Selective effect of sodium ions on a component of the stimulation- induced increase in transmitter release at the frog n...

MISSING IMAGE

Material Information

Title:
Selective effect of sodium ions on a component of the stimulation- induced increase in transmitter release at the frog neuromuscular junction
Physical Description:
ix, 148 leaves : ill. ; 29 cm.
Language:
English
Creator:
Mosier, Dennis R., 1962-
Publication Date:

Subjects

Subjects / Keywords:
Neuromuscular Junction -- physiology   ( mesh )
Synaptic Transmission -- drug effects   ( mesh )
Sodium -- physiology   ( mesh )
Rana Pipiens -- physiology   ( mesh )
Neuroscience thesis Ph.D   ( mesh )
Dissertations, Academic -- Neuroscience -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1989.
Bibliography:
Bibliography: leaves 134-143.
Statement of Responsibility:
by Dennis R. Mosier.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001136793
oclc - 22783522
notis - AFN6043
sobekcm - AA00006102_00001
System ID:
AA00006102:00001

Full Text











SELECTIVE EFFECT OF SODIUM IONS ON A COMPONENT
OF THE STIMULATION-INDUCED INCREASE IN TRANSMITTER
RELEASE AT THE FROG NEUROMUSCULAR JUNCTION











By

DENNIS R. MOSIER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1989













ACKNOWLEDGEMENTS


I would like to thank my adviser, Dr. Janet E. Zengel,

for her guidance in the techniques and ways of thinking

which are necessary for successful research. I also wish

to thank the other members of my committee, Dr. William G.

Luttge, Dr. John B. Munson, and Dr. Philip Posner, for

their helpful suggestions during the writing of this dis-

sertation.

Special thanks are due to my parents, who have sup-

ported me for years and encouraged me during my educational

career.














TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS..................................... ii

LIST OF TABLES ........................................ v

LIST OF FIGURES..................................... vi

ABSTRACT............................................. viii

CHAPTERS

1 INTRODUCTION.................................... 1

2 SELECTIVE EFFECT OF SODIUM IONS ON STIMULATION-
INDUCED CHANGES IN TRANSMITTER RELEASE......... 7

Introduction.................................... 7
Materials and Methods....................... ... 10
Experimental Procedures....................... 10
Data Analysis............................... 17
Results......................................... 19
Discussion.................................... 55
Conclusions..................................... 67

3 MODELING OF THE EFFECT OF SODIUM ON
STIMULATION-INDUCED INCREASES IN RELEASE...... 69

Introduction.......................... ...... 69
Materials and Methods........................... 70
Results ......... ............................ 79
Discussion................. ............... .... 95
Conclusions..................................... 97

4 EXPOSURE TO OUABAIN PROLONGS EVOKED TRANSMITTER
RELEASE IN MEDIA CONTAINING VERY LOW CALCIUM
CONCENTRATIONS.............................. 99

Introduction..... .......... ..... ......... 99
Materials and Methods........................... 103
Results....... ............. .............. 107
Discussion...................................... 114
Conclusions...................................... 119


iii









5 CONCLUSIONS ...... ............... .............. 121

Residual calcium and the components of
increased release.............................. 121
The sodium-calcium exchange mechanism and
processes of increased release................. 123
Contribution of intracellular calcium to
transmitter release...................... ..... 128
Significance.................................... 131

REFERENCES....... ..... ............................ .... 134

BIOGRAPHICAL SKETCH..................................... 148














LIST OF TABLES


TABLE page

2-1. EFFECT OF INCREASES IN EXTERNAL [Na+] ON
V(50 ms).............................. ........ 34

3-1. EFFECT OF REDUCED EXTERNAL [Na+] ON MODEL
PARAMETERS DESCRIBING INCREASED RELEASE ...... 83

3-2. EFFECT OF INCREASED EXTERNAL [Na+] ON MODEL
PARAMETERS DESCRIBING INCREASED RELEASE...... 85

3-3. EFFECT OF EXTERNAL [Na+] AND [Ca2+] ON
MODEL PARAMETERS DESCRIBING INCREASED
RELEASE....................................... 90













LIST OF FIGURES


FIGURE page

2-1. Effect of reduced external [Na+] on end-
plate potentials recorded extracellularly
from the frog sartorius nerve-muscle
preparation during 10 impulse conditioning
trains ................................ ..... .. 20

2-2. Effect of reduced extracellular [Na+] on
V(t), the fractional increase in EPP
amplitude during 10 impulse trains ......... 23

2-3. Ionic specificity of the effect of reduced
external [Na+] on V(t) during 10 impulse
trains................ ............. ... ..... 27

2-4. Time course and reversibility of the effect
of reduced external [Na+] on V(50 ms)....... 29

2-5. Dependence of V(50 ms) on external Na+
concentration............................... 31

2-6. Effect of increased external [Na+] on V(t)
during 10 impulse conditioning trains....... 33

2-7. Magnitude of the effect of reduced external
[Na+] on V(t) at different concentrations
of external Ca2+.............................. 38

2-8. Effect of external [Ca2+] on the magnitude
of the low-[Na+] effect on V(50 ms)......... 40

2-9. The effect on V(t) of partial replacement
of sodium with lithium ...................... 42

2-10. Intracellular measurements of the effect of
reduced external [Na+] on changes in
quantal content during 10 impulse trains.... 44

2-11. Effect of reduced external [Na+] on the
change in presynaptic action potential
width during repetitive stimulation........ 46

2-12. Effect of the Na+ ionophore monensin on V(t)
during 10 impulse trains................... 52









FIGURE Dace

2-13. Effect of the Na+/H+-exchange inhibitor
amiloride on V(t) during 10 impulse
trains........................ .......... ..... 56

2-14. Effect of increased external [Ca2+] on V(t)... 59

3-1. Components of the model of stimulation-
induced increases in transmitter release.... 75

3-2. Modeling of the effect of reduced external
[Na+] on stimulation-induced changes in
release.......................... .......... 80

3-3. Modeling of the effect of increased external
[Ca2+] on V(t) during 10 impulse trains..... 88

3-4. Effect of external [Ca2+] on the magnitude
of the low-[Na+] effect on fl*.............. 92

3-5. Modeling of the effect of the Na+ ionophore
monensin on V(t) during 10 impulse trains... 94

4-1. Ouabain pretreatment causes prolonged
release at very low external Ca2+
concentrations............................ 111

4-2. Intracellular recording of continued trans-
mitter release at very low external [Ca2+]
following pretreatment with ouabain......... 113


vii









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

SELECTIVE EFFECT OF SODIUM IONS ON A COMPONENT
OF THE STIMULATION-INDUCED INCREASE IN TRANSMITTER
RELEASE AT THE FROG NEUROMUSCULAR JUNCTION

By

Dennis R. Mosier

August, 1989

Chairman: Janet E. Zengel
Major Department: Neuroscience

Repetitive stimulation of presynaptic nerve terminals

often leads to an increase in the amount of transmitter

released by a nerve impulse. Accumulation of Ca2+ and/or

Na+ ions within the nerve terminal has been proposed to

explain this increase in release. This dissertation

describes the results of experiments performed to charac-

terize the roles of these ions in stimulation-induced

increases in neurotransmitter release at the frog (Rana

pipiens) neuromuscular junction.

Standard extra- and intracellular recording techniques

were used to record end-plate potentials (EPPs) from the

frog sartorius nerve-muscle preparation. The nerve was

conditioned using trains of 10 impulses at 20 impulses/s in

order to examine the effect of repetitive stimulation on

V(t), the stimulation-induced change in EPP amplitude.

Reduction of the external Na+ concentration produced a

concentration-dependent and reversible increase in V(t).

The Na+ effect on V(t) appeared to occur early during the


viii








conditioning train and to reach a plateau before the end of

the train, consistent with an effect of Na+ on an early

component of increased release. Monensin, a Na+-selective

ionophore which acts to increase intracellular Na+ concen-

trations, had a qualitatively similar effect on V(t), sug-

gesting that the Na+ effect is mediated through a mechanism

which is dependent on the driving force for Na+ entry into

the nerve terminal. The magnitude of the Na+ effect was

dependent upon the external concentration of Ca2+. These

observations are consistent with the hypothesis that a

Na+/Ca2+ exchange mechanism may be involved in stimulation-

induced increases in release at the neuromuscular junction.

Initiation of evoked transmitter release has tradition-

ally been attributed to Ca2+ entering the nerve terminal

from the extracellular fluid during the action potential.

However, intracellular Ca2+ has also been proposed to play

a role in evoked release. To examine this hypothesis,

nerve-muscle preparations were exposed to ouabain, a Na+/K+

pump inhibitor which secondarily elevates intracellular

Ca2+. Following treatment with ouabain, evoked release

could be maintained in the virtual absence of external

Ca2+. This finding supports the hypothesis that intracel-

lular Ca2+ may play a role in the process of evoked trans-

mitter release.













CHAPTER 1

INTRODUCTION


Rapid communication of electrical signals between most

neurons and other excitable cells is dependent upon the

process of chemical synaptic transmission. In this pro-

cess, an electrical impulse is conducted into the presynap-

tic nerve terminal, initiating events which lead to the

release of a specific chemical substance, or neurotransmit-

ter, from the nerve terminal. The neurotransmitter then

diffuses across the cleft which separates the presynaptic

and postsynaptic cells and binds to specific receptors on

the postsynaptic cell surface, producing a post-synaptic

potential, an electrical response in the membrane of the

postsynaptic cell (see Kuffler et al., 1984). Because

chemical synapses play an essential role in the nervous

system, understanding the mechanisms underlying synaptic

function is of great importance in the understanding of

nervous system organization and function.

Although much is known about the mechanisms of impulse

propagation into the nerve terminal, and about the

responses induced in the postsynaptic membrane by a neuro-

transmitter, relatively little is understood about the pro-

cess of release of neurotransmitter substances from the

nerve terminal. It is known that transmitter substances








are released from the presynaptic nerve terminal in pack-

ets, or quanta, containing similar amounts of neurotrans-

mitter (del Castillo and Katz, 1954a; Boyd and Martin,

1956). There is a great amount of evidence that the

release of a quantum of transmitter results from exocytosis

of the contents of a transmitter-containing synaptic

vesicle into the synaptic cleft (e.g., Heuser et al.,

1979), although alternative mechanisms such as carrier pro-

teins and transmitter-specific pores or gates in the cell

membrane have been proposed (Tauc, 1982). There is also

good evidence that evoked quantal transmitter release is

generally initiated by the entry of calcium ions into the

nerve terminal (Katz and Miledi, 1967a,b; Llinas and

Nicholson, 1975). Transmitter release appears to be pro-

portional to the third or fourth power of the calcium con-

centration at or near the release site, an observation

which has been interpreted as evidence for a co-operative

interaction of three or four calcium ions on the release

mechanism (e.g., Dodge and Rahamimoff, 1967). Despite

these advances in our understanding of the release process,

however, it is not known what events occur to initiate

quantal release after calcium enters the nerve terminal.

In addition to evoked quantal release, quanta of neuro-

transmitter are released spontaneously from the presynaptic

nerve terminal at a variety of central and peripheral syn-

apses including frog and crayfish neuromuscular junctions

and frog spinal motoneurons (Fatt and Katz, 1952; Dudel and








Kuffler, 1961; Katz and Miledi, 1963). At low frequencies

of spontaneous quantal release, the timing of release

events appears to be random (reviewed in Martin, 1966). At

the neuromuscular junction, the amplitude of the miniature

end-plate potential, or postsynaptic response to a single

quantum of transmitter, has been shown to be virtually

identical to the amplitude of the smallest evoked end-plate

potential, or unit potential (Fatt and Katz, 1952; Martin,

1966). The marked similarity between these two end-plate

responses led del Castillo and Katz (1954a) to propose that

the end-plate potential resulted from the synchronized

release of one or more quanta during depolarization of the

presynaptic nerve terminal (the 'quantal hypothesis' of

transmitter release).

A continuous, nonquantal release of transmitter which

results in a small (about 40 uV), tonic depolarization of

the end-plate has been described at frog and rat neuromus-

cular junctions (Katz and Miledi, 1977a). Such nonquantal

acetylcholine release may represent as much as 98% of the

total amount of acetylcholine released by the resting nerve

terminal (Katz and Miledi, 1977a; Vizi and Vyskocil, 1979).

The function of this spontaneous, nonquantal release of

transmitter is unknown, although it has been suggested to

have trophic effects on the neuromuscular synapse (e.g.,

Vizi and Vyskocil, 1979). Nonquantal release of transmit-

ter is unlikely to contribute to the generation of the end-

plate potential, however, as its effect on the membrane








potential of the muscle fiber is almost two orders of mag-

nitude smaller than the response to a single quantum of

transmitter (Katz and Miledi, 1977a). Furthermore, no

increase in the magnitude of nonquantal release could be

detected during nerve stimulation under conditions of

reduced extracellular Ca2+ concentration which were suffi-

cient to abolish evoked transmitter release (Katz and

Miledi, 1981). The discussions of transmitter release in

this dissertation will be restricted to properties of quan-

tal transmitter release.

Another intriguing and widespread property of synapses

is the modulation of the amount of transmitter released by

a nerve impulse as a result of previous synaptic activity.

