Choline Acetyltransferase and Acetylcholinesterase

CHOLINE ACETYLTRANSFERASE AND ACETYLCHOLINESTERASE

Haubrich, D. R. (1973), J. Neurochem. 21, 315. Ramasarma, J., and Wetter, L. R. (1957), Can. J. Bio- Liang, C., Segura, M., and Strickland, K. P. (1970), Can. chem. Physiol. 35, 853. J. Biocfiem. 48, 580. Sung, C., and Johnstone, R. M. (1967), Biochem. J. 105, Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, 497. R. J. (1951),J. Biol. Chem. 193, 265. Weinhold, P. A., and Rethy, V. B. (1972), Biochim. Bio- McCaman, R. E. (1 962), J. Biol. Chem. 237, 672. phys. Acta 276, 143. McCaman, R. E., and Cook, K. (1966), J. Biol. Chem. 241, Wittenberg, J., and Kornberg, A. (1953), J. Biol. Chem. 3390. 202. 431.

Choline Acetyltransferase and Acetylcholinesterase: Evidence for Essential Histidine Residues?

Robert Roskoski, Jr.

ABSTKACT: Choline acetyltransferase (EC 2.3.1.6) cata- dride also inactivates, and hydroxylamine reactivates, the lyzes the biosynthesis of acetylcholine according to the fol- partially purified electric eel acetylcholinesterase (EC lowing chemical equation: acetyl coenzyme A + choline + 3.1.1.7). High concentrations of acetylcholine substantially acetylcholine + coenzyme A. Ethoxyformic anhydride inac- protect against inactivation. The apparent pK a of the reactivates the enzyme prepared from bovine brain. Acetyl tive group is about 6.1. Inhibition by ethoxyformylation coenzyme A and coenzyme A, but not choline or acetylcho- which is reversed by hydroxylamine treatment provides evi- line, substantially protect against inactivation. The enzyme dence that histidine plays a role in the choline acetyltrans- is reactivated by hydroxylamine treatment. The apparent ferase and acetylcholinesterase reactions. pK, of the reactive group is about 6.5. Ethoxyformic anhy-

Acety~cholineis an established neurotransmitter at the choline, possible alkylating reagents, on choline acetyltrans- vertebrate neuromuscular junction and a probable, but not ferase. These studies failed to implicate histidine; instead, proven, transmitter in the vertebrate central nervous system the enzyme was inhibited by bromoacetylation of the active (cJ Iverson, 1970). Choline acetyltransferase (EC 2.3.1.6) site sulfhydryl (Roskoski, 1974b). catalyzes the following reversible reaction: acetyl coenzyme In the present studies ethoxyformic anhydride inactiva- A + choline F= acetylcholine + coenzyme A. Several stud- tion and reversal by hydroxylamine treatment implicate his- ies support the notion of an essential enzymic sulfhydryl tidine in the choline acetyltransferase and acetylcholinester- group. For example, thiol reagents inhibit choline acetyl- ase (EC 3. I .1.7) enzyme reactions. Furthermore, both en- transferase from squid head ganglia (Reisberg, 1954), pri- zymes are inhibited by N- acetylimidazole and their activity mate placenta (Schuberth, 1966), torpedo (Morris, 1967), spontaneously returns to control values within 1 hr. These and mammalian brain (Potter et al., 1968; Chao and Wolf- results are consistent with the hypothesis that choline acet- gram, 1973; Roskoski, 1974a). Experiments with the bovine yltransferase and acetylcholinesterase are inhibited by brain transferase suggest that an active site -SH reacts with chemical modification of enzymic histidine residues. acetyl coenzyme A to form an acetyl-thioenzyme intermediate (Roskoski, 1973, 1974a). This alleged thio ester inter- Experimental Section mediate, isolated by Sephadex gel filtration, further reacts Materials. Ethoxyformic anhydride and N- acetylimida- with choline to form acetylcholine. zole were purchased from Sigma Chemical Co. Electric eel Thio ester intermediates are also associated with the acetylcholinesterase (1058 units mg-I) was a product of glyceraldehyde-3-phosphate dehydrogenase (EC 