Effect of potassium propionate on free and bound acetylcholine in frog muscle

BrainResearch, 477 (1989) 109-117

109

Elsevier BRE 14155

Effect of potassium propionate on free and bound acetylcholine in frog muscle P.C. MoLmaar 1 and R . L . Polak 2 SDepartmentof Pharmacology, Uniwrsily Leiden, Leiden (TheNetherlands)and ZMedicalBiologicalLaboratory TNO, Ri]swi]k(TheNetherlands) (Accepted 28 June 1988)

Key words: Acetylcholine; Neuromuscular junction

Frog sartorius muscles were homogenized under various conditions which allowed, by means of mass spectrometry, the measurement of total ACh, and different ACh compartments in the tissue: 'bound', 'free-1' and 'free-2' ACh. Bound ACh presumably corresponded to the vesicular compartment, and the free-1 and free-2 fractions to the cytoplasmic compartments of ACh. Stimulaton of ACh release by Laa+ ions for 60 min caused a decrease of both bound and free-2 ACh, but at 20 min bound ACh was reduced much more than free-2 ACh. Stimulation of ACh release by isotonic potassium propionate (KPr) solution for only 5 min caused a decrease of bound ACh, in contrast to free-1 and free-2 ACh which were not significantly changed. When muscles after 5 min stimulation in KPr were allowed to recover in normal Ringer, f:ee-1 ACh did not change, but free-2 and bound ACh increased; after 180 min in Ringer bound ACh had recovered to control values. When ACh synthesis was prevented by hemicholinium-3 during recovery of the muscles in Ringer, bound ACh increased at the expense of free-2 ACh. In deuterium labeling experiments, in which the Ringer contained choline-d9, much more ACh-d9was formed in stimulated than in unstimulated muscles. It appeared that almost all newly formed ACh was ACh-d9, since no significant synthesis of unlabeled ACh (ACh-d0) took place. Yet again, the amount of bound ACh-d0significantlyincreased, apparently at the expense of preformed free-2 ACh-d0. The labeling ratio (ACh-dg/ACh-d9 + ACh-d0) of free-2 ACh was about twice that of bound ACh in muscles 60 min after 5 min KPr stimulation, but the labeling ratio of free-2 and bound ACh had become equal after 180 rain. The results strongly suggest that released ACh derived from the bound compartment, that the synthesis took place in the compartment recovered as the free-2 ACh, and that bound ACh was not primarily replenished by ACh synthesis, but by transport of ACh from the free ACh compartment.

INTRODUCTION Experiments with electrop!aque tissue from Torpedo marmorata show that brief nervous stimulation causes a reduction of 'free' A C h , a fraction thought to derive from the A C h dissolved in the cytoplasm of the nerve endings, whereas the ' b o u n d ' A C h derived from vesicular A C h , remains unaffected 2-4. These findings seem to contradict the theory of vesicular exocytosis of A C h during nervous transmission 1. In the electroplaque the level of bound A C h is reduced only after exhaustive stimulation 2. In the frog sartorius muscle exhaustive stimulation of neurotransmitter release, by nervous stimuli or by

incubation in 50 mM-KCI, similarly leads to a reduction of both free and bound A C h z4. However, when . frog sartorius muscle is incubated in isotonic potassium propionate (KPr) medium for a period as short as 5 rain, a procedure causing a rapid discharge of A C h quanta and a severe depletion of synaptic vesicles6, it is the bound A C h fraction which becomes preferentially depleted, whereas the free-2 A C h remains unchanged 19. It should be noted that in frog muscle two 'free' A C h fractions are distinguishedl4: one (free-1 A C h ) is hydrolyzed immediately by endogenous acetylcholinesterase (ACHE) during homogenisation, whereas the other part (free-2 A C h ) survives; the latter frac-

Correspondence: P.C. Molenaar, Department of Pharmacology, Sylvius Laboratories, P.O.B. 9503, 2300 RA Leiden, The Netherlands. 0006-8993/89/$03.50 ~) 1989 Elsevier Science Publishers B,V. (Biomedical Division)

110 tion is hydrolyzed only after excess AChE (from electric eel) has been added to the homogenate. In contrast, in the electroplaque only one free ACh fraction can be distinguished2; it is hydrolyzed instantaneously during homogenisation of the organ due to the vast amounts of AChE present in this preparation. In the present paper we extended our experiments on the effect of KPr on the bound and free-2 ACh fractions in frog muscle; among other things we studied the recovery of ACh in these fractions following stimulation and the distribution of newly formed, deuterium-labeled, ACh over free and bound ACh at different intervals following stimulation. In addition, we looked at changes of free-1 ACh, since it was conceivable that this fraction is influenced by stimulation in the same way as the free ACh in the Torpedo. Finally we studied the effect of La 3+ ions. Lanthanum was of interest because it causes a rapid, irreversible, depletion of synaptic vesicles associated with intense transmitter release 7,i2. MATERIALS AND METHODS

