Citation
The Effects of repetitive stimulation on synaptic transmission in the chick ciliary ganglion

Material Information

Title:
The Effects of repetitive stimulation on synaptic transmission in the chick ciliary ganglion
Creator:
Poage, Robert Eliot, 1964-
Publication Date:
Language:
English
Physical Description:
v, 93 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Action potentials ( jstor )
Calcium ( jstor )
Frogs ( jstor )
Ganglia ( jstor )
Nerves ( jstor )
Neurons ( jstor )
Neurotransmitters ( jstor )
Synapses ( jstor )
Time constants ( jstor )
Transmitters ( jstor )
Action Potentials ( mesh )
Calcium Channels -- physiology ( mesh )
Chick Embryo -- physiology ( mesh )
Ciliary Body -- physiology ( mesh )
Department of Neuroscience thesis Ph.D ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Neuroscience -- UF ( mesh )
Electric Stimulation ( mesh )
Electrophysiology ( mesh )
Ganglia -- physiology ( mesh )
Synaptic Transmission -- physiology ( mesh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 82-92).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Robert Eliot Poage.

Record Information

Source Institution:
University of Florida
Rights Management:
Copyright Robert Eliot Poage. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
50397168 ( OCLC )
028475030 ( ALEPH )

Downloads

This item has the following downloads:


Full Text











THE EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC TRANSMISSION
IN THE CHICK CILIARY GANGLION










By
Robert E. Poage














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













ACKNOWLEDGEMENTS

I would like to thank the many educators who have contributed to my continued pursuit of higher education, most notably my advisor, Dr. Janet Zengel. I have been very fortunate to have studied under and learned from a scientist and teacher who is as dedicated and gifted as any I have met. When I look back on years spent in graduate study I will always be grateful for the opportunities I was given and the patience that was always applied liberally. From my mentor I have gathered many of the critical skills a scientist must have to be successful. From my friend I have seen how a scientist can pursue her life's work with enthusiasm, integrity and a critical eye, turned as closely inward as outward.

My graduate work at the University of Florida has been marked by
some excellent instruction, and opportunities to teach and to present my work in front of peers. In retrospect, I am grateful for all of these things.

My supervisory committee members, Dr. Janet Zengel, Dr. Stuart
Dryer, Dr. Peter Anderson, Dr. Philip Posner and Dr. Tom Vickroy, have my gratitude for having confidence in my work when successes were slow in coming. I thank them for their many suggestions and their support.
Finally, I would like to thank my parents, who always supported my
m ,. I,,. A, }h n t r~ .n ah fiel a na

















TABLE OF CONTENTS


ACKNOWLEDGEMENTS................................. ....... ......

ABSTRACT.......................................................... iv

CHAPTERS

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

Chemical Synaptic Transmission...... ....................... 1
The Ciliary Ganglion of the Embryonic Chick.................. 2
Stimulation-Induced Changes in Synaptic Efficacy............. 4
Voltage-Dependent Calcium Channels and Transmitter Release.. 5 The Presynaptic Action Potential and Transmitter Release..... 7 Summary........................................ ........... 8

2 EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC EFFICACY
IN THE CHICK CILIARY GANGLION.......................... 10

Methods....... ............................................... 12
Results............................... ....................... 18
Discussion........................... ........................ 42

3 CHARACTERIZATION OF CA2+ CHANNELS INVOLVED IN SYNAPTIC
TRANSMISSION............................................ 45

Methods...................................................... 47
Results....................................... ............. 48
Discussion....... .............. ............... ............ 50

4 EFFECT OF REPETITIVE STIMULATION ON THE PRESYNAPTIC
ACTION POTENTIAL..... ................................... 54

Methods.......................... ................. 55
Results....... ............................................. 56
Discussion........ . ...... ............. ............ 68
Notes...... ..SC........................................ 75














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


THE EFFECTS OF REPETITIVE STIMULATION ON
SYNAPTIC TRANSMISSION IN THE CHICK CILIARY GANGLION
By
Robert E. Poage

August 1995
Chair: Janet E. Zengel, Ph.D.
Major Department: Neuroscience
Under appropriate conditions, repetitive synaptic stimulation can cause subsequent stimuli to release increased amounts of neurotransmitter. Repetitive stimulation can also change the shape of the presynaptic action potential, but there are few experimental preparations in which the relationship between these two phenomena may be studied directly. The goals of this study were to use electrophysiological recording techniques to investigate the role of Ca2+ channels in initiation of release in the embryonic chick ciliary ganglion, to characterize stimulation-induced increases of synaptic efficacy at this synapse, and to record electrical activity from the presynaptic element of this synapse under conditions conducive to facilitation of release.
It is shown that treatments reported to block Ca2+ currents in









role for these previously described Ca2+ currents in initiation of transmitter release in the chick ciliary ganglion.

It is shown that four distinct components contribute to increased

ganglionic efficacy following repetitive stimulation. Their time constants of decay (60 milliseconds, 400 milliseconds, 30 seconds and 200 seconds) are similar to those that describe the decay of the four components of stimulation-induced increases in transmitter release described in other preparations (first and second components of facilitation, augmentation and potentiation). The components described here are also similar to the above-mentioned processes in their sensitivities to Sr2+ and Ba2+. It is concluded that the components of stimulus-induced increases in release that have been described in other synaptic preparations are present in the chick ciliary ganglion.
Repetitive stimulation under conditions conducive to facilitation of transmitter release causes an increase in the duration of the presynaptic action potential and a decrease in the amplitude of its afterhyperpolarization (AHP). The time course of the effects of repetitive stimulation on the presynaptic action potential parallels the time course of facilitation in this preparation. It appears likely that the observed effects on action potential duration result from a decrease in K+ conductance.

Results are discussed in terms of possible mechanisms underlying
the individual components of stimulation-induced increases of release.














CHAPTER 1
INTRODUCTION

Chemical Synaptic Transmission

Although the release of chemical neurotransmitter substances
mediates many forms of neuronal communication, the cellular mechanisms underlying neurotransmitter release have yet to be identified. It has been established that depolarization of the presynaptic nerve terminal causes the activation of voltage-dependent Ca2+ channels and influx of Ca2+ ions (Katz and Miledi, 1967). The resulting increase in Ca2+ concentration acts through an unknown mechanism to initiate release of neurotransmitter. It is generally accepted that the resulting increase in Ca2+ concentration causes synaptic vesicles within the nerve terminal to fuse with the nerve terminal membrane and spill their contents (neurotransmitters) into the synaptic cleft via exocytosis. Diffusion of the neurotransmitter across the cleft and its binding to specific postsynaptic receptors results in the transmission of information, either in the form of an increased ionic permeability or through the action of second messenger systems. To understand neurotransmitter release and its integral role i n information processing and plasticity, it will first be necessary to understand how the release process is









terminal depolarization and release of transmitter is on the order of a millisecond (Katz and Miledi, 1965). With very few exceptions the nerve terminal is inaccessible to neurophysiological recording techniques, primarily because of the size of the presynaptic elements involved. It would be of great interest to study many aspects of the release process at a single synapse, but most experimental preparations are not amenable to a wide range of available techniques.
The Ciliary Ganqlion of the Embryonic Chick
The calyx nerve terminal in the ciliary ganglion of the embryonic

and posthatch chicken is one notable exception in as much as the structures comprising the pre- and post-synaptic elements of the synapse are large enough to be studied directly using standard electrophysiological recording techniques. The accessibility of fertile eggs and the fact that the cellular elements adapt well to cell culture enhance the value of the preparation.
The ciliary ganglion is innervated by the third cranial nerve.

Preganglionic fibers originate in the approximately 2,000 cells of the accessory motor nucleus, the avian analog of the mammalian EdingerWestphal nucleus. As the oculomotor nerve passes through the orbit, it gives off branches to the muscles controlling eye movement. The remaining fibers enter the ciliary ganglion, which is located behind the eye lateral to the optic nerve. Fibers entering the ganglion form two distinct types of synapses on two separate neuronal subpopulations.







3

nerves and pierce the sclera to innervate the striated ciliary muscles and constrictor muscles of the iris. The smaller choroid neurons are innervated by multiple bouton-type synapses and project through three to five choroid nerves to innervate the choroidal coat.

Synaptic contacts on both choroid and ciliary neurons are chemical and cholinergic in nature. Calyx/ciliary neuron synapses also display electrical coupling. The nature and size of the calical nerve terminal and synapse provide a unique experimental opportunity to observe electrical activity in the nerve terminal of a vertebrate neuronal synapse. Martin and Pilar (1963a) reported the unique nature of this synapse and successfully recorded electrical activity in both pre-and postsynaptic structures. In a series of elegant reports (Martin and Pilar, 1963a,b, 1964a-c), they described transmission in the calyx preparation, including properties of both electrical and chemical coupling. Since those early descriptions, the chick ciliary ganglion has been used by several laboratories to study neurotransmitter release and synaptic function (Bennett and Ho, 1991; Dryer and Chiappinelli, 1985; Stanley, 1989; Stanley and Goping, 1991; Yawo, 1990).
The chick ciliary ganglion is obviously a preparation of great

value to the neuroscientist. Due to the accessibility of the presynaptic element of the calyciform synapse in the chick ciliary ganglion, I have chosen to use this preparation to study the relationship between presynaptic electrical activity and transmitter release.
Thi nis a .rtati Jo Ir~sI. 11 al 4.....:n 4) su- d n e 4 n









the relationship between the presynaptic action potential and stimulation-induced increases in postsynaptic response amplitude.
Stimulation-Induced Chances in Synaptic Efficacy

Since the nervous system most often uses trains of electrical signals to convey information, one important element of the study of neurotransmission is to observe what happens to transmitter release during and following repetitive stimulation. Many studies have shown that the efficacy of synaptic transmission is affected by its prior activity (c.f. Feng, 1941), but the cellular machinery involved is almost as obscure now as it was fifty years ago.

Repetitive stimuli applied to a presynaptic axon under conditions of low levels of release can lead to a progressive increase in the amount of transmitter released by successive impulses (reviewed by Zucker, 1989). Following stimulation, this increase in release decays back to control levels with a time course that can range from milliseconds to minutes. Such stimulation-induced increases in release have been studied most extensively at the frog neuromuscular junction, where four distinct components have been described on the basis of their kinetic and pharmacological properties. These components are the first and second components of facilitation, which decay back to control levels of release with time constants of about 60 ms and 400 ms, respectively (Magleby, 1973; Mallart and Martin, 1967; Zengel and Magleby, 1982); augmentation, which decays with a time constant of approximately
7 nnS/F~ln Sn Rhain. a-a8 aaleb *n Ze*l n976I ..nd.








5

superior cervical ganglion (Zengel et al., 1980), in rat (Hubbard, 1963; Liley, 1956) and crayfish (Zucker, 1974) neuromuscular junctions, in the squid giant synapse (Charlton and Bittner, 1978a), in cat spinal cord (Curtis and Eccles, 1960; Kuno, 1964; Porter et al., 1970) and in rat hippocampus (McNaughton, 1982).
I have characterized the kinetic and pharmacological properties of
stimulation-induced increases in synaptic efficacy in the chick ciliary ganglion. These results, which are presented in Chapter 2, will show that there are 4 components contributing to stimulation-induced increases in synaptic efficacy and that these changes result from an increase in chemical synaptic transmission.
Voltage-Dependent Calcium Channels and Transmitter Release
It has been proposed that accumulation of Ca2+ in the nerve terminal may be responsible for activity-dependent increases in neurotransmitter release (Katz and Miledi, 1968; Rosenthal, 1969; Weinreich, 1971). Results of experiments designed to test Katz and Miledi's "residual Ca2+ hypothesis" support the involvement of Ca2+ as a mediator of increased release, but suggest that Ca2+ must be acting at several steps or sites in the release process to produce the observed pattern of results (Landau et al., 1973; Zengel and Magleby, 1977, 1980; Zengel et al., 1993a,b, 1994). One element that will clearly affect the concentration of Ca2+ present in the nerve terminal is Ca2+ influx through voltage-gated Ca2+ channels.
u of a2+ i cea fn h a re









compared to the N-, L- and T-type channels described in chick dorsal root ganglion cells (Fox et al., 1987a; Nowycky et al., 1985). T-type channels produce transient membrane currents and have relatively low conductances. L-type channels are noninactivating and produce longlasting currents. N-type channels are neither transient nor slowly inactivating. A more recently described class of channels that is activated by moderate depolarization appears to be most common in mammalilan neurons and has been designated "P-type" (Llinas et al., 1989).
In the chick ciliary ganglion, Ca2+ currents recorded under voltage clamp from the calyx fail to meet the criteria for a single class of Ca2+ channel, although they are most similar to the N-type group (Stanley, 1991; Stanley and Atrakchi, 1990). Calical Ca2+ currents are insensitive to blockade by dihydropyridines and are blocked by w-conotoxin, consistent with an N-type classification, but they inactivate slowly, if at all (Stanley and Goping, 1991). Stanley (1991) has called these channels "NpT-"type (for neuronal-presynaptic terminal). There is evidence that Ca2+ channels at other fast-transmitting synapses may have similar kinetic and pharmacological properties (Miller, 1987; Suszkiw et al., 1986; Yoshikami et al., 1989). These results, taken with the finding that Ca2+ channels are located, possibly in clusters, on the release face of the calical nerve terminal membrane (Stanley, 1991), imply that the calical Ca2+ currents described above may be acting in the release process. However, Ca2+ channels involved







7

complete understanding of the Ca2+ channels present in the calyx and the role these channels play in the initiation and regulation of transmitter release.
The Presynaptic Action Potential and Transmitter Release
The role of nerve terminal depolarization in triggering exocytosis has been thought to be primarily through the activation of voltageactivated Ca2+ channels and the subsequent rapid increase in local intracellular Ca2+, although a direct role for depolarization in initiating release has been proposed (Dudel et al., 1983; Hochner et al., 1989; Silinsky et al., 1995). It stands to reason that the ability to study the electrical activity of the presynaptic element of a synapse is essential if the process of transmitter release is to be well understood.
Since stimulation-induced changes in release are a general phenomenon seen at most synapses, it is of interest to discern the mechanism or mechanisms acting to produce these effects on the release process. In attempting to describe these underlying mechanisms, many investigators have examined the role of the presynaptic action potential. If the depolarization of the nerve terminal caused by the invasion of an action potential is larger in amplitude or is prolonged (Hubbard and Schmidt, 1963; Liley and North, 1953; Takeuchi and Takeuchi, 1962), the resulting increase in activation of voltage-dependent Ca2+ channels should lead to a larger influx of Ca2+ and increased release (e.g.









and potentiation. However, Martin and Pilar compared the amplitudes of individual action potentials, and it is possible that subtle changes in the action potential amplitude and/or duration would not have been detected by this type of analysis. I have characterized the effects of repetitive stimulation on the presynaptic action potential in the chick ciliary ganglion. The results of these studies are presented in chapter 4 and possible contributions of changes in presynaptic electrical activity to the processes of stimulation-induced increases in release are discussed.

Summary
The major goal of this research project was to obtain a better
understanding of the mechanisms underlying stimulation-induced changes in transmitter release. Experiments were performed using the ciliary ganglion of the embryonic chicken.
I found that the chick ciliary ganglion responds to repetitive
stimulation with four components of increased release that are analogous to the first and second components of facilitation, augmentation and potentiation, as have been described in other preparations. Both kinetic (time constants of decay) and pharmacological (response to Sr2+, Ba2+) properties of these components were used in their identification, and intracellular recording from postsynaptic cells verified their presynaptic origin.
I found that the pharmacological characteristics of Ca2+ channels
Co-e toeoearnmte reeae-r siia to th N-to chne









I also found that there are changes in the presynaptic action

potential during repetitive stimulation that have a similar time course to facilitation, but a definitive role for changes in the presynaptic action potential was not demonstrated. Possible mechanisms that could underlie stimulation-induced changes in the action potential are discussed.














CHAPTER 2

EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC EFFICACY IN THE CHICK CILIARY GANGLION

For more than fifty years it has been known that synaptic efficacy changes as a result of prior synaptic activity (for example, Feng, 1941). It has been well documented in a variety of synapses that these changes arise from a change in the amount of neurotransmitter released by each successive impulse during repetitive stimulation. At the frog neuromuscular junction, where stimulation-induced changes in transmitter release have been studied most extensively, four components of increased release have been identified on the basis of their kinetic and pharmacological properties. These components are: the first and second components of facilitation, which decay back to control levels of release with time constants of approximately 60 and 400 msec (Magleby, 1973; Mallart and Martin, 1967; Younkin, 1974; Zengel and Magleby, 1982); augmentation, which decays with a time constant of approximately 7 sec (Erulkar and Rahamimoff, 1978; Magleby and Zengel, 1976); and potentiation, which decays with a time constant of tens of seconds to minutes (Magleby and Zengel, 1975a,b; Rosenthal, 1969). Some or all of these components have been observed in the rabbit supe-







11

Liley, 1956; Nussinovitch and Rahamimoff, 1988), cat spinal cord (Curtis and Eccles, 1960; Kuno, 1964; Porter, 1970) and rat hippocampus (McNaughton, 1982).
The striking similarities in kinetics and pharmacological sensitivities of these components of increased release in different preparations indicate that they may represent general phenomena that occur at all synapses. However, the subcellular machinery subserving these modulations of transmitter release has not yet been identified. The chick ciliary ganglion is an ideal synapse to study modulation of transmitter release because stimulation-induced increases in transmitter release have been reported (Martin and Pilar, 1964c), although they have not been fully characterized, and because the preparation is amenable to a variety of experimental techniques and methodologies.
The single most unusual property of the chick ciliary ganglion is the development of a large "calyx"-type nerve terminal that can be impaled by a microelectrode, providing a rare opportunity to observe presynaptic electrical activity associated with transmitter release. As a prelude to beginning an investigation of the nerve terminal electrical events associated with transmitter release, I began by characterizing the changes in synaptic efficacy that occur during and following repetitive stimulation in the chick ciliary ganglion. Intracellular recording from postsynaptic ciliary neurons was employed to verify that increases in synaptic efficacy result from an increase in transmitter
rees I.epr her that fou cmoet s of stiultion-inde








12
similar to the components of increased transmitter release described for other preparations.

Methods

Preparation and Solutions
Fertile White Leghorn chicken eggs (Poultry Science Unit, University of Florida) were set in a forced draft rotating incubator (Petersime model 1, Gettysberg, OH) kept at 37oC, 70% humidity and candled on days 4 to 10 to determine viability. Embryos were removed at embryonic day 15-19 (stage 41-45) and sacrificed via decapitation. These ages were chosen to coincide with maturation of the large "calyx" type synapse, before the synapse becomes primarily electrical in nature (Landmesser & Pilar, 1972). The ciliary ganglia were dissected out under intermittent washing with Tyrode solution (see below for composition). Several dissection techniques were used. The most common approach was to bisect the head and free 2 to 5 mm of the oculomotor nerve proximal to the orbit. A lateral approach was then used to draw the eye aside, liberate the ganglion from surrounding connective tissue, and dissect free the ciliary nerve (3-10 mm) from both sides of the sclera.

A recording chamber was constructed entirely of Sylgard polymer

(Dow Corning, Midland, MI) poured into a small (about 6 cm diameter) Petri dish. Two chambers of approximately equal volume (1.5 ml) were connected by a 1 cm long passage through which solutions passed during perfusion. Removing the bathing solution from a chamber physically
separ~ ~a frmter-riacabrmnmle os rmsraevb







13

the recording chamber using short lengths of very fine (0.1 mm diameter) tungsten-iridium alloy wire (AlfaAESAR, Ward Hill, MA).

The recording chamber was held in place by small bits of clay and surrounded by a plexiglass base to which perfusion apparati were attached. The preparation was continuously perfused with an oxygenated Tyrode solution (saline composition [in mM]: KC1 5; NaC1 150; CaC12 1 to 5; MgC12 2 to 12; glucose 10; HEPES 10; pH adjusted to 7.2-7.4) at a rate of 1-2 ml/min (gravity driven). Fluid levels were kept constant as saline was removed by suction through a bevelled hypodermic needle; the level of the needle was adjusted to keep the preparation just below the surface of the solution. In some experiments, Ba2+ (0.1-0.5 mM) or Sr2+ (0.5-4.5 mM) was substituted for Ca2+ or added to the saline solutions. In these experiments, the concentrations of Ca2+ and Ba2+ or Sr2+ were adjusted until the extracellularly-recorded postganglionic response was approximately equal to the response in Ca2+-only Tyrode. Salts for Tyrode solutions were purchased from Sigma Chemical (St. Louis, MO). Solution changes were carried out between trials by changing the source reservoir feeding the perfusion system. All experiments were carried out at room temperature (20-23C). Stimulating and Recording
Fluid suction electrodes (Dudel and Kuffler, 1961) mounted on
mechanical micromanipulators (Narishige, Japan) were used to draw up the preganglionic (oculomotor) and postganglionic (ciliary) nerves.
Thes elctrdes eremad rE0 polythleneL. tuin of .2 m








14
a syringe, which was used to draw the nerve and bathing solution into the tapered end. A silver wire (0.005 0.01 inch diameter) was inserted through the wall of the tubing and placed within approximately

5 mm of the tapered end of the tubing. A second silver wire was wrapped around each electrode shaft to within 5 mm of the tip to serve as a ground electrode.
Short stimulus pulses (0.01-0.06 msec) were applied to the oculomotor nerve through a photoelectric stimulus isolation unit (Grass Instruments, Quincy, MA) and the stimulus amplitude was adjusted until it was clearly suprathreshold. The postganglionic responses were amplified with a Grass P-5 series pre-amplifier and displayed on a Tektronix 5113 dual beam storage oscilloscope (Beaverton, OR). In most experiments the response consisted of both an electrical and a chemical component (see Figure 2-1), although in some ganglia from younger embryos there was not a distinct peak for the electrotonically mediated component. The amplitude of the chemically mediated component is a function of the number of postsynaptic cells brought to threshold by chemical neurotransmission (Martin and Pilar, 1963a). Thus, changes in the amplitude of the chemically mediated component of the ganglionic response represent changes in the number of postsynaptic cells activated by orthodromic stimulation (Landmesser and Pilar, 1972; Poage and Zengel, 1993) The main advantage of extracellular recording is the fact that the postganglionic response represents the averaged activity
of Lah e gangl1on. Aveage ga








15
When intracellular recording was used, several changes were made to minimize vibration and excess connective tissue that could foul intracellular electrodes. The connective tissue capsule that adheres closely to the ganglion was removed with fine forceps (DuMont #5, Fine Science Tools, Belmont, CA). All intracellular experiments were performed on a Kinetic Systems 9101-11 vibration isolation table (Roslindale, MA) with the perfusion apparatus and a dissecting microscope mounted on a freestanding Faraday cage. A micromanipulator with a motorized advance attachment (460XYZ micromanipulator, 860 series motorizer, Newport, RI) held an intracellular recording probe connected to the balanced bridge input of a Dagan 8500 intracellulular amplifier (Minneapolis, MN). The output of the intracellular amplifier was sent to 2 channels of a Tektronix oscilloscope for AC and DC recording. Most experiments were also recorded onto VCR tape through a PCM recording adapter (A.R. Vetter Company, Los Angeles, CA). The basic sampling rate was 88.2 kHz and the channel rise time was 50 psec with 14 bit A/D resolution.
Microelectrodes were pulled on a horizontal pipette puller

(Brown/Flaming P-87, Sutter Instrument Co., Los Angeles, CA) using glass capillary tubing (items # 1B100F-4, TWIOOF-4; 0.54 or 0.75 mm i.d., 1 mm o.d., World Precision Instruments, Sarasota, FL) and filled with 3 M KC1 (tip resistances, 25-100 megaOhms). The microelectrode was placed above the ganglion under visual control and the micromanipulator was used to advance the electrode into the ganglion proper.
Diec viulztoo niiul el a.otncsa Mire-