1.2.1.12) Worthington Biochemical Corp. Acetylthiocholine chloride and papain (EC 3.4.4.10) reactions (Harris et a/., 1963; was purchased from Pfaltz and Bauer, Inc. Decamethonium Lowe and Williams, 1965). Crystallographic structural bromide was purchased from K and K Laboratories, Inc., analysis of the lobster muscle dehydrogenase (Buehner et and indoxyphenyl acetate, Calbiochem. al., 1973, 1974) and papain (Drenth et al., 1968) reveal a Methodology for Chemical Modification. For reaction histidine residue in close proximity to the active site -SH. with choline acetyltransferase, ethoxyformic anhydride, dis- The interaction of the imidazole with the sulfhydryl and solved in absolute ethanol, was added to give the specified with the substrates may be important in the catalytic mech- concentration of inhibitor. Ethoxyformic anhydride and N- anism of these enzymes. These findings prompted a study of acetylimidazole were dissolved in acetonitrile for the other the effects of bromoacetyl coenzyme A and bromoacetyl- experiments. These solutions were prepared immediately before use. A 1-4 solution of inhibitor was added to 100 @I

~~ ~-~~~~ ~ From the Department of Biochemistry, The University of Iowa, of enzyme solution to initiate the reaction unless specified lowa City, Iowa 52242. Received June 26, 1974. This work was sup- otherwise. Solvent alone was added to the control samples. ported by Grant NS-I 1310 of the U. s. Public Health Service. Acetylcholinesterase Assay. The spectrophotometric

BIOCHEMISTRY, VOL. 13, NO. 25, 1974 5141 KOSKOSKI

~~~~ ~

TABLI I : Ethoxyformic Anhydride Inhibition of Choline TABLE 11: Effect of Substrates on Ethoxyformic Anhydride Acetyltransferase." Inhibition of Choline Acetyltransferaw."

Enzyme Enzyme t n/ynie 4ctivity Actikity A, t IVI ty Concn (PM) (pmoli5 min) Concn (PM) (pmol/5 min) Addition Contn (".r; Control) 0 94 2 50 7s None ?(I 10 64 1 100 38 Acetyl 15 pv 0 2 25 17 9 1000 0 coenzyme A Coenzyme A 15 phl 62 "The enzyme extract (20 pg of protein) was incubated 5 Acetylcholine 2 mv '8 min at 37" in buffer C (SO mM potassium phosphate- 100 mv Choline 2, 10 mM 31 KCI-0.1 mM EDTA (pH 7.4)) with the specified concentra- -_ -- ~ tion of reagent in 100 pl. Then 10-111 aliquots were assayed The experiment was carried out a\ described in Table I for choline acetyltransferase activity as previously described except that the enzyme was preincubateti 5 min at 37" with (Roskoski, 1973) the specified iub\trate prior to the addition of ethoxyformic anhydride (25 pihi final) The control kalue was 88 pmol 5 min The apparent K, values were previously reported acetyl coenzyme A, 15 choline, 0 75 mv. coenzyme A, 20-200 procedure of Ellman and coworkers (1961) was used. Ali- p~: p~;acetylcholine 1-5 mv (Roskosbi, 1974a) quots of enzyme (to 20 PI) were added to 300 pl of 0.1 M potassium phosphate (pH 7.0), 2 mM acetylthiocholine, and 0.25 mM 5,5'-dithiobis(2-nitrobenzoic acid). The absorb- ance changes at 412 nm (1-cm path length) were followed on a Beckman 25 spectrophotometer. With indoxylacetate 98%. These results are consistent with the hypothesis that and p- nitrophenyl acetate as substrates, assays were carried enzyme inhibition is associated with ethoxyformylation of out using the methodology of O'Brien (1969) except that an enzymic histidine residue (Melchoir and Fahrney, 1970; the volumes were scaled down to 300 pl. Esterase assays Burstein et ul., 1974). were performed at ambient temperature. Determination of the pK, of the Reactive Kesidue. The Preparation of the bovine brain choline acetyltransferase pH dependence of the ethoxyformic anhydride inhibition of and the radiochemical enzyme assay were described pre- the choline acetyltransferase was measured after a 5-min viously (Roskoski, 1973). Sources of other materials are incubation. The