Preparation and incubation of muscle The experiments were made on frogs (Rana temporaria), from which sartorius muscles were dissected. Muscles were incubated at 20 °C in one or more of the following solutions (mM): Normal Ringer (NaCI, 116; KCI, 2; CaCI2 1.8; sodium phosphate, 2; pH 7.2); 'Ca2+-free Ringer' (CaCI2, 1.8, replaced by MgCI2, 2 and E G T A 1); 'KPr-Ringer' (potassium propionate, 115; calcium propionate, 1.8; HEPES buffer, 3; pH 7.4). In the experiments with LaCI3 the Ringer was buffered with 5 mM-Tris maleate, pH 7.4, because in phosphate buffer a precipitate is formed in the presence of La 3+ (see ref. 10). Homogenization of muscle Freshly dissected or incubated muscles were disrupted w~,th an UItra-Turrax homogenizer, set at 20,000 rev.min -l, either at 20 °C in 3 ml acetonitrile with 2.5% trichloroacetic acid (w/v) (CH3CN/TCA) for the determination of 'total ACh' or at 0 °C in 3 ml 'EGTA-Ringer' (1.8 mM-CaCl 2 replaced by 2 mMMgC!2 and 10 mM-EGTA). In part of the experiments the homogenate in EGTA-Ringer was deproteinated immediately (within 30 s after the beginning

of the homogenization) by the addition of TCA to a final concentration of 10% (w/v) for the determination of 'pseudo-total ACh'. In another part of the experiments the homogenate in EGTA-Ringer was divided into two equal portions one of which was deproteinated immediately for the determination of 'pseudo-total ACh', whereas the other portion was first incubated for 5 min at 0 °C after the addition of AChE from electric eel (1 #g/ml), and then deproteinated with TCA for the extraction of 'bound ACh' (bound ACh is protected against the hydrolytic action of AChE because it is occluded in subcellular structures, mainly synaptic vesiclesl4). Total ACh was often higher than pseudo-total ACh. The difference, 'free-1 ACh', was determined in contralateral muscles of the same animals (see Table I). Free-1 ACh probably represents ACh, hydrolyzed during homogenization by synaptic AChE 14. The present definition of free-1 ACh as the difference between the amounts of ACh in CH3CN/ TCA extracts and in immediately extracted homogenates in EGTA-Ringer differs from its previous deftnition 14 as the difference between the amounts of ACh extracted from homogenates in Ringer of muscles which had or had not been pretreated with an ir-

TABLE I

HomogenateA Chfractions offrog sartoriusmuscle Freshly dissected muscleswere homogenizedeither in CH3CN/ TCA (A), or in EGTA-Ringer (B). Alternatively, EGTARinger homogenates of muscles were divided into two portions, of which one was extracted with TCA immdiately (C), and the other after 5 rain incubation with 1/~g.ml-~ AChE from electric eel (D). Values are means _+S.E.M. of the mean with the number of musclesin parentheses. N.B. the fact that values of pseudoltotal ACh in B and C were at different levels was probably due to variations of ACh levels between different batches of frogs (cf. Materials and Methods).

Fraction Experiment 1 Total ACh Pseudo-total ACh Free-1 ACh Experiment 2 Pseudo-total Ach Bound ACh Free-2 ACh * Significant differences, A vs B paired t-tests).

A Ch (pmol) (A) 71 _+4.2 (19) (B) 55 4- 4.1 (19)* (A-B) 16 4- 3.1 (19) (C) 37 _+3.3 (18) (D) 22 + 2.9 (18)* (C-D) 15 4- 2.9 (18) and C vs D (/2 < 0.005~

