. The three curves represent reactions performed under identical conditions, but with the following differences: 0.05 M acetate buffer, pH 5.6 (fop curve); 0.05 M potassium phosphate buffer, pH 7.6 (middle curve) ; 0.01 M Tris-HCl buffer, pH 7.3, at room temperature (bottom curzle). The results are expressed as the tritium (counts per min) released from the starting material after incubation at each temperature. standard assay. Toxin-antitoxin floccules prepared at equivalence and washed several times in phosphate-buffered 0.87% sodium chloride solution were poor catalysts of ADP ribosylation of rat liver transferase II. The floccules were also unable to inhibit amino acid incorporation into trichloracetic acid-precipitable material by the rat liver cell-free system.
The initial rate of ADP ribosylation of transferase II is a linear function of the amount of toxin up to 0.4 pg, of the NAD concentration to 4.8 pM, and of the transferase II concentration to 1.2 mg per ml. We have confirmed the finding of Honjo et al. (15) that, in the presence of excess NAD, the extent of the reaction is a function of the amount of transferase II present.
The effect of temperature on the initial rate of ADP ribosylation of transferase II can be seen from the two curves in Fig. IA. The reaction proceeds at a significant rate at 0" and reaches a maximum at 20". The pH optimum for the forward reaction is 82 The higher Ievel of incorporation observed in the upper curve is attributed to the stimulatory effect of dithiothreitol (22). The initial rate of reversal of the reaction as a function of temperature can be seen from the three curves in Fig. 1B. The pH optimum for reversal is 5.2 As Fig. 1B shows, the removal of the ADP-ribose moiety from transferase II at pH 5.6 (top curve) occurs at all temperatures between 0 and 40"; no maximum was observed.
The rate is considerably slower at all temperatures when the pH is 7.6 (middle curve), as expected.
The two curves have the same slope, indicating that the effect of temperature on the rate of reversal is independent of the pH. The rate of rever-2 Y. Nishizuka, personal communication. Gus ~%'AM p,-r-GTP&ethylene diphosphonate and 0.37 mg per ml of ribosomes.
sal performed in Tris-HCl, pH 7.3 (at room temperature), is too low to detect at or below 20" (bottom curve). Above 20" the increase in reaction rate with increasing temperature is greater than that observed in the other two curves. This is probably caused by the added stimulation of the rate caused by the decrease in the pH of the Tris buffer with increasing temperature.
This lack of reaction in Tris buffer at low temperatures has been used in a later section to study the forward reaction under conditions where reversal does not proceed.
The turnover number of toxin under conditions of the assay (low NAD concentration) is quite low but it can be increased if the NAD concentration is raised. That toxin is catalytic, however, is indicated by the fact that 0.01 p,umole of toxin carried out the incorporation of 10 ppmoles of ADP-ribose into ADPribose-transferase II from NAD. The ADP-ribose-transferase II bond was found to be stable when the trichloracetic acid precipitate was heated for as long as 15 min at 90".
The incorporation of labeled ADP-ribose of a known specific activity into the transferase II fraction allows one to calculate the purity of the transferase II. Assuming that only 1 ADP-ribose residue is incorporated per molecule of transferase II, w.e calculate that the transferase II used in these experiments is only about 1% pure, a degree of purity of the same order as that used by Honjo et al. (15).

Mechanism of Reaction
Inhibitors of ADP Ribosylation of Transjerase II-Inhibitors in the first group are competitive with respect to both transferase II andNAD and include GMP, GDP, GTP, /3, y-GTP-methylene diphosphonate, ITP, NMN, and nicotinamide. GTP and ,B,r-GTP-methylene diphosphonate were examined because they are known to interact with transferase II (6) and to influence its association with ribosomes (7,8). Fig. 2A shows the rate of incorporation of 3H-ADP-ribose into trichloracetic acid-precipitable material in the absence of GTP, in the presence of 7.5 mM GTP, and in the presence of GTP and ribosomes. concentrations in the presence and absence of GTP. Incubation was for 10 min at 37". The initial rate v is expressed as counts per min for the lo-min period. Concentration of transferase II is expressed as milligrams per ml. VT-has been multiplied by 100 before plotting versus (IVAIl)-', and by 1000 before plotting versus (T2)-'.