16

deflection of -45 to -80 mV. Intracellular recordings were usually of short duration, with impalements usually lasting less than 15 minutes.
Cells were identified by electrophysiological means as previously published (Dryer and Chiappinelli, 1985; Martin and Pilar, 1963a, 1964a,b). By injecting hyperpolarizing current through the recording electrode, it was possible to render both electrotonic and chemical potentials subthreshold (Martin & Pilar, 1963a; see Figure 2-4 inset) so that the underlying postsynaptic potentials could be observed. Membrane responses were monitored from the balanced bridge outputs of the intracellular amplifier. The bridge balance was adjusted before recording and was verified by testing the bridge balance after the microelectrode was removed from a cell. Data Collection and Analysis

For paired pulse and 5 impulse experiments, a Grass Instruments S48 stimulator was used to generate the conditioning and testing stimuli. Responses were averaged and their amplitudes measured using either a Nicolet 1170 signal averager (Nicolet Instruments Co., Madison, WI) or 386-based data acquisition and analysis software (Axotape, Axon Instruments, Foster City, Ca). For experiments in which longer conditioning trains were applied, a HINC-11 computer (Digital Equipment Corp., Marlboro, MA) was often used to generate the stimulation patterns, measure and store the postganglionic response amplitudes, and analyze the data (Magleby and Zengel, 1976; Zengel and Magleby, 1982). Sufficient time







17

Definition of Terms

Changes in response amplitude following conditioning simulation are expressed as:
V(t) = (Vt/Vo) 1 (2-1) where Vo is the control (pre-conditioning) response amplitude and Vt is the amplitude of the response at time t following the conditioning stimulation. For analysis of different components of stimulationinduced changes in V(t), I have used the approach described by Zengel and Magleby (1980, 1982) In brief, each component is defined as the fractional change in response amplitude in the absence of other components. Since it is not always possible to measure one component in the absence of others, the magnitudes and time constants of the individual components are derived from the value of V(t) by assuming that these components have distinct non-overlapping time constants of decay. The slowest decaying component can be estimated by using data points collected after the more rapidly decaying components have decayed away. Because the faster decaying components fall on top of the slower decaying ones, estimates of the contributions of these components can be made only by assuming some relationship between the different components and using standard linear decay analyses. In this study I have used a model shown to describe stimulation-induced changes in transmitter release at the frog neuromuscular junction (Magleby and Zengel, 1982; Zengel and Magleby, 1980, 1982). Basically, this model assumes
thathere arefou independent omnet o ireased---------- .. l w.i








18

A is augmentation and P is potentiation. As reported in this chapter, this model appeared to describe stimulation-induced changes in neurotransmitter release in the chick ciliary ganglion. Statistical Analysis
Control and experimental trials were recorded in each experiment. The effect of the experimental treatment was compared to the response of the same preparation under control conditions using t-test procedures. Statistical analysis was performed using the IBM PC version of SigmaPlot 5.0 (Jandel Scientific). Averaged data are presented as mean + standard error.
Results

Description of Stimulation-Induced Increases in Synaptic Efficacy

A paired pulse paradigm was employed to examine the effects of a single conditioning stimulus on the efficacy of ganglionic transmission. Figure 2-1 presents averaged extracellular data from a single preparation that was stimulated with pairs of impulses at an interstimulus interval of 50 msec. The chemically mediated component, which is a function of the number of cells brought to threshold by chemical transmitter release, is increased following a single conditioning impulse while the shock artifact and electrotonically mediated component of the response were unaffected. Under conditions of low levels of release many of the postsynaptic cells are below threshold for action potential generation and are not contributing to either peak of the postganaln- n-.Icraesi trnmte relea se lbisom








19








C



A 100 pV
C e 10 msec


*e*
. e*





it i
: s










Figure 2-1. Facilitation of compound action potential amplitude.
The oculomotor nerve was stimulated with a pair of pulses applied at an
interstimulus interval of 50 msec and the postganglionic response was recorded extracellularly (see MATERIALS AND METHODS). In this record, each response consisted of 3 upward deflections: a shock artifact (*)
and electrically mediated (e) and chemically mediated (c) components of
the postganglionic response. The trace represents the average of 8
consecutive trials. The temporal separation between the electrical and
chemical components of each response is due to the synaptic delay present for chemical neurotransmission. Note the large increase in the








20
In order to characterize the time course of the increase in synaptic efficacy following a single conditioning impulse, testing impulses were applied at intervals ranging from 25 msec to 5 sec. The presentation of intervals was randomized during an experiment. The symbols in Figure 2-2 plot V(t), the fractional increase in post-ganglionic response amplitude (Equation (2-1)), as a function of interstimulus interval. Each symbol represents data from a single Ca2+ and Mg2+ concentration. As illustrated in Figure 2-2, facilitation of the chemically mediated component was greatest when release was reduced by decreasing Ca2+ (A) or increasing Mg2+ (B). Under these conditions, the stimulation-induced increase in response amplitude was greatest at short conditioning-testing intervals, and decreased as the interstimulus interval was increased.
Results from another preparation are presented in Figure 2-3. The
filled circles in Figure 2-3A plot V(t) as a function of time following a single conditioning impulse. When these data are plotted on a semilogarithmic scale (filled circles, Figure 2-3B), it is obvious that the decay cannot be described by a single exponential. The data are well described by a dual exponential. The contribution of the slower of the two components was estimated by fitting a regression line to the linear portion of the curve (see figure legend). In this experiment the initial magnitude of this slower component, given by the intercept of the regression line at time 0, was 0.99 and its time constant of decay was







21

A 2.0
1.8
1.6 1.4
1.2
S1.0 m C
AmMCo
> 0.8 -A 1.5 mM ca
0.6 V 2mMCo
0.4 \0 3mM Co

0.2 a0.0
I I I I I0 200 400 600 800 1000 B 2.0 Time (msec)
1.8
1.6
1.4
1.2
1.0 12 mM Mg "' #l4 A a mu ug > 0.8 A 6mMMg B mM Mg
0.6 4mMMg 0.4 2 mM Mg
0.0
0.0- - -
I i I I I I
0 200 400 600 800 1000 Time (msec)
Figure 2-2. Effects of reducing extracellular Ca2+ or increasing
extracellular Mg2+ on facilitation of compound action potential amplitude. A: Plot of the decay of V(t), the fractional change in response amplitude (Equation 2-1), a a function of time following a single conditioning impulse. The [Mg +] was held constant at 2 mM while [Ca2+]























Figure 2-3. Effect of I and 5 conditioning impulses on synaptic
efficacy. A: Plot of the decay of V(t) as a function of time following a single conditioning impulse (filled circles) and a train of 5 conditioning impulses applied at 20/sec (open circles). Single testing impulses were applied at intervals of 25 to 5000 msec after the conditioning stimulation. Conditioning-testing trials were applied about once every 30 sec. Data points represent the average of 16 trials from a single preparation. [Ca'+] = 1 mM, [Mg2+] = 2 mM. B: Semilogarithmic plot of the decay of V(t following one ?filled circles) and 5 (open circles) conditioning impulses (same data as A). The lines represent the exponential decay of the second component of facilitation
(F2), derived by fitting regression lines through the data points between 200 and 2000 msec. C: Decay of the first component of facilitation (Fl) following one (filled circles) and five conditioning impulses (open circles). Values of F1 were obtained by subtracting off from V(t) the contribution of the second component of facilitation (Equation 2-2 in Methods). The lines represent the exponential decay of Fj, derived by fitting a regression line through the data points.








23

A 12

O
0 I I
05 impulses
1 impulse

I
4-0
B

2
0%
2 e
0'"---- ----------_ -4
I .... I .... I .... I .... I ....
0 1000 2000 3000 4000 5000
Time (msec)

B C
-t/55msec
1 0 O F1 =1 1.9e
O -t/55m..e
F2= 4.1 e

4) -t/80~mse F1=5.8 e
O
1 U"





0. 1 F2=O.99e-t/s55msea

I I I I 1 I I I







24
circles in Figure 2-3C plot the estimated decay of the more rapidly decaying component, which in this experiment had an initial magnitude of 5.8 and a time constant of decay of 60 msec.

In 24 paired pulse experiments of this type, the average time constants describing the decay of V(t) were about 60 msec (n=14) and 400 msec (n=24; see Table 2-1). (In some experiments done under higher quantal conditions (>1.5 mM Ca2+), there was an apparent depression of ganglionic transmission at short interstimulus intervals, precluding precise measures of the more rapidly decaying component.) These two time constants of decay are very similar to the time constants previously reported for the first and second components of facilitation at the frog neuromuscular junction and the rabbit sympathetic ganglion (Table 2-1).
Intracellular experiments were performed under the above conditions to verify that facilitation of the ganglionic response results from an increase in EPSP amplitude. As reported using extracellular recording, paired pulse stimulation led to a facilitation of EPSP amplitude. Figure 2-4 shows results from several ciliary neurons in two different Ca2+ concentration ranges. The continuous lines describing the decay of facilitation were drawn by taking the average values for the magnitude and time constant of the two components of facilitation (using results from extracellular data) and plotting their combined decay (see figure legend) It appears that the estimates of facilitation provided







25









Table 1. Time Constants of Decay of Components of Increased Transmitter Release
Rabbit Rat Chick Frog NMJ1 Sympathetic Ganglion2 Hippocampus3 Ciliary Ganglion

Facilitation: present
F1 60+3 msec 59+14 msec 63+4 msec
F2 475+58 mseca 388+97 mseca 415+35 mseca
Augmentation 7.3+1.3 secb 7.2+1.0 secb 4.7 sec 30.7+2.3 secb
Potentiation 65+18 sec 88+25 sec 87 sec 205+24 sec


1Zengel & Magleby (1982); Magleby & Zengel (1976) 2Zengel et al. (1980)
3McNaughton (1982)
amagnitude and time constant increased in the presence of Sr2+ bmagnitude increased in the presence of BaC+
























Figure 2-4. Facilitation of EPSP amplitude in ciliary neurons.
Inset: Example of averaged data showig paired pulse facilitation of EPSP amplitude. [Ca'] = 1.4 mM, [Mg +] i 2 mM. (A) and (B) plot V(t) as a function of time following a single conditioning stimulus. The symbols connected by the dashed lines represent data from the same cell. Data points represent the average of 5 to 15 identical trials. The continuous lines describe the decay of the first and second components of facilitation (drawn using values for the magnitude and time cogstants of F1 and F2 obtained using extracellular recordings in 4 mM Mg'+ and the in d4riated [Ca+]).
A: Fl = 1.83e- /8 msec, F2 = O.58e-t/443 msec B: F1 = 0.56e-t/72 msec, F2 = 0.40e-t/366 msec









27
20 msec
2.0 l V
. 10 mV
1*
A
1.5


III
II
,-.1.0

> C ++
** 1.4 1.5 mM Ca

OII
0.5 O----- ...........................


0.0


0 100 200 300 400 500
Time (msec) B 2.0
4.+
2.5 mM Ca
1.5


.-- 1.0- 0
10

0.5



0.0 -----------------------------------







28

decayed with a time constant on the order of 65 msec. It was also reported that the observed increase in EPSP amplitude could be accounted for entirely by an increase in quantal content. The ionic conditions and stimulation paradigms used in this earlier study are nearly identical to the conditions of my own experiments. Therefore, it seems reasonable to assume that the increases in synaptic efficacy described here using paired pulse stimulation result from an increase in quantal transmitter release.
To observe more slowly decaying increases in ganglionic response amplitude, ganglia were conditioned using longer trains of stimuli (200-1200 impulses at 10-50/sec). The filled circles in Figure 2-5A plot the decay of V(t) of the chemically mediated component of the action potential following a conditioning train of 800 impulses applied at 20/sec. When the data are plotted on a semilogarithmic scale against time following the end of the conditioning stimulation (filled circles in Figure 2-5B), there appear to be two components of decay. The contribution of the slower component was estimated by fitting a regression line to the linear portion of the curve at times beyond 100 sec (see figure legend). In this experiment, the initial magnitude of the slower component was 0.32 and its time constant of decay was 200 sec. Figure 2-5C plots the decay of the faster component, obtained by correcting for the contribution of the slower component (see figure legend). This component had an initial magnitude of 1.02 and a time
rnnctant nf dayo oF 19 car Tn evnrimants in which 4+ wse nnc4ila























Figure 2-5. Effect of trains of repetitive stimulation on synaptic efficacy. A: Decay of V (t) as a function of time following a conditioning train of 800 impulses applied at 20/sec. Testing impulses were applied at 2 sec intervals for 3 impulses, then every 10 sec. The filled circles plot the fractional change in amplitude of the chemically mediated peak of the compound action potential. The open circles plot changes in the amplitude of the electrically mediated portion of the compound action potential. Data averaged from 4 identical trials from a single preparation. [Ca2+] = 1.5 mM, [Mg+] 2 mM. Note that the electrotonically mediated potential (open circles) is not increased following repetitive stimulation. B: Semilogarithmic plot of the decay of V(t) (same data as in A). The line represents the exponential decay of potentiation, obtained by fitting a regression line through the data points beyond 100 sec. C: Decay of augmentation, obtained after correcting for the contribution of potentiation using Equation 2-2. The line represents the exponential decay of augmentation, derived by fitting a regression line through the data points.








30

A 1.5
*

1.0

0
0.5

00.



-0.5
I II I I I I
0 100 200 300 400 500 600
Time (sec)

B 1
-t/12 sc
32-t/200 sec A= 1.02e
9 P=.32e
1P



S0.1 .

e. ,r ***
0.01


0.001 -I
I I I I I I I I I
0 100 200 300 400 500 0 100 200
Time (aec) Time (sec)







31

increased, and was occasionally depressed during and immediately following tetanic stimulation (open circles in Figure 2-5A).

In 27 experiments of this type, the average values for the time
constants of decay of these two more slowly decaying phases were about 30 sec (n=17) and 200 sec (n=27). These time constants are similar to those attributed to the processes of augmentation and potentiation, respectively (Table 2-1).
In order to verify that the observed changes in postganglionic

response reflect a change in EPSP amplitude, intracellular recording from ciliary neurons was used. Figure 2-6 shows the effect of 800 conditioning stimuli applied at 20/sec on the amplitude of the EPSP. The amplitude of the control EPSP (before the conditioning stimulation) was
5 mV. Test pulses were applied at 10 second intervals following the end of the train. Each test pulse applied less than 60 seconds after the train produced a large EPSP that initiated an action potential (not shown). EPSP amplitude declined over the next 6 minutes until the response reached preconditioning levels.
In three experiments of this type, EPSP amplitude increased to a
maximum of 150% to 400% of control values during the test period. The EPSP amplitude returned to a pre-conditioning level with a time constant of about 3 to 4 minutes, similar to that observed by Martin and Pilar (1964c) under similar conditions. This time course of the decay of EPSP amplitude following repetitive stimulation is also very similar
to theL ecaya o the a extanaarr recrdedl pst ngl io ic 4 nsos








32



CONTROL


ii

60 sec




90 sec


I
Swkk
120 sec



I

5 mV 250 sec

20 msec






Figure 2-6. Effect of an 800 impulse train (20/sec) on EPSP amplitude. Intracellular recording of EPSPs from a ciliary neuron shows the response to orthodromic stimulation pr or to and at the times indicated following the conditioning train. [CaZ+] = 5 raM, [Mg +] = 4 mM. The time course of potentiation in this experiment is very similar to







33

not the unit size, of spontaneous miniature EPSPs. This suggests that the potentiation of the EPSP reported here is indeed due to an increase in transmitter release, and not an effect on postsynaptic receptor sensitivity (Martin and Pilar, 1964c).
The augmentation phase of increased release was difficult to
describe using intracellular recording. Two possible reasons for this are: 1) the variability of EPSP amplitude under conditions of very low quantal content may make observing the decay of augmentation, which is described using only the first 5 to 10 test points (0 to 60 seconds after conditioning), dependent on averaging large numbers of identical trials, and 2) at higher quantal content, an apparent depression of release immediately following tetanic stimulation may confound attempts to observe effects of stimulation on increased EPSP amplitude during the time when augmentation would be observed. Effect of Number of Conditioning Impulses on the Components of
Increased Synaptic Efficacy
If the processes I have described arise from the same mechanisms
that produce stimulation-induced increases in release in other synaptic preparations, then the growth and decay of the stimulation-induced changes in synaptic efficacy in the chick ciliary ganglion should be described by the "accumulation" models which have been shown to describe release in other preparations (e.g. Magleby and Zengel, 1975a,b, 1982; Mallart and Martin, 1967; Younkin, 1974). According to these models, each conditioning impulse adds an incremental increase to








34
To test this, experiments were conducted to examine the effects of
conditioning impulse number on the two components of facilitation, augmentation, and potentiation. The open circles in Figure 2-3A plot the decay of V(t) following a 5 impulse conditioning train. At all time points tested, the increase in V(t) was much greater following the 5 impulse train (open circles) than following a single conditioning impulse (filled circles). This increase in V(t) could be attributed to an increase in the magnitudes of the second (lines through open circles in Figure 2-3B) and first components of facilitation (lines through open circles in Figure 2-3C). There was little or no effect of conditioning impulse number on the time constants of decay of the two components of facilitation. Similar results were observed in 2 additional experiments in which 1, 2 and 5 impulse conditioning trains were applied. Thus, as with the rabbit sympathetic ganglion (Zengel et al., 1980) and frog neuromuscular junction (Zengel and Magleby, 1982), increasing the number of conditioning impulses results in an increase in the magnitudes of the two components of facilitation, while having little effect on their time constants of decay.

I also examined the effect of conditioning impulse number on the magnitudes of the slower decaying components. Figure 2-7 summarizes the results of 6 experiments in which conditioning trains of various duration were applied. In each of these experiments I observed an increase in the magnitude of potentiation when the number of condition-







35













A B
4.0 4.0
3.5 3.5

3.0 3.0
o .9
2.5 2.5 o 0
2.0 c 2.0
1.5 E
S 1.5 ,* 1.5a Sr=
A -e e
0.5 0.5 -.-**
0.0 0.0
0 400 800 1200 0 400 800 1200

Number of impulses Number of impulses








Figure 2-7. Effect of the number of conditioning impulses on the
magnitudes of potentiation (A) and augmentation (B). The magnitudes of augmentation and potentiation were obtained as described for Figure 2-5. Results of 6 experiments in which the decav of V(t) was recorded








36

trains, I observed an apparent depression of ganglionic transmission immediately following the conditioning trains that made it difficult to obtain reliable estimates of augmentation. There was no consistent effect of increasing stimulus duration on the time constants of decay of these processes. These results are similar to those observed at other synapses (e.g. Magleby and Zengel, 1976; Zengel et al., 1980). Pharmacological Characterization: Effects of Strontium and Barium
At the frog neuromuscular junction and the rabbit sympathetic ganglion, the addition of certain divalent cations to the bathing solution selectively affects individual components of stimulation-induced increases in release. Barium increases the magnitude of the augmentation phase and strontium increases the magnitude and time constant of the second component of facilitation (Zengel and Magleby, 1977, 1980, 1981; Zengel et al., 1980). To further test the hypothesis that the phenomena I describe here are analogous to the four components of stimulation-induced increases in release reported in other preparations, I repeated the experiments described in Figures 2-3 and 2-5 in the presence of these divalent cations.

Figure 2-8 illustrates the effect of Sr2+ on facilitation. In the

presence of Sr2+ (open circles), V(t) was unchanged or slightly reduced at short interstimulus intervals (less than 100 msec), but there was an obvious enhancement of ganglionic efficacy at intervals of 300 to 2000 msec (A). This increase in V(t) could be attributed to an
Snrese- bot th-antd an tiecntn ofdcyotescn



























Figure 2-8. Effect of Sr2+ on facilitation. A: Plot of the decay of V(t) as a function of time following a single conditioning impulse in 1.5 nmM Ca2+ Tyrgde (filled circles) and in Tyrode containing 1.0 mM Ca2+ and 1.5 mM Sr'+ (open circles). Data points represent the average of 32 trials from a single preparation. B: Semilogarithmic plot of the decay of V(t) (same data as in A). The lines, obtained by fitting regression lines through the data points between 300 and 2000 msec, represent the exponential decay of the second component of facilitation
(F2). C: Decay of the first component of facilitation (FI), obtained after correcting for F2 as described earlier.








38


A 2.0


1.5

0
--- C2+
1.0 1.5 mM Ca2
2+ 2
O 0 1 mM Ca +/ 1.5 mM Sr2+

0.5 o
0
e O

0
0 0

0.0
I... I.... I.... I.. I
0 500 1000 1500 2000
Time (msec)

B C ,0.-t/lOm@ F1=0.93e
1 -1
1" -t/1240msec -t/1 n
F2s 47 F1 Sr=o0.46et/144m




0.1 0.1



F2 =.27 5e-t/assc
0.01 0.01
I .. .... I I L.....I
0 500 1000 1500 2000 0 500
Time (msec) Time (msec)








39
The effect of Sr2+ on the second component of facilitation, which could be reversed by washing the preparation with control Tyrode, was seen in each of 6 experiments of this type. In the presence of Sr2+ the magnitude of F2 increased from 0.50 + 0.09 to 0.92 + 0.21 (paired t-test, P<0.05), whereas the time constant of decay of F2 increased from 503 + 77 msec to 930 + 133 msec (paired t-test P<0.01 ). These effects of Sr2+ are strikingly similar to the effects of Sr2+ at the frog neuromuscular junction (compare Figure 2-8 to Figure 8 in Zengel and Magleby, 1980).

Figure 2-9 shows the effect on augmentation and potentiation of
addition of small amounts of Ba2+. Notice the large increase in V(t) during the first 90 sec of decay, the time during which augmentation is decaying, in the presence of Ba2+ (A). This effect of Ba2+ on the augmentation phase of decay is more clearly seen in Figure 2-9C, which plots the decay of augmentation in the absence (filled triangles) and presence of Ba2+ (open triangles). In contrast, Ba2+ had little or no effect on potentiation (B).
Similar results were obtained in each of 6 experiments of this

type. In these experiments the magnitude of augmentation was significantly increased in the presence of 0.1 to 0.15 mM Ba2+ (1.10 + 0.26 vs. 2.87 + 0.89; paired t-test, P


























Figure 2-9. Effect of Ba2+ on potentiation and augmentation. A:
Plot of the decay of V(t) following 800-impulse conditioning trains in the absence (filled circles) and presence of Ba2+ (open circles). Control, conditioning and testing impulses as in Figure 2-5. Data averaged from 8 trials from a single preparation. B: Semilogarithmic plot of the decay of V(t) (same data as in A). The lines, obtained by fitting regression lines through the data points beyond 100 sec, represent the exponential decay of potentiation. C: Decay of augmentation, obtained after correcting for potentiation as described earlier.







41


A 8 -0 7 .0
6- lmMCa2+
5 -Oa 0.75 mM Ca2+/0.15 mM Ba2+
O
4O

2
1
0
o


I I 1 I I I
0 100 200 300 400 500
Time (sec)

B o10 C A,,=7.22e-t/
0
V p~o.34e-t/1gOaec
10
-t/222O*
~Po=0. 27e .

