Acetylcholine Chloride

Early Postdenervation Depolarization Is Controlled by Acetylcholine and Glutamate via Nitric Oxide Regulation of the Chloride Transporter*

Frantisˇek Vyskocˇil1,2

Resting non-quantal acetylcholine (ACh) and probably glutamate (Glu) release from nerve end- ings activates M1- and NMDA receptor-mediated Ca2+ entry into the sarcoplasm with following activation of NOS and production of NO. This is a trophic message from motoneurons, which keeps the Cl— transport inactive in the innervated sarcolemma. After denervation, the secretion of ACh and Glu at the neuromuscular junction is eliminated within 3–4 h and the production of NO in the sarcoplasm is lowered. As a result, the Cl— influx is probably activated by dephosphoryla- tion of the Cl— transporter with subsequent elevation of intracellular Cl— concentration. The equi- librium Cl— potential becomes more positive and the muscle membrane becomes depolarized.

KEY WORDS: Acetylcholine; carbachol; postdenervation depolarization; glutamate.

INTRODUCTION

Motor nerves transfer signals to muscles, which in- duce contractions and maintain muscle fiber structure and function. Depolarization of the resting membrane potential (RMP) (1–3) is the first change after nerve sec- tion, which affects muscle excitability and contraction. It has been shown that denervated muscle fibers of the rat diaphragm kept in a tissue culture medium become depolarized by about 8–10 mV (10%–12% of the control RMP) within 3 to 4 hours after denervation (4–12). An inward-directed, furosemide-sensitive chloride transport is activated at this time and is believed to be the main cause not only of this early depolarization (5,13) but also of other abnormalities, such as loss of the ability of mus- cle fibers to control their volume in hypertonic solutions (14), and the appearance of extrasynaptic ACh receptors (15). The mechanisms by which the nerve influences Cl— transport and regulates RMP, or the intracellular mech- anisms involved in this control, remained unexplained until a series of our papers initiated by Alec R. Urazaev
(16) on NO donors and inhibitors was recently pub- lished. The NO role in skeletal muscles was discussed in recent reviews of Stamler and Meissner (17) and Groz- danovic (18). Because these authors mentioned only some aspects of our recent work about the contribution of NO to the membrane resting potential (RP), the pres- ent overview is aimed to summarize our ideas as to how the NO cascade is involved in early postdenervation depolarization (EPD).
We believed from very beginning that the trig- ger for RMP maintenance and EPD should be some substance released from the nerve ending. Our attention was naturally drawn to acetylcholine (ACh), which has been considered for some time as a potential presynap- tic regulator (4, for review see 19 and 20).

DEVELOPMENT OF EARLY POSTEDENERVATION DEPOLARIZATION

In our studies, 3–4-mm–wide strips of parallel in- tact muscle fibers of rat and mouse diaphragms were used, either with a long phrenic nerve stump (10–20 mm) or with no extramuscular nerve stump (considered as a short nerve stump preparation) (11). The muscle strips were pinned to a transparent dish filled with a glutamic acid-free medium. Glass microelectrodes were used for rapid recording of the RMP of 25–30 superficial mus- cle fibers of each strip. The average RMP of the mus- cles with a long nerve stump was —74.5 mV 10–15 min after dissection. This value did not change substantially during the following 3 h, when it was —74.2 mV. In the muscle strips with the short stump, the average RMP measured within 10–15 min after dissection was —74.0 and after 3 h the RMP depolarized by about 8 mV to
—66.5 mV, which was the standard value of EPD in in vitro experiments. Thus EPD developed only when the long nerve stump was not present. It was shown earlier (21–23) that the so-called nonquantal ACh release

(NQR) disappears at exactly the same time, 3–4 h after nerve section.

Nonquantal Release
It is necessary to stress here that there exists the so-called nonquantal release (NQR) of ACh in addi- tion to the well-established quantal release from nerve terminals (Fig. 1, left upper part). In electrophysiolog- ical experiments, the application of the ACh receptor– blocking agent tubocurarine (TC) to endplates previ- ously treated with an anticholinesterase causes a small hyperpolarization due to block of the depolarization produced by nonquantally released ACh (the H-effect) (24,25). We have proposed that the NQR occurs via ACh transporters, which are normally present in synaptic vesicles and become incorporated into the nerve membrane during the exocytotic release of ACh quanta (26). There are indications that, at least in ro- dents, most of ACh is released from nerve terminals nonquantally (27,28).
Several aspects of the physiological role of NQR have been considered in context with the develop- ment, maintenance, and modulation of synaptic nerve-muscle interactions (for literature see 29). Here we propose that NQR of ACh (and possibly also the release of glutamate) controls the RMP by Ca2+ entry and NO production (Fig. 1 left part). The absence of NQR after short stump nerve section results in EPD due to the NO-dependent activation of furosemide-sensitive Cl— transporter (30). First, we will discuss the ACh ef- fect (or its nonhydrolyzable derivative carbamyl- choline, CB) on EPD, then the effect of glutamate and Cl— transporter.

Acetylcholine and Carbachol Reduce EPD
As already stated above, muscle fibers became de- polarized by 8–10 mV within 3 hours after denervation. However, in the presence of 5 × 10—8 M CB or 5 × 10—8 M ACh, EPD was substantially reduced by 80%. Both drugs were used in concentrations to mimic the effect of nonquantal release of ACh (31). However, the experiments of Dlouhá et al. (32) indicated that CB and ACh might activate electrogenic membrane Na+-K+- ATPase in muscle fibers that are then hyperpolarized. The possibility therefore arises that the smaller EPD depolarization of muscle fibers is caused by a direct hyperpolarizing effect of CB (or ACh) as a result of the potentiation of electrogenic Na+-K+-ATPase. To inhibit this enzyme, 1 × 10—4 M ouabain was applied. On its own, ouabain enhanced the EPD (which is usu- ally 8 mV) and RMP decreased to —57 mV. This is a consequence of the complete elimination of electro- genic pump activity (33,34) and also of passive redis- tribution of K+ and Na+ inside the muscle fibers when the pump is blocked (2). When CB was applied to- gether with ouabain, the RMP drop was reduced close to the effect of CB on denervated muscle fibers with an intact Na+-K+ pump. In other words, the inhibition of Na+-K+-ATPase by ouabain did not change the ability of CB to reduce the development of EPD.