GTP and the analogue inhibit the ADP ribosylation of transferase II and that ribosomes enhance this inhibition.
GTP and 0, y-GTP-methylene diphosphonate inhibit the initial rate rather than the extent of the ADP ribosylation of transferase II. ITP can replace GTP as an inhibitor and is approximately as active. GMP and GDP inhibit the ADP ribosylation of transferase II by toxin and NAD, but are less effective than GTP on a molar basis. The order of inhibitory activity is p, y-GTP-methylene diphosphonate > GTP = ITP > GDP > GMP.
Inhibition of the initial rate of ADP ribosylation of transferase II by two different concentrations of GTP as a function of either the NAD or the transferase II concentration is shown by the Lineweaver-Burk reciprocal plots in Fig. 3. The inhibition is competitive with respect to both NAD and transferase II. The apparent K, for NAD is 5 pM and the Ki for GTP is 4.9 MM. A value for the K, of transferase II is not reported because of the impurity of the transferase II preparation. GTP (7.5 mM) also inhibits the re-formation of NAD from ADP-ribose-transferase II and nicotinamide by toxin. NMN, like GTP, inhibits ADP ribosylation of transferase II and is competitive with both NAD and transferase II. The Ki for NMN is 3.3 mM. NMN does not inhibit the reverse reaction, the re-formation of NAD from ADP-ribose-transferase II and nicotinamide by toxin.
Nicotinamide also inhibits the ADP ribosylation of transferase II and is competitive with both NAD and transferase II. This inhibition is observed even between 0 and 15", temperatures at which the forward reaction proceeds but at which the reverse reaction in Tris buffer does not occur (see Fig. 1B). Thus, nicotinamide inhibition is not caused by a reversal of the reaction. The kinetic experiments in which nicotinamide inhibition was studied were performed at 15" in order to preclude the possibility of any reversal of the reaction by toxin and nicotinamide.
The Ki for nicotinamide is 0.21 mM. It is thus approximately an order of magnitude more active than NMN.
Using equilibrium dialysis, we have confirmed the finding of Sperti and Montanaro (23) that nicotinamide does not bind to toxin to an appreciable extent.
Inhibitors in the second group, which includes all members of the adenine-ATP (A) series and NADH, are competitive with respect to NAD and noncompetitive with respect to transferase II. Since NAD is known to bind to toxin and AMP constitutes half of the NAD molecule, it seems likely that members of the A series inhibit the ADP ribosylation of transferase II by competing with NAD for its binding site on toxin.
Supporting evidence for such a mechansim was provided by Sperti and Montanaro (23) who showed by equilibrium dialysis and quenching of protein fluorescence that adenine competes with NAD for the same binding site on toxin and that the dissociation constants for NAD and for adenine are of the same order (10u5 M) . Members of the A series inhibit the reaction in the following order of effectiveness: adenine > adenosine > AMP > ADP > ATP. Fig. 4 shows the relative effectiveness of various concentrations of adenine, adenosine, NMN, and GTP in inhibiting ADP ribosylation of transferase II. A comparison of the concentrations needed for 50% inhibition shows that adenine is an order of magnitude more active than adenosine or NMN.
Adenosine and NMN are about 3) times more active on a molar basis than GTP.
Experiments not shown here indicated that adenosine is at least an order of magnitude more active than AMP, ADP, or ATP.
Adenine and adenosine do not cause an apparent inhibition by exchanging with the labeled adenine in the 3H-NAD and thereby diluting the label. This was determined by incubating 3H-NAD, toxin, transferase II, and adenine (or adenosine) and then chromatographing the mixture as described under "Materials and Methods." None of the radioactivity appeared in either the adenine or adenosine spots.
