Is the MEPP due to the release of one vesicle or to the simultaneous release of several vesicles at one active zone?

Regulation of quantal shape by Rab3A: evidence for a fusion pore-dependent mechanism

The function of Rab3A, a small GTPase located on synaptic vesicles, is not well understood. Studies in the Rab3A(-/-) mouse support a role in activity-dependent plasticity, but have not reported any effects on spontaneously occurring miniature synaptic currents, except that there is a decrease in resting frequency at the neuromuscular junction. Therefore we were surprised to find an increase in the occurrence of mEPCs with abnormally long half-widths at the neuromuscular junctions of Rab3A(-/-) mice. The abnormal miniature endplate currents (mEPCs), which have significantly greater charge than the average mEPCs for the same fibres, could arise from larger vesicles. However, the type of mEPC most increased in Rab3A(-/-) mice has a slow rise, which suggests it is not the result of full collapse fusion. To test if the slow mEPCs increased after loss of Rab3A could be due to malfunctioning fusion pores, we used carbon fibre amperometry to record pre-spike feet, which have been shown to correspond to the initial opening of a narrow fusion pore, in adrenal chromaffin cells of wild-type and Rab3A(-/-) mice. We found that small amplitude pre-spike feet with abnormally long durations were increased in Rab3A(-/-) cells. The correspondence between mEPC and amperometric data supports our interpretation that slow rising, long half-width mEPCs are caused by reduced diameter fusion pores that remain open longer. These data could be explained by a direct action of Rab3A on the fusion pore, or by Rab3A-dependent control of vesicles with unusual fusion pore characteristics.

The heterogeneity of vesicular acetylcholine storage in cholinergic nerve terminals

The vesicular hypothesis of quantal acetylcholine release describes the process by which discrete packages (or quanta) of the transmitter are released from nerve terminals through the exocytosis of the content of synaptic vesicles. However, cholinergic synaptic vesicles can no longer be vaguely regarded as simple membrane bound 'sacks' of the transmitter. Modern molecular, biochemical, morphological and electrophysiological research has revealed them to be complex cellular structures with a heterogeneous mixture of functions. Thus, not all synaptic vesicle populations are formed under the same circumstances and there are variations in the releasability of synaptic vesicle populations. This review briefly outlines some of the experimental research that has lead to our current thinking on the heterogeneity of vesicular acetylcholine storage in cholinergic nerve terminals. In addition, a model for vesicular acetylcholine storage and release is presented that attempts to accommodate many of the modern ideas concerning cholinergic synaptic vesicle function and interaction.

Spontaneous synaptic potentials from afferent terminals in the guinea pig cochlea

Records of spontaneous activity from units likely to be radial afferents were analyzed to find the origin of spontaneous action potentials in single auditory nerve fibers. Single synaptic events (excitatory postsynaptic potentials or EPSPs) nearly all triggered action potentials (spikes). An abrupt increase in slope during the rising phase of the EPSP often signalled the initiation of an action potential. Synaptic potentials that did not trigger spikes occurred frequently during the refractory period. These events sometimes appeared to be composed of subunits. Refractoriness appears to be the primary reason these EPSPs were ineffective. Distributions of the onset slopes of postsynaptic potentials were highly skewed. Skewing was not a consequence of refractoriness, but most likely because the amplitude distribution of spontaneous potentials is not gaussian.