Role of Ca21 in EPD
Inhibitors that block Ca2+ entry virtually inhibited the effect of ACh and CB (10): Mg2+ in concentrations which block Ca2+ channels (5 × 10—3 M) inhibited the effect of both cholinergic agonists on the RMP, which became depolarized after 3 h to —64 mV when Mg2+ was added, in spite of the presence of CB. On its own, Mg2+ did not affect the postdenervation drop of RMP. In addition to Mg2+, other inhibitors of Ca2+ channels such as diltiazem, verapamil, niphedipine, and Cd2+ also at- tenuated the CB-induced prevention of the decrease of RMP after denervation. Thus the protective effect of CB on the RMP of denervated muscles is achieved through Ca2+ influx into muscle fibers. Whether these effects of CB and ACh are due to the activation of the postsynaptic nicotinic ACh receptor, which is also permeable for calcium, was tested by applying tubocurarine (TC) to- gether with CB. TC did not prevent the protective action of CB at all. It is therefore unlikely that these effects on RMP are mediated through ACh nicotinic receptors either pre- or postsynaptically. Instead, as it will be dis- cussed later, muscarinic receptors are involved.
In many cell types, the entry of Ca2+ into cyto- plasm activates NO-synthase(s), which releases NO molecules from L-arginine (35). NO can activate solu- ble guanylyl cyclase, which in turn catalyzes produc- tion of cGMP. The cGMP activates specific protein kinases, resulting in final protein phosphorylation. To test whether this cascade might be involved in CB ac- tion on postdenervation depolarization, L-nitroarginine methylester (NAME), an inhibitor of NO-synthase, was added to the muscle bath together with CB. NAME completely eliminated the CB protection of EPD (10). The specificity of its effect was tested by adding the NO-synthase substrate, L-arginine (which competes with NAME for the enzyme), together with NAME and CB. Indeed, L-arginine can partially reverse the inhibi- tion induced by NAME. This indicates that CB may maintain the RMP by enhancing Ca2+-dependent NO synthesis. NAME by itself did not influence the EPD.
It is known that NO molecules may pass through cell membranes and thus transmit information between cells. It cannot be excluded that the targets for CB- induced production of NO are also muscle or nerve fibers other than those in which NO is produced. The effect of NO released from cells might be tested by re- duced hemoglobin, which is impermeable and binds extracellular NO molecules, neutralizing their action. The EPD in the presence of CB and hemoglobin did not differ from the controls; that is, this protein elimi- nated the effect of CB. By itself, hemoglobin was without any effect. This indicates that either NO in- creases RMP by modulating ion channels from the out- side, which is highly improbable, or NO enters nerve endings and evokes the release of another modulating factor, which might well be ACh or glutamate. In fact, we recently found that NO reduces the NQR (36,37).

Carbachol and Acetylcholine Delay the EPD through M1-Cholinergic Receptors
While the nicotinic antagonists tubocurarine and α- bungarotoxin had no effect on the action of CB (9,10), we reported that it is the muscarinic M1 subtype of cholinergic receptor that apparently mediates the protec- tive effects of cholinergic drugs on the denervated mus- cle membrane (11): oxotremorine (Oxo), a highly specific muscarinic agonist that does not have a nicotinic effect, diminished the EPD in a similar way as CB. A comparison of the actions of Oxo and CB suggests that Oxo was slightly but significantly more potent in hyper- polarizing the muscle membrane than CB (11). Atropine (Atr) blocked the protective effect of CB in a concentra- tion-dependent manner. This inhibition was already sig- nificant in the presence of Atr 5 × 10—9 M. Estimated Ki of the Atr competition for CB (or Oxo) was 1 × 10—7 M. Which type of muscarinic receptor is involved? There are at least five subtypes of muscarinic receptors. The M1, M3, and M5 subtypes can be classified as a group of re- ceptors whose activation stimulates the phosphoinositide pathway, while M2 and M4 receptors initiate the adenylyl cyclase cascade (38). The expression of M1 muscarinic receptors coupled to phospholipase C and internal cal- cium stores has been convincingly demonstrated in cul- tured rat skeletal muscle fibers (39). We studied the influence of muscarinic and nicotinic stimulation on both phosphoinositide metabolism and intracellular calcium levels and found that the muscarinic effect was mimicked by Oxo and prevented by pirenzepine. Therefore the ex- pression of M1 muscarinic receptors coupled to phospho- lipase C and to internal calcium stores in cultured skeletal muscles was proposed. In our study (11) pirenzepine (Pir), a specific antagonist of M1 receptors, blocked the effect of Oxo and CB on EPD. The apparent inhibition constant Ki for Pir was quite low, 1 × 10—7 M in the pres- ence of 5 × 10—8 M CB. Several other drugs affecting M2 and M3 receptors had no significant effect on the CB- induced decrease of EPD (11). The pharmacological ev- idence therefore suggests that Oxo, CB, and/or ACh protect the muscle membrane from EPD through the M1 subtype of muscarinic cholinergic receptors (as sche- matically indicated in Fig. 1, left). It remains to check whether these receptors are coupled either with Ca2+ channels directly or via the NO-cGMP cascade, as has been suggested indirectly by the present data.

Long-Term Cultured Muscles
There is evidence that muscarinic receptors are dis- tributed not only in the skeletal muscle membrane (39) but also in the motor nerve endings (40–43), where they may regulate both quantal and nonquantal secretion of ACh. These presynaptic receptors might be responsible for the trophic effects of ACh and could therefore be of physiological significance in maintaining the RMP in muscle fibers by modulating nonquantal secretion of ACh at the neuromuscular junction, as has already been suggested (44). Their possible participation is elimi- nated in muscles denervated for 1–3 days, where motor

nerve terminals obviously release no transmitter, either quantally or nonquantally (22,29). After denervation, CB was still able to hyperpolarize the RMP even when the nerve terminal had degenerated. This was evidently not the result of increased input resistance of the fiber membrane or activation of the electrogenic pump, be- cause both parameters were unchanged; the average input resistance of the muscle membrane incubated for 24 h in the control medium was about 0.75 M▲, and it did not change after 1 h exposure to 5 × 10—8 M CB. Similar results were obtained in muscle fibers dener- vated in vivo. Denervation experiments therefore showed that exogenous CB can slow the development of EDP directly, without stimulating the endogenous ACh release from nerve terminals through muscarinic autoreceptors. The participation of nerve muscarinic autoreceptors also seems unlikely because the minimal effective concentration of CB (which can change the Ca2+-dependent frequency of miniature endplate poten- tials through presynaptic muscarinic receptors) was es- timated to be rather high, that is, 0.6 µM (41,42). Apparently, this concentration cannot be attained non- quantally. It is also much higher than the doses of ACh and CB (5–10 × 10—8 M), which can maintain the RMP in denervated muscles through Ca2+-regulated production of NO in our experiments (7,10).