apparent first-order rate constants (K~,,,,,) previously documented (Roskoski, 1973, 1974a). were calculated using the methodolog) of Burstcin and co. workers (1974) (Table 111). A-pK, of 6.5 f 0.3 was calculated from the slope of such a plot using the data given in Results Table IV. The experiment was performed in quadruplicate. General Characteristics of Ethoxyformic Anhydride In- Ethoxyformic Anhydride Inhibition of ..lcet,ylcholitiest- hibition of Choline Acetyltransferase. Ethoxy formic anhy- erase. Because of the evidence supporting the contention of dride proved to be a potent inhibitor of transferase activity essential histidine residues in acetylcholinestcrase ((j: (Table I). A 50 /IM concentration inhibits activity 92% Cohen and Oosterbaan, l963), ethoxyformic anhydride in- under standard experimental conditions (37'. 5 min, pH activation of the electric eel enzyme mas tested. Incubation 7.4). Although the reaction conditions differ. concentra- with 0.35 m~ ethoxyformic anhydride for 5 min at 23' (pli tions used for the ethoxyformylation of creatine kinase 7.0) inhibits enzyme activity 50%. Similar results are ob- (Pradel and Kassab, 1968), pepsin, and pancreatic RNase served using the neutral substrates indoxyl acetate and p- (Melchoir and Fahrney, 1970) are three to five orders of nitrophenyl acetate. Inactivation is proportional to reagent magnitude greater than those reported here. The inactiva- concentration, and good pseudo-first-order kinetics obtain tion is proportional to the reagent concentration and cxhib- during the 5-min reaction, its good pseudo-first-order kinetics during the 5-min reac- The rate of acetylcholinesterase inhibition was measured tion. Since the inhibitor hydrolyzes with a t 112 of aborit 27 in the presence of several substrates and inhibitors. Coni- min at pH 7 (Melchoir and Fahrney, 1970), these kinetics pounds which bind to the substrate anionic site (Froede and do not obtain during long incubations. Wilson. I971 including choline (40 nibi), thiocholine (40 The effects of the substrates on the inactivation are mM), and tetramethylammonium ion (30 KIM) failed to shown in Table 11. Acetyl coenzyme A and coenzyme A protect the enzyme against ethoxyformic anhydride inacti- substantially protect against ethoxyformic anhydride inhi- vation. Decamethonium (2 X IO-' M; K, = IO-' and bition, On the other hand, choline and acetylcholine do not d-tubocurarine (8 X lo-' M; K, = 3 X IO-.' 341, com- protect. pounds which bind to '1 postulated peripheral anionic site, Hydroxylamine Reversal of Ethoxyformic Anhydride distinct from the substrate anionic site (Krupka, 1966; Inactivation. To identify the amino acid residue associated Liooser and Sigman. 1974), also failed to alter the rate of with ethoxyformic anhydride inhibition, the effect of hy- enzyme inhibition, Acetylcholine (40 mM) and acetylthio- droxylamine on enzyme inactivated with 75 PM ethoxyfor- choline (40 mM), on the other hand, protected niore than mic anhydride was tested. Treatment of the inhibited trans- 90% against enzyme inactivation. ferase with 20 mM hydroxylamine for 5 and 30 min. fol- To establish the identity of the residue ;issociated with in- lowed by dialysis for 1 hr to remove NI-IZOH, reactivates activation. hydroxylamine reversibility n;~,tested. Incuba- the enzyme 72 and 96%, respectively. Similar treatment tion of 1.7 ~g of aceiylcholinesterasc (100 PI) with ethoxy- with 100 mbf NHzOH for 5 min reactivates the enzyme formic anhydride (0.50 mM) for 5 min decreases activity to

5142 BIOCHEMISTRY. VOL 13. \C) 25, 1914 CHOLINE ACETYLTRANSFERASE AND ACETYLCHOLINESTERASE

~~ ~

TABLE 111: pH Dependence of Ethoxyformic Anhydride TABLE IV: pH Dependence of Ethoxyformic Anhydride Inhibition of Choline Acetyltransferase.' Inhibition of Acetylcholine Esterase.'

