[1] 15w Endplates after esterase inactivation in vivo" correlation between esterase concentration, functional response and fine structure

[1] 221w Mouse sternomastoid muscles were incubated with diisopropylfluorophosphate (DFP) in vivo, and the time course of recovery was studied using histochemistry, EM autoradiography and physiology. We found that: (1) the ability of the muscle to sustain tetanus in response to nerve stimulation is eliminated when the esterases at the neuromuscular junctions are saturated with DFP. This ability is regained partially when <10% of the DFP-binding sites have recovered. (2) There is a positive correlation between the frequency of stimulation at which the tetanic response can be maintained and the extent of acetylcholinesterase (ACHE) recovery. (3) Tetanic responses at fusion frequency (about 100 Hz) appear indistinguishable from controls with only about 25% of normal ACHE. (4) Butyrylcholinesterase (BuChE) possibly of Schwann cell origin recovers more rapidly than does ACHE. (5) The muscle shows fine structural changes involving Z band dissolution and the breakdown of sarcoplasmic reticulum within hours after esterase inactivation. (6) This myopathy reaches a peak at three days after esterase inactivation and is almost fully recovered by two weeks. (7) It can be eliminated if, at the time of esterase inactivation, the nerve is cut or the acetylcholine receptors at the endplate are inactivated by ~-bungarotoxin.We suggest that the myopathy, seen after DFP, is mediated by Ca 2+ fluxes due to prolonged action of acetylcholine (ACh) in the absence of esterases.

[1] 133w The distribution of esterase sites at normal neuromuscular junctions in the mouse has been determined quantitatively by autoradiography (Rogers et al., 1966(Rogers et al., , 1969;;Salpeter, 1967Salpeter, , 1969;;Salpeter et al., 1972Salpeter et al., , 1978) ) using radioactive diisopropylfluorophosphate (DFP) as the probe. In the present study, the same procedures were used to investigate the reappearance of acetylcholinesterase (ACHE) and other DFP-binding sites after inactivation of these sites by DFP in vivo. One interesting observation which has been extensively reported is that the inactivation of endplate esterases causes myopathy involving both muscle and nerve (for example, Filogamo and Gabella, 1966;Rose and Glow, 1967;Preusser, 1967;Fischer, 1968;Ariens et al., 1969;Fenichel et al., 1972;Feng et al., 1973;Lowndes et al., 1974;Laskowski et al., 1975) and seems to be independent of the type of AChE inhibitor employed.

[2] 153w Recovery occurs in 2-3 weeks depending on the experimental conditions of inactivation. We, therefore, compared the recovery of esterase sites with some physiological, histochemical and fine structural measures and correlated these with the time course of myopathy. We found that AChE recovers later than do other DFP-binding sites; that the Schwann cells may be involved in the recovery; that the ability of the muscle to sustain a tetanic contraction for a few seconds at 100 Hz is regained when less than 10% of normal AChE sites are recovered; and that a very effective way of preventing DFP-induced myopathy is to inactivate the acetylcholine receptor (AChR) prior to applying the DFP. We suggest that a major role of AChE is to restrict the action of ACh both in time and space, and thus protect the muscle from short term fatigue and from long term myopathy, which may be due to extensive Ca 2+ fluxes.

[1] 301w An in vivo preparation of the sternomastoid muscle of the albino mouse was used as previously described (Fertuck et al., 1975). Mice were anaesthetized by Nembutal (52 mg/kg body weight in 10% alcohol). The muscle was exposed by a neck incision and indirectly stimulated with a suction electrode into which a loop of the uncut nerve was drawn. A needle attached to an isotonic transducer was inserted into the surface of the muscle to monitor muscle contractions. The tetanic fusion freauency for this muscle was found to be 70-80 Hz. The exposed muscle was bathed in DFP (10 -3 M in Krebs Ringer pH 7.4) and the extent of esterase inactivation was monitored every 15 min by stimulating the nerve at 100 Hz for 2-5 s. The procedure was continued until the muscle could no longer sustain a tetanic contraction (Fig. 7). This phenomenon has been reported to occur when less than 10% of AChE activity is present (Barnes and Duff, 1953;Barstad, 1960). We found that when muscle treated this way was stained for cholinesterases by the method of Karnovsky and Roots (1964), there was no reaction product (or only a very faint one) seen by light microscopy after 30 min of staining, whereas the contralateral control muscle showed intense staining. We therefore used the inability to sustain tetanus at 100 Hz as the physiological criterion for the elimination of esterase activity. The incision was then sutured and the animal allowed to survive for varying periods up to three weeks. On different days after the operation, the incision was again opened and the muscle response tested by indirect stimulation with 0.1 ms pulses at various frequencies from 100-300 Hz for several consecutive 5 s stimulus durations. The animals were then sacrificed and treated in one or more of the following ways.