111 reversible AChE inhibitor. However, the two definitions are probably equivalent, since no significant amounts of ACh are lost during disruption in CH3CN/TCA, and since the loss which occurs during homogenization in Ringer is prevented by inactivation of the cholinesterase t4. The difference between the amounts of ACh in pseudo-total and bound ACh, i.e. the amounts of ACh disappearing during 5 rain incubation of the homogenate in EGTA-Ringer containing added AChE from electric eel, is called 'free-2 ACh'. It probably mainly represents ACh dissolved in the nerve ending cytoplasm14. In summary ~he following ACh fractions will be distinguished: (1) total ACh (CH3CN/TCA extracts of intact muscle); (2) pseudo-total ACh (EGTARinger homogenete, extracted with TCA immediately after homogenization); (3) bound ACh (EOTA-Ringer homogenate to which excess AChE has been added and which is then incubated for 5 min at 0 °C before being extracted with TCA); (4) free-1 ACh [(1)-(2)]; (5) free-2 ACh [(2)-(3)]. It is noted that there is large variation in ACh content of muscles between different batches of frogs and even of muscles from different animals within one batch n. Therefore, it was necessary to use contralateral muscles from the same animal as test and control preparations in paired experiments. Since total and bound ACh could not be determined in the same muscle (the use of CH3CI'C/TCA precluded the determination of bound ACh), it was not possible to study the effect of different conditions simultaneously on these parameters. However, pseudo-total ACh and bound ACh could be determined in the same muscle which enabled the comparison at different times between contralateral pairs of muscles. Thus for practical reasons pseudo-total ACh was often used as an approximation of total ACh.

Estimation of A Ch After addition of deuterium-k .~eled internal standard (ACh-d16), the samples were purified for measurement of ACh and its deuterated derivatives on a gas chromatograph/mass spectrometer t4'21. Endogenous ACh, ACh-d9 (containing the precursor choline-dg), and ACh-dl6 (internal standard) were monitored simultaneously at m/e 58, 64 and 66, respectively.

Materials AChE was from Boehringer Mannheim, hemicholinium-3 (HC-3) from Aldrich, recrystallized before use from ethanol, and ACh-dt6 from Merck, Sharp and Dohme. Choline-d9 and ACh-d9 were synthesized earlier2°. DEPP was kindly provided by Dr H.P. Benschop, Prins Maurits Laboratory, Institute for Chemical and Technological Research TNO, Rijswijk, The Netherlands. RESULTS

A Ch in different fractions prepared from frog sartorius muscle As presented in Table I, extracts from freshly dissected niuscles which were disrupted in CH3CNfrCA contained 71 pmol ACh ('total ACh'). Homogenates of the contralateral muscles in EGTA-Ringer which were extracted immediately with TCA contained 55 pmol ACh ('pseudo-total ACh'). Apparently, 71-55 = I6 pmol ACh ('free-1 ACh') was hydrolyzed during homogenization in EGTA-Ringer. In another series of muscles, from a different batch of frogs, 37 pmol ACh pseudo-total ACh was found whereas only 22 pmol (the 'bound ACh') was measured after treatment of the homogenates from contralateral muscles with added AChE from electric ed. Apparently, 37-22 = 15 pmol ACh ('free-2 ACh') was hydrolysed by the added enzyme. The present values of bound and free-1 ACh correspond well with those earlier reported 14, but that of free-2 ACh is lower than found earlier (about 30 pmo114). Depletion of ACh by Lc: + ions As shown in Fig. I, pseudo-total and bound ACh were strongly depleted under the influence of La3+ ions. In these experiments HC-3 was added to the medium to prevent ACh synthesis which otherwise tends to counteract the ACh depletion of the muscle in the presence of lanthanum 12. The depletion occurred after the delay which lasted at least 10 min. Pseudo-total ACh was decreased from 37 to 19 pmol at 20 rain and to 11 pmol at 60 min after application of LaCI~, and bound ACh from 22 to 8 and 6 pmol, respectively. (The latter value after 60 rain is similar to the previous value of 8 pmol, for bound ACh after 300 rain lanthanuml4.) Free-2 ACh, i.e. the difference between 'pseudo-

112

Depletion of A Ch by KPr In other experiments, we incubated muscles in KPr-Ringer for 5 min. The results, presented in Fig. 2, show that KPr caused a substantial, statistically highly significant, reduction of b o u n d A C h within 5 min. Free-2 A C h was not reduced at all, in agreement with previous findingstg. Free-1 A C h showed a tendency of reduction, but this change was not statis-

(6)

v

.¢4

"d

tically significant.

Recovery of A Ch after KPr-stimulation o

(6)

When sartorius muscles, after 5 min incubation in KPr-Ringer, were incubated again in normal Ringer,

(6)

Free-2

30

20

(7) (7)

(6)

t'

I

r-t o

Free-I

QI

20

(19)

(6) |

o

,

!

20

I

40

!

60

t

Time (min)

Fig. 1. Effect of LaCI3 on ACh in frog sartorius muscle. Pseudo-total ACh (triangles) and bound ACh (circles) were determined at various times after addition of 2 mM LaCI3to the medium. The incubation medium contained 10 ~M HC-3 which was introduced 15 min before the addition of LaCI3. Means -+ S.E.M. of mean with number of musclesas ir~ :ated.., Significant differences from corresponding contt , values at t = 0 (t'2 < 0.005, paired t-test).