> 0.1

o.~ \r
o
eq
0.01 o
00
-t/24see A= 2.29 e
0.001
0 1 00 1 30 .010_..1 00_ I0
0 100 200 300 400 500 0 100







42

Discussion
The aim of the experiments presented here was to fully characterize the kinetic properties of stimulation-induced changes in synaptic efficacy in the embryonic chick ciliary ganglion, and to investigate the sensitivity of these processes to the divalent cations Sr2+ and Ba2+. The results indicate that there are four components of stimulationinduced i ncreases in ganglionic efficacy described by time constants of about 60 msec, 400 msec, 30 sec and 200 sec (Table 2-1). In several synaptic preparations, accumulation models describing stimulus-induced increases in transmitter release have been successful in describing increases in transmitter release (e.g. Magleby and Zengel, 1975b; Mallart and Martin, 1967; Younkin, 1974). One basic attribute of these models is that each conditioning impulse increments the mechanisms underlying each of the components of increased release. The results presented in Figures 2-3 and 2-7 show that the four components observed in the present study accumulate as predicted by models describing facilitation, augmentation and potentiation (e.g. Magleby and Zengel, 1982; Zengel and Magleby, 1982). Further identification of the second component of facilitation and of augmentation was achieved by exploiting the pharmacological sensitivities of these processes to certain divalent cations. The addition of Ba2+ to the bathing solution caused an increase in the magnitude of augmentation, and partial or complete substitution of Sr2+ for Ca2+ in the bathing solution resulted in an
*nre s *n th 36aCn*tude andtetm cosat of th seon coon








43
first component of facilitation, the second component of facilitation, augmentation and potentiation as described at the frog neuromuscular junction and at other synapses.
In preparations where stimulation-induced increases in synaptic
efficacy have been studied extensively, most notably the frog neuromuscular junction, the increases have been shown to result from an increase in quantal release (del Castillo and Katz, 1954; Magleby and Zengel, 1976). In one of the first electrophysiological studies using the chick ciliary ganglion preparation, Martin and Pilar (1964c) showed that paired pulse facilitation of EPSP amplitude occurs in the embryonic chick ciliary ganglion, and that it is a result of increased quantal content which decays back to control levels with a time constant of about 65 msec. Data reported here confirm the presence of facilitatory processes and further describe two individual components of facilitation with distinct kinetic and pharmacological properties. Martin and Pilar (1964c) also reported, using intracellular recording from ciliary neurons, the presence of a more slowly decaying potentiation of EPSP amplitude, although its time course was not described in detail. The experiments reported here describe the decay of ganglionic efficacy after prolonged repetitive stimulation. A more rapidly decaying component is akin to augmentation, having very similar kinetic and pharmacological properties. The more slowly decaying component is termed potentiation.
It as now ben emnsrae tat thr4r.orcopnns n








44

question remains: What subcellular mechanisms are conserved at the synapse that produce these processes at different synapses i n a variety of species?
Since quantal content is known to be affected by manipulations of
Ca2+ buffering and entry of Ca2+ into the nerve terminal, speculations on the mechanism of stimulation-induced increases in neurotransmitter release focus on the role of intracellular Ca2+ in transmitter release (e.g. Charlton et al., 1982; Katz and Miledi, 1967, 1968; Zengel et al., 1993a,b). No single theory has been successful in accounting for all of the observed stimulation-induced changes in synaptic transmitter release. Proposed mechanisms include an increased entry of Ca2+ or an accumulation of Ca2+ in the presynaptic nerve terminal (Erulkar and Rahamimoff, 1978; Katz and Miledi, 1968; Miledi and Thies, 1971; Rosenthal, 1969; Weinreich, 1971) and nerve terminal voltage changes or processes associated with these voltage changes (e.g. Martin and Pilar, 1964c; Bittner and Baxter, 1991). The ciliary ganglion offers a versatile system in which to study these possibilities through the use of many techniques, including presynaptic intracellular recording (Dryer and Chiappinelli, 1983; Martin and Pilar, 1964c; Yawo, 1990), patch clamp recording of Ca2+ currents (Stanley, 1989), and Ca2+-imaging using fluorescent dyes. Chapter 4 will describe the effects of repetitive stimulation on presynaptic potentials under conditions conducive to facilitation of transmitter release.













CHAPTER 3
CHARACTERIZATION OF CALCIUM CHANNELS INVOLVED IN SYNAPTIC TRANSMISSION

The relationship between Ca2+ channel subtypes and transmitter release is not well understood. In most synapses evoked release of transmitter is dependent upon a coordinated influx of Ca2+ ions that elevates intracellular Ca2+ at some site in the presynaptic nerve terminal (Katz and Miledi, 1967). Voltage-activated Ca2+ conductances are most often responsible for this rapid increase in intracellular Ca2+. For this reason and because Ca2+ channels are potential targets for modulating the release process (reviewed in Scott et al., 1991), it is of interest to determine which channel type(s) are acting to initiate release in the chick ciliary ganglion.
There are clear indications for the existence of at least 4 general classifications of neuronal Ca2+ channels. Subtypes N, L and T have been described in chick dorsal root ganglion cells (Fox et al., 1987a; Nowycky et al., 1985). These channels have been identified by their sensitivities to different classes of pharmacological agents and by their kinetics of activation and inactivation. L-type channels are characterized by large unitary conductances (about 25 pS), activation voltages positive to -10 mV and sensitivity to Ca2+ channel blockers







46

mV) and are rapidly inactivated at negative holding potentials. N-type channels show an intermediate voltage activation (-40 to -30 mV), intermediate unitary conductances (about 13 pS) and sensitivity to w-conotoxin GVIA. These calcium channels may also be characterized by their sensitivity to inorganic cations. Cadmium (Cd2+) is by far the most potent, blocking L- and N-type channels with a Kd of 10 pM, but about ten times more is needed to block T-type channels (Fox et al, 1987a,b). The more recently described P-type channel (named to honor the Purkinje cell in which it was first described) shows little inactivation, has a voltage-dependence between those of N- and L-type channels and is not blocked by w-conotoxin or by dihydropyridines, but is blocked by a component of funnel-web spider venom, FTX or w-agatoxin IVA (Llinas et al., 1989).
It has been shown that N-, L- and P-type channels can all contribute to rapid Ca2+ influx associated with evoked transmitter release (see Scott et al., 1991). It has also been demonstrated that multiple types of Ca2+ channels can coexist in a single neuron (reviewed in Miller, 1987), and that more than one channel type can be involved in initiation of exocytosis from a single cell type (Artalejo et al., 1994; Regehr and Mintz, 1994; Takahashi and Momiyama, 1993).
In the presynaptic calyciform nerve terminal of the chick ciliary

ganglion, several investigators have reported the presence of "N-like" currents in the calical nerve terminal, yet no evidence for L- or
T-,na rmnrret (tanle, 191, Snl C+&rwls 9 AtraLrh4, 10; Stnley snA







47
Although Ca2+ channel classification schemes are useful in terms of defining a point of reference (for comparison and discussion), there appears to be such diversity in Ca2+ channel structure and function that overlap between these subtypes (and ensuing subclassification) is rendering these simple classifications insufficiently descriptive (as discussed in recent reports: Bertolino and Llinas, 1992; Scott et al., 1991; Stanley and Goping, 1991). It is, therefore, important to complement pharmacological classification studies with evaluations of the functional properties of presynaptic Ca2+ channels. Experiments in this chapter will address the role of presynaptic Ca2+ channels in evoked transmitter release in the chick ciliary ganglion. To test the contributions of different Ca2+ channel types to evoked transmitter release, the pharmacological sensitivity of the release process to Ca2+ channel blocking agents is investigated.
Methods
Extracellular recording techniques were employed as described in
Chapter 2 (Methods). w-conotoxin (Bachem, Torrance, CA) was dissolved in deionized H20 (stock concentration, 500 uM) and frozen in 30 pl aliquots (-20 C). Divalent cations were obtained as salts (Sigma, St. Louis). All drugs were dissolved in Tyrode solution before being applied. Due to high cost and limited availability, toxins were added directly to small amounts of Tyrode and oxygenation was maintained by bubbling 02 directly into the bath. Under control conditions, this
me o aoxygnat mn hL am t a nc tr a -ms .n.a








48

ganglionic response has been shown to reflect changes in synaptic efficacy in this preparation (Landmesser and Pilar, 1972; Marwitt, Pilar and Weakly, 1971; Poage and Zengel, 1993; Stanley and Goping, 1991). Low frequency stimulation (0.1/sec or slower) was used to obtain control values of postganglionic response amplitude and to test the effects of drugs and divalent cations. Data are also presented from experiments using more complex repetitive stimulation paradigms (see Chapter 2) In these cases, sufficient time was allowed between trials to allow the response to recover to pre-stimulus levels.

Results
The effects of Cd2+, Co2+, Ni2+ and Mg2+ on synaptic transmission through the ciliary ganglion are shown in Figure 3-1. In separate experiments, these ions were added to the bathing solution and their effects on ganglion ic transmission were observed. The addition of these divalent cations led to a decrease in the amplitude of the chemical component of the postganglionic compound action potential. Cd2+ was more than 2 orders of magnitude more potent in decreasing ganglionic transmission than Co2+, Ni2+ and Mg2+ (Figure 3-1). This order of potency of these divalent cations and the concentrations used to impede ganglionic transmission are similar to those reported to block Ca2+ currents in mouse neuromuscular junction (Penner and Dreyer, 1986), rat brain synaptosomes (Lentzner et al., 1992) and squid giant synapse (L1i nas et al., 1981) and synaptic transmission at the frog








49





2+ .. 2+ 90I

70
e 80;ri70

E 60
CC
< 50


30
U

M 20
10

0

0.001 0.01 0.1 1 10 [divalent cation] mM




Figure 3-1. Concentration-dependence of the effects of divalent
cations on postganglionic response amplitude. Values are expressed as percent control compound action potential amplitude, ith the exception
of the Mg data, which is compared to the lowest [Mg'+] applied (2 mM). Points with s standard error bjrs indicate data averaged from 2 to
4 experiments. [Ca Z] = 5 mM, [Mg 2] = 2 mM unless otherwise noted.








50
w-conotoxin (Figure 3-2, circles) led to an irreversible decrease in the amplitude of the chemically mediated portion of the compound action potential. Higher concentrations of w-conotoxin acted more rapidly (2 pM, triangles in Figure 3-2). Concentrations of 1 pM or greater usually led to a complete block of the chemical component of the postganglionic response within 90 minutes (4 of 5 experiments).
Application of 5-10 pM verapamil, a phenylalkylamine Ca2+ channel
blocker, caused a very small decrease in the amplitude of the compound action potential (4% to 7%) that was reversed by washing with control Tyrode (n = 2 experiments). While it is possible that this effect is due to a direct effect of verapamil on Ca2+ currents, it has been reported that higher concentrations of verapamil (20 to 50 pM) can block voltage-activated Na+ currents (Chang et al., 1988). Such an effect would be expected to affect the amplitude of the compound action potential.
In 2 experiments, 10 pM nifedipine, a dihydropyridine Ca2+ channel blocker, had no effect on transmission through the ciliary ganglion.

Discussion
The results presented in this chapter are consistent with other
studies (e.g. Bennett and Ho, 1991; Stanley, 1989; Yawo and Momiyama, 1993) in which it was suggested that the "N-like" voltage-activated Ca2+ currents described in the dissociated calyx preparation are responsible for initiation of transmitter release in the chick ciliary







51



110
100 -------------------------------------U go0
80



60
C
(250A
O
4' 40

YIO 2 O \ .
20 A
10
0
0 -,, , ,,;
0 10 20 30 40 50 60 70 Time (min)


Figure 3-2. Time course of the effects of 2 concentrations of
w-conotoxin on ganglionic transmission. Either 1 pM (circles) or 2 pM w-conotoxin (triangles) was added to the Tyrode bathing solution at time = 0. Each point represents the averaged amplitude of 4 to 16 consecutive responses and is normalized to the amplitude of the response before toxin addition. The block of ganglionic transmission was not reversed by 70 to 1 0 minutes of perfusion with toxin-free Tyrode. [CaZ+] = 5 mM, [Mg +] = 2 mM.







52


have been shown to completely block presynaptic Ca2+ currents in the calyx (Stanley and Goping, 1991; Yawo and Momiyama, 1993). Application of L-type Ca2+ channel blockers (verapamil and nifedipine) had little or no effect on ganglionic transmission. These results suggest that Ca2+ channels involved in release in this preparation are most similar to N-type channels. Furthermore, it is reported here that the divalent cations Cd2+, Co2+ and Ni2+ blocked ganglionic transmission with the same order of potency as described in other synaptic preparations in which N-type channels are involved in initiation of transmitter release (Lentzner et al., 1992; L1inas et al., 1981; Penner and Dreyer, 1986).
The majority of the total presynaptic Ca2+ current in the calyciform nerve terminal is w-conotoxin sensitive (Stanley and Atrakchi, 1991; Yawo and Momiyama, 1993). An w-conotoxin insensitive component to the calyx ICa has been reported (Stanley and Atrakchi, 1990; Yawo and Momiyama, 1993). This component can support low levels of transmitter release when 4-aminopyridine is added to increase the presynaptic action potential duration (Yawo and Chuhma, 1994). This w-conotoxin insensitive component is blocked by 50 uM Cd2+. The role of such a small Ca2+ current in release could not be assayed with the extracellular recording method used in this study. Under the conditions used here, little if any transmitter release is supported by this w-conotoxin insensitive current. The chemically mediated portion of the com- -, - -n nan -n A..n nIK PP mltd 1..lara







53

It is generally accepted that transmitter release from vertebrate peripheral synapses is initiated primarily by current flow through N-type Ca2+ channels (reviewed in Miller, 1987). N-type channels, but not channels of other types, have been shown to be physically connected to proteins in the presynaptic terminal that act to regulate docking of synaptic vesicles and priming of vesicles for release (reviewed in Bennett and Scheller, 1993). If these active zone proteins associate exclusively with N-type Ca2+ channels, as has been suggested (Mastrogiacomo et al., 1994), the limited role of other channel types to exocytosis may be related to their distance from the release machinery. In support of this, it has been reported (Stanley, 1993), that the coupling of Ca2+ influx to acetylcholine release in the ciliary ganglion appears to be conserved within a 3 pm2 membrane patch. Stanley further suggested that Ca2+ influx through a single channel is sufficient to trigger quantal transmitter release, implying that Ca2+ influx through N-type channels occurs extremely close to the transmitter release mechanism in the chick ciliary ganglion.
In summary, I have characterized the effects of Ca2+ channel blocking agents (divalent cations, dihydropyridine and phenylalkylamine antagonists, and w-conotoxin GVIA) on ganglionic efficacy in the embryonic chick ciliary ganglion. The results presented here further strengthen the argument that the w-conotoxin sensitive Ca2+ currents previously described in the presynaptic nerve terminal represent the
m r u o 2+lr ita o tmtt release












CHAPTER 4
EFFECT OF REPETITIVE STIMULATION ON THE PRESYNAPTIC ACTION POTENTIAL

When considering possible mechanisms for stimulation-induced
increases in release, there are several points in the excitationsecretion cascade that appear to be obvious candidates. One of these is the presynaptic action potential. If the depolarization of the nerve terminal caused by the invasion of an action potential is larger in amplitude or is prolonged, this could result in an increased recruitment of voltage-dependent Ca2+ channels and a greater influx of Ca2+, leading to increased release.
Due to their small size, presynaptic elements of most synapses are difficult to study in an intact synapse. However, there are certain synaptic preparations that have extraordinarily large presynaptic terminals. In several of these preparations the relationship between action potential parameters and transmitter release has been investigated. In the squid giant synapse, increasing the duration of the presynaptic action potential by pharmacological means leads to an increase in the amplitude of the postsynaptic response (Augustine, 1990o). In sensory neurons of Aplysia, presynaptic facilitation of release appears to be mediated by a decrease in a specific K+ current that prolongs the action potential, leading to an increase in Ca2+ influx (Klein et al.,
192nS-tefll 92. hs reslt su et tat cane nth







55
Martin and Pilar (1964a-c) employed intracellular recording from

the pre- and postsynaptic cells of the chick ciliary ganglion to investigate mechanisms of stimulation-induced increases in release. They reported no significant change i n the presynaptic nerve terminal action potential under conditions where increased release was observed. However, those studies compared the amplitudes of individual action potentials, and it is possible that subtle changes in the action potential amplitude and/or duration would not have been detected by this type of analysis. It was therefore of interest to perform experiments of a more quantitative nature to investigate the relationship between changes in the presynaptic action potential and the individual processes of stimulation-induced increases in release described an Chapter 2.
Methods
Standard intracellular recording techniques were employed

(described in detail in Chapter 2, Methods). Cells were identified as calyciform nerve endings or ciliary neurons using the electrophysiological criteria of Martin and Pilar (1963a, 1964a). Briefly, stimulation of the oculomotor nerve produced an action potential in both calyces and ciliary neurons. When spike initiation was blocked by injection of hyperpolarizing current, the ciliary neurons exhibited an underlying excitatory postsynaptic potential (EPSP) (Figure 4-1, upper record). Nerve terminals showed no EPSP following orthodromic stimulation when hyperpolarizing current was injected. The only remaining response was an attenuated action potential, which was presumably conducted electro-







56

coupling potential, representing current flow through electrical synapses (Martin and Pilar, 1963a).
In paired pulse experiments, trials consisted of a conditioning
stimulus and a testing stimulus applied some time after the conditioning stimulus. The conditioning stimulus serves as a control to which test responses are compared. Typical responses to paired stimuli from a ciliary neuron and calyx are presented in Figure 4-1. In experiments in which trains of impulses were applied, changes in the response are described relative to the first response of the train.

The parameters were defined as follows: action potential amplitude was measured from resting membrane potential to the most depolarized point of the action potential; time to action potential peak was measured from the time of the last available baseline control point (the last point at resting membrane potential before the stimulus artifact) to the time of the action potential peak; action potential duration was measured from the time of the last baseline point to the point where the repolarizing action potential crossed the original resting potential; afterhyperpolarization (AHP) amplitude was measured from the resting membrane potential (prior to the action potential) to the most hyperpolarized point during the AHP; AHP half-decay time was measured by finding the peak hyperpolarization and determining the time for the voltage to decay to half of that level.
Results
Effects of ReDetitive Stimulation on Presynaptic Potentials






57



Post-synaptic ciliary neuron





*




. 20 mV

=25 msec







Pre-synaptic nerve terminal





-'" | 10 mV
I C













e
*. .











*



.: --10 msec












Figure 4-1. Examples of electrophysiological responses of pre- and postsynaptic cells of the embryonic ciliary ganglion to paired pulse stimulation. A postsynaptic ciliary neuron is ident ified by the presence of an excitatory postsynaptic potential (EPSP) in response to orh ndrmic stilluan (linper reor. Hynprpollz 1c urr eilnF tUas .







58

second of a pair of EPSPs recorded from a postsynaptic ciliary ganglion neuron (Figure 4-1, upper record; see Chapter 2 for description of facilitation). Repetitive stimulation also caused an increase in presynaptic action potential duration (control = 3.85 msec, test = 4.05 msec) and a decrease in AHP amplitude (control = 9 mV, test = 6.5 mV; Figure 4-1, lower record). These changes are illustrated more clearly in Figure 4-2 where the control and test responses from Figure 4-1 are superimposed.

During short trains of stimulation (10 impulses applied at 20/sec), changes in the presynaptic action potential are apparent early in the train and appear to reach a steady state (Figure 4-3). These changes are more clearly seen in Figure 4-4, which plots averaged data from the same cell as Figure 4-3. Repetitive stimulation caused an increase in action potential duration (A) and decreases in both AHP amplitude (B) and AHP half decay time (C). There were no consistent effects of repetitive stimulation on action potential amplitude or time to peak (data not shown).
Ionic Mechanisms Underlying Action Potential Repolarization and AHP
It is generally accepted that the repolarizing and afterhyperpolarizing phases of the action potential are dependent on the activation of K+ channels and a subsequent K+ efflux. To test the ionic basis of the afterhyperpolarization, the membrane potential was altered by injecting current through the recording electrode and the resulting changes in AUP amnlitude were nbhcprved Finiure 4-5 shows results from a single







59









4 msec


i' 4
:. lOmV
3 *




test
.*1

a2





control






Figure 4-2. Nerve terminal action potentials evoked by paired pulse stimulation. Responses are superimposed to illustrate changes in repolarization phase. Same cell and interstimulus interval as Figure 4-1. Single responses are shown (not averaged).














60
























100 msec





1 4 4 ". 2OmV









S a







im
a *4
I
I* S S
S a











































Figure 4-3.. Effect of a short train of orthodromic stimuli on the



pre syn apt ic act ion potent ial. Each s t imul us generates an act ion poten t ial i n the pre syn apt icI ne rve t erm i nal St imul at ion rate for t hi s
example s 20/sec. [Ca+ .5 mM, [Mg2+] 4 mM.
sa U
S S* *.
** S


* .
S S .
* *





fi S ll









Fiue43.Efc of a hr ri ofotormcsiuioSh



exml s 0sc [C ] 25 mM [M2+ = SM










-- .i En 0. ==- Ifl "C
X 7X 0t O*


4 control AHP t 1/2 % control AHP amplitude % control AP du
-, a.-, ct 0-I control AHP t 1/2 op u


" P. (D --J S a a 0 mo 0 + am
-- .t" 0 (1 /V"t0 I I I I I I II 2000- f "l 0u +oo ,c oooo o o o ooo
C D W r-


H <:u =rmD

ur --D U) <:: -. .
4 .. @I a gIII I I '''''' I '' lI '

* O O 0
*n- "
MS

0*0 o "0
C* *
0 -9"'W la








D - D I
0 O 0 0






PIL:(D(D )Jl c F
8 OCT ##P Pr +00t I ( 0 0








A**~0 11 5 $>70942D O O O




:4 O # "+ fl -ln rc









L~M* 2) 1 *





L+ O .* IIr &3 O((~
CDn




0 rA + *
O(


a
(Ar *1 00 0
COO 0~


Wr@WCD C1

NflQC DQ C3 3 0
can.








62




15 15



S10 10


P.
E S
5 5
a* *
rS



0 = 0
-90 -80 -70 -60 -50
Membrane potential (mV)
CV








Figure 4-5. Effects of varying nerve terminal membrane potential on afterhyperpolarization (AHP) amplitude. Data are from a single cell in which the membrane potential was varied by injecting current through the recording electrode. Line drawn through the data points by eye. [Ca2+] = 1.5 mM, [Mg +] = 4 mM. Resting membrane potential = -52 mV.







63

point where it became unmeasurable. This finding is consistent with a role of K+ currents in generation of the AHP in the presynaptic nerve terminal.
Due to the similar mechanisms underlying action potential repolarization and AHP generation, it is not unreasonable to assume that a single underlying change in the presynaptic terminal is affecting both AHP amplitude and action potential duration. Figure 4-6 plots the relationship between changes in action potential duration and changes in AHP amplitude following paired pulse stimulation. As AHP amplitude decreases, there is a clear increase in action potential duration. The Presynaptic Action Potential and Facilitation
An ideal way to describe the correlation between changes in the

nerve terminal action potential and increases in transmitter release would be to record simultaneously from pairs of pre- and postsynaptic cells during repetitive stimulation. While there has been a report of successful simultaneous penetration of both elements in the chick ciliary ganglion, this procedure yielded only a few very brief recordings (Yawo and Momiyama, 1993). Instead, the relationship between intracellularly recorded changes in the presynaptic action potential and efficacy of release was investigated in this study using extracellular records of postsynaptic compound action potentials. It has been shown that changes in the compound action potential during repetitive stimulation accurately reflect changes in EPSP amplitude under the condi+4nne uien a kner ean thatn+e 21








64


C
o 130

4OO
IC

R= 0.76

120
O 4-O
ct
c O O

S110 o a 0 O
-- O
OO
2- 00 O O
4' C
o
a 100
a 100 80 60 40 20

Scontrol AHP amplitude





Figure 4-6. Correlation between measures of presynaptic action
potential duration and AHP amplitude. Paired pulse stimulation produced an increase in action potential duration and a decrease in AHP amplitude (see text). The effects of a single conditioning impulse on action potential duration (ordinate) are compared to its effects on AHP amplitude (abcissa). A regression line is drawn through the data, which are results of 8 intervals applied in various combinations to 7 different cells.