The regulation of quantal size

Quantal size can be altered experimentally by numerous treatments that seem to lack any common thread. The observations may seem haphazard and senseless unless clear distinctions are made from the outset. Some treatments shift the size of the entire population of quanta. These quanta are released by nerve stimulation. Other treatments add quanta of abnormal size or shape--monstrosities--to the population (4.0). Usually, perhaps even invariably, the monstrosities are not released by nerve stimulation. 6.1. POPULATION SIZE INCREASES. 6.1.1. Quantal size must be regulated. The size of the entire quantal population can be experimentally shifted to a larger size, with the mean rising two- or even four-fold. Before these observations, it was reasonable to suppose that quantal size was relatively fixed, with little room for maneuver. A logical picture is that synaptic vesicles have a maximum transmitter capacity, and usually they are filled to the brim. This picture is wrong. The quantity of transmitter packaged in the quantum must be regulated by the neuron, so depending on circumstances, quantal size can be increased or decreased. Figure 18 makes the case for regulation more strongly than words. We are beginning to identify some of the signals for up and down regulation, and the first steps have been made in discovering the signal transduction pathways, but we are far from a true understanding. This is hardly surprising, because our information about how transmitter molecules are assembled into quantal packages is still imperfect. Until we understand the engine, it may be difficult to picture the accelerator or the brake. 6.1.2. Signals that up regulate size. Stimulation of the presynaptic neuron increases quantal size at the NMJ, at synapses in autonomic ganglia and in hippocampus. The stimulus parameters necessary to elicit the quantal size increase have not been explored sufficiently in any of these cases, and all deserve further investigation. At both frog and mouse NMJs quantal size is roughly doubled following exposure to hypertonic solutions, which elevate the rate of spontaneous quantal release. This discovery, coupled with the increases caused by tetanic stimulation, suggested that the signal for up regulation is a period of greatly enhanced quantal output. The size increase takes about 15 min in hypertonic solution in mouse and about 60 min in frog. Highly hypertonic solutions do not increase the rate of quantal release in frog; they also do not increase quantal size. This supported the idea that quantal release rate is the signal for up regulation.(ABSTRACT TRUNCATED AT 400 WORDS)

Correlation between quantal secretion and vesicle loss at the frog neuromuscular junction

1. We measured the rate of occurrence of miniature endplate potentials (MEPPs) at identified endplates in frog cutaneous pectoris muscles treated with crude black widow spider venom (BWSV) or purified alpha-latrotoxin (alpha-LTX) in calcium-free solutions, and we examined the relationship between the length of the nerve terminal and the total number of quanta secreted, and the relationship between the number of quanta secreted and the number of vesicles remaining at different times. 2. The venom, or toxin, was applied in a modified Ringer solution with tetrodotoxin, 1 mM-EGTA and no divalent cations, and quantal secretion was started by applying Ca2(+)-free solutions with Mg2+. This was done to synchronize the quantal discharge at the various junctions in a muscle. Ringer solution was applied after the MEPP rate had declined to low levels, and then the muscle fibre was injected with Lucifer Yellow, the endplate stained for acetylcholinesterase and the length of the nerve terminal and the length of a sarcomere were measured on the fluorescent fibre. 3. The total number of quanta secreted by a terminal was measured under a wide variety of experimental conditions: the weights of the frogs ranged from 13 to 68 g, the temperature from 9 to 28 degrees C, and the concentration of Mg2+ from 2 to 10 mM. In one series of experiments the Mg2+ was withdrawn after 3-4 min and reapplied 35-40 min later in order to divide the total output of quanta into two approximately equal bouts of secretion that were well separated in time. 4. The total number of MEPPs recorded at a junction was loosely correlated with the length of its nerve terminal, but it was not affected by the temperature, the concentration of Mg2+ or the division of secretion into well-separated bouts of quantal release. The average total secretion per unit length was about 3700 quanta/sarcomere or about 1200 quanta/microns. 5. The average time course of quantal secretion per micrometre of terminal was determined at single junctions in muscles held at 22-23 degrees C or at 9-10 degrees C. Other muscles were fixed at various times during the course of secretion at each temperature and the number of synaptic vesicles remaining in cross-sections of the terminals were counted on electron micrographs. The number of vesicles remaining per micrometre of terminal was determined from the number per cross-section and the section thickness.(ABSTRACT TRUNCATED AT 400 WORDS)

Statistical and graphical methods for testing the hypothesis that quanta are made up of subunits

It has been proposed that at the neuromuscular junction quanta are made up of subunits, because regularly spaced peaks appear in some histograms of miniature end-plate potential (MEPP) amplitudes. Many sets of MEPP sizes fit lognormal probability distribution functions. I compared sets of MEPP amplitudes from the literature to lognormal distributions, using cumulative plots and a robust test for significance. Some data sets fit a single lognormal distribution, but most required two lognormal subpopulations. This bimodal model is a parsimonious alternative to the subunits hypothesis. I also used graphs in which cumulative MEPP sizes were plotted on the ordinate against probability on the abscissa. Model data sets, generated on the subunit assumptions, produced concave plots. End-plate potential sizes at low quantal outputs--in which we know there are subunits--show the same concave shape. Most of the tested data from the subunit literature produced convex curves.

