Isolation and characterization of an adenylyl-protein complex formed during the incubation of membranes from Dictyostelium discoideum with ATP

386

Biochimica et Biophysica Acta, 675 (1981) 386-391

Elsevier/North-HollandBiomedicalPress BBA 29659 ISOLATION AND CHARACTERIZATION OF AN ADENYLYL-PROTEIN COMPLEX FORMED DURING THE INCUBATION OF MEMBRANES FROM D I C T Y O S T E L I U M DISCOIDEUM WITH ATP EDWARD F. ROSSOMANDO,EDMUND V. CREAN and DANIEL P. KESTLER Department of Oral Biology, The University of Connecticu t, School of Dental Medicine, Farmington, CT 06032 (U.S.A.)

(Received January 5th, 1981)

Key words: Adenylyl-protein complex; Adenosine stabilization; A TP; Phosphoamide bond; (D. discoideumJ

When tritiated ATP is incubated with a membrane-enriched fraction prepared from the eukaryotic microorganism Dictyostelium discoideum significant levels of radioactivity can be precipitated with cold, 10% trichloroacetic

acid. Reaction product was formed from ATP and dATP but not from GTP, CTP and UTP. Other studies showed that the maximum amount of the acid-insoluble product was formed about 1 min after the addition of the membranes and that, with further incubation, this reaction product was degraded. The rate of degradation of the reaction product was greatly reduced when the temperature was reduced to 4°C, and when either NaF, Na2SO4 or dithiothreitol was added to the reaction mixture. These additions or conditions had no effect on the productformation reaction. The rate of degradation was also reduced following the addition of adenosine to the reaction and this result did not occur following the addition of ADP, AMP or cyclic AMP. The acid-insoluble reaction product could be solubilized with SDS and analysis by gel-f'fltration chromatography on Sephadex G-75 revealed that the radioactivity was associated with a macromolecule that was not sensitive to RNAase or DNAase but was degraded by pronase. The nucleotide-protein complex was stable at room temperature but radioactivity was released in hot acid, which, after analysis by thin-layer chromatography, was found to co-migrate with authentic AMP, suggesting the formation of an adenylyl-protein complex as the reaction intermediate. The complex bond was stable at neutral and alkaline pH, suggesting a phosphoamide linkage between the protein and the adenylyl moiety.

Introduction

In several previous papers [1-4] we investigated the fate of ATP following its incubation with a membrane-enriched fraction [5] from the eukaryotic microorganism Dictyostelium discoideum. These studies indicated that during the incubation the ATP was degraded to ADP, AMP and adenosine [3,4], a result consistent with other studies, which indicated that a number of hydrolytic activities including an ATP pyrophosphohydrolase [2] and a 5'-nucleotidase [6] were present in these membranes.

During the course of those studies, in which [aH]ATP was routinely used as substrate, we observed that some of the radioactivity could be recovered in an acid-insoluble form, suggesting the association of the radioactive nucleoside with the membrane. Since the formation of such a complex might reflect either the formation of a covalent intermediate and therefore an underlying nucleotidyl transfer reaction, or the entrapment of a non-covalently bound product arising from any one of a number of membrane hydrolytic reactions, we have undertaken a series of studies to characterize this product and the reaction responsible for its formation. The results of these studies are reported in this paper.

Abbreviation: SDS, sodium dodeeyl sulfate. 0 304-4165/81/0000-0000/$02.50 © Elsevier/North-HollandBiomedicalPress

