UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of [alpha]-bungarotoxin treatment on murine neuromuscular transmission Katul, Ziad J. 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1997-0250.pdf [ 6.18MB ]
Metadata
JSON: 831-1.0087706.json
JSON-LD: 831-1.0087706-ld.json
RDF/XML (Pretty): 831-1.0087706-rdf.xml
RDF/JSON: 831-1.0087706-rdf.json
Turtle: 831-1.0087706-turtle.txt
N-Triples: 831-1.0087706-rdf-ntriples.txt
Original Record: 831-1.0087706-source.json
Full Text
831-1.0087706-fulltext.txt
Citation
831-1.0087706.ris

Full Text

EFFECTS OF CHRONIC a-BUNGAROTOXIN TREATMENT ON MURINE NEUROMUSCULAR TRANSMISSION by ZIAD J. KATUL B.Sc., Denison University, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Pharmacology & Therapeutics Faculty of Medicine We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 30, 1997 © Ziad J. Katul, 1997 In present ing- this thesis in partial fu l f i lment of t h e requ i remen ts f o r an advanced degree at the Universi ty of British C o l u m b i a , I agree that t h e Library shall make it f reely available fo r re ference and study. I fu r ther agree that permiss ion fo r extens ive c o p y i n g of this thesis f o r scholar ly pu rposes may be g ran ted by the head o f m y d e p a r t m e n t or by his o r her representat ives. It is u n d e r s t o o d that c o p y i n g o r pub l i ca t ion of this thesis fo r f inancial gain shall n o t be a l l o w e d w i t h o u t m y w r i t t e n permiss ion . D e p a r t m e n t of ^ U ^ v ^ ^ < ^ U ^ ^ ^ ^ \V^<BJ^<ei\^\\(LS The Univers i ty o f Brit ish C o l u m b i a Vancouver , Canada Pa te 0 X 4 ^ 3 0 / / DE-6 (2/88) - ii -A B S T R A C T The development and growth of the mammalian neuromuscular junction (NMJ) is orchestrated by complex antero- and retrograde signals exchanged between the nerve terminal and the muscle. The adult NMJ has been shown to display morphological plasticity manifest under conditions of increased or decreased muscle activity, denervation (surgical and pharmacologic) and aging, as part of a continuum of responses to stresses imposed on neuromuscular transmission. Therefore, it seemed plausible to hypothesize that a reduction in postsynaptic sensitivity to ACh with consequent reduced muscle fiber activity would elicit functional changes (an increase or decrease in ACh release) mediated by anterograde and muscle-derived retrograde factors. Using mice, I produced a chronic reduction of functional AChRs by a 3-week, subparalytic treatment with the irreversible AChR antagonist, ot-bungarotoxin. Dose of toxin was adjusted based on the weight and strength of the animals. Strength was assessed by placing the mouse on a rotating rod and monitoring the time it was capable of sustaining itself without falling. Treated animals exhibited a marked reduction in their ability to stay 'roto-rod'. Subsequent electrophysiological experiments were performed on the isolated phrenic nerve/hemidiaphragm preparations to record both spontaneous and evoked ACh release under a variety of Ca and K concentrations and stimulation protocols (30 s 10 Hz period followed by 60 trains-of-ten at 70 Hz at each junction); The chronic treatment with a-bungarotoxin caused an increase in spontaneous and evoked ACh release detected by intracellular microelectrode recordings from the muscle fiber cell. However, the log normal distribution of release was unaltered and neither was its average facilitation in response to train stimulation. Train-evoked release relative to antecedent 10 Hz quantal content or by mean train quantal content was the same in both animal groups but of - I l l -greater variability in controls. The variance associated with release was found to be similar in both control and treated animals. In summary, I have provided evidence that a chronic reduction in muscle sensitivity to released ACh results in the upregulation of both spontaneous and evoked release, perhaps via the central (CNS) summation and processing of retrograde signals from individual junctions. Our data also suggest that spontaneous ACh release may have a physiological role at the adult NMJ, one of sensing the integrity and level of function of the synapse. - iv -TABLE OF CONTENTS Page Abstract ii Table of Contents iv List of Tables viii List of Figures x Acknowledgments xii Dedication xiii 1 INTRODUCTION 1 2 BACKGROUND 2.1 The Neuromuscular Junction 3 2.1.1 Events Leading to Muscle Contraction 3 2.1.2 Margin of Safety 4 2.2 Electrophysiology of the Neuromuscular Junction 5 2.2.1 Miniature End Plate Potentials 5 2.2.2 The Quantal Nature of rninEPPs 6 2.2.3 Relationship Between rninEPPs and EPPs 6 2.2.4 Quantal Content 7 2.2.5 Facilitation 8 2.3 Structure of the Neuromuscular Junction 9 2.3.1 Presynaptic Organization and Structure 9 2.3.2 Postsynaptic Organization and Structure 10 2.3.3 Other Components of the NMJ 11 2.4 Feedback Signaling at the Neuromuscular Junction 12 (cont'd) - V -2.4.1 Development of the NMJ 12 2.4.1.1 Feedback Mechanisms 13 2.4.1.2 Acetylcholine Receptor Clustering 15 2.4.2 Synapse Elimination 17 2.5 Synaptic Remodeling at the Adult NMJ 18 2.5.1 Increased Muscle Activity 19 2.5.2 Decreased Muscle Activity 20 2.5.2.1 Pre-and Postsynaptic Blockade 21 2.5.2.2 Effects of Muscle Inactivity 21 2.5.2.3 Changes in the Acetylcholine Receptor 23 2.5.2.4 Summary 24 2.5.3 Nerve Sprouting and Re-Innervation 26 2.5.4 Reduced Muscle Activity Vs. Paralysis 27 2.6 Intensive Care Unit Prolonged Paralysis 29 2.7 a-Bungarotoxin 31 3 MATERIALS AND METHODS 3.1 Experimental Models of Reduced Muscle Activity 32 3.1.1 a-Bungarotoxin Treatment 32 3.1.2 Mini-Osmotic Pumps 32 3.2 Strength and Weight Monitoring 33 3.3 Phrenic Nerve/Hemidiaphragm Preparation 34 3.4 Random Sampling 35 3.5 Physiological Solutions 35 (cont'd) - vi -3.6 |x-Conotoxin 36 3.7 Data Acquisition and Analysis 36 4 RESULTS 4.1 Miniature End Plate Potentials 38 4.2 Quantal Content with Nerve Stimulation 47 4.2.1 10 Hz Stimulation 50 4.2.2 70 Hz Train-of-Ten Stimulation 55 4.3 Normalization by 10 Hz Quantal Content 62 4.4 Normalization by Mean Train Quantal Content 62 4.5 fm Beforeand During Stimulation 70 4.6 Facilitation 75 4.7 Curve Fitting of Train Quantal Content 78 4.8 Release Histograms 84 4.9 Correlation Coefficients 84 5 DISCUSSION 5.1 Evidence of Increased Release 86 5.1.1 Spontaneous Release (fm) 86 5.1.2 Evoked Release 90 5.1.3 Discrepancies Regarding Sensitivity to Ca2+ 91 5.2 Controversy Surrounding ACh Release 92 5.3 Effects of a-BTX on Release Mechanisms 94 5.4 Interpretation of Variance of Release 95 5.5 Evaluation of a-BTX Treatment 96 (cont'd) - vii -5.5.1 Gross Morphological Changes 96 5.5.2 Degradation of a-BTX/AChR Complex 97 5.6 A Possible Mechanism for Upregulation 99 6 CONCLUSION 102 7 BIBLIOGRAPHY 103 Appendices . 120 - viii -LIST OF TABLES Table 1. A. Logfm means in [Ca2+]0 series at K+-depolarized nerve tenriinals (15 mM K4). T-tests comparing treated and control fm means (p-values). B. Standard deviations and sample size for each diaphragm at each of the six Ca2+ concentrations. Table 2. Analysis of variance between and within diaphragms (p-values) for control and treated groups. Analysis of variance of pooled logfm at each [Ca2+]0 between control and treated groups (F-test p-values). Table 3. Mean and standard deviation of control and treated log fm in 8 mM Ca2+, 10 mM K+ and 4.8 % ethanol (0.4 mM). Table 4A. Actual 10 Hz quantal content (mi0). Descriptive statistics for each diaphragm and for pooled means. Analysis of variance between and within diaphragms (p-values) for control and treated groups. Table 4B. Log 10 Hz quantal content (mi0). Descriptive statistics for each diaphragm and for pooled means. T-tests for difference between control and treated means (p-values). Analysis of variance between and within diaphragms(p-values). Table 5A. Means of actual 70 Hz train quantal content. Descriptive statistics and analysis of variance (with p-values) for control and treated diaphragms. Table 5B. Actual 70 Hz train quantal content per pulse. Means, sem, analysis of variance, and F values (comparing pooled variances) for control and treated group averages. Table 6A. Mean 70 Hz log train quantal content. Descriptive statistics, t-tests (p-values) and analysis of variance between and within diaphragms (p-values). Table 6B. Log 70 Hz train quantal content per pulse. Descriptive statistics, analysis of variance (p-values) for group data. Table 7. Actual 70 Hz train quantal content normalized by 10 Hz quantal content (per pulse): Descriptive statistics and analysis of variance between and within diaphragms (p-values). Table 8. 70 Hz log train quantal content normalized by 10 Hz quantal content (per pulse). Descriptive statistics and analysis of variance between and within diaphragms (p-values) for control and treated groups. Table 9. Actual 70 Hz train quantal content normalized by average train quantal content. Descriptive statistics and analysis of variance between and vrithin diaphragms (p-values) for control and treated data. Page 42 43 47 51 52 56 57 58 59 63 64 67 (cont'd) - IX Table 10. 70 Hz log train quantal content normalized by average train quantal content. Descriptive statistics and analysis of variance between and within diaphragms (p-values) for control and treated data. Table 11. 70 Hz log train quantal content normalized by fitted average quantal content of first pulse. Descriptive statistics and analysis of variance between and within diaphragms (p-values) for control and treated data. Table 12A. Pre-train logfm in 0.5 mM Ca2+, 5mM K+. Descriptive statistics and analysis of variance between and within diaphragms for control and treated diaphragms (p-values for t-tests comparing means and F-tests comparing variances). Table 13. Logfm during 10 Hz stimulation. Descriptive statistics and analysis of variance between and within diaphragms for control and treated diaphragms (t-tests comparing mean logfm and f-tests comparing variances of log fin between control and treated). Table 14. Log facilitation of train quantal content at the 10th pulse. Descriptive statistics, analysis of variance between and within diaphragms (t-tests comparing control and means, F-tests comparing control and treated variance, p-values). Table 15. Descriptive statistics of spontaneous and evoked release as well as of time course of release (means, SD). T-tests comparing control and treated means, F-tests comparing control and treated variance (p-values). Page 68 71 73 76 80 83 Table 16. Correlation coefficients for pooled data: spontaneous release (unstimulated, during 10 Hz, during 70 Hz), evoked release (10 Hz and 70Hz), log facilitation and time course of release (for 10 Hz and 70 Hz). 85 - X -L I S T O F F I G U R E S Page Figure 1. Log-log plot of Ca2+ titration curves at K+-depolarized nerve tenninals for control 41 and treated groups (data presented as means + sem) showing no significant differences in mean log^ values at any of the six Ca2+ concentrations. Figure 2. Top figure: Cumulative frequency distribution of treated pooled (636 junctions) log 45 fm in 8 mM Ca2+ and of theoretical normal distribution with similar mean and < standard deviation of actual log/,, data. Bottom figure: pooled log fm values plotted against theoretical distribution values (y=1.0061x-0.0045; R2 = 0.9995). Figure 3. Top figure: Cumulative frequency distribution of control pooled (~1000 junctions) 46 log fm in 8 mM Ca2+ and of theoretical normal distribution with similar mean and standard deviation of actual \ogfm data. Bottom figure: pooled log^„ values plotted against theoretical distribution values (y=1.0053x-0.0004; R2 = 0.9979). Figure 4. A. Histogram of control log fm in 8 mM Ca2+, 10 mM K+ and 4.8 % ethanol of a 48 single diaphragm (n=100 junctions). B. Histogram of treated log fm in 8 mM Ca2+, 10 mM K+ and 4.8 % ethanol of a single diaphragm (n=100 junctions). Figure 5. Histograms of pooled log m]0 data for control (A) and treated (B) diaphragms in 0.5 53 mM Ca2+, 4 mM Mg2+ showing normal distribution of both, significantly greater mean value of treated quantal content and equal standard deviation of both. Figure 6. Top figure: pooled cumulative frequency distribution of control log m10 deviations 55 from the mean and of deviations of a theoretical normal distribution with mean of zero and SD same as log m]0. Lower figure: same for treated log mi 0 . Figure 7. Top figure: Pooled cumulative frequency distribution of control log train quantal 60 content deviations from the mean with deviations of a theoretical normal distribution of mean zero and SD similar to log data. Lower figure: same for treated log train quantal content. Figure 8. Log 70 Hz train quantal content (per pulse). Means ± sem of control and treated 61 group averages. Figure 9. Train quantal content normalized by 10 Hz quantal content (per pulse) with fitted 65 curves Means ± sem. of control and treated group averages. Figure 10. Train quantal contents normalized by average quantal content of trains (per pulse) 69 with fitted curves. Means ± sem of control and treated groups. Figure 11. Train quantal content normalized by fitted quantal content of the first pulse (per 72 pulse) with fitted curves. Means ± sem of cotnrol and treated group averages. (cont'd) - xi -Page Figure 12. Top: pooled cumulative frequency distribution of control pre-train/„ deviations from 74 the mean and deviations of a theoretical normal distribution of mean zero and SD similar to log ,^. Lower: same figure for treated pooled log pre-trainj^ . Figure 13. Top: pooled cumulative frequency distribution of control log fm during 10 Hz 77 stimulation deviations from the mean and deviations of a theoretical normal distribution of mean zero and SD similar to \ogfm. Lower: same figure for treated pooled \ogfm during 10 Hz stimulation. Figure 14. Bar graphs comparing average control and treated facilitation based on two methods 79 for its calculation (p-values). Figure 15. Top: Cumulative frequency distributions of fitted control log facilitation values and 81 of a theoretical normal distribution curve of similar mean and SD. Lower: same figure for treated fitted log facilitation. ACKNOWLEDGMENTS - x i i -I wish to extend my gratitude to Dr. D. Quastel for having allowed me the opportunity to work in his laboratory, for having shared his wealth of experience and knowledge with me, as well as for having written, and re-written, several computer programs. I am indebted to the other members of my graduate committee, Drs. B. R. Sastry and D. R. Bevan for their guidance, unwavering support and encouragement. Dr. M. Walker, thank you for your '(indirect' support over the last three years. I would also like to thank Dr. M. Sutter for enriching my experience as a graduate student and for inspiring me with his comprehensive and philosophical approach to science. - X l l l --1-1. INTRODUCTION This thesis is primarily concerned with the plasticity of the adult (murine) neuromuscular junction. To that effect, we examined the effects of a chronic (3 week), postsynaptic reduction of sensitivity to acetylcholine on the release of neurotransmitter. To achieve such a compromised state of neurotransmission, we treated mice with subparalytic doses of the highly specific, irreversible nicotinic acetylcholine receptor antagonist, a-bungarotoxin (Miledi et al., 1971), which has no reported (direct) effects on nerve terminal function. A major tenet of our hypothesis was that retrograde signaling (from the muscle to the nerve), clearly implicated in the development and growth of the synapse (Hunter et al., 1989; Hynes & Lander, 1992; Bloch-Gallego et al., 1991; Arakawa et al., 1990; Fu & Poo, 1991; Harish & Poo, 1992;), continues to actively regulate the structure and function of the neuromuscular junction in the adult (Balice-Gordon & Lichtman, 1991, 1993; Rich & Lichtman, 1989). What prompted this project were the reports of prolonged post-ventilatory weakness seen in intensive care unit (ICU) patients having received paralytic doses of non-depolarizing muscle relaxants for prolonged periods of time (Sharpe, 1992). The inability of such patients to breathe on their own, upon discontinuation of relaxants, prevents their weaning from mechanical ventilation and necessitates prolonged rehabilitation. The contributions of other concomitant pathologic conditions (unrelated to neuromuscular blockade) and any possible pre-synaptic actions of these drugs notwithstanding (Chad & Lacomis, 1994), it was hypothesized that the prolonged muscle inactivity, in and of itself, may have had deleterious effects on the function and integrity of the neuromuscular junction. Therefore, we were interested in elucidating any possible roles of the postsynaptic apparatus in modulating presynaptic function. - 2 -Since an animal model of complete paralysis was beyond our means, we produced a condition which resembles myasthenia gravis; i.e., a drastic reduction in the number of functional ACh receptors (AChR). We hypothesized that under such conditions ACh release would either be upregulated to maintain the physiologic function of the diaphragm - previous reports in the literature had indicated that similar treatment resulted in an increase in AChR number and distribution in the diaphragm - or down regulated in response to diminished feedback from active muscle fibers. However, such AChRs were extrajunctional and hence, not involved in neurotransmission (Berg & Hall, 1975a, b). Such receptor upregulation may be a part of an adaptive response which does not involve increased ACh release. To test for such an increase or decrease, we measured spontaneous and evoked release as well as facilitation with different stimulation and ionic composition conditions using conventional intracellular recording. Any changes in spontaneous release frequency or facilitation would constitute evidence of presynaptic changes in release or mechanism(s), respectively. _ 3 -2. BACKGROUND 2.1 The Neuromuscular Junction In the majority of mammalian muscles, each muscle fiber has a single region of contact with the axon of its controlling motor neuron - this region constitutes the neuromuscular junction (NMJ). Its function is to relay the propagated nerve impulse from the motor nerve ending to the muscle fiber, resulting ultimately in muscle contraction. 2.1.1 Events Leading to Muscle Contraction The sequence of events leading to muscle contraction can be summarized as follows (Standaert, 1990). An action potential is propagated along the membrane of the motor neuron to the nerve terminal. This impulse causes the depolarization of the terminal and the subsequent opening of voltage-gated Ca2+ channels. The influx of Ca2+ into the terminal triggers the fusion of vesicles with the plasma membrane and the release of quanta (packets) of neurotransmitter, acetylcholine (ACh). Each quantum contains about 104 molecules and is believed to be the contents of one synaptic vesicle. The liberated ACh diffuses across the junctional cleft (a 20 nm separation between nerve and muscle) and binds to acetylcholine receptors (AChRs) on the postsynaptic muscle membrane. The binding of ACh to the AChR causes a conformational change in the latter, associated with a brief (1 ms) opening of an ion channel. This allows the passage of sodium and potassium ions down their electrochemical gradients. There is a net influx of Na+ (and efflux of K+), resulting in the depolarization of the muscle membrane. When sufficient quanta of ACh are released, the end plate potential at the NMJ reaches a threshold to activate the voltage-dependent sodium ion channels of the adjacent muscle membrane, thereby -4-initiating a muscle action potential. The muscle action potential is actively propagated down the transverse tubules, thereby initiating a sequence of events that results in muscle contraction. The action of ACh is terminated by its dissociation from the AChR; subsequent reassociation is minimized by the actions of the enzyme acetylcholinesterase (AChE). 2.1.2 Margin of Safety Under normal circumstances, the amount of ACh released and the number of AChRs activated is much larger than the minimum required to elicit a muscle action potential. This excess reflects the margin of safety of neuromuscular transmission (Waud & Waud, 1986). The margin of safety for neurotransmission in the rat (Chang et al., 1975) and mouse (Barnard et al., 1971) has been shown to be around 3-5 based on the following data: blockade of 40% of AChRs has no effect on neurotransmission, blockade of 63% of AChRs cause 5% block of twitch height, blockade of 82% of AChRs cause 75% block of twitch height, blockade of 95% of AChRs cause complete paralysis. < Any event that increases or decreases the probability of receptor opening, either due to altered release of ACh or AChR number, can change the sensitivity of the response to agonists or antagonists. For example, a decreased release of ACh in Eaton-Lambert myasthenic syndrome (ELMS; Engel, 1984) or decreased number of functional AChRs in myasthenia gravis (MG; Engel, 1984) decreases the margin of safety of neuromuscular transmission. The clinical implications of a change in the margin of safety are paramount in terms of muscle relaxant therapy. For example, MG patients are hypersensitive to neuromuscular blockade by non-depolarizing agents whereas an increase in AChR number (for example, due to immobilization or -5-burn injury) causes a reduction in sensitivity to relaxants and may lead to lethal hyperkalemia if a depolarizing (e.g., succinylcholine) is administered. 2.2 Electrophysiology of the Neuromuscular Junction 2.2.1 Miniature End Plate Potentials As previously discussed, the binding of ACh to its receptors on the postsynaptic membrane causes a brief depolarization in the muscle cell membrane, known as an end plate potential (EPP). However, even in an unstimulated preparation, Fatt and Katz (1952) observed "small" EPPs with postsynaptic intracellular recording. These miniature end plate potentials (minEPPs), as they were termed, were about 0.5-1.0 mV in amplitude, 2 msec in duration and were seen to occur at random at an average frequency of about one per second. Except for their spontaneous, random occurrence and their small amplitude, minEPPs were indistinguishable from the EPPs produced by nerve stimulation; for example, they were found to be suppressed by curare and potentiated by cholinesterase inhibitors. That minEPPs are elicited by the postsynaptic actions of nerve-released ACh was deduced from the following observations: they disappear several days after denervation at the same time that nerve endings disintegrate, and they are abolished by botulinum toxin, which inhibits ACh release (Brooks, 1956; Thesleff, 1960). Also, the frequency of minEPPs was found to be directly controlled by the membrane potential of the nerve terminal but not by the potential of the muscle fiber (del Castillo & Katz, 1954). -6-2.2.2 The Quantal Nature of rninEPPs Results from local application of ACh to muscle showed that the postsynaptic response was graded according to dose and whose time course varied with the distance and speed of application. The observation that the potentials evoked by a few molecules of ACh were well below the detection limit led to the conclusion that discrete potential changes like the rninEPPs, with regular size and time course, can only arise from a synchronous action of a "packet" of ACh containing a large number (thousands) of molecules (Katz, 1969). Moreover, such packets must be highly concentrated and delivered at a very short distance from the receptors, for their time course is not compatible with a diffusion path of more than about a micron. Morphological evidence for the quantal theory was provided by electron microscope observations of nerve terminals, in which 50 nm membrane-bound structures within the terminal cytoplasm were noted (Palade & Palay, 1954). These structures were termed synaptic vesicles and their existence led to the linking of vesicles and quanta; the vesicular hypothesis of quantal transmission. Each packet of ACh (whose release causes a minEPP) is preformed within a synaptic vesicle in the nerve terminal. The vesicle actively accumulates ACh and maintains it at a concentration considerably higher than that in the surrounding cytosol (Elmqvist & Quastel, 1965a). That release results from the fusion of a vesicle with the nerve terminal membrane was substantiated with the advent of the electron microscope. 2.2.3 Relationship Between rninEPPs and EPPs By varying the external Ca2+/Mg2+ ratio del Castilllo & Katz (1954b) established that rninEPPs and their underlying membrane conductance change are the basic unit of transmitter action and that EPPs are made up of an integral multiple of such unit components. Calcium is - 7 -known to be required for electrically evoked transmitter release (Harvey & Macintosh, 1940; del Castillo & Katz, 1954a) such that lowering its concentration in the bath reduces the amount of ACh released. Conversely, Mg 2 + competes with calcium and acts as an inhibitor of its entry and therefore release (Jenkinson, 1957; Dodge & RahamimofF, 1967). Stimulating the nerve in low Ca2+/high Mg 2 + revealed that EPPs can be reduced in discrete steps, which corresponded to the dropping out of individual minEPP units. Statistical analysis of such a response found that the distribution of EPP amplitudes is fitted accurately by a Poisson distribution, whose "unit class" is identical to the minEPP. This means that: (1) each response is made up of an integral number of units whose mean size and variance are identical with the mean size and variance of the spontaneous potentials, and (2) that, in low Ca27high Mg 2 + each "unit response" (the release of any one packet of ACh) is an event of very low statistical probability p, where p « l - each time a nerve impulse arrives at the terminal it causes the release of a few ACh packets out of a very large available store; the chance of any one unit being released remains small at all times and is not contingent upon the release of any other member of the store. 2.2.4 Quantal Content The fact that release is Poisson distributed tells us that if the mean number of packets released by an impulse is m, then the chance p of observing any particular number x (0, 1, 2, 3, etc.) ispx = (m*/x!)e'm. Therefore, for a sufficiently large number of observations N, the value NjPx should come close to the actually observed number of responses which contain x quanta and which are consequently made up of a summation of x rninEPPs. Similarly, m = (mean amplitude of response) / (mean amplitude of minEPP). By the same token, e"m = rio / N = (number of failures) / (number of impulses), or m = ln (no. of impulses / no. of failures). This equation has -8-been tested over a wide range with frog and mammalian nerve-muscle preparations and in all cases excellent agreement was found (del Castillo & Katz, 1954a; Boyd & Martin, 1956; Liley, 1956). 