Effect of Atropine on the Time-Course of Muscle Depolarization
If muscarinic receptors are involved in the regu- lation of EPD, then the in vitro incubation of dener- vated muscles with muscarinolytic drugs should facilitate the development of EPD due to the protec- tion of fibers from the action of ACh still being re- leased for a period of several hours after nerve section (22,29). We incubated the muscle strips in the culture medium and measured the RMP every 30 min in the absence and presence of Atr. In control muscles, the RMP became depolarized significantly by 2.5 mV after the 90 min and this depolarization continued to increase for 3 h of muscle incubation. In the presence of both Atr and Pir, the same level of EPD developed more rapidly, already within 30 min (11). This accel- erating effect of Atr and Pir was highly significant dur- ing the first hour after denervation.

Glutamate Reduces EPD
It is known that glutamate-like molecules are pres- ent in the cytoplasm of motoneurons (45) and that radioactively labeled L-glutamate can be taken up and

secreted at the frog neuromuscular junction (46). Glutamate-like activity was also demonstrated at rat motor nerve terminals by means of immunocytochem- istry (45,47,48), mRNA coding for glutamate transporter was found in rabbit motoneuron perikarya (45), and NMDAR-1 glutamate receptor subunit was found on the postsynaptic membrane of the rat diaphragm (48). The physiological reason for the participation of gluta- mate synaptic activity has not been obvious until now. We put forward the hypothesis (8,16) that Glu can also participate in the regulation of muscle RMP through nitric oxide, similarly as Ach (Fig. 1, left). It has been shown that incubation of denervated muscles in the presence of exogenous Glu and N-methyl-D-aspartate (NMDA) induces Ca2+-dependent synthesis of NO in the sarcoplasm, which in turn slows down the early post- denervation depolarization (1,8,9). Moreover, we found that Glu and NMDA were inactive in the presence of several known inhibitors (8): The specificity of NMDA receptor-mediated sensitivity of the EPD was followed using NMDA-antagonists of both competitive type (1 × 10—3 M 2-amino-5-phosphonovaleric acid APV and 7-Cl kynuretic acid) (49) and noncompetitive type Zn2+ (50– 52,56) and MK-801 (53) and also a glycine-free medium (6). Let us consider their effects in more detail.

The Effect of Glu and NMDA on Long Nerve Stump Preparations
Both amino acids, Glu and NMDA, depolarized the muscle membrane in acute experiments by about 2 mV. As will be further discussed, this depolarization was removed to a great extent by the NMDA receptor- specific inhibitors and can thus be considered as the expression of Na+ and Ca2+ entry into the cell through NMDA-regulated channels. However, there is no avail- able information about the actual density of these re- ceptors on the rodent muscle fiber sarcolemma. The question is whether the opening of channels which causes depolarization by only a few mV is sufficient to enable Ca2+ entry in amounts that might activate NO-synthase. According to the data from central neu- rons, the Ca2+ fraction of the ionic current through the NMDA receptor channels is about 7% (54). Direct measurement of ionized calcium might provide further information on this subject (55).

The Effect of Glu and NMDA on Short Nerve Stump Preparations
The EPD (which was 8–10 mV within 3 h in vitro) was substantially smaller (3 mV) when muscles were bathed with 1 × 10—3 M Glu or 1 × 10—3 M NMDA in

the presence of glycine and Mg2+. The protective effect of Glu was also observed when its concentration was de- creased to 5 × 10—4 M, but the 1 × 10—4 M dose was already ineffective. This concentration dependence of the Glu action on EPD in the millimolar range is similar to central neurons (51,52,56). Because Mg2+ decreases the efficacy of Glu action on the NMDA-type receptor (57), it could be expected that lower doses of Glu would be more effective if Mg2+ were decreased in Hank’s medium. Interestingly, the early effect of Glu and NMDA on the RMP already mentioned for long-stump preparations (depolarization by about 2–3 mV was ob- served 15–20 min after application of these amino acids to freshly isolated muscle strips) did not develop in the presence of APV and MK-801. Thus the marginal depolarization might be caused by direct opening of receptor-coupled channels to Na+ and Ca2+.

Competitive and Noncompetitive Inhibition
The effects of Glu and NMDA on EPD were in- hibited in a dose-dependent manner by the competitive inhibitor 2-amino-5-phosphonovaleric acid (APV) with Ki 6.3 × 10—4 calculated for the most effective con- centration of Glu (1 × 10—3 M)(6). APV also prevented the less marked protective action of lower doses of Glu (5 × 10—4 M).
It is known that neuronal NMDA receptors have two cation-binding sites, one for Mg2+ and another for Zn2+ and Cd2+, which hinder channel opening when occupied by a particular ion. The Mg2+ site is located inside the transmembrane domain of the receptor pro- tein. Activation of NMDA receptors opens cation- selective ionic channels which permit Ca2+ entry into the cell. In neurons, this influx is inhibited by Mg2+ (57). In a similar manner, an increase of Mg2+ from
0.3 mM to 5 mM completely eliminated the effect of Glu on EPD of the denervated diaphragm (8). Besides Mg2+, the noncompetitive antagonist MK-801 was found to inhibit the cationic receptor channels in dif- ferent brain neurons with different efficacy during Glu action in relatively low concentrations (53,58). MK-801 also prevented the Glu effect on EPD when applied in the 2 × 10—7 M concentration. This obser- vation further supports the idea that the NMDA-type of Glu receptors is a target for Glu action on the mus- cle membrane. But when tested alone, 1 × 10—7 M MK-801 surprisingly hyperpolarized the RMP during the first 3 h after denervation, similarly as Glu itself. Moreover, MK-801 in this concentration even in- creased the hyperpolarizing effect of 1 × 10—3 M Glu. The possibility that ACh subsynaptic receptors are activated was excluded by experiments with concomitant application of tubocurarine (59). This classical in- hibitor of the nicotinic type of ACh receptors failed to change the action of MK-801. Another possibility con- cerning opening of the NMDA-receptor channel by this particular concentration of MK-801 was, never- theless, confirmed. The competitive inhibitor APV (1 × 10—3 M) restored the EPD when present together with 1 × 10—7 M MK-801. We have no obvious ex- planation for this phenomenon.
Another cation-binding site is apparently located on the outer part of the NMDA receptor and can be occupied by Zn2+ and Cd2+ (50,51). Zn2+ ions bind to some cysteine and glycine groups of the receptor moiety
(52) and can decrease or eventually inhibit NMDA responses in different neurons in concentrations ranging from 5 × 10—6 to 1 × 10—3 M (50,51,56) and possibly function as modulators of NMDA receptors when taken up and released from neurons (60–62). In muscles, Zn2+ inhibited NMDA action and deepened the EPD in con- centrations not exceeding its effect in other tissues (6).

