In the present work, we have investigated the molecular. in view of the hypothesis of Koelle et al., what relationship

Proc. Natl. Acad. Scf. USA Vol. 75, No. 6, pp. 2588-2592, June 1978 Biochemistry "Nonspecific" cholinesterase and acetylcholinesterase in rat tissues: Molecular forms, structural and catalytic properties, and significance of the two enzyme systems (sedimentation coefficients/quatemary structure/excess substrate inhibition/active site titration) MARC VGNY*, VCTOR GSGRt, AND JAN MASSOULO* * Laboratoire de Neurobiologie, cole Normale Superieure, 46, rue d'ulm, 75230 Paris Cedex 05, France; and t Unite 153 de l'nstitut National de la Sant6 et de la Recherche MWdicale, Hapital de la Salp6triere, 47, boulevard de 'Hopital, 75005 Paris, France Communicated by David Nachmansohn, March 3,1978 ABSTRACT "Nonspecific" cholinesterase (acylcholine acylhydrolase; C 3.1.1.8) from various rat tissues has been found to exist in several stable molecular forms that appear as exact counterparts of molecular forms of acetylcholinesterase (acetylcholine hydrolase; C 3.1.1.7) The sedimentation pattern of cholinesterase was similar to that of acetylcholihesterase with a small but significant shift between the sedimentation coefficients of the corresponding forms. xtraction yields in different media also demonstrated a close parallelism between the two enzyme systems. Other properties, such as thermal stability and catalytic characteristics, indicated both differences and similarities. n spite of the structural resemblance implied by their physicochemical properties, cholinesterase did not crossreact with antibodies against acetylcholinesterase. The nature of the relationships revealed by these studies and their bearing on the physiological significance of cholinesterases are discussed. ver since their discovery in mammalian serum, in 1943, by Mendel and Rudney (1), the so-called nonspecific cholinesterases (Chol) (acylcholine acylhydrolase; C 3.1.1.8) have remained a puzzle. The nature of their differences from "true" or specific acetylcholinesterases (AcChol) (acetylcholine hydrolase; C 3.1.1.7), which hydrolyze acetylcholine faster than other choline esters, and, more importantly, the physiological significance of these cholinesterases have not yet been fully understood (2, 3). Cholinesterases with different specificities occur in the serum of vertebrates and have been extensively studied by Augustinsson (4). The presence of Chol in sympathetic ganglia has also been recognized for a long time (5). Klingman et al. have shown that in rat sympathetic ganglia (e.g., the superior cervical ganglion, SCG) a large proportion of the acetylcholine hydrolysis is due to Chol (6). Koelle et al. (7) have recently reported that AcChol and Chol have the same cytochemical localization in cat sympathetic ganglia (superior cervical and stellate ganglia) and that after irreversible inactivation by phosphorylating agents, the recovery of AcChol was dependent upon the activity of Chol (8, 9). These authors expressed the hypothesis that Chol is a precursor of AcChol, and that the synthesis of this enzyme is regulated by the level of Chol. The catalytic site of Chol is not essentially different from that of AcChol, since it possesses both an anionic subsite and an active serine-esteratic subsite (10, 11). AcChol is, however, thought to possess additional anionic centers (12), included perhaps in a peripheral site which appears responsible for excess substrate inhibition (for a review, see ref. 13). The difference between the two classes of enzymes, Ac- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. 