Proximodistal gradients of the postjunctional folds at the frog neuromuscular junction: a scanning electron microscopic study

Frog endplates were studied with the scanning electron microscope following the removal of the presynaptic terminal by collagenase and acid treatments. Endplates had 2-14 branches of primary cleft. The longest branches were parallel to the muscle fiber. Short branches oblique or perpendicular to the muscle fiber were also present near the central region of the endplates. The openings of postjunctional folds in the primary cleft were clearly visible at the bottom of the primary cleft and could be counted and measured. The longest primary cleft branches of each endplate were divided into segments of 20 microns (length corrected for shrinkage). The number of postjunctional folds per micrometer of primary cleft, the average postjunctional fold length (i.e. across the primary cleft) and the total postjunctional fold's length per micrometer of primary cleft were evaluated for each 20-microns segment of primary cleft. Negative proximodistal gradients were observed for these three parameters for the long branches of primary cleft, i.e. values were higher in the proximal region (near the motor axon) than in the distal region. These postsynaptic gradients probably reflect similar or smaller proximodistal presynaptic gradients for the active zones along the nerve.

Reversible effect of depolarization by K-propionate on sub-miniature endplate potential to bell-miniature endplate potential ratios, on miniature endplate potential frequencies and amplitudes, and on synaptic vesicle diameters and densities in frog neuromuscular junctions

Miniature endplate potentials were recorded from edge muscle fibers of frog sartorius muscles during high frequencies induced with K-propionate and during recovery. The identified neuromuscular junctions were studied with the electron microscope and their ultrastructure was correlated with amplitude and numbers of miniature endplate potentials generated. Miniature endplate potential amplitudes were maintained during the first 10 min of depolarization. They then decreased during the next 2-3 h until the mode was lost to the noise. Miniature endplate potential frequency was greatly increased during the first hour and there was initial depletion of vesicles. Miniature endplate potential frequencies remained high (5 x 10(5)/h) for 3 h but vesicle densities returned to nearly normal values during the second to third hour of treatment. The conspicuous infolding of the presynaptic membrane noted during the first hour of treatment suggests that recycling of vesicles is initially slower than fusion. Calculated recycling time is shorter than 25 min. During recovery after prolonged K-propionate treatment, the sub-miniature endplate potential class reappeared within minutes but about 20 min were required before it returned to control size. Subsequently, the bell-miniature endplate potentials reappeared and slowly increased in amplitude. The ultrastructure returned to a normal state. There was no change in vesicle diameters. No significant difference was found between the diameters of "touching vesicles" (vesicles touching the presynaptic membrane) and the non-touching vesicles. By comparison, lanthanum ions (1 mM) released a smaller number of quanta which did not exceed the number of vesicles present at the start of the experiment. Variations of the subunit hypothesis of the quantum of transmitter release are discussed.

Effects of an inhibitor of the synaptic vesicle acetylcholine transport system on quantal neurotransmitter release: an electrophysiological study

The drug 2-(4-phenylpiperidino)cyclohexanol (AH5183), which potently inhibits the active transport of acetylcholine (ACh) into synaptic vesicles, was used as a pharmacological tool to study the functional role of synaptic vesicles in quantal transmitter release. Using microelectrode recording techniques, miniature endplate potentials (mepps) and nerve-evoked endplate potentials (epps) were recorded from frog cutaneous pectoris neuromuscular junctions in low Ca2+/high Mg2+ Ringer solution, and in normal Ringer with added D-tubocurarine (D-TC). Stimulation in the presence of AH5183 caused a 40% reduction in quantal size (mepp amplitude), depressed tetanic potentiation, and decreased the number of quanta released with each nerve impulse in the presence of D-TC. All of these effects appeared gradually and only after extended stimulation of the nerve, during which several hundred thousand quanta were released. Consequently, these findings suggest a serial one-time usage of vesicles, with little or no re-entry of recycled vesicles until after a large fraction of the original vesicles has been exhausted. The results primarily show that filling of synaptic vesicles with ACh is crucial for sustaining synaptic transmission, and gives further evidence that the ACh released by nerve impulses originates from these organelles.

Synaptic vesicle relationships with the presynaptic membrane as shown by a new method of fast chemical fixation

Brief vascular perfusion of the rat brain with a mixture of concentrated aldehydes completely insolubilized the brain protein in less than 30 s and yielded excellent ultrastructural preservation. Abundant synaptic vesicles closely and specifically attached to the presynaptic membrane were constantly detected. These vesicles appeared to undergo progressive transformation into amorphous, electron-dense material. No evidence of vesicle exocytosis was detected in the brains perfused in vivo but fixations performed 1 h after death showed abundant exocytotic-like images. The results suggest that the vesicles may not be exocytotically released to the intersynaptic cleft but disintegrate intracytoplasmically in the immediate vicinity of the presynaptic membrane.

