sign in sign up
<<
Home , Alanine, Alanine dehydrogenase

Proc. Nat. Acad. Sci. USA Vol. 68, No. 5, pp. 916-919, May 1971

Covalent Attachment of Diethylstilbestrol to : Implications for

JOHN KALLOS* AND K. P. SHAW The Research Laboratory, Department of Pathology, St. Mary's Hospital, Montreal, Canada Communicated by C. B. Anfinsen, December 17, 1970

ABSTRACT An affinity labeling reagent for the estro- glutamate dehydrogenase by means of site-specific reagents genic- of bovine liver L-glutamate dehydro- and to be published). The present paper deals genase (EC 1.4.1.3) was prepared by conversion of diethyl- (9, manuscript stilbestrol to its alkylating analogue, bromoacetyldiethyl- with the covalent attachment of diethylstilbestrol to GDH stilbestrol. Under standard assay conditions, the analogue and its implication in allosteric transition. With regard to the acted as a reversible allosteric with regulatory problem of allosteric transition, the present paper describes activity much like that of diethylstilbestrol. However, experiments pertinent to the following question: (a) Since incubation of the with the alkylating agent in the presence of DPNH resulted in a permanent decrease in allosteric ligands exert a controlling function, without them- glutamate (X form) and an increase in (Y form) selves participating in the reaction, will covalent attachment activities, and in covalent attachment of diethylstilbestrol of ligand to the produce the same, similar, or different in the ratio of 1 mol per subunit (of particle weight effects as the reversible ligand-protein interaction? (b) What is 52,000). The brominated analogue behaved as an affinity the degree of conformational stabilization resulting from label that mimicked the allosteric effects of diethyl- stilbestrol. Diethylstilbestrol protection of the enzyme permanently freezing the protein in one form as a result of the against alkylation by bromoacetylated sterol suggested chemical attachment of the ligand to the protein? competition for the same binding site, while ADP protec- Preliminary studies with various DES derivatives indicated tion indicated a shift of protein equilibrium into the X that one of the two hydroxyl groups can be modified by acet- form. The diethylstilbestrol-enzyme compound was de- sensitized (relative to the native enzyme) to allosteric ylation without loss of the capacity to bind to GDH. This ob- reagents such as ADP and GTP. The results were con- servation prompted us to attach a bromoacetyl group to one of sistent with conformational freezing of the modified the hydroxyl groups of the DES molecule. The close structural protein molecule into the Y form. analogy between bromoacetyl and acetyl DES led us to the expectation that a reversible ligand-protein interaction should the structure the Estrogenic hormones alter and catalytic produce a reversible allosteric effect, and that such