65

A 3.0

20
C1.5
> 1.0 o o
0.5 o
0.0 o oo
-0.5
100 110 120 130 %control AP duration

B 3.0
2.5
2.0 0
1.5
too


0.5
0.0 -o 0 o

100 80 60 40 20 % control AHP amplitude




Figure 4-7. Correlation between facilitation of extracellularly recorded compound action potential and changes in intracellularly recorded presynaptic action potential. Each point represents the average of 4 to 16 responses from a single cell. Data pres nted here are from experi ents done under low quantal conditions ([Ca'+] = 1.25 to 1.5 mM, [Mg=] 4 mM) where simultaneous measures of intracellularly recorded presynaptic action potential and extracellularly recorded compound action potentials were obtained (V(t) calculated as described in Chapter 2). Data are results of 8 different interstimulus intervals applied to 7 different cells. (A): Relationship between extracellularly
reoddVt a and hes in .rsnoi acio *oeta duraton.







66

from 7 cells at interstimulus intervals between 25 and 2000 msec. Figure 4-7A plots V(t) as a function of action potential duration. Figure 4-7B plots V(t) as a function of AHP amplitude. Although changes in synaptic efficacy appear to correlate with both action potential duration and AHP amplitude, the correlation with AHP amplitude was stronger, probably because of the greater reliability of measures of AHP amplitude (see Note 1 at the end of this chapter). These results show that stimulation-induced changes in the presynaptic action potential and increases in transmitter release occur simultaneously in the chick ciliary ganglion.
The effect of repetitive stimulation on the presynaptic action

potential is most pronounced in the first 150 msec of a 20/sec train (Figure 4-5). The first component of facilitation (F1), which has a time constant of decay of about 60 msec in the chick ciliary ganglion (see Chapter 2), seems to accumulate in a similar manner to that described in the frog neuromuscular junction (Figure 2-3; Magleby and Zengel, 1982). If this is the case, then the time course of accumulation of Fl would be quite similar to the time course of the effects of repetitive stimulation on the presynaptic action potential shown in Figure 4-5. To determine whether changes in the presynaptic action potential correlate with facilitation of transmitter release, a paired pulse paradigm (like that used to describe facilitation in Chapter 2) was employed to characterize changes in the presynaptic action potent Jial n unde a j 4. A rrL shw to pou aciitan.
























67














II 1 1


*
It I a

I ft U

a I

*


I a *


24msec 93 msec
ft
ft e

S
* S

*
a
ft
4* S

* S
ft S

a
a
ft

ft S ft


S 5 ft
S
ft S S


ft ft

S
ft ft a a *
S
U
S
4 S ft

a
ft ft
ft
5 ft ft
ft

S

ft
ft ft *
S

ft 5 ft
ft ft

ft ft ft S
S

ft
ft
S ft ft C S a
S aft
a
-~

ft
ft. a
ft S U
U ft
ft 50 msec

ft













lOmV










IU


S
S 4
a
ft a

ft
a
ft


S
460 msec


/
U ft U ft S ft ft
ft
ft

*
ft
ft S S ft

ft ft
S
ft S
U


ft

S
C
5 ft

S

ft

S
S
C C

a
S
a ft
S


a ft ft


ft



CC
S





ft

ft

S U ft

C





ft ft ft U






I








68

potential duration was increased and AHP amplitude was dramatically decreased, whereas at longer intervals the effects on AHP amplitude and action potential duration were less pronounced. The time course of the effects on AHP amplitude are more clearly seen in Figure 4-9. Like facilitation of transmitter release, the effects of a single conditioning impulse on AHP amplitude were greatest at short interstimulus intervals (less than 150 msec) and less pronounced at longer intervals (Figure 4-gA). Figure 4-9B plots the same data as in (A), expressed as percent inhibition of AHP amplitude. In three experiments where 5 or more intervals were tested, the decay of the effect on AHP amplitude, plotted as percent AHP inhibition, could be described by a dual exponential decay with time constants of 64+20 msec and 1219+292 msec. In experiments where fewer than 5 intervals were applied, the shortest interstimulus intervals (<200 msec) consistently produced the greatest effects on the presynaptic action potential, similar to the results presented in Figure 4-9.

Discussion
The goal of this study was to record from the presynaptic nerve
terminal of the chick ciliary ganglion during repetitive stimulation and to investigate the relationship between the presynaptic action potential and transmitter release. It is shown that repetitive stimulation, under conditions that produce facilitation of transmitter release, gives rise to changes in the presynaptic action potential that







69


A 100
a 90
S80 -.0 -----*- '- '''
70
I 70 )- O'0
o 60
50 O
o 9
a 40
30 ,
0 200 400 600 800 1000
Interatimulus interval (msec)
B 100
O

o. O -t/1661ms
0 "rt=-']2e

C
P~ E O slow
-- I.
-a
< -t/62ma
7 =53e
fast
10 I.. I
0 200 400 600 800 1000
Interatimulus interval (msec)


Figure 4-9. Effect of a single conditioning impulse on presynaptic action potential AHP amplitude. Data points represent averages of 4 to 10 identical trials. (A): Effects of paired pulse stimulation on AHP amplitude. Data are expressed as percent of the control AHP amplitude.
(B): Same data as (A) expressed as percent maximal inhibition of AHP amplitude and plotted on a semilogarithmic scale. A regression line drawn through the linear portion of the data (points beyond 150 msec) gives a line described by the equation Tslow. The values of the slow regression line at earlier points (<150 msec) were calculated and the contributions of the slower regression were subtracted assuming an







70

to be a common phenomenon in neurons and other electrically excitable cells (e.g. Bourque, 1991; Crest and Gola, 1993; Quattrocki et al., 1994) and may represent a common mechanism for modulation of release during and following repetitive stimulation. The Action Potential
Before considering the implications of this study, a brief description of the currents'that comprise the action potential is in order. Several overlapping currents comprise the action potential (Hodgkin and Huxley, 1939). The depolarizing phase of the action potential results from an increase in Na+ permeability and subsequent entry of Na+ ions down a strong electrochemical gradient. The falling or repolarizing phase of the action potential involves a decrease in Na+ permeability (Na+ channel inactivation) and an increase in permeability to K+ due to the opening of voltage-dependent K+ channels. The increase in K+ permeability can last for several milliseconds, so that in many cells K+ efflux can hyperpolarize the membrane beyond the resting potential, producing an afterhyperpolarization (AHP). Repetitive Stimulation and the Action Potential
Martin and Pilar (1964c) looked at the effects of repetitive stimulation on the presynaptic action potential of the chick ciliary ganglion. A paired pulse paradigm like the one used in the current study was employed. The only effect they reported was a depression of action potential amplitude at interstimulus intervals of 5 msec or less. (see








71

Several factors may have contributed to differences between the earlier results of Martin and Pilar and those presented here. When Martin and Pilar performed their classic experiments describing electrical and chemical transmission through the ciliary ganglion, they compared single action potentials. The results presented here are taken from averaged data, which should make small changes in action potential waveform more easily apparent. Also, in the 1960s, current theories about the role of the action potential in facilitation seem to have focussed primarily on changes in action potential amplitude (e.g. Hubbard and Schmidt, 1963; Takeuchi and Takeuchi, 1962). For this reason, perhaps, the effects of repetitive stimulation on action potential amplitude were more vigorously examined than other aspects of the action potential.
In recent studies of a number of neuronal preparations, repetitive stimulation has been shown to have effects on the action potential waveform (reviewed in Nicholls et al., 1992). For example, Charlton and Bittner (1978b) reported that repetitive stimulation of the squid giant synapse can lead to an increase in the amplitude of the presynaptic action potential, but only during the first few responses in a train. The effects of repetitive stimulation on the duration of the action potential were not reported, although their data appear to show a decrease in AHP amplitude during tetanic stimulation (Figures 3 and 11, Charlton and Bittner, 1978b). Similarly, at the crayfish neuromuscular







72

Mechanisms for Action Potential Broadening During Repetitive Stimulation
It is not clear what electrical events bring about the increases in action potential duration reported in this chapter. One possible mechanism involves activation of Ca2+-dependent cation channels (Partridge and Swandulla, 1988). Although these channels have been described in chick sensory neurons, their activation range (over 1 pM; RazaniBoroujerdi and Partridge, 1993) raises doubts as to their role in neuronal function under more normal ionic conditions. A more likely mechanism for broadening the presynaptic spike is a stimulation-induced inhibition of K+ current(s).

Figure 4-7 shows that the reversal potential for the calyx AHP lies close to the predicted equilibrium potential for K+. In the ciliary ganglion, several K+ currents have been observed in the presynaptic nerve ending, including delayed rectifier (Dryer and Chiappinelli, 1985), Ca2+-activated (Fletcher and Chiappinelli, 1992b, 1993) and inwardly rectifying K+ currents (Dryer and Chiappinelli, 1985; Fletcher and Chiappinelli, 1992a).
There are several mechanisms through which repetitive stimulation
could affect K+ currents. One possibility is that the accumulation of [Ca2+]i during stimulation (e.g. Charlton et al., 1982; Smith and Augustine, 1988; Stinnakre and Tauc, 1973) could act to decrease K+ efflux. There have been reports of Ca2+-inhibited K+ channels in many cell types (see Marty, 1989). For example, increases in intracellular
a2 s d e h o en n pi oea







73

There are also reports that describe cumulative stimulation-induced inactivation of K+ currents. In molluscan neurons, cumulative inactivation of a fast delayed rectifier K+ current contributes to stimulation-induced action potential broadening (Aldrich et al., 1979; Thompson, 1977). A delayed rectifier channel cloned from rat brain (Marom and Levitan, 1994) also shows similar inactivation properties. EdrySchiller and Rahamimoff (1993), on the basis of data obtained from the fused Torpedo synaptosome preparation, have proposed a "potassium inactivation hypothesis" for frequency modulation of transmitter release. They suggest that reactivation of slow K+ channels following an action potential may be responsible for action potential broadening and facilitation of release. Mallart (1985) proposed a similar role for the inactivating Ca2+-activated K+ channel in the motor nerve terminal of the mouse. In the chick ciliary ganglion, cumulative K+ current inactivation has been observed in the postsynaptic ciliary neuron (S. Dryer, personal communication) and it is reasonable to assume that similar K+ channels could be present in the presynaptic nerve terminal
Action Potential Duration: Effects on Release
Although the mechanisms underlying stimulation-induced increases in action potential duration in the chick ciliary ganglion are not fully understood, the fact remains that many studies have shown that changes in presynaptic action potential duration can have large effects on synaptic. tansmitter. relea e-g. Augs., at I u a neLa 10 1 et .6







74

a ten percent increase in action potential duration resulted in a 60 percent increase in the amplitude of the postsynaptic current (Augustine, 1990). Hochner and colleagues (1986) report similar results in Aplysia sensory ganglia: a 10% to 30% increase in presynaptic action potential duration was correlated with a 25% to 120% increase in postsynaptic potential amplitude. A similar relationship between presynaptic action potential duration and increased synaptic efficacy can be seen in the chick ciliary ganglion (Figure 4-7). Current Hypotheses: Stimulation-Induced Increases in Release
Some of the earliest theories concerning mechanisms of stimulationinduced increases in transmitter release involved the presynaptic action potential (B1loedel et al., 1966; Hubbard and Schmidt, 1963; Miledi and Slater, 1966; Takeuchi and Takeuchi, 1962). Experimental testing of this hypothesis has shown that changes in the presynaptic action potential cannot account for all of the observed increases in transmitter release during and following repetitive stimulation (Charlton and Bittner, 1978b; Zucker, 1974). However, these results should not be interpreted as eliminating the possibility that changes in the action potential play a role in mediating the effects of repetitive stimulation on transmitter release. Several individual components, which appear to arise from separate mechanisms, contribute to stimulation-induced increases in release. It has been suggested that different Ca2+-dependent processes are involved in initiation and modulation
r release A l (B and usl,12 Z e t al., 1l3a. The








75

Recent reports from this laboratory suggest that facilitation of transmitter release is related to entry of Ca2+ through voltageactivated channels (Zengel et al., 1993a,b), whereas augmentation appears to be particularly sensitive to the intraterminal concentration of Ca2+ (Zengel et al., 1994). Other mechanisms underlying stimulation-induced modulation of release (for example, changes in Ca2+ channel kinetics [Lee, 1987], Ca2+ buffering [L1inas et al., 1992; Neher and Augustine, 1990], phosphorylation of nerve terminal and vesicular proteins [reviewed in Greengard et al., 1993]) may be contributing to stimulation-induced increases in transmitter release in the ciliary ganglion. Unfortunately, these processes are not readily observed by electrophysiological means.
The "residual Ca2+" hypothesis (Katz and Miledi, 1968) proposes
that Ca2+ accumulates in the nerve terminal during repetitive stimulation, and that this cumulative increase in Ca2+, or some Ca2+-sensitive process, contributes to stimulation-induced increases in release. The findings reported here are consistent with a role for residual Ca2+ in facilitation. Stimulation-induced broadening of the presynaptic action potential should cause a significant increase in Ca2+ influx during repetitive stimulation. The time course of stimulation-induced changes in action potential duration and AHP amplitude indicate that, if these changes are acting to bring about increases in neurotransmitter release, they are most likely to be involved in facilitation, as








76
itive stimulation on this parameter was a primary goal of this study. Unfortunately, due in part to the length of the preganglionic nerve and the rapid conduction of the action potential in this preparation, the early rising phase of the presynaptic action potential was often temporally superimposed on the stimulus artifact (see Figure 4-2), making precise measurement of action potential duration difficult. This technical problem may have added an element of variability to measures of action potential duration, even though changes were always expressed relative to a paired control response. Due to the rapid time course of the stimulus artifact, measures of AHP amplitude were not affected.


















CHAPTER 5




CONCLUSIONS


As recently as 100 years ago, energetic debate surrounded hypotheses concerning the functional organization of the nervous system. The proponents of cell theory suggested that nerves were independent units, a proposal that went against the prevalent theory that all nerves were a single continuous structure, part of a syncytium, interconnected by protoplasmic bridges (see Nicholls et al., 1992, p. 185). Later, it was accepted that nerves were discontinuous, separated by small gaps across which information was passed through unknown methods. Otto Loewi, in 1921, performed a simple and convincing series of experiments showing that stimulation of the vagus nerve acts to slow heart rate by releasing a diffusible substance (acetylcholine). Dale and others soon established the role of acetylcholine as a neurotransmitter at the neuromuscular junction and in autonomic ganglia (e.g. Dale and Feld-







78

In the 1960s, Bernard Katz and others showed that initiation of

transmitter release is dependent upon the presence of Ca2+ ions in the extracellular medium (Katz and Miledi, 1965, 1967). These studies and others led to the formulation of the "Ca2+ hypothesis", which proposed that the entry of Ca2+ ions into the presynaptic terminal is an essential step linking membrane depolarization to transmitter release.

Despite the wide acceptance of the Ca2+ hypothesis, recent reports have suggested that other factors may be sufficient to initiate exocytosis in the absence of extracellular Ca2+. For example, Stuenkel and Nordmann (1993) report Na+-dependent neuropeptide release from the rat neurohypophysis in the absence of a rise in intracellular Ca2+. Parnas
and co-workers have investigated the role of nerve terminal depolarization in initiation of transmitter release at the frog neuromuscular junction and have suggested that a depolarization-dependent factor promotes release in cooperation with intracellular Ca2+ (e.g. Hochner et al., 1989; Parnas et al., 1986). Other investigators have also reported depolarization-induced transmitter release that appears to occur in the absence of Ca2+ influx (Mosier and Zengel, 1993; Silinsky et al., 1995). These reports suggest that although intracellular Ca2+ is necessary for most forms of exocytosis, it is not the sole factor acting to initiate transmitter release.
As described in previous chapters, four components of stimulationinduced increases in synaptic transmitter release have been reported at
many~ f sgn, 7 Mar t di







79
tion in the chick ciliary ganglion (Chapter 2). Many investigators have suggested that separate underlying mechanisms bring about the four components of increased release (e.g. Landau et al., 1973; Lev-Tov and Rahamimoff, 1980; Magleby, 1973; Magleby and Zengel, 1982). It has been suggested that the effects of repetitive stimulation are a consequence of an increase in residual intracellular Ca2+ or a calcium activated factor (Ca*) that causes a given presynaptic depolarization to release an increased amount of transmitter (Katz and Miledi, 1965, 1968). This "residual calcium hypothesis" is the most widely accepted theory to account for stimulation-induced increases in release. In its simplest form, however, this model fails to adequately account for all of the properties of stimulation-induced increased increases in release (e.g. Zengel and Magleby, 1980, 1981, 1982; Bain and Quastel, 1992a). Although Ca2+ ions appear to play an important role in the facilitation phase (Katz and Miledi, 1968; Zengel et al., 1993a,b) and augmentation phase (Erulkar and Rahamimoff, 1978; Magleby and Zengel, 1976; Zengel et al., 1994) of increased release, the potentiation phase does not appear to involve Ca2+; instead, it has been suggested that an accumulation of Na+ ions in the nerve terminal may be involved in this phase of increased release (e.g. Birks and Cohen, 1968; Nussinovitch and Rahamimoff, 1988). If the mechanisms underlying stimulation-induced increases in release can be described, the release process itself will be more completely understood.
Data oresented in Chanter 4 of this disrtation show that facili-







80

duration of the presynaptic depolarization recruits a greater percentage of voltage-dependent Ca2+ channels in the calyciform nerve terminal and that the resulting increase in Ca2+ influx underlies facilitation of transmitter release. The time constant of activation for Ca2+ currents in the calyx is 0.9 1.6 msec at 230 C (Stanley and Goping, 1991) and calcium currents recorded using whole-cell patch-clamp techniques do not reach a peak level for several milliseconds following sustained depolarization (Stanley, 1989). The calyx action potential only produces a depolarization sufficient to activate Ca2+ channels for a very short time (1 to 2 msec, Martin and Pilar, 1963a; personal observations), which would not activate the full complement of presynaptic Ca2+ channels. This conclusion is not without precedent. Earlier studies have also suggested that increased Ca2+ influx by consecutive stimuli might cause facilitation (e.g. Stinnakre and Tauc, 1973).
Several other areas of study bear mentioning when discussing the
effects of repetitive stimulation on neurotransmitter release. It has been reported in the frog and crayfish neuromuscular junctions and other preparations that the relative contribution of the various components of stimulation-induced changes in release may vary across cells within an experimental preparation (Collins et al., 1984; Fadoga and Brookhart, 1962; Frank, 1973; Meriney and Grinnell, 1991). One use of these systems would be to identify the biochemical or electrophysiological differences between, for example, facilitating and nonfcil :Itatnn







81

involved in synaptic vesicle trafficking, mobilization and exocytosis. Studies have already begun to describe the role of these proteins in release of neurotransmitter under control conditions (e.g. Lledo et al., 1993; Pevsner et al., 1994) and following repetitive stimulation (e.g. Tarelli et al., 1992). More recently, the use of antisense nucleotides and transgenic mice has enabled researchers to study synaptic transmission in synaptic preparations that have modified versions of these proteins (e.g. Alder et al., 1992) or lack them entirely (e.g. Rosahi et al., 1993).

The presynaptic nerve terminal is a highly specialized and complex
structure. It contains mechanisms for uptake, storage and synthesis of transmitter substances, as well as voltage-sensitive proteins that can alter ionic conditions in response to changes in membrane potential, and ligand-activated complexes that can respond to cytoplasmic or extracellular chemical signals. Consequently, there are many ways by which modulation of the release process could occur, including increases in intracellular Ca2+ and intracellular Na+, activation of presynaptic ligand gated receptors, changes in the kinetics of voltageactivated ion channels and changes in the presynaptic action potential.
Recent technological advances in many elements of neuroscience
research (cell culture, molecular biological, genetic and imaging techniques) are increasing the precision with which neurons can be studied. With this enhanced ability to observe and describe neuronal physiology,
an o.ng-stand ng. qetins will cnnn be answered. Tt would 2ppear















REFERENCES

Alder, J., Xie, Z.-P., Valtorta, F., Greengard, P., and Poo, M.-m.
(1992) Antibodies to synaptophysin interfere with transmitter
secretion at neuromuscular synapses. Neuron, 9:759-768.
Aldrich, R.W., Getting, P.A., and Thompson, S.H. (1979) Mechanism of
frequency-dependent broadening of molluscan neurone somata. L
Physiol (London), 291:531-544.
Artalejo, C.R., Adams, M.E. and Fox, A.P. (1994) Three types of calcium
channels trigger secretion with different efficacies in chromaffin
cells. Nature 367:72-76.

Augustine, G.J. (1990) Regulation of transmitter release at the squid
giant synapse by presynaptic delayed rectifier potassium current.
J. Physiol. (London), 431:343-364.
Augustine, G.J., Smith, S.J., and Zucker. R.S. (1982) Role of presynaptic calcium ions and channels in synaptic facilitation and depression at the squid giant synapse. J. Physiol. (London), 323:173-193.
Bain, A.I., and Quastel, D.M.J. (1992) Quantal transmitter release
mediated by strontium at the mouse neuromuscular junction. J. Physiol. (London), 450:63-87.

Bain, A.I., and Quastel, D.M.J. (1992) Multiplicative and additive
Ca2+-dependent components of facilitation at mouse endplates. L.
Physiol. (London), 455:383-405.
Bennett, MK., and Scheller, R.H. (1993) The molecular machinery for
secretion is conserved from yeast to neurons. PNAS (USA),
90:2559-2563.
Bennett, M.R., and Ho, S. (1991) Probabilistic secretion of quanta from
nerve terminals in avian ciliary ganglia modulated by adenosine. J.
Physiol., 440:513-527.
Bertolino, M., and L1inas, R. (1992) The role of voltage activated and
receptor operated calcium channels in neuronal cells. Ann. Rev.








83

Bittner, G.D., and Baxter, D.A. (1991) Synaptic plasticity at crayfish
neuromuscular junctions: facilitation and augmentation. Synapse,
7:235-243.

Bloedel, J., Gage, P.W., Linas, R., and Quastel, D.M. (1966) Transmitter release at the squid giant synapse in the presence of tetrodotoxin. Nature, 212:49-50.

Bourque, C.W. (1991) Activity-dependent modulation of nerve terminal
excitation in a mammalian peptidergic system. TINS, 14:28-31.

Chang, C.C., Lin, S.-O., Hong, S.J., and Chiou, L.C. (1988) Neuromuscular block by verapamil and diltiazem and inhibition of acetylcholine release. Brain Res., 454:332-339.

Charlton, M., and Bittner, G.D. (1978a) Facilitation of transmitter
release at squid synapses. J. Gen. Physiol., 72:471-486.

Charlton, M., and Bittner, G.D. (1978b) Presynaptic potentials and
facilitation of transmitter release at squid synapses. J. Gen.
Physiol., 72:487-511.
Charlton, M.P., Smith, S.J., and Zucker, R.S. (1982) Role of presynaptic calcium ions and channels in synaptic facilitation and depression at the squid giant synapse. J. Physiol. (London), 323:
173-193.

Collins, W.F., Honig, M.G., and Mendell, L.M. (1984) Heterogeneity of
group la synapses on homonymous alpha- motoneurons as revealed by
high-frequency stimulation of la afferent fibers. J. Neurophysiol.,
52:980-993.