PH Kn,,r,(min-9 PH K:,,,,,(min- ._____. 9 5.7 0.019 7.2 0,079 6.2 0.211 7.2 0.460 6.2 0.041 7.7 0.086 6.7 0 335 7.7 0.517 6.7 0.062 7.0 0.420

a The ethoxyformic anhydride treatment (20 p~)was car- ' Acetylcholinesterase (80 pg in 100 p1 of 5 mM potassium ried out as described in Table I except that 10 mM phosphate phosphate-100 mM KCI) was incubated with 0.75 mM ethoxy- was used at the specified pH. To,adjust the pH after the 5-min formic anhydride for 2.5 min at 23" at the specified pH. incubation, aliquots (20 p1) of 0.3 M potassium phosphate Then 204aliquots were assayed as described in the Experi- (pH 7.4) were added to the solution, and transferase activity mental Section. The rate constants and pK, were determined was measured in 10-p1 samples (Roskoski, 1973). The rate by the procedure of Burstein er af.(1974). constants and pK, were calculated by the procedure of Bur- stein and coworkers (1974).

t-amino, (Melchoir and Fahrney, 1970), and phenolate groups (Burstein et al., 1974) are not regenerated by hy- 30% of control. Aliquots of the enzyme were incubated with droxylamine treatment under the conditions of the present 20 mM NH20H for 5 and 30 min. Enzyme activity was 60 experiments. Al'though these functional groups may react and 92% of the control after this treatment. Dilution of the with ethoxyformic anhydride, hydroxylamine reversal indi- enzyme reaction mixture 50-fold prior to assay prevents sig- cates that such chemical modification is not inhibitory. The nificant hydroxylaminolysis of acetylthiocholine thereby reactive sulfhydryl groups of arginine and creatine kinase eliminating adventitious thiocholine generation. The pH de- (Pradel and Kassab, 1968) and glyceraldehyde-3-phosphate pendence on the rate of inactivation was measured (Table dehydrogenase (Ovldi and Keleti, 1970) are not modified IV). From these data, using the methodology of Burstein et by ethoxyformylation. Although acetylcholine protects al. (1 974), the pK, of the reactive group is 6.1 f 0.3. against thiol reagent inactivation of choline acetyltransfer- N-Acetylimidazole Inhibition of Choline Acetyltransfer- ase (1974a), it fails to protect against ethoxyformic anhy- ase and Acetylcholinesterase. Reaction of N- acetylimidaz- dride inactivation, indicating that modification of the active ole with polypeptidic histidine and cysteine produces ad- site sulfhydryl is not the basis of the latter inhibition. The ducts which undergo spontaneous hydrolytic regeneration. characteristics of the ethoxyformylation of the active site Dilute hydroxylamine treatment, however, is necessary to serine hydroxyl of a-chymotrypsin (Melchoir and Fahrney, cleave 0-acetyltyrosine formed during the N- acetylimidaz- 1970) differ substantially from the type of inhibition ob- ole reaction (Pontremoli and coworkers, 1966). served for acetylcholinesterase reported here. For example, N-Acetylimidazole (1 .O mM) treatment of choline acet- the rate of spontaneous hydrolysis and reactivation at pH 7 yltransferase for 10 min at 23' (pH 6.7) inactivates the en- of a-chymotrypsin and acetylcholinesterase are 29 min and zyme 40%. Enzyme activity spontaneously returns to con- about 48 hr, respectively. Secondly, hydroxylamine reversal trol values within 1 hr. Acetyl'coenzyme A (10 pM) fully of the inhibited chymotrypsin requires a few seconds and protects against inactivation. Similarly, after N- acetylim- that of acetylcholinesterase, 45 min. These experiments idazole (0.25 mM) inhibition of acetylcholinesterase (23O, support the notion that inhibition of the acetylcholinester- 10 min, pH 7.0), enzyme activity spontaneously reactivates ase is associated with histidine, and not serine, modifica- from 52 to 96% within 1 hr. Acetylcholine (40 mM) almost tion. completely protects