[2] 113w We found tfiat the mouse sternomastoid muscle responds to the inactivation of its endplate esterases by becoming what has been termed necrotic by previous authors (for example, Preusser, 1967;Fischer, 1968;Ariens et al., 1969;Feng et al., 1973). The necrosis begins and is most extensive in the region of the endplate. The peak of necrosis is seen between one and three days at which time muscle striations in the region of the endplate are absent, and there is extensive fibre destruction with cellular ifivasion. At seven days, signs of muscle recovery are evident. Many fibres appear normal and in some others, centrally placed strings of nuclei suggest muscle regeneration. Full recovery occurs by 2-3 weeks.

[3] 316w Early fine structural alterations in the region of the neuromuscular junction could be seen immediately after the 2 h period required for complete inactivation by DFP (zero day). Among the earliest changes noted were vesiculation and destruction of the sarcoplasmic reticulum, especially in the region of the triad, mitochondrial damage and changes in the Z band. The Z bands first appear to spread and become diffuse before they 'disappear'. The time course of the spread of damage and recovery is variable, but generally by one day after DFP incubation, the above-listed changes are striking (Figs. 1 and 2). One often sees muscle fibres with intact thick and thin filaments, and normal A and I bands, but with no Z bands and no triads (see also Preusser, 1967;Laskowski et al., 1975Laskowski et al., , 1977)). After the dissolution of the Z band, the myofilaments become disorganized. The filament disorganization reaches a peak between 1-3 days, although filaments frequently still remain aligned along the long axis of a fibre which is devoid of other organelles (Figs. 3 and 4). The destruction spreads from the endplate region deeper into and along the muscle fibre. Schwann cells often appear enlarged with prominent rough endoplasmic reticulum (Fig. 5). Muscle recovery frequently begins in the cytoplasm near the endplate by seven days after DFP. The damage is now greatest in the neuromuscular junction itself (Fig. 5). The terminal axoplasm is sometimes disrupted, containing myelin figures. The axonal membranes are often separated from the PJMs by interposed processes of Schwann cells, or by connective tissue plus clusters of small vesicles appearing free in the cleft. The secondary clefts are widened, the basal lamina is disrupted and granular, and the postjunctional folds are distorted, decreased in size and apparently vesiculated. By two weeks, except for occasional remnants of cleft vesicles and myelin figures, the muscle and endplate morphology appear essentially recovered (Figs. 6a, b).

[4] 125w Of the five conditions designed to study the cause of the necrosis (see Methods), we found that three days after treatment, the muscles and nerves appeared completely normal in those animals that were subjected to the 'sham operation' and in those treated with o~-BTX prior to the application of DFP. In the other three conditions (that is, 2-PAM reactivations after DFP, denervation after DFP and cx-BTX after DFP) some minor damage was sometimes seen similar to that described at day 0. This was probably due to some functional interaction between ACh and receptor in the intact neuromuscular junction during the 2 h inactivation period or, in the case of the denervated preparation, to the time it takes for the nerve to cease releasing ACh. g

[5] 74w By light microscopy no stain (or a very faint one) was seen on day 0 when stained for either AChE or BuChE. BuChE could be stained at 1-3 days and AChE only after 3--7 days. The stain became indistinguishable from normal controls by 2-3 weeks (the data are summarized in Table 1). We therefore conclude that the AChE recovers more slowly than does BuChE. (A similar conclusion was reached by Clouet and Waelsch, 1961).