Description of the sub-miniature endplate potential distribution, determination of subunit size and number of subunits in the adult frog neuromuscular bell-miniature endplate potential

Miniature endplate potentials were recorded from the isolated pectoralis cutaneous muscle of the frog during very stable recording conditions and low noise levels. Two to 5 x 10(3) miniature endplate potentials were filmed at eight unstressed junctions that met rigorous experimental criteria for analysis. Seven junctions generated enough sub-miniature endplate potentials (1-3%) to produce a bell-shaped amplitude distribution. The sub-miniature endplate potential means were usually 9-10 times smaller than the modal bell-miniature endplate potential values. The standard deviation of the sub-miniature endplate potential class was calculated by subtracting the noise and measurement error from the measured sub-miniature endplate potential distribution. The coefficients of variation of the sub-miniature endplate potential distribution were 9-16%. Half of the bell-miniature endplate potential amplitude distributions showed 4-6 integral peaks in the central part of the distributions and the positions of these peaks were maintained with increasing sample size. The remaining distributions were not smooth and suggested integral peaks. The intervals between the peaks were about the same size as the sub-miniature endplate potential mode. These data provide further evidence for the subunit hypothesis of the quantum of transmitter release and describe the amplitude distribution of the sub-miniature endplate potential class of the adult preparation.

Effect of hypertonic saline on quantal size and synaptic vesicles in identified neuromuscular junction of the frog

Miniature endplate potential amplitude distributions, miniature endplate potential frequencies and the percentage of sub-miniature endplate potentials were studied during treatment with hypertonic saline (with sucrose) during the initial high frequencies of release and after fatigue. Small muscle fibers were selected which had normal miniature endplate potential frequencies of 0.1/s to 1/s so that the miniature endplate potential amplitude distributions could be determined at the height of the hypertonic effect (first 5-15 min) at which time the miniature endplate potential frequency increased two-hundredfold. During the first few minutes of the effect, there was little change in miniature endplate potential amplitude or in the profiles of their amplitude histograms. Subsequently, after the occurrence of as few as 10(4) miniature endplate potentials, the size of the mean bell-miniature endplate potentials decreased. Later (25 min) the amplitude profiles became uniform, and finally (45 min) the percentage of sub-miniature endplate potentials and smaller miniature endplate potentials increased until many miniature endplate potentials (30-70%) were of the sub-miniature endplate potential class and the overall distributions were skewed. The mean sub-miniature endplate potential amplitude did not appear to change. After the initial high frequency of release, many miniature endplate potentials showed a definite break on the rising phase and the amplitude of the break was usually that of the sub-miniature endplate potential. The rapid decrease in miniature endplate potential size, change in miniature endplate potential amplitude profile and breaks on the miniature endplate potential rising phase can be explained with the subunit hypothesis. The edge fibers of the sartorius muscle were used so that physiologically studied edge junctions that were producing various miniature endplate potential histograms could be identified for electron microscopy. Synaptic vesicle diameters and the coefficient of variation of vesicle diameters were not changed either during high miniature endplate potential frequencies or in those junctions that generated mainly sub-miniature endplate potentials. Thus, the quantal class (i.e. sub-miniature endplate potential or bell-miniature endplate potential) cannot be determined from the vesicle diameter.

Time course and frequency dependence of synaptic vesicle depletion and recovery in electrically stimulated sympathetic ganglia

The mammalian superior cervical sympathetic ganglion has been extensively used to study the kinetics of ACh metabolism and release. The present investigation examined the time course of changes in the number of synaptic vesicles and abundance of plasma membrane at preganglionic nerve terminals using stimulation protocols similar to those used in previous biochemical and electrophysiological studies. Continuous stimulation of the preganglionic trunk to the cat superior cervical ganglion in vivo produced an initially rapid fall in the number of clear synaptic vesicles followed by a subsequent plateau. Reciprocal changes in plasma membrane occurred with a similar time course. The plateau phase is interpreted as a steady-state where vesicle exocytosis is balanced by the rate of vesicle reformation from plasma membrane. During quiescent recovery, restoration of normal resting ultrastructure is initially rapid but slows with time as vesicle number and plasma membrane abundance approach pre-stimulation values, indicating that the rate of vesicle reformation at the end of stimulation is high and proportional to the number of vesicles incorporated into the plasma membrane. These results are interpreted as consistent with the 'vesicle hypothesis' of neurotransmitter release.