Stimulation-induced increases in synaptic responses at

peripheral synapses have been shown to result from

increases in transmitter release (Liley and North, 1953;

del Castillo and Katz, 1954b; Martin and Pilar, 1964; Mal-

lart and Martin, 1967). Increases in synaptic responses

during repetitive stimulation have also been described at a

number of central synapses (Curtis and Eccles, 1960;

Porter, 1970; McNaughton, 1980; Hess et al., 1987). As the

nervous system uses trains of impulses to transmit signals,

understanding the effect of repetitive stimulation on

transmitter release is particularly relevant to the under-

standing of nervous system function. For instance, changes

in the magnitude of the synaptic response as a result of

repeated activation of a synapse have been suggested to








underlie many aspects of learning and memory (e.g., Hebb,

1949). Activity-induced modification of synaptic transmis-

sion may also play a major role in nervous system disease

processes such as the epileptic seizure focus, in which

abnormalities of synaptic function such as triggering of

prolonged bursts of action potentials and paroxysmal depo-

larizing shifts in postsynaptic membrane potentials have

been described (reviewed in Taylor, 1988).

As with the mechanism of transmitter release itself,

the mechanisms responsible for changes in transmitter

release due to repetitive stimulation are still poorly

understood. Katz and Miledi (1965, 1968) have suggested

that the phenomenon of increased release may occur as a

result of the accumulation of calcium within the nerve ter-

minal during repetitive stimulation. However, this

hypothesis cannot easily account for all of the processes

of increased release which have been described at the

neuromuscular junction (Zengel and Magleby, 1982).

Understanding the mechanisms responsible for changes in

transmitter release due to repetitive stimulation may give

some insight into the mechanism of the release process

itself. Any model proposed to describe the release process

will have to account for the numerous properties of release

which have been reported. Among these are the calcium

dependence of release (Dodge and Rahamimoff, 1967), the

divalent cation selectivity of release (Ca2+ > Sr2+ > Ba2+;

Miledi, 1966; Blioch et al., 1968; Augustine and Eckert,





6


1984), the very short delay for initiation of release (less

than 200 us; Llinas et al., 1981), and the occurrence of

stimulation-induced changes in release. The following

chapters will describe the results of investigations of two

such properties, the increase in transmitter release due to

repetitive stimulation and the calcium dependence of

release, as a means of better understanding the release

process.













CHAPTER 2

SELECTIVE EFFECT OF SODIUM IONS
ON STIMULATION-INDUCED CHANGES IN TRANSMITTER RELEASE


Introduction


At the frog neuromuscular junction, repetitive stimula-

tion, under conditions of low quantal content, results in

increases in evoked quantal release during the stimulus

train. This increase in evoked release has been attributed

to accumulation of "residual" calcium or a calcium-

activated factor within the nerve terminal, which is only

slowly removed from the cytoplasm after passage of a nerve

action potential (Katz and Miledi, 1968; Rahamimoff, 1968;

Zucker and Lara-Estrella, 1983). The influx of calcium

accompanying the subsequent nerve action potential would

add to the small increase in calcium persisting from the

previous action potential. As transmitter release appears

to be related to the third or fourth power of the calcium

concentration at or near the release site, even a small

increase in residual cytosolic calcium could lead to a sig-

nificant increase in release (Dodge and Rahamimoff, 1967;

Charlton et al., 1982).

The stimulation-induced increase in release at the frog

neuromuscular junction appears to have at least four compo-

nents: two short-acting components of facilitation (Mallart








and Martin, 1967; Magleby, 1973; Younkin, 1974), a long-

term process called potentiation (Rosenthal, 1969; Magleby

and Zengel, 1975), and an intermediate component termed

augmentation (Magleby and Zengel, 1976; Erulkar and Rahami-

moff, 1978; Zengel and Magleby, 1982). Although residual

intracellular calcium has been suggested to play a role in

each of these processes, their ionic dependence is not

clearly understood. Many authors have suggested that

increases in intracellular sodium concentrations may con-

tribute to one or more of the processes of increased

release, either by a direct effect of intraterminal sodium

or by an indirect effect of sodium on intracellular calcium

concentrations (e.g., Rahamimoff et al., 1980).

Although the experiments of Katz and Miledi (1967b)

have shown that Na+ entry into the nerve terminal is not

necessary for release to occur, there is a great deal of

evidence that manipulations known to affect intracellular

Na+ can modulate release and stimulation-evoked changes in

release. Blockade of the Na+/K+ pump by ouabain, which

elevates [Na+]i, can increase evoked release (Birks and

Cohen, 1968b, Baker and Crawford, 1975). Introduction of

Na+ into motor nerve terminals using Na+-filled liposomes

increases both evoked and spontaneous transmitter release

at the frog neuromuscular junction (Rahamimoff et al.,

1978). Accumulation of [Na+]i has been implicated in the

process of potentiation at frog neuromuscular junctions

(Lev-Tov and Rahamimoff, 1980; Meiri et al., 1981), while








introduction of Na+ into the nerve terminal with the Na+

ionophore monensin increases a slow time course process

called "long-term facilitation" at the crayfish neuromuscu-

lar junction (Charlton et al., 1980; Atwood et al., 1983).

However, there has been no systematic investigation of the

effects of manipulations of [Na+]i on shorter time course

components of increased release.

Whether Na+ ions affect the release mechanism directly,

or act indirectly by increasing intracellular [Ca2+], is

unknown. Originally it was proposed that Na+ may compete

with Ca2+ for entry into the nerve terminal (Colomo and

Rahamimoff, 1968), although this effect may be insignif-

icant at low Ca2+ concentrations in the extracellular

medium. Decreases in external [Na+] or increases in inter-

nal (Na+] may also alter intracellular [Ca2+] by inhibiting

Ca2+ extrusion via the sodium-calcium exchanger, which is

driven by the electrochemical gradient for Na+ across the

plasma membrane (e.g., Blaustein and Oborn, 1975; Dipolo

and Beauge, 1983). Elevated internal [Na+] has been sug-

gested to trigger release of Ca2+ from intracellular stores

(Lowe et al., 1976; Rahamimoff et al., 1980).

In this study, reductions in extracellular [Na+] are

shown to selectively increase an early component of the

stimulation-induced increase in transmitter release at the

frog neuromuscular junction under conditions of reduced

quantal content. Addition of monensin, an ionophore known

to increase intracellular [Na+] (Pressman and Fahim, 1982),








has a qualitatively similar effect on increased release.

The magnitude of the Na+ effect on this early component of

increased release appears to be dependent on extracellular

[Ca2+]. These observations suggest that a sodium-calcium

exchange mechanism may be involved in stimulation-induced

increases in transmitter release at the frog neuromuscular

junction. Portions of this work have been published in

abstract form (Mosier et al., 1986; Mosier and Zengel,

1987).


Materials and Methods


Experimental Procedures

The results reported in this study are based upon

observations made on more than 120 sartorius nerve-muscle

preparations from the frog Rana pipiens. Healthy frogs of

2 to 2 1/2 inch length were obtained from Nasco and Charles

Sullivan and housed in a basin at 20-22 oC prior to usage.

Frogs were decapitated and the sartorius muscle with the

attached nerve was dissected in normal Ringer's solution

with special care taken to minimize trauma to the prepara-

tion. The nerve to the sartorius muscle was cut at the

point of its separation from the sciatic nerve, approxi-

mately 4-5 mm from the entry site of the sartorius nerve

into the muscle. The sartorius muscle was stretched

slightly by pinning the connective tissue at the borders of

the muscle to wax at the bottom of the recording chamber.








The muscle nerve was stimulated with a fluid suction

electrode (Dudel and Kuffler, 1961). One end of a PE-60

polyethylene tube of 0.76 mm inside diameter (Clay Adams,

no. 7416) was drawn after gentle heating to an inside

diameter of approximately 0.3 mm. The other end of the

tube was connected to a 10 cc syringe, which was used to

draw the nerve and bathing solution into the tapered end.

The nerve was stimulated by the passage of a supramaximal

stimulus of 0.01-0.1 ms duration between a silver-silver

chloride (Ag-AgC1) wire inserted into the suction electrode

and a similar Ag-AgC1 wire in the bath.

After observing a muscle twitch upon stimulation of the

sartorius nerve, which verified that action potential con-

duction was intact in the nerve, the bathing medium was

changed to a solution containing 5 mM Mg2+ and reduced

concentrations of Ca2+ (0.4-0.6 mM) in order to reduce the

quantal content, or number of quanta of transmitter

released per nerve impulse (del Castillo and Stark, 1952;

del Castillo and Katz, 1954a). Under these conditions, the

mean quantal content of release was typically 0.2 to 2 for

the first impulse in a conditioning train, and rarely

exceeded 3.5 during a 10 impulse train. Nerve stimulation

under conditions of reduced quantal content does not evoke

a muscle contraction, thereby allowing stable recordings of

end-plate responses to be made. Reduction of the quantal

content also minimizes the effect of stimulation-induced

depression of transmitter release, which has been attrib-








uted to depletion of the amount of transmitter available

for release (e.g., Liley and North, 1953; Thies, 1965;

Elmqvist and Quastel, 1965, Mallart and Martin, 1968).

Furthermore, indirect methods of estimating quantal content

(discussed later in this section) are based on the assump-

tion of a Poisson distribution of quantal content, which is

valid only at low probabilities of release (corresponding

to average quantal contents of less than 4; del Castillo

and Katz, 1954a; Martin, 1966).

Extracellular recordings of end-plate potentials (EPPs)

were obtained with a surface electrode from end-plate

regions of the frog sartorius muscle. Under conditions of

reduced quantal content, changes in the amplitudes of

extracellularly recorded EPPs have been shown to be a good

measure of changes in transmitter release (Mallart and Mar-

tin, 1967; Magleby, 1973). The surface electrode was con-

structed from PE-60 tubing similar to that used for the

stimulating electrode. A 3.5 cm length of tubing was

heated gently at one end to form a smooth flange around the

tube opening, and a Ag-AgCl wire was inserted into the tube

from the other end. The surface electrode was filled with

the bathing solution used in the experiment, and the

flanged end was visually positioned close to an end-plate

region of the muscle. End-plate potentials were recorded

differentially between the Ag-AgCl wire in the surface

electrode and a similar Ag-AgCl wire placed in the bath.

The electrode was moved over the surface of the muscle to








locate the position of the maximum end-plate response to

nerve stimulation. After placing the electrode tip very

close to the muscle at a site of maximal end-plate

response, the electrode was not shifted during the course

of the experiment.

The nerve to the sartorius muscle was stimulated using

a Grass S48 or S88 stimulator and a Grass SIU5 stimulus

isolation unit. To rule out increases in transmitter

release due to recruitment of axons, preparations were

stimulated at five times the voltage required to evoke a

maximally sized single EPP. For most experiments, trains

of 10 impulses applied at a stimulus frequency of 20

impulses/s were used to condition the nerve. The condi-

tioning trains were delivered to the nerve once every 55 s.

In a few experiments, paired pulses were delivered to the

muscle nerve every 40 s. These intervals should be of suf-

ficient duration to allow the nerve terminal to recover to

a control state between trains or pairs of conditioning

stimuli (Magleby and Zengel, 1975).

Extracellular responses obtained with the surface elec-

trode were amplified with a Grass P511 A.C. preamplifier

and displayed on a Tektronix model 5113 dual beam storage

oscilloscope. Amplified responses were filtered at 10 kHz.

In most experiments, the responses to 4 to 128 trains of

impulses delivered under each experimental condition were

averaged for analysis using a Nicolet 1170 signal averaging

computer. End-plate potential amplitudes were measured








directly from the digitized data displayed on the screen of

the signal averaging computer during the course of the

experiment. In some experiments, amplified responses from

the surface electrode were digitized using the 12-bit

analog-to-digital converter of a MINC-11 computer which

sampled and recorded EPP amplitudes for subsequent averag-

ing and analysis (Magleby and Zengel, 1976). As the abso-

lute magnitude of the end-plate potential measured with a

surface electrode may vary due to electrode placement and

the location within the muscle of the end-plates sampled,

all EPP amplitudes within an experiment were normalized to

the average amplitude of the first EPP of trains recorded

in the control bathing medium. Comparisons of EPP ampli-

tudes between preparations were made on the basis of these

normalized data.

Intracellular recordings were performed using glass

microelectrodes filled with 3M KCl, with tip resistances of

5-15 megohms. Microelectrodes were pulled on a David Kopf

model 720 vertical pipette puller using glass capillaries

obtained from World Precision Instruments (WPI #TW150F).

End-plates were located by inserting a microelectrode into

muscle fibers in a visually identified end-plate region.

The microelectrode was inserted along the length of a

muscle fiber to the position which gave the maximum EPP

amplitude and minimum EPP rise time (typically less than

1.0 ms), and at which the mean miniature end-plate poten-

tial (MEPP) amplitude was at least 0.4 mV. Although the








majority of end-plates of twitch muscle fibers in the adult

frog are innervated by only a single axon, polyneuronal

innervation of end-plates has been demonstrated (e.g.,

Trussell and Grinnell, 1985). To prevent effects due to

recruitment of nerve fibers, the nerve to the sartorius was

stimulated at 5 times the threshold voltage required to

evoke an end-plate response. No differences in average EPP

amplitude were noted at different stimulation intensities

above threshold, suggesting that a significant contribution

of polyneuronal innervation was unlikely among the end-

plates sampled in this study.

End-plate potentials recorded using intracellular tech-

niques were typically less than 1.5 mV in amplitude in the

absence of repetitive stimulation and seldom exceeded 3 mV

during the conditioning trains. No corrections were made

for nonlinear summation of unitary EPPs, since experimental

evidence suggests that such corrections are unnecessary if

EPP amplitudes are less than 10 mV (McLachlan and Martin,

1981). Because of the conditions of reduced quantal con-

tent in these experiments, quantal fluctuation was usually

too great to estimate responses from single preparations in

which intracellular recording techniques were employed.