2.2.5 Facilitation The increase in quantal content of an EPP by antecedent nerve impulses is referred to as facilitation, and is distinguished from augmentation and tetanic potentiation (which also increase EPP quantal content) on the basis of the rate of growth and decay (Zengel & Magleby, 1982). Katz and Miledi (1968) found that extracellular Ca2+ must be present not only for a stimulus to evoke ACh release, but also for it to facilitate subsequent release. They suggested that facilitation might be due to residual Ca2+ within the nerve terminal. Two observations suggest that facilitation depends on Ca2+ entry: (1) it is blocked after the intracellular application of Ca2+ chelators to nerve terminals; and (2) it disappears at low extracellular Ca2+ / high Mg2+, and is restored by 4-aminopyridine (K+ channel blocker). However the multiplicative interactions of facilitation with augmentation and tetanic potentiation at the frog NMJ (Zengel & Magleby, 1982) cannot fit such a residual Ca2+ model for more than one of these processes. On the other hand, Bain and Quastel (1992a) found that when Sr2* is substituted for Ca2+ co-facilitation of EPPs and minEPP frequency (fm) becomes consistent with the residual ion model (i.e., that facilitation occurs only because accumulated ST2+ adds to the amount of ion brought in by each impulse. Bain and Quastel (1992b) subsequently reported that in addition to residual Ca2+, facilitation includes a multiplicative component, not seen with Sr2*, that probably depends upon entry of Ca2+ into the nerve terminal; i.e., facilitation, although dependent upon Ca2+, does not occur primarily because the Ca2+ that enters with each presynaptic -9-action potential is supplemented by Ca2+ that has accumulated within the nerve terminal. They also found that facilitation 'saturates' at low levels of Ca2+, such that further increase in external [Ca2+] does not result in more facilitation although it does increase quantal content of EPPs. 2.3 Structure of the Neuromuscular Junction Like other chemical synapses, the neuromuscular junction consists of two primary elements, a presynaptic one, the nerve terminal, and a postsynaptic structure, formed at a discrete region on the skeletal muscle fiber. However, unlike other chemical signaling systems, such as the endocrine system, the neuromuscular junctions differ in two important aspects: they are considerably faster (milliseconds vs., seconds), and they are focal, with the signal from a single nerve terminal directed to a single target cell that is physically attached to it. Electron micrographs of the NMJ indicate its highly specialized nature, both pre- and postsynaptically (Engel, 1986; Ogata, 1988). 2.3.1 Presynaptic Organization and Structure The nerve terminal contains large numbers of 50-nm diameter synaptic vesicles that contain ACh (Kelly, 1993). These vesicles are clustered at specialized sites known as active zones and within the active zones is found much of the molecular machinery that regulates vesicle docking, fusion, and neurotransmitter secretion, such as calcium channels (Robitaille et al., 1990; Cohen et al., 1991), calcium-gated potassium channels (Robitaille et al., 1993), neurexins (Ushkaryov et al., 1992), and syntaxin (Bennett et al., 1992). A few larger, dense-core vesicles are also present; these presumably contain adenosine tri-phosphate (ATP) and neuropeptides such - 10-as calcitonin gene-related peptide (CGRP: Matteoli et al., 1990) which are involved in the formation and maintenance of the NMJ (see Feedback Signaling at the Neuromuscular Junction). 2.3.2 Postsynaptic Organization and Structure The specializations on the postsynaptic component of the endplate region formed by the muscle fiber are characterized by shallow gutters on the fiber surface into which the nerve terminal fits, 1 urn deep invaginations of the membrane termed junctional folds that indent the gutters, membrane thickenings that cover the crests of the folds and extend partway down their sides, and clusters of nuclei that lie beneath the postsynaptic membrane. The thickened membranes are the ACh receptor (AChR)-rich chemosensitive surfaces (~104/um2) and are precisely aligned with the presynaptic active zones. However, the AChR levels fall by >90% in the depths of the lum deep folds, and there are fewer than 10 AChRs per jam2 in extrasynaptic membrane (Salpeter et al., 1988). The AChRs are glycoproteins that traverse the membrane with five subunits, each approximately 400-500 amino acids long. Four proteins, termed a, (3, s and 8 subunits in order of increasing molecular weight, are present in a ratio of 2:1:1:1 (Yu & Hall, 1991). The two binding sites for ACh and other ligands are located in the a subunit and are not identical, but rather of high and low affinity. The source of this difference has been suggested to be the direct contribution of subunits flanking the a subunit, specifically 6 or y and 5 (Yu & Hall, 1991; Pedersen & Cohen, 1990) (the y subunit is the fetal analog to the adult e subunit). -11 -2.3.3 Other Components of the NMJ Other membrane proteins, including neural cell adhesion molecule (N-CAM), voltage-gated sodium channels (Covault & Sanes, 1986; Boudier et al., 1992) and cytoskeletal anchoring proteins, such as 43K acetylcholine receptor associated protein (Neubig et al., 1979; Porter & Froehner, 1983), are also concentrated in the postsynaptic apparatus of the motor endplate. And although precisely aligned with each other, the nerve terminal does not contact the muscle fiber but is separated by a 50-nm synaptic cleft. The basal lamina traverses this cleft and extends into the junctional folds. It is part of a continuous sheath that encircles the muscle fiber and is fused to the Schwann cell's basal lamina. Although the synaptic basal lamina is indistinguishable from the extrasynaptic basal lamina to which it is attached, it is biochemically specialized: it contains acetylcholinesterase (AChE), the enzyme which inactivates ACh, and components responsible for the strong adhesion of nerve to muscle and factors that mediate developmental interactions, such as agrin, s-laminin and enactin (Hall, 1973; Reist et al., 1987; Hunter et al., 1989; Chiu & Ko, 1994). Furthermore, even the Schwann cells are specialized at the neuromuscular junction. Whereas pre-terminal Schwann cells form myelin, the synapse-associated Schwann cells cap the nerve terminals presumably to protect them from chemical and mechanical insults. Three sets of observations suggest, however, that terminal Schwann cells may play other roles as well. First, they elaborate extensive processes and phagocytose nerve terminals following axonal damage and thus may participate in axonal remodeling (Birks et al., 1960; Reynolds & Wolf, 1992). Second, they exhibit a calcium transient in response to axonal action potentials, showing that they can sense electrical signals (Jahromi et al., 1992). Finally, they acquire the ability to synthesize and secrete ACh following denervation (Birks et al., 1960; Brockes, 1984). - 12 -2.4 Feedback Signaling at the Neuromuscular Junction The function of the neuromuscular junction is contingent upon the precise alignment of its pre- and postsynaptic elements which is achieved during development, and maintained in the adult, through the exchange of feedback signals between the nerve terminal and the muscle. Although some of the presynaptic mediators of this signaling have been identified (CGRP, ARIA, agrin), much less certainty surrounds the muscle-derived retrograde factors that control the nerve terminal. Furthermore, given the inherent overlap in the sequence of developmental events and their inductive signals, no attempt will be made to separate the antero- and retrograde signals as such. Rather, the formation and development of the NMJ will be reviewed and feedback signaling contributing to each developmental aspect will be appropriately identified. 2.4.1 Development of the Neuromuscular Junction Prior to the development of a neuromuscular junction the neuronal growth cone has to make contact with the myotube which it will be innervating; however, neither are differentiated yet. Neuronal growth cones are unspecialized in that they have a sparse supply of vesicles, lack active zones and terminal branching. However, they do possess a release mechanism, underdeveloped as it may be, and release ACh, both spontaneously (Hume et al., 1983) and in response to stimulation (Young & Poo, 1983), as well as other neurotrophic factors, most prominent of which is CGRP. Similarly, the myotube has no membrane specializations and no basal lamina, but the surface of the cell contains randomly inserted, unanchored, immature (or fetal) AChRs. These receptors differ from those present at the adult synapse in two respects: they have a different subunit composition (o^ PyS vs. o^ PsS; Gu & Hall, 1988) and different ionic channel properties (smaller single-channel conductance and longer opening time; Schuetze & -13-Role, 1987). And even before contact is made the myotube appears to be receptive to neurally-released substances (ACh, CGRP etc.) and the growth of the neural cone itself appears to be directed by factors released from the myotube. 2.4.1.1 Feedback Mechanisms The first contact-mediated change is the arrest of the neuronal cone's growth and its adhesion to the myotube surface, presumably affected by a basal lamina-associated protein, S-laminin (Hunter et al., 1989). Subsequent to this arrest in growth, is a rapid increase in the rate of transmitter release, synaptic currents (Xie & Poo, 1986; Sun & Poo, 1987; Chow & Poo, 1985) and neuritic growth (Henderson et al., 1984); all of which have been shown to be mediated by the contact-mediated increase in intracellular Ca2+ concentration. This increase in calcium concentration is specific in that it is not induced by contact with nontarget cell membranes, can be rapidly reversed if contact is terminated and may be mediated through the activity of cAMP-dependent protein kinase (Zoran et al., 1993). Furthermore, calcium-induced calcium release has been implicated as a general regulator of neuronal differentiation (Holliday et al., 1991) and although Ca2+ does not appear to be required for formation of a functional nerve-muscle contact, it is required for its subsequent maturation (Henderson et al., 1984). One such factor suggested to be part of the system involved in the contact-mediated differentiation is N-CAM, a cell adhesion protein (Hynes & Lander, 1992), which has been shown to: (1) be downregulated at extrajunctional regions during development, (2) be restricted to neuromuscular junction in the adult, (3) result in a G-protein-dependent increase in intracellular Ca2+(Doherty et al., 1991) and protein phosphorylation in growth cone membrane (Atashi et al., 1992). -14-In addition to the contact-mediated signaling, other effects exerted by the muscle on the nerve may be mediated by diffusible factors. For example, embryonic muscle extracts support the survival of motor neurons (Bloch-Gallego et al., 1991) as do several other identified growth and neurotrophic factors, including ciliary neurotrophic factor (CNTF) (Oppenheim et al., 1991) and basic fibroblast growth factor (bFGF) (Arakawa et al., 1990). Also, bFGF, in which skeletal muscle is relatively rich, has been shown to induce changes in nerve terminals that resemble those associated with nerve-muscle contact, such as increased ACh release accompanied by a rise in intra-terminal cytosolic Ca2+ concentrations as well as clustering of synaptic vesicles. The bFGF receptor is a tyrosine kinase (Pazin & Williams, 1992) whose activation in other systems leads to an increase in intracellular Ca2+ concentrations. Furthermore, heparan sulfate proteoglycan (HSP) has been shown to be required for bFGF signaling (Ornitz et al., 1992). Immunohistochemcial studies have revealed that NMJs are enriched in HSP (Bayne et al., 1984); however, the presence of bFGF at the NMJ has not yet been established. Finally, in addition to growth factors, products of muscle metabolism may also act as retrograde signals. For example, while muscle cells release ATP (Smith & Lindgren, 1986), direct application of ATP potentiates spontaneous ACh release at developing NMJs (Fu & Poo, 1991) not by direct depolarization of the nerve terminal but through activation of a protein kinase and a concomitant increase in intracellular Ca2+ concentration; also 5-hydroperoxy-6,8,11,14-eicosatetraenoic acid (5-HPETE), a metabolite of arachidonic acid released from muscle, has been shown to have a similar effect (Harish & Poo, 1992). i Therefore although no definitive retrograde signal mediating presynaptic differentiation has yet been identified, it appears as though contact-mediated interactions and soluble factors are involved in triggering an influx of calcium (into the cone; presumably by activating a cAMP-dependent protein kinase (Funte & Haydon, 1993)) resulting in a rapid increase in transmitter release and the induction of a slower remodeling of the cone into a branched, varicose nerve -15-terminal (with an increased number of synaptic vesicles, the appearance of active zones, and the maturation of the neurotransmitter release mechanism (all of which occur in the order specified) (Kelly & Zacks, 1969; Nakajima et al., 1980; Takahashi et al., 1987). During this phase of neuronal growth cone differentiation the muscle cell exhibits several developmental changes as well. The membrane starts thickening at regions of contact with neuronal tissue, a basal lamina starts forming, AChRs cluster in the muscle membrane across the nerve terminal and retract from non-synaptic sites as well as undergo a change in metabolic properties, and an increase in transcription of AChR mRNA encoding mature AChRs occurs. Although these events are not directly mediated by muscle-derived signals, a brief overview is required because AChR number and distribution will be evaluated in light of the ct-bungarotoxin treatment in the Discussion section. 2.4.1.2 Acetylcholine Receptor Clustering The initial expression of AChRs is independent of nerve control and is part of a general "program" of muscle differentiation controlled in part by MyoD and its homologs (Li & Olson, 1992), a family of muscle-specific transcription factors which appear to activate most muscle-specific genes. However, subsequent to nerve-muscle contact the number and distribution of AChRs falls under the control of the nerve terminal. Innervation regulates the accumulation of synaptic AChRs in two ways: it causes their clustering around the newly formed synapse (and indirectly suppresses AChRs at non-synaptic regions by eliciting muscle electrical activity) and stimulates the nuclei near the synapse to synthesize AChR mRNA encoding a different subunit composition. -16-The clustering of AChRs is brought about by neurally-derived agrin molecules (a basal lamina-associated protein bearing homologous domains to protease inhibitors, epidermal growth factor, and laminin; Tsim et al., 1992) that are inserted into the basal lamina of the postsynaptic membrane (Reist et al., 1992). Agrin interacts with a distinct receptor (Nastuk et al., 1991) via a mechanism that may involve heparan sulfate proteoglycans and the phosphorylation of a tyrosine in the cytoplasmic loop of the P subunit of the AChR (Wallace et al., 1991). On the other hand, the increase in the local synthesis of AChRs may be induced by calcitonin gene-related peptide (CGRP; Fontaine et al., 1986) which is co-released with ACh from the nerve terminal (Matteoli et al., 1990) and increases AChR mRNA and protein levels apparently through a cAMP-mediated pathway, or by acetylcholine receptor-inducing activity protein (ARIA). To increase transcription in nuclei near synapses, such factors may act through a conventional second messenger system or by causing a more prolonged alteration in the nuclei. Evidence for differences between synaptic and extrasynaptic nuclei has been provided in terms of morphology and attachment to the cytoskeleton (Couteaux, 1973) as well as gene transcription (Brenner et al., 1990). The release of the above mentioned factors may also trigger the accumulation of AChRs even prior to nerve-muscle contact (Dahm & Landmesser, 1991), thereby potentiating the effects of the spontaneously released ACh. The density of AChRs is controlled through a third pathway that depends on the electrical activity evoked by synaptic transmission. The subsequent changes in AChR properties at synapses are also nerve dependent and appear to reflect a combination of activity-dependent and -independent influences. -17-2.4.2 Synapse Elimination Following the differentiation of the pre- and postsynaptic elements of the neuromuscular junction, the next discrete stage in its formation hinging on muscle-derived retrograde signals is that of synapse elimination, or the transition from a poly-innervated to a singly- innervated state. Following the establishment of the first synaptic site, other axons converge and innervate the same muscle fiber for a transient period, of about 1-2 weeks, after which all synapses are lost except for one (Redfern, 1970; Van Essen et al., 1990). This reduction does not correspond to a decrease in the number of motor neurons innervating each muscle; rather, each neuron decreases the number of muscle fibers that it innervates (Brown et al., 1976). This process of elimination involves activity-dependent, competitive interactions between closely appositioned synapses. This "relative activity" pattern of control is similar to that observed in the visual system (Goodman & Shatz, 1993). Synapse elimination has been shown to be reduced, or halted, by a reduction in neuronal activity (Callaway & van Essen, 1989; Thompson et al., 1979) and enhanced when synaptic activity was increased (O'Brien et al., 1978; Thompson, 1983). Some theories involving both trophic and suppressive retrograde factors presume that the competitive interactions occur presynaptically, with the muscle cell serving as a source, and that they only require separate inputs to be differentially sensitive to a uniform retrograde signal (Oppenheim, 1991; Lo & Poo, 1991). However, evidence for a more direct involvement of the postsynaptic apparatus as a mediator of competition comes from studies in which synapse elimination was slowed by postsynaptic blockade of muscle activity (Callaway & van Essen, 1989). Furthermore, Balice-Gordon and Lichtman (Balice-Gordon & Lichtman, 1991) showed that at neuromuscular junctions in which a gradient of activity had been established by focal postsynaptic blockade with ot-bungarotoxin, presynaptic and postsynaptic elements were gradually removed from the silenced portions of the -18-end plate. Such elimination of postsynaptic structures did not occur when the entire end plate was inactivated by either toxin or denervation. Thus the activity level at and within a synapse can directly influence the formation and maintenance of synaptic sites. A muscle-derived retrograde factor would be a likely candidate to mediate this synaptic remodeling since muscle activity has a well-documented role in controlling the transcription and expression of several muscle fiber proteins (Booth et al., 1984; Goldman et al., 1988; Eftimie et al., 1991; Yang et al., 1991). Further evidence for a role of the muscle fiber in synaptic maintenance comes from examination of the stability of end plates after removal of target muscle fibers. In the mouse, selective removal of muscle fibers results in a dramatic loss of nerve terminal branches (Rich & Lichtman, 1989). Additionally, loss of AChRs at synaptic sites has been shown to precede nerve terminal retraction (Balice-Grodon & Lichtman, 1993). This suggests (1) a competitive model in which more active axons remove the postsynaptic specializations of their less active neighbors thereby further "weakening" them and ultimately leading to their withdrawal, and (2) that the fate of competing inputs in determined by short-range interactions within the postsynaptic cell. 2.5 Synaptic Remodeling at the Adult Neuromuscular Junction In most adult animals, skeletal muscle fibers remain innervated by the same motor neuron for the life of the animal. However, studies of the stability of the adult neuromuscular junction, in fixed tissues (Wernig & Herrera, 1986; Robbins, 1988) and in vivo (Lichtman et al., 1987; Herrera et al., 1990), are consistent with the observations made in growing neuromuscular junctions in that the degree of remodeling displayed by adult junctions varied depending on muscle fiber type and species. -19-One possible explanation of such differences in synaptic stability may be the level of synaptic activity and activity-dependent production of retrograde factors by muscle. Wernig et al. (1984) and Hill et al. (1991) found that while the overall or gross structure of NMJs in mouse soleus muscle remained stable, the lengths and complexity of the NMJs changed over time in mature and developing mice. Similarly, muscle fiber growth resulting in the expansion of the postsynaptic element elicits a concomitant increase in presynaptic apparatus, as would be expected if the efficiency of synaptic transmission is to be maintained (Balice-Gordon & Lichtman, 1990). Therefore, it appears that changes in the structure of the NMJ are initially elicited by the postsynaptic endplate, which in turn, evokes alterations in the nerve terminal. Similar observations have also been made in aged animals (Kelly & Robbins, 1983; Robbins & Fahim, 1985; Stebbins et al., 1985). Furthermore, it is important to note that while these cited reports had used slow-twitch muscle fibers to demonstrate synaptic plasticity, others (Lichtman et al., 1987), who did not find evidence of such remodeling, had used fast-twitch muscles. This is significant because slow twitch fibers are recruited to a much greater extent and, thus, are more active than fast twitch fibers (Henning & Lomo> 1985). This further reinforces the regulatory role that the muscle exerts over the nerve at the NMJ. However, no studies have sought to document alterations in either neuromuscular transmission or muscle fiber contractile function, in healthy animals, during time frames similar to those used in NMJ remodeling studies or in the adult. 2.5.1 Increased Muscle Activity Much of the current research regarding the effects of increased muscle activity on NMJ morphology has yielded conflicting results and appears to be dependent on muscle fiber type - 20 -(Andonian & Fahim, 1988), age (Rosenheimer, 1985; Stebbins et al., 1985; Tomas et al., 1989), and intensity (Deschenes et al., 1993). In fast-twitch extensor digitorum longus (EDL) muscles of young adult mice, Dorlochter et al. (1991) reported no changes in tetanic force (generated by the muscle) but a significant increase in fatigue resistance and quantal content (as evidenced by greater maximum and plateau endplate potentials during trains of direct nerve stimulation). This increase in evoked ACh release was attributed to an increase in nerve terminal size as reported by others (Andonian & Fahim, 1988; Deschenes et al., 1993; Waerhaug et al., 1992). Also, the activities of choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) have been found to be muscle fiber type dependent (Crockett et al., 1976) and to be affected differently by increased muscle activity, such that AChE, but not ChAT, is increased (Crockett et al., 1976; Herscovich & Gershon, 1987). Furthermore, Lomo et al. (1985) found that muscle activity has a greater influence on total AChE activity and expression than nerve-muscle interactions. 2.5.2 Decreased Muscle Activity Although a considerable amount of work has been done on the effects of muscle inactivity on the NMJ, a clear distinction has to be made between those employing presynaptic blockade of muscle activity and those using postsynaptic blockade; primarily because, and as will be further discussed, they affect different changes implicating distinct feedback pathways. Also, presynaptic blockade precludes the ability to distinguish effects due to impaired nerve terminal function (trophic and/or suppressive effects of ACh and other co-released neuropeptides) from those due to muscle inactivity per se. -21-2.5.2.1 Pre- and Post-Synaptic Blockade Nonetheless, muscle inactivity due to presynaptic impairment of neurotransmission has been achieved with a number of experimental techniques, including surgical denervation, spinal cord section (Johns et al., 1961), application of local anesthetics (Robert & Oester, 1970), botulinum toxin (Thesleff, 1960), P-bungarotoxin (Hoffman & Thesleff, 1972), hemicholinium (Chang et al., 1975) or tetrodotoxin (Brown & Ironton, 1970) to the motor nerve; thereby either reducing impulse frequency in the nerve or its ability to release ACh. Conversely, postsynaptic blockade of muscle activity has been achieved with d-tubocurarine (Hogue et al., 1992; Kim et al., 1995) or ct-bungarotoxin (Chang et al., 1975; Berg & Hall, 1975a,b; Plomp et al., 1992). And although there is controversial evidence regarding the possible presynaptic effects of d-tubocurarine (Auerbach & Betz, 1971), a-BTX has been shown not to affect the function of the nerve or the release of ACh (Berg & Hall, 1975b; Gelsema, 1980; Katz & Miledi, 1978). 2.5.2.2 Effects of Muscle Inactivity That there is a difference in effect due to pre- and postsynaptic blockade can be readily inferred from the resulting changes in NMJ morphology. On the other hand, that there is overlap in certain effects implies a certain extent of redundancy in the role(s) of some antero- and retrograde signals; however, in the absence of one or the other the integrity of the NMJ cannot be sustained solely by either. Therefore, to begin with, both types of blockade stimulate the development of ACh supersensitivity at non-endplate regions (Axelson & Thesleff, 1959; Miledi, 1960; Albuquerque & Mclsaac, 1970) by an increase and spread of extrajunctional AChRs along the whole length of muscle fiber (Lee et al., 1967; Miledi & Potter, 1971; Berg et al., 1972; Chang et al., 1973) as J -22-well as decrease the resting membrane potential of muscle fibers (Berg & Hall, 1975a). Ultimately, muscle inactivity, regardless of method of induction, leads to (in the following order): the development and spread of extrajunctional AChRs (as mentioned), an increase in nerve terminal area, branch number and complexity (Tomas et al., 1989), motor endplate expansion (Eldridge et al., 1981), nerve terminal degeneration, abandoned junctional folds and synaptic gutters (Fahim, 1989; Fahim & Robbins, 1986), and finally nerve terminal sprouting (see later; Fahim, 1989; Fahim & Robbins, 1986; Holland & Brown, 1980; Pestronk & Drachman, 1978). However, only denervation and presynaptic blockade (interference with ACh synthesis or release with hemicholinium or P-bungarotoxin (Chen & Lee, 1970)) cause an increase in AChR at junctional sites. It is also worthy of note that the extent of both the neuromuscular blockade and structural changes at the muscle fiber membrane caused by the maximum tolerable dose of hemicholinium were less than those produced by the local application of P-bungarotbxin (or denervation) suggesting that the extent and nature of the blockade might be a determining factor in the extent to which such changes occur, the sequence and severity of which had been outlined above. Similarly, postsynaptic blockade has been shown to cause subsequent increase in the margin of safety of neuromuscular transmission (Chang et al., 1975, Barnard et al., 1971). Such an effect was seen within 24 hours with a single paralytic dose of a-bungarotoxin (150 u,g/kg; which most animals survived, presumably by relying on auxiliary respiratory muscles). Within 12 hours more than 90% of AChR were still blocked but neurotransmission was restored to 77-85% of normal (tetanic stimulation at 50 Hz also gave a rather well sustained contraction) in about 80% of the junctions. As early as two days after the initial injection, neuromuscular transmission was shown to be fully restored but the diaphragm continued to be extremely vulnerable to further - 23 -treatment with a-bungarotoxin. An additional two injections did not increase the number of extrajunctional AChRs, nor did they alter the numbers at the junctional sites. However, the development of extrajunctional AChRs continued for about six days. Since a-bungarotoxin does not cause any inhibition of acetylcholine release (Chang & Lee, 1963, Tsai, 1975) the increase in the number of extrajunctional receptors by a-bungarotoxin could not be due to a deficiency of any neurotrophic substances as had been suggested by Miledi (1960) but rather due to muscle inactivity, as had been previously stated. 