The most active member of the series, adenine, was used for the kinetic studies shown in Fig. 5. Adenine is competitive with respect to NAD and noncompetitive with respect to transferase II. The K; for adenine is 36 pM. Supporting evidence is provided by Sperti and Montanaro (23) who report Ki for adenine as 38.5 PM from equilibrium dialysis experiments performed with toxin, NAD, and adenine in the absence of transferase II. The apparent K, for NAD and the Ki for adenine did not change when the kinetics were performed in the presence of 40 mM dithiothreitol, a concentration which raises the initial rate of the reaction about 5-fold (22). Adenine (1.04 mM) does not inhibit the reverse reaction.
Montanaro and Sperti (24)  Incubation was for 10 min at 30". The initial rate is expressed as counts per min for the IO-min period.
Concentration of transferase II is expressed as milligrams per ml. V-1 has been multiplied by 1000 before plotting versus (NAD)-i or (TZ)-i.
fluorescence that 2 moles of NADH bind per mole of toxin and that the affinity of toxin is higher for NADH than for NAD. Nonetheless, Goor and Pappenheimer (3) demonstrated that NADH is incapable of replacing NAD in the inhibition by toxin of amino acid incorporation in cell-free systems. These observations lead to the prediction that NADH should be a very potent inhibitor of the ADP ribosylation of transferase II by toxin and NAD.
This prediction is confirmed. NADH is competitive with respect to NAD and noncompetitive with respect to transferase II. The Ki for NADH is 0.23 PM; it is approximately two orders of magnitude more active than adenine on a molar basis and is the most active inhibitor tested. NADH (0.55 mu) also inhibits the reverse reaction.
The pyrimidines, cytosine, CMP, CDP, uridine, UMP, and UDP did not inhibit the forward reaction at all at a concentration of 7.5 InM.
CTP and UTP showed slight inhibitory activity at this concentration consistent with their binding MgCla in the reaction mixture.
The concentration of MgCla normally present in the reaction mixture caused a slight stimulation of the initial rate of ADP ribosylation. Table I provides a summary of the studies with inhibitors of the ADP ribosylation of transferase II by toxin and NAD.
If this represents the first step in the reaction, then the succeeding reaction may follow either of two different mechanisms.
In one case the mechansim may be concerted, i.e. the toxin-NAD complex will not react until transferase II is added and only the toxin-NAD-transferase II complex is active in splitting NAD to give the reaction products, nicotinamide and ADP-ribose-transferase II.  Toxin in a large molar excess relative to NAD slowly cleaves the NAD in the absence of transferase II, releasing nicotinamide which can be identified by paper chromatography. Dithiothreitol fails to stimulate the rate of this cleavage even at a concentration that stimulates ADP ribosylation.

Model
Both adenine and nicotinamide inhibit the cleavage. If this cleavage is part of the reaction mechanism, then ADP-ribose-toxin should be formed concomitant with the release of nicotinamide.
No evidence for an ADP-ribose-toxin intermediate could be found. Thus, 3H-NAD labeled in the adenine moiety was incubated with toxin under conditions leading to cleavage and the reaction mixture then passed through a Sephadex G-50 column.
No tritium eluted in the excluded volume with the proteins and all the label was found in the included volume.
These results suggest that the 3H-ADPribose that is split from NAD by toxin is probably released and that the cleavage is probably hydrolytic.
It may be either a side reaction of toxin carried out at a very slow rate or a reaction catalyzed by a contaminant NADase present in the toxin preparation in extremely low concentrations.
Recent results2 support the latter possibility.
As toxin was purified to homogeneity, the hydrolytic cleavage of NAD in the absence of transferase II decreased to zero and the ratio of trichloracetic acid precipitable ADP-ribose-transferase II formed in the presence of transferase II to nicotinamide released approached one. Addition of transferase II to the reaction mixture causes the formation of 3H-ADP-ribose-transferase II which is excluded from Sephadex G-50. Subsequent incubation of the 3H-ADPribose-transferase II with nicotinamide leads to the re-formation of 3H-NAD, identified by paper chromatography.