Consequently, data from a number of preparations were aver-

aged for analysis.

The standard bathing solution had the following compo-

sition (in mM): NaC1 116, KC1 2, CaC12 1.8, HEPES 2, glu-

cose 5. This solution was modified by reducing [Ca2+] to










0.4-0.7 mM and adding 5 mM Mg2+ to decrease transmitter

release. Osmolality was maintained by reducing NaCl. The

pH was adjusted to 7.2-7.4 using NaOH. Alterations of pH

of the bathing medium within this range did not appear to

have any obvious effect on control EPP amplitude or on

stimulation-induced increases in EPP amplitude. Exper-

iments were performed at 20-22 oC. Solution changes were

performed with Pasteur pipettes and typically required 1-2

min to complete. Amiloride, N-methylglucamine, ouabain,

and monensin were obtained from the Sigma Chemical Company.

Changes in the osmolality of the bathing medium have

been shown to have an effect on transmitter release (Fatt

and Katz, 1952; Furshpan, 1956). In solutions in which

extracellular Na+ was reduced by the removal of NaCl,

osmolality was maintained by the addition of sucrose in the

proportion: 183 mM sucrose = 100 mM NaCl (Birks and Cohen,

1968b). Solutions prepared to control for the effect of

Cl1 removal employed equimolal substitution with N-methyl-

glucamine chloride to maintain osmolality. In experiments

performed to ascertain the effect of increased extracellu-

lar [Na+] on increases in release, osmolality was increased

in the normal [Na+] bathing media by adding sucrose. The

order of exposure to different bathing solutions was varied

to control for possible long-term drift that can occur in a

preparation over time.








Data Analysis

The fractional change in EPP amplitude, V(t), is given

by the following formula:



V(t) = EPP(t)/EPP(0) 1 (eqn. 2-1)


where EPP(t) represents the end-plate potential amplitude

at time t, and EPP(O) represents the control (unfacili-

tated) EPP amplitude in the absence of repetitive stimula-

tion. A similar formula,



Vm(t) = m(t)/m(0) 1 (egn. 2-2)


is used to describe changes in quantal content (m) with

repetitive stimulation.

Three methods of estimating quantal content were used

(del Castillo and Katz, 1954a, Martin, 1966). Quantal con-

tent could be calculated directly using the equation



m = EPP/MEPP (eqn. 2-3)


where EPP and MEPP are the mean EPP and MEPP (miniature

end-plate potential) amplitudes, respectively. Indirect

estimates of quantal content were obtained from the method

of coefficient of variation and the method of failures as

shown in the equations below:








m = 1/(CV)2 (eqn. 2-4)

m = In (N/F) (eqn. 2-5)


where CV is the coefficient of variation, N is the total

number of stimulations and F is the number of times a stim-

ulation failed to evoke an EPP (del Castillo and Katz,

1954a). Data were discarded from analysis if the quantal

content estimates obtained by the methods described above

differed by more than 15%. Obtaining consistent estimates

of quantal content was a particular problem under condi-

tions of reduced external [Na+], which leads to a reduction

of MEPP amplitude (Fatt and Katz, 1952) and an increase in

quantal content (e.g., Birks and Cohen, 1965, 1968b). Both

of these effects of reduced [Na+]o reduce the accuracy of

the above methods of estimating quantal content (e.g., del

Castillo and Katz, 1954a).

In most experiments, differences between means were

tested for statistical significance using Student's t

tests. In the series of experiments conducted to assess

the effect of increased external [Na+] on V(t), the

increased variability of V(t) introduced by increases in

osmolality precluded the use of parametric tests of statis-

tical significance (see Results). In this situation, a

nonparametric test, the Wilcoxon rank-sum test for paired

experiments, was used to test for significant differences

between population distributions (Mendenhall and Scheaffer,

1973).








In addition to statistical testing of data collected at

discrete times during the conditioning trains, increases in

V(t) during 10 impulse conditioning trains were modeled as

described in Chapter 3. Differences between the model par-

ameters used to describe V(t) during 10 impulse trains were

also assessed for statistical significance in order to pro-

vide an additional test of the significance of effects

described in this chapter. These results will be discussed

in detail in Chapter 3.


Results

Effect of Reducing Extracellular Sodium Concentration
on V(t)

Reduction of the Na+ concentration of the solution

bathing the frog sartorius neuromuscular junction increased

both the control EPP amplitude (the amplitude of the first

EPP in a train) and the magnitude of the increase in EPP

amplitude which normally occurs during repetitive stimula-

tion under conditions of reduced quantal content. These

effects are illustrated in Figure 2-1, which shows EPPs

recorded during 10 impulse conditioning trains applied at a

stimulus frequency of 20 impulses/s in the presence of nor-

mal levels of extracellular Na+ (also referred to as '100%

Na+'; Figure 2-1A) and following reduction of [Na+] in the

bathing medium to 33% of normal (also referred to as '33%

Na+'; Figure 2-1B). In this experiment, the control EPP

amplitude was increased by 82% in the 33% Na+ Ringer's

solution. An increase in control EPP amplitude in 33% Na+
















A







I Ii I

1Q t* 1 | *i .











50 uV

50 ms

FIGURE 2-1. Effect of reduced external [Na+] on end-plate
potentials recorded extracellularly from the frog sartorius
nerve-muscle preparation during 10 impulse conditioning
trains. Each record is traced from a photograph of the
averaged responses from 32 trains of suprathreshold stimuli
delivered to the muscle nerve. Bathing media contained
0.4 mM Ca2+. The preparation was sequentially exposed to
bathing media containing 100% of normal [Na+] (A) and 33%
of normal [Na+] with osmolality maintained by sucrose
substitution (B). Both recordings were made at the same
voltage gain.








solutions was seen in 46 of 47 experiments. Control EPP

amplitudes measured in 33% Na+ solutions were an average of

2.5 times greater than control EPP amplitudes measured in

100% Na+ solutions. An increase in EPP amplitude under

conditions of reduced extracellular [Na+] has been

described previously at the frog neuromuscular junction,

and shown to be due to an increase in quantal content, or

the average number of quanta of transmitter released per

impulse (Birks and Cohen, 1965; Kelly, 1965; Colomo and

Rahamimoff, 1968).

In addition to the effect of reduced external [Na+] on

control EPP amplitude, there was a greater stimulation-

induced increase in EPP amplitude during 10 impulse trains

recorded in low-[Na+] bathing solutions. This latter

effect is more easily seen in Figure 2-2A, which presents

plots of V(t), the fractional increase in EPP amplitude

(equation 2-1 in Materials and Methods), as a function of

time during 10 impulse conditioning trains.

In the presence of normal levels of external Na+

(filled circles), V(t) was 0.28 at the time of the second

impulse (V(50 ms)), and reached a value of 0.84 by the

tenth impulse (V(450 ms)). When the concentration of Na+

in the bathing solution was reduced to 33% of normal (open

triangles), V(50 ms) increased to 0.54, and V(t) reached a

value of 1.40 by the tenth impulse. The time course of the

Na+ effect on V(t) during the 10 impulse train is more

clearly seen in Figure 2-2B, which shows the difference in


























FIGURE 2-2. Effect of reduced extracellular [Na+] on V(t),
the fractional increase in EPP amplitude during 10 impulse
trains. (A) Data from a single representative experiment
performed at a bath concentration of 0.6 mM Ca2+. Bathing
solutions contained 100% of normal [Na+] (filled circles)
or 33% of normal [Na+] (open triangles), with osmolality
maintained by the addition of sucrose. (B) Time course of
the increase in V(t) attributable to reduced external
[Na+]. Data are from the same experiment as (A), with the
ordinate representing the difference between V(t) measured
at 33% and 100% of normal external [Na+].








A
1.6-
1.4-A
1.2
.. 1.0 A1A
- 0.8- .0-*8
0.6- A *0
0.4-
0.2 /
0.0
0 100 200 300 400
TIME (ms)
B


0 0.6
LL UJ
, *m --*-m-- m

-- /
OE 0.4




0.0
0 100 200 300 400
TIME (ms)








V(t) between the trains of EPPs depicted in Figure 2-2A.

The increase in V(t) seen in the low [Na+] medium appears

early during the 10 impulse train and approaches a plateau

level by the end of the train. A qualitatively similar

increase in V(t) during exposure to 33% Na+ bathing media

was seen in 46 of 47 preparations. The remarkably consis-

tent occurrence and time course of this effect of reduced

external Na+ on V(t), presumably reflecting an effect on an

underlying component of stimulation-increased release, will

be analyzed in greater detail in the subsequent sections of

this chapter.


Specificity of the Effect of Extracellular Sodium
Reduction

In the experiments described in the preceding section,

NaCl was removed from the bathing solutions, resulting in

decreases in both [Na+] and [Cl-]. To rule out effects on

V(t) of changes in extracellular Cl-, sodium acetate was

substituted for NaCl to reduce extracellular [Cl-] to 33%

of the normal concentration. Substitution of acetate for

Cl1 had little or no effect on control EPP amplitude. In 4

of 4 experiments, no discernible effect on V(50 ms) of

reduced extracellular [Cl-] was seen (mean S.D.,

0.237 + 0.011, 100% NaC1; 0.231 + 0.020, 33% Cl1). In

acetate-containing media, there was a small but statisti-

cally insignificant increase in V(t) near the end of the 10

impulse train (data not shown). In these same experiments,

V(50 ms) was substantially increased by reduction of both








Na+ and Cl- to 33% of normal (mean + S.D., 0.301 + 0.022;

P < 0.005). Thus, reductions in extracellular [C1-] do not

appear to be involved in the increase in V(t) seen with

reductions in NaC1 in the bathing solution.

To rule out an effect of sucrose on V(t) and to corro-

borate the results of the anion substitution experiments,

Na+ was replaced isosmotically with the impermeant organic

cation N-methylglucamine. In 6 of 6 experiments, reduction

of NaC1 to 33% of normal levels resulted in a similar

increase in V(t) whether sucrose or N-methylglucamine chlo-

ride was used to maintain osmolality. This can be clearly

seen in Figure 2-3A, which shows data from a single exper-

iment in which NaCl was reduced to 33% of the normal level,

with isosmolal substitution by either sucrose (open

triangles) or N-methylglucamine chloride (inverted open

triangles). Substitution for Na+ by either of these com-

pounds produced virtually equivalent increases in V(t),

which became evident early during the conditioning train

and appeared to reach a steady state by the end of the

train. These results argue that neither reduction in

external [Cl1] nor osmotic substitution with sucrose are

responsible for the increase in V(t) seen in solutions con-

taining reduced amounts of NaC1.

Figure 2-3B shows the effect of extracellular [Na+]

reduction on EPP amplitude during the stimulus train in the

same preparation. It is obvious that while sucrose substi-

tution caused a marked increase in control EPP amplitude,


























FIGURE 2-3. Ionic specificity of the effect of reduced
external [Na+] on V(t) during 10 impulse trains. Data from
a representative experiment in which external [Na+] was
reduced to 33% of normal by isosmotic substitution of NaCI
with sucrose (open triangles) or N-methylglucamine chloride
(inverted open triangles). Data obtained in media contain-
ing 100% of normal NaCl are represented by filled circles.
All solutions contained 0.5 mM Ca2+. (A) Effect of reduced
external [Na+] on V(t). (B) Effect of reduced external
[Na+] on the amplitude of extracellularly recorded EPPs.
EPP amplitudes are normalized to the amplitude of the first
EPP of the train collected in 100% Na+ bathing medium.








A
1.2
1.0- 7=-X-
0.8- _o
, s .-.-*-
, 0.6- 0 /*
0.4- /
0.2-
0.0
0 100 200 300 400
TIME (ms)
B
260
_.A __A ---A-

S220- /

_1 180 A

< 140 -
a. 100

60t---
S100 200 300 400
TIME (ms)








N-methylglucamine substitution produced little or no

change in control EPP amplitude. A similar effect on con-

trol EPP amplitude was noted in 6 of 6 experiments in which

N-methylglucamine was substituted for Na+ in the bathing

medium. These results indicate that an increase in V(t)

may be produced relatively independently of an increase in

control EPP amplitude (see Discussion). The increase in

control EPP amplitude seen in low-[Na+] solutions contain-

ing sucrose may be due to a specific effect of sucrose, as

substitution of external [C1-] by acetate had little effect

on control EPP amplitude (see above).


Time Course and Reversibility of the Effect of
Extracellular Sodium Reduction on V(t)

In approximately two-thirds of experiments, an increase

in EPP amplitude and V(50 ms) could be detected during the

first conditioning train delivered following a reduction in

extracellular [Na+]. The increase in V(50 ms) required

5-10 minutes to reach a maximum. Figure 2-4 shows a plot

of V(50 ms) during a single experiment in which the Na+

concentration of the bathing medium was alternated between

100% and 33% of the control level. In this experiment,

paired pulses were delivered to the muscle nerve every

40 s. Reduction of extracellular [Na+] from 100% to 33% of

normal led to an approximate doubling of V(50 ms). The

effect of reduced [Na+] on V(t) could be readily reversed

by returning the preparation to the control Ringer's solu-













0.7 100% 33% 100% 33% 100%
NaCI N NaC NaCI NaCI NaCI

0.5 o il

E 0.4 0 *
0
O 0 0 0
c) *
*
> **
> 0.2- % N ".*

0.1-



TIME (min)


FIGURE 2-4. Time course and reversibility of the effect of
reduced external (Na+] on V(50 ms). Data from a single
representative experiment in which paired pulses, 50 ms
apart, were delivered to the sartorius muscle nerve every
40 s. All bathing media contained 0.6 mM Ca2+. In 33%
NaCl solutions, osmolality was maintained by the addition
of sucrose. Each data point represents the average of 4
trials, smoothed using a 3-point moving-bin method as
described in Rahamimoff and Yaari (1973), with a delta bin
of 1 point. Although in this experiment an initial peak of
V(50 ms) was seen shortly after changing from 100% Na+ to
33% Na+ solutions, this effect was not consistently
observed in other experiments.








tion, and reproduced by a repeated exposure of the prepara-

tion to a 33% Na+ bathing medium.