a complex activity of crystalline glutamate dehydrogenase (EC 1.4.1.3) could be induced to form a covalent bond (by elimination of (GDH); they simultaneously inhibit glutamate and stimulate bromoacetic ) and attach the DES molecule to the protein. activity of the enzyme (1-3). ADP can reverse the effect of these (4). The action of regulatory molecules, such as diethylstilbestrol (DES), GTP, and ADP, is MATERIALS AND METHODS explained by a shift of the equilibrium between two different forms (X=Y) of the enzyme. The X form possesses high Bovine liver -glutamate dehydrogenase (type 1, glutamate activity; the Y form possesses alanine dehydro- sulfate suspension), DPNH, ADP, and GTP were obtained genase activity. Estrogen and GTP favor the Y form, while from Sigma Chemical Co. The enzyme suspension was usually ADP shifts the equilibrium to the X form. Such behavior has centrifuged before use, and the crystals were redissolved in an been rationalized in terms of allosteric interaction by Monod appropriate buffer. The enzyme solution was equilibrated with et al. (5) and Koshland (6); the term allosteric is now generally buffer either by gel filtration through Sephadex G-25 or by employed in connection with regulatory effects that are due to dialysis. Protein concentration was measured at 278 nm on the conformational changes in the protein molecule induced by a assumption that 1 mg/ml has an absorbance of 1 in a 1.0-cm ligand (7, 8). The precise molecular mechanism by which the cuvette. DES was obtained from Mann Research Laboratories allosteric ligand acts remains obscure. Clearly, more specific (New York, N.Y.). Bromoacetyldiethylstilbestrol (BADES) information is needed about the chemistry of the regulatory was prepared in our laboratory and dissolved in acetonitrile- and active sites before allosteric regulation is understood. -ethylene glycol 2: 1: 1 before use. During the past several years we have been engaged in a Tritium-labeled BADES was prepared in our laboratory systematic chemical study of the active and regulatory sites of from ['HIDES (400 Ci/mol, Amersham/Searle Corp.). All radioactive materials were mixed with Bray's solution or -2,5-diphenyloxazole solution and counted in a Nu- Abbreviations: DES, diethylstilbestrol; BADES, bromoacetyl- clear-Chicago liquid scintillation counter. All the buffer salts diethylstilbestrol; GDH, glutamate dehydrogenase. used were of analytical grade. in a * Present address: Department of Cell Biology, Faculty of Medi- Enzyme assays were performed at room temperature cine, Centre Hospitalier Universitaire, Sherbrooke, P.Q., Canada. Bausch and Lomb 505 spectrophotometer; the disappearance To whom request for reprints should be addressed. of DPNH at 340 nm was measured in 1-cm quartz cuvettes. 916 Downloaded by guest on September 25, 2021 Proc. Nat. Acad. Sci. USA 68 (1971) Modification of Glutamate Dehydrogenase 917