[6] 88w By EM histochemistry at three days after DFP inactivation, the reaction product due to BuChE is predominantly in the teloglial cap with much less in the synaptic zone (Fig. 4a), whereas that due to AChE is mainly in the synaptic zone (Fig. 4b). Thus during recovery, the earliest reaction product seen by light microscopy appears to be due to BuChE in the teloglial cap. A large variability was seen in the histochemical localization, and no conclusion regarding quantitative aspects of esterase localization can be drawn from these results.

[7] 82w Table 1 and Fig. 7 illustrate the ability of muscles to sustain tetanic contractions in response to nerve stimulation for 5 s every 20 s (0.1 ms stimulus pulse duration, at different stimulation frequencies). The tetanic response is an indirect measure of function. However, it is of interest that the ability to sustain repeated 5 s tetanic responses at increasing frequencies of stimulation is directly correlated with the time after esterase inactivation and, as will be seen below, with AChE site density.

[8] 266w The normal muscle has the ability to maintain the amplitude of a 5 s contracture upon numerous repetitions at 100 Hz and 200 Hz stimulation periods. A slight Fig. 1. Mouse sternomastoid muscle one day after esterase inactivation by DFP in vivo (prefixed by 4% paraformaldehyde). Note spreading and destruction of Z bands (arrow), vesiculation of sarcoplasmic reticulum and mitochondrial disruption in the muscle near the neuromuscular junction. The neuromuscular junction looks essentially normal at this stage. There is considerable variability in the development of muscle breakdown and at one day muscle frequently looks similar to that illustrated in Fig. 2. x 17 500. Fig. 2. Muscle three days after DFP at peak of necrosis (prefixed by 4% paraformaldehyde). Note a relatively normal fibre (at left) not near an endplate, adjacent to the necrotic fibre. Considerable axonal disruption and junctional fold vesiculation is evident, x 17 500. decline was observed only at 300 Hz. The three-day post-DFP muscle sustained responses at 100 Hz partially, but could not sustain any contractures at 200 Hz and 300 Hz. By 14 days, a deviation from normal was seen only at 300 Hz (Fig. 7). Furthermore, when the periods of stimulation at different frequencies were prolonged (about 30 s), we found that the normal muscle could sustain a contraction for this period without difficulty, showing a slight (30%) drop in amplitude only at 300 Hz. The 14-day post-DFP animal could sustain a contracture at 100 Hz in a manner undistinguishable from normal but at 200 Hz and 300 Hz the amplitude dropped by 80% within the 30 s test period.

[9] 204w On day 0, no label above background could be introduced at the endplate by [3 H] DFP. The muscle background was comparable to 0.5 x 103 sites//~m 3 of tissue and equivalent to values previously obtained with normal tissue. Considering first the general distribution of DFP sites, we found that at three days after inactivation, the muscle background was unchanged but the label at the endplate was increased about four-fold. The highest grain density was in the Schwann cytoplasm (Fig. 3) and about 50% of the endplate label was in the entire teloglial cap region (Table 1). The teloglial cap grain density did not increase further but the grain density over the synaptic zone continued to increase with time. Thus, by seven days (Fig. 5), the distribution of developed grains had shifted, and only 27% of the total endplate grains were in the teloglial cap region (Table 1). By two weeks (Fig. 6), approximately 17% of the grains were in the teloglial cap, which is approaching that in the normal animal, where 10-15% of the label is in the teloglial cap. As will be seen below, however, the absolute DFP site density at the PJM was not yet fully recovered even at 14 days.