Bound (7)

t (19) IQ

(7)

f total' and ' b o u n d ' ACh, showed a tendency of reducrio_-, at 20 min (a 5 pmol reduction), but this was not statistically significant. Free-2 was, however, reduced from 16 + 1.8 pmol at t = 0 min to 6 + 1.3 pmol at 60 min in contralateral muscles from 6 animals (P2 < 0.005, paired t-test). In these experiments b o u n d and free-2 ACh were depleted at 60 min to values approaching the levels of b o u n d and free-2 ACh in 15 days' denervated muscles (ca 3 and 5 pmol, respectivelyl9).

0

5

0

5

0 Time (min) Fig. 2. Effect of KPr on ACh in frog sartorius muscle. Contralateral pairs of muscles were fractionated or extracted before (0) and after 5 min incubation in KPr medium (5). The incubations were in the presence of 10 ~M HC-3, which was introduced 30 min before the homogenization. The data of free-1 ACh at t = 0 are the same as those in Table I. Means ± goE.M. with number of muscles as indicated. *, Significantlydifferent from value at t = 0 (P2 < 0.005, paired t-test).

113 they did not twitch for a while in response to either indirect or direct stimulation of the muscle fibres. As illustrated in Table It, both the twitch response to direct stimulation and that to indirect stimulation returned after about 20 min incubation in Ringer, even if the synthesis of A C h was blocked by 10/~M HC-3. Apparently, the transmitter left in the nerve terminals after 5 rain incubation in KPr-Ringer was sufficient to support neuromuscular transmission in response to single impulses. In another experiment sartorius muscles which had been incubated for 5 rain in KPr-Ringer, were either incubated for 2.5 min in a Ca2+-free Ringer (containing 2 mM-Mg 2+ and 1 m M - E G T A ) and then extracted or fractionated for the determination of ACh, or they were incubated for 5 rain in a Ca2+-free Ringer and subsequently for 60 min in normal Ringer before being extracted or fractionated. The 2.5- or 5rain 'washing' period in Ca2+-free Ringer served to stop the evoked release of transmitter as rapidly and completely as possible. In this experiment total A C h increased from 40 + 3.3 pmol at t - 7.5 min to 54 + 5.6 pmol at t = 70 rain (n = 6; P2 = 0.06, Student's ttest). Free-1 ACh was unchanged at t = 70 rain but free-2 ACh increased from 12 to 17 pmol and bound ACh from 10 to 14 pmol in the course of 62.5 rain (Table lII, experiment 1). The latter differences, which were statistically significant, show that ACh synthesis took place following stimulation with KPr. In the experiment 2 of Table Ill we measured free2 and bound ACh in muscles which were incubated for 5 min in KPr, for a subsequent 5-rain period in

TABLE II Return of twitch response after paralysis of frog sartorius muscles by potassium propionate

Muscles were incubated in KPr medium for 5 rain and subsequently in a normal, low K+, Ringer in the absence or presc,,ce of l0 gM HC-3. The preparations were stimulated once every 30 s indirectly, via the nerve, or directly at the muscle. The data refer to the time when the first twitch re-appeared as judged by visual inspection (time after change of KPr to low K + Ringer medium). Values are means + S.E,M. of mean, 8 muscles. Incubation medigm

Time of first twitch (rain) after stimulation via Muscle

Nerve

Ringer Ringer + 10gM HC-3

20 + 1.6 19 -+ 1.4

22 + 1.3 20 -+ 1.5

TABLE III Recovery of A Ch in homogenate fractions of frog sartorius muscles after KPr-induced depletion of A Ch

In experiment 1, muscles were first incubated for 5 rain in KPr medium. Subsequently, they were either washed for 2.5 rain in a Ca2+-free Ringer and then fractionated, or they were washed for 5 rain in Ca2+-free Ringer and then incubated for 60 rain in normal Ringer before being fractionated. The different ACh fractions measured at t = 70 rain were compared with the corresponding fraction at t = 7.5 rain in contralateral muscles. In experiment 2, muscles stimulated for 5 rain in KPr, were incubated for 180 rain in normal Ringer ('stimulated'); the contralateral muscles were incubated for 190 rain in normal Ringer, without exposure to KPr ('unstimulated'). Means 4- S.E.M. of mean with number of muscles as indicated. Acerylcholine (pmol)

Experiment 1 7.5 min, stimulated 70 min, stimulated Experiment 2 190 min, unstimulated 190 min, stimulated

Free-1

Free-2

Bound

8_+1.7(6)

12_+2.1(16)

10_+2.1(16)

8_+2.2(6)

17_+1.7(16)** 14_+2.2(16)*

n.d.