c 0 0.3 c f._ 0 a c a E 0.5 Alanine Reaction E 4 an < E 0 E 04 0.2 >,

t. a.o 0-E -U 0 -U 0.3

0 0 -a

Concn of FIG. 1. Effects of concentration of DES and BADES on GDH activity. GDH activity was measured in the assay mixture described in Methods; the GDH concentration was 0.2 jug/ml. The mixture for determining alanine dehydrogenase activity contained 0.1 M Tris (pH 8.6) and 5 X 10-2 M pyruvate in lieu of the a-ketoglutarate; the enzyme concentration was 5.15 pg/ml. Activity is expressed as A340 units per cuvette per 2 min.

The reactions were started by the addition of the enzyme; Finally, it was dissolved in 0.25 ml of Nuclear-Chicago 2-min incubations were used for calculations of enzymatic solubilizer and counted in a liquid scintillation spectrometer. activity. GDH activity was measured, in the direction of a-keto- Hydroxylamine treatment of ['HJDES-GDH glutarate , with a solution containing 1 ml Three samples, 1 ml each, were incubated at 220C; they con- of 0.05 M phosphate (pH 7.8); 1 X 10-4 M EDTA; 2 X tained [8H]DES-GDH (0.38 mg/ml), 0.15 M hydroxyl- 10-4 M DPNH; 0.2 M NH4Cl; and 2.5 X 10-2 M a-ketc- -HCl, 0.025 M Tris, and 0.1 M KCl (pH 7.8). The reac- glutarate. Alanine dehydrogenase activity was measured, tion was stopped by the addition of ice-cold 2 M HCl to pH 3 in the direction of pyruvate reductive amination, with a and the free [3HIDES was extracted with ether. The ether solution containing 1 ml of 0.1 M Tris * HCl (pH 8.6); 1 X 10-4 layer was washed with diluted NaHCO3, dried over MgSO4, M EDTA; 2 X 10-4 M DPNH; 0.2 M NH3Cl; and 5 X 10-2 M pyruvate. I - 100 100 > 0-e 0 of GDH with BADES 5' 0 GDH (0.3 mg) was incubated in 1 ml of a reaction mixture _ 80 80 c containing 0.05 M phosphate buffer (pH 7.6), 5 X 10-5 M BADES, and 1 X 10-3 M DPNH at room temperature. The iO rate of irreversible inhibition was determined by assay of '04 60 60 o 0~ 0 aliquots. The reaction was terminated by gelfiltration through to )) a G-25 column X 40 the was 0 Sephadex (1.5 cm); 0' C eluted with 0.05 M phosphate buffer (pH 7.6). 0C Preparation of [SH]DES-GDH .Ad 20 0 20 GDH (6 mg) was incubated in a 10-ml reaction mixture con- E taining 0.05 M phosphate (pH 7.6), 8 X 10-5M bromoacetyl- 0 O-I ['H]DES (536,000 cpm/mg), and 1 X 10-3 M DPNH: after url 30 60 90120-c 2 hr of incubation, an aliquot (5 ml) of the reaction mixture Time in Minutes was passed through a Sephadex G-25 (fine) column (2.5 X 40 to remove the excess For an FrG. 2. Progress of the chemical reaction of GDH with cm) reagent. counting, aliquot (1 BADES. GDH (0.3 mg/ml) was incubated in phosphate buffer ml) of modified protein was mixed with 10 ml of Bray's solution (0.05 M, pH 7.6) with 5 X 10-4 M BADES in the presence of 1 X and counted. Another aliquot (2 ml) of modified protein was 10-3 M DPNH. Aliquots were taken at various time intervals precipitated with 4 ml of trichloroacetic acid (30%), then and passed through a column of Sephadex G-25, and the alanine centrifuged at 5000 rpm for 15 min. The precipitate was and glutamate activities were determined. 0, Alanine; A, gluta- washed four times with absolute and once with ether. mate. Downloaded by guest on September 25, 2021 918 : Kallos and Shaw Proc. Nat. Acad. Sci. USA 68 (1971) TABLE 1. Labeling of GDH with bromoacetyl-['H IDES TABLE 3. Hydroxylamine treatment of ['H ]DES-GDH