[10] 306w For a better comparison with normal animals, we determined the site density (sites/gin 2 of PJM surface area) as described in Methods. Although some endplates were not easy to include in this tabulation because of the disrupted morphology, the results (presented in Table 2) were generally consistent with the regional analysis. We found that, compared with the normal animal, the number of [3 H] DFP binding sites per/am 2 of PJM is roughly 3% at three days, 26% at seven days, and 51% at two weeks after inactivation. The specific 2-PAM reactivated sites had a somewhat slower recovery rate. In the normal adult, AChE represents about 30% of total DFP sites. During recovery a significant amount of AChE was first seen by autoradiography only at seven days. At that time, it represented about 7% of all DFP sites present, and was 6% of the AChE sites normally present in the adult animal. At two weeks, 14% of all DFP sites were reactivated by 2-PAM. This constitutes 25% of the AChE site normally present at normal adult endplates. In addition, we found that at seven 100 Hz SALPETER, KASPRZAK, FENG and FERTUCK 200 Hz 300 Hz Normal 0 day 3 days ~~-:.:.~ 7 days ~ .z 7 ----- 1/. days Fig. 7. Tetanic contractions produced by nerve stimulations at 100,200 and 300 Hz (0.1 ms pulse duration) recorded in the normal sternomastoid muscle and in muscles at different times after inactivation of all esterases with DFP. At each stimulation frequency the muscle was given three 5 s trains of stimuli separated by a 20 s rest period. A 15 rain rest period was used between sets of different frequencies. The three consecutive muscle responses for each frequency were traced and superimposed in this figure: first response ( ), second response ( ..... ) and third response (

[11] 51w ). Note that in the normal animal the amplitude of the tetanic response to repeated stimulations remains constant but that the ability to sustain tetanus decreases with increasing stimulation frequency. Note also that the ability to sustain tetanus at different stimulation frequency depends on the length of recovery time after DFP.

[12] 69w Table 1. Survey of recovery after in vivo inactivation of esterases. tFor details of region analysis see Methods. 'Sites in TC'; refers to the total number of grains over the teloglial cap (Schwann cell plus connective tissue) expressed as a percent of total number of endplate grains. 'Site density Schwann/JF' gives the ratio of grain density (grains/unit volume of tissue) in Schwann cytoplasm to that over the junctional fold.

[13] 13w ~For the 2-PAM reactivated (ACHE) sites the value was 16% at this time.

[14] 58w 108 SALPETER, KASPRZAK, FENG and FERTUCK -(a) (a)Each value is an average site density based on 1-6 experimental animals(numberof animals given in parentheses). For each animal, six different regions of the muscle were sampled and a total of 200-300 developed grains were obtained. Each determination of site density is limited by the autoradiographic accuracy to +25% (see Methods).

[15] 30w (b)The range of site densities for the three 14-day animals was 3000-7000 sites/#m 2. For the three day and seven day animals, the animal to animal variation was within -+25%.

[16] 154w (C)Normal values are pooled from Salpeter et al. (1972Salpeter et al. ( , 1978)). (d)Autoradiography was not done since histochemistry suggested very little AChE recovery at this time (see Table 1). days only 16% of all the AChE sites was in the teloglial cap compared with 27% of the total DFP binding sites (Table 1). Thus from the earliest recovery time, the AChE sites are distributed between the teloglial cap and the synaptic region in a ratio closer to that seen in the normal animal. The autoradiography was thus consistent with the histochemistry in as far as the general distribution of DFP reactive sites is concerned. A comparison between the results obtained by the two procedures, however, emphasizes the absence of quantitation in the histochemical data. Note that the extensive reaction product seen in Fig. 4b is due to the activity of less than 5% of the AChE sites normally present at the endplate.

[1] 197w In normal adult mouse sternomastoid muscle there are about 9000 DFP-binding sites per/lm 2 of postjunctional surface area. DFP phosphorylates not only esterases but also a variety of proteases (see discussion by Cohen et al., 1959). The studies by Rogers et al. (1966Rogers et al. ( , 1969) ) and Salpeter et al. (1972Salpeter et al. ( , 1978) ) used a combination of DFP and 2-PAM to label various esteratic and non-esteratic sites under condition designed to avoid ageing of the phosphorylated enzyme, and have shown that only 30% of the DFP-binding sites at the endplate are reactivated by 2-PAM and thus are ACHE. BuChE constitutes another 10% (Rogers et al., 1969) or possibly 30% (Barnard et al., 1971) of the DFP-binding sites. The remaining of DFP-binding sites (about 50%) are still of unknown nature. The constant ratio of AChE to all the other DFP-binding sites and their similar localization in normal animals suggest that the different DFP-binding sites at the endplate may be anatomically and functionally coupled. Our current finding that after esterase inactivation the various DFPbinding sites recover at different rates, indicates that at least during recovery these DFP-binding sites are not functionally coupled.