387 Materials and Methods

Materials. [aH]ATP (34.8 Ci/mmol), [3H]GTP (35 Ci/mmol), [aH]CTP (28 Ci/mmol), [3H]UTP (35.5 Ci/mmol) were from New England Nuclear. ATP, dATP, GTP, UTP, cyclic AMP, AMP, ADP, adenosine and phosphodiesterase I (venom phosphodiesterase, type II), were from Sigma Chemical Co. RNAase and DNAase (both from bovine pancreas) were purchased from Worthington Biochemical Co. Pronase was from Calbiochem-Behring Co. PEI-cellulose sheets (F5504) and cellulose sheets (No. 13254)with fluorescent indicator, were obtained from Merck and Eastman, respectively. Sephadex G-75 and G-15 were obtained from Pharmacia. Organisms. The axenic derivative (Ax-3) of D. discoideum was grown to the stationary phase of growth ((1-2)" 107 cells/ml) and harvested using methods previously described [5]. Preparation of plasma membrane-enrichedfraction. Membrane-enriched fractions, prepared as described [5] from 1 • 101° cells, were stored as a pellet at -20°C. For digestion with RNAase, one frozen membrane pellet was resuspended in 2 ml of 10 mM MgC12/10 mM Tris-HCl (pH 7.5) and 200 p.l of an RNAase solution (3 mg/ml) added. After incubation at 30°C for 30 min, the reaction was diluted with 15 ml of 0.01 M Tris.HC1 (pH 7.5), centrifuged at 30000Xg for 30 min in the cold, and the pellet washed twice with the same buffer. Reaction conditions. A reaction mixture usually contained in a final volume of 50 /.d: 5 panel Tris. HC1, pH 7.5, 0.18 nmol of [3H]ATP (1.6 • 104 cpm/ pmol) and up to 400 ktg membrane protein. Reactions were usually incubated at 220C and terminated by the addition of about 2 ml ice-cold 10% trichloroacetic acid. After a 30 min incubation in an ice-water bath, precipitates were collected on.glass-fiber ftlters (Whatman GF/C) using vacuum t'titration. Filters were dried and radioactivity determined. Isolation of nucleotide-protein complex. The all-labeled membrane pellet was solubilized com. pletely following an incubation for 16 h at 37"C in 2 ml of 8.5% SDS/5% 2-mercaptoethanol/12.5 mM EDTA/5 mM Tris-HC1 (pH 7.5). The dissolved sample was applied to a Sephadex G-75 (100X 1.5 em) column and the column eluted with a buffer containing 10 mM Tris-HC1 (pH 7.5)/0.1% SDS/2.5 mM

EDTA. Fractions, from the excluded region of the column, were pooled and dialyzed for 5 days against repeated changes of distilled water containing 0.1 M NaC1 to reduce the SDS concentration, and lyophi. lized. Enzyme digestions. For treatment with pronase, the lyophilized sample was resuspended in 1 ml buffer containing 0.01 M Tris-HC1, pH 7.5/10 mM CaC12 and I mg pronase was added and the solution covered with 100 pl toluene. The solution was incubated at 37°C for 30 h at which time an additional 1 mg of pronase was added and the incubation continued for 48 h. The reaction mixture was clarified by centrifugation, and the supernatant was lyophilized, resuspended in 2ml 0.1 M pyridine/acetate buffer, pH 5.6, and loaded onto a Sephadex G-15 column and eluted with the same buffer. Snakevenom phosphodiesterase reactions were in 100 mM Tris-HC1, pH 8.5, and 30 mM magnesium acetate for 1 h at 37°C and were terminated with trichloroacetic acid. Results

Characterization of overall reaction. In a previous study we found that the incubation of ATP with a membrane-enriched fraction, prepared from D. discoideum, resulted in the formation of ADP, AMP and, given sufficient incubation time, its complete conversion to adenosine [3,4]. These incubations, carried out with [aH]ATP, were terminated by the addition of acid and the acid-soluble mixture analyzed by thin-layer or high-pressure liquid chromatography [4]. In the course of our studies we found that a significant amount of the radioactivity was preeipitable by acid and could not be solubilized by extensive washing or by extraction with chloroform/ methanol. The precipitate could be collected by filtration and the amount of radioactivity present determined by liquid scintillation counting. A typical time course for the formation of this acid-precipitable product at 22"C with [3H]ATP as substrate is illustrated in Fig. 1. The values for zerotime controls were obtained by adding [3H]ATP to a complete reaction mixture after the addition of the trichloroaeetic acid. This type of control was used to detect any 3H-labeled impurities which are acidprecipitable or which bind to denatured membranes.

388 I

0,50-

]

TABLE I

I

EFFECT OF ADENINE NUCLEOSIDES ON RATE OF PRODUCT DEGRADATION

~ 0.50

o

0.1 > 0.25

0.4

Membranes

O

o

;

Each compound was added at concentration of 5 mM 30 s after reaction was started by adding membranes. Values represent the amount of acid-insoluble radioactivity recovered 10 min after addition of compound.