20_+4.1 (6)

17_+1.6(6)

n.d.

25_+5.0(6)

15_+2.1(6)

*P., < 0.05; **/2 < 0.01 (paired t-tests); n.d., not determined.

Ca2+-free Ringer and thereafter for 3 h in Ringer. The contralateral control muscles were incubated for 190 rain in Ringer, without being exposed to KPr. No significa.nt differences between muscles were found indicating that the bound ACh, reduced by KPr (see Fig. 2), recovered, because of increased ACh synthesis, during a subsequent 3-h period in Ringer to values close to those of the unstimulated controls. The question arose whether bound ACh, depleted by KPr, could subsequently increase at the expense of free ACh if ACh synthesis was blocked by HC-3. As illustrated in Fig. 3 this was indeed the case. The bound ACh showed a small, but statistically not significant, increase after 70 rain and a considerable, significant, increase after 190 min. On the other hand, pseudo-total ACh was not increased either after 70 or after 190 min indicating (1) that the synthesis of ACh was effectively blocked by HC-3 (as found earlier t2) and (2) that the free-2 ACh became thoroughly depleted in favour of bound ACh.

114 P s e u d o - t o r n ~. 30

Pseudo-total

(5)

O v

Bound

O~ ,,:4

(6)

20.

U i,,,4

Bound (11)

c,I

C6) 10-

I 7.5

70

7.5

70

7.5 190

'

7.5 190 T:trae (mtv)

Fig. 3. Influence of HC-3 on the recovery of ACh of frog sartorius muscles after Kpr-induced depletion of ACh. Muscles were incubated in KPr and Ringer as described under Table Ill, but 10/~M HC-3, given 15 rain before the addition of KPr, was present throughout the incubation in all experiments. Presented are pseudo-total ACh and bound ACh. Means _ S.E.M. with number of muscles as indicated. **, Significant increase (/)2 < 0.005, paired t-test).

Incorporation of deuterium-labeled choline into A Ch The recovery after 5 rain KPr stimulation was further analyzed by stud3~ng the incorporation of deuterium-labeled choline into ACh. In the experiment of Table IV, muscles which had been incubated for 5 min in KPr, were incubated for 2.5 min in cae+-free medium, and then fractionated for the de*.ermination

of bound and free-2 ACh, while the contralateral muscles were incubated for 5 min in KPr, for 5 min in Ca2+-free medium and then for 60 rain in normal Ringer before being fractionated. In the latter, choline-d9 (50/~M) was present in the Ca2+-free medium and in the Rirlger during the first 50 rain of the 60-rain recovery periad. In the experiment of Table IV 10.8

TABLE IV Incorporation o f choline-d 9into A Ch after KPr.induced depletion o f A Ch

Experiment 1: similar experiment as in Table HI, but the incubation medium contained 50/~M-choline-d9 during the 5-min washing period in Ca2+-~ee Ringer and during the first 50 rain of the 60-min period of incubation in normal Ringer. The muscles homogenized at t = 7.5 rain were not exposed to ch°line'dg".2+Experiment. 2: similar experiment as in Table III, but the medium contmned' 50/~M- cholined 9 during the 5-min washing period in Ca -free Ringer and during the first 170 min of the subsequent 180-rain period of incubation in normal Ringer both in KPr-pretreated ('stimulated') and control ('unstimulated') muscles. Presented are ACh-d0 and ACh-d 9, and the corresponding labeling ratios (L.R., ACh-dg/(ACh-d9 + ACh-d0). Means + S.E.M., ACh values in pmol. Free-2

Experiment 1 7.5 min, stimulated 70 rain, stimulateda Experiment 2 190 rain, unstimulated 190 min, stimulated

Bound

N

ACh-d o

ACh-d 9

L.R.

7 7

10 _+0.9 8 + 1.8

0 8 + 1.0

0 0.53 + 0.066b

6 6

18 + 3.8 16 + 3.9

2 + 1.1 9 + 1.3"

0.11 _+0.039 0.36 +_0.037**

ACh-d o

5 + 1.2 9 + 1.8"* 17 + 2.3 9 + 1.4

ACh-d o

L.R.