[2] 123w Our observation that after DFP inactivations BuChE recovers faster than AChE has also been reported for other inactivating agents such as Paraoxon (Davison, 1953;Welsch and Dettbarn, 1972). However, since the time course of recovery after Paraoxon is generally faster than after DFP it is not obvious that the mechanism involved is the same. We believe that the recovery after inactivation with DFP is due to de novo synthesis of the esterases. This conclusion is based primarily on recent preliminary tZindings (Kasprzak and Salpeter, unpublished) that after esterase inactivation with [3H] DFP, the time course of removal of radioactivity from the endplates is much faster than is the reappearance of new DFP reactive sites, and that actinomycin D can slow down the physiological recovery.

[3] 322w We had hoped that this study might shed light on the site of synthesis of the different DFP-binding sites. However, the search for the site of synthesis using EM autoradiography as the tool is complicated by the fact that the esterases reside in the basal lamina of the clefts (Hall and Kelly, 1971;Betz andSakmann, 1971, 1973;McMahan et al., 1978). Thus, whatever the cellular source, the final product has to be secreted, and may then migrate a considerable distance from its site of synthesis. Thus the proximity of developed grains to a given structure in an autoradiogram, although suggestive, cannot be unequivocal evidence for a cellular source. The earlier literature on the sites of esterase synthesis is contradictory. Some authors suggest local axonal synthesis (Koenig and Koelle, 1961;Koenig, 1965) others suggest transport by axoplasmic flow (for example, Ranish and Ochs, 1972), while still others implicate muscle synthesis (Lubinska, 1966;Koenig and Vigny, 1978). Schwann cell involvement was rejected by Koelle (1962). Our data show that during the early recovery periods, a relatively higher autoradiographic grain density and histochemical reaction product are located in the teloglial cap (Table 1), due predominantly to non-AChE sites. These observations suggest that the Schwann cell cannot be excluded from consideration as a source of at least the non-AChE sites. It has recently been suggested (G. B. Koelle et al., 1977;W. A. Koelle et al., 1977) that BuChE may function as a precursor in the synthesis of ACHE. This suggestion would be compatible with our observations of the earlier recovery of BuChE after DFP inactivation, and would further implicate the Schwann cell. However, since species differences are particularly common in esterase distributions (see reviews by Csillik, 1965;Barnard et al., 1971;Silver, 1974) and since Tennyson et al. (1973) find no BuChE during the development of rabbit muscle, verification of BuChE conversion to AChE is required for the mouse sternomastoid muscle before an unequivocal conclusion, applicable to our study, can be drawn.

[4] 66w If the muscle is also a source of new endplate esterases, then the experimental muscle necrosis may influence the time course of their recovery. Studies on the relative rate of recovery of different esterases under conditions which prevent myopathy (for example, o~-BTX pretreatment) may help identify the esterase species if any, which are derived from muscle, and may help resolve some of the questions raised above.

[5] 127w It is of interest to note that the ability to sustain tetanic contraction at stimulation frequencies of 100 Hz has already recovered to some extent at three days, (Fig. 7). Of five animals tested three days after inactivation, four could sustain a tetanic contraction for 2-5 s period, although that ability fatigued quickly with repeated stimulation and no tetanic response could be maintained at higher frequencies. By 14 days post-DFP, with 25-50% of the esterases recovered, the muscle shows difficulties only at about 300 Hz (Fig. 7), which is very similar to the normal response of this muscle, at least in the time scale of this test stimulus. Only with prolonged stimulation (about 30 s), did a significant difference between the normal and 14 day animal emerge.

[6] 124w What then is the function of the 'excess' of AChE in the sternomastoid muscle, which has a tetanic fusion frequency of <80 Hz? Normally, muscle contraction is a triggered event and, when AChE is fully active, the ACh life time is short relative to the normal contraction-relaxation cycle. We would like to argue that the high AChE concentration, and thus the rapid removal of ACh, is necessary not for repetitive activity on the ms time scale, but for normal nerve muscle interactions over a more prolonged time scale. The major action of normally functioning AChE is to restrict the action of ACh in space and time. The significance of this action is best seen in relation to the molecular organization of the neuromuscular iunction.