Interrelation between MEPP amplitude and MEPP frequency in different regions along the frog neuromuscular junction

Experiments were done to study the relationship between miniature endplate potential (MEPP) frequency and MEPP amplitude in different regions along the frog neuromuscular junction (NMJ). The position of the release site producing a MEPP and the amplitude of a MEPP at its release site were evaluated by using the spatial decay method after simultaneous intracellular recordings at each distal end of the NMJ. The NMJ was divided in 10 regions of equal length. It was found that MEPP amplitude and MEPP frequency are smaller in distal regions than in proximal ones. Moreover, we observed a logarithmic relationship between MEPP frequency of a given region and the mean MEPP amplitude of the same region. These results suggest that the probability to produce a MEPP and the quantity of neurotransmitter liberated are two synaptic functions controlled or affected by common mechanisms.

Spontaneous quantal and subquantal transmitter release at the Torpedo nerve-electroplaque junction

Focal electrodes were used to record the spontaneous miniature potentials generated on delimited patches of innervated membrane in the Torpedo electric organ. The main population of miniature potentials followed a bell-shaped amplitude distribution. In addition, we observed a second class of spontaneous events that were smaller and whose amplitude distribution was skewed. These subminiatures formed an homogenous population together with the regular miniatures with respect to their time course versus amplitude relationship. They were thus probably generated at the same sites. The proportion of potentials that were subminiature was less than 10% in resting, freshly excised tissue, but it increased markedly: (i) when the tissue was kept for 24-28 h in vitro after excision; (ii) in the period following a brief heat challenger or (iii) stimulation to exhaustion; and (iv) in the presence of dinitrophenol or dinitrofluorobenzene. In all these conditions, we measured the acetylcholine, adenosine 5'-triphosphate and creatine phosphate content of the tissue and found a correlation between the relative number of subminiature potentials and the lack of energy rich molecules. It is concluded that subminiature potentials are present in the electric organ as in neuromuscular junctions. They are probably produced at the same sites as the regular miniature potentials and their relative occurrence seems to increase greatly when the nerve terminals are in a state of energy deficiency.

Non-uniform release at the frog neuromuscular junction: evidence of morphological and physiological plasticity

The frog neuromuscular junction (NMJ) is a fusiform structure parallel to the muscle fiber with a few secondary and tertiary branches. Both sprouting and regression can occur on the same nerve terminal, suggesting a continuous on-going remodelling of the mature neuromuscular junction. Thus, the frog NMJ is a dynamic structure. Ultrastructural observations of the nerve terminal suggest that the active zones are distributed equally along the mature nerve terminal. Disorganized active zones have however been observed in distal regions. The density of synaptic vesicles is also uniform throughout the whole structure. However, mitochondria appear to be more abundant in the very distal regions of the nerve terminal. The postjunctional folds and the cholinergic receptors are also uniformly distributed along the NMJ. However, during remodelling periods, the distributions of postjunctional folds and of cholinergic receptors are not uniform in the degenerating and regenerating regions. Fig. 1 summarizes these morphological data. The frequency of spontaneous release (MEPPs) at the NMJ is higher in the proximal region than in the distal regions and recent evidence suggests that the mean MEPP amplitude is higher in the proximal than in the distal portions. Evoked transmitter release is also non-uniform along the frog NMJ. As for spontaneous release, it is higher in the proximal regions than in the distal regions. Failures of the active propagation of the PNAP at low safety points, such as the end of the myelinated axon and the branching points, may be one of the mechanisms responsible for unequal evoked release. It is also possible that the PNAP does not actively invade the whole extend of the nerve terminal since Na+ channels are absent from the distal regions. Fig. 2 summarizes these physiological data.