/~

~:,, 0.25

,b

,;

(rag)

20

Compound

Product recovered (pmol)

None ADP AMP Cyclic AMP Adenosine

0.062 0.102 0.057 0.056 0.34

TIME (MINUTES)

Fig. 1. Time course of formation of 3H-labeled product. Reaction mixture containing 200 ~g of membrane protein, 5 ,umol of Tris-HC1 (pH 7.5) and 0.18 nmol [3H]ATP (5 t.,Ci) in a total volume of 50 /A were included for the indicated times at 22°C. Reactions were terminated by the addition of trichloroacetie acid (10%) and the precipitated collected by filtration. The amount of radioactivity was measured by liquid scintillation counting. (inset) Effects of membrane concentration on the rate of product formation. Reaction mixtures containing various concentrations of membranes were incubated for 30 s at 30°C under the conditions described.

The amount of recoverable product varied considerably over the 20-min incubation period, reaching a peak in less than 1 min. As shown in the inset of Fig. 1, the amount of product formed in 30 s was proportional to the concentration of membrane protein. Within 1 0 - 2 0 min after the start of the incubation, the amount of reaction product declined to a value less than 5% of the peak value (Fig. 1). This decline suggested not only a cessation of product synthesis but a process of product degradation as well. When the incubations were carried out at 4°C the product-degradation reaction was found to be reduced about 75% while the amount of product formed after the first minute remained unchanged. Also, the addition of 0.1 M NaF reduced the rate of degradation over 50%, while both.Na2SO4 (10 mM) and dithiothreitol (0.3%) reduced it about 20 and 30%, respectively. None of the three had any effect on the amount of product formed after the first minute. The addition of 5 mM of either ADP, or AMP or cyclic AMP to the reaction mixture 30 s after the

start of the reaction did not reduce the rate of product degradation (Table I). In contrast, if 5 mM adenosine were added at this time, a significant reduction in the rate of product degradation was observed such that about 70% of the product could still be recovered after 10 min (Table I). While no significant amount of product formation occurred following incubation of the membranes with GTP, CTP or UTP, the amount of product formed, was reduced 80% in the presence of a 50-fold excess

I

0.50 "d

)-D--D

I

I

D

[]

[

0 0 t~

o

E .9 0.25

0 ~ O

>I-.-

_>

"-O

I-

o

0

0

lb

2b

sb

40

P R E I N C U B A T I O N TIME (MINUTES)

Fig. 2. Effects of preincubation temperature of membranes on 3H-labeled product formation. Samples containing approximately 200 Izg membrane protein were preincubated at temperatures indicated in 50 mM Tris-HC1 buffer (pH 7.5). At the indicated time [3H]ATP was added and the amount of radioactivity incorporated after 30 s determined as in Fig. I legend. D, 4°(2; o, 22 and 30°C.

389

of either dATP, 60% with GTP and 50% with UTP (Table II). If excess unlabeled ATP was added to the reaction mixture, following the formation of product, no reduction in the amount of radioactivity associated with the product was observed, suggesting that the membrane-associated label was not easily displaced. Preincubation of the membranes in an icewater bath (4°C) for up to 40 rain had no effect on the amount of product formed after 30 s of incubation (Fig. 2). In contrast, preincubation of the membranes at either 22 or 30°C resulted in a decrease in the amount of product formed after a 30 s reaction, with the magnitude of the decrease proportional to the time of pre-incubation. Thus, as shown in Fig. 2, after 20 min at either 22 or 30°C, only about 40% of the product-forming activity remained.

Characterization o/acid-insoluble reaction product. Treatment of the membranes with SDS resulted in their complete solubilization. Analysis of the SDSsoluble reaction product by gel-f'fltration chromatography revealed that the radioactivity remained associated with a molecule with a molecular weight in excess of 60 000 as estimated by its elution volume of Sephadex G-75 (Fig. 3A). The 3H-labeled complex was also subjected to digestion by a number of enzymes and the products rechromatographed on Sephadex G-75. No change in elution position of the radioactivity was noted following incubation with either RNAase or DNAase. Digestion with pronase, however, resulted in an increase in elution volume of

TABLE II EFFECT OF NUCLEOTIDE TRIPHOSPHATES ON PRODUCT FORMATION Values for substrate represent product formed/30 s using [3H]labeled compound as substrate using reaction condition described in Materials and Methods. Values for inhibitor are inhibition produced when 0.1 mM of each compound added to reaction containing [3 H] ATP as substrate, n.d., not determined. Compound

As substrate pmol

As inhibitor(%)