Glycine Site
It is known that for the activation of NMDA- sensitive ion channels by Glu, glycine has to be bound to a particular site on the receptor moiety (63,64). If the effect of Glu on the early postdenervation depolariza- tion is mediated by the NMDA-sensitive ion channel, then incubation of muscles in a medium without glycine or with a specific competitor for the glycine site, for ex- ample, 7-Cl-kynuretic acid, should eliminate the Glu action. Indeed, the effects of Glu were inhibited by both glycine-free solutions and 7-Cl-kynurenic acid (6). This means that NMDA receptors in the diaphragm, simi- larly to neurons, bear a modulatory site that has to be occupied by glycine to permit NMDA action.
In conclusion, the RMP of rat diaphragm fibers is substantially lower in the presence of Glu. Properties of the receptive substance are very similar, but not identi- cal, to the NMDA subtype of neuronal glutamate recep- tor. The activation of this particular type of glutamate receptor by glutamate (supposedly released from the motor nerve, see 45,47,48) can be considered to be one of the pathways by which Ca2+ enters the muscle cyto- plasm and activates NO-synthase, which in turn modu- lates the Cl— transporter in the sarcolemma (see later).

Muscle NMDA Receptors Regulate the Resting Membrane Potential through NO-Synthase
Activation of NMDA receptors opens cation- selective ionic channels that permit Ca2+ entry into the

cell. In neurons, this influx is inhibited by Mg2+ ions (57). In denervated muscles, an increase of Mg2+ from
0.3 mM to 5 mM completely eliminated the effect of Glu. In other words, the early depolarization present in denervated muscles develops if the NMDA receptor- mediated Ca2+ entry is inhibited. The entry of Ca2+ into the cytoplasm via NMDA receptors can activate the NO-cascade. To test whether this cascade is in- volved in Glu action, NAME, an inhibitor of NO- synthase, was added to the muscle bath together with Glu. At 0.1 mM, NAME completely eliminated the Glu protection of RMP. The specificity of its effect was tested by adding the NO-synthase substrate, L-arginine, together with NAME. In a concentration of 1 mM, L-arginine partially reversed the inhibition induced by NAME. This indicates that Glu may maintain the RMP by enhancing NO synthesis in a similar way to ACh. NAME by itself did not influence the postdenervation drop of RMP. The effect of nitroprusside, which might serve as a direct donor of the NO group in an aqueous medium (65) also speaks in favor of this mechanism. The application of sodium nitroprusside (66), which is degraded to 2 NO in solution, inhibits early post- denervation depolarization of isolated rat diaphragm fibers. The observation that “old” solutions of sodium nitroprusside (that have been allowed to decompose) are without effect and that hemoglobin, oxadiazolo quinox- alinone (ODQ), and methylene blue (both guanylyl cy- clase inhibitors) can antagonize the inhibition normally produced by sodium nitroprusside also suggests that the inhibitory effects of sodium nitroprusside on EPD are mediated by NO and guanylyl cyclase. The RMP of denervated muscles kept for 3 hours in the presence of
0.1 mM sodium nitroprusside was significantly less de- polarized (—70 mV) than in the controls (—66 mV). The evidence for the presence of Glu in motoneurons and its effect on EPD suggests that Glu is released quan- tally or nonquantally (67) from the motor nerve ending at the neuromuscular junction together or independently of ACh. The Glu triggers the production of NO with subsequent phosphorylation of membrane proteins, channels, or transporters involved in RMP maintenance. Nerve section might impair the Glu release—similarly as has been observed for nonquantal ACh release—and subsequent Ca2+-dependent NO production.
Present results are in accord with observations that rat skeletal muscles express neuronal type nitric oxide synthase (68) and that the NO is involved in smooth muscle relaxation and hyperpolarization because of opening of specific types of K+ channels (69,70). The structural similarity between imidazole and the NO syn- thase inhibitor 7-nitroindazole (71) and the findings that glutamate and CB regulate muscle postdenervation depolarization through NO production points to the possibility that imidazole, carnosine, and anserine also inhibit NO production and hence they decrease— similarly to 7-nitroindazole—the hyperpolarizing ef- fect of glutamate and CB. If this is so, then the exoge- nous source of NO, sodium nitroprusside (71), could mimic the effect of glutamate and CB in lowering the EPD and imidazole compounds would not prevent it. The experiments with SNP and 7-nitroindazole, carno- sine, anserine, and imidazole supported this idea (9). Imidazole and imidazole-containing dipeptides, carno- sine, and anserine, interfere with glutamate and CB de- polarization-preventing effect much as 7-nitroindazole does, apparently by inhibiting the skeletal muscle NO synthase. Because carnosine and anserine are released from the nerve during neuromuscular transmission (72), one can speculate that they may have a local balancing function on muscle NO-synthase during liberation of neurotransmitters at the endplate and play a role in the trophic dialogue between the nerve and muscle, besides their antioxidant function in brain and muscles (73).

Carbachol and Glutamate Effects Are Directed toward the Muscle Chloride Transporter
An inward-directed, furosemide-sensitive chlo- ride transport is activated after nerve section and is apparently responsible for EPD and also of the other abnormalities such as the loss of ability of the muscle fibers to control their volume in hypertonic solutions
(14) and probably also the appearance of extrasynap- tic ACh receptors (15). The phosphorylation via the NO cascade might keep the transporter inactive, and the lack of nonquantal ACh and possibly of glutamate leads to dephosphorylation and activation of chloride entry. We addressed this experimentally and studied the CB and glutamate action on muscles with active and inhibited Cl— transporter (7).

Chloride Transport Inhibition
First, we measured muscles with the long nerve stump, where no EPD is observed within 3–4 h. As al- ready mentioned, 1 mM Glu depolarized the strips with the long nerve stump from the initial —74.5 to —72.0, whereas superfusion of these fibers with a solution containing 5 × 10—8 M CB hyperpolarized the RMP by
1.4 mV to —77.3 mV within 15–20 min. The hyper- polarization by CB was observed in previous studies and can be explained by activation of the electrogenic ouabain-sensitive sodium pump. (5,23,32,33). In the

denervated strips with no stump, early postdenervation depolarization of 7.8 mV developed within 3 h, the RMP dropped from —74.5 mV to —66.6 mV. When Glu (48) or CB in a concentration mimicking nonquantal ACh release (23,31,33,74,75) was present in the bath, the depolarizations were reduced to —71.3 and —72.2 mV, respectively. ACh acted similarly, and the effects of CB and ACh were not influenced by tubocurarine.
The protective effects of Glu and CB during inhi- bition of Cl— transport induced by bathing the muscle strips in either a Cl— -free medium or a medium con- taining 1 × 10—4 furosemide significantly decreased the EPD. In a Cl— -free medium or in the presence of furosemide the RMP was —74.2 mV after dissection and only dropped to —70.2 mV within 3 h. This indi- cates that inhibition of Cl— transport prevents to a great extent EPD, apparently by preventing the load- ing of muscle fiber with Cl—. (Loading with Cl— can otherwise decrease the Cl— transmembrane gradient and shift the Cl— equilibrium potential towards posi- tive values, that is, depolarize the RMP).