against this inactivation. The failure of The pseudo-first-order kinetics of inhibition suggests that Ellman's and other thiol reagents to inactivate acetylcholi- a modification of a single residue is inhibitory. Experiments nesterase suggests that N- acetylimidazole inhibition is un- with highly purified enzymes, however, are required to rig- related to cysteine modification with spontaneous reactiva- orously establish this hypothesis (Burstein and coworkers, tion. Although an active site thiol has been postulated for 1974). The apparent pK, of each reactive residue (6.5 for choline acetyltransferase, acetylation of this group leads to choline acetyltransferase and 6.1 for acetylcholinesterase) a chemically competent enzyme which undergoes deacety- is reasonable for an enzymic histidine (Tanford and Hauen- lation in the presence of acceptor choline (Roskoski, 1973; stein, 1956). Because enzymes of only 10% purity were used 1974b). This deacetylation, which involves a single turn- in the present experiments, spectral confirmation of histi- over, is assumed to occur rapidly. Therefore, the N- acetyli- dine modification was unfeasible (Burstein et al., 1974). midazole inhibition of choline acetyltransferase probably N- Acetylimidazole inhibition and spontaneous reactivation involves a residue other than the active site -SH group. provide additional indirect evidence for the involvement of Spontaneous reactivation is consistent with the hypothesis an enzymic histidine. that histidine is the chemically modified residue in these en- The inhibition of enzyme activity by chemical modifica- zymes. tion of histidine does not establish that the imidazole is in- volved in the chemistry of the catalytic process. The modifi- Discussion cation may, for example, alter the tertiary structure of the Ethoxyformic anhydride inhibits choline acetyltransfer- enzyme. That the reactive histidine forms a part of the ac- ase and acetylcholinesterase. Reactivation by hydroxyl- tive site of choline acetyltransferase is indicated by the find- amine suggests that histidine is the chemically modified res- ing that acetyl coenzyme A and coenzyme A substantially idue (Melchoir and Fahrney, 1970). Guanidino, a- and protect against inhibition. In the case of acetylcholinester-

BIOCHEMISTRY, VOL. 13, YO. 25, 1974 5143 ROSKOSKI ase, high concentrations of acetylcholine also protect. Sub- References stances which bind to the substrate anionic site (choline. tetramethylammonium ion) and to the peripheral anionic Buehner, M., Ford, G. C., Moras, D., Olsen, K. W., and site (decamethonium and d- tubocurarine), on the other Rossman, M. G. (1973), Proc. Nut. Acad. Sci. U. S. 70, hand, do not alter the rate of inactivation. That the residue 3052. modified by ethoxyformylation is unrelated to binding cat- Buehner, M., Ford, G. C., Moras, D., Olsen, K. W., and ionic substrates is also substantiated by the finding that the Rossmann, M. G. (1974), J. Mol. Biol. (in press). rate of enzymic hydrolysis by uncharged substrates (p-ni- Burstein, Y., Walsh, K. A,, and Neurath, H. (l974), Bio- trophenyl acetate and indoxyl acetate) is decreased to the chemistry 13, 205. same extent as that of acetylthiocholine. This is in contrast Chao, L. P., and Wolfgram, F. (1973), J. Neurochem. 20, to the bovine erythrocyte acetylcholinesterase inhibition by 1075. aLiridinium substances where acetylcholine hydrolysis is in- Cohen, J. A,, and Oosterbaan, R. A. (1963), in Cholinester- hibited 93% and p- nitrophenyl acetate hydrolysis is inhibit- ase and Anticholinesterase Agents, Eichler, O., and ed only 48% (O'Brien, 1969). Farah, A,, Ed., West Berlin, Springer-Verlag, pp 299- The acetylcholinesterase reaction proceeds by a two-step 373. chemical mechanism (cf Froede and Wilson. 