Incubation Incorporation Incubation time Radioactivity (cpm) time Activity (% of initial) of ['H]DES (min) Enzyme-bound In ether extract (min) Glutamate Alanine (mol/subunit) 0 1620 210 15 40 100 0.69 20 460 1250 60 31 108 0.85 60 320 1490 120 18 110 1.08 Free [3H]DES was extracted into ether and the protein-bound GDH (0.3 mg/ml) was incubated in phosphate buffer (0.05 M, ['HIDES was precipitated with trichloroacetic acid; the radio- pH 7.6) with 5 X 10-5 M [3H]BADES in the presence of 1 X activity of both fractions was measured. 10-3 M DPNH. Aliquots were taken at various time intervals and passed through a Sephadex G-25 column, and the glutamate removing aliquots, passing them through a Sephadex G-25 and alanine activities and the radioactivity were determined. column, and measuring the analine and glutamate activities of the modified protein (Fig. 2). There was a progressive loss of filtered, evaporated to dryness, and dissolved in 1 ml of glutamate activity to about 80% inhibition in the first 1.5 hr, toluene, and the radioactivity was determined. and a slight activation of the alanine activity. This modifica- tion appears to be an irreversible one, since attempts to re- RESULTS AND DISCUSSION store native activity by extensive dialysis against 0.05 M The reversible BADES-enzyme interaction phosphate buffer or gel-filtration through Sephadex G-25 were The reversible allosteric effect of BADES (added to the en- not successful. These results indicate that a covalent bond zyme without incubation with DPNH) was examined by with DES has been formed at the estrogen-binding site of determining the glutamate and alanine dehydrogenase the enzyme. To substantiate this conclusion, and to quanti- activities of the enzyme (Fig. 1). Clearly, BADES, like DES tate the incorporation into protein, we performed the reaction itself, stimulates alanine and inhibits glutamate activity and with bromoacetyl-['HIDES (Table 1). About one DES behaves as a reversible allosteric ligand. Note that the stimu- molecule was bound per subunit of 52,000 daltons, in good latory and the inhibitory effects were reciprocal, as has been agreement with the kinetic data showing that this binding previously observed for other GDH effectors such as ADP and was irreversible. The covalent binding is further demonstrated GTP. The interaction between BADES and GDH, under by the experiments summarized in Table 2. standard assay conditions, appears to be a reversible one. This DES-GDH was examined by acrylamide gel electrophoresis conclusion was based on the following observations: (a) Dilu- and was found to be a single component. The DES-GDH tion of the assay mixture with bovine serum albumin solution derivative was purified by affinity chromatography (manu- resulted in complete recovery of glutamate activity (9). (b) script in preparation) on DES-Sepharose (a column of Sepha- When the assay was performed with [3H]BADES, there was rose conjugated with DES). By this technique, it was possible no significant radioactivity incorporated into the protein to separate an unretarded peak (88%) containing the modified after Sephadex filtration or trichloroacetic acid precipitation protein (and all the radioactivity) from a second component, (to be published). These results indicate that the enzyme which showed much stronger UV absorption than the DES- responds identically to BADES and DES; furthermore, Sepharose that contained native enzyme. These techniques the interaction is reversible. The most reasonable explanation demonstrated that the reaction of BADES with GDH was is that BADES and DES exert their effect by shifting the specific and stoichiometric. equilibrium from the X (glutamate) to the Y (alanine) form. Since DES is probably attached to the protein by a car- The structure and the catalytic activity of the enzyme are boxylic ester linkage, we expected that such a bond would be rapidly altered by its interaction with the ligand without any stable at neutral pH, but easily cleaved by mild alkaline treat- apparent chemical alteration in the protein. ment or by exposure to nucleophilic reagents such as hydroxyl- amine. This was found to be true (Table 3). However, it was Chemical reaction of GDH with BADES impossible to regenerate enzyme activity (glutamate assay) We turn our attention to the next question, the irreversible after the displacement of DES by hydroxylamine. effect of BADES on GDH. The enzyme was incubated with The results of the affinity-labeling experiments are as fol- BADES, and the progress of the reaction was followed by lows. (a) We have definite evidence for the covalent attach- ment of DES to the protein; the stoichiometric ratio was ap- proximately 1 mol of DES per subunit. (b) The inactivation of TABLE 2. Stability of the [3H]DES-GDH compound glutamate, but the retention of analine, activity of the enzyme 3H per mg of by BADES is compatible with the conclusion that this com- Expt. Treatment of sample enzyme (cpm) pound acts as an affinity label that mimics DES. (c) A reason- A 3H-labeled enzyme obtained by gel filtra- able conclusion is that the covalently bound DES acts simi- tion (from experiment in Table 1) 1600 larly to the noncovalently bound DES. B Aliquot of A dialyzed for 48 hr against 7 M Direct and indirect protection against the BADES reaction 1460 C Aliquot of A precipitated with 2 vols of To explore the mechanism of BADES reaction, we examined 20% trichloroacetic acid, centrifuged, two methods of protecting the enzyme against modification by and washed with alcohol and ether 1740 BADES. analogues are known to protect the of an enzyme against chemical modification. Therefore, Downloaded by guest on September 25, 2021 Proc. Nat. Acad. Sci. USA 68 (1971) Modification of Glutamate Dehydrogenase 919