Concentration Dependence of the Extracellular Sodium
Effect on V(t)

The magnitude of the increase in V(50 ms) appeared to

be dependent upon the amount of reduction in extracellular

[Na+]. This is shown in Figure 2-5, which summarizes data

obtained from 27 preparations in 0.5 mM Ca2+ bathing solu-

tions in which V(50 ms) was measured at extracellular Na+

concentrations of 33-100% of normal. Although there was

considerable variability in both the control EPP amplitude

and the magnitude of the Na+ effect on V(t) from prepara-

tion to preparation, V(50 ms) was clearly increased as

extracellular [Na+] was reduced. The effect of reducing

extracellular [Na+] on V(50 ms) was also measured at extra-

cellular Ca2+ concentrations of 0.4 and 0.6 mM (these

measurements are later presented in more detail in Figures

2-7 and 2-8). The magnitude of the Na+ effect on V(50 ms)

was clearly dependent on the magnitude of the reduction in

external [Na+] for each concentration of [Ca2+]o tested.


Effect of Increased Extracellular Sodium on V(t)

To further characterize the concentration dependence of

the effect of extracellular sodium on V(t), experiments

were performed in which the [Na+] in the bathing medium was

increased above normal levels. However, increasing exter-

nal [Na+] also increases the osmolality of the bathing

solution. To control for this increase in osmolality,























*5*@*
T ..-**
0***


100


80


1..
I... /


60


PERCENT NORMAL [Na+]o



FIGURE 2-5. Dependence of V(50 ms) on external Na+ concen-
tration. Points represent mean + S.E.M. of data averaged
from experiments performed at an external Ca2+ concentra-
tion of 0.5 mM, using 10 impulse conditioning trains. In
solutions containing reduced Na+, osmolality was maintained
by the addition of sucrose. The following numbers of
experiments were averaged for each data point: 100% Na+
27; 50% Na+, 7; 33% Na', 18. The data point for V(50 ms)
at 67% Na+ represents a single experiment. Differences
between means are statistically significant for the 100%
and 50% Na+ data (P < 0.01), the 50% and 33% Na+ data
(P < 0.05), and the 100% and 33% Na+ data (P < 0.005).


0.36


CIO
E

Lo


0.32


I
0-*"


0.284


0.24


40


_ I








sucrose was added to solutions containing normal [Na+]. In

100% Na+ media, an increase in osmolality caused an

increase in control EPP amplitude and a marked reduction in

V(t) which was most pronounced near the end of the train

(data not shown). In solutions of equivalent osmolality,

increases in extracellular [Na+] cause a reduction in V(t)

which is obvious early in the course of the train, and

appears to reach a steady state by the end of the train.

This can be more clearly seen in Figure 2-6, which summa-

rizes data obtained from 6 preparations exposed to 100% Na+

bathing solutions (osmolality adjusted with sucrose; filled

squares) and 150% Na+ bathing solutions (open squares). An

easily reversible reduction in V(t) was noted in all six

preparations upon exposure to bathing solutions containing

150% of the normal Na+ concentration.

The magnitude of the reduction in V(t) due to increased

external [Na+] appeared to be concentration-dependent.

This can be more clearly seen in Table 2-1, which summa-

rizes the effects of increased external [Na+] on V(50 ms)

in 6 experiments in which [Na+]o was increased to 125% of

normal and in 6 experiments (also depicted in Figure 2-6)

in which [Na+]o was increased to 150% of normal. Increas-

ing [Na+]o to 125% of normal resulted in an average

reduction of 9.2% in V(50 ms), while increasing [Na+]o to

150% of normal resulted in an average reduction in V(50 ms)

of 19%. The results described above suggest that the con-

centration dependence of the Na+ effect on V(t) extends to
















0.6-


0.5- I 0-



>_ 0
> 0.3 1 "

0.2 -D

0.1-

0.0,
0 100 200 300 400
TIME (ms)

FIGURE 2-6. Effect of increased external [Na+] on V(t)
during 10 impulse conditioning trains. Data points repre-
sent mean + S.E.M. of V(t) measured in 6 experiments con-
ducted at an external [Ca +] of 0.5 mM. Bathing solutions
contained either 150% of normal NaCl (open squares) or 100%
of normal NaC1, with osmolality increased to 150% of normal
by the addition of sucrose (filled squares). The differ-
ence between the population distributions of the V(50 ms)
measurements in 100% and 150% NaC1 bathing solutions is
statistically significant (Wilcoxon rank-sum test for
paired experiments, P < 0.025).




















TABLE 2-1. EFFECT OF INCREASES IN EXTERNAL [Na+] ON
V(50 ms)


[Na+] Osmolality V(50 ms)
(% normal) (% normal) (mean + S.D.)

100a 125 0.229 + 0.031

125 125 0.208 + 0.020b


100a 150 0.182 + 0.074

150 150 0.147 + 0.022C

aOsmolality increased to that of the corresponding high-Na+
solution by the addition of sucrose.
bDifference from 100% Na+, 125% normal osmolality data is
statistically significant (Wilcoxon rank-sum test for
paired experiments, P = 0.05).
CDifference from 100% Na+, 150% normal osmolality data is
statistically significant (Wilcoxon rank-sum test for
paired experiments, P < 0.025).








conditions in which the external [Na+] has been increased

above normal concentrations. However, the magnitude of the

change in V(t) for a given increase in [Na+]o is smaller

than that seen with a similar reduction in extracellular

[Na+].


Effect of Extracellular Calcium Concentration on the
Magnitude of the Increase in V(t) Due to Reduced
Extracellular Sodium

Small elevations of extracellular [Ca2+] under low

quantal content conditions have been shown in a series of

related experiments performed in this laboratory to

increase V(t) late in the course of 10 impulse conditioning

trains (Mosier et al., 1986; see Discussion). This effect

of raised extracellular [Ca2+] on V(t) appears to be

distinguishable in its time course from the effect of

reduced extracellular [Na+] on V(t). Comparison of the

Ca2+ dependence of the Na+ effect on V(t) with the direct

effect of increased extracellular [Ca2+] on V(t) could

provide additional evidence as to whether these two effects

represent distinguishable processes underlying increased

release. Furthermore, a demonstrable Ca2+ dependence of

the Na+ effect on V(t) would be consistent with the

hypothesis that changes in extracellular Na+ may alter

release by affecting levels of Ca2+ or a Ca2+-activated

factor within the nerve terminal.

It was possible to measure the magnitude of the sodium

effect on V(t) at extracellular Ca2+ concentrations from

0.4 to 0.6 mM. At higher Ca2+ concentrations, EPPs often








evoked action potentials in the muscle during a condition-

ing train; at lower Ca2+ concentrations, EPP amplitudes

were often so small and variable as to preclude accurate

estimates of V(t). As is shown in Figure 2-7, the effect

of reduced extracellular [Na+] on V(t) is increased at

higher levels of extracellular [Ca2+]. At all concentra-

tions of extracellular Ca2+, the increase in V(t) is mani-

fested early during the conditioning train, and appears to

reach a steady state before the end of the train, similar

to the time course of the Na+ effect depicted in Figure

2-2B.

The Ca2+ dependence of the Na+ effect is more clearly

shown in Figure 2-8, which plots the dependence of V(50 ms)

on extracellular [Na+] at three different calcium concen-

trations. In this figure, which shows data averaged from a

number of experiments, the ordinate represents the increase

in V(50 ms) at a given extracellular Na+ concentration over

the value of V(50 ms) measured at normal Na+ concentra-

tions. The magnitude of the extracellular [Na+] effect on

V(50 ms) was virtually the same in bathing media containing

0.4 and 0.5 mM Ca2+. Increasing extracellular [Ca2+] to

0.6 mM causes a marked increase in the magnitude of the Na+

effect on V(50 ms). This effect of increased extracellular

[Ca2+] cannot easily be explained by a direct effect of

Ca2+ on V(t), as measurements of V(50 ms) differed by less

than 10% among the three Ca2+ concentrations tested at

normal Na+ concentrations.



















FIGURE 2-7. Magnitude of the effect of reduced external
[Na+] on V(t) at different concentrations of external Ca2+.
Data represent mean + S.E.M. Osmolality in all reduced-
[Na+] solutions was maintained by the addition of sucrose.
Solutions contained 100% (filled circles), 50% (open dia-
monds), or 33% (open triangles) of normal [Na+]. (A) Data
obtained in 0.4 mM Ca2+ bathing solutions. Data were
averaged from 14 experiments at 100% Na+, 7 experiments at
50% Na+, and 6 experiments at 33% Na+. Differences between
means of V(50 ms) data were statistically significant for
the 100% and 50% Na+ data (P < 0.025), and for the 100% and
33% Na+ data (P < 0.005), but not for the 50% and 33% Na+
data (P > 0.10). (B) Data obtained in 0.5 mM Ca2+ bathing
solutions. Data were averaged from 27 experiments at 100%
Na+, 7 experiments at 50% Na+, and 18 experiments at 33%
Na+. Differences between means of V(50 ms) values were
statistically significant for the 100% and 50% Na+ data
(P < 0.01), for the 50% and 33% Na+ data (P < 0.05), and
for the 100% and 33% Na+ data (P < 0.005). (C) Data
obtained in solutions containing 0.6 mM Ca2+. Data were
averaged from 17 experiments at 100% Na+, 7 experiments at
50% Na+, and 9 experiments at 33% Na+. Differences between
means of V(50 ms) values were statistically significant for
the 100% and 50% Na+ data (P < 0.005), for the 50% and 33%
Na+ data (P < 0.05), and for the 100% and 33% Na+ data
(P < 0.005).










A 0.4 mM Ca2+
1.2

1.0
0.8 T
-4-T
> 0.6 ---

0.4
0.2

0 100 200 300 400


B 0.5 mM Ca2+
1.2

1.0 ...

0.8
-4-,
0.6

0.4

0.2- LI

0.


100


200 300
TIME (ms)


400















C 0.6 mM Ca2+
1.2- T



T0.6 A---



0.02
o.oi- -- ^/--- --- --- ---

0.2 .



0 100 200 300 400
TIME (ms)


FIGURE 2-7 continued














E 60
60
0 /
Ld> 40 '



z *__ _
L 0 20

0 u)0 *

100 80 60 40 20

PERCENT NORMAL [Na+]o



FIGURE 2-8. Effect of external [Ca2+] on the magnitude of
the low-[Na+] effect on V(50 ms). Values of V(50 ms) were
obtained from the same data as Figure 2-7. Points
represent the percent increase of V(50 ms) over the value
of V(50 ms) recorded at 100% of normal [Na+], for each
concentration of external Ca2+. Data were obtained at
external Ca2+ concentrations of 0.4 (filled circles), 0.5
(open diamonds), and 0.6 mM Ca2+ (open triangles).








Lithium Partially Replaces Sodium in Maintaining V(t)

Lithium ions can substitute for sodium ions in a number

of sodium-dependent enzymes and transport mechanisms,

whereas lithium is a competitive antagonist of sodium in

other systems. Replacement of varying amounts (10-50%) of

extracellular sodium with lithium ions had only a small

effect on extracellularly recorded V(t). Figure 2-9 summa-

rizes data collected from 6 experiments in which 50% of the

Na+ in the bathing medium was replaced with Li+. In these

experiments, V(50 ms) measured in solutions containing 50%

Na+/50% Li+ was slightly higher than in control (100% Na+)

bathing solutions (mean S.D., 0.178 + 0.023, 100% Na+;

0.203 + 0.021, 50% Na+/50% Li+; P < 0.05). Reduction of

extracellular [Na+] to 50% of normal, with osmolality

maintained by the addition of sucrose, resulted in a larger

increase in V(50 ms) over control (0.228 + 0.044;

P < 0.025). These results suggest that lithium ions can

substitute for sodium ions in the mechanism underlying

early changes in V(t), albeit at somewhat reduced effic-

iency.


Intracellular Measurements of the Effect of Reduced
Extracellular Sodium on V(t)

Intracellular experiments were performed to confirm

that the observed increments in V(t) were presynaptic in

nature. Quantal content was estimated as described in the

Materials and Methods. In addition, MEPP amplitude and

resting MEPP frequency were measured directly from the












0.6-

0.5-

S0.4 .^.--

0.3-

0.2-

0.1-

0.01
0 100 200 300 400

TIME (ms)


FIGURE 2-9. The effect on V(t) of partial replacement of
sodium with lithium. Data averaged from 6 preparations in
bathing solutions containing 0.4 mM Ca2+. Bathing solu-
tions were alternated between media containing normal Na+
concentrations (filled circles), 50% of normal [Na+] with
isosmotic replacement by Li+ (open diamonds), and 50% of
normal [NaC1] with osmolality maintained by addition of
sucrose (open triangles).








screen of the storage oscilloscope. Reduction of extracel-

lular [Na+] to 50% of normal had no obvious effect on

resting MEPP frequency, but decreased MEPP amplitude (as

would be expected, since a large portion of the end-plate

current is carried by Na+; Takeuchi and Takeuchi, 1960).