Chloride Transport Activation
It is known that hypertonicity activates Cl— trans- port in muscle fibers (30). Both impermeable sucrose and NaCl, when used for increasing the osmolarity to 500 mosmol/L, significantly potentiated EPD to
—62 mV, respectively. But the inhibition of Cl— trans- port by furosemide in hypertonic solutions of both types completely prevented the depolarization, and RMPs were about —73.5 mV. Because the massive de- polarization might be due an increased Na+ trans- membrane gradient and resulting Na+ entry into the sarcoplasm while bathing the muscle in the NaCl hy- pertonic solutions, we added tetrodotoxin, a sodium channel blocker. This drug did not prevent the massive postdenervation depolarization potentiated by hyper- osmotic solution. Thus, hyperosmolarity of both types equally augmented Cl—-transport, which forms the background of the EPD.
Glu and CB completely lost their efficacy to de- crease the early depolarization in a hypertonic sucrose solution. RMPs were about —62 mV, and these values did not differ from those found in a hypertonic solu- tion only.

CONCLUSION

These experiments show that early postdenerva- tion depolarization of muscle fibers is due to activation of chloride inward transport, which is furosemide sensitive. The application of furosemide holds the RMP near —70 mV, as in Cl—-free incubation medium. On the other hand, hypertonic solutions enhanced this de- polarization in the absence of furosemide. Once Cl— transport is blocked by furosemide, the hypertonic so- lution loses this effect. This seems to be a plausible explanation of the depolarization, which is in accord with previous studies (5,13). The hyperpolarizing effect of Glu and CB on the early postdenervation depolariza- tion can be ascribed to their potency to inhibit, at least partly, the active Cl— transport into muscle fibers. The involvement of another transport system, which can be potentiated by CB, namely sodium electrogenic trans- port, was excluded by experiments with ouabain (7). CB was used here in concentrations which were found to mimic the effect of nonquantally released ACh at adult endplates (31). The ACh leakage may cause sustained depolarization in muscles with inhibited cholinesterase and can be mimicked by the bath application of ACh or its nonhydrolyzable derivative CB. This depolarization is definitely receptor-mediated and can be eliminated by curarization, α-bungarotoxin, or desensitization (26,31). There is also evidence that nonquantal leakage might be responsible for 1–3 mV hyperpolarization of the endplate area of muscle fiber with intact cholinesterase (23,33). This effect is ouabain-dependent and could therefore be explained by activation of the Na+ electro- genic pump by ACh. Here, another role of nonquantal ACh is being suggested. ACh, when released from the nerve terminal, can apparently maintain the Cl— inward current at a low value and RMP at a high level. After the nerve section, this protection from depolarization disappears within several hours (Fig. 1, right). We pro- pose that the main effect is due to nonquantal release (see also 4, 75), because it disappears very soon after denervation (3–4 h, 21,22,29,33) in both rats and mice, in parallel with development of EPD.
Glutamate is also a candidate for the regulation of transmission at the neuromuscular junction (16). We demonstrated that the NMDA-subtype of Glu receptors might be involved in the regulation of RMP in muscle fibers, through the NO-synthase system. nNOS is en- riched at the neuromuscular junction, where it is ap- parently co-localized with the NMDA receptor (76). The furosemide-sensitive Cl— transporter might be a target of phosphorylation and change its function after the action of Glu probably released by the nerve. As far as NMDA receptors are concerned, Glu might initiate Ca2+ fluxes into the sarcolemma followed by activation of Ca2+-dependent NO synthase. NO would then acti- vate a soluble guanylyl cyclase type 1, which was found in fast muscle fibers (68). cGTP is produced, which in turn activates specific proteinkinase(s) and eventually the Cl—-transporter protein is phosphory- lated and activated. NO can also enter nerve terminals and inhibit the NQR of ACh as we recently demon- strated recently (36,37,77) and thus decrease its own production. This feedback (supported by experiments with NO scavenger hemoglobin) tunes the trophic in- fluence of the nerve on muscle fibers and keeps them in an optimal viable state.
Another regulatory pathway might involve ATP. We recently demonstrated that ATP (but not adeno- sine), which is released alone or together with ACh from presynaptic vesicles, decreases the NQR and in- hibits very effectively (78) the nonquantal acetyl- choline release at the mouse neuromuscular junction. This might act as a link between quantal and non- quantal ACh release and be of definite physiological significance.
Although the role of Glu in the proposed neu- rotrophic mechanism could be ascribed to opening of NMDA-sensitive channels for Ca2+, the pathway by which CB and ACh regulate the RMP is less under- standable. Most probably the latter compounds influ- ence Ca2+ channels (apparently through M1 receptors), as Mg2+, verapamil, diltiazem, nifedipine, and Cd2+
(10) can eliminate their effect on EPD. Direct proof of this hypothesis, however, is still lacking at the level of single channel currents.
Thus, resting nonquantal ACh and probably Glu release from nerve endings activate Ca2+ entry into the sarcoplasm, with the subsequent activation of nNOS and production of NO molecules. This is a “tropic” message from motoneurons that keeps the Cl— trans- port inactive in the innervated sarcolemma (5,13,79). After denervation, the secretion of ACh and Glu at the neuromuscular junction is eliminated and the produc- tion of NO in the sarcoplasm is lowered (Fig. 1, right). As a result, the Cl— influx is activated with subsequent elevation of intracellular Cl— concentration. The equi- librium Cl— potential becomes more positive and the muscle membrane becomes depolarized. Which part of the chloride transporter protein undergoes the phos- phorylation/dephosphorylation process remains to be elucidated.
Experiments are now in progress to check whether NO can also regulate other properties of the muscle membrane (cf. 80,81,82). In particular, in rats that were treated with intraperitoneal injections of L-NAME (a nonselective NOS inhibitor) and S-ethylthiourea (which is a very potent selective inhibitor of the in- ducible NOS) for several days, both inhibitors evoked in innervated muscles another typical postdenervation sign connected with Na+ channels: tetrodotoxin resistance and anodal break excitation (83).
An intriguing idea seems to be that the direct stim- ulation of denervated muscles (even by subthreshold pulses, 84,85), which can partly overcome the absence of nerve, increases the intracellular Ca2+ by release from the endoplasmic cisterns or by promoting the Ca2+ inward fluxes during depolarization; NO cascade is then maintained in the active state.