0 2 + 0.26

0 0.26 + 0.046 b

2 + 0.27 6 + 1.7'*

0.09 + 0.012 0.39 + 0.026**

*/)2 < 0.05; **/)2 < 0.01, paired t-tests of t = 70 vs t = 7.5 min, or 'stimulated' vs 'unstimulated'. a Data from Molena~r and Polakt9; b L.R. of free-2 ACh statistically significantly higher than L.R. of bound ACh (P2 < 0.05).

115 +_ 1.1 pmol ACh-d9 (n = 7) was synthesized and contained in the pseudo-total ACh extracts, whereas only 2.6 + 1.3 pmol ACh-d0 was formed, an amount not statistically significantly different from zero. Apparently, the synthesis of ACh-d0 was suppressed by the relatively high concentration of choline-d9 in the incubation medium. Yet there was a net increase of 4.2 pmol ACh-d 0 in the bound ACh fraction (P2 < 0.01). This observation is in agreement with the observation of Fig. 3, in which bound ACh also increased when the synthesis of ACh-d 0 was inhibited, this time by HC-3. However, it should be noted that, in contrast to the effect of HC-3, the total synthesis of ACh (ACh-d 0 + ACh-dg) was not affected by the presence of choline-d9: in the 60 min following stimulation the sum of ACh-d0 and ACh-d 9 was increased from 9.8 _+ 0.9 to 16.6 + 1.8 pmol (n = 7, P2 < 0.01), a result similar to that presented in Table III. The main part of the newly formed ACh-d9 was found in the free-2 ACh. Its labeling ratio (0.53) was twice that in the bound ACh (0.26), indicating that newly formed ACh was generated in the free-2 ACh compartment. In other experiments (not illustrated) the experiment of Table IV was repeated after complete AChE inLibition by DEPP. In this case the true total ACh was measured (see Materials and Methods) and the difference between total and bound ACh represented the sum of free-1 and free-2 ACh. The result was similar to that of the experiment in Table IV: the labeling ratios were 0.43 +_ 0.022 and 0.24 +_0.025, n = 7, for bound and free-1 + free-2 ACh, respectively (P2 < 0.01, paired t-test). In experiment 2 of Table IV, muscles stimulated for 5 min by KPr and unstimulated contralateral control muscles were incubated for 175 rain in the presence of choline-dg. Table IV shows that KPr greatly stimulated the subsequent incorporation of cholined9 into the ACb. This shows again that the rate of ACh synthesis which followed the stimulation was much higher than the 'resting' synthesis. The labeling ratio of bound ACh was about the same as that of free-2 ACh (ca 0.4), demonstrating that, in contrast to the situation after 60 rain (see experiment 1 in Table IV), after 180 min equilibration was reached. Apparently, the transfer of ACh from the compartment of free-2 ACh to that of bound ACh took place at a relatively slow rate, at least under resting conditions.

DISCUSSION In the present work KPr was used instead of KC! in order to bring about transmitter release at a very high rate 16. KPr probably depoladses the nerve terminal membrane to a greater extent tba~ KC! because the cell membrane is practically impermeable to the Pranion; Pr- ions per se have no effect on the process of ACh release ms. The finding that bound ACh was the only fraction which was significantly reduced after 5 min incubation of frog muscle in KPr, appears to be in agreement with the theory of vesicular exocytosis of ACh. Previously it was fot,nd that incubation of frog muscle for 60 rain in a medium containing 50 mM KCI, caused a decrease not only of bound ACh, but also of free-2 ACh 14. In the present experiments La 3+ ions also reduced both bound and freeo2 ACh. However, the level of bound ACh reached its minimum after 20 rain, and that of free-2 ACh only after 60 min, just as one would expect if vesicular ACh is released and depleted first, and then supplemented from the cytoplasmic ACh eompariment. The 'recovery' experiments discussed below indicate that such is indeed the course of events. During recover, of frog muscle in normal Ringer following stimulation in KPr, synthesis of ACh took place at an increased rate. In the first 60 rain of recovery, free-2 ACh increased by 5 pmol (Table III) and by 6.7 pmol (Table IV); ff the medium contained choiine-dg, 8.8 pmol ACh-d9 (Table IV) was found in the fraction of free-2 ACh. Consequently, the increase of free-2 ACh must have been due wholly to newly formed ACh. The bound ACh also increased, by 5 pmol (Table III) and by 6.6 pmol (Table IV). But in the experiments of Table IV this increase consisted of only Z4 pmol labeled ACh, whereas the remainder of the ACh increase, 4.2 pmol, was unlabeled. Apparently, the increase of bound ACh was not directly dependent on ACh synthesis. Instead, there must have been a net transport of (predominantly old) ACh from the free to the bound ACh compartment. The following results further support the view that bound ACh, depleted by KPr, was replenished at the expense of free-2 ACh. When the synthesis of ACh was inhibited by HC-3, the pseudo-total ACI' did not change during the recovery period, but the bound ACh nevertheless increased considerably.