[7] 239w Based on the high AChR concentration at the postjunctional receptive surface, Fertuck and Salpeter (1976) and Matthews-Bellinger and Salpeter (1978) have argued that when AChE is fully active each quantum of ACh probably interacts with a small postjunctional area (about 0.3 ~m 2 ). From calculations of the surface area of the total postjunctional membrane, it follows that the sum of the postsynaptic areas activated by about 300 quanta of ACh released by a single nerve impulse constitutes <10% of the total postjunctional receptive surface. This model, therefore, suggests that, when AChE is fully active, the probability that the ACh quanta released by a second nerve impulse would activate exactly the same postjunctional areas is very low. This may explain why desensitization is not seen at neuromuscular junctions with repeated nerve stimulation at tetanic fusion frequency, even though each quantum is believed to act at very high local ACh concentration. As the AChE site density is lowered, however, the life time and action of ACh is prolonged (Eccles and McFarlane, 1949;Kuba and Tomita, 1971;Gage and McBurney, 1975) and ACh is expected to spread in the cleft to activate increasingly larger postiunctional areas (see also Katz and Miledi, 1973;Hartzel et al., 1975). The probability of partially activating the same areas by repeated stimulation therefore increases significantly, and some desensitization could occur. This may be the cause of the fatigue seen in the tetanic response during the stages of esterase recovery.

[8] 129w Another major consequence of AChE action is probably associated with the muscle damage seen after esterase inactivation. A neurotrophic etiology for this damage has been suggested by Dettbarn and his collaborators (for example, Fenichel et al., 1974;Laskowski et al., 1977;Wecker and Dettbarn, 1976), possibly mediated by Na + fluxes (Laskowski et al., 1977). The observations by Ariens et al (1969) that injecting curare ameliorates the necrosis, and our present observations that inactivating the AChR with c~-BTX prior to the DFP, completely eliminates even the earliest sign of muscle fine structural changes is consistent with the arguments that neither the DFP per se, nor the persistent presence of ACh, causes the damage. The development of muscle destruction depends on the events initiated by the functional interaction between ACh and AChR.

[9] 305w Based on our ultrastructural data, and on current physiological and biochemical observations reported in the literature we argue that not Na + (as suggested by Laskowski et al., 1977), but Ca 2+ may be the mediator of the detrimental effect of DFP on the muscle structure. Our suggestion is based on the facts that: (1) one of the earliest signs of the experimental necrosis produced by esterase inactivation is the destruction of the Z bands in the muscle fibres near the nerve terminal; (2) several studies have recently demonstrated that in skeletal muscle there is a proteolytic enzyme which is specific for Z band material (Reddy et al., 1975) and which is activated by elevated levels of Ca 2+ (Busch et al., 1972;Dayton et al., 1976;Reville et al., 1976); (3) membrane depolarization causes Ca 2+ influx and, when the depolarization is produced by agonist receptor interaction, the Ca 2+ influx is greatest in the region of the endplate (Jenkinson and Nicholls, 1961;Ahmad and Lewis, 1962;Csillik, 1965;Evans, 1974;Beaty and Stefani, 1976;Miledi et al., 1977). Recent in vitro studies further show that Ca 2+ is released from the AChR itself as a result of ACh binding (Chang and Neumann, 1976;Rubsamen et al., 1976) thus further adding to the Ca 2+ fluxes in the region of the endplate. Finally, the sarcoplasmic reticulum (SR) is a major site of Ca 2+ binding (see review by Martonosi, 1972). If the vesiculated sacroplasmic reticulum that we see during the period of early necrosis is indicative of disruption of the reticulum this may further enhance the elevation of free Ca 2+ levels in the cytoplasm. The early fine structural changes after esterase inactivation are compatible with the idea that the Ca 2+ fluxes during the resultant prolonged agonist-receptor interactions are sufficient to initiate Z band destruction which subsequently triggers the muscle damage.

[10] 86w The AChE at the neuromuscular junction normally turns over ACh much faster than is needed merely to prevent a buildup of ACh at fusion frequencies. The apparently 'excessive' AChE may in part, however, have a protective and regulative function through its control of free ACh. By restricting the spread of ACh over the post-synaptic membrane, AChE may prevent desensitization and fatigue and, by restricting the lifetime of ACh in the cleft, it may limit Ca 2+ influxes thus controlling the structural integrity of the muscle cell.