1734 solely to indicate this fact. Chol and Chol, would thus reflect regulatory interactions between catalytic and peripheral sites, which might result from post-transcriptional modifications of a single polypeptide, as suggested by Koelle et al. (8, 9). n the present work, we have investigated the molecular properties of rat ganglion (SCG) and muscle Chol and compared them with those of AcChol. t was interesting to know, in view of the hypothesis of Koelle et al., what relationship Chol might bear to the multiple molecular forms of AcChol, which we have recently characterized by sedimentation analysis in rat brain (14), muscle (15), and sympathetic ganglia (16). These forms are stable molecular entities. They differ in their sedimentation coefficients and also in their solubility: the 4S ganglion and brain AcChol are readily soluble in aqueous media, while the 10S form may be quantitatively solubilized in a detergent-containing buffer. A 6.5S and a 16S form, which also require detergent for complete solubilization, were also found in the SCG. n muscle, the 16S form is restricted to the endplate regions (15), as observed by Hall (17). n addition, the lighter forms appear to be precursors of the heavier molecules (18, 19). Since the hydrolytic activity per active site is identical for all forms (20), we believe that the multiplicity of molecular forms possesses a structural rather than a catalytic significance. n the present work we describe the occurrence of multiple molecular forms of Chol, the properties of which closely parallel those of AcChol. We also compare molecular and catalytic properties of the two enzymes. These observations are relevent to the possible relationships between "nonspecific" Chol and AcChol. MATRALS AND MTHODS All reagents used in this study were of analytical grade. 1,5- Bis(4-allyldimethylanonium phenyl)pentan-3-one dibromide (BW 284 C 51) was obtained from Wellcome Reagents Ltd. and 10-(2-diethylaminopropyl)phenothiazine hydrochloride (ethopropazine; Parsidol) from Rh6ne-Poulenc. Hydrolytic activities were measured by the method of llman et al. (21), with either acetylthiocholine or butyrylthiocholine as substrate. The assay medium contained 0.75 mm substrate, 0.5 mm di- Abbreviations: AcChol, acetylcholinesterase (C 3.1.1.7); Chol, "nonspecific" cholinesterase (C 3.1.1.8); SCG, superior cervical ganglion; DFP, diisopropylfluorophosphonate; BW 284 C 51, 1,5- bis(4-allyldimethylammoniumphenyl)pentan-3-one dibromide; ethopropazine (Parsidol), 10-(2-diethylaminopropyl)phenothiazine hydrochloride; MTP, O-ethyl-S-(diisopropylamino-2)methanethiophosphonate; GTA, ethylene glycol bis(2-aminoethyl ether)-n,n'- tetraacetic acid. 2588

- Biochemistry: Vigny et al. 100[. 801 r._ 60V 08 U 40L 20 Ao fr 10-i 10-7 10-6 10-5 -8 10-7 106 10-5 BW284C51, M thopropazine, M FG. 1. ffect of selective inhibitors BW 284 C 51 and ethopropazine on acetylthiocholine and butyrylthiocholine hydrolysis by SCG extracts. Activities were assayed by the llman method (see Materials and Methods) in the presence of various concentrations of ethopropazine and BW 284 C 51, and are expressed as percentages of control rates. 0, Hydrolysis of butyrylthiocholine (7.5 X 10-4 M); A, hydrolysis of acetylthiocholine (7;5 X 10-4 M). thiobisdinitrobenzoic acid, and 50 mm phosphate buffer (ph 7.0) and the reaction was monitored at 280, 412 nm, in cuvettes of 1-cm pathlength containing 1 ml of medium. We used BW 284 C 51 and ethopropazine as specific inhibitors of AcChol and Chol. Unless otherwise stated, extracts from rat SCG or muscle were obtained by homogenizing the tissues in 1 M NaCl/50 mm MgCl2/10 mm Tris-HCl (ph 7)/1% Triton X-100 (Triton/ saline medium) as in ref. 14 (for SCG, 10 ganglia were homogenized in 0.4 ml of medium). These extracts contained more than 95% of AcChol or Chol activities. Solubilization in the absence of detergent was also studied by homogenizing the tissues in water. The homogenates were centrifuged at 17,000 X g for 20 min before the enzymic activities were assayed or the sedimentation patterns were analyzed. Sedimentation analyses were performed as described (14), in 5-20% (wt/vol) sucrose gradients in the Triton/saline medium. Horse liver alcohol dehydrogenase (4.8 S), beef liver catalase (11.3 S), and scherichia coli f-galactosidase (16 S) were