This decrease in MEPP amplitude, together with the increase

in quantal content observed with decreases in [Na+]o (Birks

and Cohen, 1965, 1968b), made accurate measurements of

quantal content extremely difficult in low-Na+ solutions.

As a result, in only 3 of 20 intracellular experiments

could the effect of reduced external [Na+] on V(t) be

characterized (see Materials and Methods).

Figure 2-10 depicts the stimulation-induced increase in

both quantal content and intracellularly recorded EPP

amplitude in a representative experiment during 10 impulse

trains. Only a slight increase in quantal content (filled

circles) and EPP amplitude (open circles) was noted during

the 10 impulse train in solutions containing 100% of normal

[Na+]. However, in 50% Na+ solutions, a clear increase in

both quantal content (filled triangles) and EPP amplitude

(open triangles) was noted during the train. Under both

experimental conditions, increases in EPP amplitude are

paralleled by increases in quantal content, confirming that

changes in V(t) represent changes in transmitter release.

Therefore, the increase in V(t) due to reduction of exter-

nal [Na+] can be accounted for by an increase in transmit-

ter release. As was noted in the extracellularly recorded

















A
&:::.A


1.4-
1.2-
1.0-
0.8-
0.6-
0.4-
Oi-
0.2-
0.01
-0.2-
-0.4-
-0.6


100


.*A.. .** A>. .A -..A



S. 8::


"0 ."' 'O.


200


300


400


TIME (ms)



FIGURE 2-10. Intracellular measurements of the effect of
reduced external [Na+] on changes in quantal content during
10 impulse trains. Ordinates represent Vm(t), the change
in the quantal content with repetitive stimulation, and
V(t), the change in the corresponding EPP amplitude with
repetitive stimulation. Quantal content was measured by
the method of failures and the method of coefficient of
variation as described in the Materials and Methods sec-
tion. Data are from a single representative experiment
performed at a bath [Ca2+] of 0.5 mM. The filled circles
represent Vm(t) and the open circles represent V(t) in a
100% Na+ bathing medium. The filled triangles represent
Vm(t) and the open triangles represent V(t) in a 50% Na+
bathing medium, with osmolality maintained by the addition
of sucrose.


-4-,

E

>


-4-
\J


k .0.

.0 .-'
*-v '
IS .


1 1 1








data, much of the increase in quantal content due to

reduced external [Na+] occurred at the beginning of the

stimulus trains.


Effects of Manipulations of Extracellular Sodium on the
Presvnaptic Action Potential

One way that reduced extracellular [Na+] could affect

V(t) is through a stimulation-dependent change in the nerve

terminal action potential. In particular, an increase in

the width of the presynaptic action potential, causing a

longer duration of depolarization and a consequent increase

in Ca2+ entry, has been proposed to underlie increases in

transmitter release in Aplysia sensory neurons (Klein and

Kandel, 1978). In a small number of experiments, due to a

fortuitous placement of the extracellular recording elec-

trode, it was possible to record a presynaptic action

potential as well as the end-plate potential. Thus it was

feasible in these experiments to measure the width and

peak-to-peak amplitude of the extracellularly recorded pre-

synaptic action potential to determine if changes in V(t)

could be explained by changes in the shape of the presynap-

tic action potential.

Figure 2-11 plots the change in action potential width

during 10 impulse conditioning trains in solutions contain-

ing 100% of normal [Na+] (filled circles) and in solutions

containing 33% of normal [Na+] with osmolality maintained

by sucrose (open triangles) or N-methylglucamine (inverted

open triangles). There was little or no difference between













0.2-


0.1-


0.0 A


S--0.1 ,/ -- 2

-0.2


-0.35 -,, ,
0 100 200 300 400

TIME (ms)


FIGURE 2-11. Effect of reduced external [Na+] on the
change in presynaptic action potential width during repeti-
tive stimulation. Nerve terminal potentials were recorded
using a surface electrode. The ordinate represents the
ratio of the width of a given action potential to the width
of the first action potential in each train. Data are
averaged from a number of preparations, all at bath Ca2+
concentrations of 0.5 mM. The Na+ concentration of the
bathing solution was varied between 100% of normal (filled
circles, n = 10), 33% of normal with osmolality maintained
by sucrose (open triangles, n = 9), and 33% of normal with
osmolality maintained by N-methylglucamine chloride
(inverted open triangles, n = 4).








the stimulation-induced changes in action potential width

recorded in 100% Na+ or 33% Na+ bathing media. In these

same experiments, the average V(50 ms) increased from 0.274

to 0.358 in 100% and 33% Na+ media, respectively. Like-

wise, no obvious differences in the amplitude of the action

potential were noted with repetitive stimulation in 100% or

33% Na+ solutions (data not shown), although a small but

statistically insignificant decrease in action potential

amplitude was noted during the conditioning train under

both conditions. These findings are consistent with the

findings of previous workers (Hubbard and Schmidt, 1963;

Katz and Miledi, 1965; Lev-Tov and Rahamimoff, 1980), who

also demonstrated small decreases in the amplitude of

extracellularly recorded presynaptic action potentials with

repetitive stimulation. The above results suggest that

increases in V(t) due to reduction of external [Na+] cannot

be accounted for by stimulation-induced changes in action

potential width or amplitude.

As expected, reduction of extracellular [Na+] to 33% of

normal caused a slight reduction in the amplitude of the

control action potential (the first action potential of the

conditioning train) and an increase in the width of the

control action potential in 9 experiments. With prolonged

exposure of the preparation to reduced extracellular Na+,

there was a progressive increase in the width of the con-

trol action potential, but no additional increase in V(t)

beyond that seen immediately after the solution change.








Similar effects of reduced extracellular [Na+] on action

potential width and amplitude were seen in 4 preparations

in which N-methylglucamine rather than sucrose was used to

maintain osmolality, even though these two methods of Na+

substitution are markedly different in their effects on

control EPP amplitude. In a single experiment, the extra-

cellular action potential was measured during reduction of

NaCl to 33% of normal, with isosmotic substitution by lith-

ium chloride and sucrose. In this experiment, a similar

increase in action potential width was seen in the presence

of low-Na+ solutions substituted with either sucrose or

Li+, although Li+ substitution caused little or no increase

in either control EPP amplitude or V(t). These experiments

suggest that the changes in control action potential width

or amplitude seen in the presence of reduced [Na+]o have

little or no effect on either control EPP amplitude or

V(t).

The effect of elevated extracellular [Na+] on the

presynaptic action potential was also examined. No changes

in action potential amplitude or width with repetitive

stimulation were apparent in 3 experiments in which [Na+]o

was alternated between 100% and 150% of normal. In these

experiments, osmolality was held constant at 150% of nor-

mal. Likewise, there was little or no effect of raised

[Na+]o on control action potential amplitude or width (data

not shown). These experiments provide additional evidence

that changes in the presynaptic action potential play








little or no role in the changes in V(t) seen with alter-

ations of the extracellular Na+ concentration.


Effect of Monensin on V(t)

The experiments described thus far cannot distinguish

between an effect on V(t) due to a change in internal or

external [Na+] or an effect resulting from a change in the

driving force for Na+ entry into the nerve terminal.

Increasing the concentration of intracellular Na+ offers a

way to distinguish between these possibilities by reducing

the driving force for Na+ entry without affecting the

external Na+ concentration. If the effect of reduced

external [Na+] on V(t) is mediated by a corresponding

decrease in [Na+]i (Thomas, 1972; Deitmer and Schlue,

1983), then an increase in internal [Na+] should have the

opposite effect on V(t) as a reduction in external [Na+].

However, if the effect of reduced [Na+]o on V(t) is

mediated by a reduction in the driving force for Na+ across

the cell membrane, then an increase in internal [Na+]

should have the same effect on V(t) as a decrease in exter-

nal [Na+]. One way to examine the effect of increased

intracellular [Na+] on stimulation-increased release is the

addition of monensin, a sodium ionophore which inserts into

the plasma membrane and acts as a carrier to transport Na+

across the plasma membrane (Pressman and Fahim, 1982).

Addition of monensin (dissolved in ethanol) to the

bathing medium had no consistent effect on control EPP

amplitudes recorded extracellularly in 17 experiments (7








increased; 6 decreased; 4 little or no effect). However,

in 13 of the 17 experiments, a striking increase in

V(50 ms) was noted upon addition of 0.5-10 uM monensin dis-

solved in ethanol. In the remaining 4 experiments, addi-

tion of monensin had little or no effect on V(t). A simi-

lar concentration of ethanol alone (0.8%, v/v) had no dis-

cernible effect on control EPP amplitude or V(t) in 5

experiments. Results of a representative experiment are

shown in Figure 2-12A. In this experiment, the magnitude

of V(50 ms) is 0.35 in a normal [Na+] bathing solution, and

is increased to 0.98 after the addition of 2 uM monensin to

the bathing solution. Similar to the effect of reduced

extracellular [Na+] depicted in Fig. 2-1B, most of the

monensin-induced increase in V(t) occurred early during the

conditioning train, and reached a plateau by the end of the

train (Figure 2-12B). The effect of monensin on V(t) usu-

ally took 8-20 min to reach stable levels, although an

effect was sometimes noticeable during the first two or

three trains following its addition. The effect of monen-

sin on V(t) could be reversed readily by washing several

times with the control Ringer's solution, and could be

reproduced upon re-addition of monensin to the bathing

solution.

Effect of Ouabain on V(t)

Inhibition of the Na+/K+ pump with cardiac glycosides

has been shown to result in accumulation of intracellular

Na+ in a number of preparations (Thomas, 1972; Ellis and



























FIGURE 2-12. Effect of the Na+ ionophore monensin on V(t)
during 10 impulse trains. (A) Data from a representative
experiment performed at a bath [Ca2+] of 0.5 mM. The
bathing medium was alternated between a control solution
(filled circles) and a solution containing 2 uM monensin
(open diamonds). (B) Time course of the increase in V(t)
attributable to monensin. Data are from the same exper-
iment as (A), representing the difference between measures
of V(t) with and without monensin in the bathing medium.










.. 0 0---0
S^ + MON
0




MON

^____________


2.0

1.6

1.2

0.8

0.4-

0.0
0


B
1.61


200


300


400


TIME (ms)


100


200


360


400


TIME (ms)


A


100


1.2-


0.8-


0.4-


0.0


0


/ *








Deitmer, 1978; Deitmer and Schlue, 1983). As Na+/K+ pump

inhibition offers an alternative method of increasing

[Na+]i at the frog neuromuscular junction, experiments were

performed to ascertain the effect of the cardiac glycoside

ouabain on increased release during 10 impulse conditioning

trains.

Surprisingly, addition of 2.8-20 uM ouabain to the

bathing medium had little or no effect on V(t) during 10

impulse trains in 8 of 10 experiments, although V(t)

appeared to increase in 1 experiment and decrease in 1

experiment. After a 15-60 minute delay, ouabain also led

to a rapid increase in EPP amplitude, similar to the find-

ings of Birks and Cohen (1968a). More prolonged (35-90

min) exposure to ouabain resulted in the sudden onset of

failure of neuromuscular transmission. This is probably a

result of failure of impulse conduction in the nerve to the

sartorius muscle, as has been described previously (Birks

and Cohen, 1968a,b; Baker and Crawford, 1975).

The apparent lack of effect of ouabain on V(t) may be

due to the effects of other factors which may have obscured

any effect of increased [Na+]i on V(t). Inhibition of the

Na+/K+ pump not only leads to an increase in internal [Na+]

but also to a reduction of internal [K+], with concomitant

membrane depolarization (e.g., Banks, 1967). The latter

effects of Na+/K+ pump inhibition may affect transmitter

release independently of changes in external or internal

Na+. Furthermore, the striking increase in quantal release








observed with ouabain treatment may have led to a depletion

of quanta available for release, causing depression of

transmitter release during conditioning trains (e.g.,

Thies, 1965). Finally, measurements of V(t) during ouabain

exposure were only feasible in the interval preceding the

phase of rapid increases in control EPP amplitude, when the

accumulation of internal Na+ may have been insufficient to

affect the driving force for Na+ entry. Further exper-

iments employing alternative methods of increasing internal

[Na+], such as the introduction of Na+ into the nerve

terminal with liposomes (e.g., Rahamimoff et al., 1978),

would be desirable in order to clarify these difficulties.

Effect of Amiloride on V(t)

The effects on V(t) of reduced [Na+]o, which leads to

reduced [Na+]i, and monensin, which increases [Na+]i, are

qualitatively similar, suggesting that a mechanism depen-

dent upon the transmembrane Na+ gradient may underlie the

expression of V(t) during 10 impulse trains (see Discus-

sion). One mechanism which may depend upon the transmem-

brane Na+ gradient for its operation is the Na+/H+

exchanger, a carrier protein which regulates intracellular

pH by Na+-gradient-driven removal of H+ from the cytosol.

Intracellular pH changes are known to affect cytosolic Ca2+

handling and a number of membrane electrical properties in

excitable cells (reviewed in Moody, 1984), and could con-

ceivably play a role in the expression of stimulation-

induced increases in release.








To determine whether a Na+/H+ exchange mechanism is

involved in the expression of V(t) during short condition-

ing trains, a number of preparations were exposed to the

Na+/H+ exchange inhibitor amiloride. Since amiloride has

been reported to have little or no effect on Na+/Ca2+

exchange in squid axons (Allen and Baker, 1986), the use of

amiloride offers a way to distinguish between the effects

on V(t) of Na+/H+ exchange and Na+/Ca2+ exchange. The

amiloride concentration required to achieve half-maximal

inhibition of the Na+/H+ exchanger has been reported to be

in the range of 2-5 uM in a number of non-renal cells (Fre-

lin et al., 1987). As can be seen in Figure 2-13, which

presents averaged data from 6 preparations, exposure of the

frog sartorius nerve-muscle preparation to amiloride con-

centrations as high as 50 or 500 uM resulted in little or

no change in V(t) during 10 impulse conditioning trains.