Crest, M., and Gola, M. (1993) Large conductance Ca2+-activated K+
channels are involved in both spike shaping and firing regulation
in Helix neurones. J. Physiol. (London), 465:265-287.

Curtis, D.R., and Eccles, J.C. (1960) Synaptic action during and after
repetitive stimulation. J. Physiol. (London), 150:374-398.
Dale, H.H., and Feldberg, W. (1936) The chemical transmission of secretory impulses to the sweat glands of the cat. J. Physiol. (London),
82:121-128.

DeCamilli, P. (1993) Exocytosis goes with a SNAP. Nature, 364:387-389.
flaLrenzo, A. T ITO0 Tahe 4in etrun nf synanes in the ciliary








84

Dryer, S.E., and Chiappinelli, V.A. (1983) Kappa-bungarotoxin: an
intracellular study demonstrating blockade of neuronal nicotinic
receptors by a snake neurotoxin. Brain Res., 289:317-321.

Dryer, S.E., and Chiappinelli, V.A. (1985) Substance P depolarizes
nerve terminals in an autonomic ganglion. Brain Res., 336:190-194.
Dudel, J., and Kuffler, S.W. (1961) The quantal nature of transmission
and spontaneous miniature potentials at the crayfish neuromuscular
junction. J. Physiol. (London), 155:514-529.
Dudel, J., Parnas, I., and Parnas, H. (1983) Neurotransmitter release
and its facilitation in crayfish, VI. Release determined by both
intracellular calcium and depolarization of the nerve terminal.
Pfluqers Archly., 339:1-10.
Edry-Schiller, J., and Rahamimoff, R. (1993) Activation and inactivation of the bursting potassium channel from fused Torpedo synaptosomes. J. Physiol. (London), 471:659-678.
Erulkar, S.D. (1983) The modulations of neurotransmitter release at
synaptic junctions. Rev. Biochem. Pharmacol., 98:63-175.
Erulkar, S.D., and Rahamimoff, R. (1978) The role of calcium ions in
tetanic and post-tetanic increase in miniature end-plate potential
frequency. J. Physiol. (London), 287:501-511.
Fadoga, E., and Brookhart, J.M. (1962) Interactions of excitatory postsynaptic potentials generated at different sites on the frog motoneuron. J. NeuroDhysiol., 25:790-804.

Feng, T.P. (1941) Studies on the neuromuscular junction. XXVI. The
changes of the end-plate potential during and after prolonged stimulation. Chin. J. Physiol., 16:341-372.
Fletcher, G.H., and Chiappinelli, V.A. (1992a) An inward rectifier is
present in presynaptic nerve terminals in the chick ciliary ganglion. Brain Res., 575:103-112.
Fletcher, G.H., and Chiappinelli, V.A. (1992b) Spontaneous miniature
hyperpolarizations of presynaptic nerve terminals in the chick ciliary ganglion. Brain Res., 579:165-168.
Fletcher, G.H., and Chiappinelli, V.A. (1993) The actions of the kappa
1 opioid agonist U-50,488 on presynaptic nerve terminals of the
ch L..ick1 ci.lar anliona.a .N ue 5a: hmn29*








85

Fox, A.P., Nowycky, M.C. and Tsien, R.W. (1987b) Single-channel recordings of three types of calcium channels on chick sensory neurons.
J. Physiol. (London), 394:173-200.
Frank. E. (1973) Matching of facilitation at the neuromuscular junction
of the lobster. J. Physiol. (London), 233:635-658.
Greengard, P., Valtorta, F., Czernik, A.J., and Benfenati, F. (1993)
Synaptic vesicle phosphoproteins and regulation of synaptic function. Science, 259:780-785.

Hess, A. (1965) Developmental changes in the structure of the synapse
on the myelinated cell bodies of the chicken ciliary ganglion. 3L.
Cell Biol., 25:1-19.

Hochner, B., Klein, M., Schacher, S., and Kandel, ER. (1986) Action
potential duration and the modulation of transmitter release from
the sensory neurons of Aplysia in presynaptic facilitation and
behavioral sensitization. PNAS (USA), 83:8410-8414.
Hochner, B., Parnas, H., and Parnas, I. (1989) Membrane depolarization
evokes neurotransmitter release in the absence of calcium entry.
Nature, 342:433-435.

Hodgkin, A.L., and Huxley, A.F. (1939) Action potentials recorded from
inside a nerve fibre. Nature 144:710-711.
Hubbard, J.I. (1963) Repetitive stimulation at the neuromuscular junction, and the mobilization of transmitter. J. Physiol. (London),
169:641-662.
Hubbard, J.I., and Schmidt, R.F. (1963) An electrophysiological investigation of mammalian motor nerve terminals J. Physiol. (London),
166:145-165.

Katz, B., and Miledi, R. (1965) The measurement of synaptic delay, and
the time course of acetylcholine release at the neuromuscular
junction. Proc. Roy. Acad. B161:483-495

Katz, B., and Miledi, R. (1967) The timing of calcium action during
neuromuscular facilitation. J. Physiol. (London), 189:535-544.
Katz, B., and Miledi, R. (1968) The role of calcium in neuromuscular
facilitation. J. Phvysiol. (London), 195:481-92.
Klin a.ardo J. an K.~aE.R. r(1982)n 4 Sertonlnf jmod~au.ates~ a









86

Landau, E.M., Smolinsky, A., and Lass, Y. (1973) Post-tetanic potentiation and facilitation do not share a common calcium-dependent mechanism. Nature New Biolovgy, 244:155-157.

Landmesser, L. and Pilar, G. (1972) The onset and development of transmission i n the chick ciliary ganglion. J. Physiol. (London),
222:691-713.

Lee, K.S. (1987) Potentiation of the calcium-channel currents of
internally perfused mammalian heart cells by repetitive depolarization. PNAS (USA), 84:3941-3945.
Lentzner, A., Bykov, V., and Bartschat, D.K. (1992) Time-resolved
changes in intracellular calcium following depolarization of rat
brain synaptosomes. J. Physiol. (London), 450:613-628.

Lev-Tov, A., and Rahamimoff, R. (1980) A study of tetanic and posttetani c potentiation of miniature end-plate potentials at the frog
neuromuscular junction. J. Physiol. (London), 309:247-273.

Liley, A.W. (1956) The quantal components of the mammalian end-plate
potential. J. Physiol. (London), 133:571-587.
Liley, A.W., and North, K.A.K. (1953) An electrical investigation of
effects of repetitive stimulation on mammalian neuromuscular junction.J. NeuroDhvys., 16:509-527.
L1edo, P.-M., Vernier, P., Vincent, J.-D., Mason, W.T., and Zorec, R.
(1993) Inhibition of Rab3B expression attenuates calcium-dependent
exocytosis in rat anterior pituitary cells. Nature, 364:540-544.
L1inas, R., Steinberg, I.Z., and Walton, K. (1981) Presynaptic calcium
currents in squid giant synapse. Biophys. J. 33:289-321.

L1inas, R., Sugimori, M., Lin, J.W., and Cherksey, B. (1989) Blocking
and isolation of a calcium channel from neurons in mammals and
cephalopods utilizing a toxin fraction (FTX) from funnel web spider
poison. PNAS (USA), 86:1689-1693.
L1inas, R., Sugimori, M., and Silver, R.B. (1992) Microdomains of high
calcium concentration in a presynaptic terminal. Science,
256:677-679.
Loewi, O0. (1921) Uber humorale Ubertragbarkeit der Herznervenwirkung.
Pfluqers Archiv. 189:239-242.








87

Magleby, K.L., and Zengel, J.E. (1975a) A dual effect of repetitive
stimulation on post-synaptic potentiation of transmitter release at the frog neuromuscular junction. J. Physiol. (London), 245:163-182.

Magleby, K.L., and Zengel, J.E. (1975b) A quantitative description of
tetanic and post-tetanic potentiation of transmitter release at the
frog neuromuscular junction. J. Physiol. (London) 245:183-208.
Magleby, K.L., and Zengel, J.E. (1976) Augmentation: a process that
acts to increase transmitter release at the frog neuromuscular
junction. J. Physiol. (London), 257:449-470.
Magleby, K.L., and Zengel, J.E. (1982) A quantitative description of
stimulation-induced changes in transmitter release at the frog
neuromuscular junction. J. Gen. Physiol., 80:613-638.
Mallart, A. (1985) A calcium-activated potassium current in motor nerve
terminals of the mouse. J. Physiol., 368:577-591.
Mallart, A., and Martin, A.R. (1967) An analysis of facilitation of
transmitter release at the neuromuscular junction of the frog. L
Physiol (London), 193:679-694.
Marom, S., and Levitan, I. B. (1994) State-dependent inactivation of the
Kv3 potassium channel. Biophys. J., 67:579-589.
Martin, A.R., and Pilar, G. (1963a) Dual mode of synaptic transmission
in the avian ciliary ganglion. J. Physiol. (London), 168:443-463.
Martin, A.R., and Pilar, G. (1963b) Transmission through the ciliary
ganglion of the chick. J. Physiol. (London), 168:464-475.
Martin, A.R., and Pilar, G. (1964a) An analysis of electrical coupling
at synapses in the avian ciliary ganglion. J. Physiol. (London),
171:4545-475.
Martin, A.R., and Pilar, G. (1964b) Quantal components of the synaptic
potential in the ciliary ganglion of the chick. J. Physiol. (London), 175:1-16.
Martin, A.R., and Pilar, G. (1964c) Presynaptic and post-synaptic
events during post-tetanic potentiation and facilitation in the
avian ciliary ganglion. 3. Physiol. (London), 175:17-30.
Marty, A. (1989) The physiological role of calcium-dependent channels









88

Marwitt, R., Pilar, G., and Weakly, J.N. (1971) Characterization of two
cell populations in avian ciliary ganglion. Brain Res., 25:317-334.

Mastrogiacomo, A., Evans, C.J., and Gundersen, C.B. (1994) Antibody
peptides against a cysteine-string protein. J. Neurochem.
62:873-880.

McNaughton, B.L. (1982) Long-term enhancement and short-term potentiation in rat fascia dendata act through different mechanisms. L
Physiol (London), 324:249-262.
Meriney, S.D., and Grinnell, A.D. (1991) Endogenous adenosine modulates
stimulation-induced depression at the frog neuromuscular junction.
J. Physiol. (London), 443:441-455.
Miledi, R., and Slater, C.R. (1966) The action of calcium on neuronal
synapses in the squid. J. Physiol. (London) 198:473-498.
Miledi, R., and Thies, R. (1971) Tetanic and post-tetanic rise in
frequency of miniature end-plate potentials in low-calcium solutions. J. Physiol. (London) 212:245-257.
Miller, R.J. (1987) Multiple calcium channels and neuronal function.
Science, 235:46-51.
Mosier, D.R., and Zengel, J.E. (1994) Evoked transmitter release at the
frog ng uromuscular junction in the presence of very low extracellular Ca +. Neurosci. Letters, 174:1-4.
Neher, E., and Augustine, G.J. (1990) Calcium gradients and buffers in
bovine chromaffin cells. J. Physiol. (London), 450:273-301.
Nicholls, J.G., Martin, A.R., and Wallace, B.G. (1992) From Neuron to
Brain. third edition, Sinauer Press, New York, USA.

Nowycky, M.C., Fox, A.P., and Tsien, R.W. (1985) Three types of neuronal calcium channel with different calcium agonist sensitivity.
Nature, 316:440-443.

Nussinovitch, I., and Rahamimoff, R. (1988) Ionic basis of tetanic and
post-tetanic potentiation at a mammalian neuromuscular junction. 3.
Physiol. (London) 396:435-455.
Parnas, H., Dudel, J., and Parnas, I. (1986) Neurotransmitter release
and its facilitation in crayfish VII. Another voltage-dependent
-rcs eieclimetycnrl h tim cors ofoa








89

Penner, R., and Dreyer, F. (1986) Two different presynaptic calcium
currents in mouse motor nerve terminals. Pfluq. Arch., 406:190-197.
Pevsner, J., Hsu, S.-c., Braun, J., Calakos, N., Ting, A.E., Bennett,
M.K., and Scheller, R.H. (1994) Specificity and regulation of a
synaptic vesicle docking complex. Neuron, 13:353-361.
Poage, R.E., and Zengel, J.E. (1993) Kinetic and pharmacological examination of stimulation-induced increases in synaptic efficacy in the
chick ciliary ganglion. Synapse, 14:81-89.
Porter, R. (1970) Early facilitation at corticomotoneuronal synapses.
J. Physiol (London), 207:733-745.

Quattrocki, E.A., Marshall, J., and Kaczmarek, L.K. (1994) A Shab
potassium channel contributes to action potential broadening in
peptidergic neurons. Neuron, 12:73-86.
Razani-Boroujerdi, S., and Partridge, L.D. (1993) Activation and modulation of calcium-activated non-selective cation channels from
embryonic chick sensory neurons. Brain Res., 623:195-200.

Regehr, W.G., and Mintz, I.M. (1994) Participation of multiple calcium
channels types in transmission at single climbing fiber to Purkinje
cell synapses. Neuron, 12:605-613.

Robitaille, R., and Charlton, M. (1992) Presynaptic calcium signals and
transmitter release are modulated by calcium activated potassium
channels. J. Neurosci., 12:297-305.

Rosahl, T.W., Geppert, M., Spillane, D., Herz, J., Hammer, R.E., Malenka, RC., and Sudhof, T.C. (1993) Short-term synaptic plasticity
is altered in mice lacking synapsin 1. Cell, 75:661-670.

Rosenthal, J. (1969) Post-tetanic potentiation at the neuromuscular
junction of the frog. J. Physiol. (London), 203:121-133.
Scheller, M.K., and Bennett, M.K. (1994) Molecular correlates of synaptic vesicle docking and fusion. Curr. Opin. Neurobiol., 4:324-9.

Scott, R.H., Pearson, H.A., and Dolphin, A.C. (1991) Aspects of vertebrate neuronal voltage-activated calcium currents and their regulation. Prog. Neurobiol. 36:485-520.
Silinsky, E.M., Watanabe, M., Redman, R.S., Qui, R., Hirsh, J.K., Hunt,
J..Slna '.. Alford. S. and Ma~na.. ..r in5) Neuro-








90

Stanley, E.F. (1989) Calcium currents in a vertebrate presynaptic nerve
terminal: the chick ciliary ganglion calyx. Brain Res.,
505:341-345.
Stanley, E.F. (1991) Single calcium channels on a cholinergic presynaptic nerve terminal. Neuron, 7:585-591.
Stanley, E.F. (1993) Single calcium channels and acetylcholine release
at a presynaptic nerve terminal. Neuron, 11: 1007-1011.
Stanley, E.F. and Atrakchi, A.H. (1990) Calcium currents recorded from
a vertebrate presynaptic nerve terminal are resistant to the dihydropyridine nifed ipine. PNAS (USA), 87:9683-9687.
Stanley, E.F. and Goping, G. (1991) Characterization of a calcium current i n a vertebrate cholinergic presynaptic nerve terminal. 2
Neurosci., 11:985-993.
Stinnakre, J., and Tauc, L. (1973) Calcium influx in active Aplysia
neurons detected by injected aequorin. Nature, 293:115-117.
Stuenkel, E.L., and Nordmann, J.J. (1993) Sodium-evoked, calciumindependent vasopressin release from rat isolated neurohypophysial
nerve endings. J. Physiol. (London), 468:357-78.

Sugita, S., Goldsmith, J.R., Baxter, D.A., amd Byrne, J.H. (1992)
Involvement of protein kinase C in serotonin-induced spike broadening and synaptic facilitation in sensorimotor connections of Aplysia. J. Neurophysiol., 68:643-651.
Suszkiw, J.B., O'Leary, M.E., Murawsky, M.M., and Wang, T. (1986) Presynaptic calcium channels in rat cortical synaptosomes: fast kinetics of phasic calcium influx, channel inactivation, and relationship to nitrendipine receptors. J. Neurosci., 6:1349-1357.
Takahashi, T., and Momiyama, A. (1993) Different types of calcium channels mediate central synaptic transmission. Nature 366:156-158.
Takeuchi, A., and Takeuchi, N. (1962) Electrical changes in pre- and
post-synaptic axons of the giant synapse of Loligo. J. Gen. Physiol. 45:1181-1193.
Tarelli, F.T., Bossi M., Fesce, R., Greengard, P., and Valtorta, F.
(1992) Synapsin i partially dissociates from synaptic vesicles during exocytosis induced by electrical stimulation. Neuron, 9:
a14 - 53









91

Vergara, C., and Latorra, R. (1983) Kinetics of Ca2+-activated K+
channels from rabbit mus le incorporated into planar bilayers.
Evidence for Ca'+ and BaZ+ blockade. J Gen. Physiol., 82:543-568.

Weinreich, D. (1971) Ionic mechanism of post-tetanic potentiation at
the neuromuscular junction of the frog. J. Physiol. (London)
212:431-446.

Yawo, H. (1990) Voltage-activated calcium currents in presynaptic nerve
terminals of the chicken ciliary ganglion. J. Physiol. (London),
428:199-213.

Yawo, H., and Chuhma, N. (1994) Omega-conotoxin-sensitive and
-resistant transmitter release from the chick ciliary presynaptic
terminal. J. Physiol. (London), 477:437-448.

Yawo, H., and Momiyama, A. (1993) Re-evaluation of calcium currents in
pre-and postsynaptic neurones of the chick ciliary ganglion. J.
Physiol. (London), 460:153-172.
Yoshikami, D., Bagobaldo, Z., and Olivera, B.M. (1989) The inhibitory
effects of omega conotoxin on calcium channels and synapses. Ann.
NY Acad. Sci., 560:230-248.

Younkin, S.G. (1974) An analysis of the role of calcium in facilitation
at the frog neuromuscular junction. J. Physiol. (London), 237:1-14.

Zengel, J.E., Lee, D.T., Sosa, M.A. and Mosier, D.R. (1993a) Effects of
calcium channel blockers on stimulation-induced increases in transmitter release at the frog neuromuscular junction. Synapse,
15:251-262.

Zengel, J.E., and Magleby, K.L. (1977) Transmitter release during
repetitive stimulation: selective changes produced by Sr + and
Ba' Science, 197:67-69.

Zengel, J.E., and Magleby, K.L. (1980) Differential effects of Ba, Sr,
and Ca on stimulation-induced changes in transmitter release at the
frog neuromuscular junction. J. Gen. Physiol., 76:175-211.

Zengel, J.E., and Magleby, K.L. (1981) Changes in miniature end-plate
potent ial frequency during repetitive stimulation in the presence
of CaZ+, Ba +, and SrZ+ at the frog neuromuscular junction. J.
Gen. Physiol., 77:503-529.
Zenel rJ.E..a nles a.L a182 Au meatati. nfaciiationof








92

Zengel, J.E., Magleby, K.L., Horn, J.P., McAfee, D.A., and Yarowski,
P.J. (1980) Facilitation, augmentation and potentiation of synaptic
transmission at the superior cervical ganglion of the rabbit. L.
Gen. Physiol., 76:213-231.

Zengel, J.E., Sosa, M.A., and Poage, R.E. (1993b) Omega-conotoxin
reduces facilitation of transmitter release at the frog neuromuscular junction. Brain Res., 611:25-30.

Zengel, J.E., Sosa, M.A., Poage, R.E., and Mosier, D.R. (1994) Role of
intracellular Ca2+ in stimulation-induced increases in transmitter
release at the frog neuromuscular junction. J. Gen. Physiol. 104:
337-355.

Zucker, R.S. (1974) Characteristics of crayfish neuromuscular facilitation and their calcium dependence. J. Physiol. (London),
241:91-110.

Zucker, R.S. (1989) Short-term synaptic plasticity. Ann. Rev. Neurosci., 12:13-31.












BIOGRAPH I CAL SKETCH

Robert Eliot was born, a month early, to James David Poage and
Scottie Johnston Poage on August 6, 1964, in Winston-Salem, North Carolina. The second of three boys, Bob moved with the family to Baltimore and then to Jacksonville and Orange Park, Florida, where he spent his formative years. Dr. Jim Poage received his doctoral degree from the University of Florida in counselor education, and has spent the better part of his life working in rehabilitation of drug- and alcoholdependent persons. Scottie Poage-Hotchkiss has a successful practice marriage counseling in Yuma, AZ.

His parents' careers interested young Mr. Poage in the field of
psychology, in which he pursued a bachelor's degree from the University of Florida in 1987. Rather than continue in that vein, Bob decided that understanding the underlying causes of mental illness and nervous system dysfunction represented the future of psychology. He began working in the field of neuroscience with primates, then for a while at the cellular and biochemical level, before finally settling in for a long stay as an electrophysiologist under the guidance of Janet Zengel. Upon completing his dissertation, Bob will be working with Dr. Steve Meriney at the University of Pittsburgh, where it is said to snow on occasion. He currently lives in Gainesville, Florida, with his wife, Heidi, his children, Shea Elizabeth and Scott Eliot, and pets too












I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Jait E. Zenge, C ir
Associate Professor of Neuroscience


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quali as a di sertation for the degree of Doctor of Philosophy.


Pter A.V. nderson
Professor of Neuroscience


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Thomas W. Vickroy
Associate Professor of geoscience


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.

4H/1









I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Pilp Posne
Professor of Physiology




This dissertation was submitted to the Graduate Faculty of the College of Medicine and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.
August, 1995 (
a Co lege of Medicine



Dean, Graduate School




Full Text
o
100 200 300 400 500
Time (msec)


64
c
% control AHP amplitude
Figure 4-6. Correlation between measures of presynaptic action
potential duration and AHP amplitude. Paired pulse stimulation pro
duced an increase in action potential duration and a decrease in AHP
amplitude (see text). The effects of a single conditioning impulse on
action potential duration (ordinate) are compared to its effects on AHP
amplitude (abcissa). A regression line is drawn through the data,
which are results of 8 intervals applied in various combinations to 7
different cells.