used as sedimentation standards. Chol activity was detected after sedimentation in sucrose gradients either by using butyrylthiocholine as substrate (in the presence or absence of 10 MM BW 284 C 51) or by using acetylthiocholine in the presence of 10MM BW 284 C 51. AcChol activity was determined with acetylthiocholine in the presence of 10 MM ethopropazine. Active sites were titrated with the irreversible, stable inhibitor O-ethyl-S-(diisopropylaminoethyl-2)-methanethiophosphonate._ Cu DFP, M FG. 2. nhibition of SCG Chol by DFP. The extract of SCG in Triton/saline was incubated for 1 hr with the indicated DFP concentrations at 20 before being assayed for hydrolysis of acetylthiocholine, in the absence of additional inhibitor (A) (AcChoJ + Chol) and in the presence 10MuM ofbw 284 C 51 (A) (Chol), and for hydrolysis of butyrylthiocholine (0) (Chol). c *13 0 0. cm Proc. Natl. Acad. Sci. USA 75(1978) 2589 / Clo >.5 0> 10-5 10-4 10-3 10-2 Substrate, M FG. 3. Hydrolysis of acetylthiocholine by SCG AcChol and Chol as a function of substrate concentration. Rat SCG were homogenized in 1 M NaCi/50 mm Tris-HCl (ph 7)/1% Triton X-100, and 10-il aliquots of the extracts (2.5 mg of protein/ml) were assayed in media containing 0.5 mm dithiobisdinitrobenzoic acid, 50 mm phosphate (ph 7.0), 10MM specific inhibitor (BW 284 C 51 or ethopropazine), and variable concentrations of acetylthiocholine iodide. *0*, AcChol; --- o-, Chol. (MTP) as described (20). This inhibitor was allowed to react with the enzymes for 4 hr at 200 in the presence of either ethopropazine or BW 284 C 51 (10MgM). There was no change after a prolonged incubation (24 hr). An immunoglobulin preparation from the serum of rabbits that had been immunized against mouse AcChol was kindly donated by. Adamson (University of Oxford) (22). This preparation was free of Chol activity. mmunoprecipitation experiments were performed as follows: equal volumes of SCG extract and diluted immunoglobulin preparation were allowed to react overnight at 40, the-mixtures were centrifuged (10,000 X g for 30 min), and the supernatants were assayed for both AcChol and Chol activities and analyzed by sucrose gradient centrifugation. RSULTS stimation of AcChol and Chol Activities. Rat SCG extracts (obtained by homogenizing the ganglia in Triton X- 100/saline buffer) hydrolyze both acetylthiocholine and butyrylthiocholine. These reactions are completely inhibited in the presence of 10 MM eserine and must therefore be due to AcChol and Chol. The effects of selective inhibitors of AcCho (BW 284 C 51) and Chol (ethopropazine) are shown in Fig. 1. BW 284 C 51, up to 10-5 M, does not inhibit more than 2-3% of the hydrolysis of butyrylthiocholine, indicating that AcChol does not contribute significantly to this reaction. nhibition by 10-5 M ethopropazine is, on the contrary, quantitative, showing that it is entirely due to Chol. The hydrolysis of acetylthiocholine, on the other hand, is partially inhibited by both inhibitors, reaching a plateau at about 10-5 M in each case. The residual activities observed in the presence of 10-5 M BW 284 C 51 and ethopropazine add up to 100%, suggesting that they reflect the relative contributions of AcChol and Chol to acetylthiocholine hydrolysis. This conclusion was confirmed by diisopropylfluorophosphonate (DFP) inhibition experiments (Fig. 2), which were based upon the ability of DFP to block Chol irreversibly at lower concentrations than AcChol. The inhibition curves obtained thus showed two phases, which corresponded quantitatively to the contributions of AcChol and