No obvious change in V(t) during exposure to amiloride was

seen in any of the individual experiments which were aver-

aged for Figure 2-13. It can be concluded that the Na+/H+

exchange mechanism probably plays little role in the

expression of the increase in V(t) seen with reductions of

extracellular [Na+].


Discussion

Effect on V(t) of Alterations in Sodium is Distinguishable
from the Effect of Altered External Calcium

At least four components of increased transmitter

release have been defined at the frog neuromuscular junc-



















-4-,


1.2-

1.0-

0.8-

0.6-


0.4

0.2

0.0
0


100 200 300 400


TIME (ms)


FIGURE 2-13. Effect of the Na+/H+-exchange inhibitor
amiloride on V(t) during 10 impulse trains. Data are aver-
aged from a number of experiments in which V(t) was meas-
ured in solutions containing 0.5 mM Ca2+. Solutions are:
normal [Na+] without amiloride (filled circles, n = 6);
normal [Na+] with 50 uM amiloride added (inverted open
triangles, n = 4); normal [Na+] containing 500 uM amiloride
(open diamonds, n = 2); and 33% of normal [Na+] without
amiloride (open triangles, n = 2).








tion. These are believed to reflect four distinct biophys-

ical or biochemical processes which occur within the nerve

terminal during repetitive stimulation. While changes in

intra-terminal calcium pools are thought to underlie many,

if not all, of the components of increased release, sodium

ions have also been proposed to play a key role in regulat-

ing one or more of these pools. The data presented in this

dissertation suggest that the first component of facilita-

tion, a process with a decay constant of 50-60 ms (Zengel

and Magleby, 1982), may be selectively affected by manipu-

lations of extracellular and intracellular [Na+].

Alterations in external [Ca2+] have little effect on

V(50 ms) in solutions containing normal concentrations of

external Na+, while they appear to increase a longer-

lasting component of increased release (Mosier et al.,

1986). The effect of increases in external [Ca2+] on V(t)

is demonstrated in Figure 2-14A, which plots averaged V(t)

data obtained in this laboratory. As the external [Ca2+]

is increased from 0.4 mM (filled squares) to 0.6 mM (open

squares), V(t) is increased. However, the increase in V(t)

is small early in the course of the 10 impulse train, and

reaches its greatest magnitude at the end of the train.

This can be seen more clearly in Figure 2-14B, which plots

the difference between the V(t) data shown in Figure 2-14A.

It is obvious that the change in V(t) attributable to

increased external [Ca2+] continues to increase during the




























FIGURE 2-14. Effect of increased external [Ca2+] on V(t).
(A) Averaged V(t) values from 10 impulse conditioning
trains. Data represent mean + S.E.M. of V(t) from 15
experiments in media containing 0.4 mM Ca2 (filled
squares) and 10 experiments in media containing 0.6 mM Ca2+
(open squares). (B) Time course of the increase in V(t)
attributable to increased external [Ca2+]. Data are from
the same experiments as (A), with ordinates representing
the differences between V(t) measured at external Ca2+
concentrations of 0.4 and 0.6 mM.
























200 300
TIME (ms)


200
TIME


300
(ms)


1.0

0.8

0.6

0.4

0.2


-I-


400


0.3-


0.2-


0.1-


0.0


100


400


IA.


/z*








course of the 10 impulse conditioning train, and shows no

evidence of approaching a plateau during the train.

On the other hand, changes in external Na+ exert their

principal effect on V(t) early during short conditioning

trains delivered to the muscle nerve. The Na+ effect on

V(t) appears to reach a steady state by the end of the 10

impulse train. Comparing Figure 2-2B, which plots the

increase in V(t) attributable to reduced external [Na+],

with Figure 2-14B, which plots the increase in V(t) attrib-

utable to increased external [Ca2+], shows that the effect

of reduced [Na+]o on V(t) is markedly distinct from the

previously described effect of increased [Ca2+]o on V(t).


Localization of the Site of Action of Sodium Ions

As all preparations were stimulated at five times the

threshold voltage required to produce a maximal control EPP

amplitude as recorded by the extracellular electrode, a

change in the measured increase in V(t) is unlikely to be

due to recruitment of axons. Furthermore, the increase in

V(t) due to reduced external [Na+] could be demonstrated

with intracellular as well as extracellular recording tech-

niques, consistent with the hypothesis that the Na+ effect

on V(t) is homosynaptic, i.e., capable of being expressed

at a single synapse (see Materials and Methods).

Stimulation-induced changes in the presynaptic action

potential are also an unlikely explanation for the observed

increases in transmitter release. Facilitation and longer-

term components of stimulation-increased transmitter








release are not associated with any consistent changes in

the amplitude or width of the presynaptic action potential

during repetitive stimulation at a number of crustacean,

amphibian, and mammalian synapses (Martin and Pilar, 1964;

Ortiz, 1972; Zucker, 1974; Charlton and Bittner, 1978;

Zucker and Lara-Estrella, 1979; Baldo et al., 1983). Fur-

thermore, measurements of presynaptic action potential

amplitude have actually shown a decrease with repetitive

stimulation in some preparations (Katz and Miledi, 1965;

Braun and Schmidt, 1966; Zucker, 1974; Lev-Tov and Rahami-

moff, 1980), which cannot account for the stimulation-

induced increase in EPP amplitude observed. Injection of

Ca2+ into the presynaptic terminal of the squid giant

synapse results in facilitation of transmitter release

without a change in spike waveform (Charlton et al., 1982).

Facilitation of release can also be demonstrated in syn-

apses in which the action potential has been blocked with

tetrodotoxin and the nerve terminal stimulated focally with

repetitive impulses of constant amplitude and duration

(Zucker, 1974).

Consistent with the aforementioned findings, no

stimulation-induced changes in presynaptic action potential

amplitude or width which could account for an increase in

V(t) were noted in this study. Furthermore, in a number of

experiments, marked changes in control action potential

amplitude could be detected without a significant change in

V(t). Extracellular measurements of the presynaptic action








potential are not sufficiently sensitive to exclude subtle

changes in calcium currents occurring at the nerve termi-

nal; however, the lack of a direct effect of changes in

extracellular [Ca2+] on the early component of facilitation

argues against this possibility.

Postsynaptic changes are also unlikely to contribute

to the low-Na+-induced increase in V(t). Desensitization,

the only stimulation-dependent postsynaptic change that has

been described at neuromuscular junctions, requires tens of

seconds of repetitive stimulation or of continuous exposure

to acetylcholine for its expression (e.g., Thesleff, 1955;

Magleby and Pallotta, 1981; Ruzzier and Scuka, 1986), and

decreases rather than increases the effect of acetylcholine

on the postsynaptic membrane. In experiments using intra-

cellular recording techniques, changes in V(t) could be

fully explained by changes in quantal release during repet-

itive stimulation.

The effects of manipulations of [Na+] on V(t) do not

appear to be a function of their effects on control levels

of transmitter release. While substitution of NaCl with

sucrose often markedly increases both control EPP amplitude

and V(t) in this preparation, substitution of NaCl with

N-methylglucamine chloride increases V(t) with no signifi-

cant effect on control EPP amplitude. Monensin, a Na+

ionophore, dramatically increases V(t) in many preparations

with no consistent effect on control EPP amplitude. On the

other hand, prolonged exposure to ouabain results in little








or no increase in V(t) while producing a marked increase in

control EPP amplitude.


Specificity of the Sodium Effect on V(t)

The results of ion substitution experiments indicate

that the increase in V(t) due to NaCl reduction is specific

for the sodium ion. Replacement of Cl1 by organic anions

such as methylsulfate has been shown to have little effect

on muscle membrane potential or MEPP frequency at frog

neuromuscular junctions, although replacement with more

lyotropic anions such as bromide or nitrate may increase

MEPP frequency (Muchnik and Gage, 1968). Substitution of

Cl1 by acetate ion has little or no effect on V(50 ms),

although it may affect a slower component of increased

release. Substitution of Na+ by N-methylglucamine, an

impermeant cation, has a similar effect on V(t) as does

substitution of NaC1 by sucrose. These results suggest

that the increase in V(t) is due to a reduction in external

[Na+] rather than a reduction of Cl" or a specific effect

of sucrose. The results of substitution with Li+, a per-

meant cation, indicate that isosmotic substitution with Li+

can partially but not completely prevent the increase in

V(t) which was seen in the presence of reduced external

Na+. This may indicate that Li+ interacts, albeit somewhat

weakly, at the site of action of external Na+; alterna-

tively, Li+ may have other actions which could modify V(t),

e.g., an effect on the phosphatidylinositol system

(reviewed in Downes, 1983).








Possible Mechanisms for the Sodium-Induced Increase in V(t)

A stimulation-dependent increase in Ca2+ entry into the

nerve terminal could theoretically underlie an increase in

V(t); however, evidence supporting such a mechanism is

lacking at the frog neuromuscular junction. The amount of

Ca influx per impulse into molluscan or leech neurons or

squid nerve terminals does not seem to increase during

short trains of repetitive stimulation (Smith and Zucker,

1980; Charlton et al., 1982; Ross et al., 1987; but see

Barrett and Morita, 1987). In the experiments described

above, increases in extracellular Ca2+ had little or no

direct effect on V(50 ms) during short conditioning trains,

although they did seem to increase a longer-lasting compo-

nent of increased release.

The observed effects of extracellular [Na+] changes and

the Na+ ionophore monensin are most consistent with a

mechanism which depends upon the transmembrane Na+ gradient

of the nerve terminal. Relief of competition between

extracellular Na+ and Ca2+ for entry into the nerve

terminal, as was first proposed by Colomo and Rahamimoff

(1968), could conceivably account for the effect of reduced

extracellular [Na+] on increased release. However, the

similar increase in V(t) seen with application of monensin,

an ionophore known to increase intracellular [Na+], is more

consistent with a mechanism driven by the transmembrane Na+

gradient. The effect of monensin, which increases intracel-

lular [Na+], could possibly be explained by a mechanism








dependent upon the concentration of intracellular Na+

(reviewed in Rahamimoff et al., 1980). However, the simi-

lar effect on V(t) of reduced extracellular [Na+], which

causes a slow decrease in intracellular [Na+] (Thomas,

1972), cannot be easily explained by this hypothesis. The

similar effects on V(t) of manipulations which decrease

extracellular [Na+] or increase intracellular [Na+] are

best explained by a mechanism involving the reduction in

the Na+ gradient across the nerve terminal membrane.

One difficulty with the above interpretation is the

seeming lack of effect of ouabain, which should increase

intracellular [Na+], on V(t). However, ouabain may have

other effects which act to obscure any increase in V(t) due

to elevated intraterminal [Na+], e.g., synaptic depression

due to the marked increase in transmitter release which

occurs after a short exposure to ouabain.

If the effect on V(t) of manipulations of extracellular

and intracellular Na+ concentrations is mediated through a

mechanism dependent upon the transmembrane Na+ gradient,

what mechanisms might be proposed to underlie this effect?

An effect mediated by voltage-dependent sodium channels is

unlikely due to the marked dependence of the magnitude of

the Na+ effect on extracellular Ca2+. Intracellular pH

changes mediated by Na+/H+ exchange have been proposed as a

prerequisite for intracellular Ca2+ mobilization in other

systems (e.g., Siffert and Akkerman, 1987), and decreases

in intracellular pH have been measured during trains of








action potentials in molluscan neurons (Ahmed and Connor,

1980). Conceivably a Na+/H+ exchanger could affect

increased release either directly or indirectly by an

alteration in intracellular pH. However, the Na+/H+

exchange mechanism is not known to be significantly

affected by extracellular [Ca2+] (Grinstein and Cohen,

1987). Furthermore, reduction of external [Na+] and appli-

cation of the Na+ ionophore monensin, which produced quali-

tatively similar increases in V(t) in this study, are known

to have opposite effects on intracellular pH in other sys-

tems (acidification and alkalinization, respectively; e.g.,

Ellis and MacLeod, 1985; Grinstein and Cohen, 1987;

Lichtshtein et al., 1979). Finally, application of the

Na+/H+ exchange inhibitor amiloride had little or no effect

on V(t) even at concentrations sufficient to achieve a com-

plete block of the Na+/H+ exchanger.

A Na+/Ca2+ exchange mechanism has been proposed to be

involved in one or more stimulation-induced changes in

transmitter release at vertebrate neuromuscular junctions

(Parnas et al., 1982; Meiri et al., 1986; Misler et al.,

1987). Na+/Ca2+ exchange is believed to play an important

role in the regulation of intracellular Ca2+ levels in many

biological systems (reviewed in Blaustein, 1988), and a

Na+/Ca2+ exchange mechanism has been characterized in

preparations of nerve endings from rat brain (Nachshen et

al., 1986; Sanchez-Armass and Blaustein, 1987). The

Na+/Ca2+ exchanger in rat brain synaptosomes may be able to








extrude Ca2+ sufficiently quickly to match Ca2+ influx at

firing frequencies up to 18/s (Sanchez-Armass and Blaus-

tein, 1987; Blaustein, 1988). Such a rate of Ca2+ extru-

sion would be sufficiently rapid to play a role in even the

shortest component of facilitation at the nerve terminal.