REFERENCES

1. Albuquerque, E. X., Schuh, F. T., and Kauffman, F. C. 1971. Early membrane depolarization of the fast mammalian muscle after denervation. Pflugers Arch. 328:36–50.
2. Shabunova, I. and Vyskocˇil, F. 1982. Postdenervation changes of intracellular potassium and sodium measured by ion selective mi- croelectrodes in rat soleus and extensor digitorum longus muscle fibers. Pfluegers Arch. 394:161–164.
3. Lorkovic, H. and Tomanek, R. J. 1977. Potassium and chloride conductances in normal and denervated rat muscles. Am. J. Physiol. 232:C109–C114.
4. Bray, J. J., Hawken, M. J., Hubbard, J. I., Pockett, S., and Wil- son, L. 1976. The membrane potential of rat diaphragm muscle fibers and the effect of denervation. J. Physiol. 255:651–667.
5. Urazaev, A. Kh., Chikin, A. V., Volkov, E. M., Poletaev, G. I., and Khamitov, Kh. S. 1987. Effect of acetylcholine and carba- chol on the resting membrane potential in diaphragm muscle of the rat after denervation [in Russian]. Sechenov’s Physiol. J. 40:360–362.
6. Urazaev, A. Kh., Naumenko, N. V., Poletayev, G. I., and Nikol- sky, E. E., F. 1998. The effect of glutamate and inhibitors of NMDA receptors on postdenervation decrease of membrane po- tential in rat diaphragm. Mol. Chem. Neuropathol. 33:163–174.
7. Urazaev, A. Kh., Naumenko, N. V., Nikolsky, E. E., and Vyskocˇil, F. 1999. The glutamate and carbachol effects on the early post-denervation depolarization in rat diaphragm are di- rected towards furosemide-sensitive chloride transport. Neu- rosci. Res. 33:81–86
8. Urazaev, A. Kh., T., Poletaev, G. I., Nikolsky, E. E., and Vyskocˇil,
F. 1995. Muscle NMDA receptors regulate the resting membrane potential through NO-synthase. Physiol. Res. 44:205–208.
9. Urazaev, A. Kh., Naumenko, N. V., Poletayev, G. I., Nikolsky,
E. E., and Vyskocˇil, F. 1996. Nitroprusside decrease the early postdenervation depolarization of diaphragm muscle fibers of the rat. Eur. J. Pharmacol. 316:219–222.
10. Urazaev, A. Kh., Naumenko, N. V., Poletayev, G. I., Nikolsky,
E. E., and Vyskocˇil, F. 1997. Acetylcholine and carbachol pre- vent muscle depolarization in denervated rat diaphragm. Neu- roReport 8:403–406.
11. Urazaev, A. Kh., Naumenko, N. V., Malomough, A., Nikolsky,
E. E., and Vyskocˇil, F. 2000. Carbachol and acetylcholine delay the early postdenervation depolarization of muscle fibers through M1-cholinergic receptors. 2000. Neurosci. Res. 37:255–263.
12. Urazaev, A. Kh., Surovtsev, V. A., Chikin, A. V., Volkov,
E. M., Poletaev, G. I., and Khamitov, Kh. S. 1987. Neurotrophic control of transmembrane Cl—-pump in mammalian muscle fibers [in Russian]. Neurophysiology, 19:766–771.
13. Betz, W. J., Caldwell, J. H., and Harris, G. L. 1986. Effect of denervation on a steady electric current generated at the end- plate region of rat skeletal muscle. J. Physiol. 373:97–114.
14. Sitdikov, R. F., Urazaev, A., Volkov, E. M., Poletaev, G. I., and Khamitov, Kh. S., 1991. Neurotrophic control of the ionic regulation mechanisms for intracellular water content in muscle fibers of mammals [in Russian]. Neirofiziologiia 23: 625–628.
15. Chikin, A. V., Urazaev, A., Volkov, E. M., Poletaev, G. I., and Khamitov, Kh. S. 1987. Effect of disruption of chloride con- ductance on the development of denervation changes in muscle fiber membranes of the rat [in Russian]. Fiziol. Zh. SSSR 73: 51–55.
16. Malomouzh, A. I., Mukhtarov, M. R., Nikolsky, E. E., Vyskocˇil, F., Lieberman, E. M., and Urazaev, A. Kh. 2003. Glutamate reg- ulation of non-quantal release of acetylcholine in the rat neuro- muscular junction. J. Neurochem. in press.
17. Stamler, J. S. and Meissner, G. 2001. Physiology of nitric oxide in skeletal muscle. Physiol. Rev. 81:209–237.
18. Grozdanovic, Z. 2001. NO message from muscle. Microscop. Res. Technique 55:148–153.
19. Gutmann, E. 1976. Neurotrophic relations. Annu. Rev. Physiol. 38:177–216.
20. McArdle, J. J. 1983. Molecular aspects of the trophic influence of nerve on muscle. Prog. Neurobiol. 21:135–138.
21. Nikolsky, E. E., Voronin, V. A., and Oranska, T. I. 1985. Dy- namics of quantal and nonquantal acetylcholine release after motor nerve cutting [in Russian]. Dokladi Akademii Nauk SSSR, 281:762–764.
22. Zemková, H., Vyskocˇil, F., and Edwards C. 1987. A study on early postdenervation changes of nonquantal and quantal acetylcholine release in the rat diaphragm. Pflueg. Arch. 409: 540–546.
23. Vyskocˇil, F., Nikolsky, E. E., Zemková H., and Kru˙sˇek, J. 1995. The role of nonquantal release of acetylcholine in regula- tion of postsynaptic membrane electrogenesis. J. Physiol. (Paris) 89:157–162.
24. Katz, B. and Miledi, R. 1977. Transmitter leakage from motor nerve endings. Proc. Royal Soc. Lond. Ser. B 196:59–72.
25. Vyskocˇil, F. and Illes, P. 1977. Nonquantal release of transmit- ter at mouse neuromuscular junction and its dependence on the activity of Na-K-ATPase. Pflueger. Arch. 370:295–297.
26. Edwards, C., Dolezˇal, V., Tucˇek, S., Zemková, H., and Vyskocˇil, F. 1985. Is an acetylcholine transport system respon- sible for nonquantal release of acetylcholine at the rodent myoneural junction? Proc. Natl. Acad. Sci. USA 82:3514–3518.