-2-590 0 r. Biochemistry: Vigny et al. Proc. Nati. Acad. Sci. USA 75 (1978) C041.0 o 0 1 2 3 nhibitor, pmol FG. 4. Active site titration of AcChol and Chol. SCG extracts; were incubated for 4 hr at 200 with varying concentrations of the irreversible inhibitor MTP in the presence of 10 MAM ethopropazine (AcChol) or 10 MM BW 284 C 51 (Chol). The samples were then assayed for hydrolysis of acetylthiocholine at 28 in the presence of the same specific reversible inhibitors. Residual activity is plotted against MTP quantities (in 1 ml). *0*, AcChol; --- -1, Chol. The decrease in activity was linear with MTP concentration in both cases, indicating that the experimental conditions used allowed active site titrations. Chol. n addition, the hydrolysis of butyrylthiocholine and of acetythiocholine in the presence of 10 ;tm BW 284 C 51 was totally inhibited at a low concentration of DFP, confirming that they are indeed catalyzed by Chol almost exclusively. By using the specific anti-acchol inhibitor, we were thus able to examine the substrate specificity of Chol. At 10-3 M substrate concentration, in the presence of 10 AM BW 284 C 51, the rate of hydrolysis of acetylthiocholine by the enzymes contained in a detergent/saline SCG extract was studied as a function of substrate concentration (Fig. 3). AcChol activity exhibited the characteristic excess substrate inhibition above 10-3 M acetylthiocholine; Chol activity deviated from michaelian kinetics and did not reach saturation below 2 X 10-2 M. n the 10-2 M concentration range, the two activities were approximately equal. >~~~~~~~~~~~~~~~ 'r-0.5 0 10 20 FG. 6. Sedimentation pattern of muscle AcChol and Chol. The extract was obtained from a neural zone of sternocleidomastoidian muscle. The experimental conditions were identical to those of Fig. 5A. Activities ofacchol (0-0) and Chol (o- - -0) are plotted on an arbitrary scale. n muscle extracts, the Chol contribution to acetylthiocholine hydrolysis was relatively smaller (approximately 10-15% instead of about 30%, under our assay conditions), and in brain extracts we did not detect any Chol activity. AcChol and Chol Active Site Concentrations. An SCG extract was incubated with various known concentrations of the irreversible inhibitor (MTP) in the presence of ethopropazine or BW 284 C 51, which inhibit its reaction with Chol and AcChol, respectively (20). After completion of the reaction, the remaining activity (e.g., AcChol after incubation with MTP plus ethopropazine), representing an excess of active sites over inhibitor, was plotted as a function of MTP (Fig. 4). The straight titration lines intersect the horizontal axes at a point that represents the number of active sites. Turnover numbers were ½1 20 10 20 _ ~0.5 ~0.5- :t ~~~~~~~~ > a - FG. 5.Sedimetationpatternof SCGAc~hoan htearosidctte poito of th markrezms oa,-aats 0 0 10 20 10 20 FG. 5. Sedimentation pattern of SCG AcChol and Chol. The arrows indicate the position of the marker enzymes. coli fl-galactosidase (16 S), beef liver catalase (11.3 S), and horse liver alcohol dehydrogenase (4.8 S). (A) Centrifugation in a Beckman SW 56 rotor at 32,000 rpm for 15 hr at 40. * *, AChol activity was measured with acetylthiocholine as substrate, in the presence of 10MM ethopropazine. The indicated absorbance variation corresponds to an incubation time of 24 min (50-gl aliquot in 1 ml of assay medium). --- -0, Chol activity was measured with butyrylthiocholine as substrate, in the presence of 10MtM BW 284 C 51. The absorbance indicated corresponds to an incubation time of 110 min (50-Al aliquot in 1 ml of assay medium). -..-a, Chol activity, determined after 340 min of incubation. The 16.6S peak was totally inhibited when 10MM ethopropazine was added to the Chol assay medium, excluding the possibility that it might be due to AcChol activity, which would escape BW 284 C 51 inhibition. n this experiment, the abundance of the heavy forms was exceptionally high (8% for 16.6S AcChol and 0.9% for 17S Chol). Since seasonal variations have been noticed (16), it may be relevant that the extract used was prepared in early October; it was centrifuged immediately after homogenization. (B) Centrifugation at 49,000 rpm for 15 hr at 4. The faster migration allows a better separation and characterization of the minor 6.3S (AcChol) and 7S (Chol) peaks. Activities are plotted on an arbitrary scale.