42
Discussion
The aim of the experiments presented here was to fully characterize
the kinetic properties of stimulation-induced changes in synaptic effi
cacy in the embryonic chick ciliary ganglion, and to investigate the
sensitivity of these processes to the divalent cations Sr2+ and Ba2+.
The results indicate that there are four components of stimulation-
induced increases in ganglionic efficacy described by time constants of
about 60 msec, 400 msec, 30 sec and 200 sec (Table 2-1). In several
synaptic preparations, accumulation models describing stimulus-induced
increases in transmitter release have been successful in describing
increases in transmitter release (e.g. Magleby and Zengel, 1975b; Mal-
lart and Martin, 1967; Younkin, 1974). One basic attribute of these
models is that each conditioning impulse increments the mechanisms
underlying each of the components of increased release. The results
presented in Figures 2-3 and 2-7 show that the four components observed
in the present study accumulate as predicted by models describing
facilitation, augmentation and potentiation (e.g. Magleby and Zengel,
1982; Zengel and Magleby, 1982). Further identification of the second
component of facilitation and of augmentation was achieved by exploit
ing the pharmacological sensitivities of these processes to certain
divalent cations. The addition of Ba2+ to the bathing solution caused
an increase in the magnitude of augmentation, and partial or complete
substitution of Sr2+ for Ca2+ in the bathing solution resulted in an
increase in the magnitude and the time constant of the second component
of facilitation. Because of the similarity in kinetic and pharmacolog
ical properties to those described previously for other synapses, I
believe that the processes I have described here are analogous to the


BIOGRAPHICAL SKETCH
Robert Eliot was born, a month early, to James David Poage and
Scottie Johnston Poage on August 6, 1964, in Winston-Salem, North Caro
lina. The second of three boys, Bob moved with the family to Baltimore
and then to Jacksonville and Orange Park, Florida, where he spent his
formative years. Dr. Jim Poage received his doctoral degree from the
University of Florida in counselor education, and has spent the better
part of his life working in rehabilitation of drug- and alcohol-
dependent persons. Scottie Poage-Hotchkiss has a successful practice
marriage counseling in Yuma, AZ.
His parents' careers interested young Mr. Poage in the field of
psychology, in which he pursued a bachelor's degree from the University
of Florida in 1987. Rather than continue in that vein, Bob decided
that understanding the underlying causes of mental illness and nervous
system dysfunction represented the future of psychology. He began
working in the field of neuroscience with primates, then for a while at
the cellular and biochemical level, before finally settling in for a
long stay as an electrophysiologist under the guidance of Janet Zengel.
Upon completing his dissertation, Bob will be working with Dr. Steve
Meriney at the University of Pittsburgh, where it is said to snow on
occasion. He currently lives in Gainesville, Florida, with his wife,
Heidi, his children, Shea Elizabeth and Scott Eliot, and pets too
numerous and ill-behaved to mention. In his spare time, he used to
like the outdoors, but now he just stares blankly at the TV, or the
wall, or his kids.
93


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
PhiTip Posnerv
Professor of Physiology
This dissertation was submitted to the Graduate Faculty of the
College of Medicine and to the Graduate School and was accepted as
partial fulfillment of the requirements for the degree of Doctor of
Philosophy.
August, 1995
K
lege of Medicine
Dean, Graduate School


32
CONTROL
60 sec
90 sec
120 sec
5 mV
20 msec
250 sec
Figure 2-6. Effect of an 800 impulse train (20/sec) on EPSP ampli
tude. Intracellular recording of EPSPs from a ciliary neuron shows the
response to orthodromic stimulation prior to and at the times indicated
following the conditioning train. [Ca2+] = 5 mM, [Mg2+] = 4 mM. The
time course of potentiation in this experiment is very similar to
results reported by Martin and Pilar (1964c, Figure 5).


48
ganglionic response has been shown to reflect changes in synaptic effi
cacy in this preparation (Landmesser and Pilar, 1972; Marwitt, Pilar
and Weakly, 1971; Poage and Zengel, 1993; Stanley and Goping, 1991).
Low frequency stimulation (0.1/sec or slower) was used to obtain con
trol values of postganglionic response amplitude and to test the
effects of drugs and divalent cations. Data are also presented from
experiments using more complex repetitive stimulation paradigms (see
Chapter 2). In these cases, sufficient time was allowed between trials
to allow the response to recover to pre-stimulus levels.
Results
The effects of Cd2+, Co2+, Ni2+ and Mg2+ on synaptic transmission
through the ciliary ganglion are shown in Figure 3-1. In separate
experiments, these ions were added to the bathing solution and their
effects on ganglionic transmission were observed. The addition of
these divalent cations led to a decrease in the amplitude of the chemi
cal component of the postganglionic compound action potential. Cd2+
was more than 2 orders of magnitude more potent in decreasing gan
glionic transmission than Co2+, Ni2+ and Mg2+ (Figure 3-1). This order
of potency of these divalent cations and the concentrations used to
impede ganglionic transmission are similar to those reported to block
Ca2+ currents in mouse neuromuscular junction (Penner and Dreyer,
1986), rat brain synaptosomes (Lentzner et al., 1992) and squid giant
synapse (Llinas et al., 1981) and synaptic transmission at the frog
neuromuscular junction (Zengel et al., 1993a). The effects of these
ions were reversible by washing with control Tyrode solution.
The effects of w-conotoxin GVIA were tested by adding small amounts
of stock solution to a static bath (see Methods). Application of 1 jjM


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
THE EFFECTS OF REPETITIVE STIMULATION ON
SYNAPTIC TRANSMISSION IN THE CHICK CILIARY GANGLION
By
Robert E. Poage
August 1995
Chair: Janet E. Zengel, Ph.D.
Major Department: Neuroscience
Under appropriate conditions, repetitive synaptic stimulation can
cause subsequent stimuli to release increased amounts of neurotrans
mitter. Repetitive stimulation can also change the shape of the pre-
synaptic action potential, but there are few experimental preparations
in which the relationship between these two phenomena may be studied
directly. The goals of this study were to use electrophysiological
recording techniques to investigate the role of Ca2+ channels in ini
tiation of release in the embryonic chick ciliary ganglion, to charac
terize stimulation-induced increases of synaptic efficacy at this
synapse, and to record electrical activity from the presynaptic element
of this synapse under conditions conducive to facilitation of release.
It is shown that treatments reported to block Ca2+ currents in
presynaptic terminals of this preparation cause concentration-dependent
blockade of synaptic transmission. These results are consistent with a
iv


70
to be a common phenomenon in neurons and other electrically excitable
cells (e.g. Bourque, 1991; Crest and Gola, 1993; Quattrocki et al.,
1994) and may represent a common mechanism for modulation of release
during and following repetitive stimulation.
The Action Potential
Before considering the implications of this study, a brief descrip
tion of the currents that comprise the action potential is in order.
Several overlapping currents comprise the action potential (Hodgkin and
Huxley, 1939). The depolarizing phase of the action potential results
from an increase in Na+ permeability and subsequent entry of Na+ ions
down a strong electrochemical gradient. The falling or repolarizing
phase of the action potential involves a decrease in Na+ permeability
(Na+ channel inactivation) and an increase in permeability to K+ due to
the opening of voltage-dependent K+ channels. The increase in K+
permeability can last for several milliseconds, so that in many cells
K+ efflux can hyperpolarize the membrane beyond the resting potential,
producing an afterhyperpolarization (AHP).
Repetitive Stimulation and the Action Potential
Martin and Pilar (1964c) looked at the effects of repetitive stimu
lation on the presynaptic action potential of the chick ciliary gan
glion. A paired pulse paradigm like the one used in the current study
was employed. The only effect they reported was a depression of action
potential amplitude at interstimulus intervals of 5 msec or less, (see
Figures 8 and 9, Martin and Pilar, 1964c); no measures of action poten
tial duration or AHP amplitude were reported. During stimulation at
higher frequencies than were applied here (50/sec for 20 sec), action
potential amplitude decreased (Figure 11, Martin and Pilar, 1964c).


24
circles in Figure 2-3C plot the estimated decay of the more rapidly
decaying component, which in this experiment had an initial magnitude
of 5.8 and a time constant of decay of 60 msec.
In 24 paired pulse experiments of this type, the average time con
stants describing the decay of V(t) were about 60 msec (n=14) and 400
msec (n=24; see Table 2-1). (In some experiments done under higher
quantal conditions (>1.5 mM Ca^+), there was an apparent depression of
ganglionic transmission at short interstimulus intervals, precluding
precise measures of the more rapidly decaying component.) These two
time constants of decay are very similar to the time constants pre
viously reported for the first and second components of facilitation at
the frog neuromuscular junction and the rabbit sympathetic ganglion
(Table 2-1).
Intracellular experiments were performed under the above conditions
to verify that facilitation of the ganglionic response results from an
increase in EPSP amplitude. As reported using extracellular recording,
paired pulse stimulation led to a facilitation of EPSP amplitude. Fig
ure 2-4 shows results from several ciliary neurons in two different
Ca^+ concentration ranges. The continuous lines describing the decay
of facilitation were drawn by taking the average values for the magni
tude and time constant of the two components of facilitation (using
results from extracellular data) and plotting their combined decay (see
figure legend). It appears that the estimates of facilitation provided
by extracellular records can provide a reliable estimate of the effects
of paired pulse stimulation on EPSP amplitude. These findings are
quite similar to the results of Martin and Pilar (1964c), who reported
a facilitation of EPSP amplitude in the chick ciliary ganglion that


58
second of a pair of EPSPs recorded from a postsynaptic ciliary ganglion
neuron (Figure 4-1, upper record; see Chapter 2 for description of
facilitation). Repetitive stimulation also caused an increase in pre-
synaptic action potential duration (control = 3.85 msec, test = 4.05
msec) and a decrease in AHP amplitude (control = 9 mV, test = 6.5 mV;
Figure 4-1, lower record). These changes are illustrated more clearly
in Figure 4-2 where the control and test responses from Figure 4-1
are superimposed.
During short trains of stimulation (10 impulses applied at 20/sec),
changes in the presynaptic action potential are apparent early in the
train and appear to reach a steady state (Figure 4-3). These changes
are more clearly seen in Figure 4-4, which plots averaged data from the
same cell as Figure 4-3. Repetitive stimulation caused an increase in
action potential duration (A) and decreases in both AHP amplitude (B)
and AHP half decay time (C). There were no consistent effects of
repetitive stimulation on action potential amplitude or time to peak
(data not shown).
Ionic Mechanisms Underlying Action Potential Repolarization and AHP
It is generally accepted that the repolarizing and afterhyperpolar
izing phases of the action potential are dependent on the activation of
K+ channels and a subsequent K+ efflux. To test the ionic basis of the
afterhyperpolarization, the membrane potential was altered by injecting
current through the recording electrode and the resulting changes in
AHP amplitude were observed. Figure 4-5 shows results from a single
cell in which the resting membrane potential was varied from -48 mV to
-95 mV. As the membrane potential approached the expected equilibrium
potential for K+ (-90 mV), the amplitude of the AHP was reduced to a


15
When intracellular recording was used, several changes were made to
minimize vibration and excess connective tissue that could foul intra
cellular electrodes. The connective tissue capsule that adheres closely
to the ganglion was removed with fine forceps (DuMont #5, Fine Science
Tools, Belmont, CA). All intracellular experiments were performed on a
Kinetic Systems 9101-11 vibration isolation table (Roslindale, MA) with
the perfusion apparatus and a dissecting microscope mounted on a free
standing Faraday cage. A micromanipulator with a motorized advance
attachment (460XYZ micromanipulator, 860 series motorizer, Newport, RI)
held an intracellular recording probe connected to the balanced bridge
input of a Dagan 8500 intracellulular amplifier (Minneapolis, MN). The
output of the intracellular amplifier was sent to 2 channels of a Tek
tronix oscilloscope for AC and DC recording. Most experiments were
also recorded onto VCR tape through a PCM recording adapter (A.R. Vet
ter Company, Los Angeles, CA). The basic sampling rate was 88.2 kHz
and the channel rise time was 50 psec with 14 bit A/D resolution.
Microelectrodes were pulled on a horizontal pipette puller
(Brown/Flaming P-87, Sutter Instrument Co., Los Angeles, CA) using
glass capillary tubing (items # 1B100F-4, TW100F-4; 0.54 or 0.75 mm
i.d., 1 mm o.d., World Precision Instruments, Sarasota, FL) and filled
with 3 M KC1 (tip resistances, 25-100 megaOhms). The microelectrode
was placed above the ganglion under visual control and the micromanipu
lator was used to advance the electrode into the ganglion proper.
Direct visualization of individual cells was not necessary. Microelec
trode penetration of a cell was achieved by capacitance "ringing" or by
gently tapping the micromanipulator and was evident as a voltage


30
Time (sec)
B
10
0.1
0.01
0.001
1/1 2
0 100 200 300 400 500
Time (sec)
0 100 200
Time (sec)
see


20
In order to characterize the time course of the increase in synap
tic efficacy following a single conditioning impulse, testing impulses
were applied at intervals ranging from 25 msec to 5 sec. The presenta
tion of intervals was randomized during an experiment. The symbols in
Figure 2-2 plot V(t), the fractional increase in post-ganglionic
response amplitude (Equation (2-1)), as a function of interstimulus
interval. Each symbol represents data from a single Ca^+ and Mg^+
concentration. As illustrated in Figure 2-2, facilitation of the chem
ically mediated component was greatest when release was reduced by
decreasing Ca^+ (A) or increasing Mg^+ (B). Under these conditions,
the stimulation-induced increase in response amplitude was greatest at
short conditioning-testing intervals, and decreased as the interstimu
lus interval was increased.
Results from another preparation are presented in Figure 2-3. The
filled circles in Figure 2-3A plot V(t) as a function of time following
a single conditioning impulse. When these data are plotted on a semi-
logarithmic scale (filled circles, Figure 2-3B), it is obvious that the
decay cannot be described by a single exponential. The data are well
described by a dual exponential. The contribution of the slower of the
two components was estimated by fitting a regression line to the linear
portion of the curve (see figure legend). In this experiment the ini
tial magnitude of this slower component, given by the intercept of the
regression line at time 0, was 0.99 and its time constant of decay was
565 msec. To obtain an estimate of the more rapidly decaying compo
nent, the data were corrected for the contribution of the slower compo
nent using the model described in Equation 2-2 to determine the rela
tive contribution of each component (see figure legend). The filled


16
deflection of -45 to -80 mV. Intracellular recordings were usually of
short duration, with impalements usually lasting less than 15 minutes.
Cells were identified by electrophysiological means as previously
published (Dryer and Chiappinelli, 1985; Martin and Pilar, 1963a,
1964a,b). By injecting hyperpolarizing current through the recording
electrode, it was possible to render both electrotonic and chemical
potentials subthreshold (Martin & Pilar, 1963a; see Figure 2-4 inset) so
that the underlying postsynaptic potentials could be observed. Mem
brane responses were monitored from the balanced bridge outputs of the
intracellular amplifier. The bridge balance was adjusted before
recording and was verified by testing the bridge balance after the
microelectrode was removed from a cell.
Data Collection and Analysis
For paired pulse and 5 impulse experiments, a Grass Instruments S48
stimulator was used to generate the conditioning and testing stimuli.
Responses were averaged and their amplitudes measured using either a
Nicolet 1170 signal averager (Nicolet Instruments Co., Madison, WI) or
386-based data acquisition and analysis software (Axotape, Axon Instru
ments, Foster City, Ca). For experiments in which longer conditioning
trains were applied, a MINC-11 computer (Digital Equipment Corp., Marl
boro, MA) was often used to generate the stimulation patterns, measure
and store the postganglionic response amplitudes, and analyze the data
(Magleby and Zengel, 1976; Zengel and Magleby, 1982). Sufficient time
was allowed between trains to ensure that release had returned to pre
conditioning levels (8 to 20 minutes, depending on train duration and
stimulation rate).


35
A B
Number of impulses Number of impulses
Figure 2-7. Effect of the number of conditioning impulses on the
magnitudes of potentiation (A) and augmentation (B). The magnitudes of
augmentation and potentiation were obtained as described for Figure
2-5. Results of 6 experiments in which the decay of V(t) was recorded
following conditioning stimulation at 20 impulses/sec. Lines connect
data from the same preparation.


87
Magleby, K.L., and Zengel, J.E. (1975a) A dual effect of repetitive
stimulation on post-synaptic potentiation of transmitter release at
the frog neuromuscular junction. J. Physiol. (London). 245:163-182.
Magleby, K.L., and Zengel, J.E. (1975b) A quantitative description of
tetanic and post-tetanic potentiation of transmitter release at the
frog neuromuscular junction. J. Physiol. (London! 245:183-208.
Magleby, K.L., and Zengel, J.E. (1976) Augmentation: a process that
acts to increase transmitter release at the frog neuromuscular
junction. J. Physiol. (London). 257:449-470.
Magleby, K.L., and Zengel, J.E. (1982) A quantitative description of
stimulation-induced changes in transmitter release at the frog
neuromuscular junction. J. Gen. Physiol.. 80:613-638.
Mallart, A. (1985) A calcium-activated potassium current in motor nerve
terminals of the mouse. J. Physiol.. 368:577-591.
Mallart, A., and Martin, A.R. (1967) An analysis of facilitation of
transmitter release at the neuromuscular junction of the frog. L
Physiol. (London). 193:679-694.
Marom, S., and Levitan, I.B. (1994) State-dependent inactivation of the
Kv3 potassium channel. Biophvs. J.. 67:579-589.
Martin, A.R., and Pilar, G. (1963a) Dual mode of synaptic transmission
in the avian ciliary ganglion. J. Physiol. (London). 168:443-463.
Martin, A.R., and Pilar, G. (1963b) Transmission through the ciliary
ganglion of the chick. J, Physiol. (London). 168:464-475.
Martin, A.R., and Pilar, G. (1964a) An analysis of electrical coupling
at synapses in the avian ciliary ganglion. J. Physiol. (London),
171:4545-475.
Martin, A.R., and Pilar, G. (1964b) Quantal components of the synaptic
potential in the ciliary ganglion of the chick. J. Physiol. (Lon
don). 175:1-16.
Martin, A.R., and Pilar, G. (1964c) Presynaptic and post-synaptic
events during post-tetanic potentiation and facilitation in the
avian ciliary ganglion. J. Physiol. (London). 175:17-30.
Marty, A. (1989) The physiological role of calcium-dependent channels
Trends Neurosci.. 12:420-424.
Marty, A., Tan, Y.P., and Trautmann, A. (1984) Three types of calcium-
dependent channels in rat lacrimal cells. J. Physiol. (London).
357:293-325.


13
the recording chamber using short lengths of very fine (0.1 mm
diameter) tungsten-iridium alloy wire (AlfaAESAR, Ward Hill, MA).
The recording chamber was held in place by small bits of clay and
surrounded by a plexiglass base to which perfusion apparati were
attached. The preparation was continuously perfused with an oxygenated
Tyrode solution (saline composition [in mM]: KC1 5; NaCl 150; CaCl2 1
to 5; MgCl2 2 to 12; glucose 10; HEPES 10; pH adjusted to 7.2-7.4) at a
rate of 1-2 ml/min (gravity driven). Fluid levels were kept constant
as saline was removed by suction through a bevelled hypodermic needle;
the level of the needle was adjusted to keep the preparation just below
the surface of the solution. In some experiments, Ba2+ (0.1-0.5 mM) or
Sr2+ (0.5-4.5 mM) was substituted for Ca2+ or added to the saline
solutions. In these experiments, the concentrations of Ca2+ and Ba2+
or Sr2+ were adjusted until the extracel 1 ularly-recorded postganglionic
response was approximately equal to the response in Ca2+-only Tyrode.
Salts for Tyrode solutions were purchased from Sigma Chemical (St.
Louis, MO). Solution changes were carried out between trials by chang
ing the source reservoir feeding the perfusion system. All experiments
were carried out at room temperature (20-23C).
Stimulating and Recording
Fluid suction electrodes (Dudel and Kuffler, 1961) mounted on
mechanical micromanipulators (Narishige, Japan) were used to draw up
the preganglionic (oculomotor) and postganglionic (ciliary) nerves.
These electrodes were made from PE-60 polyethylene tubing of 1.22 mm
outer diameter and 0.8 mm inner diameter (Becton, Dickinson & Co., Par-
sippancy, NJ). One end of the tubing was heated and pulled to an inner
diameter of 0.5 to 1.0 mm. The other end of the tube was connected to


62
Membrane potential (mV)
Figure 4-5. Effects of varying nerve terminal membrane potential on
afterhyperpolarization (AHP) amplitude. Data are from a single cell in
which the membrane potential was varied by injecting current through
the recording electrode. Line drawn through the data points by eye.
[Caz+] = 1.5 mM, [Mg2+] = 4 mM. Resting membrane potential = -52 mV.


75
Recent reports from this laboratory suggest that facilitation of
transmitter release is related to entry of Ca2+ through voltage-
activated channels (Zengel et al., 1993a,b), whereas augmentation
appears to be particularly sensitive to the intraterminal concentration
of Ca2+ (Zengel et al., 1994). Other mechanisms underlying stimula
tion-induced modulation of release (for example, changes in Ca2+ chan
nel kinetics [Lee, 1987], Ca2+ buffering [Llinas et al., 1992; Neher
and Augustine, 1990], phosphorylation of nerve terminal and vesicular
proteins [reviewed in Greengard et al., 1993]) may be contributing to
stimulation-induced increases in transmitter release in the ciliary
ganglion. Unfortunately, these processes are not readily observed by
electrophysiological means.
The "residual Ca2+" hypothesis (Katz and Miledi, 1968) proposes
that Ca2+ accumulates in the nerve terminal during repetitive stimula
tion, and that this cumulative increase in Ca2+, or some Ca2+-sensitive
process, contributes to stimulation-induced increases in release. The
findings reported here are consistent with a role for residual Ca2+ in
facilitation. Stimulation-induced broadening of the presynaptic action
potential should cause a significant increase in Ca2+ influx during
repetitive stimulation. The time course of stimulation-induced changes
in action potential duration and AHP amplitude indicate that, if these
changes are acting to bring about increases in neurotransmitter
release, they are most likely to be involved in facilitation, as
described in this and other preparations.
Notes
1 Because transmitter release has been shown to be a sensitive
function of action potential duration, describing the effects of repet-


47
Although Ca2+ channel classification schemes are useful in terms of
defining a point of reference (for comparison and discussion), there
appears to be such diversity in Ca2+ channel structure and function
that overlap between these subtypes (and ensuing subclassification) is
rendering these simple classifications insufficiently descriptive (as
discussed in recent reports: Bertolino and Llinas, 1992; Scott et al.,
1991; Stanley and Goping, 1991). It is, therefore, important to com
plement pharmacological classification studies with evaluations of the
functional properties of presynaptic Ca2+ channels. Experiments in
this chapter will address the role of presynaptic Ca2+ channels in
evoked transmitter release in the chick ciliary ganglion. To test the
contributions of different Ca2+ channel types to evoked transmitter
release, the pharmacological sensitivity of the release process to Ca2+
channel blocking agents is investigated.
Methods
Extracellular recording techniques were employed as described in
Chapter 2 (Methods), w-conotoxin (Bachem, Torrance, CA) was dissolved
in deionized H2O (stock concentration, 500 uM) and frozen in 30 jjI
aliquots (-20 C). Divalent cations were obtained as salts (Sigma, St.
Louis). All drugs were dissolved in Tyrode solution before being
applied. Due to high cost and limited availability, toxins were added
directly to small amounts of Tyrode and oxygenation was maintained by
bubbling O2 directly into the bath. Under control conditions, this
method of oxygenation had no effects on ganglionic transmission.
Orthodromic stimulation of the ciliary ganglion and subsequent
extracellular recording of the postganglionic compound action potential
were used to assay synaptic activity. The amplitude of the post-


61
A
B
C
Figure 4-4 Parameters describing nerve terminal action potentials
during 10 impulse trains. The average of three identical trials from a
single cell (see Figure 4-3) are presented. Parameters were measured
as described in Methods. Changes in successive action potentials are
expressed as the percent of values obtained for the first response in
the train. (A): action potential duration. (B): AHP amplitude. (C):
AHP half-decay time. [Ca2+] = 2.5 mM, [Mg2+] = 4 mM.