2.5.2.3 Changes in the Acetylcholine Receptor Generally speaking, although receptor number regulation has been explained by the classic model of agonist/antagonist-induced upregulation and downregulatioh of receptors (Kenakin, 1987), even in the CNS benzodiazepine/y-aminobutyric acidA, dopamine and cholinergic synapses (Goss et al., 1991; Herman & Sominska-Zurek, 1979; Miller et al., 1988a,b), such a model cannot account for muscle inactivity-induced changes at the neuromuscular junction; primarily because extrajunctional AChRs are not involved in neurotransmission (see later). However, such a model probably pertains at the NMJ in the effects of chronic treatment with neostigmine, which increases synaptic concentrations of ACh above physiological levels: muscle weakness and a decrease in both synaptic AChR number and ACh release occur (Chang et al., 1973). It has already been said that both neurally-derived factors and muscle activity-dependent factors have been found to influence the distribution of AChRs on the postsynaptic membrane. CGRP, agrin and ARIA induce AChR clustering at synaptic sites, while muscle electrical activity influences the gene activity of muscle membrane proteins, including AChR (Goldman et al., - 2 4 -1988). Therefore, when the muscle is deprived of ACh and other neural factors the ensuing muscle inactivity triggers the increased transcription of AChRs, both at synaptic and extrasynaptic sites. However, when muscle activity is solely due to postsynaptic blockade, only extrajunctional AChRs proliferate as the presence of neural factors still exerts an influence on the synaptic region. The transactivating factors that mediate the effects of activity probably include members of the MyoD family of transcription activators, which are also involved in the initial induction of AChR genes during myogenesis (as previously mentioned), as levels of mydD and myogenin mRNA have been found to be regulated by electrical activity in parallel with the levels of the AChR subunit mRNAs (Eftimie et al., 1991; Witzemann & Sakmann, 1991). As such, the first steps in the pathway from electrical activity to the nucleus may involve regulation of transcription factors by protein kinase C (PKC). Electrical stimulation of denervated muscle has been shown to increase nuclear PKC activity, whereas inhibition of kinase blocks the effects of denervation (Huang et al., 1992) (i.e., the development of extrajunctional AChRs). Also, PKC activity could be increased by intracellular calcium or diacylglycerol, released by the action of phospholipase C, as both have been found to be elevated by sustained muscle activity (Vergara et al., 1985), with alterations of intracellular calcium concentrations having been shown to alter AChR expression (Birnbaum et al., 1980). However, whether changes in inositol triphosphate concentration are secondary to changes in intracellular calcium or arise independently has not been elucidated. 2.5.2.4 Summary Regardless of the exact mechanisms involved in AChR regulation, the physiological relevance can be rationalized as follows: under conditions of presynaptic blockade the number of -25-junctional AChRs becomes increased because the functional AChRs are not "sensing" any transmitter. On the other hand, a postsynaptically paralyzed muscle which cannot respond to ACh reacts by increasing AChR number; however, neurally-released substances (which act through a variety of pathways) suppress the production of synaptic AChRs and maintain the integrity and constitution of the junction. Therefore, only extrajunctional AChRs appear. With regards to extrajunctional AChRs, when neurotransmission cannot be restored at existing neuromuscular junctions, the system attempts to establish new synapses. However, the only available mechanism is that encoded in the cells' genome, which involves immature, or embryonic, AChRs. Reverting to a pseudo-developmental state, the muscle can now accept re-innervation at any point on its surface (keeping in mind that innervated, adult skeletal muscle cannot be re-innervated in its normal state; Frank et al., 1975). Although it is true that nerves preferentially make contacts at original synaptic sites in the case of denervation (Rich & Lichtman, 1989), in the event that this is not possible (due to AChR blockade or to the presence of the original nerve terminal) any other site of the muscle surface will suffice. Therefore, that extrajunctional AChRs are not physiologically relevant (they are not aligned with nerve terminals, have different subunit make-up (a2Py8; Gu & Hall, 1988) resulting in smaller single-channel conductance andjonger opening times (Schuetze & Role, 1987) suitable for depolarizing the smaller myotubes (Jaramillo et al., 1988)) is perfectly understandable; they are precursors for junctions yet to be formed. The transcription of the relevant mRNA results in expression of such receptors across the entire surface of the muscle cell. In the event that transmission is restored (as was the case in the above mentioned cases of a single paralytic injection of ct-BTX) the resulting muscle activity suppresses the further insertion of embryonic (extrajunctional) AChRs. Even the time course of appearance and removal of such receptors is consistent with the persistence and -26-alleviation of the threat to transmission by a-BTX. Transmission is restored within 24 hours within which the genetic machinery would have been activated. However, the number of junctional AChRs is still marginal and continues to constitute a threat to transmission. But within 6 days, or so, the a-BTX/AChR complex is degraded (as will be discussed in Discussion) and the threat is eliminated at which time the extrajunctional receptors are removed. 2.5.3 Nerve Sprouting and Re-Innervation A brief discussion of muscle re-innervation will be undertaken for two reasons: one, to re-inforce the previously outlined connection between extrajunctional AChRs and reinnervation in the adult, and two, to attempt to further elucidate the postsynaptic element's role in the differentiation and maintenance of the nerve terminal. Denervation and/or (postsynaptic) muscle paralysis not only stimulates the development of extrajunctional AChRs but also triggers nerve terminal sprouting (reviewed by Brown et al., 1981). However, direct electrical stimulation of denervated muscle, either in vivo (Lomo & Rosenthal, 1972) or in organ culture (Purvis & Sakmann, 1974), can prevent the development of such extrajunctional AChRs as well as block nerve innervation (Jansen et al., 1973). Moreover, the observation that neurons sprout in response to inactivation of their own or neighboring muscle fibers suggests that both contact-mediated and short-range diffusible signals originating in muscle are involved in the maintenance of the neuromuscular junction (Brown et al., 1981). Such studies of motor nerve terminal regeneration have established a role of muscle activity as well as molecules associated with the synaptic basal lamina in providing cues that direct neuronal growth and differentiation. In fact, at least two such adhesion molecules that can promote nerve-muscle interactions in vitro, N-CAM and N-cadherin, are concentrated at synaptic sites in innervated -27-muscle (Bixby et al., 1987) and are re-expressed at extrasynaptic regions after dennervation (or paralysis) (Covault et al., 1986). The idea that a signal from the muscle affects the behavior of the nerve terminal arises from the observation that partial denervation of a muscle causes sprouting of the remaining nerve terminals, an effect which can also be blocked by electrical stimulation (Brown et al., 1981). Because such an effect would occur over a distance, a factor secreted by denervated or inactive fibers is presumed to be involved. Possible candidates for such a role include insulin-like growth factor 2 (IGF-2), basic fibroblast growth factor (bFGF) and ciliary neurotrophic factor (CNTF), all of which induce motor nerve sprouting (Caroni & Grandes, 1990; Gurney et al., 1992), as well as FGF-5 and CNTF that are produced by the muscle (Hughes et al., 1993; Leung et al., 1992). 2.5.4 Reduced Muscle Activity Vs. Complete Paralysis Sub-paralytic doses of a-bungarotoxin (Plomp et al., 1992, 1994, 1995; Berg & Hall, 1975a,b) and d-tubocurarine (Kim et al., 1995; Hogue at al., 1992) have been shown to stimulate the development of extrajunctional receptors with further prolongation of administration of the toxin (Chang et al., 1975) not increasing the extent of development of extrajunctional receptors. Therefore, previous attribution of extrajunctional AChR development to muscle inactivity per se (Drachman & Witzke, 1972; Drachman, 1974; Lomo & Rosenthal, 1972; Jones & Vrbova, 1970) is not completely accurate. Furthermore, since reported treatments have ranged from complete paralysis (with mechanical respiration; Berg & Hall, 1975a,b) to single a-BTX injections (Fertuck et al., 1975) to 3-week subparalytic doses of either d-TC (Kim et al., 1995; Hogue et al., 1992) or a-BTX (Plomp et al., 1995) it is clear that complete muscle inactivity is not a prerequisite for the upregulation of AChRs (i.e., the development of extrajunctional -28-AChRs). Also, the extent (magnitude) of neuromuscular blockade does not appear to be critical, for it too varies with the different agents and duration of treatment; provided it taxes the neuromuscular transmission sufficiently to establish a threat; i.e., more than 80% blockade in the case of the rodent diaphragm. In support of the latter estimate, reports of chronic d-tubocurarine infusions (for 2 weeks) with a steady state blood level producing 60% twitch blockade did not affect diaphragmatic AChR distribution or numbers, although they did cause an upregulation of AChR numbers of other skeletal muscles. Therefore, reduced postsynaptic sensitivity appears to be a sufficient stimulus for the induction of such ultrastructural changes while complete muscle inactivity leads to further modifications, primarily nerve terminal sprouting. Additionally, it has been shown that chronic treatment with paralytic doses of either D-tubocurarine, succinylchdline or oc-bungarotoxin does not impair the function of the phrenic nerve but that the resting membrane potentials and input resistance of muscle fibers became similar to those of denervated muscles (Berg & Hall, 1975a). However, and in contrast to denervation or muscle paralysis, a mere reduction in postsynaptic sensitivity has been shown not to affect resting muscle membrane potential (Plomp et al., 1994; experiments for this thesis). This observation further supports the clear difference in effect of reduced postsyanptic sensitivity and complete paralysis; primarily, muscle activity and its regulatory role in the maintenance of the adult NMJ. The observations made thus far suggest that loss of postsynaptic structures can affect the retraction, elimination or sprouting of nerve terminals. Muscle-fiber derived "retrograde" factors are involved in the maintenance of the mature neuromuscular junction. However, the molecular mechanism for the regulation of expansion, addition and retraction of synaptic sites within an individual junctions and for the loss of entire nerve terminal arbors during synapse elimination remains a mystery. And more importantly, the evidence is implicated in gross structural -29-remodeling but not in finer functional modulation. Therefore, that retrograde signals between muscle and nerve terminal exist has been rather well established but the involvement of the latter in the physiological function of nerve terminals has not. 2.6 ICU Prolonged Paralysis Over the past 10-15 years two conditions have come to the forefront in the management of patients in the intensive care unit (ICU) receiving neuromuscular blocking agents. The first is the observation of prolonged blockade of neuromuscular transmission following extensive administration of muscle relaxants. The second is the development of profound weakness that prevents voluntary breathing, necessitating the continuation of mechanical ventilation. The cause of the prolonged blockade, which persists after discontinuation of muscle relaxant administration, has been difficult to establish and believed to be multifactorial; however, some potential contributors to the condition have been other drugs used in conjunction with the relaxants as well as pathophysiological conditions associated with diseased state of the patient. They have included, acidosis and hypothermia (which are known to prolong neuromuscular blockade), antibiotics, local anesthetics, calcium channel blockers (as anti-arrhytlimic), cyclosporin, electrolyte disorders (hypocalcemia, hypkalemia, hypermagnesemia, hypernatremia), metabolic disturbances, renal and/or hepatic dysfunction, large peripheral compartments for and active metabolites of administered muscle relaxants (Sharpe, 1992; Stoelting, 1991; Bevan et al., 1988). - 30 -Some of the common factors associated with the extensive weakness, also referred to as post ventilatory weakness, have included: critical illness, sepsis, asthma, ventilation, and the use of muscle relaxants and/or steroids. However, not all of these factors are present in any one case. Furthermore, the complexity of the situation has been compounded by the idiopathic pathophysiologies found in these patients, such as polyneuropathies, myopathies and denervation (Chad et al., 1994; Gorson & Ropper, 1993; Coronel et al., 1990, Douglas et al., 1992). All these factors combined with the lack of controlled studies have resulted in an inability to define a particular process or coexisting processes which may be responsible for the post ventilatory weakness. Nonetheless, the development of sepsis has been linked to the polyneuropathies, which have been termed critical illness polyneuropathies (CD?) characterized by primary axonal degeneration of motor and sensory fibers. Histological examinations reveal muscular atrophy and nodal and ultraterminal nerve terminal sprouting along with enlargement of motor end plates . Following such a neuropathy, re-innervation appears to be preferential to upper limbs first, followed by lower limbs and finally the diaphragm. The development of the neuropathy in and of itself will lead to ultrastructural changes at the neuromuscular junction, as previously mentioned. However, the administration of muscle relaxants and/or steroids to patients with neuropathies may in fact further aggravate the insult delivered to the NMJ. On the other hand, there have been other reports of post ventilatory weakness in patients who were not septic but had received muscle relaxants and, in some cases, steroids. That drugs or disease may be affecting other synaptic processes has not been ruled out either. ACh synthesis, storage and release may be especially vulnerable. -31-2.7 a-Bungarotoxin a-bungarotoxin (a-BTX) is a polypeptide ( of ~8000 molecular weight) isolated from the banded krait snake, Bungarus multicinctus, whose venom selectively targets neuromuscular function - it contains AChE, 6 toxins which block ACh release and 4 which block ACh receptors (Miledi et al., 1971). a-BTX has been shown to block irreversibly the action of ACh at the neuromuscular junction without affecting the resting or action potentials in the nerve or muscle, the release of ACh or the activity of AChE ((Lee et al., 1963; Lee, 1972; Berg et al., 1972; Sarvey et al., 1973). Also, Katz & Miledi (1978) showed that a-BTX does not cause a shortening of the lifetime of the open channel of the AChR like d-TC does. Nonetheless, a-BTX binds two types of endplate sites as indicated by the only partial (~50%) protection from toxin binding by d-tubocurarine (Porter et al., 1973). Parallel evidence was provided by the complementary partial protection from the toxin's binding by perhydrohistrionicotoxin, which affects the ionic conductance modulator properties of the postsynaptic membrane (Albuquerque et al., 1973). And unlike d-TC, a-BTX has been shown to bind both junctional and extrajunctional AChRs to the same extent (Albuquerque et al., 1973; Lapa et al., 1974). -32-3. MATERIALS AND METHODS 3.1 Experimental Models of Reduced Postsynaptic Sensitivity 3.1.1 a-Bungarotoxin Treatment Lyophilized a-BTX (Sigma, Mississauga, O N ) was reconstituted with distilled water. Aliquots were subsequently diluted with 0.9% saline to a concentration of 53.75xl0"8 M and frozen such that only the amount of toxin to be injected per day was thawed (thus avoiding repeated cycles of thawing and freezing). Toxin solutions were allowed to reach room temperature before they were injected. Injections were made intraperitoneally at 0.1 mL/10 grams weight from the 53.75x10"8 M solution. Control animals received saline injections of similar volume which were shown not to affect parameters we were measuring. Mice were injected on alternate days at the same time for three weeks. The optimal dosing regimen was determined to be 4 injections of 0. lmL/lOg followed by 65% of that dose, 75% of that dose, 85% of that dose, 100% and so on. As will be later discussed in Discussion, changes in dosing appear to parallel the time course of development and degradation of extrajunctional AChRs which in effect act as a sink for injected toxin without directly affecting neuromuscular transmission and the viability of the animal. 3.1.2 Mini-Osmotic Pumps In addition to the a-bungarotoxin injections we attempted the use of Alzette mini-osmotic pumps (Alza Corp., San Paolo, CA) to deliver non-depolarizing muscle relaxants, with or without prednisolone, via a constant infusion method. Such a technique would minimize the stress - 3 3 -experienced by the animals in response to repeated handling injections. After performing dose-response experiments with intraperitoneal injections of vecuronium, pancuronium and d-tubocurarine, we established the maximum tolerable dose (per hour). Knowing the delivery rate of the mini-osmotic pumps (0.15 uL/hour), we calculated the required amount of muscle relaxant for a two week period (for which the pumps were designed). Animals were anesthetized with pentobarbital, their upper backs shaved and a 1.5 cm midline incision was made starting 0.5 cm from the bottom of the neck. Using blunt dissection, tissue was teased away to form a "pocket" into which the pump was inserted, delivery port first. The incision was sutured with 6.0 silk suture and the animals allowed to recover. However, none of the animals developed any signs of weakness with vecuronium, pancuronium, or vecuronium and prednisolone throughout the two-week period (as measured by the roto-rod test; see next section). Furthermore, miniature end plate potential frequency, measured from these diaphragms, showed no signs of treatment-induced changes, compared to controls, in terms of mean value or standard deviation. In retrospect, the concentrations of the administered agents could have been increased in the osmotic pumps to determine whether lack of effect was due to insufficient dose. Nonetheless, and in light of these negative results, the mini-osmotic pumps were abandoned as an experimental model in favor of a-bungarotoxin injections. 3.2 Strength and Weight Monitoring Animals were weighed and their strength tested on a daily basis. Strength was assessed by placing the animal on a rotating rod (roto-rod) and recording the time it was capable of sustaining itself without falling. Normal mice are capable of staying on the roto-rod for several minutes. Both weight and time spent on the rotating rod were used to assess the extent of weakness due to - 34 -toxin treatment and to adjust the subsequent does. Control ariimals were handled in a similar manner. The animals did not begin exhibiting signs of weakness until several hours after an injection. During the postinjection 24-hour period treated animals lost weight compared to their pre-injection weight. Control animals steadily gained weight throughout the three-week period. 3.3 Phrenic Nerve/Hemidiaphragm Preparation All experiments were performed on the mouse diaphragm preparation, Under halothane anesthesia, the left hemidiaphragm, and part of the left phrenic nerve, were quickly dissected out and immediately placed in a bubbled, modified Krebs solution. Excess tissue was trimmed from both diaphragm and nerve and the preparation subsequently pinned out on a Sylgrad disc. Care was taken not to cut the muscle fibers from the rib cage (that causes a change in their resting membrane potential), not to damage the nerve (by pinching stretching or cutting) and to ensure that every preparation was stretched to similar extents such that the tension on the muscle was comparable. The disc was then mounted and superfused with physiological solution at room temperature using a system allowing fast (~ls) exchange of solution at superficial junctions (Cooke & Quastel, 1973) and visualization using a light microscope. This system of rapid solution changes was used for the experiments in 15 mM K + in which MEPP frequency was measured in six Ca2+ from each junction. Since the order in which the calcium concentrations was delivered was random, whenever a switch was made from a higher to a lower concentration the preparation was washed with a 1 mM EDTA/50uM Ca2+ solution (to "mop up" residual Ca2+) and allowed to equilibrate at the new [Ca2+]0 before recording was commenced. Conventional intracellular recording with microelectrodes filled with 3M KC1 was used. The phrenic nerve was stimulated by a suction glass electrode, into which the nerve was sucked, -35-and which contained a AgCl-coated wire connected to the stimulator. Prior to the commencement of each experiment the stimulation threshold for the nerve was determined and subsequent stimulation was carried out at 3x the threshold. 3.4 Random Sampling A consistent attempt was made to sample the diaphragm for junctions in a random fashion. It is true that areas next to the main branches of the phrenic nerve are more abundant in junctions than'distal portions of the diaphragm, especially next to the rib cage or near the central tendon region. Therefore, within the junction-rich areas, sampling was random. Attempts at locating junctions were started at one end of the diaphragm and proceeded horizontally to the other end. Several such passes would be made across the diaphragm throughout the experiment without concentrating at any one region more than others. 3 . 5 Physiological Solution Standard solution contained (mM): 150 Na+, 5 K+, 24 HC0 3\ 125 Cl", 1 H2P04", 11 glucose, bubbled with 95% 0 2 - 5% C0 2 with added Ca2+ and Mg 2 + as required. For/m experiments we used 15 mMK+, 1 mMMg2+ and a series of Ca2+ concentrations (mM=0.25, 0.5, 1, 2, 4, 8). For the nerve stimulation experiments we used 0.5 mM Ca2+ and 4 mM Mg2+. In such a solution EPP's were subthreshold in normal mouse diaphragm and muscle twitching upon nerve stimulation was therefore eliminated. -36-3.6 u.-Conotoxin Although normal diaphragm muscle did not twitch in our 0. 5mMCa2+/4mMMg2+ solutions (due to the reduced EPP quantal content with these external ionic concentrations), the treated ones did (perhaps due an increase in quantal content). Therefore, to eliminate muscle action potentials we incubated all preparations (control and treated) in 1.2 mM |i-conotoxin (Sigma, Mississauga, ON) for 60 minutes prior to the commencement of the experiment. During this period, the superfusion was stopped and the bath bubbled with 95% 0 2 - 5% C02. jx-Conotoxin, or geographutoxin II, is a 22 amino acid peptide toxin isolated from the venom of Comis geographus that inhibits the contraction and abolishes the action potential of skeletal muscles by preferentially blocking muscle Na+ channels (Hong & Chang, 1989). Thus the toxin can be used to paralyze skeletal muscle without affecting motor nerve terminal function. 3.7 Data Acquisition and Analysis MEPP frequency and EPP quantal content were monitored on-line using a PC program with a continuous print-out of the recorded data at 10 second intervals. All data were also recorded on standard VHS videotapes using a PCM-1 (Medical Systems Corp.) manifold and subsequent off-line analysis was done on a PC. For a more accurate count of MEPPs and quantal components of EPPs we used a computer program, written in "C" and Assembler languages and executed on an AT-386 microcomputer, which in effect deconvoluted each section of the digitized record by serial identification of "bumps" (corresponding to individual MEPPs), within a certain range of shape, followed by subtraction of a "template" derived by averaging isolated MEPPs. Another program identified failures to provide an average stimulus artifact and field potential to be subtracted at - 3 7 -appropriate points in the record (Bain & Quastel, 1992a). The accuracy of this program had been previously checked by superimposing artificial MEPPs, with amplitude variation in the normal range, on typical recorded noise, of varied amplitude. Subsequent programs were written to provide the number and size of MEPPs, their time of occurrence, the size, quantal content and variance of EPPs. Other programs were used to average the 60 trains of stimulation delivered at each junction and to output average quantal content per pulse, variances, and histograms of release following each stimulus. All subsequent data analysis and transformation was done on Microsoft Excel 5.0. When the variability of the data between diaphragms of a group (control or treated) was found to be less than that within those diaphragms, data were normalized and pooled. By doing so, we increased our sample size from 4 (treated) and 8 (control) diaphragms to ~120 junctions in each. Pooled data was used to construct cumulative distribution curves or histograms and to calculate the variance of the data with F-tests. Statistical comparisons of means were done on diaphragm means with t-tests or paired t-tests. 