Biochemistry: Vigny et al. Table 1. Proc. Natl. Acad. Sci. USA 75(1978) 2591 Activities of AcChol and Chol and relative proportions of the different molecular forms xtract AcChol Chol Aqueous extract Activity 720 270 Relative proportion 3.6S 6.3S 10.OS 16.6S 4.2S 7S 10.8S -17S 80% "-0 20% -0 85% --0 15% "-0 Saline/detergent extract Activity 1485 630 Relative proportion 3.6S + 6.3S 10.OS 16.6S 4.2S + 7S 10.8S -17S 51% 45% 4% 55% 45% <1% Activities are expressed as nmol of acetylthiocholine hydrolyzed per hr per ganglion. The contributions of the two lighter forms have not been separately estimated in the total Triton/saline extract because the corresponding peaks are not sufficiently well resolved. The apparent sedimentation coefficients were determined by centrifugation in a 5-20% (wt/vol) sucrose gradient containing 1 M NaCl/50 mm MgCl2/10 mm Tris-HCl, ph 7/1% Triton X-100. computed, and the value obtained for AcChol was identical with that determined previously for brain AcChol (i.e., in the absence of ethopropazine) (20). The SCG extracts contained approximately 2.8 pmol of Ac- Chol active sites and 2.2 pmol of Chol active sites per mg of protein. Under our assay conditions (7.5 X 10-4 M acetylthiocholine in 0.1 M phosphate buffer, ph 7, at 28 ), the corresponding turnover numbers for acetylthiocholine hydrolysis were, respectively, 1.35 X 107 mol/hr per site and 0.8 X 107 mol/hr per site. Molecular Forms. Figs. 5 and 6 show that the sedimentation pattern of Chol activity is remarkably similar to that of Ac- Chol activity in SCG and muscle extracts. n SCG extracts there are three main peaks of Chol which sediment slightly faster (about 0.5 S) than the main 3.6S, 6.3S, and 10S AcChol molecular forms. The relative proportions of these three components are almost identical for both activities (Table 1). n the bottom region of the gradient, after prolonged incubation times, we observed a small 17S Chol peak, which amounts to less than 1% of the total Chol activity (Fig. 5A). The same relationship between Chol and AcChol profiles was observed in muscle extracts, but the 6.3S AcChol and its corresponding 7S Chol form possessed a much lower relative activity. The lower Chol activity did not allow us to detect any heavy 17S form. n the same manner as for AcChol molecular forms (15, 16) we found that the Chol peaks correspond to stable entities, which retain their characteristic sedimentation after isolation from the gradients. Solubilization Properties. The AcCho and Chol activities that were solubilized from SCG by homogenization in water and in the detergent/saline medium are reported in Table 1, together with the relative proportions of molecular forms in both extracts. We observed that the aqueous extract contained approximately 60% of the activity observed in the detergent/ saline extracts for Chol as well as for AcChol. Moreover, in the absence of detergent, most of the solubilized activity corresponds to the light molecular forms of both Chol (4.2 S) and AcChol (3.6 S). An important proportion of Chol was thus only solubilized in the detergent/saline medium. Thermal nactivation. We studied the thermal inactivation of the AcChol and Chol activities contained in SCG extracts. Their inactivation rates are different, and they are modified in opposite directions in the presence of the divalent cation Mg2+ (Fig. 7). n contrast to the molecular properties we have examined so far, this experiment illustrates a marked difference between the two enzymes. However, only the 1OS form of Ac- Chol and the corresponding 10.8S form of Chol remained after 90% inactivation in every case. The fact that the inactivation did not follow first-order kinetics may be related to the molecular heterogeneity of the enzymes. mmunological Characterization. A Triton/saline extract of SCG was incubated with various dilutions of rabbit antimouse AcChol immunoglobulin (22). Fig. 8 shows that Ac- Chol was precipitated, whereas Chol was not. Sucrose gradient analysis of the supernatants demonstrated further that ' ċj.4 - Q._ 0 5 10 15 0 5 10 15 Time, min FG. 7. Thermal inactivation of SCG AcChol and Chol. Rat SCG were homogenized with a detergent/saline buffer without divalent cations [a small concentration of ethylene glycol bis(2-aminoethyl ether)-nn'-tetraacetic acid (GTA) was included in order to chelate endogenous ions] and the extracts were submitted to thermal inactivation at 55. AcChol (0-0) and Chol (o--- -0) activities were measured as a function of time and plotted on a semilogarithmic scale. (A) n the homogenization buffer (1 M NaCl/5 mm GTA/50 mm Tris-HCl, ph 7.0/1% Triton X-100). (B) After addition of 50 mm MgCl2. 1 1:2 14 1:8 1:16 132 164 1:128 1:256 Dilutions FG. 8. mmunoprecipitation of AcChol and Chol. The supernatant activity is plotted as a function of immunoglobulin dilution, after incubation of SCG extracts in Triton/saline, with equal volumes of a rabbit anti-mouse AcChol immunoglobulin preparation (overnight at 40), and centrifugation (10,000 X g, 30 min). 0-0, AcChol; o--- -o, Chol.