The partial efficacy of Li+ in preventing an increase in

V(t) due to reduction of external [Na+] would also be

consistent with an effect on a Na+/Ca2+ exchanger, as Li+

has been shown to reduce the Vmax of the exchange reaction

in purified synaptic plasma membrane vesicles (Hermoni et

al., 1987). Such a mechanism would appear to be the most

parsimonious explanation for the presumably Na+ gradient-

dependent process characterized in this study, although

some other mechanism, as yet uncharacterized, cannot be

ruled out.


Conclusion

At least four processes act to increase transmitter

release during repetitive stimulation; some or all of these

may involve the buildup of residual Ca2+ within the nerve

terminal between successive impulses. Sodium ions have been

proposed to play a role in one or more of these four pro-

cesses. The data presented in this study support the

hypothesis that an early component of increased release,

with a time course similar to the previously defined first

component of facilitation (Zengel and Magleby, 1982), is

selectively affected by manipulations of extracellular Na+.





68

The apparent dependence of this process on the transmem-

brane Na+ gradient and external Ca2+ concentration suggest

a Na+-Ca2+ exchange mechanism, although other explanations

may be advanced. Further research will be necessary to

test this hypothesis, as well as to elucidate the mecha-

nisms underlying the other components of increased release

at the neuromuscular junction.













CHAPTER 3

MODELING OF THE EFFECT OF SODIUM ON
STIMULATION-INDUCED INCREASES IN RELEASE


Introduction


As previously discussed in Chapter 2, four processes of

increased release have been described at the frog neuromus-

cular junction: two components of facilitation with time

constants of decay of approximately 60 and 500 ms, augmen-

tation which decays with a time constant of about 7 s, and

potentiation which has a time constant of decay of tens of

seconds to minutes (Zengel and Magleby, 1982). It is

likely that the different components of increased release

reflect different underlying biochemical or biophysical

processes which affect transmitter release. Besides marked

differences in the kinetic properties of the components of

increased release, several findings have added to the evi-

dence that these components are separable and act rela-

tively independently of one another. At frog neuromuscular

junctions, externally applied Sr2+ and Ba2+ ions selec-

tively increase the second component of facilitation and

augmentation, respectively, of both evoked release and

miniature end-plate potential (MEPP) frequency (Zengel and

Magleby, 1980, 1981).








In the first part of this dissertation, reduction of

external [Na+] was shown to have a selective effect on

increased release early in the time course of short condi-

tioning trains at the frog neuromuscular junction. In this

chapter, a slightly modified version of the model of

stimulation-induced increases in transmitter release devel-

oped by Magleby and Zengel (1982) will be used to quantita-

tively describe sodium-induced changes in release in terms

of the different components of increased release. The

process of increased release described previously, which is

selectively affected by manipulations of extracellular and

intracellular Na+ concentrations, will be shown to corre-

spond in its kinetic properties to the first component of

facilitation.



Materials and Methods

The preparation, solutions, stimulation paradigms, and

recording techniques were the same as those described in

Chapter 2 of this dissertation.

Modeling of V(t) data was done using a simplified form

of the quantitative model for stimulation-induced increases

in transmitter release described by Magleby and Zengel

(1982). In this model, the effect of repetitive stimula-

tion on the four components of increased release are

assumed to result from changes in underlying factors in the

nerve terminal which increase transmitter release. An

impulse invading the nerve terminal is assumed to add an








increment to each of the underlying factors, which then

decay with time.

Several models of increased release during short trains

have been proposed (e.g., Mallart and Martin, 1967; Barrett

and Stevens, 1972; Balnave and Gage, 1977; Zengel and

Magleby, 1982). In general, facilitation can best be

described by a power model of order 3 or 4 (Barrett and

Stevens, 1972; Younkin, 1974; Zengel and Magleby, 1982).

Power models of facilitation bear a close resemblance to

the third- or fourth-power relationships between calcium

and transmitter release that have been described at a vari-

ety of synapses by a number of workers (e.g., Dodge and

Rahamimoff, 1967; Katz and Miledi, 1970). In this disser-

tation, a third-power relationship between the factors

responsible for the first and second components of facili-

tation, or F1 and F2, and the observed increase in EPP

amplitude due to facilitation, or F, was assumed for pur-

poses of data analysis, giving the equation



F + 1 = (F1* + F2* + 1)3 (eqn. 3-1)


where FI* and F2* represent the factors responsible for F1

and F2, the first and second components of facilitation.

The relationship between facilitation and augmentation

was assumed to be multiplicative, as Zengel and Magleby

(1982) have shown quantitatively at the frog neuromuscular

junction. Potentiation has also been described by a multi-








plicative relationship to facilitation (Landau et al.,

1973; Magleby, 1973). These relationships can be described

by the equation



V + 1 = [F1* + F2* + 1]3 [A + 1] [P + 1] (eqn. 3-2)


where A is the magnitude of augmentation, and P is the

magnitude of potentiation. The relationships described by

this equation can successfully account for stimulation-

induced changes in release under a wide variety of stimulus

frequencies and conditioning train lengths (Magleby and

Zengel, 1982).

Since the small magnitude and long time course of

potentiation (tens of seconds to minutes) would allow no

distinction from augmentation during the 10 impulse trains

used in this study, these two processes of increased

release were combined into a single term for modeling pur-

poses. This assumption results in the simplified equation



V + 1 = [F1* + F2* + 1]3 [A' + 1] (eqn. 3-3)


in which A' represents the magnitude of the combined aug-

mentation and potentiation terms (hereafter referred to as

augmentation).

Increments in the factors underlying the two components

of facilitation, which are added at the time of each

impulse, were assumed to be constant during the train (Mal-








lart and Martin, 1967; Barrett and Stevens, 1972; Zengel

and Magleby, 1982). The increment in augmentation added by

each impulse has been found to increase with successive

impulses during long conditioning trains of over 100

impulses (Zengel and Magleby, 1982). For analysis of 10

impulse trains, however, the increment in augmentation was

assumed to remain constant as no significant difference in

the magnitude of augmentation would have been apparent

during such a short conditioning train.

Following a nerve impulse, the magnitudes of the fac-

tors underlying the two components of facilitation and

augmentation decay with first-order kinetics with time

constants of about 55 ms, 475 ms, and 7 s, respectively

(Zengel and Magleby, 1982). Each component builds up dur-

ing repetitive stimulation because there is insufficient

time for the increment of each underlying factor added

during a nerve impulse to decay before the next impulse

invades the nerve terminal.

The basic assumptions of the model described above are

summarized in Figure 3-1. In this diagram of the increase

in total facilitation during a conditioning train, or F,

the underlying factors Fl* and F2* are each assumed to

acquire a constant increment in magnitude (fl* and f2*,

respectively) with each nerve impulse. Between impulses,

the total magnitude of each component decays exponentially

with a time constant which is approximately 55 ms for Fl*

and 475 ms for F2*. As can be seen in Figure 3-1A, the



















FIGURE 3-1. Components of the model of stimulation-
induced increases in transmitter release. Each component
of increased release is assumed to be related to accumula-
tion of some underlying factor in the nerve terminal, which
is increased by a constant amount during a nerve impulse
and which decays with first-order kinetics between nerve
impulses. (A) Accumulation of Fl*, the factor assumed to
underlie the first component of facilitation. With each
nerve impulse, a constant increment (fl*) is added to Fl*.
Model parameters are: fl 0.173; time constant of decay of
Fl, 55 ms. (B) Accumulation of F2*, the factor assumed to
underlie the second component of facilitation. As in (A)
above, a constant increment (f2*) is added to F2 with each
nerve impulse. Model parameters are: f2*, 0.0119; time
constant of decay of F2, 475 ms. (C) Facilitation during
10 impulse trains, predicted by the power facilitation
relation of equation 3-1, in which total facilitation, F,
is related to the third power of FI* and F2*. (D) Modeling
of increased release during 10 impulse conditioning trains.
The filled circles depict data from the representative
experiment shown in Figure 2-2A, which were measured during
exposure of the preparation to a bathing solution contain-
ing 0.6 mM Ca2+ and 100% of normal [Na+]. The best fit of
the model (equation 3-3) to the data is shown by the solid
line. Model parameters are: fl*, f2* as in (A) and (B);
increment in augmentation (a'), 0.0118; time constant of
decay of augmentation (A'), 7.0 s.










A
0.3-


0.2
\
LL
0.1


0.0- .. .

B
0.3


0.2

LL-
0.1


0.0 ,
0 100 200 300 400 500 600
TIME (ms)












1.6


1.2

LL 0.8


0.4

0.0-

D

0.8-

0.6

-> 0.4
0.4


0.2

0.0.


FIGURE 3-1 continued


V(t)

F(t)


Fl(t)


0 100 200 300 400 500 600
TIME (ms)








magnitude of the increment fl* (represented by the vertical

deflection at the time of each impulse) remains constant

during the train. The magnitude of Fl* present at the time

of each stimulus has reached a plateau before the end of

the 10 impulse train. Likewise, the magnitude of f2*,

represented by the vertical deflection at the time of each

impulse in Figure 3-1B, remains constant. However, because

of the slower decay of F2*, the magnitude of F2* continues

to increase throughout the train. The result of combining

the effects of these two factors in a power facilitation

model (see equation 3-1) is seen in Figure 3-1C. The

increase of A', or the term describing the combined effect

of augmentation and potentiation, is not depicted but obeys

similar considerations.

Data on V(t) obtained from the experiments presented in

Chapter 2 were modeled using Lotus 1-2-3 graphic templates

on an IBM PC computer. Unless otherwise noted, time con-

stants of decay of the components of increased release were

fixed at 55 ms for F1, 475 ms for F2, and 7.0 s for A'

(Zengel and Magleby, 1982), while the magnitudes of the

increments fl*, f2*, and a', the increment in the factor

underlying the effects of augmentation and potentiation,

were varied. As will be shown in this chapter, changes in

the time constants of decay of the components of increased

release could not adequately describe changes in V(t) due

to experimental manipulations. The magnitudes of fl*, f2*,

and a' were chosen to minimize the sum of squared error








(SSE) of the difference between the model predictions and

the actual V(t) data. The values for fl*, f2*, and a'

determined by this method were similar to published values

for these parameters (Zengel and Magleby, 1982).

In some experiments, there was considerable variability

in the data, such that a close fit giving reliable esti-

mates of the model parameters was impossible to achieve.

Data from an experiment were discarded from analysis if the

minimum attainable SSE/V(450 ms) ratio exceeded 0.01. In

order to determine whether this criterion for exclusion

introduced a selection bias, modeling was also done on the

averaged data from all experiments performed under a given

experimental condition and the results compared with data

from experiments analyzed separately. Sets of model par-

ameters derived from these two methods of analysis differed

by less than 5% in all cases, indicating that a selection

bias was unlikely.

An example of data analyzed using the simplified model

(equation 3-3) is shown in Figure 3-1D. The filled circles

plot V(t) data from the representative experiment shown in

Figure 2-2A, which were recorded from a preparation in a

bathing solution containing 0.6 mM Ca2+ and 100% of normal

[Na+]. The solid line shows the predicted rise of V(t)

obtained by using values of fl*, f2*, and a' selected to

give the best fit of the model predictions to the data

(fl* = 0.173, f2* = 0.0119, a' = 0.0118). The relative

contributions of the two components of facilitation and of








augmentation to V(t) are shown by setting f2* and a' to

zero (long-dashed line) and by setting only a' to zero

(short-dashed line). It is obvious that at 50 ms, the

first component of facilitation makes the predominant con-

tribution to V(t), while at later times during the train,

the effects of the longer term components of increased

release become more prominent.

Differences between the model parameters derived from best

fits to experimental data were tested for statistical signifi-

cance using Student's t tests.


Results


Effect of Reduced Extracellular Sodium on Components of
Increased Release

In the previous chapter, reduction of extracellular

[Na+] was shown to cause an increase in V(t) which was most

prominent near the beginning of a 10 impulse conditioning

train and appeared to approach a plateau by the end of the

train. As the time course of this increase in V(t) is

similar to that of the F1 component of facilitation (Mal-

lart and Martin, 1967; Zengel and Magleby, 1982), the V(t)

data from the low-Na+ experiments were modeled in order to

test the hypothesis that the increase in V(t) seen with

reductions in external [Na+] can be explained by a selec-

tive increase in the magnitude of the F1 component of

facilitation.

Figure 3-2 plots V(t) data collected in 100% Na+

(filled circles) and 33% Na+ solutions (open triangles)











1.6

1.4-

1.2-

1.0 .
4-'
S0.8






0 100 200 300 400

TIME (mns)

FIGURE 3-2. Modeling of the effect of reduced external
[Na+] on stimulation-induced changes in release. V(t) data
are from a single representative experiment (also shown in
Figure 2-2) in which V(t) was measured under conditions of
100% (filled circles) and 33% (open triangles) of normal
external [Na+)]. The bath contained 0.6 mM Ca+. The solid
line represents the best fit of the model (described in the
Methods) to the 100% Na+ data, with the following parame-
ters: fl* = 0.173, f2 = 0.0119, a' = 0.0118. The
long-dashed line represents the best fit to the 33% Na+
data, with f* = 0.341, f2* = 0.0119, and a' = 0.0118.
0.2 -














Selectively increasing the magnitude of 2* from 0.0119 to
0.0885 to account for the increase in V(50 as) due to 33%
Na+ media resulted in a marked overprediction of all
subsequent data points (dotted line). Assuming a selective
increase in the time constant of decay of Fm (from 55 ms to
225 i s) sufficient to account for the rise (n V(50 i s) seen
in the 33% Na+ solution resulted in an even greater over-
prediction of subsequent data points (short-dashed line).








from the representative experiment shown in Figure 2-2A,

along with model predictions fitted to the V(t) data. The

100% Na+ data can best be described by the following

parameters: fl* = 0.173, f2* = 0.0119, a' = 0.0118 (solid

line). The model parameters are then changed to give the

best fit to the V(t) data measured in a bathing medium

containing 33% of normal NaC1. The low-[Na+] data can best

be described by the following parameters: fl*= 0.341, f2* =

0.0119, a' = 0.0118 (long-dashed line). The change in V(t)

attributable to reduced [Na+]o could be described by

changing only the magnitude of fl*, which was increased by

97%. It was typical of most preparations that the effect

of reduced external [Na+] could be described quite well by

a change in the magnitude of fl*, with only minor changes in

the magnitudes of f2* and a' to improve the fit to the last

few impulses of the train (these findings will be discussed

in detail later in this chapter).