ATP dATP GTP CTP UTP

0.65 n.d. 0.035 0.010 0.054

0 80 60 n.d. 50

i

3000.

i

&

i

(n)

(A)

-800 -600

iooo-

"400

500-

"2O0

~

6b

80

ib

2b

3'0

,~0

4o

FRACTION

Fig. 3. A, Gel chromatography of 3H-labeled products. A 3H-labeled membrane pellet was incubated for 16 h at 37°(2 in 2 ml of 8.5% SDS/5% 2-mercaptoethanol/12.5 ram EDTA/ 5 mM Tris-HCl (pH 7.5). The dissolved sample was then applied to a 100× 1.5 cm column of Sephadex G-75 and eluted with 2.5 mM EDTA/0.1% SDS/10 mM Tris-HC1buffer (pH 7.5). Fractions of 2.8 ml were collected and 50-~1 aliquots were counted in Aquasol. Vo and Vi indicate excluded and included volume, respectively. B, Gel chromatography of pronase digestion products. The soluble products of Pronase digestion were dissolved in 2 ml of 0.1,M pyridine acetate buffer (pH 5.6) and applied to a 1.5 × 42 cm column of Sephadex G-15 and eluted with the same buffer. Fractions of 2 ml were collected and 0.1 ml of each was counted in Aquasol.

the radioactivity (Fig. 3B), consistent with the conclusion that the complex containing the radioactive nucleoside was protein.

Properties of nucleotide-protein bond. While the 3H-labeled complex was stable in 10% trichloroacetic acid at both 0°C and room temperature, radioactivity was released if the acid solution was warmed to 90°C. These results suggested that the nucleoside was not attached to the protein through an O-alkyl phosphoester, since bonds of this type are usually stable in hot acid. In contrast, the bond remained intact at neutral and alkaline pH (0.5 M KOH) even after 1 h at 50°C. This alkaline stability (together with the acid lability mentioned above) suggested that a phosphoamide bond was involved in the complex. When samsamples of the radioactivity released by treatment with acid were anlyzed by thin-layer chromatography as previously described [3], over 95% of the radioactivity co-eluted with authentic AMP. Discussion

In this paper we show that during the incubation of a membrane-enriched fraction with ATP a reaction

390 product is formed that is precipitable by cold acid. A number of other observations suggest that the formation of the product is a result of the formation of a strong, probably covalent, bond. For example, the label could not be displaced from the membrane by excess unlabeled ATP, a 'cold chase', suggesting that the radioactive ATP was not merely bound or trapped in a vesicle. Second, extraction of the acid-precipitable product with organic solvents and detergents, which solubilized the membrane, did not result in release of the radioactivity from the membrane protein. And third, when gel-filtration experiments were carried out on the complex, after solubilization in strongly denaturing conditions, the radioactivity co-eluted with a protease-sensitive macromolecule, suggesting the presence of a bond between the nucleoside and a membrane protein. The time course of interactions observed suggest that the formation of the 3H-labeled protein bond is followed by a process in which the bond is cleaved. The time-dependent net loss of reaction product also suggests further that the process underlying the formation of the reaction product is of limited duration. The cessation of a product formation, however, does not appear to be the result of a reduction in the level of substrate, since addition of more ATP at the onset of the degradation process does not result in the production of additional reaction product whereas its addition after 30 min does. However, the amount of product recovered after the first minute of reaction is affected by both preincubation time and temperature, suggesting that the process underlying the formation of the product is unstable or is inactivated. The reaction does not occur with a number of other nucleotides such as GTP, indicating that catalysis requires the chemical configuration present in adenosine. Other nucleotides, however, can compete with ATP, suggesting that the requirements for substrate binding are not as specific as for catalysis. While these data do not allow for an unequivocal statement concerning the nature of the bond joining the phosphate group of the adenylyl intermediate to the protein, the difference in properties of the three types of previously demonstrated bonds do allow some comment. For example, the complex studies by Bader et al. [7] was stable at pH 1 - 2 but was rapidly hydrolyzed in strong acid (performic acid) and at neutral pH and above. Based on the similarity