27. Vizi, E. S. and Vyskocˇil, F. 1979. Changes in total and quantal release of acetylcholine in the mouse diaphragm during activa- tion and inhibition of membrane ATPase. J. Physiol. 286:1–14.
28. Dolezˇal, V. and Tucˇek, S. 1983. The synthesis and release of acetylcholine in normal and denervated rat diaphragms during incubation in vitro. J. Physiol. 334:461–474.
29. Nikolsky, E. E., Oranska, T. I., and Vyskocˇil, F. 1996. Non- quantal acetylcholine release in the mouse diaphragm after phrenic nerve crush and during recovery. Exp. Physiol. 81: 341–348.
30. Sitdikov, R. F., Urazaev, A., Volkov, E. M., Poletaev, G. I., and Khamitov Kh. S. 1989. Effects of hyperosmolarity and furosemide on resting membrane potentials and skeletal muscle fiber volume in rats [in Russian]. Bull. Eksp. Biol. Med. 108: 563–566.
31. Vyskocˇil, F., Nikolsky, E. E., and Edwards, C. 1983. An analy- sis of the mechanisms underlying the nonquantal release of acetylcholine at the mouse neuromuscular junction. J. Neurosci. 9:429–435.
32. Dlouhá, H., Teisinger, J., and Vyskocˇil, F., 1979. Activation of membrane Na+/K+-ATPase of mouse skeletal muscle by acetylcholine and its inhibition by alpha-bungarotoxin, curare and at- ropine. Pflügers Arch. 380:101–104.
33. Nikolsky, E. E., Zemková, H., Voronin, V. A., and Vyskocˇil, F. 1994. Participation of nonquantal acetylcholine in release in surplus polarization of the mouse diaphragm fibers at the end- plate zone. J. Physiol. 477:497–502.
34. Vyskocˇil, F. and Gutmann, E. 1977. Anabolic effect of testos- terone on the levator ani muscle of the rat. Pflügers Arch. 371:3–8.
35. Vincent, S. R. and Hope B. T. 1992. Neurons that say NO. Trends Neurosci. 15:108–113.
36. Mukhtarov, M. R., Vyskocˇil, F., Urazaev, A. K., and Nikolsky,
E. E. 1999. Nonquantal acetylcholine release is increased after nitric oxide synthase inhibition. Physiol. Res. 48:315–317.
37. Mukhtarov, M. R., Urazaev, A. K., Nikolsky, E. E., and Vyskocˇil, F. 2000. Effect of nitric oxide and NO synthase inhi- bition on nonquantal acetylcholine release in the rat diaphragm. Eur. J. Neurosci. 12:980–986.
38. Hosey, M. M. 1992. Diversity of structure, signaling and regu- lation within the family of muscarinic cholinergic receptors. FASEB J. 6:845–852.
39. Reyes, R. and Jaimovich, E. 1996. Functional muscarinic re- ceptors in cultured skeletal muscle. Arch. Biochem. Biophys. 331:41–47.
40. Vizi, E. S. and Somogyi, G. T. 1989. Prejunctional modulation of acetylcholine release from the skeletal neuromuscular junc- tion: Link between positive (nicotinic)- and negative (mus- carinic)-feedback modulation. Brit. J. Pharmacol. 97:65–70.
41. Nikolsky, E. E., Bukharaeva, E. A., Strunsky, E. G., and Vyskocˇil, F. 1991. Depression of miniature endplate potential frequency by acetylcholine and its analogues in frog. Brit. J. Pharmacol. 104:1024–1032.
42. Wessler, I. 1989. Control of transmitter release from the motor nerve by presynaptic nicotinic and muscarinic autoreceptors. Trends Pharmacol. Sci. 10:110–114.
43. Wessler, I. 1989. Acetylcholine at motor nerves: Storage, re- lease, and presynaptic modulation by autoreceptors and adreno- ceptors. Int. Rev. Neurobiol. 34:283–384.
44. Zemková, H., Vyskoˇcil, F., and Edwards, C. 1990. The effects of nerve terminal activity on nonquantal release of acetylcholine at the mouse neuromuscular junction. J. Physiol. 423:631–640.
45. Meister, B., Arvidsson, U., Zhang, X., Jacobsson, G., Villar,
M. J., and Hokfelt, T. 1993. Glutamate transporter mRNA and glutamate-like immunoreactivity in spinal motoneurons. Neuro- report 5:337–340.
46. Kerkut, G. A., Shapiro, A., and Walker, R. J. 1967. The trans- port of C-labelled material from CNS to muscle along a nerve trunk. Comp. Biochem. Physiol. 23:729–748.
47. Waerhaug, O. and Ottersen, O. P. 1993. Demonstration of glu- tamate-like immunoreactivity at rat neuromuscular junctions by quantitative electron microscopic immunocytochemistry. Anat. Embryol. (Berl.) 188:501–513.
48. Berger, U. V., Carter, R. E., and Coyle, J. T. 1995. The im- munocytochemical localization of N-acetylaspartyl glutamate, its hydrolysing enzyme NAALADase, and the NMDAR-1 re- ceptor at a vertebrate neuromuscular junction. Neuroscience 64: 847–850.
49. Davies, J. and Watkins, J. C. 1982. Actions of D and L forms of 2-amino-5-phosphonovalerate and 2-amino-4-phosphonobutyrate in the cat spinal cord. Brain Res. 235:378–386.
50. Mayer, M. L., Vyklicky, L. J., and Westbrook, G. L. 1989. Modu- lation of excitatory amino acid receptors by group II B metal cations in cultured mouse hippocampal neurons. J. Physiol. 415: 329–350.
51. Westbrook, G. L. and Mayer, M. L. 1987. Micromolar concen- tration of Zn2+ antagonize NMDA and GABA responses of hip- pocampal neurons. Nature 328:640–643.
52. Legendre, P. and Westbrook, G. L. 1990. The inhibition of sin- gle N-methyl-D-aspartate-activated channels by zinc ions on cultured rat neurons. J. Physiol. 429:429–449.