116 The question arises as to whether there is transport of ACb in the reverse direction, i.e. from the bound towards the free-2 ACh compartment, which together with the uptake transport would lead to mixing of ACh between the two compartments. However, the labeling ratio of ACh in the bound fraction at t = 80 rain was different by a factor of two from that of ACh in the free-2 ACh, indicating that such mixing, if any, took place at a relatively low velocity. The labeling of ACh by choline-d9 was due to synthesis and not to (enzymatic or non-enzymatic) exchange of label between choline-d9 in the medium and ACh-d0 in the tissue, because 3 h incubation of unstimulated muscles in a medium containing choline-d9 resulted in the formation of as little as 4 pmol ACh-dg. This amount is readily accounted for by resting turnover of ACh, which takes place at about 2 pmol.h -! (see ref. 13). The present conclusion that ACh is synthesized in the free-ACh compartment is in agreement with observations in the electroplaque2 and in line with the findings that both in brain and in the electroplaque choline acetyltransferase is localized in the cytoplasm of the cholinergic nerve ending5"9. On the other hand our finding that bound ACh is preferentially depleted upon stimulation disagrees with other findings in the electroplaque by Dunant et al. 2"3and notably by Dunant et a l : who found a 20% decrease of total ACh after delivery of only 10 electrical field stimuli; in these experiments they found that the bound ACh remained unaltered and therefore they concluded that the free ACh decreased under the influence of stimulation. Obuiously, it would be interesting if the ACh content of frog sartorius muscle could also be affected by such a short train of nervous impulses. However, this is most unlikely. Assuming that a nervous impulse releases 300 quanta per endplate s, each quantum consisting of 13,000 molecules ACh 15, a train of 10 stimuli would cause the release of 3000 quanta per endplate, corresponding to 0.06 pmol in the whole muscle (which contains 1000 endplates). This is, of course, far too little to cat~se a measurable decrease of the ACh content of the muscle. Indeed, in experiments on R.esculenta (Molenaar, unpublished experiments) it was found that 20 nervous stimuli applied in 0.2 s did not cause a measurable change of total ACh (CHaCN extracts). In con-

tralateral pairs of muscles the ACh content was 73 + 3.4 pmol after 20 stimuli and 67 __. 4.8 pmol in unstimulated controls (n = 6). The present results strongly suggest that bound ACh contains the (readily) releasable ACh of the tissue. Therefore, it is important to consider the evideace that the homogenate fractions were representative for specific ACh compartments in the motor nerve ending. The evidence that bound ACh derives from the ACh in synaptic vesicles comes from work using two important tools, lanthanum and chloride ions. La 3+ ions cause a transient release of ACh, followed by an irreversible depletion of synaptic vesicles in frog muscle7"l°a2; La3+ initially causes a reduction of the ACh content of the muscle, but the reduction is transient and the ACh content subsequently rises to a high level if the incubation is continued for 5 h. Thus the ACh present after 5 h is mainly located in the cytoplasm of the nerve endings 12,~5. Bound ACh is rapidly and irreversibly reduced by La 3+ to a level as low as found after denervation, whereas free-1 and free-2 ACh reach relatively high values after 5 h t4. Incubation for 30 rain in KPr medium, followed by 3 h incubation in low K + medium, in which chloride has been replaced by propionate leads to results similar to those obtained with La3+: a persistent depletion of vesicles and bound ACh, and an increase of free ACh tT. From these findings we have concluded that cytoplasmic ACh does not contribute significantly to bound ACh, and that free-1 and free2 ACh together contain all ACh dissolved in the cytoplasm. Although these fractions are contaminated by some ACh of non-neural origin 11, the changes in bound and free ACh probably exclusively reflect changes in the ACh content of the synaptic vesicles and the motor nerve ending cytoplasm, respectively, because non-neural ACh has beep ,~ound to be rather inert, not being influenced by high K + and La 3+ in the incubation medium 12A3.

ACKNOWLEDGEMENTS We thank Mrs J.W.M. Tas, Mrs G.Th.H. van Kempen and Mr A.L. van der Laaken for excellent technical assistance. Financial support by MEDIGON/ZWO (formerly FUNGO/ZWO) is gratefully acknowledged.