between the properties of their complex and that of acetyl phosphate [8], Bader et al. concluded that the bond was most probably a mixed anhydride and most likely an acyl phosphate. In another study the bond linking an adenylyl group to an enzyme was characterized [9]. This bond, while able to withstand boiling at neutral and alkaline pH and treatment with acid at 0°C, was cleaved by exposure to dilute acid at room temperature and above [9]. This property of acid lability and alkaline stability was consistent with the conclusion that the nucleotide was bound to the enzyme by a phosphoamide bond [9]. In contrast, the O-alkyl phosphoester bond, such as phosphoserine, is stable in acid at temperatures of 25°C or above [10], and is cleaved by bovine intestine phosphodiesterase [11]. Since, in the present study, the nucleotide-protein bind is not stable in acid at room temperature or at 50°C, we conclude that an O-alkyl phosphodiester, such as adenylyl-O-serine, may not be involved. The bond, however, is stable at neutral and basic pH, suggesting a bond of the phosphoamide type [12]. It is interesting to note that the properties of the bond studied in this case differ from those of adenylyl-Otyrosine which is bout in Escherichia coli glutamine synthetase [13]. As discussed by Stadtman [14], in the cases where adenylyl groups have been covalently bonded to proteins, the formation of such a covalent intermediate has a regulatory function. Whether or not the formation of the adenylyl-protein complex that we observed has such a role and what reaction is being regulated remains to be established. With this comment in mind the results obtained with adenosine should be discussed. The role of adenosine in this reaction sequence appears to be limited to an effect on the degradation process, since the addition of adenosine to the incubation mixture at the start of the experiments had no effect on the formation of the adenylyl-protein complex but reduced, significantly, the rate of its degradation. This result was not obtained with ADP, AMP or cyclic AMP. This striking effect of adenosine could be accounted for by its acting as an uncompetitive inhibitor and stabilizing the intermediate. Based on this effect of adenosine in prolonging the life of the adenylylated complex, and on previous observations showing an effect of adenosine on adenylate cyclase catalytic activity [15], it is tempt-

391 ing to speculate that the formation of this intermedi. ate has a role in the regulation of the activity of this enzyme. Also it should be noted that since a hydrolytic reaction catalyzed by ATP pyrophosphohy. drolase has been shown to be present in this membranes [2], the adenylylated protein isolated here may be the covalent intermediate in this reaction. If this is the case, the intermediate is quite similar to the one demonstrated in the analogous enzyme from bovine intestine [11 ].

Acknowledgements This research was supported in part by grants from the National Institutes of Health (DE-03715) and the National Science Foundation (20615). We thank Mary Ann Hesla and Barbara Maldonado Holbink for technical assistance during the early stages o f this work.

References 1 Rossomando, E.F. and Sussman, M. (1972) Bioehem. Biophys. Res. Comm. 47,604-610 2 Rossomando, E.F. and Sussman, M. (1973) Proe. Natl. Acad. Sei. U.S.A. 70, 1255-1257

3 Rossomando, E.F. and Hesla, M.A. (1976) J. Biol. Chem. 251,6568-6573 4 Hodge, J. and Rossomando, E.F. (1980) Anal Bioehem. 102, 59-62 5 Rossomando, E.F. and Cutler, L.S. (1975) Exp. Cell Res. 95,67-78 6 Rossomando, E.F. and Maldonado, B. (1976) Exp. Cell Res. 100, 383-388 7 Bader, H., Sen, A.K. and Post, R.L. (1966) Biochim. Binphys. Acta 118,106-115 8 Lipmann, F. and Turtle, L.C. (1944) J. Biol. Chem. 153, 571-582 9 Gumport, R.I. and Lehman, E.F. (1971) Proc. Natl. Acad. Sci. U.S.A. 68, 2559-2563 10 Wong, L.-J. and Rose, I.A. (1976) J. Biol. Chem. 251, 5431-5439 11 Landt, M. and Butler, L.G. (1978) Biochemistry 17, 4130-4135 12 Shabarova, Z.A. (1970) in Progress in Nucleic Acid Research and Molecular Biology (Davidson, N.J. and Cohn, W.E., eds.) Vol. X, pp. 145-182, Academic Press, New York 13 Shapiro, B.M., Kingdon, H.S; and Stadtman, E.F. (1967) Proe. Natl. Acad. Sci. U.S.A. 58,642-649 14 Stadtman, E.F. 0973) in The Enzymes (Boyer P.D., ed.), Vol. VIII, part A, pp. 1-49, Academic Press, New York 15 Baer, H.P. and Drummond, G.I. (eds.) (1979) Physiological and Regulatory Functions of Adenosine and Adenine Nucleotides, Raven Press, New York