5
superior cervical ganglion (Zengel et al., 1980), in rat (Hubbard,
1963; Liley, 1956) and crayfish (Zucker, 1974) neuromuscular junctions,
in the squid giant synapse (Charlton and Bittner, 1978a), in cat spinal
cord (Curtis and Eccles, 1960; Kuno, 1964; Porter et al., 1970) and in
rat hippocampus (McNaughton, 1982).
I have characterized the kinetic and pharmacological properties of
stimulation-induced increases in synaptic efficacy in the chick ciliary
ganglion. These results, which are presented in Chapter 2, will show
that there are 4 components contributing to stimulation-induced
increases in synaptic efficacy and that these changes result from an
increase in chemical synaptic transmission.
Voltage-Dependent Calcium Channels and Transmitter Release
It has been proposed that accumulation of Ca2+ in the nerve termi
nal may be responsible for activity-dependent increases in neurotrans
mitter release (Katz and Miledi, 1968; Rosenthal, 1969; Weinreich,
1971). Results of experiments designed to test Katz and Miledi's
"residual Ca2+ hypothesis" support the involvement of Ca2+ as a
mediator of increased release, but suggest that Ca2+ must be acting at
several steps or sites in the release process to produce the observed
pattern of results (Landau et al., 1973; Zengel and Magleby, 1977,
1980; Zengel et al., 1993a,b, 1994). One element that will clearly
affect the concentration of Ca2+ present in the nerve terminal is Ca2+
influx through voltage-gated Ca2+ channels.
The ubiquitous role of Ca2+ in cellular function has led to a great
interest (and subsequently a large body of literature) in describing
and classifying voltage-activated Ca2+ channels according to their
kinetics and pharmacology. Neuronal Ca2+ channels are routinely


31
increased, and was occasionally depressed during and immediately
following tetanic stimulation (open circles in Figure 2-5A).
In 27 experiments of this type, the average values for the time
constants of decay of these two more slowly decaying phases were about
30 sec (n=17) and 200 sec (n=27). These time constants are similar to
those attributed to the processes of augmentation and potentiation,
respectively (Table 2-1).
In order to verify that the observed changes in postganglionic
response reflect a change in EPSP amplitude, intracellular recording
from ciliary neurons was used. Figure 2-6 shows the effect of 800 con
ditioning stimuli applied at 20/sec on the amplitude of the EPSP. The
amplitude of the control EPSP (before the conditioning stimulation) was
5 mV. Test pulses were applied at 10 second intervals following the
end of the train. Each test pulse applied less than 60 seconds after
the train produced a large EPSP that initiated an action potential (not
shown). EPSP amplitude declined over the next 6 minutes until the
response reached preconditioning levels.
In three experiments of this type, EPSP amplitude increased to a
maximum of 150% to 400% of control values during the test period. The
EPSP amplitude returned to a pre-conditioning level with a time con
stant of about 3 to 4 minutes, similar to that observed by Martin and
Pilar (1964c) under similar conditions. This time course of the decay
of EPSP amplitude following repetitive stimulation is also very similar
to the decay of the extracel1ularly recorded postganglionic response
(Figure 2-5; Table 2-1). The increases in EPSP amplitude reported by
Martin and Pilar were accompanied by an increase in the frequency, but


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E9WJO06Q2_WJGSF6 INGEST_TIME 2015-03-25T19:23:28Z PACKAGE AA00029763_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


17
Definition of Terms
Changes in response amplitude following conditioning simulation are
expressed as:
v(t) = (Vt/V0) 1 (2-1)
where V0 is the control (pre-conditioning) response amplitude and
is the amplitude of the response at time t following the conditioning
stimulation. For analysis of different components of stimulation-
induced changes in V(t), I have used the approach described by Zengel
and Magleby (1980, 1982). In brief, each component is defined as the
fractional change in response amplitude in the absence of other compo
nents. Since it is not always possible to measure one component in the
absence of others, the magnitudes and time constants of the individual
components are derived from the value of V(t) by assuming that these
components have distinct non-overlapping time constants of decay. The
slowest decaying component can be estimated by using data points col
lected after the more rapidly decaying components have decayed away.
Because the faster decaying components fall on top of the slower decay
ing ones, estimates of the contributions of these components can be
made only by assuming some relationship between the different compo
nents and using standard linear decay analyses. In this study I have
used a model shown to describe stimulation-induced changes in transmit
ter release at the frog neuromuscular junction (Magleby and Zengel,
1982; Zengel and Magleby, 1980, 1982). Basically, this model assumes
that there are four independent components of increased release which
interact according to the following equation:
Vt/V0 = (Fj + F2 + 1)(A + 1)(P + 1) (2-2)
where Fj and F2 are the first and second components of facilitation,


63
point where it became unmeasurable. This finding is consistent with
a role of K+ currents in generation of the AHP in the presynaptic
nerve terminal.
Due to the similar mechanisms underlying action potential repolar
ization and AHP generation, it is not unreasonable to assume that a
single underlying change in the presynaptic terminal is affecting both
AHP amplitude and action potential duration. Figure 4-6 plots the
relationship between changes in action potential duration and changes
in AHP amplitude following paired pulse stimulation. As AHP amplitude
decreases, there is a clear increase in action potential duration.
The Presynaptic Action Potential and Facilitation
An ideal way to describe the correlation between changes in the
nerve terminal action potential and increases in transmitter release
would be to record simultaneously from pairs of pre- and postsynaptic
cells during repetitive stimulation. While there has been a report of
successful simultaneous penetration of both elements in the chick cil
iary ganglion, this procedure yielded only a few very brief recordings
(Yawo and Momiyama, 1993). Instead, the relationship between intracel-
lularly recorded changes in the presynaptic action potential and effi
cacy of release was investigated in this study using extracellular
records of postsynaptic compound action potentials. It has been shown
that changes in the compound action potential during repetitive stimu
lation accurately reflect changes in EPSP amplitude under the condi
tions used here (see Chapter 2).
Figure 4-7 plots the relationship between changes in intracellu-
larly recorded presynaptic action potential parameters and facilitation
of the postganglionic compound action potential. Data were collected


11
Liley, 1956; Nussinovitch and Rahamimoff, 1988), cat spinal cord (Curtis
and Eccles, 1960; Kuno, 1964; Porter, 1970) and rat hippocampus
(McNaughton, 1982).
The striking similarities in kinetics and pharmacological sensiti
vities of these components of increased release in different prepara
tions indicate that they may represent general phenomena that occur at
all synapses. However, the subcellular machinery subserving these
modulations of transmitter release has not yet been identified. The
chick ciliary ganglion is an ideal synapse to study modulation of
transmitter release because stimulation-induced increases in transmit
ter release have been reported (Martin and Pilar, 1964c), although they
have not been fully characterized, and because the preparation is amen
able to a variety of experimental techniques and methodologies.
The single most unusual property of the chick ciliary ganglion is
the development of a large "calyx"-type nerve terminal that can be
impaled by a microelectrode, providing a rare opportunity to observe
presynaptic electrical activity associated with transmitter release. As
a prelude to beginning an investigation of the nerve terminal electri
cal events associated with transmitter release, I began by characteriz
ing the changes in synaptic efficacy that occur during and following
repetitive stimulation in the chick ciliary ganglion. Intracellular
recording from postsynaptic ciliary neurons was employed to verify that
increases in synaptic efficacy result from an increase in transmitter
release. I report here that four components of stimulation-induced
increases in transmitter release are present in the chick ciliary gan
glion. These components have kinetic and pharmacological properties


33
not the unit size, of spontaneous miniature EPSPs. This suggests that
the potentiation of the EPSP reported here is indeed due to an increase
in transmitter release, and not an effect on postsynaptic receptor
sensitivity (Martin and Pilar, 1964c).
The augmentation phase of increased release was difficult to
describe using intracellular recording. Two possible reasons for this
are: 1) the variability of EPSP amplitude under conditions of very low
quantal content may make observing the decay of augmentation, which is
described using only the first 5 to 10 test points (0 to 60 seconds
after conditioning), dependent on averaging large numbers of identical
trials, and 2) at higher quantal content, an apparent depression of
release immediately following tetanic stimulation may confound attempts
to observe effects of stimulation on increased EPSP amplitude during
the time when augmentation would be observed.
Effect of Number of Conditioning Impulses on the Components of
Increased Synaptic Efficacy
If the processes I have described arise from the same mechanisms
that produce stimulation-induced increases in release in other synaptic
preparations, then the growth and decay of the stimulation-induced
changes in synaptic efficacy in the chick ciliary ganglion should be
described by the "accumulation" models which have been shown to
describe release in other preparations (e.g. Magleby and Zengel,
1975a,b, 1982; Mallart and Martin, 1967; Younkin, 1974). According to
these models, each conditioning impulse adds an incremental increase to
each component of increased release. The components then decay with
their characteristic time constants between impulses. Increasing the
number of conditioning impulses would be expected to increase the mag
nitudes of the various components of increased ganglionic efficacy.


19
C
Figure 2-1. Facilitation of compound action potential amplitude.
The oculomotor nerve was stimulated with a pair of pulses applied at an
interstimulus interval of 50 msec and the postganglionic response was
recorded extracellularly (see MATERIALS AND METHODS). In this record,
each response consisted of 3 upward deflections: a shock artifact (*)
and electrically mediated (e) and chemically mediated (c) components of
the postganglionic response. The trace represents the average of 8
consecutive trials. The temporal separation between the electrical and
chemical components of each response is due to the synaptic delay pre
sent for chemical neurotransmission. Note the large increase in the
amplitude of the chemical component of the response to the second stim
ulus. This facilitation was typically observed under low quantal con
ditions. [Ca]2+ = 1.5 mM, [Mg+] *= 2 mM.


CHAPTER 4
EFFECT OF REPETITIVE STIMULATION ON THE PRESYNAPTIC ACTION POTENTIAL
When considering possible mechanisms for stimulation-induced
increases in release, there are several points in the excitation-
secretion cascade that appear to be obvious candidates. One of these
is the presynaptic action potential. If the depolarization of the
nerve terminal caused by the invasion of an action potential is larger
in amplitude or is prolonged, this could result in an increased
recruitment of voltage-dependent Ca2+ channels and a greater influx of
Ca2+, leading to increased release.
Due to their small size, presynaptic elements of most synapses are
difficult to study in an intact synapse. However, there are certain
synaptic preparations that have extraordinarily large presynaptic ter
minals. In several of these preparations the relationship between
action potential parameters and transmitter release has been investi
gated. In the squid giant synapse, increasing the duration of the pre
synaptic action potential by pharmacological means leads to an increase
in the amplitude of the postsynaptic response (Augustine, 1990). In
sensory neurons of Aplvsia, presynaptic facilitation of release appears
to be mediated by a decrease in a specific K+ current that prolongs the
action potential, leading to an increase in Ca2+ influx (Klein et al.,
1982; Sugita et al., 1992). These results suggest that changes in the
presynaptic action potential may represent a viable mechanism for regu
lation of synaptic efficacy (see Augustine, 1990).
54


2
terminal depolarization and release of transmitter is on the order of a
millisecond (Katz and Miledi, 1965). With very few exceptions the
nerve terminal is inaccessible to neurophysiological recording tech
niques, primarily because of the size of the presynaptic elements
involved. It would be of great interest to study many aspects of the
release process at a single synapse, but most experimental preparations
are not amenable to a wide range of available techniques.
The Ciliary Ganglion of the Embryonic Chick
The calyx nerve terminal in the ciliary ganglion of the embryonic
and posthatch chicken is one notable exception in as much as the struc
tures comprising the pre- and post-synaptic elements of the synapse are
large enough to be studied directly using standard electrophysiological
recording techniques. The accessibility of fertile eggs and the fact
that the cellular elements adapt well to cell culture enhance the value
of the preparation.
The ciliary ganglion is innervated by the third cranial nerve.
Preganglionic fibers originate in the approximately 2,000 cells of the
accessory motor nucleus, the avian analog of the mammalian Edinger-
Westphal nucleus. As the oculomotor nerve passes through the orbit, it
gives off branches to the muscles controlling eye movement. The
remaining fibers enter the ciliary ganglion, which is located behind
the eye lateral to the optic nerve. Fibers entering the ganglion form
two distinct types of synapses on two separate neuronal subpopulations.
The larger, more prevalent ciliary neurons receive single synaptic
inputs in the form of a "calyx" or cup-like nerve terminal (DeLorenzo,
1960). These nerve endings can cover 70% of the ciliary neuron's soma
(Hess, 1965). The axons of ciliary neurons exit in one or two ciliary


CHAPTER 3
CHARACTERIZATION OF CALCIUM CHANNELS
INVOLVED IN SYNAPTIC TRANSMISSION
The relationship between Ca2+ channel subtypes and transmitter
release is not well understood. In most synapses evoked release of
transmitter is dependent upon a coordinated influx of Ca2+ ions that
elevates intracellular Ca2+ at some site in the presynaptic nerve
terminal (Katz and Miledi, 1967). Voltage-activated Ca2+ conductances
are most often responsible for this rapid increase in intracellular
Ca2+. For this reason and because Ca2+ channels are potential targets
for modulating the release process (reviewed in Scott et al., 1991), it
is of interest to determine which channel type(s) are acting to initi
ate release in the chick ciliary ganglion.
There are clear indications for the existence of at least 4 general
classifications of neuronal Ca2+ channels. Subtypes N, L and T have
been described in chick dorsal root ganglion cells (Fox et al., 1987a;
Nowycky et al., 1985). These channels have been identified by their
sensitivities to different classes of pharmacological agents and by
their kinetics of activation and inactivation. L-type channels are
characterized by large unitary conductances (about 25 pS), activation
voltages positive to -10 mV and sensitivity to Ca2+ channel blockers
(dihydropyridines (BAY K 8644, nifedipine, nimodipine) and phenylalky-
lamines (verapamil)). T-type channels have small unitary conductances
(near 8 pS), are activated with weak depolarization (positive to -70
45


4
the relationship between the presynaptic action potential and stimula
tion-induced increases in postsynaptic response amplitude.
Stimulation-Induced Changes in Synaptic Efficacy
Since the nervous system most often uses trains of electrical sig
nals to convey information, one important element of the study of
neurotransmission is to observe what happens to transmitter release
during and following repetitive stimulation. Many studies have shown
that the efficacy of synaptic transmission is affected by its prior
activity (c.f. Feng, 1941), but the cellular machinery involved is
almost as obscure now as it was fifty years ago.
Repetitive stimuli applied to a presynaptic axon under conditions
of low levels of release can lead to a progressive increase in the
amount of transmitter released by successive impulses (reviewed by
Zucker, 1989). Following stimulation, this increase in release decays
back to control levels with a time course that can range from millise
conds to minutes. Such stimulation-induced increases in release have
been studied most extensively at the frog neuromuscular junction, where
four distinct components have been described on the basis of their
kinetic and pharmacological properties. These components are the first
and second components of facilitation, which decay back to control lev
els of release with time constants of about 60 ms and 400 ms, respec
tively (Magleby, 1973; Mallart and Martin, 1967; Zengel and Magleby,
1982); augmentation, which decays with a time constant of approximately
7 seconds (Erulkar and Rahamimoff, 1978; Magleby and Zengel, 1976); and
potentiation, which decays with a time constant of tens of seconds to
minutes (Magleby and Zengel, 1975a,b; Rosenthal, 1969). Some or all of
these components of increased release have been observed in the rabbit


Figure 2-5. Effect of trains of repetitive stimulation on synaptic
efficacy. A: Decay of V(t) as a function of time following a condi
tioning train of 800 impulses applied at 20/sec. Testing impulses were
applied at 2 sec intervals for 3 impulses, then every 10 sec. The
filled circles plot the fractional change in amplitude of the chemi
cally mediated peak of the compound action potential. The open circles
plot changes in the amplitude of the electrically mediated portion of
the compound action potential. Data averaged from 4 identical trials
from a single preparation. [Ca^+] =1.5 mM, [Mg^+] = 2 mM. Note that
the electrotonically mediated potential (open circles) is not increased
following repetitive stimulation. B: Semilogarithmic plot of the decay
of V(t) (same data as in A). The line represents the exponential decay
of potentiation, obtained by fitting a regression line through the data
points beyond 100 sec. C: Decay of augmentation, obtained after cor
recting for the contribution of potentiation using Equation 2-2. The
line represents the exponential decay of augmentation, derived by fit
ting a regression line through the data points.


78
In the 1960s, Bernard Katz and others showed that initiation of
transmitter release is dependent upon the presence of Ca2+ ions in the
extracellular medium (Katz and Miledi, 1965, 1967). These studies and
others led to the formulation of the "Ca2+ hypothesis", which proposed
that the entry of Ca2+ ions into the presynaptic terminal is an essen
tial step linking membrane depolarization to transmitter release.
Despite the wide acceptance of the Ca2+ hypothesis, recent reports
have suggested that other factors may be sufficient to initiate exocy-
tosis in the absence of extracellular Ca2+. For example, Stuenkel and
Nordmann (1993) report Na+-dependent neuropeptide release from the rat
neurohypophysis in the absence of a rise in intracellular Ca2+. Pamas
and co-workers have investigated the role of nerve terminal depolariza
tion in initiation of transmitter release at the frog neuromuscular
junction and have suggested that a depolarization-dependent factor pro
motes release in cooperation with intracellular Ca2+ (e.g. Hochner et
al., 1989; Pamas et al., 1986). Other investigators have also
reported depolarization-induced transmitter release that appears to
occur in the absence of Ca2+ influx (Mosier and Zengel, 1993; Sil insky
et al., 1995). These reports suggest that although intracellular Ca2+
is necessary for most forms of exocytosis, it is not the sole factor
acting to initiate transmitter release.
As described in previous chapters, four components of stimulation-
induced increases in synaptic transmitter release have been reported at
many synapses (e.g. Erulkar and Rahamimoff, 1978; Mallart and Martin,
1967; Rosenthal, 1969; Zengel and Magleby, 1982). All four of these
processes (two components of facilitation, augmentation and potentia
tion) are acting to increase EPSP amplitude during repetitive stimula-


CHAPTER 2
EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC EFFICACY
IN THE CHICK CILIARY GANGLION
For more than fifty years it has been known that synaptic efficacy
changes as a result of prior synaptic activity (for example, Feng,
1941). It has been well documented in a variety of synapses that these
changes arise from a change in the amount of neurotransmitter released
by each successive impulse during repetitive stimulation. At the frog
neuromuscular junction, where stimulation-induced changes in transmit
ter release have been studied most extensively, four components of
increased release have been identified on the basis of their kinetic
and pharmacological properties. These components are: the first and
second components of facilitation, which decay back to control levels
of release with time constants of approximately 60 and 400 msec
(Magleby, 1973; Mallart and Martin, 1967; Younkin, 1974; Zengel and
Magleby, 1982); augmentation, which decays with a time constant of
approximately 7 sec (Erulkar and Rahamimoff, 1978; Magleby and Zengel,
1976); and potentiation, which decays with a time constant of tens of
seconds to minutes (Magleby and Zengel, 1975a,b; Rosenthal, 1969).
Some or all of these components have been observed in the rabbit supe
rior cervical ganglion (Zengel et al., 1980), squid giant synapse
(Charlton and Bittner, 1978a), crayfish (Bittner and Baxter, 1991;
Zucker, 1974) and rat neuromuscular junction (Hubbard, 1963;
10


21
B
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
200 400 600 800 1000
Time (msec)
12 mM Mg
200 400 600 800 1000
Time (msec)
Figure 2-2. Effects of reducing extracellular Ca2+ or increasing
extracellular Mg2+ on facilitation of compound action potential ampli
tude. A: Plot of the decay of V(t), the fractional change in response
amplitude (Equation 2-1), as a function of time following a single con
ditioning impulse. The [Mg2+] was held constant at 2 mM while [Ca2+]
was varied from 1 to 3 mM. Conditioning-testing trials were applied
once every 20 sec. Data points represent the average of 8 to 16 iden
tical trials from a single preparation. B: In another preparation,
plot of the decay of V(t) as above, only with [Ca2+] held constant at 3
mM and [Mg2+] varied from 2 to 12 mM. Data points represent the aver
age of 8 to 24 identical trials.


69
Interstimulus interval (msec)
Figure 4-9. Effect of a single conditioning impulse on presynaptic
action potential AHP amplitude. Data points represent averages of 4 to
10 identical trials. (A): Effects of paired pulse stimulation on AHP
amplitude. Data are expressed as percent of the control AHP amplitude.
(B): Same data as (A) expressed as percent maximal inhibition of AHP
amplitude and plotted on a semilogarithmic scale. A regression line
drawn through the linear portion of the data (points beyond 150 msec)
gives a line described by the equation Ts-|ow. The values of the slow
regression line at earlier points (<150 msec) were calculated and the
contributions of the slower regression were subtracted assuming an
additive relationship between the two decays. A regression line
through the resulting points (squares) is described by the equation
Tfast- tCa2+] = 1-2 mM, [Mg2+] = 4 mM.


28
decayed with a time constant on the order of 65 msec. It was also
reported that the observed increase in EPSP amplitude could be
accounted for entirely by an increase in quantal content. The ionic
conditions and stimulation paradigms used in this earlier study are
nearly identical to the conditions of my own experiments. Therefore,
it seems reasonable to assume that the increases in synaptic efficacy
described here using paired pulse stimulation result from an increase
in quantal transmitter release.
To observe more slowly decaying increases in ganglionic response
amplitude, ganglia were conditioned using longer trains of stimuli
(200-1200 impulses at 10-50/sec). The filled circles in Figure 2-5A
plot the decay of V(t) of the chemically mediated component of the
action potential following a conditioning train of 800 impulses applied
at 20/sec. When the data are plotted on a semilogarithmic scale
against time following the end of the conditioning stimulation (filled
circles in Figure 2-5B), there appear to be two components of decay.
The contribution of the slower component was estimated by fitting a
regression line to the linear portion of the curve at times beyond 100
sec (see figure legend). In this experiment, the initial magnitude of
the slower component was 0.32 and its time constant of decay was 200
sec. Figure 2-5C plots the decay of the faster component, obtained by
correcting for the contribution of the slower component (see figure
legend). This component had an initial magnitude of 1.02 and a time
constant of decay of 12 sec. In experiments in which it was possible
to accurately measure the amplitude of both components of the postgan
glionic response, the amplitude of the electrically mediated peak never


CHAPTER 1
INTRODUCTION
Chemical Synaptic Transmission
Although the release of chemical neurotransmitter substances
mediates many forms of neuronal communication, the cellular mechanisms
underlying neurotransmitter release have yet to be identified. It has
been established that depolarization of the presynaptic nerve terminal
causes the activation of voltage-dependent Ca^+ channels and influx of
Ca^+ ions (Katz and Miledi, 1967). The resulting increase in Ca^+
concentration acts through an unknown mechanism to initiate release of
neurotransmitter. It is generally accepted that the resulting increase
in Ca^+ concentration causes synaptic vesicles within the nerve termi
nal to fuse with the nerve terminal membrane and spill their contents
(neurotransmitters) into the synaptic cleft via exocytosis. Diffusion
of the neurotransmitter across the cleft and its binding to specific
postsynaptic receptors results in the transmission of information,
either in the form of an increased ionic permeability or through the
action of second messenger systems. To understand neurotransmitter
release and its integral role in information processing and plasticity,
it will first be necessary to understand how the release process is
initiated and how it is modulated.
Several aspects of synaptic transmission make it difficult to
study. Exocytosis is an extremely rapid event. The time between nerve
l


Figure 2-3. Effect of 1 and 5 conditioning impulses on synaptic
efficacy. A: Plot of the decay of V(t) as a function of time follow
ing a single conditioning impulse (filled circles) and a train of 5
conditioning impulses applied at 20/sec (open circles). Single testing
impulses were applied at intervals of 25 to 5000 msec after the condi
tioning stimulation. Conditioning-testing trials were applied about
once every 30 sec. Data points represent the average of 16 trials from
a single preparation. [Ca2+]? = 1 mM, [Mg2+]g = 2 mM. B: Semilogarith-
mic plot of the decay of V(t) following one (filled circles) and 5
(open circles) conditioning impulses (same data as A). The lines rep
resent the exponential decay of the second component of facilitation
(F2) derived by fitting regression lines through the data points
between 200 and 2000 msec. C: Decay of the first component of facili
tation (Ft) following one (filled circles) and five conditioning
impulses (open circles). Values of Fj were obtained by subtracting off
from V(t) the contribution of the second component of facilitation
(Equation 2-2 in Methods). The lines represent the exponential decay
of Fj, derived by fitting a regression line through the data points.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Janet E. Zengel, C
Associate Professor
ir
of
Neuroscience
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Peter A.V. Anderson
Professor of Neuroscience
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
kjo -Met
Thomas W. Vickroy
Associate Professor of
oscience
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Stuart E. Dryer
Associate Professor of Biological Science
Florida State University


59
4 msec
10 mV
test
Figure 4-2. Nerve terminal action potentials evoked by paired pulse
stimulation. Responses are superimposed to illustrate changes in repo
larization phase. Same cell and interstimulus interval as Figure 4-1.
Single responses are shown (not averaged).