117 REFERENCES 1 Del Castillo, J. and Katz, B., La base 'quantale' de la transmission neuromuseulaire. In Microphysiologie Compar~e des Elements Excimbles. Colloques Int. du Centre de la Recherche Scientifique. No. 67, Gif-sur-Yvette, Paris, 1957, pp. 245-258. 2 Dunant, Y., Gautron, J., Isra61, M., Lesbats, B. and Manaranche, R., Les compartiments d'ac~tylcholine de rorgane ~lectrique de la Torpille et leurs modifications par la stimulation, J. Neurochem., 19 (1972) 1987-2002. 3 Dunant, Y., Israel, M., Lesbats, B. and Manaranche, R., Oscillations of acetylcholine during nerve activity in the Torpedo electric organ, Brain Research, 125 (1977) 123-140. 4 Dunant, Y., Jones, G.J. and Loctin, F., Acetylcholine measured at short time intervals during transmission of nerve impulses in the electric organ of Torpedo, J. Physiol. (Lond.), 325 (1982) 441-460. 5 Fonnum, F., Is choline acetyltrar.:ferase present in synaptic vesicles.'?,Biochem. Pharmacol., 15 (1966) 1641-1643. 6 Gennaro, J.F., Nastuk, W.L. and Rutherford, D.T., Reversible depletion of synaptic vesicles induced by application of high e×ternal potassium to the frog neuromuscular junction, J. Physiol. (Lond.), 280 (1978)237-247. 7 Heuser, J.E. and Miledi, R., The effect of lanthanum ions on function and structure of frog neuromuscular junctions, Proc. R. Soc. Lond. Ser. B, 179 (1971) 247-260. 8 Katz, B. and Miledi, R., Estimates of quantal content during 'chemical potentiation' of transmitter release, Proc. R. Soc. Lond. Ser. B, 205 (1979) 369-378. 9 Marchbanks, R.M. and Isra61, M., Aspects of acetylcholine metabolism in the electric organ of Torpedo marmorata, J. Neurochem., 18 (1971) 439-448. 10 Miledi, R., Molenaar, P.C. and Polak, R.L., Acetylcholine compartments in frog muscle. In D.J. Jenden (Ed.), Chohnergic Mechanisms and Psychopharmacology, Plenum, New York, 1977, pp. 377-386. 11 Miledi, R., Molenaar, P.C. and Polak, R.L., An analysis of

acetylcholine in frog muscle by mass fragmentography, Pro& R. Soc. Lond. Ser. B, 197 (1977) 285-297. 12 Miledi, R., Molenaar, P.C. and Polak, R.L., The effect of lanthanum ions on acetylchol!ne in frog muscle, J. Physiol. (Lond.), 309 (1980) 199-214. 13 Miledi, R., Molenaar, P.C. and Polak, R.L., Early effects of denervation on acetylcholineand choline acetyltransferase in skeletal muscle. In G. Pepeu and H. Ladinsky (Eds.), Cholinergic Mechanisms, Plenum, New York, 1981, pp. 205-214. 14 Miledi, R., Molenaar, P.C. and Po!ak, R.L., Free and bound acetylcholine in frog muscle, J. Physiol. (Lond.), 333 (1982) 189-199. 15 Miledi,R., Molenaar, P.C. and Polak, R.L., Electrophysiological and chemical determination of acetylcbolinerelease at the frog neuromuscular junction, J. Physiol. (Lond.),334 (1983)245-254. 16 Molenaar, P.C. and Oen, B.S., Analysisof quantalacetylcholine noise at end-plates of frog muscle during rapid transmitter secretion, J. Physiol. (Lond.), 400 (1988) 335-348. 17 Molenaar, P.C., Oen, B.S. and Polak, R.L., Effect of CIions on giant miniature end-plate potentials in frog muscle, J. Physiol. (Lond.), 366 (1985) 88P. 18 Molenaar, P.C., Oen, B.S. and Polak, R.L., The effect of chloride ions on giant miniature end-plate potentials at the frog neuromuscular junction, J. Physi.~l. (Lond.), 383 (1987) 143-152. 19 Molenaar, P.C. and Polak, R.L., Potassium proprionate causes preferential loss of 'bound' acetylcholine in frog muscle, Neurosci. Lett., 43 (1983) 209-213. 20 Polak, R.L. and Molenaar, P.C., Pitfalls in the determination of acetylcholine from brain by pyrolysis-gas chromatography/mass spectrometry, J. Neuroehem., 23 (1974) 1295-1297. 21 Polak, R.L. and Molcnaar, P.C., A method for the determination of acctylcholine by slow pyrolysis combined with mass fragmcutography on a packed capillary column, J. Neurochem., 32 (1979) 407-412.