[1] 119w The sternomastoid muscles were first prefixed with (1) 4% paraformaldehyde or (2) 1.5% glutaraldehyde in 0.06 M phosphate buffer either by whole body perfusion or by topical application onto the exposed muscle, followed by additional h'nmersion fixation for 2 h in the same fixative. Perfusion with 4% paraformaldehyde was found to give the best fixation. The tissue was stored in 'rinse buffer' (0.06 M phosphate buffer + 6% sucrose, pH 7.4) overnight, then postfixed in 1% OsO4, block stained in 2% uranyl acetate and embedded in Epon 812, (Luft, 1961). Light and EM studies were made of such preparations at each of the following postinactivation times: O, 1, 3, 7 and 14 days, using 2-3 animals for each time.

[2] 422w Radioactive labelling of all DFP reactive sites or specifically of AChE sites was performed as previously described (Rogers et al., 1966(Rogers et al., , 1969;;Salpeter, 1967Salpeter, , 1969;;Salpeter et al., 1972Salpeter et al., , 1978)). Sternomastoid muscles from 1-3 animals per time period were fixed for 2 h as above on 0, 3, 7, and 14 days after inactivation by DFP in vivo. Each muscle was divided into two pieces. The first was incubated in [3H] DFP (10 -4 M in rinse buffer pH 7.4) for 1 h at room temperature, thus labelling all DFP reactive sites. Since DFP is not specific for AChE but phosphorylates other esterase-like sites, the second group was treated with pyridine 2-aldoxime methiodide (2-PAM), a specific reactivator of phosphorylated ACHE, (Wilson et al., 1958;Wilson and Froede, 1971), so as to distinguish AChE from other DFP-reactive sites (for example, BuChE, aliesterase). This was done by one of the two following incubation sequences: (a) incubating in [3H] DFP, followed by 2-PAM, (10-3) M in rinse buffer at pH 7.9, at room temperature for 40 min), thus labelling all DFP-reaetive sites except ACHE; or (b) incubating in non-radioactive DFP then in 2-PAM to reactivate ACHE, followed by [3 H] DFP, thus labelling only ACHE. Ageing of phosphorylated sites prior to 2-PAM treatment was avoided by keeping the washing period between DFP and 2-PAM reactivation to less than 30 rain. The assumption that 2-PAM adequately reactivated all the AChE sites was validated by Rogers et al. (1966Rogers et al. ( , 1969) ) who showed that the same fraction of DFP-binding sites was reactivated by 2-PAM as was protected from DFP labelling by the specific AChE inhibitor BW 284C51 (Austin and Berry, 1953). In all the incubation sequences, the tissue was always washed first in buffer (three changes for a total of 15 rain) and then in several washes of non-radioactive DFP (10-3 M for 2 x 5 min and 10-4 M for 3 x 20 rain) to remove non-specific binding. After completing the full incubation sequence the tissue was postfixed in 1% i3sO4, stained in 2% aqueous uranyl acetate, and embedded in Epon 812. Pale gold sections were cut and prepared for autoradiography by the 'flat substrate' method of Salpeter andBaehmann (1964, 1972), using a closely packed monolayer of Ilford L4 emulsion (purple to blue interference eolour). Interferometric measurements monitored and thickness of the sections. After about 3 months exposure, the slides were developed in a gold-latensification Elon ascorbie acid sodium sulphate developer (GEAS) described by Salpeter and Szabo (1972).

[3] 151w Two approaches to the analysis of the autoradiograms were employed. The first involved a general survey of the radioactivity in different regions of the tissue. The endplate was divided into two regions: (1) the teloglial cap (Couteaux, 1972), which consisted of the Schwann cell cytoplasm and a zone of connective tissue about 2 #m wide around the top of the axon terminal and its surrounding Schwann cell; and (2) the synaptic zone, which consisted of the axon terminal, the primary and secondary clefts and a zone 2 gm wide on the muscle side of the postjunctional membrane (PJM). For each endplate, all the grains seen in these regions were counted. A grid (calibrated to relate grid point intersections to ~m 2) was used to determine the areas of the regions tabulated. A background correction was made by subtracting grains found over an area of similar size ;>4 #m from an endplate.