53. Wong, E. H. F., Kemp, J. A., Priestly, T., Knight, A. R., Woodruff, G. N., and Iversen, L. L. 1986. The anticonvulsant MK-801 is a potent N-methyl-D-aspartate antagonist. Proc. Natl. Acad. Sci. USA 83:7104 –7108.
54. Schneggenburger, R., Zhou, Z., Konnerth, A., and Neher, E. 1993. Fractional contribution of calcium to the cation current through the glutamate receptor channels. Neuron 11:133–143.
55. Stankovicˇová, T., Zemková, H., Breier, A., Amler, E., Burkhard, M., and Vyskocˇil, F. 1995. The effect of calcium and calcium channel blockers on sodium pump. Pfluegers Arch. Eur. J. Physiol. 429:716–721.
56. Peters, S., Koh, J., and Choi, D. W. 1987. Zinc selectively blocks the action of N-methyl-D-aspartate on cortical neurons. Science 236:589–593.
57. Jahr, C. E. and Stevens, C. F. 1990. Voltage dependence of NMDA-activated macroscopic conductances predicted by single- channel kinetics. J. Neurosci. 10:3178–3182.
58. Hahn, J. S., Aizenmann, E., and Lipton, S. A. 1988. Central mammalian neurons normally resistant to glutamate toxicity are made sensitive by elevated extracellular Ca2+: Toxicity is blocked by the N-methyl-D-aspartate antagonist MK-801. Proc. Natl. Acad. Sci. USA 85:6556–6560.
59. Amador, M. and Dani, J. A. 1991. MK-801 inhibition of nico- tinic acetylcholine receptor channels. Synapse 7:207–215.
60. Howell, G. A., Welch, M. G., and Frederickson, C. J. 1984. Stimulation-induced uptake and release of zinc in hippocampal slices. Nature 308:736–738.
61. Assaf, S. Y. and Chung, S.-H. 1984. Release of endogenous Zn2+ from brain tissue during activity. Nature 308:734–736.
62. Masters, B. A., Quaife, C. J., Erickson, J. C., Kelly, E. J., Froelick, G. L., Zambrowicz, B. P., Brinster, R. L., and Palmiter,
R. D. 1994. Metallothionein III is expressed in neurons that se- quester zinc in synaptic vesicles. J. Neurosci. 14:5844 –5857.
63. Johnson, J. W. and Ascher, P. 1987. Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature 325: 529–531.
64. Wood, E. R., Bussey, T. J., and Philips, A. G. 1994. A glycine antagonist 7-chlorkynuretic acid attenuates ischemia-induced learning deficits. Neuroreport 4:151–154.
65. Moncada, S., Palmer, R. M. J., and Higgs, E. A. 1991. Nitric oxide: Physiology, pathophysiology and pharmacology. Phar- macol. Rev. 43:1709–1715.
66. Bohme, G. A., Bon, C., Stutzmann, J. M., Doble, A., and Blan- chard, J. 1991. Possible involvement of nitric oxide in long-term potentiation, Eur. J. Pharmacol. 199:379–381.
67. Antonov, S. M. and Magazanik, L. G. 1988. Intense nonquantal release of glutamate in an insect neuromuscular junction. Neu- rosci. Lett. 93:204–208.
68. Kobzik, L., Reid, M. B., Bredt, D. S., and Stamler, J. S. 1994. Nitric oxide in skeletal muscle. Nature, 372:545–548.
69. Kilpatrick, E. V. and Cocks, T. M. 1994. Evidence for dif- ferential roles of nitric oxide (NO) and hyperpolarization in endothelium-dependent relaxation of pig isolated coronary ar- tery. Br. J. Pharmacol. 112:557–565.
70. Miyoshi, H. and Nakaya, Y. 1994. Endotoxin-induced non- endothelial nitric oxide activates the Ca(2+)-activated K+ channel in cultured vascular smooth muscle cells. J. Mol. Cell Cardiol. 26:1487–1495.
71. Böhme, G. A, Bon, C., Stutzmann, J. M., Doble, A., and Blan- chard, J. C. 1991. Possible involvement of nitric oxide in long- term potentiation. Eur. J. Pharmacol. 199:379–381.
72. Poletaev G. I. 1970. Mechanism of imidazole action on frog neuromuscular junctions [in Russia]. Sechenov’s Physiol. J. 56: 64 –67.
73. Kohen, R., Yamamoto, Y., Cundy, K. C., and Ames, B. N. 1988. Antioxidant activity of carnosine homocarnosine and anserine present in muscle and brain. PNAS 85:3175–3179
74. Ciani, S. and Edwards, C. 1963. The effect of acetylcholine on neuromuscular transmission in the frog. Pharmacol. Exp. Ther. 142:21–29.
75. Bray, J. J., Forrest, J. W., and Hubbard, J. I. 1982. Evidence for the role of nonquantal acetylcholine in the maintenance of the membrane potential of rat skeletal muscle. J. Physiol. F326: 285–296.
76. Grozdanovich, Z. and Grossrau, R. 1998. Co-localization of nitric oxide synthase I (NOS I) and NMDA receptor subunit 1 (NMDAR-1) at the neuromuscular junction in rat and mouse skeletal muscle. Cell Tissue Res. 291:57–63.
77. Mukhtarov, M. R., Urazaev, A. K., Nikolski, E. E., and Vyskocˇil, F. 2001. Modulation by nitric oxide (NO) of the in- tensity of nonquantal mediator secretion in neuromuscular junc- tions in rats. Neurosci. Behav. Physiol. 31:451–455.
78. Galkin, A. V., Giniatullin, R. A., Mukhtarov, M. R., Sˇvandová, I., Grishin, S. N., and Vyskocˇil, F. 2001. ATP but not adenosine inhibits nonquantal acetylcholine release at the mouse neuro- muscular junction. Eur. J. Neurosci. 13:2047–2053.
79. Dulhunty, A. F. 1978. The dependence of membrane potential on extracellular chloride concentrations in mammalian muscle fibers. J. Physiol. 276:67–82.

80. Tews, D. S. 2001. Role of nitric oxide and nitric oxide synthases in experimental models of denervation and reinnervation. Mi- crosc. Res. Tech. 55:181–186.
81. Sˇvandová, I., Vyskocˇil, F., and Ujec, E. 2001. Nitric oxide syn- thase inhibition partially imitates postdenervation tetrodotoxin resistance and anodal break excitation in innervated rat muscles. Physiol. Res. 50:P29.
82. Melzer, W., Schneider, M. F., Simon, B. J., and Szucs, G. 1986. Intramembrane charge movement and calcium release in frog skeletal muscle. J. Physiol. 373:481–511.
83. Bray, J. J. and Harris, A. J. 1975. Dissociation between nerve- muscle transmission and nerve trophic effects on rat diaphragm using type D botulinum toxin. J. Physiol. 253:53–77.
84. Adámek, S., Schutzner, J., Seidle, Z, Smat, V., Pit`ha, J. 1996. Correlation of computer tomography findings with surgical findings in patients with myasthenia gravis. Rozhl. Chir. 75: 237–239.
85. Schutzner, J., Smat, V., Patko, P., Adámek, S., Sláma, J. 1999. Surgical therapy of thymomas. Sb. Lek. 100:27–31.Acetylcholine Chloride