Quantal currents evoked by graded intracellular depolarization of crayfish motor axon terminals

1. Quantal transmitter release was examined at nerve terminals of the excitatory motor axon of the crayfish opener muscle. The magnitude of synaptic currents, recorded with macro-patch electrodes at a nerve terminal, served as a measure of quantal size. Transmitter release was initiated by pulses of depolarizing current applied intracellularly to the axonal terminals after application of tetrodotoxin. Quantal release was altered by a variety of methods and the resulting quantal output and quantal size were measured. 2. Amplitude distributions of quantal events were obtained during experimental manipulations which altered the rate of quantal release by up to 25-fold. These manipulations consisted of: varying pulse amplitude or pulse duration; facilitating the release by prolonged depolarization; and application of a potassium channel blocker, 4-aminopyridine. 3. The amplitude of quantal events is impervious to marked changes in presynaptic depolarization and is not affected by experimental procedures which promote accumulation of calcium ions in the terminals. The vesicular mechanism of release, in which transmitter substance is prepackaged in vesicles which individually undergo exocytosis at a release zone, could account for the observed results.

Calcium-insensitive miniature endplate potentials at the neuromuscular junction

Several different types of acetylcholine secretion have been shown to coexist at the neuromuscular junction along with the Ca2+-dependent quantal release producing miniature endplate potentials (mepps) and endplate potentials. One of these, the Ca2+-insensitive, slow-rising mepps (slow mepps), is present in normal untreated muscles but is most prominent in many conditions where the Ca2+-dependent quantal release mechanism is not functioning properly. Slow mepps occur at a frequency of less than 0.1 Hz in normal muscles, with large variability between fibres and muscles, and can reach frequencies of 1-2 Hz in several pathological conditions. The potentials are also highly variable in size and shape, being generally of high amplitude (0.1-15 mV) and prolonged time course (1-15 ms rise time). Most importantly, slow mepps are not affected by procedures which increase the intraterminal Ca2+ concentration, including nerve stimulation, thus being unable to contribute to the function of synaptic transmission. The cellular source of the Ca2+-insensitive mepps has been determined to be the nerve terminal and not the Schwann cells or nerve sprouts. The release process producing slow mepps is generally insensitive to many drugs, ions, and procedures, stimulation being observed with vinblastine, cytochalasin B, and caffeine. Depression of this secretion is effected by uncouplers of oxidative phosphorylation and by a drug (AH5183) which inhibits the vesicular active acetylcholine transport system. It is concluded that the slow mepps are due to an exocytic fusion of unique synaptic vesicles with the plasma membrane near the active zones, in a process insensitive to many intracellular ions and regulators. Since slow mepps are prominent in many pathological conditions of nerve and muscle, it is speculated that they play some role in the recovery or development of synaptic function.

Effects of excitatory amino acid antagonists on evoked and spontaneous excitatory potentials in guinea-pig hippocampus

Evoked and spontaneous excitatory post-synaptic potentials (e.p.s.p.s) at the mossy fibre input to CA3 pyramidal neurones were recorded intracellularly in slices from the guinea-pig hippocampus. The effects of several amino acid antagonists on these responses were examined. L-2-amino-4-phosphonobutyrate (L-AP4), L-serine-O-phosphate (L-SOP), kynurenate, and N-(p-bromobenzoyl)piperazine-2,3-dicarboxylate (pBB-PzDA) reduced the amplitude of evoked mossy fibre e.p.s.p.s without affecting membrane potential or input resistance. Antagonism of mossy fibre spontaneous miniature e.p.s.p.s (m.e.p.s.p.s) by these compounds fell into two groups. L-AP4 and L-SOP applied at concentrations that blocked evoked e.p.s.p.s did not affect amplitude distributions of spontaneous m.e.p.s.p.s. Kynurenate and pBB-PzDA significantly affected the amplitude distributions and reduced the mean amplitude of spontaneous m.e.p.s.p.s. These results are consistent with a presynaptic site of action for L-AP4 and L-SOP and a post-synaptic site of action for kynurenate and pBB-PzDA as antagonists of e.p.s.p.s at the guinea-pig mossy fibre-CA3 pyramidal neurone synapse.

Miniature endplate potential amplitudes corrected for spatial decay are not normally distributed

Miniature endplate potentials (MEPPs) were recorded simultaneously with two intracellular electrodes placed in the muscle fiber near the ends of the frog neuromuscular junction (nmj). The MEPP amplitudes are different in each electrode and are not normally distributed. A method is proposed to correct MEPP amplitudes for spatial decay. The MEPP amplitudes corrected for spatial decay are also not normally distributed. When the class interval of the distribution is small, multiple peaks are observed.