[4] 282w The second analysis was as previously described (Salpeter, 1967(Salpeter, , 1969;;Salpeter et al., 1972Salpeter et al., , 1978)). This involved first counting all grains in an area of emulsion expected to be exposed to radiation emanating from a radioactive line source coincident with the PJM. (The expected distribution was calculated from the resolution curves of Salpeter et al., 1969.) These grains, after background corrections, were then divided by the measured surface area of the PJM in the autoradiographs. The grain density was then used to calculate the sites per/~m 2 of PJM surface area using the formula given previously (Salpeter, 1969;Salpeter et al., 1972). The accuracy of each final calculated site density is limited in the autoradiographic technique by the following factors: (a) variations in section thickness (which we measured by the interferometer to be <10%). (b) Fluctuations in autoradiographic sensitivity (which we have established over many years to be <~20% if we recalibrate the sensitivity of each emulsion batch and include a sensitivity standard with each experiment). (c) Percent fluctuations in the grain counts which, by Poisson statistics is (1/~/N) x 100% where N is the number of grains counted. (Since for each animal in the present study the site density is based on 200-300 grains, grain sampling introduces an error of <10%.) (d) Our measurement of the membrane lengths were compared with determinations using computer graphics and a map measurer and was found to be within +10%. Since the final accuracy of any value is the square root of the sum of the squares of all the independent errors, each of our calculated site densities per animal is accurate to about +25%. Systematic errors are not included in this evaluation.

[5] 51w In another group of animals, sternomastoid muscles were removed, fixed in 10% formalin, and stained for cholinesterase by the procedure of Karnovsky and Roots (1964), using acetylthiocholine as the substrate. Since this substrate is not completely specific for AChE but will be hydrolysed also by BuChE, we included the following controls.

[6] 46w (1) Prior to staining, some tissue was incubated either in tetra-isopropylpyrophosphoramide (iso-OMPA) to inhibit specifically BuChE (Austin and Berry, 1953), or in DFP followed by 2-PAM, to reactivate the ACHE. Both of these preincubation procedures were thus designed to stain AChE as distinct from other cholinesterases.

[7] 15w (2) Some tissue was stained using butyrylthiocholine as substrate in order to stain selectively BuChE.

[8] 27w The amount of staining was assessed qualitatively by light microscopy 0, 3, 7, 14 and 21 days after DFP. Several preparations were also examined in the EM.

[9] 238w The possible cause of the myopathy produced by the in vivo DFP treatment was investigated in several groups of two animals each as follows. (a) Animals were subjected to 'sham operations'. The sternomastoid muscle was exposed and bathed in Krebs Ringer (but without DFP) while the nerve was stimulated periodically over the 1.5-2 h period that is normally involved in the esterase inactivation procedure. (b) The nerve to the sternomastoid muscle was cut immediately after the DFP incubation. (c) AChE was reactivated in vivo after DFP treatment by the application of 2-PAM (10 -3 M in 0.06 M phosphate buffer pH 7.9) for 15--20min which was sufficient to fully restore the tetanic response. (d) The muscle of the in vivo preparation was bathed with topically applied a-bungarotoxin (0e-BTX) (10 -6 M in Krebs Ringer), a specific inhibitor of the acetylcholine receptor (AChR) (Chang and Lee, 1963). The 0t-BTX was applied as previously described (Fertuck et al., 1975) just prior to, or just after the esterase inactivation by DFP. When a-BTX was used prior to the DFP, the muscle no longer responded to nerve stimulation and we could not use our standard criterion for the endpoint of in vivo esterase inactivation by DFP, that is, the inability to sustain tetanus. In that case DFP was applied to the paralysed muscle for 1.5-2 h, that is, the time found to be adequate for complete inactivation of esterases in normal preparations.

[10] 41w In all of the four control groups described above, the incisions were subsequently sutured and the animals allowed to survive for 1-3 days, at which time the myopathy had been found to be at its peak in the DFP-treated experimental animals.