4. R E S U L T S -38-4.1 Miniature End Plate Potentials The first aspect of transmitter release to be measured was miniature endplate potential frequency (fm) in both control and treated diaphragms; The reason for this choice was that/„ can be easily and rapidly measured at large numbers of junctions, especially in a high K + (15 mM) (modified) physiological solution which causes a considerable increase in MEPP frequency. fm s were measured in a series of calcium concentrations (0.25, 0.5, 1, 2, 4, and 8 mM) in the presence of 15 mMK+ (to depolarize nerve tenriinals and increase rate of spontaneous release, fm) and 1 mM Mg 2 + . Muscle fibers were sampled in a random fashion making several passes across the diaphragm until the required number of junctions had been achieved. Although the actual neuromuscular junctions could not be visualized under the light microscope, the general vicinity in which they tend to be clustered with respect to the main branch of the phrenic nerve was known. To be able to accurately describe^ distributions 200 junctions were sampled from each diaphragm in both 0.5 and 8 mM Ca2+ and 30 junctions in each of the other 4 Ca2+ concentrations. The distributions were chosen to be constructed from the 0.5 and 8 mM data for two reasons: (1) in control diaphragms, logfm is about the same at these two Ca2+ concentrations and it was desirable to find out whether treated diaphragms behave in a similar manner; (2) junctions may behave differently at low than at high [Ca2+]e and 8 mM is considered to be well above the level at which_/5n is maximal and release should be unaffected by any variation in the "EC50" of Ca2+. First, it was found that/m was stable over the few hours it took to sample the 200 junctions. Mean \ogfm was plotted against the time at which the junction was sampled (obtained -39-from the computer printout along with/,,) and was found to be a straight line (with a slope of zero) for all diaphragms (data not shown). That was an indication of the sustained viability of the preparation over the duration of the experiments; Lapa et al. (1974) reported that the mouse diaphragm preparation remains viable for over 15 hours without developing contractures, impairment of neuromuscular transmission or significant changes in resting membrane potential or membrane excitability. However, that MEPPs recorded from treated diaphragms were smaller in size (amplitude) than those from control diaphragms, made it more difficult to capture signals from all sampled junctions. In some instances, MEPPs could be just recognized on the monitor but were not picked up by the acquisition program. It is also possible that some MEPPs were invariably missed because they were inextractable from baseline recording noise (however, see ethanol results as a means to ascertain that there were no sampling biases when recording from treated diaphragms). Another interesting observation was that a considerable number of fibers sampled in treated diaphragms were "silent"; that is, no MEPPs could be observed. It is true that if recording is from a point in the muscle fiber too far from the neuromuscular junction MEPPs are too small to be discerned; however, the frequency with which such silent fibers were encountered might reflect a true physiological phenomenon; i.e., certain fibers might be completely blocked by the toxin such that there was no postsynaptic response to released ACh, or transmitter release might be so reduced that no MEPP's are observed. Although the number of such fibers varied between different diaphragms it was roughly about 50% of impaled fibers. The actual numbers were not recorded mainly because it is practically impossible to differentiate between a silent fiber and distant impalement of a fiber causing recorded MEPPs to be too small to be picked out of the baseline noise (see Discussion section for further elaboration of potential relevance of such silent fibers). -40-Thefms in the different [Ca2+]0 were used to construct Ca2+ titration curves to compare the responses of treated and control junctions to the varying Ca2+ concentrations. Figure 1 illustrates the relationship between fm and [Ca2+]0, plotted as logfm vs. log [Ca2+]0, at both control and treated junctions. Such group (treated, control) titration curves were constructed using the means logfm of each diaphragm at each [Ca2+]0. (Note that the number of junctions sampled per diaphragm was 200 at 0.5 and 8 mM Ca2+.) There were no statistically significant differences in mean logfm between treated and control animals (see Tables 1A and B for means, p-values, SDs, and sample sizes) at the different [Ca2+]0. Because in both groups, the variability in log./™ between diaphragms was found to be significantly greater than that within (Table 2; p<0.01) t-tests were performed on the means for each diaphragm. At 8 mM Ca2+, the difference between mean logfm for treated and control was 0.107 (t=2.18, p=0.05). This indicates a p<0.05 that a true difference was more than 0.2. On the other hand, the variance in logfm of the pooled data (which was normalized with respect to its respective diaphragm mean) was found to be significantly different between the two groups (treated greater) at 1, 4, and 8 mM Ca2+, but not at 0.25, 0.5 and 2 mM Ca2+. Treated diaphragms may in effect contain two sub-populations of junctions, one of which has been altered by the a-bungarotoxin treatment, and exhibits altered Ca 2 + sensitivity, while the other has not (thereby accounting for the increased variability in log/m within any one treated diaphragm). In that case, the distribution of \ogfm within a diaphragm may shed some light on such possible sub-populations either deviating from that of control diaphragms, perhaps by a bimodal presentation. Therefore, cumulative distribution curves of log fm in 8 and 0.5 mM Ca2+ were constructed for each diaphragm, as well as for pooled data, and plotted along with a -41 -• Treated o Control _ 1 . . 1 1 10 Ca 2 + Concentration (mM) Figure 1. Log-log plots of Ca2+ titration curves for treated and control groups (n=5 diaphragms each; n=200 junctions in 0.5 and 8mM Ca2 + and 30 junctions in all other Ca2+concentrations). Values plotted are averages (± sem) of dia-phragm means per group .Bathing solution contained 15mM K + and ImM Mg2+.No significant differences between treated and control group means were at any [Ca2+]0 (t-test; p»0.05). Numbers above data points indicate [Ca2+]0 at which each set offm were measured. 1 4-0.1 sr a S 3 cr a -t o s a 5 3 «3 a cu CO o sr o o 3 CD a Cu CL ET 3 CD ft <• Ln .fe LO S J <—i 3 p o © -fe. S J - fe L O VO O o O L O s> L O -~J -fe O S 3 -fe ON L O L O S 3 L O P> J O J 3 0 J © " S J " S J "LO " L O 0 0 VO O S 3 p o o o S 3 S 3 S 3 VO S 3 LO > — * 0 0 S 3 - J o o O O S 3 -fe Lo S 3 LO o VO VO 0 0 oo o LO LO S 3 S 3 vP vP J 3 0 J * 0 "sJ " S 3 " S 3 "LO 0 0 VO U l o O o O O S 3 S 3 S 3 S 3 S 3 - fe O - J Ui O N so S 3 LO O O O O O ' . fe LO LO '.fe 0 0 - J LO LO S 3 Ln S 3 Os K 3 O ^ ^ © ° ' p o sS 53 LO Os S 3 -fe S J 0 0 Ul ^ W S) M 3 © © © © © '«— S 3 O N LO - J so SO i — • O i — 1 VO Os ~ J 0 0 O © O Ln S 3 S O SO LO S J LO S 3 LO O O O U l M K) W LO © © o O O Lo Lo • f e Lo S 3 - J S 3 LO LO SO • f e >—» S J Ln © © o o Lo Ln LO Lo o Os • f e S 3 LO Ln o S J ~ g -o VO 2 vo VO ON J-* vo ^ "vo ">- S S 3 o p O o S 3 Lo LO S 3 Ln LO • f e o 0 0 Ln O o p O o S 3 Lo Lo • f e -fe S J Os Ln 0 0 -o --J LO LO S 3 S J © I—» VO oo " S J "LO " S J " S J 0 0 o ON VO H 5 Cu o 5 3 </> O 1^ P 3 i c o 8 CD 3> Cu O 3 T3 5 3 I I a a 3 O o 3 a w Cu OQ O c a t» o 3* r — - i O o Ln -fe L O . S 3 — © © © © p 'vo ife ON oo O N Ln -fe Ln i—• Ln Ul A v) e U t — o S 3 f^e Ul VO Os ~~J i— O >— <i o SO i— S J ON >— vo L O i—' 0 0 _ o ^ O LO Ln S 3 VO -fe 4^. O - J LO S J S J S J LO o *—' Os Os so so '—1 0 0 ON U I S J LO O fe. S J ON S J S J Ui O 0 0 ON S J i—' S J Ul -fe 0 0 0 0 - ^ J fe. — O N O N -fe LO S 3 Ul Ul — O LO O S 3 fe. O O N S 3 -fe Ul S 3 Ui ON ON >—> O i — 1 0 0 U i Ul O Ul fe. 0 0 0 0 i — LO S J LO fe. O O i — o o ON S J Ui 0 0 p—> LO Ul Ul LO 0 0 S ) H - S 3 O O LO SO 0 0 S 3 S 3 ~-J — o S 3 Ul O Ln to 0 0 • f l "> < o II a*> O < O (6 5 g o y o 0 1 f 13 0 1 3 O o a g a. a cu w 3 O ro CO 3 a -a o o CD o. P> 3> rt - i 3 O I o 3 a* s cr § 3 OQ a. T 3 09 OQ 3 3 0> 3 O OQ o 3 tr & o r> 3* t .. C 3 o r+ 5' 3 ro < T , a V — © un oo 6r v _ o to VO o to o U J p w O ON O to V ~ o ~ VO 00 00 *0 o a. r to 00 u> M H K S2. co' O 3. w 09 o 3-o <8 2 S lfi XA Cjg o o 3 e.E.2 - S c e o 3- Pi £ on aa. 1 ET. 3 C U cr I a a 3, C U O a. " 3 co C o £ o vo b b o oo A o b ON o o un Is o Zl 8 8 A -VO 3 ffi 3 O O 00 KJ v l —' - J IO W <J Vl Wl ^, VO 00 © © to ON U J Ul S l M — O U J to VO p © © '>—• b In i—• vo un VO 00 oo un p © I—• '>—• bo ON to 00 O0 v l w KJ ^ UJ O - J © © i— N - 00 v l o\ w A C \ vJ ON ON I — » H - U) © p to ~ i—• Ul oo ON un to VO U J U J U J to U J a* c a b o ~ J un A ^ P w o 2 A o ^1 § to A -P ON 2 O N A ^ U J ^ vo ° vo o _ sr B: — 3. O O O to to vo to 00 t o o -W W ON to o o — b -~J U J vo ON vo 00 VO H - » I—» U J p o o b H - VO U J to o I—» H - U J v l -fc* p p p b b vo vJ Ul w oo o\ *• * H - » U J ON U J O O O b b v i v l v] U J •fc. ON O O K H - * O O ON o o un " H - ' H - v l un U J o ON U J 00 VO vo VO vo -fc. 4^ O O l to cn a. • f l • f l • f l 00 o. • f l p In Cu •fl H to o o r o (Q 0 p 1 to n 3 P> 3 Cu 3* 5" ro 3 o >-« O o 09 3 Cu c ' O on 09 ff-W 09 o 3* O -44-theoretical, normal distribution, of a data set with similar mean and standard deviation as that of each respective diaphragm. (For the pooled data, normalized logfm were plotted with a theoretical curve of mean equal to zero and SD equal to that displayed by the log/m data.) Two such distributions, of the pooled data (one treated and one control) in 8 mM Ca2+, are illustrated in Figures 2 and 3. Similar distributions in 0.5 mM Ca2+ are presented in Appendices I and n. The cumulative frequency distributions for treated diaphragms show no signs of bimodal or multimodal distribution. That is, as with the controls, there appears to be a log normal distribution. However, and as previously mentioned, the spread of the distributions was found to be significantly different at 8 mM Ca2+ (Table 2). Whether this increased variability was due to the toxin treatment or to recording difficulties resulting from the reduced MEPP amplitude was addressed next. Since it was found that MEPP amplitudes were smaller in treated diaphragms, it was reasoned that some MEPPs may have been indiscernible from recording baseline noise. Therefore, similar experiments were conducted in another batch of animals (n=3 each) in the presence of 4.8% ethanol (0.4 mM) to test the latter hypothesis. Ethanol, and as expected, prolonged the time course as well as raised the rate and amplitudes of MEPPs. MEPPs were measured solely in 8 mM Ca2+ since it was believed that such a high concentration would minimize the differences between junctions. As Table 3 shows, ethanol treatment decreased the variability within and between both groups to a similar level. Thus the range of log fm became the same for both groups. Table 3, Mean and Standard Deviations (SD) of Control and Treated Log 4 in 8 mM rCa2 +L with 4.8% Ethanol (0.4 mM) and 10 mM K + . 8 mM[Ca2+]o Control Treated Diaphragm 1 2. 3 1 2. 3 Mean 1.595 1.594 1.485 1.512 1.630 1.418 SD 0.244 0.284 0.256 0.224 0.232 0.251 -45-Theoretical Deviations from the Mean Figure 2. Top figure: cumulative frequency distribution of pooled* treated logfm in 8mM Ca2 + (n=636 junctions) and of theoretical normal distribution with mean of zero and SD similar to actual log/m data (mean=1.21; SD=0.43). Lower figure: deviations of lcg/m values from their mean plotted against deviations of theoretical distribution values (y = 1.0061x-0.0045). (Similar data in 0.5 mM Ca2+ presented in Appendix I.) * data were normalized with respect to population mean in each diaphragm before pooling. -46-1 T ^ » 0.8 -H e 0.6 -c o ••e 0.4 -u 2 0.2 --2 + Observed Log(fin) Theoretical Values •1.5 -1 -0.5 0 0.5 1 1.5 Observed (& Theoretical) Deviations from the Mean 1.5 T -1 M Theoretical Deviations from the Mean Figure 3. Top figure: pooled* cumulative frequency distribution of control logfm in 8mM Ca2 + (n= 1015 junctions; 5 diaphragms) and of theoretical normal distribution with mean of zero and SD similar to that of logfm data set (mean=1.07; SD=0.26). Lower Figure: deviations of \ogfm values from their mean plotted against the deviations of the theoretical distribution (y = 1.0053x-0.0004). (Similar data in 0.5 mM Ca2+ presented in Appendix U.) * data were normalized with respect to population mean in each diaphragm before pooling. -47-However, with ethanol the logfm distributions deviated from normality to become skewed with a distinct group of/mS appearing in the left end of the distribution (Figures 4A and B, for control and treated diaphragms, respectively). In effect, the ethanol treatment did uncover a population of junctions that had not been discernible in its absence, but it did so for both groups equally. Therefore, even if heterogeneous populations of muscle fibers do in fact exist in treated diaphragms, the experimental methods used were not capable of verifying that. Moreover, that the observed increase in variance of treated fm was found at the higher Ca2+ concentrations (1,4, and 8 mM), all of which retained a log normal distribution, suggests that this result is in fact true, and not merely a consequence of biased sampling. Therefore, that distinct muscle fiber populations may exist could not be verified. It is also plausible to interpret the results as emanating from an altered sensitivity to Ca2+ of treated junctions in high K +, especially since the variability was found to increase with increasing [Ca2+]0. Since variance was not significantly different at 2 mM Ca2+, and if that concentration is considered an approximation of in vivo levels, that may also lend further indirect support to the hypothesis of altered Ca2+ sensitivity. 4.2 Measurement of Quantal Content with Nerve Stimulation In a different batch of animals, receiving treatment identical to that of the previous batch, nerve stimulation was performed (as discussed in Methods) at 10 Hz and 70 Hz (trains of 10, repeated every second) so that evoked release (quantal content of end plate potentials, EPPs), in addition to spontaneous release (fm), could be measured. The number of junctions sampled in each diaphragm ranged from 10 to 40. The 10 Hz stimulation period lasted for 30 seconds and was designed to produce a "baseline" measurement of quantal content with respect to the 70 Hz train-of-ten stimulation (which was delivered for 1 minute, a few seconds after the end of the 10 Hz period). The reason -48-© u .Q s 9 30 "125 Control e • B20 SB * J 1 0 * 0 5 + 0 0.96 1.09 1.22 1.35 1.48 1.61 1.73 1.86 1.99 2.12 Upper Limit of Bin Value (Log f„) 0.80 0.92 1.05 1.18 1.31 1.44 1.57 1.69 1.82 1.95 Upper Limit of Bin Value (Log fm) Figures 4A&B. Histograms of log/m in 8mM Ca2+ (lOmM K +, ImM Mg2+) with 4.8% ethanol. Top figure: control diaphragm (n = 100 junctions). Lower figure: treated diaphragm (n = 100 junctions). Note the skewed nature of the distributions in ethanol, the similarity between treated and control in terms of mean and standard deviation. -49-for requiring such a baseline value is that train stimulation at 70 Hz invokes the phenomenon of facilitation (as discussed in the Background), which itself may have been affected by the toxin treatment independently of any other possible increase in quantal content. Therefore a measure of quantal content independent of facilitation was desirable. MEPPs were also recorded for 120 seconds before and after the stimulation periods, as well as during. In these experiments the physiological saline solution contained modified concentrations of Ca2+(0.5 mM), K+(5 mM), and Mg 2 + (4 mM). These concentrations were chosen to produce quantal contents of about 1; i.e,. a level at which quantal content may be determined fairly accurately. The reason for not measuring EPPs in normal concentrations of Ca2+ and Mg 2 + was the very large potential error associated with such measurements when quantal content becomes high; with high Ca2+ concentrations there will be no failures (i.e., lack of an EPP in response to a nerve stimulus) which renders it impossible to subtract the stimulation artifact and extracellular field potential from the EPP, and there is the added complication of non-linear summation, which can be compensated for only if recordings are exactly focal. On the other hand, in lower Ca2+ concentrations, release is too low to allow accurate determination of its magnitude without stimulating for very long periods of time. An additional advantage of our low Ca2+/high Mg 2 + solution is that such conditions cannot sustain a muscle action potential, thereby obviating the complication(s) of muscle twitching. However, the first striking observation was that treated diaphragms did in fact twitch in the above solution (while the stimulation threshold was being determined). That was the first clear indication that the a-bungarotoxin had indeed affected a change in the physiology of neuromuscular transmission at the mouse diaphragm. To block these action potentials (which obstruct electrophysiological recordings by causing muscle twitching and the subsequent displacement or breaking of the glass microelectrode tips), the preparations were incubated in p> conotoxin (1.2 mM; as discussed in Methods), a skeletal muscle Na+ channel blocker, for about -50-an hour. This toxin was chosen because it does not interfere with presynaptic electrical conduction. Usually, such an incubation prevented action potential firing for the duration of the experiment. 4.2.1 Measurement of End Plate Potential Quantal Content with 10 Hz Stimulation Much like the \ogfm data, the actual (non-log) quantal contents of EPPs elicited with the 30 second, 10 Hz stimulation period exhibited greater variability within, and between, treated diaphragms compared to controls. However, the variability between diaphragms exceeded that within for both groups (Table 4A, analysis of variance). Additionally, the actual mean treated 10 Hz quantal content (mio) was about double that of the control, but the data could not be subjected to a t-test because it was found to be non-normally distributed (Appendix III). Log transforming the data made it amenable to parametric statistical testing and it was found to be significantly different between the two groups (p<0.01; Table 4B). The variability of the log quantal contents was not significantly different between the treated and control groups; neither within nor between diaphragms (Table 4B), although between- diaphragm variance remained greater than within-diaphragm, for both groups. An F-test performed on the variances of the pooled data (after normalizing with respect to diaphragm means; n=131 and 122 junctions for control and treated, respectively) revealed that it too was not significantly different between the two groups (p>0.1). The distributions of log mio were shown to be normally distributed for each diaphragm as well as for the pooled data for controls and treated (Figures 5A, B and 6A, B). That variance of log/m in 15 mM K + was (significantly) different between controls and treated (at some, but not all, [Ca2+]0) but mio was not, suggests that the conditions (cation concentrations, spontaneous vs. evoked -51-Table 4A. Actual 10 Hz Quantal Content: Descriptive Statistics and Analysis of Variance Control Treated Diaphragm Mean SD n1 Mean SD n1 1 0.502 0.189 13 1.057 0.561 36 2 0.531 0.303 17 1.551 0.74 15 3 0.327 0.178 27 0.748 0.436 39 4 0.296 0.13 13 0.660 0.401 32 5 0.323 0.202 23 6 0.691 0.363 21 7 0.968 0.655 10 8 0.389 0.323 7 Group2 0.503 1.004 Pooled3 0.478 0.351 131 0.915 0.581 122 n=number of junctions sampled in respective diaphragm, mean of diaphragm means (ncontroi^ , ntre,ted=4). data were pooled after normalization with respect to diaphragm means, for calculation of SD. Analysis of Variance of 10 Hz Quantal Content. dF SS MS Control Between 7 5.058 0.723 8.12!! Within 123 10.941 0.089 Total 130 15.999 Treated Between 3 9.963 3.321 12.69!! Within 118 30.860 0.262 Total 121 40.822 dF: degrees of freedom; SS: sum of squares; MS: mean square. F # = (MS Between / MS Within). !! p<0.001 - 5 2 -Table 4B. 10 Hz Lop Quantal Content: Descriptive Statistics and Analysis of Variance. Control Treated Diaphragm Mean SD n1 Mean SD n1 1 -0.337 0.208 13 -0.028 0.213 36 2 -0.330 0.22 17 0.148 0.198 15 3 -0.548 0.253 27 -0.185 0.223 39 4 -0.572 0.208 13 -0.251 0.253 32 5 -0.563 0.255 23 6 -0.222 0.242 21 7 -0.096 0.278 10 8 -0.558 0.418 7 Group2 -0.403 -0.079 t# 2.718* 10(dF)3 Pooled J7J -0.417 1.29 0.294 131(dF) -0.115 0.259 122(dF) n=number of junctions sampled in respective diaphragm. 2 mean of diaphragm means (rw^S, ntreatcd=4). dF=degrees of freedom. t# t-test of difference of means between treated and control; * p<0.05. F s F-test of difference of variances between treated and control (p>0.10). Analysis of Variance of 10 Hz Log Quantal Content. dF SS MS p.$ Control Between 7 3.442 0.492 7.740!! Within 123 7.814 0.064 • Total 130 11.255 Treated Between 3 2.089 0.696 13.660!! Within 118 6.016 0.051 . Total 121 8.105 dF: degrees of freedom; SS: sum of squares; MS: mean square. F $ = (MS Between) / (MS Within), !! p<0.001 -53 -M e o « es e u JO 35 30 25 20 + 15 10 5 0 • Control H 1 h a jo SB t 9i cn o o CU s s -2.3 40 35 30 -25 -20 -15 10 + 5 0 -2.1 -1.9 -1.7 -1.5 -1.3 -1.1 -0.9 -0.7 -0.5 -0.3 -0.1 0.1 OJ 0.5 0.7 0.9 1.1 13 1.5 Middle of Bin Value (Log mio) I Treated H h -2.3 -2.1 -1.9 -1.7 -1.5 -1.3 -1.1 -0.9 -0.7 -0.5 -0.3 -4).l 0.1 03 03 0.7 0.9 1.1 13 1.5 Middle of Bin Value (Log mio) Figure 5. Histograms of pooled log 10 Hz quantal content in 0.5 mM Ca2+, 5 mMK+, 4 mM Mg . Top figure: control group (8 diaphragms; 131 junctions, mean=-0.417; SD= 0.245). Bottom figure: treated group (4 diaphragms; 122 junctions; mean=-0.115; SD=0.223). Both distributions are normal; however, treated have a significantly higher quantal content (p«0.05). Moreover, the two pooled populations are homoscedastic (F-test; p>0.1). Also, the variance in log m i 0 was greater betwen diaphragms than within, for both controls and treated. Individual diaphragms exhibit the same results. -54--0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Observed (& Theoretical) Deviations from the Mean -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Observed (& Theoretical) Deviations from the Mean Figures 6 A and B. Top figure: pooled cumulative frequency distribution of control log mio deviations from the mean (131 junctions, mean=-0.417, SD=0.245) and of deviations of a theoretical normal distribution with mean of zero and SD similar to that of log mi0 data. Lower figure: pooled cumulative frequency distribution of treated log mio deviations from the mean (122 junctions, mean=-0.115, SD=0.223) and of deviations of a theoretical normal distribution with mean of zero and SD similar to that of log mio data. 10 Hz stimulation delivered in 0.5 mM Ca2+, 5 mM K +, 4 mM Mg 2 + (preparation incubated in 1.2 mM p>conotoxin for 60 minutes prior to experiment). -55-release) under which release is measured may either unmask or confound treatment-induced differences. A discussion of the possible role of minEPPs in the maintenance of the neuromuscular junction and in response to a-bungarotoxin is presented in the Discussion section. 4.2.2 Measurement of End Plate Potential Quantal Content with 70 Hz Trains. The 10 Hz stimulation was followed by a 60-second train-of-ten stimulation delivered at 70 Hz (as previously explained).. In agreement with the mi0 data, actual 70 Hz train quantal contents (m7o) were greater in the treated diaphragms (about double) and exhibited a greater variance as well (Tables 5 A and B). Also in accord with the mio results, log transforming the m70 data yielded (1) almost identical variance within and between diaphragms for both groups (Table 6A), (2) significantly greater variance between diaphragms than within, in both groups, and (3) normal distributions of the log of average quantal contents for the whole train (Figures 7 A and B). (Appendix IV illustrates the non-normal distributions of actual Table 5 A mean train quantal contents.) These observations are true for log quantal content of the entire train (average m7o of the 10 pulses) as well as for mean log quantal content at each pulse (Table 6B). A plot of mean log m70 vs. pulse number illustrates the clear and significant increase in quantal contents in treated diaphragms (Figure 8). Having established that treated diaphragms had greater mio and m7o values (than controls), I was interested in the relationship between mio and m7 0, and whether that was altered in response to treatment. To investigate that, m70 data were normalized either by their respective mio values or by the mean quantal content of their respective trains. -56-Table 5A. Mean Actual 70 Hz Train Quantal Content Descriptive Statistics and Analysis of Variance. Control Treated Diaphrag m Mean SD n l Mean SD n 1 0.869 0.386 13 1.767 0.957 36 2 0.798 0.325 17 2.942 1.85 15 3 0.540 0.29 . 27 1.209 0.741 39 4 0.464 0.18 13 1.186 0.738 32 5 0.509 0.289 23 6 1.048 0.549 21 7 1.526 1.17 10 8 0.467 0.533 7 Group2 Pooled3 0.778 0.746 0.557 131 1.776 1.581 1.14 122 1 n=num 2 mean o 3 data wc Der of junctions sampled in respective diaphragm. f diaphragm means (ncontroi^, nueated^). ;re pooled after normalization with respect to diaphragm means. Analysis of Variance of Train Quantal Content. dF SS MS Control Between 7 12.256 1.751 7.683!! Within 123 28.032 0.228 Total 130 40.288 Treated Between 3 39.406 13.135 13.174!! Within 118 117.65 0.997 Total 121 157.057 . ' - ' . dF: degrees of freedom; SS: sum of squares; MS: mean square F $ = (MS Between)/(MS Within), !! p<0.001 -57-Table 5B. Actual 70 Hz Train Quantal Content Descriptive Statistics and Analysis . of Variance. CONTROL Mean SEM MS Btwn1 MSWthri2 F!! Pulse 1 0.435 0.032 0.596 0.109 5.446 Pulse2 0:550 0.037 1.101 0.123 8.919 Pulse3 0.624 0.041 1.224 0.162 7.575 Pulse4 0.710 0.051 1.931 0.248 7.800 Pulse5 0.758 0.051 1.946 0.249 7.820 Pulse6 0.809 0.056 1.979 0.327 6.046 Pulse7 0.862 0.059 2.502 0.336 7.454 Pulse8 0.858 0.052 1.893 0.261 7.250 Pulse9 0.929 0.059 2.639 0.337 7.825 . PulselO 0.925 0.060 2.479 0.350 7.074 Train3 0.746 0.049 1.751 0.228 7.683 Controls N= 131 dF=7(Between Groups), 123 (Within Groups). ! ! F = (MS Between)/(MS Within), p<0.001 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. TREATED Mean SEM MS Btwn1 MS Wthn2 F!! Pulse 1 0.910 0.064 3.301 0.424 7.789 Pulse2 1.129 0.077 6.094 . 0.591 10.314 Pulse3 1.310 0.088 8.048 0.767 10.486 Pulse4 1.468 0.099 11.525 0.924 12.476 Pulse5 1.589 0.102 13.587 0.961 14:145 Pulse6 1.716 0.115 14.492 1.295 11.190 Pulse7 1.828 0.120 17.178 1.372 12.518 Pulse8 1.886 : 0.120 17.954 1.350 13.299 Pulse9 1.929 0.125 23.128 1.357 17.037 PulselO 2.042 0.137 25.839 1.696 15.239 Train3 1.581 0.103 13.135 0.997 13.174 Treated N= 122 dF=3(Between Groups), 118(Within Groups). !! F = (MS Between) / (MS Within), p<0.001. 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. -58-Table 6A. Mean 70Hz Log Train Quantal Content Descriptive Statistics and Analysis of Variance. Control - Treated Diaphragm Mean SD n1 Mean SD n 1 -0.099 0.192 13 0.194 0.217 36 2 -0.137 0.199 17 0.403 0.237 15 3 -0.326 0.235 27 0.022 0.219 39 4 -0.367 0.186 13 -0.003 0.263 32 5 -0.358 0.242 23 6 -0.037 0.231 21 7 0.091 0.288 10 8 -0.496 0.379 7 Group2 t# -0.216 2.769* 10(dF)3 0.154 Pooled -0.220 0.283 131 0.113 0.267 122 1.12 n=number of junctions sampled in respective diaphragm. 2 means of diaphragm means (ncontroi=8, t W e a ^ ) . 3 dF=degrees of freedom. t# t-test on difference of means between control and treated, * p<0.05. F $ F-test on difference of variances between control and treated (p>0.10). Analysis of Variance of Log Train Quantal. dF SS MS Control Between 7 3.524 0.503 8.988!! Within .123 6.89 0.056 Total 130 10.414 Treated Between 3 2.249 0.750 13.831!! Within 118 6.395 0.054 Total 121 8.644 dF: degrees oi •freedom; SS: sum of squares; MJ5: mean square. F s = (MS Between) / (MS Within), !! p<0.001. -59-Table 6B. Log 70 Hz Train Quantal Content Descriptive Statistics and Analysis of Variance. CONTROL Mean SEM MS Btwn1 MS Wthn2 F!1 J Pulsel -0.473 0.027 0.525 0.074 7.093 Pulse2 -0.366 0.028 0.655 0.068 9.576 Pulse3 -0.305 0.026 0.538 0.065 8.273 Pulse4 -0.259 0.027 0.589 0.070 8.360 Pulse5 -0.218 0.026 0.515 0.062 8.304 Pulse6 -0.200 0.027 0.569 0.071 8.000 Pulse7 -0.168 0.027 0.530 0.069 7.716 Pulse8 -0.155 0.025 0.450 0.059 7.602 Pulse9 -0.128 0.027 0.537 0.068 7.894 PulselO -0.127 0.025 0.476 0.058 8.147 Train3 -0.220 0.025 0.503 0.056 8.988 Control N= 131, dF=7(Between Groups), 123(Within Groups). !! F= (MS Between) / (MS Within), p<0.001 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. TREATED Mean SEM MS Btwn1 MS Wthn2 F!! Pulsel -0.138 0.026 0.626 0.069 9.095 Pulse2 -0.039 0.025 0.679 0.062 10.966 Pulse3 . 0.026 0.025 0.679 0.062 10.889 Pulse4 0.076 0.025 0.730 0.060 12.245 Pulse5 0.117 0.024 0.753 0.053 14.270 Pulse6 0.145 0.025 0.688 0.058 11.908 Pulse7 0.171 0.025 . 0.788 0.059 13.396 Pulse8 0.191 0.024 0.721 0.055 12.996 Pulse9 0.195 0.026 0.933 0.058 16.012 PulselO 0.217 0.025 0.866 0.059 14.779 Train3 0.113 0.024 0.750 0.054 13.831 Treated N=122; dF=3(Between Groups), 118(Within Groups). !! F = (MS Between) / (MS Within), pO.OOl. 