Changes in f2* or a' could not account for the data.

For example, if a change in F2 is assumed, and values for

f2* chosen to fit the increase in V(50 ms) attributable to
reduced external [Na+], the model drastically overpredicts

V(t) by the end of the 10 impulse train, as is shown by the

dotted line in Figure 3-2. Changes in a' chosen to account

for the observed increase in V(50 ms) also lead to massive

overpredictions by the end of the stimulus train (data not

shown). A selective change in the time constant of decay

of F1 was also unable to account for the observed increase








in V(t), as is shown by the short-dashed line of Figure

3-2. A change in the time constant of decay of F1 from 55

ms to 225 ms, chosen to account for the increase in V(50

ms) due to reduced external [Na+], consistently overpre-

dicts V(t) for all subsequent impulses during the train.

As in the previous chapter, the effect of reduced [Na+]

on V(t) appeared to be ion-specific. Modeling of V(t) data

obtained in reduced [Na+] solutions in which N-methylgluca-

mine chloride was used to maintain osmolality showed a

similar result to modeling of V(t) data obtained when

sucrose was used to maintain osmolality. The results are

presented in Table 3-1, which presents estimates of the

magnitudes of fl*, f2*, and a' derived from 8 experimental

preparations in bathing media containing 0.5 mM Ca2+.

These data show an increase of 38% in the magnitude of fl*

(from 0.147 to 0.203) following reduction of extracellular

[Na+] to 33% of normal with isosmotic substitution by

sucrose. An increase (49%) was also noted upon reduction

of [Na+]o to 33% of normal with isosmotic substitution by

N-methylglucamine. Small but statistically insignificant

increases were also noted in the estimated magnitudes of

f2* for the N-methylglucamine data, the importance of which
is unclear at the present time.

It can be concluded that increases in V(t) due to

reduced external [Na+] can most easily be explained by

assuming a selective increase in the magnitude of the F1

component of facilitation. No other single change in the



















TABLE 3-1.


EFFECT OF REDUCED EXTERNAL [Na+] ON MODEL
PARAMETERS DESCRIBING INCREASED RELEASE.


[Na+] Substituted Parameter Estimates (mean)
(% normal) n bya fl* f2* a'

100 8 -- 0.147 0.0086 0.0110

33 6 Sucrose 0.203b 0.0086 0.0110

33 8 NMG+ 0.220c 0.0126 0.0119

aExternal [Na+] was reduced by removing NaCI and substituting
with an isosmotic amount of either sucrose or N-methylgluca-
mine chloride.
bSignificantly different from 100% Na+ data (P < 0.005).
CSignificantly different from 100% Na+ data (P < 0.025). Not
significantly different from 33% NaCl/sucrose data (P > 0.10).








model parameters, either in the magnitudes of longer term

components of increased release or in the time constant of

decay of the F1 component, appears to be adequate to

describe the effect of low external [Na+] on V(t) during 10

impulse conditioning trains. A scenario in which more than

one significant change occurs in the magnitudes or time

constants of decay of the processes of increased release is

not excluded by this analysis, but would be a less parsimo-

nious explanation of the observed data.


Effect of Increased External Sodium on Components of
Increased Release

Increasing the concentration of Na+ in the bathing

medium led to a small decrease in V(t), which could be

described by a selective decrease in fl as is shown in

Table 3-2. This table summarizes data from 6 preparations

in bathing media containing 0.5 mM Ca2+. The concentration

of external Na+ was alternated between 150% of normal and

100% of normal (with osmolality maintained by the addition

of sucrose). The results were consistent with a selective

reduction in the magnitude of fl* by an increase in

external [Na+].


Comparison of the Effect of Reduced External Sodium with
the Effect of Increased External Calcium on the
Components of Increased Release

The effect of decreased [Na+] on V(t) appears to be

qualitatively different from the effect of increased [Ca2+]

on V(t), as has been reported previously (Mosier et al.,






85









oo
o 0
oo
*

0
gr +1 +1 -4



E4 0 0 0g
0 0 a

;.C



M o o c m
r.- 0%
0 o 0 o o
0 0 0 0
O 0 0 rq 0

0 +1 *
o0 C +1 +1 0 v
go IV cl
+ O) w o +
0 0 0 0(
'- .o 4J
1) < 0 0 0 +0
SM e OO
*0
4.) I0 O0
1+ +1+
SQ 0) 0


E- % 0 o wA
20 00 0 V 0
wQ 0 0 0



HW a)
H 0 H
4M +1 +u
OH o o
P z a $4



144 H 0


z 0 *d 0 V
H4)





M 44O0C
-0 ( J o4
S'-'01 r-4 04
+ () m 0 0 0

044 CO








1986). In contrast to the sodium effect, the increase in

V(t) caused by elevations in external [Ca2+] could best be

described by a selective increase in f2*. To further

illustrate this point, data from this laboratory showing

V(t) during 10 impulse trains at two different external

Ca2+ concentrations are shown in Figure 3-3A. This figure

represents averaged data from 15 experiments conducted with

0.4 mM Ca2+ in the bathing solution (filled squares), and

10 experiments performed with 0.6 mM Ca2+ in the bathing

solution (open squares). The best fit of the model (solid

line) to the 0.4 mM Ca2+ data is obtained with the follow-

ing parameters: fl*, 0.103; f2*, 0.0082; a', 0.0110. The

model parameters were then changed to give the best fit to

the 0.6 mM Ca2+ data, resulting in the following values:

fl*, 0.114, f2*, 0.0206; a', 0.0106. The greatest change

observed was in the magnitude of f2*, which increased by

151%, while little or no change was observed in the magni-

tudes of the other incremental factors.

Changes in F1 could not account for the Ca2+ effect.

If the magnitude of fl* was selectively increased to

account for the increase in V(50 ms) seen in 0.6 mM Ca2+

solutions, all subsequent values of V(t) during the train

were underpredicted (dotted line in Figure 3-3B). Alterna-

tively, if a selective increase in the time constant of

decay of the F1 component were assumed to occur, a better

fit could be obtained to V(t) for the first 5 impulses

(short-dashed line in Figure 3-3B), but subsequent values






















FIGURE 3-3. Modeling of the effect of increased external
[Ca2+] on V(t) during 10 impulse trains. Data represent
averaged V(t) values from 15 experiments in media contain-
ing 0.4 mM Ca2+ (filled squares) and 10 experiments in
media containing 0.6 mM Ca2+ (open squares), also shown in
Figure 2-14. (A) Best fit of the model to the 0.4 mM Ca2+
data, with parameters fl* = 0.103, f2 = 0.0082, and
a' = 0.0110 (solid line). Parameters were then changed to
give the best fit to the 0.6 mM Ca2+ data, with
fl* = 0.114, f2* = 0.0206, a' = 0.0106 (long-dashed line).
(B) Attempts to account for the Ca2+-induced increase in
V(t) by selective changes in other model parameters. The
solid line represents the best fit of the model to the
0.4 mM Ca2+ data as in (A). Selectively increasing the
magnitude of fl* (from 0.103 to 0.142) to account for the
Ca +-induced increase in V(50 ms) resulted in underpredic-
tion of all subsequent data points (dotted line). A selec-
tive increase in the time constant of decay of F1 (from 55
to 85 ms) to account for the Ca2+-induced increase in
V(50 ms) gave a good fit to data points early in the train,
but underpredicted the last four points (short-dashed
line).










A

1.0

0.8
EEr
-. 0.6 -B
>4-j

S0.4-

0.2

0.01
0 100 200 30i

B TIME (ms)

1.0

0.8

, 0.6 ..a

> 0.4

0.2-

0.0
0 100 200 30


,-*


0 400


0 400


TIME (ms)


w








of V(t) were underpredicted. In contrast, increasing the

magnitude of a' could also account for the increase in V(t)

during during 10 impulse trains, but changes in a' did not

result in as good a fit (see Materials and Methods) as did

changes in f2* (data not shown). Thus it appears that the

increase in V(t) due to increased extracellular [Ca2+] can

be explained by an increase in the magnitude of the second

component of facilitation or of a longer-lasting component

of increased release, and that this effect is distinct from

the increase in V(t) which occurs in reduced extracellular

[Na+].


External Calcium Dependence of the Effect of Reduced
External Sodium on Components of Increased Release

In Figure 2-7 of the previous chapter, an increase in

extracellular [Ca2+] was shown to increase the magnitude of

the effect of reduced external [Na+] on V(t) in a manner

which appeared to differ from the direct effect of

increased external [Ca2+] on V(t). To determine whether

these apparently disparate effects on V(t) could be

explained by a change in a single underlying factor, the

data presented in Figures 2-7 were modeled using the sim-

plified version of the Zengel and Magleby model. The

results obtained from modeling, which are presented in

Table 3-3, show that at any given concentration of external

Ca2+, the magnitude of the estimated fl* increment appears

to increase in a concentration-dependent manner with

decreasing external [Na+]. In addition, the magnitude of

















%0
0
0
0
*
0

0 +1



0
0

V *

S 0
0
(0 0

+1 0
C i cN +|
0a C0
0 r'i
0


00



) 0



4. '- +1
a 0
M 0n

0 0


U)








0




2
0



+
N
0










z
x 0







I-4
OM









!0
+r1
(0

CQ
M





0 Q





2 M





NW
M
U (
M
m


VC p


200





+-I
z 4)


0 N
0 0
0 0
* *

+1 +1


I-4 0
0 0
0 0





0 0
0 0
0 0

+1 +1
N m
In m
0 0
0 0


0 AQ
N v*
0 0
0 0

+1 +1



* *0
0 0











in n


N
N
0
0

+1


r-4
0
0


In r-
o H
0 0



+1 +1
0 %0
Hi 0
0 0
0 0o
c C
0 0
o o
0 0
0 0

+1 +1
ON N

0 0
0 0
0 m

0 0
0 C;
0 0
0 0

+1 +1
In r-i


0 0


N u-I


0 m
in m


m r-4 cO
-4 0N 0


0 0 0
+1 +1 +1
CO N
0 0 0
0 0 0
0 0 0


0 -4 N
in m w
0 0 0
S0 0


+1 +1 +1
N N (A

0 0 0
0 0 o0


) 4.4
N 0 N
0 0 0
0 0 0

+1 +1 +1




C; N N
0 0 0





S-


I
+ p
IV -M I



ai *1 n
-,400
I O0 c




Wd oC
-10 0 Uto
0a in 4 *4

S$ 00





M-. .I.
0 -0 + -0
S c a00
44 Q) 00
-l $4 ft dp *
r -) -0 0 V



-- ;oo
N M 0


0 0-
a a m 41
*o-+ 4- 0 (




0 o e0 (go 0



m a
*-c- 0 + 2
0 4-4 0 -* t0
-.r-4 rO in Z d
v 9 N 0

+0 0 '0 40










oooo
0C -4 V 0



+ .1 4 C


40 4) t0
dI4 0 4 4 .
0 + t0o
00 >i 0I 44 '0
r- r-4 Z44
VJ -4+


4-1 r-- OW
0 4-4 0 4-) 0


$4 -I -4 0
W+ W V4J4 4
4-4 C C
-4 5-4 0 PO dp 4) -4 C
0 *-.-e C4
o0 4n 4i)a)
0n 4.40 54



O 4 A -4I n
--4-4 4V 0
4 V- t o V-
c 9 ou 4
--4 4 4 4-4 V 0
ca 0 (0-4 U
(0 44 1O C N -


C .
1-4








the low-[Na+]o effect on fl* appears greater at higher

concentrations of external Ca2+. This can be more clearly

seen in Figure 3-4, in which the ordinate represents the

increase in fl* at a given concentration of external Na+

over the value of fl* measured at normal external Na+

concentrations. Although there is no apparent difference

in the magnitude of the effect of reduced external [Na+] on

fl* as external [Ca2+] rises from 0.4 mM to 0.5 mM, a

further increase in [Ca2+]o to 0.6 mM produces a striking

increase in the magnitude of the low-[Na+] effect on fl*

This figure bears a remarkable resemblance to Figure 2-8,

which shows an increase in the Na+ effect on V(50 ms) in

0.6 mM Ca2+ media.

There appear to be no consistent changes in the effect

of low external [Na+] on the magnitudes of f2* or a' during

increases in external [Ca2+]. Thus, the simplest explana-

tion for the effect of increased external [Ca2+] on the

magnitude of the low-[Na+] effect on V(t) appears to be a

selective Ca2+-dependence of the effect of reduced external

[Na+] on the magnitude of fl*, the increment in the

proposed factor underlying the first component of facilita-

tion.


Effect of Monensin on Components of Increased Release

Addition of 2 uM monensin, an ionophore selective for

Na+ (Pressman and Fahim, 1982) to a bathing solution

containing 0.5 mM Ca2+ caused a marked increase in V(t)

which appeared early during a 10 impulse conditioning train