TABLE 4. Effect of DES and ADP on the rate of reaction of TABLE 5. Effects of alosteric modifiers (ADP and GTP) on GDH with BADES the activities of native GDH and DES-GDH Treatment Glutamate activity (%) Alanine activity Enzyme of GDH 100 Ligand (M) (% control) GDH + BAIJES 33 GDH ... 100 GDH + DES + BADES 65 GDH 5 X 10-6 GTP 269 GDH + ADP + BADES 78 DES-GDH 5 X 10-5 GTP 127 GDH I X 10-4ADP 35 GDH (0.3 mg/ml) was incubated in phosphate buffer (0.05 M, DES-GDH 1 X 10-4ADP 94 pH 7.6) with 5 X 10-6 M BADES and 1 X 10-8 M DPNH, in the presence or absence of 5 X 10-4 M DES or 6 X 10-4 M ADP. Aliquots were taken after a period of 30 min. passed through Sephadex G-25, and assayed for glutamate activity. Our results are consistent with the assertion that the cova- lent attachment of DES to GDH produces a permanent shift DES should protect the enzyme against the BADES reaction. of the equilibrium from the X to the Y state, i.e., conforma- Also, in the case of an allosteric enzyme, there is an alternative tional freezing of the protein in the Y form. This covalent possibility to control chemical modification at the estrogen- attachment appears to freeze the enzyme in a conformation binding site, based on the following assumption: the Y form of that mimics the noncovalent DES -protein state. There are the enzyme has an affinity for the estrogen molecule while the several examples of enzyme-catalyzed modifications, by X state does not. This line of reasoning leads to the idea that covalent mechanisms, that regulate enzyme activity. The an effector that will shift the equilibrium into the X form will classical ones are the regulation of glycogen phosphorylase prevent chemical modification at the estrogen-binding site. (12) in mammalian tissues and the regulation of ADP is known to put GDH into the X form and, thus, reverse synthetase of E. coli by adenylation (13). Although the the action of the steroid (4). Thus, we would expect that both covalent attachment of ligands such as DES to a protein may DES and ADP should protect the enzyme against reaction not have direct physiological significance, chemical manipula- with BADES, DES directly by competition for the same site tions involving conformational freezing, subunit exchange of and ADP indirectly by a conformational shift in the protein. modified native protein, and synthetic introductions of func- The results (Table 4) tend to confirm our predictions; both tional groups at the active sites may yield insight into the de- ADP and DES protect against the BADES reaction, as sign of the protein molecule and subunit interactions in allo- shown by glutamate activity and the incorporation of radio- steric control. activity into the protein. This investigation was supported by grants from the Medical Conformational freezing Research Council of Canada, National Cancer Institute of Can- Can the enzyme be frozen into a given conformation as a ada, and NATO. result of the covalent attachment of DES? The response of the 1. Tomkins, G. M., K. L. Yielding, J. F. Curran, M. R. Sum- modified enzyme to effectors such as ADP and GTP is shown mers, and M. W. Bitensky, J. Biol. Chen., 240, 3739 (1965). in Table 5. The DES-GDH did not respond to allosteric ef- 2. Tomkins, G. M., K. L. Yielding, N. Talal, and J. F. Curran, fectors such as ADP and GTP. Cold Spring Harbor Symp. Quant. Biol., 28, 461 (1963). These 3. Frieden, C. J., J. Biol. Chemn., 238, 3286 (1965). results indicate that modification with BADES 4. Yielding, K. L., and G. M. Tomkins, Proc. Nat. Acad. Sci. desensitizes the enzyme to effectors such as ADP and GTP. USA, 46, 1483 (1960). We have not examined whether the actual affinity of the en- 5. Monod, J., J. Wyman, and J. P. Changeux, J. Mol. Biol., 12, zyme for these effectors has been altered. Desensitization is 88 (1965). one of the characteristics of 6. Koshland, D. E., Jr., G. Nemethy, and D. Filmer, Bio- interesting allosteric ; chemistry, 5, 365 (1966). this term refers to the loss of the regulatory, but not the cata- 7. Koshland, D. E., Jr., in Current Topics in Cellular Regulation, lytic, function. Recently, we have observed that covalent at- ed. B. L. Horecker and E. R. Stadtman (Academic Press, tachment of a pyruvate group to the alanine site of GDH New York, 1969), p. 1. resulted in a loss of the regulatory response to ADP and GTP. 8. Koshland, D. E., Jr., and K. E. Neet, Annu. Rev. Biochem., It has also been 37, 359 (1968). reported that treatment of GDH with mercury 9. Shaw, K. P., and J. Kallos, Can. Fed. Proc., 13, 92 (1970). resulted in desensitization to various effectors (10). To ex- 10. Bitensky, M. W., K. L. Yielding, and G. M. Tomkins, J. plain this effect, it was suggested that the change in structure Biol. Chem., 240, 668 (1965). of the enzyme induced by mercury may interfere with the 11. Colman, R. F., and C. Frieden, J. Biol. Chem., 241, 3661 binding of regulatory molecules. Colman and Frieden (11) (1966). 12. Holzer, H., in Advances in Enzymology, ed. F. F. Nord studied the effect of of GDH; they reported a loss (Interscience, New York, 1969), Vol. 32, p. 297. of enzyme activity and extensive desensitization to GTP. 13. Ginsburg, A., Biochemistry, 8, 1726 (1969). Downloaded by guest on September 25, 2021