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. -60-Observed (& Theoretical) Deviations from the Mean Figures 7A and B. Top figure: pooled cumulative frequency distribution of control log train quantal content deviations from the mean (131 junctions, mean=-0.22, SD=0.23) with deviations of a theoretical normal distribution with mean of zero and Sd similar to that of log quantal content data. Lower figure: pooled cumulative frequency distribution of treated log train quantal content deviations from the mean (122 junctions, mean=-0.113, SD=0.23) with deviations of a theoretical normal distribution with mean of zero and SD similar to that of log quantal content data. Mean log train quantal content was found to significantly greater in treated group (p«0.05). The variance in log train quantal content was found to be greater between diaphragms than within, for either group. However, no significant difference was found in the varaince of the pooled data between the two groups (p>0.1). -61 -0.4 0.2 f •Treated ^ $ I I * 0 + -0.2 -0.4 + -0.6 4 -0.8 o Control i 5 x 5 5 5 -f——- 1 : 1— — i 4 6 8 10 Pulse Number Figure 8. Log 70Hz, train quantal content (1100,^01=8 diaphragms; tWea^ diaphragms -with 10-40 junctions sampled in each). Data presented are means ± sem. Stimulation delivered in 0.5 mM Ca2+, 5 mM K+, 4 mM Mg 2 + (and 1.2 mM |j,-conotoxin). To note: significant increase in treated quantal content but apparent lack of difference in time course of quantal content rise during train between groups. Also, no difference in varaince of log data was found between control and treated groups, at every pulse. But variance between diaphragms, of a group, was found to be consistently larger than that within (for control and treated). - 62 -4.3 Normalization by 10 Hz Quantal Content ( m i n ) The results of normalizing nvro values, by dividing each value by its respective m i 0 value, for each junction, i.e., means, SEMs, analysis of variance, and F values of normalized and log normalized train quantal contents, are presented in Tables 7 and 8, respectively. Although there were no significant differences in the variances of the normalized data (actual or log) between the two groups, and no significant difference in variance between and within diaphragms in either group, the treated mean normalized quantal contents appear slightly greater (Figure 9). This suggests that treated junctions may release slightly more at 70 Hz, with respect to release at 10 Hz, than do control junctions. However, normalization by train means showed no difference in facilitation between treated and controls, see below. It may be noted that the variances of the normalized log data (mean square within, Table 8) do not change with pulse number and are almost identical in both groups. This indicates that (1) any treatment-induced changes were uniform at all junctions, and (2) facilitation may have been uniform at all junctions. 4.4 Normalization by Train Means (rntrain) An alternative way of looking at facilitation was to normalize m70 at each pulse of the train by the mean quantal content of the entire train. Such a normalization in effect allows comparison of the quantal content at each pulse to that of the entire train; i.e., the contribution, and variability of the latter, at each unitary pulse to the overall increase in quantal content throughout the train. Another advantage of normalizing by m^ m rather than, for example, the quantal content of the first pulse, is that the latter is the least accurate measure of release in the data set. This derives from the fact that quantal contents are Poisson distributed, that the quantal contents of the first -63 -Table 7. Actual 70 Hz Train Quantal Content Normalized by 10 Hz Quantal Content: Descriptive Statistics and Analysis of Variance. CONTROL Mean SEM MS Btwn1 MS Wthn2 p$ Pulse 1 , 0.933 0!029 0.122 0.107 1.143 Pulse2 1.185 0.033 0.186 0.138 1.348 Pulse3 1.361 0.036 0.275 0.166 1.659 Pulse4 1.511 0.040 0.417 0.201 2.074 Pulse5 1.659 0.047 0.382 0.279 1.367 Pulse6 1.742 0.056 0.610 0.393 1.551 Pulse7 1.851 0.047 0.331 0.285 1.161 Pulse8 1.898 0.049 0.359 0.316 1.136 . Pulse9 2.041 0.056 0.688 0.391 1.760 PulselO 2.042 0.060 0.355 0.471 0.753 Train3 1.622 0.036 0.269 0.164 1,647 Control N= 131, dF=7(Between Groups), 123 (Within Groups). F $ = (MS Between) / (MS Within), no significant difference.. 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. TREATED Mean SEM MS Btwn1 MS Wthn2 F Pulse 1 0.991 0.028 0.159 0.094 1.692 Pulse2 1.237 0.031 0.119 0.118 1.008 Pulse3 1.435 0.035 0.206 0.151 1.361 Pulse4 1.602 0.036 0.194 0.156 1.240 Pulse5 1.771 0.045 0.312 0.244 1.278 Pulse6 1.877 0.043 0.301 0.226 1.330 Pulse7 1.996 0.046 0.217 0.265 0.819 Pulse8 2:090 0.050 0.325 0.300 1.083 Pulse9 2.121 0.054 0.861 0.345 2.494 PulselO 2.219 0.051 0.504 0.310 1.628 Train3 1.734 0.036 0.241 0.156 1.544 Treated N=122; dF=3(Between Groups), 118(Within Groups). F $ = (MS Between) / (MS Within), no significant difference. 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. Table 8. 70 Hz Log Train Quantal Content Normalized by 10 Hz Quantal Content: Descriptive Statistics and Analysis of Variance. CONTROL Mean SEM MS Btwn1 MS Wthn2 F Pulsel -0.056 0.013 0.041 0.022 1.840 Pulse2 0.052 0.013 0.045 0.019 2.322* Pulse3 0.113 0.012 0.047 0.018 2.538* Pulse4 0.158 0:012 0.050 0.018 2.738* Pulse5 0.199 0.012 k 0.033 0.017 1.961 Pulse6 0.217 0.012 0.061 0.018 3.375! Pulse7 0.249 0.011 0.021 0.016 1.320 Pulse8 0.262 0.010 0.019 0.013 1.402 Pulse9 0.290 0.012 0.052 0.017 3.104! PulselO 0.291 0.011 0.017 0.016 1.059 Train3 0.198 0.009 0.025 0.010 2.563* Control N= 131, dF=7(Between Groups), 123(Within Groups). * p<0.05, F = (MS Between) / (MS Within). ! p<0.01, F = (MS Between) / (MS Within). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. TREATED Mean SEM MS Btwn1 MS Wthn2 F Pulsel -0.023 0.012 0.022 0.017 1.347 Pulse2 0.076 0.01.1 0.013 0.015 . 0.873 Pulse3 0.141 0.011 0.015 0.014 1.114 Pulse4 0.191 0.010 0.015 0.012 1.322 Pulse5 0.232 0.011 0.017 0.014 1.198 Pulse6 0.260 0.010 0.014 0.012 1.207 Pulse? . 0.286 0.010 0.011 0.012 0.873 Pulse8 0.306 0.010 0.014 0.013 1.095 Pulse9 0.310 0.011 0.040 0.014 2.756 PulselO 0.332 0.010 0.021 0.012 1.792 Train3 0.228 0.009 0.014 0.009 1.510 Treated N=122; dF=3(Between Groups), 118(Within Groups). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. F s = (MS Between) / (MS Within), no significant difference. -65-2.5 T 0 H : • : 1 1 = 1 1 ' 1 0 2 4 6 8 10 Pulse Number Figure 9. Pooled, normalized train quantal content. Data were normalized by 10 Hz quantal content preceeding train (per junction). Since variability between diaphragms was less than that within, data were pooled. Data presented as means ± sem. Sample size is 131 and 122 junctions for control and treated groups, respectively. Stimulation delivered in 0.5mM Ca2+, 5mMK+, 4mM Mg 2 + bathing solution. Preparation was incubated in 1.2mM u.-conotoxin (to block action potentials) for about 60 minutes prior to stimulation. Values presented in Tables 7 and 8 as well. No significant differences in variance of normalized quantal content at each pulse were found between the two groups. Also, variance between diaphragms was found to be similar to that within, in both groups. However, there does appear to be a slight increase in treated values. Fitting was done based on the same time constant. -66-pulse (Pi) are rather low, and that 60 trains were delivered at each junction. For example, if Pi is 0.1, the total release over the entire minute of stimulation becomes 60*0.1=6 with an associated error of square root of 6; i.e., 6+ 2.5. The magnitude of this uncertainty can lead to gross overestimation or underestimation of the amount of facilitation observed per junction. Tables 9 and 10 contain the actual and log normalized data, respectively (means, SEMs, variance within and between, and F values). Again, the variance within and between diaphragms was found to be the same for control and treated diaphragms whether diaphragms were averaged or pooled. Moreover, unlike the mio normalized data, mtrain normalized values were almost identical in both groups. Figure 10 illustrates the lack of difference in (actual) normalized train quantal contents between the two groups. This suggests that the relationship between individual pulse quantal content and that of the entire train remains unaltered by treatment; i.e., the modulatory effects of train stimulation on evoked release was not affected. Furthermore, the variance of the control data is very similar to that of the treated suggesting that, like the relationship between mio and m?o, the quantal content contributed by each pulse and the manner in which that increases over the train, is extremely similar in both groups. This latter observation is synonymous with "facilitation", which will be further investigated and discussed later on. All these aforementioned normalized data are further supported by another normalized value, that of quantal content divided by the fitted quantal content value of the first pulse, m7o,i, of the train (fitting the data will be discussed in a following section). Unlike data normalized by train mean, these data indicate an upward shift in the treated values; however, this is probably accounted for by an artifact of the fitting process since the difference is not statistically significant (p>0.10). Nonetheless, they also lend further support to what has been suggested thus far: that the are no significant differences in variances between the two groups (either within or between diaphragms; -67-Table 9. Actual 70 Hz Train Quantal Content Normalized by Average Train Quantal Content: Descriptive Statistics and Analysis of Variance. CONTROL Mean SEM MS Btwn1 MS Wthn2 F Pulse 1 0.579 0.014 0.010 0.025 0.398 Pulse2 0.732 0.014 0.027 0.025 1.089 Pulse3 0.842 0.016 0.028 0.033 0.867 Pulse4 0.931 0.016 0.043 0.031 1.404 Pulse5 1.019 0.015 0.032 0.030 1.059 Pulse6 1.066 0.018 0.061 0.042 1.465 Pulse7 1.145 0.018 0.044 0.042 1.070 Pulse8 1.174 0.016 0.086 0.031 2.763* Pulse9 1.260 0.020 0.024 0.052 0.461 PulselO 1.253 0.017 0.059 0.036 1.642 Train3 1.000 0.000 0.000 0.000 0.000 Control N= 131, dF=7(Between Groups), 123(Within Groups). * p<0.05, F = (MS Between) / (MS Within). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. TREATED Mean SEM MS Btwn1 MS Wthn2 F $ Pulse 1 0.572 0.010 0.020 0.013 1.546 Pulse2 0.715 0.011 0.008 0.016 0.509 Pulse3 0.828 0.011 0.012 0.014 0.874 Pulse4 0.927 0.011 0.023 0.016 1.443 Pulse5 1.018 0.012 0.001 0.018 0.063 Pulse6 1.084 0.012 0.020 0.018 1.141 Pulse7 1.152 0.012 0.011 0.019 0.592 Pulse8 1.204 0.013 0.010 0.021 0.492 Pulse9 1.220 0.016 0.092 0.030 3.109 PulselO 1.280 0.013 ' 0.038 0.020 1.893 Train3 1.000 0.000 0.000 0.000 0.000 Treated N= 122; dF=3 (Between Groups), 118(Within Groups). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. F $ = (MS Between) / (MS Within), no significant difference. - 68-Table 10. 70 Hz Log Train Quantal Content Normalized bv Average Train Quantal Content: Descriptive Statistics and Analysis of Variance. CONTROL Mean SEM MS Btwn1 MS Wthn2 F Pulsel -0.254 0.011 0.007 0.015 0,480 Pulse2 -0.146 0.009 0.013 0.010 1.203 Pulse3 -0.085 0.009 0.010 0.010 1.027 Pulse4 -0.040 0.008 0.010 0.008 1.289 PulseS 0.002 0.007 0.006 0.006 1.034 Pulse6 0.019 0.008 0.014 0.007 1.874 Pulse? 0.052 0.007 0.006 0.007 0.823 Pulse8 0.064 0.006 0.011 0.004 2.648* Pulse9 0.092 0.008 0.007 0.009 0.752 PulselO 0.093 0.006 0.007 0.004 1.640 Train3 -0.000 0.000 0.000 0.000 1.632 Control N= 131, dF=7(Between Groups), 123(Within Groups). * p<0.05, F = (MS Between) / (MS Within). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. TREATED Mean SEM MS Btwn1 MS Wthn2 Pulsel -0.251 0.008 0.013 0.008 1.652 Pulse2 -0.152 0.007 0.003 0.006 0.479 Pulse3 -0.087 0.006 0.002 0.004 0.514 Pulse4 -0.037 0.005 0.005 0.004 1.378 Pulse5 0.004 0.005 0.000 0.003 0.045 Pulse6 0.032 0.005 0.003 0.003 1.078 Pulse7 0.058 0.005 0.002 0.003 0.686 Pulse8 0.077 0.005 0.002 0,003 0.563 Pulse9 0.082 0.006 0.014 0.004 3.127 PulselO 0.104 0.004 0.004 0.002 1.881 Train3 -0.000 0:000 0.000 0.000 0.190 Treated N=122; dF=3(Between Groups), 118(Within Groups). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. F s = (MS Between) / (MS Within), no significant difference. -69-1.4 T | 1.2 o U "3 1 a 95 a 0.8 | 0.6 -a cu •a o.4 O0.1 o Control • Treated Control fitted Cure Treated Fitted Curve + 4 6 Pulse Number 10 Figure 10. Comparison of pooled treated and control 70 Hz train quantal contents, normal-ized by dividing by means of trains (m i^n). Data presented as means + sem. Sample sizes are 8 diaphragms, 131 junctions and 4 diaphragms, 122 junctions for control and treated groups, respectively. Stimulation delivered in 0.5 mM Ca2+, 5 mM K +, 4 mM Mg 2 + after a 60 minute incubation with jx-conotoxin to abolish muscle action potentials. If error bars are not visible, then they are hidden by the data point. Data points for either group may be overlapping. That no significant difference between control and treated data was found suggests that the relationship between release after each pulse and the mean release of the entire entire is the same at control and treated junctions (i.e., unaltered by toxin treatment). No significant difference was found between mean normalized quantal content or variance at each pulse between the two groups. Also, no significant difference in variance between and within diaphragms was found in either group. - 70 -Table 11, and Figure 11). The issue of the "relationship" between mi0, m7o and m^u, will be further explored by the calculation of the respective correlation coefficients. 4.5 /m Before and During Stimulation In addition to having measured EPP quantal contents, I also recorded fm before, during and after the stimulation periods. fm is expected to rise in response to stimulation (Bain & Quastel, 1992a) and the question of whether such an increase would be altered in the treated group was posed. Since these measurements were carried out in 5 mM K +, such./™ would more closely reflect true spontaneous release, as compared to previous measurements in high K + (15 mM). Therefore, \ogfm, before the stimulation period(s), were compared between the two groups (Table 12). Similar to the 15 mM K + results, such logo's were also normally distributed (Figure 12) while actual values were not. However, unlike 15 mM K + data, both pre-train mean logfm values were significantly greater at treated junctions (p<0.01). With regards to the variability of spontaneous release, and unlike previous observations in 15 mMK+, no significant differences were found between and within diaphragms, in either group, and no significant difference in pooled variance was found between the two groups. Also, it was found that the extent to which fm was increased iii response to stimulation was not altered by treatment. Therefore, the observations made at K+-depolarized nerve terminals are not applicable under lower K + concentrations, although the potential contributions of varying [Ca2+]0 could not have been excluded (as only 0.5 mM Ca2+ was used with 5 mM K*). Miniature endplate potential frequency (fm) was also recorded during the 30-second, 10 -71 -Table 11. 70 Hz Log Train Quantal Content Normalized bv Fitted Average Quantal Content of First Pulse; Descriptive Statistics and Analysis of Variance. CONTROL Mean SEM MS Btwn,1 MS Wthn2 • F $ a • Pulsel -0.030 0.008 0.005 0.008 0.585 Pulse2 0.078 0.009 0.009 0.011 0.752 Pulse3 0!l39 0.008 0.008 0.009 0.957 Pulse4 0.184 0.010 0.005 0.013 0.417 Pulse5 0.226 0.010 0.011 0.012 0.885 Pulse6 0.244 0.011 0.012 0.016 0.764 Pulse7 0.276 0.011 0.011 0.015 0.772 Pulse8 0.288 0.010 0.022 0.011 1.983 Pulse9 0.316 0.012 0.006 0.020 0.313 PulselO 0.317 0.010 0.015 0.013 1.160 Train3 0.224 0.006 0.004 0.006 0.680 Control N= 131, dF=7(Between Groups), 123(Within Groups). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. F $ = (MS Between) / (MS Within), no significant difference. TREATED Mean SEM MS Btwn1 MS Wthn2 F Pulsel -0.013 0.005 0.009 0.003 2.638* Pulse2 0.086 0.006 0.004 0.005 0.899 Pulse3 0.151 0.007 0.004 0.006 0.691 Pulse4 0.201 0.007 0.012 0.007 1.773 Pulse5 0.242 0.008 0.011 0.007 1.605 Pulse6 0.270 0.008 0.005 0.007 0.750 Pulse7 0.297 0.009 0.018 0.010 1.844 Pulse8 0.316 0.009 . 0.009 0.009 0.988 Pulse9 0.320 0.011 0.046 0.013 3.443! PulselO 0.343 0.009 0.028 0.010 2.714* Train3 0.238 0.006 0.011 0.005 2.353 Treated N= 122; dF=3(Between Groups), 118(Within Groups). * p<0.05, ! p<0.01 for F = (MS Between) / (MS Within). 1 mean square Between diaphragms. 2 mean square Within diaphragms. 3 average quantal content of entire train. -72-+» B cu e © U "e3 s es 3 O (MO © J cu .a "es © fc 0.4 T 0.35 0.3 0.25 0.2 0.15 + 0.1 0.05 + 0 -0.05 0 O Control • Treated Control Fitted Curve Treated Fitted Curve 4 6 Pulse Number 10 Figure 11. Comparison between control and treated log train quantal content normalized with respect to the fitted quantal content value of the first pulse in the train. Data presen-ted as means + sem of 8 control and 4 treated diaphragms, with 131 and 122 junctions in each group, respectively. No significant differences in mean or variance of normalized quantal content at each pulse were found between the two groups. Also, no significant difference in variance was found between and within diaphragms, in either group. Experiments were conducted in 0.5 mM Ca2+, 5 mM K +, and 4 mM Mg 2 + with a 60-minute incubation in 1.2 mM p.-conotoxin. These results are in agreement with those presented in the previous two figures and are also presented in Table 11. -73-Table 12A. Descriptive Statistics of Pre-Train Log fm and Analysis of Variance. Control Treated Diaphragm Mean SD n1 Mean SD n 1 -0.845 0.438 13 -0.638 0.45 36 2 -1.061 0.36 17 -0.477 0.262 15 3 -0.873 0.183 . 27 -0.642 0.306 39 4 -0.862 0.194 13 -0.634 0.272 32 5 -0.795 0.185 23 6 -0.754 0.258 21 7 -0.750 0.593 10 8 -0.740 0.272 7 Group2 t# -0.835 3.736! 10(dF)3 -0.598 Pooled -0.844 0.312 131 -0.618 0.343 122 t# : pj 5.492!! 1.21 251(dF) n=number of junctions sampled in respectivediaphragm. means of diaphragm means (n=8, 4 control and treated, respectively). dF=degrees of freedom. t t-test for difference of means between control and treated, ! = p<X01, !! = p<0.001. F $ F-test for difference of variances between control and treated, not significant. Analysis of Variance of Log^, dF SS MS p$ Control Between 7 1.220 0.174 1.882 .; Within 123 11.396 0.093 Total 130 12.616 Treated Between 3 0.343 0.114 0.972 Within 118 13.873 0.118 Total 121 14.216 dF: degrees of freedom; SS: sum of squares; MS: mean square. F $ = (MS Between) / (MS Within), no significant difference. -74--1.5 -1 -0.5 0 0.5 1 1.5 2 Observed (& Theoretical) Deviations from the Mean -1.5 -1 -0.5 0 0.5 1 Observed (& Theoretical) Deviations from the Mean Figure 12. Top figure: pooled cumulative frequency distribution of control pre-train logfm deviations from the mean (131 junctions, mean=0.844, SD=0.296) and deviations of a theoretical normal distribution with mean of zero and SD similar to that of logfm data set. Lower figure: pooled cumulative frequency distribution of treated pre-train logfm deviations from the mean (122 junctions, mean=0.618, SD=0.339) and deviations of a theoretical normal distribution with mean of zero and SD similar to that of log^, data set. The treated pre-train mean log fm was found to be significantly greater than the corresponding control value. However, no significant difference was found between the variance of the pooled data (p>0.1) or between that between diaphragms and that within, in either group. The deviation from the theoretical distribution exhibited by the control data may be attributed to the presence of the outlier data points clearly visible at the extreme ends of the distribution. -75-Hz stimulation period, the means, SDs, and analysis of variance of which are presented in Table 13. Again, mean logfm was greater in treated diaphragms (p<0.01). However, and unlike pre-stimulation \ogfm data, variability between diaphragms was found to be significantly greater than that within, in both groups. Moreover, the within diaphragm variance of the pooled data was also significantly different, but with that of the control being larger. In both groups fm still presented with a log normal distribution (Figure 13; Appendix V illustrates the non-normal distribution of the actual (non-log) fm data). With regards tofm during the 70 Hz train which was also measured, fm in the treated diaphragms was significantly greater than control (although both present with similar variances). 4.6 Facilitation The major reason for looking at facilitation is that it had been shown that release at neonatal rat junctions facilitated considerably more than in the adult (Quastel, personal communication). The hypothesis that toxin treatment may affect changs at the neuromuscular junction such that release would resemble that observed at immature junctions was tested by measuring facilitation. A more direct measure of facilitation than previously discussed was to calculate it as the ratio of the quantal content at the end of the train to that at its beginning (i.e., the ratio of the quantal content of the tenth pulse to that of the first). However, and as previously mentioned, to avoid overestimating or underestimating the quantal content of the first pulse, and subsequently underestimate or overestimate facilitation, an alternate simple method was also tried. Facilitation was calculated as the ratio of the average quantal content during the second half of the train (pulses 6, 7, 8, 9, 10) to that during the first half (pulses 1, 2, 3, 4, 5). Facilitation was calculated for every junction by this method and subsequently either averaged per diaphragm or -76-Table 13. Log 4 During 10 Hz Stimulation: Descriptive Statistics and Analysis of Variance. Control Treated Diaphragm Mean SD n1 Mean SD n 1 -0.544 0.369 13 -0.203 0.304 36 2 -0.774 0.371 17 -0.155 0.274 15 3 -0.745 0.401 27 -0.333 0.231 39 4 -0.799 0.274 13 -0.392 0.328 32 5 -0.661 0.264 23 6 -0.414 0.241 21 7 -0.404 0.518 10 8 -0.605 0.2 7 Group2 t* -0.618 3.394! 10(dF)3 -0.271 Pooled4 -0.633 0.361 131 -0.288 0.296 122 1.49* n=number of junctions sampled in respective diaphragm, means of diaphragm means (ncpntn.i=8, ntreated=4). dF=degrees of freedom. data were pooled after they were normalized by their respective diaphragm mean. t t-test comparing the mean log/m between treated and control; ! p<0.01. F $ F-test comparing the variance in logfm between treated and control; * p<0.05. Analysis of Variance of Logfm During 10 Hz. dF SS MS F Control Between 7 2.693 0.385 3.318! Within 123 14.259 0.116 Total 130 16.952 Treated Between 3 0,951 0.317 3.378* Within 118 9.647 0.082 Total 121 10.598 dF: degrees of freedom; SS: sum of squares; MS: mean square F = (MS Between / MS Within) * p<0.05 ! p<0.01 -77-Observed (& Theoretical) Deviations from the Mean Observed (& Theoretical) Deviations from the Mean Figure 13. Top figure: pooled cumulative frequency distribution of control log/m during 10 Hz stimulation deviations from the mean (131 junctions, mean=-0.633, SD=0.331) and deviations of a theoretical normal distribution with mean of zero and SD similar to that of log/n data set. Lower figure: pooled cumulative frequency distribution of treated pre-train logfm deviations from the mean (122 junctions, mean=-0.288, SD=0.282) and deviations of a theoretical normal distribution with mean of zero and SD similar to that of logfm data set. The treated pre-train mean logfm was found to be significantly greater than the corresponding control value. Also, the variance of the pooled data was found to be significantly different between the two groups, and that between diaphragms was greater than that within, in both groups. -78-pooled per group. As initially suspected, the facilitation during the train was indeed the same for both groups of animals (for averaged data, p=0.3; for pooled data, p=0.17; Figure 14). 4.7 Curve Fitting of Train Quantal Contents In an attempt to obtain an accurate measure of facilitation at each junction each data set (quantal content at each pulse) was fitted to a curve using the least-squares method. The equation used was of the form, . Iog(y0 = a + b ( l - e « " * ) , where "i" represents each of the pulses in the train of 10, "y" is the quantal content per pulse, "a" the intercept, "b" the slope, and "T" the time constant. This equation had previously been found to fit well to averaged data (lines drawn in Figures 9 and 10). It was found that "T" could in fact not be determined for each data set - 'chi-square' was seldom reduced significantly by allowing "T" to float rather than fixing it at the value best fitting the data as a whole (3.7 pulses or 52.9 ms). That is, allowing "x" to vary to a "best fit" did not result in significantly altered fits of the data to the generated curves. Therefore it appeared as though the estimate of the time constant was adequate given the more robust determination of the other parameters of the equation. With this fitting, "a" is the interpolated value of log quantal content at the first pulse. Facilitation was defined as b(l-e'9/T); i.e., the fitted increment of quantal content at the tenth pulse. These values for log facilitation showed no excess of variance between diaphragms, no significant difference of means between treated and controls, and no significant difference in variance between treated and controls (Table 14). Figures 15A and B show that for both treated and controls, log facilitation fitted a normal distribution. Figure 14. Facilitation of quantal content during 70 Hz train stimulation; calculated as the average of the sixth through the tenth pulse over that of the first through the fifth or as the ratio of the quantal content of the tenth pulse to that of the first. The former method is a more a ccurate deterrnination (as discussed in text). No significant difference between control and treated values found (t-test; p>0.05). Since the variability between diaphragms was found to be less than that within, data were subsequently pooled for each group Pooled data also showed no significant difference in facilitation and were homoscedastic (F-test; p»0.05). - 8 0 -Table 14. Log Facilitation of Train Quantal Content at the 10th Pulse: Descriptive Statistics and Analysis of Variance. Control Treated Diaphragm Mean SD n1 Mean SD n1 1 0.298 0.0866 13 0.359 0.0876 36 2 0.315 0.118 17 0.384 0.121 15 3 0.324 0.0868 27 0.329 0.0828 39 4 0.351 0.127 13 0.324 0.0928 32 5 0.344 0.0943 23 6 0.309 0.112 21 7 0.332 0.0636 10 8 0.397 0.123 7 Group2 0.334 0.349 t# 0.791 10(dF)3 Pooled 0.329 0.102 133 0.343 0.0933 122 t# 1.183 1.19 251(dF) n=number of junctions sampled in respective diaphragm, mean and SD calculated for diaphragm means (rWror ,^ iWed=4). dF=degrees of freedom. t-test for difference of means between control and treated (p>0.20). F-test for difference of variances between control and treated (p>0.20). Analysis of Variance of Log Facilitation. dF SS MS F s Control Between 7 0.068 0.010 0.943 Within 123 1.275 0.010 Total 130 1.344 Treated dF SS MS F Between 3 0.053 0.018 2.082 Within 118 1.000 0.008 Total 121 1.053 dF: degrees of freedom; SS: sum of squares; MS: mean square. F$= (MS Between)/(MS Within), p>0.50. -81--0-3 -0.2 -0.1 0 0.1 0.2 0.3 Observed (&Theoretical) Values Observed (& Theoretical) Values Figures 15A and B. Top Figure: cumulative frequency distributions of fitted control log facilitation values (131 junctions, mean=0.329, SD=0.099) and of a theoretical distribution of similar mean and SD. Lower figure: cumulative frequency distributions of fitted treated log facilitation values (122 junctions, mean=0.343, SD=0.091) and of a theoretical distribution of similar mean and SD. There was no significant difference in the mean facilitation value between the groups, and no significant difference in the variance between and within diaphragms, for either group, or for the pooled data (p>0.1). -82-Keeping in mind that the actual quantal content data was more variable in treated diaphragms, that the facilitation itself is so consistent indicates that the mechanism(s) of facilitation remains unaltered with the toxin treatment. Therefore, the suggested binding sites and activated second messenger system(s) underlying facilitation were not upregulated as release was. Whether those junctions that facilitate more actually release more will be answered by calculating the relevant correlation coefficients (data presented later). 