2.5-92 Biochemistry: Vigny et al. in all cases the Chol peaks remained unaffected, although AcChol peaks were shifted even at immunoglobulin dilutions that precipitated only 30% of AcChol activity. Proc. Nati. Acad. Sci. USA 75 (1978) DSCUSSON "Nonspecific" cholinesterase (Chol) from rat ganglia (SCG) or muscle has been characterized by using specific inhibitors. t does not exhibit excess substrate inhibition. We have shown that Chol from these tissues can be fractionated by sucrose gradient centrifugation into distinct molecular forms which closely parallel the AcChol molecular forms. n spite of the remarkable similarity of the two sedimentation profiles, AcChol and Chol activities are carried by distinct macromolecules, as shown by a difference in apparent sedimentation coefficients (about 0.5 S), which is small but systematic and significant. Thus, every AcChol molecular form possesses its Chol counterpart, in SCG as well as in muscle extracts, and except for the heavy 16-17S components, the relative abundance of the corresponding forms is practically identical. We also observed that muscle extracts possess a smaller activity in the 6-7S region for both enzymes. The heavy 17S form of Chol was, however, markedly less abundant, in relative proportion, than 16.6S AcChol. We were able to observe it in SCG extracts, but its detection in muscle extracts was difficult. t would be interesting to know whether this heavy form, if it does occur in muscle at all, is restricted to the endplate region like its AcChol counterpart. The parallelism between AcChol and Chol molecular forms extends to their solubility properties: in SCG, the lighter forms, 3.6S AcChol and 4.2S Chol, were preferentially extracted in water, in the same proportions. On the contrary, the other molecular forms were quantitatively solubilized only in the presence of the nonionic detergent Triton X-100. Thus, some of the Chol molecular forms, like those of AcChol, cannot be considered soluble. Our previous studies suggested that multiple forms of Ac- Chol represent quaternary assemblies which contain equivalent catalytic subunits at different levels of integration, representing either soluble or membrane-bound states. At first sight, this might seem consistent with the hypothesis of Koelle et al. (8,9) that AcChol and Chol specificities are obtained by some post-transcriptional modification of a single basic subunit without altering its participation in the different quaternary interactions which result in the macromolecular heterogeneity of the enzyme. This idea is not supported, however, by some of our observations. First, the thermal inactivation experiments demonstrated a marked difference between the two proteins. Second, we have never observed any interconversion between the two activities under various incubation conditions. Finally, there was no immunological crossreaction between the two enzyme systems. n addition to the lack of common antigenic determinants, our results imply that hybrid assemblies containing the two kinds of catalytic subunits do not occur. Furthermore, our previous results suggested that the lighter forms of AcChol behave as precursors of the heavier ones, and this would not seem compatible with the precursorproduct relationship between the two series of molecular forms which would be implied in the hypothesis of Koelle et al. (8, 9). t thus appears likely that AeChol and Chol catalytic units correspond to distinct genes. This hypothesis offers a simple basis for the wide differences observed in Chol specificity and level among