The above results suggest that the reaction mechanism is concerted and that the reactive intermediate Incubation was for 10 min at 34". The initial rate v is expressed as counts per min for the 10.min period. Concentration of transferase II is expressed as milligrams per ml. V-l has been multiplied by lo4 before plotting versus (iVAD)-1 or (T.Q-l. consisting of toxin, NAD, and transferase II (Model A). Supporting evidence is provided by the results of kinetic studies (25)(26)(27)(28) in which the initial rate was measured at a series of different concentrations of one of the two substrates while the other substrate concentration was held constant (Fig. 6). A ping-pong mechanism (Model B) can be ruled out since the lines are intersecting and not parallel.
Intersection of the lines on the abscissa is consistent with a random rather than an ordered mechansim.
In this case, toxin-NAD and toxin-transferase II complexes are formed rapidly and reversibly and the dissociation constant for either substrate is unaffected by the prior binding of the other to toxin.
Finally, the kinetic analysis predicts that the dissociation constant for each toxin-substrate complex equals the apparent Km for that substrate, as determined by the point at which the lines meet the abscissa. In this system the agreement is quite good; thus, the apparent K, for NAD is 5 pM and the dissociation constant for NAD, independently determined by equilibrium dialysis, is 10 pM (24).
Results of isotope exchange reactions confirm the existence of a ternary intermediate (Model A) and rule out the possibility of a partial reaction involving ADP-ribose-toxin (Model B). If the reaction occurred by a ping-pong mechanism (Model B), isotope exchange would be expected between 3H-nicotinamide and NAD upon incubation with toxin in the absence of transferase II. This was never observed.
On the contrary, evidence was obtained for a reaction mechanism of the Model A type by the following isotope exchange experiment.
3H-ADP-ribose-transferase II labeled in the adenine moiety was incubated with toxin and NAD in the presence and absence of transferase II. According to Model B, exchange of tritium into NAD would occur whether or not transferase II was present.
Model A, however, predicts exchange only in the presence of transferase II. The results in Table II show the exchange of tritium into NAD occurred only when transferase II was added to the incubation mixture.
The low level of radioactive NAD formed in the absence of added transferase II is probably the result of a small

DISCUSSION
Inhibition of the initial rate of ADP ribosylation by GTP and 0, y-GTP-methylene diphosphonate is enhanced by ribosomes (Fig. 2, A and B). Skogerson and Moldave (7,8) showed that transferase II binds to ribosomes only in the presence of GTP and Raeburn et al. (6) found that GTP binds to partially purified transferase II. p, y-GTP-Methylene diphosphonate allows transferase II to bind to ribosomes (8) and it probably can replace GTP in binding to transferase II. As already discussed, the inhibition of ADP ribosylation by GTP and 0, y-GTPmethylene diphosphonate probably results from their binding to transferase II. Since the binding of GTP to transferase II is reversible while the ADP ribosylation is irreversible under the conditions of the experiment, inhibition of ADP ribosylation by GTP (/3,-y-GTP-methylene diphosphonate) is expected to affect only the initial rate and not the extent of the reaction.
The addition of ribosomes to the reaction changes the situation considerably. Goor and Pappenheimer (19) and Gill et al. (14) showed that ribosomes compete with the toxin-NAD complex for transferase II, which is protected from toxic inactivation when bound to ribosomes.
When ribosomes are included in the reaction mixture with GTP, transferase II, toxin, and NAD, the fraction of transferase II that is protected from ADP ribosylation by toxin increases since a certain percentage of the transferase II is now bound to ribosomes where toxin cannot inactivate it. However, transferase II splits GTP in the presence of ribosomes (6,12) and may turn over, i.e. re-enter the pool of free transferase II where it is once more susceptible to toxin.
Thus, even in the presence of ribosomes and GTP, all of the transferase II added to the reaction eventually will be ADP ribosylated.
The GTP analogue 0, y-GTP-methylene diphosphonate, however, is incapable of being split by transferase II on the ribosomes, and dave, using ultracentrifugal techniques, have obtained direct