4.8 Release Histograms The release process was further investigated by looking at the time at which release occurred after the stimulus and whether this was shifted by treatment. It is well known that there is a delay from the time an action potential is elicited to when it arrives at the nerve terminal and then to the time an EPP is elicited by the action of ACh on its postsynaptic receptors (Katz & Miledi, 1965 ) . Factors, such as temperature or the prolongation of nerve terminal action potential (producing a slower rise and decay in the probability of transmitter release), can affect a change in the latency of release (Katz & Miledi, 1 9 6 5 , 1967 ; Benoit & Mambrini, 1 9 7 0 ) . Latency was determined at each junction as the time between stimulation and release of 1 0 % of the total quantal content. Those for the 10 Hz stimulation are called tio (in Table 15) and the values for the trains (histograms of release for all pulses combined, at each pulse) termed ttnin (also Table 15) These numbers varied considerably between junctions, but 1) tio and t^m at the same junction were always nearly identical, and 2) there was no significant difference between treated and controls in either means or variance (of tio and ttrain). There was in both treated and controls a significantly higher variance between diaphragms than within. 00 r ** > f l — "-a II II A r + P A- 2 o 8 » r ^ a : -& « A Eft § ° s - 8 8 CO 0 ^ 1 s Ct »n Cf Ct cn 3 o o a P a o. ct 8. i 2 •5" 3 to UJ UJ UJ to to UJ UJ UJ 1—1 UJ UJ 1 1 — * _ 1—1 o 1 o 1 o 1 © ' 1 o 1 o UJ UJ UJ to to ON 00 -fc>- to UJ 00 o VO VO •o 00 u*> p o © o o o © © *—* © to to J> UJ UJ o © 00 VO Ul ON 1—1 ON to UJ •u o 00 1—1 0.943 00 113! 884! 0.943 988! 740! _ UJ ^ ON ^ 00 vo Jr oo t o t o t o t O t o t o t o t o o o t o t o t o o o t o 1—4 p p 1 © o j © 1 o UJ UJ UJ i — i UJ to 00 ON to to -C>. 1—1 o ON VO UJ UJ un VO o o o o o © o © b b b to to un to UJ 00 00 vo ON un O UJ UJ UJ UJ vo 00 o UJ to UJ H - l , — . to to UJ UJ o b\ b bo ON VO ON 00 UJ ON VO ON to O 00 — >—' >—i^p—ip—I*—ii—ihH o o - C o o o o X l o o w w 0 0 y Vi o U 5> * ^ W * oo ON oo • »-* •*»• ft -J )R un oo UJ vo vo £ un N> o s ct 5 o •* c ? ct p a t/3 O lb ct" o ct CO O a. <• ct l-t> p a CO O •-•> </J O o a ct o c cn O a o. H 3 ct o o c •-I co Ct o ct. ct* p co Ct 00 UJ 5 a - 84 -4.9 Correlation Coefficients The question regarding the possible correlation of quantal content and magnitude of facilitation called for an answer. Therefore, correlation coefficients were determined for all the measured parameters (Table 16) including tio and ttrain, the correlations being determined for deviations from the the means within each diaphragm. This excludes any possible between diaphragm correlations. The correlations that were found were similar in both groups of diaphragms, and hence unchanged by toxin treatment. Basically, facilitation was not correlated with quantal content (mio or mtrain). Spontaneous release (logfo, logfio, logftrain) was found to be correlated with evoked release logmio, logmtrain which were highly correlated with each other. The time at which 10% of release occurred during the train was highly correlated with the time at which 10% of release occurred during antecedent 10 Hz stimulation, i.e., release latency was consistent at each junction. There also appear unexpected correlations in the treated diaphragms between tio and ttrain and release parameters, with muam (high quantal content is associated with high latency) and with facilitation (relatively high facilitation associated with relatively high tio and ttrain)--85-Table 16. Correlation Coefficients for Pooled Data. logfo logfio lOgftnun logmio lOgmtn in logfacil t,0 ttrain logfo 0.36!! 0.25! 0.24! 0.24! -0.13 -0.12 -0.05 0.50!! 0.23* 0!27! 0.26! -0.09 -0.17 -0.18* logfio 0.36!! 0.45!! 0.51!! 0.46!! 0.02 0.06 0.03 0.50!! 0.57!! 0.38!! 0.41!! 0.07 -0.03 -0.05 lOgftndn 0.25! 0.45!! 0.47!! 0.47!! 0.05 0.03 0.05 0.23* 0.57!! 0.48!! 0.54!! 6.18 0.28! 0.25! logmio 0.24! 0.51!! 0.47!! 0.92!! -0.05 0.02 0.03 0.27! 0.38!! 0.48!! 0.91!! -0.12 0.17 0.17 lOgmtrain 0.24! 0.461! 0.47!! 0.92!!. 0.01 0.02 0.02 0.26! 0.41!! 0.54!! 0.91!! -0.08 0.20* 0.22* logfacil -0.13 0.02 0.05 -0.05 0.01 0.08 0.01 -0.09 0.07 0.18 -0.12 -0.08 0.27! 0.25! tio -0.12 0.06 0.03 0.02 0.02 0.08 0.96!! -0.17 -0.03 0.28! 0.17 0.20* 0.27! 0.95!! ttrain -0.05 0.03 0.05 0.03 0.02 0.01 0.96!! -0.18* -0.05 0.25! 0.17 0.22* 0.25! 0.95!! * p<0.05 ! p<0.01 !! p<0.001 : In each case, value on top pertains to controls and value below to treated. LEGEND: logf0=log(pre-train/m) l6gfio=logr/m10Hz) logftr,i„=log(m train) ldgmio=log(mio) l0grntr»in=l0g(mtalin) logfacil=log(facilitation) tio=time (ms) at which 10% of release occurred (10 Hz) Itrain time (ms) at which 10% of release occurred (70 Hz) - 8 6 -5 DISCUSSION 5.1 Evidence of Increased Acetylcholine Release 5.1.1 Spontaneous Miniature End Plate Frequency The 3-week a-bungarotoxin treatment effected clear and significant changes in neurotransmitter release at the murine neuromuscular junction. Spontaneous rniniature endplate potential frequency,/,, (in 5 mM K+; as distinguished from that measured in 15 mMK*), was significantly higher in treated diaphragms, before and during nerve stimulation. Cull-Candy et al. (1980) reported on similar results (without stimulation) in (human) myasthenic muscles which exhibited higher/,, in 5 mM K + , as well as higher potassium concentrations (up to 20mM), in the presence of 2 mM Ca2+Q. However, at 15 mM K+, I found no significant increase infm between treated and control junctions in a series of [Ca2+]0. On the other hand, Plomp et al. (1992) found a reduction infm due to a-BTX treatment (similar to ours in duration and dose) although they did not specify the Ca2 + concentration in which their/m experiments were done. They argued that the decrease infm was not due to loss of MEPPs in baseline recording noise, but rather due to treatment. They maintained that the control and treated junctions they compared were similar since they had binned them according to MEPP amplitude. Their argument is flawed primarily because they had also reported a toxin-induced shift to the left of the MEPP amplitude distribution (i.e., treated junctions had MEPP amplitudes significantly smaller than controls). In effect, what this shift really means is that treated MEPPs with amplitudes comparable to control ones were originally of greater amplitude, which in turn proves that the junctions they had compared based on similar levels of release were not comparable. This fact makes their -87-assurnption, regarding comparing/, of junctions with similar MEPP amplitudes, invalid. Since I have also shown that/m is log normally distributed at 5 and 15 mM K + in a series of Ca 2 + concentrations with a standard deviation of about 0.3, i.e., 1/3 of junctions have a value more than double or less than half the (geometric) mean, that they had sampled a very small number of junctions (around 10 per diaphragm) may account for their result. They had based their reasoning on the earlier findings that in normal frog (Kuno et al., 1971) and mouse muscle (Harris & Ribchester, 1979) there is an inverse relationship between MEPP amplitude and quantal content or /„ , the rationale being that larger muscle fibers have lower input resistance and therefore smaller MEPP amplitude. This is compensated for by a larger nerve terminal, higher fm and increased quantal content. However, in our hands, that too was not the case. We found a significant positive correlation between log/, (log(fo), log(fio), log(fuain); in 5 mM K 4) and quantal content in both groups of animals. However, it needs to be emphasized that the conditions (cation concentrations, drugs, and methods of muscle action potential abolition) under which electrophysiological measurements are made are far from standardized and the discrepancies in the literature may be a reflection of that (a more thorough discussion will follow). That the increase in/„, which I observed is probably due to a mechanism (activated in response to toxin treatment) probably unrelated to increased nerve terminal size, is indirectly supported by histochemical evidence provided by others. No changes in the activity of choline transferase (ChAT) (Plomp et al., 1992) or acetylcholinesterase (AChE) (Molenaar et al., 1991),the latter haying been shown not to interact with or be affected by a-bungarotoxin (Robaire & Kato, 1974), were found after chronic, sub-paralytic treatment of rats with a-bungarotoxin. (The activity of the two enzymes is considered an indicator of nerve terminal size.) Additionally, direct measurement and observation of toxin-treated junctions showed no evidence for an increase in size, complexity or sprouting (Pestronk & Drachman, 1978). That others had reported nerve terminal sprouting in response to -88-a-bungarotoxin, denervation and botulinum can be explained as follows. Although nerve sprouting, irrespective of the mode of induction (pre- or postsynaptic), has been reported to be highly correlated with the extent of development of extrajunctional AChRs (Pestronk et al., 1976; Lavoie et al., 1976), it is dependent on muscle inactivity. So, although a mere reduction in postsynaptic sensitivity (as opposed to muscle paralysis) will lead to the development of : extrajunctional AChRs, complete inactivity is what triggers nerve sprouting (Holland & Brown, 1980). This suggests that the level of activity of the neuromuscular junction is an induction signal for such genetically-transcribed ultrastructural changes. Therefore, incomplete postsynaptic blockade coupled with the high turn over rate of extrajunctional AChRs does not induce nerve sprouting. In support of the latter statement have been the observations that even small levels of activity suppress development of ACh super sensitivity (Lomo & Westgaard, 1976) and that non-nerve derived muscle activity (i.e., direct electrical stimulation of muscle) decreases the development of extrajunctional AChRs and sprouting (Brown et al., 1976) in botulinum-poisoned muscles. Additionally, Hogan et al. (Hogan et al., 1976) showed that muscle activity also reduces the rate of degradation of ct-bungarotoxin-bound extrajunctional AChRs thereby providing indirect evidence that muscle activity slowed the progression of denervation-related changes. Therefore extrajunctional AChRs appear to be the first step in a cascade of events that (can) culminate in nerve terminal sprouting. However, blockade of extrajunctional AChRs would inhibit any feedback mechanism they either initiate or mediate between the nerve and muscle and in effect inhibit muscle re-innervation.. Interestingly enough, Jansen and van Essen had reported that a-BTX treatment (causing 99.9% blockade of junctional and extrajunctional AChRs for 3 days) during the period of reinnervation after nerve crush did not alter the degree of reinnervation nor did it significantly affect either the appearance of newly formed synapses nor the amount of transmitter they released during nerve stimulation (Jansen & Van Essen, 1975). This can be -89-explained as follows: denervation-triggered nerve terminal sprouting is a genetically activated process which cannot be readily terminated. Therefore the lack of postsynaptic activity over the duration of the experiment will not interfere with the actual nerve sprouting which had been induced earlier. That extrajunctional AChRs were blocked during the experimental period has little impact since their turn-over rate is about a day (Berg & Hall, 1975). That means that within ~24 hours of discontinuation of toxin application the nerve will be able to make functional contacts with the postsynaptic cell. Further evidence that the genetic response to altered neuromuscular function is not readily switched off was provided when it was shown that even after neuromuscular transmission had been restored extrajunctional AChRs persisted for about a week. Therefore sprouting is to be distinguished from re-innervation. The former will occur in response to muscle inactivity regardless of whether functional contacts can be made with the muscle or not. On the other hand, the latter is the actual contact of nerve and muscle and the formation of new synapses. On the other hand, spontaneous release of ACh has been shown to play a critical role in the initial induction of neuromuscular junction formation (Hume et al., 1983; Young & Poo, 1983; Fontaine et al., 1986; Dahm & Landmesser, 1991). And although the reasons for the persistence of such spontaneous quantal release at the adult neuromuscular junction are not clear, the possibility of a continued role (similar to their developmental one of sensing initial nerve-muscle contact and potentiating it) as a physiological sensor of neuromuscular junction integrity (i.e., successful nerve-muscle contact) and function is plausible. Furthermore, the frequency of spontaneous release has been shown to be increased prior to the actual differentiation of the myotube and neuronal growth cone after initial contact (Denis, 1981; Xi'e & Poo, 1986). Hence, that spontaneous release frequency has been shown to be increased with toxin treatment (this -90-thesis) and in MG muscle (Cull-Candy et al., 1980) is suggestive of a physiological role for MEPPs. Sensing a reduction in postsynaptic sensitivity, the nerve terminal responds with an increase \n/m in an attempt to "re-capture" the muscle, as is done during development. Therefore, an increase mfm may be the first response to reduced postsynaptic sensitivity. 5.1.2 Evoked Acetylcholine Release with Nerve Stimulation Since spontaneous release does not mediate the physiologic role of the neuromuscular junction, a concomitant increase in evoked release to maintain the efficiency of neuromuscular transmission would be expected. Indeed, I also found that evoked ACh release was upregulated such that quantal content measured with 10 Hz stimulation, m i 0 , and with 70 Hz train stimulation, m7o, was also increased at treated junctions. Although the number of animals used for this study was slightly greater than that reported in other studies, the total number of junctions sampled far exceeded any of theirs. In turn, that lent an increased level of accuracy to our results. In both control and treated diaphragms, a very strong correlation was found between m i 0 and m7o values establishing that junctions which have higher release at 10 Hz will continue to release more at 70 Hz. And in spite of the increased release, treated junctions still had log normally distributed evoked-release values, much like controls (see later). This reported increase in quantal content is consistent with other electrophysiological reports performed on ct-BTX-treated animals (Plomp et al., 1992, 1994, 1995), human myasthenic endplates (Cull-Candy 1978, 1980) and EAMG animals (Kelly et al., 1978). The above findings of increased release lend further credence to the hypothesis that a mere (albeit significant) reduction in postsynaptic sensitivity to released ACh may be sufficient to affect -91 -changes in release, as opposed to complete inactivity or paralysis which have been shown to trigger more dramatic morphological and physiological changes at the neuromuscular junction (i.e., nerve terminal sprouting). 5.1.3 Discrepancies Regarding Ca2* Sensitivity of Acetylcholine Release Although an increase in ACh release has been found by several investigators, its sensitivity to external [Ca2+] remains a point of contention. Plomp et al. (1994) reported that at low Ca2 + concentrations junctions from rats treated with a-BTX release less than controls, while at higher concentrations they release more. Conversely, Cull-Candy et al. showed that the relationship between quantal content and Ca2 + concentration (plotted on logarithmic co-ordinates) at normal and myasthenic neuromuscular junctions is identical but shifted to the left for MG end plates (reflecting the five-fold increase) with the linear portions of the curves having slopes of 3.3 and 3.4 for normal and myasthenic junctions, respectively, in the range of 0.25 and 0.7 mM (with 2 mM Mg2+). At higher Ca2+ concentrations this difference was reduced. (Their values were also comparable with a reported slope value of 3.0 which was obtained for single human endplates when followed over the range of three or four Ca2+ concentrations, Cull-Candy et al., 1978). Although release properties/mechanisms may be different at human and murine neuromuscular junctions, thereby accounting for these dramatic differences, that Plomp et al. found no parallel sensitivity to Mg 2 + at the same junctions brings into question the validity of their results. Another complication associated with the measurement of quantal content at elevated external Ca2+ concentrations is that when release is high, the analysis of data entails a significantly large source of error (see Results). With such high Ca2 + concentrations, AChR antagonism by either d-tubocurarine or hexamethonium would reduce EPP amplitude and enhance the accuracy -92-of such measurements. The major source of error with high quantal contents is the validity of correction for non-linear summation, which depends critically upon how 'focal' the recording is (Quastel, personal communication) with potentially major underestimation occurring with large quantal units and non-focal recording. Also, calculation of quantal content by the variance method depends even more critically on correction for non-linear summation and assumes that transmitter release is described by Poisson statistics. Although there is evidence to indicate that transmitter release follows Poisson statistics when release is depressed by lowering the Ca27Mg2+ ratio in the bathing solution (del Castillo & Katz, 1954) and remains applicable in normal Ca27Mg2+ when release is depressed by continuous stimulation (Elmqvist & Quastel, 1965b) some studies maintain that a binomial distribution fits the results better when release is not depressed (Johnson & Wernig, 1971; Glavinovic, 1979; Bennet et al., 1975). Hence, the analysis of data based on the latter (incorrect) belief will result in inaccurate results and further confound the interpretation of the available literature. 5.2 Controversial Findings Regarding Acetylcholine Release Not all reports are in agreement regarding the effects of a-BTX and MG on ACh release. Although the autoimmune response in MG appears to be directed at a site different from, but in close proximity to, the AChR recognition site (Almon & Appel, 1975; Lindstrom et al., 1976; Plittag et al., 1976) the primary pathophysiological mechanism of the defect of neuromuscular transmission is a decreased amount of active AChRs (Fambrough et al., 1973; Vincent, 1980). Therefore, the underlying assumption in a-bungarotoxin, MG and EAMG studies has been one of a postsynaptic reduction in ACh sensitivity, due to loss of AChRs, triggering a retrograde signal-mediated upregulation of release from nerve terminal to overcome the deficit in transmission. - -93-Although Elmqvist et al. (1964) and Cull-Candy et al. (1979) had reported a reduction in MEPP amplitude (in intercostal muscles) and a decrease in MEPP current at human myasthenic endplates (Cull-Candy et al., 1979), Elmqvist et al. found no change in quantal content of EPPs when measured in normal Ringer solution (2 mM Ca2+, 1 mM Mg2+). Several other studies on ACh release in autoimmune myasthenic muscle produced controversial results. A considerable increase was found by Molenaar et al (1979) in a biochemical study and by Cull-Candy et al. (1980) in an electrophysiological study (Cull-Candy et al., 1980) while a decrease in quantal content was reported by Lindstrom and Lambert (using intercostal muscle; Lindstrom & Lambert, 1978) and by Maselli using anconeus muscle (Maselli et al., 1991). A similar controversy exists surrounding the release of ACh in muscles of EAMG rats, measured with electrophysiological, biochemical, or bioassay techniques. Both an increase in release (Molenaar et al., 1979; Takamori et al., 1984; Lambert et al., 1976) and a lack of effect (Gallant, 1982; Kelly et al., 1978) have been reported. Additionally, increased neurotransmitter release has also been reported in dystrophic mice (Kelly et al., 1986) although a concomitant reduction in a-BTX binding sites was not found. This apparent conflict in the literature regarding the effect of decreased postsynaptic sensitivity on ACh release may be, at least in part, justifiably attributed to the different investigative techniques used. Biochemical methods measure ACh collected from the entire diaphragm, as opposed to from individual endplates, they require relatively high concentrations of anticholinesterases to preserve the released ACh, and depend on basal levels of ACh release which vary considerably under different experimental and pharmacologic conditions. On the other hand, electrophysiological measurement of quantal ACh release is highly sensitive to the methods used to eliminate muscle action potentials, to the calculation (and correction) of quantal content, as well as to the external concentrations of Ca 2 + and Mg 2 + . Another problem with the electrophysiological studies has been the consistently low number of junctions sampled (10-12) from individual diaphragms which is further exacerbated by the small number of animals used (4-5). Our results clearly established that variance (of all measurements of release) was significantly greater between diaphragms than within. Again, general experimental conditions appear to exert a dramatic effect on the results, and until such disparate conditions are standardized and reconciled, the difficulty in comparing results in the literature will persist. 5.3 Effect of a-BTX Treatment on Release MechanismCsl Spontaneous and evoked ACh release were found to be highly correlated at treated and control junctions. Although treatment caused an increase in both spontaneous and evoked release, it did so without disrupting the nature of their distributions, the time course of release ( t i o tuain) or the magnitude of increase of release in response to stimulation. Facilitation of evoked release was the same in mean value and variance in both groups. Further reinforcement of these results came from the correlation of measurements of release as they were not different between the two groups. That release histograms were similar in both groups in terms of time of occurrence of release (either 10% of total or an EPP) is further evidence that mechanisms of release (save for those responsible for its upregulation) were not altered. And although Cull-Candy (1980) found no change in the slope of the relationship between release and [Ca2 +]0 between M G and control junctions, they did report an increased variability and a shift to the right in the latency histograms of release at M G end plates. Their finding may be interpreted as a change in the release - 95 -mechanism or as a change in the duration and/or intensity of the nerve terminal action potential. The possible immunogenic effects responsible for MG on the nerve terminal cannot be ruled out though. However, they also found no changes in facilitation (in 1 mM Ca2+, 2 mM Mg2+) or in depression of quantal content (in 2 mM Ca2+, 1 mM Mg2+). 5.4 Interpretation of Variance in Neuromuscular Transmission Toxin treatment increased evoked and spontaneous ACh release, but did not increase the variance associated with these measurements. This may be interpreted as a reflection of an equal upregulation (multiplication) of release at all synapses given our finding that the (log normal) distribution of spontaneous and evoked release, and of facilitation, was not affected by treatment. The distribution describes a relationship between different junctions and it is highly likely that if treatment preferentially affected some, then this relationship would be altered That the variability offm of treated junctions was found to be higher than that of controls in 15 mM K + at higher Ca2+ concentrations may be an indication of altered sensitivity to Ca2 + which was not manifest in terms of amount of release. A larger number of sampled junctions under such conditions may help better define this difference in terms of distribution(s). In normal, or control, mice there is intrinsic variability in release at any one junction as quantal content is best described by a Poisson distribution (del Castillo & Katz, 1954). In this report I have not investigated this phenomenon and have assumed that Poisson statistics hold true for all junctions, including treated ones. Beyond this fundamental level of variability there is that between junctions of a single diaphragm. This variability, however, may very well be due to the different types of muscle fibers constituting the diaphragm. Such fibers have different properties -96-and are recruited to different frequencies and intensities. Moreover, the sizes and shapes of nerve tenninals associated with these muscle fibers types have been shown to be dramatically different (Padykula & Gauthier, 1970; Ogata & Yamasaki, 1985; Oki et al., 1990; Tomas et al., 1992). Therefore, the observed log normal distributions of spontaneous and evoked release may in fact be a reflection of the distribution of the muscle fiber type constituency of the diaphragm. In this respect, it becomes even more crucial to ensure that animals being compared are indeed of the same age and weight because the neuromuscular junction has been reported to undergo age-related changes. According to our results, the greatest source of variability is due to differences between animals. This clearly establishes the need for large sample sizes. And if the effects of a certain treatment are being investigated (rather than an attempt to describe the distribution of release), then a large number of animals is more important than the number of junctions sampled from each. 5.5 Evaluation of a-BTX Treatment in Present Experiments 5.5.1 Gross Morphological Changes In determining the optimal toxin treatment regimen it was observed that animals could not survive 4 injections of 0.43 jig/lOgr administered on alternate days. And although the fourth injection had to be reduced to about 65% of the first, the fifth and sixth injections had to be re-increased to about 85%. Furthermore, and for the duration of the treatment, it was found that the dose had to be continuously adjusted so as not to kill the animals. These empirical observations coincide well with findings of the time course of development and degradation of extrajunctional AChRs (see next section). Additionally, treatment of similar duration and comparable toxin dose had been shown to induce the development of extrajunctional AChRs (Plomp et al., 1994). -97-Therefore, it is highly likely that our treatment also induced the proliferation of extrajunctional AChRs; which in turn would imply that neuromuscular transmission had been blocked by more than 80% and that its margin of safety had been increased (Chang et al., 1975). Hogue et al. (1992) had shown that steady state blood levels of d-tubocurarine producing 60% block for a period of 2 weeks did not affect AChR proliferation in the diaphragm but did so in other skeletal muscles. The relation between extrajunctional AChRs and upregulation of release as early steps i a mechanism against a deficit in neuromuscular transmission will be further discussed in the Conclusion. 