animal species. For instance, we have found that bovine SCG extracts, although containing the same AcChol molecular forms as rat SCG extracts, do not possess any significant Chol. The number of Chol active sites in rat SCG is approximately equal to the number of AcChol sites. Leaving aside the question of substrate specificity, which may be more relevant to enzymologists than to physiologists, we observe that in rat SCG extracts the contribution of Chol to acetylthiocholine (and most likely acetylcholine) hydrolysis is about half that of AcChol at 7.5 X 10-4 M substrate and equivalent at 10-2 M substrate. t is conceivable that, because of diffusion barriers, the effective concentration of acetylcholine at the active site under physiological conditions may reach values at which Chol is as efficient as, or more efficient than, AcChol. The structural similarities between AcChol and Chol, as demonstrated in this work as well as by the experiments of Koelle et al. (8, 9), which indicated cross-regulatory phenomena between the two systems, seem to imply that Chol possesses a precise physiological role, closely related to that of AcChol. ndeed, the multimolecular associations of these enzymes, which appear to be exactly homologous, probably play an essential role in their function in situ. We would therefore like to suggest that the two enzyme systems fulfill very similar physiological functions. The precise requirements for acetylcholine hydrolysis would be met, in various systems, by different combinations of the two genetically distinct, yet structurally similar, enzymes. We are very grateful to Dr. ileen Adamson for the generous gift of rabbit anti-mouse AcChol immunoglobulins. This work was supported by grants from the Centre National de la Recherche Scientifique, the DMlegation Gen6rale a la Recherche Scientifique et Technique, and the nstitut National de la Sant6 et de la Recherche h4dicale. 1. Mendel, S. & Rudney, H. (1943) Biochem. J. 37,59-63. 2. Koelle, G. B. (1963) Cholinesterases and Anticholinesterase Agents (Springer, Berlin). 3. Silver, A. (1974) The Biology of Cholinesterases (North-Holland, Amsterdam). 4. Augustinsson, K. B. (1959) Acta Chem. Scand. 13, 571-592, 1081-1096. 5. Koelle, G. B. (1953) Biochem. J. 5,217-226. 6. Klingman, G.., Klingman, J. D. & Poliszczuk, A. (1968) J. Neurochem. 15, 1121-1130. 7. Koelle, G. B., Davis, R., Smyrl,. G. & Fine, A. V. (1974) J. Histochem. Cytochem. 22,252-259. 8. Koelle, W. A., Smyrl,. G., Ruch, G. A., Siddons, V.. & Koelle, G. B. (1977) J. Neurochem. 28,307-311. 9. Koelle, G. B., Koelle, W. A. & Smyrl,. G. (1977) J. Neurochem. 28,313-319. 10. Bergman, F. & Wursel, M. (1954) Biochim. Biophys. Acta 13, 251-259. 11. Bergman, F., Segal, R., Shimoni, A. & Wurzel, M. (1956) Biochem. J. 63,684-690. 12. Bergman, F., Wilson,. B. & Nachmansohn, D. (1950) Blochim. Biophys. Acta 6,217-224. 13. Rosenberry, T. L. (1975) Adv. nzymol., 43, 103-218. 14. Rieger, F. & Vigny, M. (1976) J. Neurochem. 27, 121-129. 15. Vigny, M., Koenig, J. & Rieger, F. (1976) J. Neurochem. 27, 1347-1353. 16. Gisiger, V., Vigny, M., Gautron, J. & Rieger, F. (1978) J. Neurochem. 30,501-516. 17. Hall, Z. (1973) J. Neurobiol. 4,343-361. 18. Rieger, F., Faivre-Bauman, A., Benda, P. & Vigny, M. (1976) J. Neurochem. 27,1059-1063. 19. Koenig, J. & Vigny, M. (1978) Nature 271,75-77. 20. Vigny, M., Bon, S., Massoulie, J. & Letefrier, F. (1978) ur. J. Biochem. 85,317-323. 21. llman, G. L., Courtney, K. D., Andres, V. & Featherstone, R. M. (1961) Biochem. Pharmacol. 7,88. 22. Adamson,. (1977) J. Neurochem. 28,605-615.