1971). The Drenth, J., Jansonius, .I. N., Koekoek, R., Swen, H. M., and active site serine -OH reacts with acetylcholine to form an Wolthers, €3. G. (l968), Nature (London) 218, 929. acetyl--enzyme intermediate. Water then reacts with the in- Ellman, G.L., Courtney, K. D., Andres, V. Jr., and Feath- termediate to form acetate and the regenerated enzyme. erstone, R. M. (1961), Biochem. Pharmacol. 7, 88. Histidine has previously been implicated in the enzyme Froede, H. C., and Wilson, 1. B. (l97l), Enzymes, 3rd Ed. mechanisni (cf Cohen and Oosterbaan, 1963). The evi- 5, 87. dence supporting this contention is largely based on pH de- Harris, J. I., Meriwether, B. P., and Park, J. H. (1963), pendence of the enryme reaction with substrate (Mounter Nature (London) 198, 154. et ul.. 1957; Krupka, 1966) and inhibitors such as 2.4-dini- Iverson, L.. L. (l970), in The Neurosciences, 2nd ed, trolluoroberuene and diazo compounds (Mounter et al., Schmitt, F. 0.. Ed., New York, N.Y., Rockefeller Uni- 1957). The present experiments provide additional evidence versity Press, pp 768-78 1. for a role of histidine in the acetylcholinesterase reaction. Jencks, W. P., and Carriuolo, J. (1 959), J. Biol. Chem. 234, In addition to possible structural and binding roles, histi- 1272. dine may play two other, and not mutually exclusive, roles Krupka, R. M. (1966), Biochemistry 5, 1983. in the acyl transfer reactions of choline acetyltransferase Lowe, G.,and Williams, A. (1965), Biochem. J. 96, 189. and acetylcholinesterase. First, it may enhance the reactivi- Melchoir, W. B., Jr., and Fahrney, D. (1970). Biochemistry ty of the respective active site nucleophiles (--Sii in the for- 9, 251. mer and ~ OH in the latter). Second, it may function as a Mooser, G.,and Sigman, D. S. (1974), Biochemistry 13, general acid-base catalyst by donating or accepting protons 2299. from substrates or fragments thereof. Although transient Morris. D. (1967), J. Neurochem. 14, 19. acetylimidaroles have not been exhaustively excluded from Mounter, L. A. Alexander, H. C., Tuck, K. D., and Dien, L. the catalytic mechanisms (Cohen and Oosterbaan, 1963), T. H. (I 957), J. Biol. Chem. 226, 867. there is little, if any, direct support for this contention. In O'Brien, R. D. ( 1969), Biochem. J. 113, 7 13. the case of choline acetyltransferase, the alleged acetyl- Ovidi, .J., and Keleti, T. (1970). Arch. Biochem. Biophys. ent.ynie intermediate is isolable by Scphadex gel filtration. Acad. Sci. Hung. 4. 365. 7'he bond between the acetyl group and enqmc is stable in Pontremoli, S., Grazi, E., and Accorsi, A. (l966), Biochem- 10(vn trichloroacetic acid (90'. 20 min). but is readily i,ctry 5, 30 72. cleaved by dilute alkali (pti IO) (Koskoaki, 1973). This is Potter. L.T., Glover, V. A. S., and Saelens, J. K. (1968). J. riot cori\i>tent with an acctqlirnidazole bond, which is unsta- Biol. Chem. 243, 3864. ble in acid and alkali, and even undergoes hydrolysis at pH Pradel, L.-A,, and Kassab, R. (l968), Biochim. Biophys. 7 uith a I ,? of 2 hr (Jencks and Carriuolo. 1959). The pos- Acta 167, 3 1 7. tulatcd acetyl en/.yme bond is cleaved by performic acid Reisberg, R. B. (1 954), Biochim. Biophys. Acta 14, 442. oxidation arid these chemical characteristics are those of a Roskoski, R., Jr. (1 973), Biochemistry 12, 3709. thio ester (Koskoski, 1973). Further experiments are re- Roskoski, R., Jr. (1974a), J. Biol. Chem. 249, 2156. quired to show that the rate constants for the formation and Roskoski, R., Jr. (1974b), Biochemistr,v 13, 2295. further reaction of the postulated thio ester intermediate Schuberth, J. (1966), Biochim. Biophys. Acta 122, 470. are adequate to account for the observed rate of the en7yme Tanford. C., and Hauenstein, J. P. (1 956). J. .4mer. C'hrm. reaction So(,.78. 5287.