5.5.2 Degradation of a-BTX/AChR Complex It has been shown, in vivo and in culture, that 24 hours after an injection of a-[125I]bungarotoxin, normal diaphragm muscle retains 80% of bound toxin, specifically associated with the end plate regions, but denervated muscle retains only 20% at extrajunctional AChRs (Berg & Hall, 1974). Berg and Hall (1975b) also showed that loss of toxin from junctional regions of normal muscles occurred with a half-time of approximately 6 days and was accompanied by a corresponding increase in the number of free toxin-binding sites. In contrast, 65% of toxin bound to extrajunctional regions of denervated muscle or neonatal rats was lost in 24 hours at a rate close to that found by Chang and Huang (1975). In an in vivo mouse sternomastoid preparation, Fertuck et al. (1975) showed that no internalization of radioactivity was found throughout the period of recovery following an a-bungarotoxin injection. The higher rate of loss of toxin at extrajunctional AChRs was found to occur as a first-order process with a half-time of approximately 8 hours and to be dependent on the rate of protein synthesis as it was significantly increased (to 61%) after inhibition of the latter with cyclohexamide, a known protein -98-synthesis inhibitor which at lOug/ml inhibits 95% of protein synthesis in the rat diaphragm (Sakmann, 1975; Fambrough, 1970) and almost completely blocked by sodium cyanide and dinitrophenol (Berg & Hall, 1975b). (Rate of toxin loss at junctional receptors of denervated muscle was the same as that of normal muscle). Inhibition of protein synthesis, by actinomycin D, was shown to be effective at inhibiting the increase in ACh sensitivity (Grampp et al., 1972) and the increase in the number of extrajunctional AChRs (Chang & Tung, 1974) only when administered within 24 hours of denervation. Therefore, it appears as though such changes require an induction stimulus but take several days to become fully manifest. Such a sequence of events is also supported by the previous observation that the number of extrajunctional AChRs continued to increase well after neuromuscular transmission had been restored. Consistent with the above was the finding that extrajunctional regions of denervated muscles have a more rapid recovery of ACh sensitivity after a-bungarotoxin blockade as compared to recovery by the end plate regions of normal muscle. In any case, radioactivity recovered from the culture medium is largely as free iodotyrosine, indicating that the toxin had indeed been degraded; moreover, only toxin molecules originally associated with receptor are degraded. The characteristics of the loss of toxin from the muscle are consistent with a mechanism of intracellular degradation of toxin-receptor complex that could reflect a process of normal receptor turnover, or one of degradation of an "abnormal" receptor structure induced by toxin binding (Pine, 1967; Lin & Zabin, 1972; Berg & Hall, 1974). However, in cultures of chick and rat muscle, Devreotes and Fambrough (Devreotes & Fambrough, 1975) have provided indirect evidence that toxin loss solely reflects receptor degradation and that the rate of degradation is not influenced by the presence of the toxin. The latter results can be explained by a difference in turnover rates for junctional and extrajunctional AChRs. Studies comparing purified receptors of both types have shown that -99-several of their properties, including the rates of dissociation of toxin-receptor complex (Brockes & Hall, 1975a) are very similar. However, the two types of receptors have different affinities for d-tubocurarine and can be separated by isoelectric focusing, indicating that they are structurally distinct (Brockes & Hall, 1975b). This structural difference may be the basis for the different rates of toxin loss. Alternatively, different membrane environments for the AChRs could affect the stability of the two receptors in the membrane. 5.6 A Possible Mechanism for Upregulation of Acetylcholine Release As had been previously presented, the neuromuscular junction responds to complete muscle inactivity by a spread of extrajunctionl AChRs and ultimately re-innervation (for detailed sequence of events refer to Background Section). Also, the degree of such nerve terminal sprouting has been shown to be correlated with the extent of extrajunctional AChRs (Pestronk & Drachman, 1976; Lavoie et al., 1976). Therefore, that a reduction in postsynaptic sensitivity is, in essence, one physiologic step short of complete muscle inactivity, it is not surprising that it too would elicit responses identical to those first seen after muscle inactivity. In effect, there appears to be a graded response. Hence, the response to such a reduction is graded to the extent of neuromuscular transmission compromise. Also, in the event of complete inactivity, the response is merely a further progression from that seen with reduction of postsynaptic sensitivity; i.e., both elicit the same response except the former merely takes it further. One may hypothesize that such a sequence of events would be initiated when neuromuscular transmission is compromised (thereby leading to reduced muscle activity) with retrograde inputs from individual junctions processed in the CNS where the extent of reduction in postsynaptic sensitivity (i.e., the number of junctions whose postsynaptic, apparatus is not -100 -responding to released transmitter) is somehow evaluated. Once the number of inactive junctions exceeds the margin of safety of neuromuscular transmission a central response is triggered. Quantal acetylcholine release is increased to maintain the physiologic function of the muscle, since fewer fibers have to mediate the contraction. Meanwhile, those junctions at which there is no postsynaptic activity will respond to the same central drive, via mediators released by nerve terminals, by activation and transcription of extrajunctional mRNA. Immature receptors are inserted throughout the entire muscle membrane surface in preparation for receiving neuronal innervation. The reason for suggesting a distinct threshold for this to occur is the finding that a chronic reduction in postsynaptic sensitivity (equivalent to a 60% block of twitch response) will not induce the development of extrajunctional AChRs. Therefore, if the development of extrajunctional receptors were controlled locally at each junction, then each blocked junction would develop such receptors. However, such is not the case. Only when there remains no margin of safety for transmission will fibers respond by inserting new receptors. Only a central signal can make such a distinction by processing feedback. In a condition such as ours, where the muscle is not completely paralyzed, no re-innervation of the blocked fibers occurs. However, should the reduction in postsynaptic sensitivity progress to complete paralysis, then nerve terminal sprouting will occur as the muscle membranes have already been primed for re-innervation. Therefore, local retrograde signaling which controls the development and growth of individual neuromuscular junctions becomes subservient to a central control which takes into account retrograde signals. However, to explain our results, a central drive would have to affect the different muscle fibers constituting the diaphragm to the same extent, independent of muscle fiber type and, presumptively, the extent to which different junctions are blocked. This is most easily explained in terms of generally increased CNS 'drive' in response to a deficiency of muscle contraction, as - 101 -sensed by stretch receptors, resulting in adaptation of nerve terminals. That is, it seems likely that quantal release can be upregulated as the result of chronic activity. This implies that if spontaneous release in the adult mediates a function similar to its developmental one (the "sensing" of the junction), the increase in/m allows junctions to "hold on" to their functional muscle contacts despite subsynaptic receptor blockade. It may be argued that since there is a reduction in the number of functional muscle fibers with chronic receptor blockade, the remaining few should undergo use hypertrophy. Muscle fiber growth would result in the growth of the nerve terminal, which would account for the increase in the frequency of spontaneous release and in the quantal content of evoked release. However, that the activities of AChE and ChAT have been shown not to be increased under conditions of reduced postsynaptic sensitivity (Plomp et al, 1992; Molenaar et al., 1991) suggests that no such presynaptic growth occurs. Furthermore, direct observation of nerve terminals after a-bungarotoxin treatment has shown that there is.no increase in terminal size or complexity (Pestronk & Drachmann, 1978; Salpeter et al., 1988). Once again though, it needs to be said that a distinctions to be made between effects of denervation and those of reduced postsynaptic sensitivity for they both proceed through the same initial steps but the former continues to more profound and morphological changes: 6. CONCLUSION - 102-Through this thesis I have provided further evidence that nerve terminal function, specifically spontaneous and evoked quantal ACh release, becomes modified with chronic postsynaptic receptor blockade. In diaphragms from animals treated for 3 weeks with injections of a-bungarotoxin to cause weakness, quantal transmitter release was increased over controls to about double, the same increase being manifest in spontaneous MEPP frequency, MEPP frequency in trains of stimulation, and in quantal content of EPP's. There was no change in facilitation in trains of 10 stimuli at 70 Hz. In both controls and treated diaphragms each measure of transmitter release was log normally distributed with a standard deviation close to 03 (logio) units. Variance between diaphragms (different animals) was significantly greater than within in both control and treated groups. This necessitated statistical tests of the difference caused by treatment on the basis of mean values from diaphragms, rather than considering each individual junction as a sample. In a separate set of experiments MEPP frequency in raised K +, at various Ca2+, failed to show significant differences in quantal transmitter release. Since the increase in quantal transmitter release in a-bungarotoxin treated diaphragms was not associated with any increase in the variance of log release for any of the aspects of release that were measured, it is concluded that release is upregulated in response to impaired postsynaptic function via a change in overall presynaptic activity regulated in the CNS rather than through local regulatory factors. The mechanism of this upregulation, at the level of the motoneuron nerve terminal, remains to be elucidated. The purpose of such a centrally-driven, feedback-mediated regulation would be the maintenance of the physiological function of the diaphragm (and the viability of the animal) and not the integrity or function of any one neuromuscular synapse. B I B L I O G R A P H Y - 103 -Albuquerque, E. X., Barnard, E. A., Chiu, T. R, Lapa, A. J., Dolly, J. O. et al. (1973). Acetylcholine receptor and ion conductance modulator sites at the murine neuromuscular junction: evidence from specific toxin reactions. Proceedings of the National Academy of Science USA 70, 949-953. Albuquerque, E. X. & Mclsaac, R. J. (1970). Fast and slow mammalian muscle after denervation. ExperimentalNeurology26, 183-202. Almon, R. R. & Appel, S. H. (1975). Interaction of myasthenic serum globulin with the acetylcholine receptor. Biochim Biophys Acta 393, 66-77. Andonian, M. H. & Fahim, M. A. (1988). Endurance exercise alters the morphology of fast- and slow-twitch rat neuromuscular junctions. International Journal of Sports Medicine 9, 218-223. Arakawa, Y., Sendtner, M. & Thoenen, H. (1990). Survival effect of ciliary neurotrophic factor (CNTF) on chick embryonic motoneurons in culture: comparison with other neurotrophic factors and cytokines. Journal of Neuroscience 10,3507-3 515. Atashi, J. R., Klinz, S. G., Ingraham, C. A., Matten, W. T., Schachner, M. & Maness, P.F. (1992). Neural cell adhesion molecules modulate tyrosine phosphorylation of tubulin in nerve growth cone membranes. Neuron 8, 831-842. Auerbach, A. & Betz, W. (1971), Does curare affect transmitter release? Journal of Physiology 213,691-705. Axelson, J. & Thesleff, S. (1959). A study of supersensitivity in denervated mammalian skeletal muscle. Journal of Physiology 147, 178-193. Bain, A. I. & Quastel, D. M. J. (1992a). Quantal transmitter release mediated by strontium at the mouse motor nerve terminal. Journal of Physiology 450, 63-87. Bain, A. I. & Quastel, D. M. J. (1992b). Multiplicative and additive Ca2+-dependent components of facilitation at the mouse endplates. Journal of Physiology 455, 383-405. Balice-Gordon, R.J. & Lichtman, J. W. (1990). In vivo visualization of the growth of pre- and postsynaptic elements of neuromuscular junctions in the mouse. Journal of Neuroscience 10, 894-908. Balice-Gordon, R. J. & Lichtman, J. W. (1991). Motor nerve terminals are selctively eliminated from regions of neuromuscular junctions focally blocked with alpha-bungarotoxin. Society of Neuroscience Abstracts 11', 1287'. - 104-Balice-Grodon, R. J. & Lichtman, J. W. (1993). In vivo observations of pre- and postsynaptic changes during the transition from multiple to single innervation at developing neuromuscular junctions. Journal ofNeuroscience 13, 834-855. Barnard, E. A., Wieckowski, J. & Chiu, T. H. (1971). Cholinergic receptor molecules and cholinesterase molecules at mouse skeletal muscle junctions. Nature (Lond) 234, 207-209. Bayne, E. K., Anderson, M. J. & Fambrough, D. M. (1984). Extracellular matrix organization in developing muscle: correlation with acetylcholine receptor aggregates. Journal of Cell Biology 9 9 , 1486-1501. Bennett, M. K., Calakos, N. & Scheller, R. H. (1992). Syntaxin: a synaptic protein implicated in docking of synaptic vesicles at presynaptic active zones. Science 257, 255-259. Bennet, M. R., Florin, T. & Hall, R. (1975). The effect of calcium ions on the binomial statistical parameters that control acteylcholine release at synapses in striated muscle. Journal of Physiology 247, 429-446. Benoit, P. R. & Mambrini, J. (1970). Modification of transmitter release by ions which prolong the presynaptic action potential. Journal of Physiology lib, 681-695. Berg, D. K. & Hall, Z. W. (1974). Fate of a-bungarotoxin bound to acetylcholine receptors of normal and denervated muscle. Science 184, 473-475. Berg, D. K. & Hall, Z. W. (1975a). Increased extrajunctional acetylcholine sensitivity produced by chronic post-synaptic neuromuscular blockade. Journal of Physiology'244, 659-676. Berg, D. K. & Hall, Z. W. (1975b). Loss of a-bungarotoxin from junctional and extrajunctional acetylcholine receptors in rat diaphragm in vivo and in organ culture. Journal of Physiology 252, 771-789. Berg, D. K., Kelly, R. B., Sargent, P. R., Williamson, P., & Hall, Z. W. (1972). Binding of a-bungarotoxin to acetylcholine receptors in mammalian muscle. Proceedings of the National Academy of Sciences, USA 6 9 , 147-151. Bevan, D. Bevan, J. & Donati, F. (1988). Muscle relaxants in clinical anesthesia. Year Book Medical Publishers, Inc. Birks, R., Katz, B. & Miledi, R. (1960). Physiological and structural changes at the amphibian myoneural junction, in the course of nerve degeneration. Journal of Physiology 150, 145-168. Birnbaum, M., Reiss, M. & Shainberg, A. (1980). Role of calcium in the regulation of acetylcholine receptor synthesis in cultured muscle cells. Pflugger'sArch 385, 37-43. - 105 -Bixby, J. L., Pratt, R. S., Lilien, J. & Reichardt, L. F. (1987). Neurite outgrowth on muscle cell surfaces involves extracellular matrix receptors as well as Ca2+-dependent and -independent cell adhesion molecules. Proceedings of the National Academy of Science USA 84, 2555-2559. Bloch-Gallego, E., Huchet, M., El M'Hamdi, H., Xie, F. K. et al. (1991). Survival in vitro of motoneurons identified or purified by novel antibody-based methods is selectively enhanced by muscle-derived factors. Development 111, 221-232. Booth, C. M., Brown, M. C , Keynes, R. J. & Barclay, A. N. (1984). Muscles of adult rat and mice express the Thy-1 glycoprotein on denervation. Brain Research 308, 380-382. Boudier, J. L., Le, T. T. & Jover, E. (1992). Autoradiographic localization of voltage-dependent sodium channels on the mouse neuromuscular junction using 125I-alpha scorpion toxin. Journal of Neuroscience 12, 454-466. Boyd, I. A. & Martin, A. R. (1956). The end-plate potential in mammalian muscle. Journal of Physiology132, 74-91. Brenner, H. R., Witzemann, V. & Sakmann, B. (1990). Imprinting of acetylcholine receptor messenger RNA accumulation in mammalian neuromuscular synapses. Nature 344, 544-547. Brockes, J. P. (1984). Assays of cholinergic properties in cultured rat Schwann cells. Proceedings of the Royal Society of London (Biol.) 222, 121-134. Brockes, J. P. & Hall, Z. W. (1975a). Acetylcholine receptors in normal and denervated rat diaphragm muscle. I. Purification and interaction with [125I]a-bungarotoxin. Biochemistry 14, 2092-2099. Brockes, J. P. & Hall, Z. W. (1975b). Acetylcholine receptors in normal and denervated rat diaphragm muscle. II. Comparison.of junctional and extrajunctional receptors. Biochemistry 14, 2100-2106. . Brown, M. C , Goodwin, G. M. & Ironton, R. (1976). Prevention of motor nerve sprouting in botulinum toxin poisoned mouse soleus muscles by direct stimulation of the muscle. Journal of Physiology 267, 42P. Brown, M. C , Holland, R. 1. & Hopkins, W. G. (1981). Motor nerve sprouting. Annual Review of Neuroscience 4, 17-42. Brown, M. C. & Ironton, R. (1977). Motor neuron sprouting induced by prolonged tetrodotoxin block of nerve action potentials. Nature 265, 549-561. Brown, M. C , Jansen, J. K. S. & Essen, D. V. (1976). Polyneuronal innervation of skeletal muscle in new-born rats and its elimination during maturation. Journal of Physiology 261, 387-424. - 106-Callaway, E. M. & Van Essen, D. C. (1989). Slowing synapse elimination by a-bungarotoxin superfusion of neonatal rabbit soleus muscle. Developmental Biology 131, 356-365. Caroni, P. & Grandes, P. (1990). Nerve sprouting in innervated adult skeletal muscle induced by exposure to elevated levels of insulin-like growth factor. Journal of Cell Biology 110, 1307-1317. Chad, D. A. & Lacomis, D. (1994). Ill patients with newly acquired weakness: the clinicopathological spectrum. Annals of Neurology 35, 257-258. Chang, C. C , Chen, T. F., & Chuang, S. T. (1973) N,0-D; and N,N,0-tri[3H]acetyl a-bungarotoxins as specific labeling agents of cholinergic receptors. British Journal of Pharmacology 47, 147-160. Chang, C. C , Chen, T. F. & Huang, S. T. (1973). Influence of chronic neostigmine treatment on the number of receptors and the release of acetylcholine from the rat diaphragm. Journal of Physiology 230, 613-618. Chang, C. C , Chuang, S-T, & Huang, M. C. (1975). Effects of chronic treatment with various neuromuscular blocking agents on the number and distribution of acetylcholine receptors in the rat diaphragm. Journal of Physiology 250, 161-173. Chang, C. C. & Huang, M. C. (1975). Turnover of junctional and extrajunctional acetylcholine receptors of the rat diaphragm. Nature, Lond. 253, 643-644. Chang, C. C. & Lee, C. Y. (1963). Isolation of neurotoxins from the venom of Bungarus multicinctus and their modes of neuromuscular blocking action. Archslnt. Pharmacodyn. Ther. 144, 241-257. Chang, C. C. & Tung, L. H. (1974). Inhibition by actinomycin D of the generation of ACh receptors induced by denervation in skeletal muscle. European Journal of Pharamcology 26, 386-388. Chen, I. L. & Lee, C. Y. (1970). Ultrastructural changes in the motor nerve terminals caused by P-bungarotoxin. Virchows Acrh. B. Zellpath. 6, 318-325. Chiu, A. Y. & Ko, J. (1994). A novel epitope of enactin is present at the mammalian neuromuscular junction. Journal of Neurscience 14, 2809-2817. Chow, I. & Poo, M-M. (1985). Release of acetylcholine upon contact with muscle cell. Journal of Neuroscience 5, 1076-1082. Cohen, M. W., Jones, O. T. & Angelides, K. J. (1991). Distribution of Ca2+ channels on frog motor nerve terminals revealed by co-conotoxin. Journal of Neuroscience 11, 1032-1039. - 107-Coronel, B., Mercatello, A. et al. (1990). Polyneuropathy: potential cause of difficult weaning. Critical Care Medicine 18, 486-489. Couteaux, R. (1973). Motor endplate structure. In Structure and Function of Muscle, Vol. 2, GH Bourne, ed. New York, Academic Press Inc., pp.483-530. Covault, J., Merlie, J. P., Goridis, C. & Sanes, J. R. (1986). Molecular forms of N-CAM and its RNA in developing and denervated skeletal muscle. Journal of Cell Biology 102, 731-739. Covault, J. & Sanes, J. R. (1986). Distribution of N-CAM in synaptic and extrasynaptic portions of developing and adult skeletal muscle. Journal of Cell Biology 102, 716-730. Crockett, J. L., Endgerton, V. R., Max, S. R. & Barnard, R. J. (1976). The neuromuscular junction in response to endurance training. Experimental Neurology 51, 207-215. Cull-Candy, S. G., Miledi, R. & Trautmann, A. (1978). Acetylcholine-induced channels and transmitter release at human endplates. Nature (Land) 271, 74-75. Cull-Candy, S. G., Miledi, R. & Trautamnn, A. (1979). End-plate currents and acetylcholine noise in normal and myasthenic erid-plates. Journal of Physiology 287, 247-265. Cull-Candy, S. G., Mildei, R., Trautmann, A., & Uchitel, O. D. (1980). On the release of transmitter at normal, myasthenia gravis and myasthenic syndrome affected human end-plates. Journal of Physiology 299,621-638. Dahm, L. M. & Landmesser, L. T. (1991). The regulation of synaptogenessis during normal development and following activity blockade. Journal of Neuroscience 11, 328-255. del Castillo, J. & Katz, B. (1954a). Quantal components of the end-plate potential. Journal of Physiology 124, 560-573. del Castillo, J. & Katz, B. (1954b). The effect of magnesium on the activity of nerve endings. Journal of Physiology 124, 553-559. Dennis, M. J. (1981). Development of the neuromuscular junction: inductive nteractions between cells. Annual Review of Nneuroscience 4, 43-68. Deschenes, M. R., Maresh, C. M., Crivello, J. F., Armstrong, L. E., Kraemer, W. J. et al. (1993). The effects of exercise training of different intensities on neuromuscular junction morphology. Journal of Neurocytology 22, 603-615. Devreotes, P. & Frambough, D. M. (1975). Acetylcholine receptor turnover in membranes of developing muscle fibers. Journal of Cell Biology 65, 335-358. Dodge, F. A. & Rahamimoff, R. (1967). Co-operative action of calcium ions in transmitter release at the neuromuscular junction. Journal of Physiology 193, 419-432. - 108-Doherty, P., Ashton, S. V., Moore, S. E, & Walsh, F. S. (1991). Morphoregulatory activities of NCAM and N-cadherin can be accounted for by G protein-dependent activation of L- and N-type neuronal Ca2+channels. Cell 61, 21-33. Dorlochter, M., Irinthcev, A., Blinkers, M. & Wernig, A. (1991). Effects of enhanced activity of synaptic transmission in mouse extensor digitorum longus muscle. Journal of Physiology 436, 283-292. Douglas, J. A., Tuxen, D., Home, H., Scheinkestel, C , Weinmann, M., Czarny, D., Bowes, G. (1992). Myopathy in severe asthma. American Review of Respiratory Disease 146, 517-519. Drachman, D. B. (1974). The role of acetylcholine as a neurotrophic transmitter. Annals of the New York Academy of Science 228, 160-175. Drachman, D. B. & Witzke, F. (1972). Trophic regulation of acetylcholine sensitivity of muscle: effect of electrical stimulation. Science N. Y. 176, 514-516. Eftimie, R., Brenner, H. R. & Buonanno, A. (1991). Myogenin and MyoD join a family of skeletal muscle genes regulated by muscle activity. Proceedings of the National Academy of Science USA 88, 1349-1353. Eldridge, L., Liebold, M. & Steinbach, J. H. (1981). ALterations in cat skeletal neuromuscular junctions following prolonged inactivity. Journal of Physiology 313, 529-545. Elmqvist, D. & Quastel, D. M. J. (1965a). Presynaptic action of hemicholinium at the neuromuscular junction. Journal of Physiology 111, 463-482. Elmqvist, D. & Quastel, D. M. J. (1965b). A quantitative study of endplate potential in isolated human muscle. Journal of Physiology178, 505-529. Engel, A. G. (1984). Myasthenia gravis and myasthenic syndromes. Annals of Neurology 16, 519-534. Engel, A. G. (1986). The neuromuscular junction. InMyology, A. G. Engel & B. Q. Banker, eds. New York, McGraw Hill, pp. 209-253. Fatt, P. & Katz, B. (1952). Spontaneous subthreshold activity at motor nerve endings. Journal of Physiology 117, 109-128. Fahim, M. A. (1989). Rapid neuromuscular remodellling following hindlimb immobiliation. Anatomical Record 224, 102-109. Fahim, M. A. & Robbins, N. (1986). Remodelling of the neuromuscular junction after subtotal disuse. Brain Research 383, 353-356. Fambrough, D. M. (1970). Acetylcholine sensitivity of muscle fiber membranes: mechanism of regulation by motoneurons. Science 168, 372-373. - 109 -Fambrough, D. M., Drachmann, D. B. & Satyamurti, S. (1973). Neuromuscular junction in myasthenia gravis: decreased acetylcholine receptors. Science, NY. 182, 293-295. Fatt, P. & Katz, B. (1952). Spontaneous subthreshold activity at motor nerve endings. Journal of Physiology 117, 109-128. Fertuck, H. C , Woodward, W. & Salpeter, M. M. (1975). In vivo recovery of muscle contraction after a-bungarotoxin binding. Journal of Cell Biology 66, 209-213. Fischbach GD and Robbins N. Effect of chronic disuse of rat soleus neuromuscular junctions on postsynaptic membrane. Journal of Neurophysiology 34:562-569, 1971 Fontaine, B., Klarsfeld, A., Hokfelt, T. & Changeux, J-P. (1986). Calcitonin gene-related peptide, a peptide present in spinal cord motoneurons, increases the number of acetylcholine receptors in primary cultures of chick embryo myotubes. Neuroscience Letters 71, 59-65. Frank, E., Jansen, J. K. S., Lomo, T. & Westgaard, R. H. (1975). The interaction between foreign and original motor nerves innervating the soleus muscle of rats. Journal of Physiology 247,725-743. Fu, W-M & Poo, M-m. (1991). ATP potentiates spontaneous transmitter release at developing neuromuscular synapses. Neuron 6, 837-843. Funte, L. R. & Haydbn, P. G. (1993). Synaptic target contact enhances presynaptic calcium influx by activating cAMP-dependent protein kinase during synaptogenesis. Neuron 10, 1069-1078. Gallant, P. E. (1982). The relationship between anti-acetylcholine receptor antibody levels and neuromuscular function in chronically myasthenic rats. Journal of Neurological Sciences 54, 129-141. Gelsema, A. J. (1980). Apparent lack of effect of a-bungarotoxin on the spike-induced release of acetylcholine at the mammalian motor end-plate. Neuroscience Letters 20, 189-193. Glavinovic, M. I. (1979). Change of statistical parameters of transmitter release during various kinetic tests in unparalysed voltage-clamped rat diaphragm. Journal of Physiology 290, 481-497. Goldman, D., Brenner, H. R. & Heinemann, S. (1988). Acetylcholine receptor a, P, y and y subunit mRNA levels are regulated by muscle activity. Neuron 1,329-333. Goodman, C. & Shatz, C. (1993). Developmental mechanisms that generate precise patterns of neuronal connectivity. Cell 72(Suppl.), 77-98. Gorson, K, & Ropper, A. (1993). Acute respiratory failure neuropathy: a variant of critical illness polyneuropathy. Critical Care Medicine 21, 267-271. - 110-Goss, J. R., Kelly, A. B., Johnson, S. A. & Morgan, D. G. (1991). Haloperidol treatment increases D dopamine receptor independently of RNA levels in mice. Life Science 48, 1015-1022. Grampp, W., Harris, J. B. & Thesleff, S. (1972). Inhibition of denervation changes in skeletal muscle by blockers of protein synthesis. Journal of Physiology 221, 743-754. Gu, Y. & Hall, Z. W. (1988). Immunological evidence for a change in subunits of acetylcholine receptor in developing and denervated rat muscle. Neuron 1, 117-125. Gurney, M. E., Yamamoto, H. & Kwon, Y. (1992). Induction of motor neuron sprouting in vivo by ciliary neurotrophic factor and basic fibroblast growth factor. Journal of Neuroscience 12, 3241-3247. Hall, Z. W. (1973). Multiple forms of acetylcholinesterase and their distribution in the endplate and non-endplate regions of rat diaphragm muscle. Journal of Neurbiology 4, 343-361. Harish, O. E. & Poo, M-m. (1992). Retrograde modulation at developing neuromuscular synapses: involvement of G protein and arachidonic acid cascade. Neuron 9, 1201-1209. Harris, J. B. & Ribchester, R. R. (1979). The relationship between end-plate size and transmitter release in normal and dystrophic muscles of the mouse. Journal of Physiology 296, 245-265. Harvey, A. M. & Macintosh, F. C. (1940). Calcium and synaptic transmission in a sympathetic ganglion. Journal of Physiology 97, 408-416. Henderson, L. P., Smith, M. A. & Spitzer, N. C. (1984). The absence of calcium blocks impulse-evoked release of acetylcholine but not de novo formation of functional neuromuscular synaptic contacts in culture. Journal of Neuroscience 4, 3140-3150. Hennig, R. & Lomo, T. (1985). Firing patterns of motor untis in normal rats. Nature 314, 164-166. Herman, Z. S. & Sominska-Zurek, J. (1979). Central cholinergic receptor supersensitivity after long term atropine administration. Psychopharmacology 64, 337-340. Herrera, A. A., Banner, L. R. & Nagaya, N. (1990). Repeated, in vivo observation of frog neuromuscular junctions: remodeling involves concurrent growth and retraction. Journal of Neurocytology 19, 85-99. Herscovich, S. & Gershon, D. (1987). Effects of aging and physical training on the neuromuscular junction of the mouse. Gerontology 33, 7-13. Hill, R. R., Robbins, N. & Zang, Z. P. (1991). Plasticity of presynaptic and postsynaptic elements of neuromuscular junctions repeatedly observed in living adult mice. Journal of Neurocytology 20, 165-182. - I l l -Hoffman, W. W. & Thesleff, S. (1972), Studies on the trophic influence of nerve on skeletal muscle. European Journal ofPharmacology 20,256-260. Hogan, P. G., Marshall, J. M. & Hall, Z. W. (1976). Muscle activity decreases rate of degradation of a-bungarotoxin bound to extrajunctional acetylcholine receptors. Nature 261, 328-330. Hogue, C. W, Ward, J. M., Itani, S. & Martyn, J. A. J. (1992). Toelrance and upregulation of acetyclholine receptors follow chronic infusion of d-tubocurarine. Journal of Applied Physiology 72, 1326-1331 Holland, R. L. & Brown, M. C. (1980). Postsynaptic transmission block can cause terminal sprouting of a motor nerve. Science 207, 649-651. Holliday, J., Adams, R. J., Sejnowski, T. J. & Spitzer, N. C. (1991). Ca2+-induced release of Ca2+ regulates differentiaion in cultured spinal neurons. Neuron 7, 787-796. Hong, S. J. & Chang, C. C. (1989). Use of georaphutoxin II (|x-conotoxin) for the study of neuromuscular transmission in mouse. British Journal of Pharamcology 97, 934-940. Huang, C-F, Tong, J. & Schmidt, J. (1992). Protein kinase C couples membrane excitation to acetylcholine receptor gene iinactivation in chick skeletal muscle. Neuron 9, 671-678. Hughes, R. A., Sendtner, M., Goldfarb, M., Lindholm, D. & Thoenen, H. (1993). Evidence that fibroblast growth factor 5 (FGF-5) is a major muscle-derived survival factor for cultured spinal motoneurons. Neuron 10, 369-377. Hume, R. I., Role, L. W. & Fischbach, G. D. (1983). Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes. Nature 305, 632-634. Hunter, D. D., Porter, B. E., Bulock, J. W., Adams, S. P., Merlie, J. P. & Sanes, J. R. (1989). Primary sequence of a motor neuron-selective adhesive site in the synaptic basal lamina protein S-laminin. Cell 59, 905-913. Hynes, R. O. & Lander, A. D. (1992). Contact and adhesive specificities in the associations, migrations, and targeting of cells and axons. Cell 68, 303-322. Jahromi, B. S., Robitaille, R. & Charlton, M. P. (1992). Trasnmitter release increases intracellular calcium in perisynaptic Schwann cells in situ. Neuron 8, 1069-1077. Jansen, J. K. S., Lomo, T., Nicolaysen, K. & Westgard, R. H. (1973). Hyperinnervation of skeletal muscle fibers: Dependence on muscle activity. Science 181, 559-561. Jansen, J. K. S. & Van Essen, D. C. (1975). Re-innervation of rat skeletal muscle in the presence of a-bungarotoxin. Journal of Physiology 250, 651-667. - 112-Jaramillo, F., Vicini, S. & Schuetze, S. M. (1988). Embryonic acetylcholine receptors guarantee spontaneous contraction in rat developing muscle. Nature 355, 66-68. Jenkinson, D. H. (1957). The nature of the antagonism between calcium and magnesium ions at the neuromuscular junction. Journal of Physiology 138, 434-444. Johns, T. R. & Thesleff, S. (1961). Effects of motor inactivation on the chemical sensitivity of skeletal muscle. Acta Physiol. Scand. 51, 136-141. Johnson, E. W. & Wernig, A. (19171). The binomial nature of transmitter release at the crayfish neuromuscular junction. Journal of Physiology 218,1 SI-161. Jones, R. & Vrbova, G. (1970). Effect of muscle activity on denervation hypersensitivity. Journal of Physiology 210, 144-145P. Katz, B. (1969). The Release of Neural Transmitter Substances. Charles C. Thomas, Springfield, IL. Katz, B. & Miledi, R (1965), The measurement of synaptic delay, and the time course of acetylcholine release at the neuromuscular junction. Proceedings of the Royal Society of London Biol. 161, 483-495. Katz, B. & Miledi, R (1967). Modification of transmitter release by electrical interference with motor nerve endings. Proceedings of the Royal Society of London Biol. 167, 1-7. Katz, B. & Miledi, R (1968). The role of calcium in neuromuscular facilitation. Journal of Physiology 195, 481-492. Katz, B. & Miledi, R (1978). A re-examination of curare action at the motor endplate. Proceedings of the Royal SocietyLond B. 203,119-133. Kelly, J. J., Lambert, E. H. & Lennon, V. A. (1978). Acetylcholine release in diaphragm of rats with chronic experimental autoimmune mysthenia gravis. Annals of Neurology 4, 67-72. Kelly, S. S., Morgan, G. P. & Smith, J. W. (1986). The origin of (+)-tubocurarine resistance in dystrophic mice. British Journal of Pharamcology 89, 47-53. Kelly, S. S. & Robbins, N. (1983). Progression of age changes in synaptic transmission at mouse neuromuscular junctions. Journal of Phsyiology 343, 375-3 83. Kelly, A. M. & Zacks, S. I. (1969). The fine structure of motor endplate morphogenesis. Journal of Cell Biology 108, 2459-2466. Kenakin, T. P. (1987). Pharmacological Analysis of Drug-Receptor Interactions. New York, Raven Press. -113 -Kim, C , Hirose, M. & Martyn, J. A. J. (1995). D-tubocurarine accentuates the burn-induced upregulation of nicotinic acetylcholine receptors at the muscle membrane. Anesthesiology 83, 309-315. Kuno, M., Turkanis, S. A. & Weakly, J. N. (1971). Correlation between nerve terminal size and transmitter release at the neuromuscular junction of the frog. Journal of Physiology 213, 545-556. Lambert, E. H., Lindstrom, J. M., & Lennon, V. A. (1976). End-plate potentials in experimental autoimmune myasthenia gravis in rats. Annals of the New York Academy of Science 274, 300-318. Lapa, A. J., Albuquerque, E X . & Daly, J. (1974). An electrophysiological study of the effects of d-tubocurarine, atropine and a-bungarotoxin on the cholinergic receptor in innervated and chronicaly denervated mammalian skeletal muscle. Experimental Neurology 43, 31'5-398. Lavoie, P. A., Collier, G. & Tenehouse, A. (1976). Comparison of alpha-bungarotoxin binding to skeletal muscles after inactivity or denervation. Nature (Lond) 260, 349-350. Lee, C. Y. (1972). Chemistry and pharamcology of polypeptide toxins in snake venoms. Annual Review of Pharmacology 12, 265-286. Lee, C. Y., Tseng, L. F., & Chiu, T. H. (1967). Influence of denervation on localization of neurotoxins from elapid venoms in rat diaphragm. Nature 215, 1177-1178. Leung, D. W., Parent, A. S., Cachianes, G., Esch, F., Coulombe, J. N. et al. (1992). Cloning, expression during development, and evidence for release of a trophic factor for ciliary ganglion neurons. Neuron 8, 1045-1053. Li, L. & Olsen, E. N. (1992). Regulation of muscle cell growth and differentiation by the MyoD family of helix-loop-helix proteins. Advances in Cancer Research 95-119. Lichtman, J. W., Magrassi, L. & Purves, D. (1987). Visualization of neuromuscular junctions over periods of several months in living mice. Journal of Neuroscience 7, 1215-1222. Liley, A. W. (1956). The quanatl components of the mammalian end-plate potential. Journal of Physiology 133, 571-587. Lin, S. & Zabin, I. (1972). Beta-galactosidase. Rates of synthesis and degradation of incomplete chains. Journal of Biological Chemistry. 247, 2205-2211. Lindstrom, J. M. & Lambert, E. H. (1978). Content of acetylcholine receptor and antibodies bound to receptor in myasthenia gravis, experimental autoimmune mysathenia gravis, and Eaton-Lambert syndrome. Neurology 28, 130-138. Lo, Y-J & Poo, M-M. (1991). Activity-dependent synaptic competition in vitro: heterosynaptic suppression of developing synapses. Science 254, 1019-1022. - 114-Lomo, T., Massoulie, J. & Vigny, M. (1985). Stimulation of denervated rat soleus muscle with fast and slow activity patterns induces different expression of acetylcholinesterase molecular forms. Journal of Neuroscience 5, 1180-1187. Lomo, T. & Rosenthal, J.(1972). Control of acetylcholine sensitivity by muscle activity in the rat. Journal of Physiology 221, 493-513. Lomo, T. & Westgaard, R. (1976). Control of acetylcholine sensitivity in rat muscle fibers. Cold Spring Harbour Symposioum on Quantitative Biology 40, 263-27.4. Maselli, R. A., Richman, D. P. & Wollmann, R. L. (1991). Inflammation at the neuromuscular junction in myasthenia gravis. Neurology 41, 1497-1504. Matteoli, M., Balbi, S., Sala, C , Chini, B., Cimino, M., Vitadello, M. & Fumagelli, G. (1990). Developmentally regulated expression of calcitonin gene-related peptide at mammalian neuromuscular junction. Journal of Molecular Neuroscience 2, 175-184. Miledi, R. (1960). The acetylcholine sensitivity of frog muscle fibers after complete or partial denervation. Journal of Physiology 151, 1-23. Miledi, R. & Potter, L. T. (1971). Acetylcholine receptors in muscle fibers. Nature 233, 599-603. Miller, L. G., Greenblatt, D. G , Barnhill, J. G. & Shader, R. I . (1988a). Chronic benzodiazepine administration: I. Tolerance is aasociated with benzodiazepine receptor downregulation and decreased y-aminobutyric acidA receptor function. Journal of Pharmacology and Experimental Therapeutics 246, 170-176. Miller, L. G., Greenblatt, D. G., Roy, R. B., Summer, W. R. & Shader, R. I. (1988b). Chronic benzodiazepine administration: I I . Discontinuation syndrome is associated with upregulation of y-aminobutyric acidA receptor complex binding and function. Journal of Pharmacology and Experimental Therapeutics 246, 111-182. Molenaar, P. C , Oen, B. S., Plomp, J. J., Van Kempen, G. T. & Jennekens, F. G. (1991). A non-immunogenic myasthenia gravis model and its application in a study of transynaptic regulation at the neuromuscular junction. European Journal of Pharamcology 196, 93-101. Molenaar, P. C , Polak, R. L., & Miledi, R. (1979). Acetyclholine in intercostal muscle from myasthenia gravis patients and in rat diaphragm after blockade of acetylcholine receptors. Progress in Brain Research 49, 449-458. Nakajima, Y., Kidokoro, Y. & Klier, F. G. (1980). The development of neuromuscular junctions in vitro: an ultrastructural and physiological study. Developmental Biology 11, 52-72. Nastuk, M. A., Lieth, E., Ma, J., Cardassis, C. A., Moynihan, E. B., McKenzie, B. A. & Fallon, J. R. (1991). The putative agrin receptor binds ligands in a calcium-dependent manner and aggregates during agrin-induced acetylcholine receptor clustering. Neuron 1, 807-818. - 115 -Neubig, R. R., Krodel, E. K., Boyd, N. D. & Cohen, J. B. (1979). Acetylcholine and local anesthestic binding to Torpedo nicotinic post-synaptic membranes after removal of non-receptor peptides. Proceedings oj the National Academy of Science USA 76, 690-694. O'Brien, R. A. D., Ostberg, A. J. C. & Vrbova, G. (1978). Observations on the ehmination of polyneuronal innervation in developing mammalian skeletal msucle. Journal of Physiology 282, 571-582. Ogata, T. (1988). Structure of motor endplates in different fiber types of vertebrate skeletal muscles. Archives of Histology and Cytology 51, 385-424. Ogata, T. & Yamasaki, Y. (1985). The three-dimensional structure of mtor endplates in different fiber types of rat inetrcostal muscle. Cell and Tissue Research 241, 465-472. Oki, S., Matsuda, Y., Kitaoka, K., Nagano, Y., Nojima, M. et al. (1990). Scanning electron microscope study of neuromuscular junctions in different muscle fiber types in the zebra finch and rat. Archives of Histology and Cytology 53, 327-332. Oppenheim, R. W. (1991), Cell death during development of the nervous system. Annual Review of Neuroscience 14, 453-501. Oppenheim, R. W., Prevette, D., Qin, W. Y., Collins, F. & MacDonald, J. (1991). Control of embryonic motoneuron survival in vivo by ciliary neurotrophic factor. Science 251, 1616-1618. Padykula, H. A. & Gauthier, G. F. (1970). The ultrastructure of the neuromuscular junctions of mammalian red, white, and intermediate skeletal muscle fibers. Journal of Cell Biology 36, 27-41. Palade, G. E. & Palay, S. L. (1954). Electron microscope observations of interneuronal and neuromuscular synapses. Anatomical Records 118, 335-336. Pazin, M. J. & Williams, L. T. (1992). Triggering signaling cascades by receptor tyrosine kinase. Trends in Biochemical Sciencse 17,374-378. Pedersen, S. E. & Cohen, J. B. (1990). D-Tubocurarine binding sites are located at the a-y and a-5 subunit interfaces of the nicotinic acetylcholine receptor. Proceedings of the National Academy of Science USA 87, 2785-2789. Pestronk, A., Drachman, D. B, & Griffin, J. W. (1976). Effect of muscle disuse on acetylcholine receptors. Nature (Lond) 260,352-353. Pestronk, A. & Drachman, D. B. (1978). Motor nerve sprouting and acetyclholine receptors. Science 199, 1223-1225. Pine, M. J. (1967). Response of intracellular proteolysis to alteration of bacterial protein and the implications in metabolic regulation. Journal of Bacteriology 93, 1527-1533. - 116-Plittag, T., Kornfeld, P., Tormay, A. et al. (1976). Detection of anti-acetylcholine receptor factors in serum and thymus from patients with myasthenia gravis. New England Journal of Medicine 294, 691-694. Plomp, J. J., van Kempen G. ThH., De Baets, M., Graus, Y. M. F., Kuks, J. B. M. & Molenaar, P. C. (1995). Acetylcholine release in mysthenia gravis: regulation at single end-plate level. Annals of Neurology 37, 621'-636. Plomp, J. J., van Kempen, GThH & Moelnaar, P. C (1992). Adaptation of quantal content to decreased postsynaptic sensitivity at single endplates in a-bungarotoxin-treated rats. Journal of Physiology 458, 487-499. Plomp, J. J., van Kempen G. ThH. & Moelnaar, P. C. (1994). The upregulation of acetylcholine release at endplates of a-bungarotoxin-treated rats: its dependency on calcium. Journal of Physiology 478, 125-136. Porter, C. W., Chiu, T. H., Wieckowski, J. & Barnard, E. A. (1973). Types and locations of cholinergic receptor-like molecules: muscle fibers. Nature (New Biol.) 241, 3-7. Porter, S. & Froehner, S. C. (1983). Characterization and localization of the M, 43,000 proteins associated with acetylcholine receptor-rich membranes. Journal of Biological Chemistry 258, 10034-10040. Purves, D. & Sakmann, B. (1974). The effect of contractile activity on fibrillation and extrajunctional acetylcholine sensitivity in rat muscle maintained in organ culture. Journal of Physiology 237, 157-182. Redfern, P. A. (1970). Neuromusculat transmission in newborn rats. Journal of Physiology 209, 701-709, 1970. Reist, N. E., Werle, M. J. & McMahan, U. J. (1992). Agrin released by motor neurons induces the aggregation of acetylcholine receptors at neuromuscular junctions. Neuron 8, 865-868. Reynolds, M. L. & Wolf, C. J. (1992). Terminal Schwann cells elaborate extensive processes following denervation of the motor endplate. Journal of Neurocytology 21, 50-66. Rich, M. M. & Lichtman, J. W. (1989). In vivo visualization of pre- and postsynaptic changes during synpase elimination in reinnervated mouse muscle. Journal of Neurosciene 9,1781-1805. Robaire, B. & Kato, G. (1974). Acetylcholinesterase and a-bungarotoxin: a study of their possible interaction. FEBS Letters 46, 218-223. Robataille, R., Adler, E. M. & Charlton, M. P. (1990). Startegic localization of calcium channels at neurotransmitter release sites of the frog neuromuscular synapses. Neuron 5, 773-779. - 117-Robataille, R., Garcia, M L., Kaczorowski, G. J. & Charlton, M. P. (1993). Functional colocalization of calcium and calcium-gated potassium channels in control of trasnrnitter release. Neuron 11, 645-655. Robert, E. D. & Oester, Y. T. (1970). Absence of supersensitivity to acetylcholine in innervated muscle subjected to a prolonged pharmacological nerve block. Journal of Pharmacology and ExperimentalTherapeutics 11'4, 133-140. Robbins, N. (1988). Plasticity of motor nerve terminals. In: Nerve-Target Cell Trophic Communication. CRC Press, Boca Raton, FL, pp. 199-215. Robbins, N. & Fahim, M. A. (1985). Progression of age changes in mature mouse motor nerve terminals abd its relation to locomotor activity. Journal of Neurocytology 14, 1019-1036. Rosenheimer, J. (1985). Effects of chronic stress and exercise on age-related changes in endplatearchitecture. Journal of Neurophysiology 53, 1582-1589. Sakmann, B. (1975). Reappearance of extrajunctional acetylcholine sensitivity in denervated rat muscle after blockage with a-bungarotoxin. Nature 255,,415-417. Salpeter, M. M., Marchaterre, M. & Harris, R. (1988). Distribution of extrajunctional acetylcholine receptors on a vertebrate muscle: evaluated by using a scanning electron microscope autoradiographic procedure. Journal of Cell Biology 106, 2087-2093. Sarvey, J. M., Lapa, A. J. & Albuquerque, E. X. (1973). Effect of a-bungarotoxin on blockade of cholinergic receptors. Pharamcologist 15, 222-228. Schuetze, S. M. & Role, L. W. (1987). Developmental regulation of nicotinic acetylcholine receptors. Annual Review oj"Neuroscience 10,403-451. Sharpe, M. (1992). The use of muscle relaxants in the intensive care unit: a case of overdose. Canadian Journal oj"Anaesthesia 39, 949-962. Smith, D. O. & Lindgren, C. A. (1986). Modulation of presynaptic transmitter release by ATP derived from postsynaptic sources. In Calcium, Neuronal Function and Transmitter Release. Martinus Nijhoff, Boston, MA, pp. 357-374. Standaert, F. G. (1990). Neuromuscular Physiology. In Anesthesia, 3rd edition. Edited by Miller, R. D. New York, Chrchill-Livingstone, pp. 659-684. Stebbins, C. L., Schultz, E., Smith, R. T. & Smith, E. L. (1985). Effects of chronic exercise during aging on muscle and end-plate morphology in rats. Journal of Applied Physiology 58, 45-51. Stoeling, R. K. (1991). Pharmacology and physiology in anesthetic practice. 2nd Edition, J. B. Lippincott Co. - 118-Sun, Y-A & Poo, M-M. (1987). Evoked release of acetylcholine from the growing embryonic neuron. Proceedings of the National Academy of Science USA 84, 2540-2544. Takahashi, T., Nakajima, Y., Hirosawa, S. & Onodera, K. (1987). Structure and physiology of developing neuromuscular synapses in culture. Journal of Neuroscience 7, 473-481. Takamori, M., Sakato, O., & Okumura, S. (1984). Presynaptic function modified by acetylcholine-receptor interaction in experimental autoimmune myasthenia gravis. Journal of Neurological Sciences 66, 245-253. Thesleff, S. (1960). Supersensitivity of skeletal muscle produced by botulinum toxin. Journal of Physiology 151, 598-607. Thompson, W. (1983). Synapse elimination in neonatal rat muscle is sensitive to pattern of muscle use. Nature 302, 614-616. Thompson, W., Kuffler, D. P. & Jensen, J. K. S. (1979). the effect of prolonged reversible block of nerve impulses on the elimination of polyneuronal innervation of new-born rat skeletal muscle fibers. Neuroscience 4, 271-281. Tomas, J., Fenoll, R., Santafe, M., Battle, J. & Mayayo, E. (1989). Motor nerve terminal morphologic plasticity induced by small changes in the locomotor activity of the adult rat. Neuroscience Letters 106, 137-140. Tomas, J., Santafe, M., Fenoll, R., Mayayo, E. & Battle, J. (1992). Pattern of arborization of the motor nerve terminals in the fast and slow mammalian muscles. Biology of the Cell 74, 299-305. Tsai, M. C. (1975). Studies on the ultrastructural changes in the motor nerve terminal induced by P-bungarotoxin. Ph.D. Dissertation, National Taiwan University, Republic of China. Tsay H-J and Schmidt J. Skeletal muscle denervation activates acetylcholine receptor genes. J. Cell Biol. 108:1523-1526, 1989. Tsim, K. W. K., Ruegg, M. A., Escher, G., Kroger, S. & MacMahan, TJ. J. (1992). cDNAthat encodes active agrin. Neuron 8, 677-689. Ushkaryov, Y. A., Petrenko, A. G., Geppert, M. & Sudhoff, T. C. (1992). Neurexins: synaptic cell surface proteins related to the a-latratoxin receptor and laminin. Science 257, 50-56. Van Essen, D. C , Gordon, H., Soha, J. M. & Fraser, S. E. (1990). Synaptic dynamics at the neuromuscular junction: mechanisms and models. Journal of Neurobiology 21, 223-249. Vergara, J., Tsien, R. Y. & Delay, M. (1985). Inositol 1,4,5-triphosphate: a possible chemical link in excitation-contraction coupling in muscle. Proceedings of the National Academy of Science USA.S2, 6352-6356. - 119 -Vincent, A. (1980). Immunology of acetylcholine receptors in relation to myasthenia gravis. Physiological Reviews 60, 756-824. Waerhaug, O., Dahl, H. A. & Kardel, K.(1992). Different effects of physical training on the morphology of the motor nerve terminals in the rat extensor digitorum longus and soleus muscles. Anatomy and Embryology 186, 125-128. Wallace, B. G., Qu, Z. & Huganir, R. L. (1991). Agrin induces phosphorylation of the nicotinic acetylcholine receptor. Neuron 6, 869-878. . • * Waud, B. E. & Waud, D. R. (1986). Tubocurarine sensitivity at the diaphragm after limb immobilization. Anesthesia and Analgesia 65, 493-495. Wernig, A., Carmody, J. J., Anzil, A. P., Hanser, E. & Marciniak, M. (1984). Persistence of nerve sprouting with features of synapse remodelling in soleus muscles of adult mice. Neuroscience 11, 241-253. Wernig, A. & Herrera, A. A. (1986). Sprouting and remodelling at the neuromuscular junction. Progresses in Neurobiology27,251-291. Witzemann, V. & Sakmann, B. (1991). Differential regulation of MyoD and myogenin mRNA levels by nerve-induced muscle activity. FEBS Letter: 282, 259-264. Xie, Z-P & Poo, M-M. (1986). Initial events in the formation of neuromuscular synapse: rapid induction of acetylcholine release from embryonic neuron. Proceedings of the National Academy of Science USA 83,7069-7073. Yang, JS-J., Sladjy, J, T., Kallen, R. G. & Barchi, R. L. (1991). TTX-sensitive and TTX-insensitive sodium channel mRNA transcripts are independently regulated in adult skeletal muscle after denervation. Neuron 7, 421-427. Young, S. H. & Poo, M-M. (1983). Spontaneous release of transmitter from growth cones of embryonic neurons. Nature 305, 634-637. Yu, X. M. & Hall, Z. W. (1991). Extracellular domains mediating e subunit interactions of muscle acetylcholine receptor. Nature 352, 64-67. Zengel, J. E. & Magleby, K. L. (1982). A quantitative description of stimulation-induced changes in transmitter release at the frog neuromuscular junction. Journal of General Physiology 80, 613-638. Zoran, M. J., Funte, L. R., Kater, S. B. & Haydon, P. G. (1993). Contact with a synaptic target causes an elevation in a presynaptic neuron's resting calcium set-point. Developmental Biology 158,163-171. - 120 -O -1.5 x Theoretical Deviations from the M e a n Appendix I. Log/„in 15 mMK+, 0.5 mMCa2+. Treated diaphragms. Top figure: cumulative frequency distribution of pooled* treated logfm deviations from the mean (331 junctions) and of deviations of a theoretical normal distribution with mean of zero and SD similar to actual log/n data (mean=l .30; SD=0.36). Lower figure: deviations of logfm values from their means plotted against deviations of theoretical distribution (y = 1.0065x-0.0033). * data were normalized with respect to population mean in each diaphragm before pooling. - 1 2 1 -Appendix II. Log/m in 15 mMK+, 0.5 mM Ca2+. Control diaphragms. Top figure: pooled* cumulative frequency distribution of control \og/m deviations from the mean (954 junctions; 5 diaphragms) and of deviations of a theoretical normal distribution with mean of zero and SD similar to that of log/n data (mean= 1.07; SD=0.36). Lower Figure: deviations of \ogfm values from their mean plotted against the deviations of the theoretical distribution (y = 0.9995x-0.0015). * data were normalized with respect to population mean in each diaphragm before pooling. - 122-Observed (& Theoretical) Deviations from the Mean -1 -0.5 0 0.5 1 1.5 2 Observed (& Theoretical) Deviations from the Mean Appendix III. Top figure: pooled cumulative frequency distribution of control actual (non-log) m i o data deviation from the mean (131 junctions, mean= 0.478,SD=0.29) with deviations of a theoretical normal distribution of mean of zero and SD similar to that of m i o data. Lower figure: pooled cumulative frequency distribution of treated actual (non-log) m i 0 data deviation from the mean (122 junctions, mean=0.915, SD=0.505) deviations of a theoretical normal distribution of mean of zero and SD similar to that of m i o data. Neither distribution is normal. Variance was found to be significantly different between pooled treated and control values, and greater between diaphragms than within. - 123 --2 -1 0 1 2 3 4 5 Observed (& Theoretical) Deviations from the Mean Appendix IV. Top figure: pooled cumulative frequency distribution of control actual (non-log) mean m70 data deviation from the mean (131 junctions, mean=0.746.,SD=0.464) with deviations of a theoretical normal distribution of mean of zero and SD similar to that of mio data. Lower figure: pooled cumulative frequency distribution of treated actual (non-log) m™ data deviation from the mean (122 junctions, mean=1.581, SD=0.986) deviations of a theoretical normal distribution of mean of zero and SD similar to that of nty data. Neither distribution is normal. Variance was found to be significantly different between pooled treated and control values, and greater between diaphragms than within in both groups. - 124-Observed (& Theoretical) Deviations from the Mean Observed (& Theoretical) Deviations from the Mean Appendix V. Top Figure: cumulative frequency distribution of control fm (during 10 Hz stim-ulation) deviations from the mean with deviations of a theoretical normal distrib-ution of mean zero and SD similar tofm data (131 junctions, mean=0.338, SD= 0.416). Lower Figure: cumulative frequency distribution of treated/,, (during 10 Hz stimulation) deviations from the mean with deviations of a theoretical normal distribution of mean zero and SD similar tofm data (122 junctions, mean=0.644, SD=0.446). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0087706/manifest

Comment

Related Items