18
A is augmentation and P is potentiation. As reported in this chapter,
this model appeared to describe stimulation-induced changes in neuro
transmitter release in the chick ciliary ganglion.
Statistical Analysis
Control and experimental trials were recorded in each experiment.
The effect of the experimental treatment was compared to the response
of the same preparation under control conditions using t-test proce
dures. Statistical analysis was performed using the IBM PC version of
SigmaPlot 5.0 (Jandel Scientific). Averaged data are presented as mean
+ standard error.
Results
Description of Stimulation-Induced Increases in Synaptic Efficacy
A paired pulse paradigm was employed to examine the effects of a
single conditioning stimulus on the efficacy of ganglionic transmis
sion. Figure 2-1 presents averaged extracellular data from a single
preparation that was stimulated with pairs of impulses at an interstim
ulus interval of 50 msec. The chemically mediated component, which is a
function of the number of cells brought to threshold by chemical trans
mitter release, is increased following a single conditioning impulse
while the shock artifact and electrotonically mediated component of the
response were unaffected. Under conditions of low levels of release
many of the postsynaptic cells are below threshold for action potential
generation and are not contributing to either peak of the postgan
glionic response. Increases in transmitter release will bring some
previously unresponsive cells above threshold. These cells contribute
to the observed increase in compound action potential amplitude in
Figure 2-1.


Figure 2-8. Effect of Sr2+ on facilitation. A: Plot of the decay
of V(t) as a function of time following a single conditioning impulse
in 1.5 mM Ca2+ Tyrode (filled circles) and in Tyrode containing 1.0 mM
Ca2+ and 1.5 mM Sr2+ (open circles). Data points represent the average
of 32 trials from a single preparation. B: Semilogarithmic plot of the
decay of V(t) (same data as in A). The lines, obtained by fitting
regression lines through the data points between 300 and 2000 msec,
represent the exponential decay of the second component of facilitation
(F2). C: Decay of the first component of facilitation (Fj), obtained
after correcting for F2 as described earlier.


73
There are also reports that describe cumulative stimulation-induced
inactivation of K+ currents. In molluscan neurons, cumulative inacti
vation of a fast delayed rectifier K+ current contributes to stimula
tion-induced action potential broadening (Aldrich et al., 1979; Thomp
son, 1977). A delayed rectifier channel cloned from rat brain (Marom
and Levitan, 1994) also shows similar inactivation properties. Edry-
Schiller and Rahamimoff (1993), on the basis of data obtained from the
fused Torpedo synaptosome preparation, have proposed a "potassium
inactivation hypothesis" for frequency modulation of transmitter
release. They suggest that reactivation of slow K+ channels following
an action potential may be responsible for action potential broadening
and facilitation of release. Mallart (1985) proposed a similar role
for the inactivating Ca2+-activated K+ channel in the motor nerve
terminal of the mouse. In the chick ciliary ganglion, cumulative K+
current inactivation has been observed in the postsynaptic ciliary
neuron (S. Dryer, personal communication) and it is reasonable to
assume that similar K+ channels could be present in the presynaptic
nerve terminal.
Action Potential Duration: Effects on Release
Although the mechanisms underlying stimulation-induced increases in
action potential duration in the chick ciliary ganglion are not fully
understood, the fact remains that many studies have shown that changes
in presynaptic action potential duration can have large effects on syn
aptic transmitter release (e.g. Augustine, 1990; Hochner et al., 1986;
Robitailie and Charlton, 1992). In the squid giant synapse, applica
tion of the K+ channel blocker diaminopyridine had concentration-
dependent effects on action potential duration and transmitter release:


6
compared to the N-, L- and T-type channels described in chick dorsal
root ganglion cells (Fox et al., 1987a; Nowycky et al., 1985). T-type
channels produce transient membrane currents and have relatively low
conductances. L-type channels are noninactivating and produce long-
lasting currents. N-type channels are neither transient nor slowly
inactivating. A more recently described class of channels that is
activated by moderate depolarization appears to be most common in mam
malian neurons and has been designated "P-type" (Llinas et al., 1989).
In the chick ciliary ganglion, Ca2+ currents recorded under voltage
clamp from the calyx fail to meet the criteria for a single class of
Ca2+ channel, although they are most similar to the N-type group (Stan
ley, 1991; Stanley and Atrakchi, 1990). Calical Ca2+ currents are
insensitive to blockade by dihydropyridines and are blocked by w-cono-
toxin, consistent with an N-type classification, but they inactivate
slowly, if at all (Stanley and Goping, 1991). Stanley (1991) has
called these channels "Npj-"type (for neuronal-presynaptic terminal).
There is evidence that Ca2+ channels at other fast-transmitting syn
apses may have similar kinetic and pharmacological properties (Miller,
1987; Suszkiw et al., 1986; Yoshikami et al., 1989). These results,
taken with the finding that Ca2+ channels are located, possibly in
clusters, on the release face of the calical nerve terminal membrane
(Stanley, 1991), imply that the calical Ca2+ currents described above
may be acting in the release process. However, Ca2+ channels involved
in release at this synapse have not been fully characterized.
I have investigated the effects of Ca2+ channel blockers on initia
tion of evoked transmitter release in the chick ciliary ganglion. The
results of this study, presented in Chapter 3, contribute to a more


Figure 2-9. Effect of Ba2+ on potentiation and augmentation. A:
Plot of the decay of V(t) following 800-impulse conditioning trains in
the absence (filled circles) and presence of Ba2+ (open circles).
Control, conditioning and testing impulses as in Figure 2-5. Data
averaged from 8 trials from a single preparation. B: Semilogarithmic
plot of the decay of V(t) (same data as in A). The lines, obtained by
fitting regression lines through the data points beyond 100 sec, repre
sent the exponential decay of potentiation. C: Decay of augmentation,
obtained after correcting for potentiation as described earlier.


56
coupling potential, representing current flow through electrical syn
apses (Martin and Pilar, 1963a).
In paired pulse experiments, trials consisted of a conditioning
stimulus and a testing stimulus applied some time after the condition
ing stimulus. The conditioning stimulus serves as a control to which
test responses are compared. Typical responses to paired stimuli from
a ciliary neuron and calyx are presented in Figure 4-1. In experiments
in which trains of impulses were applied, changes in the response are
described relative to the first response of the train.
The parameters were defined as follows: action potential amplitude
was measured from resting membrane potential to the most depolarized
point of the action potential; time to action potential peak was meas
ured from the time of the last available baseline control point (the
last point at resting membrane potential before the stimulus artifact)
to the time of the action potential peak; action potential duration was
measured from the time of the last baseline point to the point where
the repolarizing action potential crossed the original resting poten
tial; afterhyperpolarization (AHP) amplitude was measured from the
resting membrane potential (prior to the action potential) to the most
hyperpolarized point during the AHP; AHP half-decay time was measured
by finding the peak hyperpolarization and determining the time for the
voltage to decay to half of that level.
Results
Effects of Repetitive Stimulation on Presvnaptic Potentials
As shown in Figure 4-1, both presynaptic and postsynaptic responses
were affected by paired pulse stimulation. At low levels of release,
paired pulse stimulation led to an increase in the amplitude of the


86
Landau, E.M., Smolinsky, A., and Lass, Y. (1973) Post-tetanic potentia
tion and facilitation do not share a common calcium-dependent mech
anism. Nature New Biology. 244:155-157.
Landmesser, L. and Pilar, G. (1972) The onset and development of trans
mission in the chick ciliary ganglion. J. Physiol. (London).
222:691-713.
Lee, K.S. (1987) Potentiation of the calcium-channel currents of
internally perfused mammalian heart cells by repetitive depolari
zation. PNAS (USA). 84:3941-3945.
Lentzner, A., Bykov, V., and Bartschat, D.K. (1992) Time-resolved
changes in intracellular calcium following depolarization of rat
brain synaptosomes. J. Physiol. (London). 450:613-628.
Lev-Tov, A., and Rahamimoff, R. (1980) A study of tetanic and post-
tetanic potentiation of miniature end-plate potentials at the frog
neuromuscular junction. J. Physiol. (London). 309:247-273.
Liley, A.W. (1956) The quantal components of the mammalian end-plate
potential. J. Physiol. (London). 133:571-587.
Liley, A.W., and North, K.A.K. (1953) An electrical investigation of
effects of repetitive stimulation on mammalian neuromuscular junc
tion. J^-NeunjfihYii, 16:509-527.
Lledo, P.-M., Vernier, P., Vincent, J.-D., Mason, W.T., and Zorec, R.
(1993) Inhibition of Rab3B expression attenuates calcium-dependent
exocytosis in rat anterior pituitary cells. Nature. 364:540-544.
Llinas, R., Steinberg, I.Z., and Walton, K. (1981) Presynaptic calcium
currents in squid giant synapse. Biophvs. J. 33:289-321.
Llinas, R., Sugimori, M., Lin, J.W., and Cherksey, B. (1989) Blocking
and isolation of a calcium channel from neurons in mammals and
cephalopods utilizing a toxin fraction (FTX) from funnel web spider
poison. PNAS (USA). 86:1689-1693.
Llinas, R., Sugimori, M., and Silver, R.B. (1992) Microdomains of high
calcium concentration in a presynaptic terminal. Science.
256:677-679.
Loewi, 0. (1921) Uber humorale Ubertragbarkeit der Herznervenwirkung.
Pflugers Archiv. 189:239-242.
Magleby, K.L. (1973) The effect of repetitive stimulation on facilita
tion of transmitter release at the frog neuromuscular junction. jL
Physiol. (London). 234:327-352.


74
a ten percent increase in action potential duration resulted in a 60
percent increase in the amplitude of the postsynaptic current
(Augustine, 1990). Hochner and colleagues (1986) report similar
results in AdIvsia sensory ganglia: a 10% to 30% increase in presynap-
tic action potential duration was correlated with a 25% to 120%
increase in postsynaptic potential amplitude. A similar relationship
between presynaptic action potential duration and increased synaptic
efficacy can be seen in the chick ciliary ganglion (Figure 4-7).
Current Hypotheses: Stimulation-Induced Increases in Release
Some of the earliest theories concerning mechanisms of stimulation-
induced increases in transmitter release involved the presynaptic
action potential (Bloedel et al., 1966; Hubbard and Schmidt, 1963;
Miledi and Slater, 1966; Takeuchi and Takeuchi, 1962). Experimental
testing of this hypothesis has shown that changes in the presynaptic
action potential cannot account for all of the observed increases in
transmitter release during and following repetitive stimulation (Charl
ton and Bittner, 1978b; Zucker, 1974). However, these results should
not be interpreted as eliminating the possibility that changes in the
action potential play a role in mediating the effects of repetitive
stimulation on transmitter release. Several individual components,
which appear to arise from separate mechanisms, contribute to stimula
tion-induced increases in release. It has been suggested that differ
ent Ca^+-dependent processes are involved in initiation and modulation
of release (Bain and Quastel, 1992b; Zengel et al., 1993a,b). The
effects of repetitive stimulation on action potential duration probably
only represent one of many effects of repetitive stimulation.


ACKNOWLEDGEMENTS
I would like to thank the many educators who have contributed to my
continued pursuit of higher education, most notably my advisor, Dr.
Janet Zengel. I have been very fortunate to have studied under and
learned from a scientist and teacher who is as dedicated and gifted as
any I have met. When I look back on years spent in graduate study I
will always be grateful for the opportunities I was given and the
patience that was always applied liberally. From my mentor I have
gathered many of the critical skills a scientist must have to be suc
cessful. From my friend I have seen how a scientist can pursue her
life's work with enthusiasm, integrity and a critical eye, turned as
closely inward as outward.
My graduate work at the University of Florida has been marked by
some excellent instruction, and opportunities to teach and to present
my work in front of peers. In retrospect, I am grateful for all of
these things.
My supervisory committee members, Dr. Janet Zengel, Dr. Stuart
Dryer, Dr. Peter Anderson, Dr. Philip Posner and Dr. Tom Vickroy, have
my gratitude for having confidence in my work when successes were slow
in coming. I thank them for their many suggestions and their support.
Finally, I would like to thank my parents, who always supported my
me, and my continuing education, and got me started in the field, and
my wife and family, who, whether they know it or not, are beautiful.


71
Several factors may have contributed to differences between the
earlier results of Martin and Pilar and those presented here. When
Martin and Pilar performed their classic experiments describing elec
trical and chemical transmission through the ciliary ganglion, they
compared single action potentials. The results presented here are
taken from averaged data, which should make small changes in action
potential waveform more easily apparent. Also, in the 1960s, current
theories about the role of the action potential in facilitation seem to
have focussed primarily on changes in action potential amplitude (e.g.
Hubbard and Schmidt, 1963; Takeuchi and Takeuchi, 1962). For this rea
son, perhaps, the effects of repetitive stimulation on action potential
amplitude were more vigorously examined than other aspects of the
action potential.
In recent studies of a number of neuronal preparations, repetitive
stimulation has been shown to have effects on the action potential
waveform (reviewed in Nicholls et al., 1992). For example, Charlton and
Bittner (1978b) reported that repetitive stimulation of the squid giant
synapse can lead to an increase in the amplitude of the presynaptic
action potential, but only during the first few responses in a train.
The effects of repetitive stimulation on the duration of the action
potential were not reported, although their data appear to show a
decrease in AHP amplitude during tetanic stimulation (Figures 3 and 11,
Charlton and Bittner, 1978b). Similarly, at the crayfish neuromuscular
junction (Bittner and Baxter, 1991) it has been shown that repetitive
stimulation causes an increase in the duration and amplitude of the
presynaptic action potential for the first 3 or 4 impulses in a train.


81
involved in synaptic vesicle trafficking, mobilization and exocytosis.
Studies have already begun to describe the role of these proteins in
release of neurotransmitter under control conditions (e.g. Lledo et
al., 1993; Pevsner et al., 1994) and following repetitive stimulation
(e.g. Tarelli et al., 1992). More recently, the use of antisense
nucleotides and transgenic mice has enabled researchers to study synap
tic transmission in synaptic preparations that have modified versions
of these proteins (e.g. Alder et al., 1992) or lack them entirely (e.g.
Rosahl et al., 1993).
The presynaptic nerve terminal is a highly specialized and complex
structure. It contains mechanisms for uptake, storage and synthesis of
transmitter substances, as well as voltage-sensitive proteins that can
alter ionic conditions in response to changes in membrane potential,
and ligand-activated complexes that can respond to cytoplasmic or
extracellular chemical signals. Consequently, there are many ways by
which modulation of the release process could occur, including
increases in intracellular Ca^+ and intracellular Na+, activation of
presynaptic ligand gated receptors, changes in the kinetics of voltage-
activated ion channels and changes in the presynaptic action potential.
Recent technological advances in many elements of neuroscience
research (cell culture, molecular biological, genetic and imaging tech
niques) are increasing the precision with which neurons can be studied.
With this enhanced ability to observe and describe neuronal physiology,
many long-standing questions will soon be answered. It would appear
that the Ca^+-dependent "trigger" mechanism that initiates the process
of neurotransmitter release will be characterized within half a century
of its postulation.


41
8 r
7 -o
6 -
5
4
3
2 -
1 -
O

o
2 +
1 mM Ca
O a 0.75 mM Cq2+/0.15 mM Ba2+
0 100 200 300 400 500
Time (see)
Time (see)
Time (see)


76
itive stimulation on this parameter was a primary goal of this study.
Unfortunately, due in part to the length of the preganglionic nerve and
the rapid conduction of the action potential in this preparation, the
early rising phase of the presynaptic action potential was often tem
porally superimposed on the stimulus artifact (see Figure 4-2), making
precise measurement of action potential duration difficult. This tech
nical problem may have added an element of variability to measures of
action potential duration, even though changes were always expressed
relative to a paired control response. Due to the rapid time course of
the stimulus artifact, measures of AHP amplitude were not affected.


7
complete understanding of the Ca2+ channels present in the calyx and
the role these channels play in the initiation and regulation of
transmitter release.
The Presvnaptic Action Potential and Transmitter Release
The role of nerve terminal depolarization in triggering exocytosis
has been thought to be primarily through the activation of voltage-
activated Ca2+ channels and the subsequent rapid increase in local
intracellular Ca2+, although a direct role for depolarization in ini
tiating release has been proposed (Dudel et al., 1983; Hochner et al.,
1989; Sil insky et al., 1995). It stands to reason that the ability to
study the electrical activity of the presynaptic element of a synapse
is essential if the process of transmitter release is to be
well understood.
Since stimulation-induced changes in release are a general phenome
non seen at most synapses, it is of interest to discern the mechanism
or mechanisms acting to produce these effects on the release process.
In attempting to describe these underlying mechanisms, many investiga
tors have examined the role of the presynaptic action potential. If
the depolarization of the nerve terminal caused by the invasion of an
action potential is larger in amplitude or is prolonged (Hubbard and
Schmidt, 1963; Liley and North, 1953; Takeuchi and Takeuchi, 1962), the
resulting increase in activation of voltage-dependent Ca2+ channels
should lead to a larger influx of Ca2+ and increased release (e.g.
Augustine, 1990; Hochner et al., 1986).
In the ciliary ganglion, Martin and Pilar (1964c) found no gross
changes in the presynaptic action potential or resting membrane poten
tial under conditions conducive to what I will describe as facilitation


50
w-conotoxin (Figure 3-2, circles) led to an irreversible decrease in
the amplitude of the chemically mediated portion of the compound action
potential. Higher concentrations of w-conotoxin acted more rapidly
(2 juM, triangles in Figure 3-2). Concentrations of 1 juM or greater
usually led to a complete block of the chemical component of the post
ganglionic response within 90 minutes (4 of 5 experiments).
Application of 5-10 juM verapamil, a phenyl alkyl amine Ca2+ channel
blocker, caused a very small decrease in the amplitude of the compound
action potential (4% to 7%) that was reversed by washing with control
Tyrode (n = 2 experiments). While it is possible that this effect is
due to a direct effect of verapamil on Ca2+ currents, it has been
reported that higher concentrations of verapamil (20 to 50 pM) can
block voltage-activated Na+ currents (Chang et al., 1988). Such an
effect would be expected to affect the amplitude of the compound
action potential.
In 2 experiments, 10 pM nifedipine, a dihydropyridine Ca2+ channel
blocker, had no effect on transmission through the ciliary ganglion.
Discussion
The results presented in this chapter are consistent with other
studies (e.g. Bennett and Ho, 1991; Stanley, 1989; Yawo and Momiyama,
1993) in which it was suggested that the "N-like" voltage-activated
Ca2+ currents described in the dissociated calyx preparation are
responsible for initiation of transmitter release in the chick ciliary
ganglion. Concentrations of w-conotoxin shown to have a maximal effect
on presynaptic Ca2+ currents (Stanley and Atrachki, 1991) blocked
chemically mediated synaptic transmission through the ciliary ganglion.
Similarly, Cd2+ blocked synaptic transmission at concentrations that


12
similar to the components of increased transmitter release described
for other preparations.
Methods
Preparation and Solutions
Fertile White Leghorn chicken eggs (Poultry Science Unit, Univer
sity of Florida) were set in a forced draft rotating incubator (Peter-
sime model 1, Gettysberg, OH) kept at 37C, 70% humidity and candled on
days 4 to 10 to determine viability. Embryos were removed at embryonic
day 15-19 (stage 41-45) and sacrificed via decapitation. These ages
were chosen to coincide with maturation of the large "calyx" type syn
apse, before the synapse becomes primarily electrical in nature (Land-
messer & Pilar, 1972). The ciliary ganglia were dissected out under
intermittent washing with Tyrode solution (see below for composition).
Several dissection techniques were used. The most common approach was
to bisect the head and free 2 to 5 mm of the oculomotor nerve proximal
to the orbit. A lateral approach was then used to draw the eye aside,
liberate the ganglion from surrounding connective tissue, and dissect
free the ciliary nerve (3-10 mm) from both sides of the sclera.
A recording chamber was constructed entirely of Sylgard polymer
(Dow Corning, Midland, MI) poured into a small (about 6 cm diameter)
Petri dish. Two chambers of approximately equal volume (1.5 ml) were
connected by a 1 cm long passage through which solutions passed during
perfusion. Removing the bathing solution from a chamber physically
separate from the recording chamber minimized noise from surface vibra
tion. A ciliary ganglion, complete with the preganglionic (oculomotor)
and postganglionic (ciliary) nerve trunks, was pinned to the bottom of


38
2.0
1.5
>
1.0
0.5
0.0
1.5 mM Ca2+
O 1 mM Ca2+/ 1.5 mM Sr2+
0 500 1000 1500 2000
Time (msec)
0 500 1000 1500 2000 0 500
Time (msec) Time (msec)


89
Penner, R., and Dreyer, F. (1986) Two different presynaptic calcium
currents in mouse motor nerve terminals. Pfluo. Arch.. 406:190-197.
Pevsner, J., Hsu, S.-c., Braun, J., Calakos, N., Ting, A.E., Bennett,
M.K., and Scheller, R.H. (1994) Specificity and regulation of a
synaptic vesicle docking complex. Neuron. 13:353-361.
Poage, R.E., and Zengel, J.E. (1993) Kinetic and pharmacological exami
nation of stimulation-induced increases in synaptic efficacy in the
chick ciliary ganglion. Synapse. 14:81-89.
Porter, R. (1970) Early facilitation at corticomotoneuronal synapses.
J. Physiol. (London). 207:733-745.
Quattrocki, E.A., Marshall, J., and Kaczmarek, L.K. (1994) A Shab
potassium channel contributes to action potential broadening in
peptidergic neurons. Neuron. 12:73-86.
Razani-Boroujerdi, S., and Partridge, L.D. (1993) Activation and modu
lation of calcium-activated non-selective cation channels from
embryonic chick sensory neurons. Brain Res.. 623:195-200.
Regehr, W.G., and Mintz, I.M. (1994) Participation of multiple calcium
channels types in transmission at single climbing fiber to Purkinje
cell synapses. Neuron. 12:605-613.
Robitaille, R., and Charlton, M. (1992) Presynaptic calcium signals and
transmitter release are modulated by calcium activated potassium
channels. J. Neurosci.. 12:297-305.
Rosahl, T.W., Geppert, M., Spillane, D., Herz, J., Hammer, R.E., Mal-
enka, R.C., and Sudhof, T.C. (1993) Short-term synaptic plasticity
is altered in mice lacking synapsin 1. Cell. 75:661-670.
Rosenthal, J. (1969) Post-tetanic potentiation at the neuromuscular
junction of the frog. J. Physiol (London). 203:121-133.
Scheller, M.K., and Bennett, M.K. (1994) Molecular correlates of synap
tic vesicle docking and fusion. Curr. Qpin. Neurobiol.. 4:324-9.
Scott, R.H., Pearson, H.A., and Dolphin, A.C. (1991) Aspects of verte
brate neuronal voltage-activated calcium currents and their regula
tion. Prog. Neurobiol. 36:485-520.
Silinsky, E.M., Watanabe, M., Redman, R.S., Qui, R., Hirsh, J.K., Hunt,
J.M., Solsona, C.S., Alford, S., and MacDonald, R.C. (1995) Neuro
transmitter release evoked by nerve impulses without calcium entry
through calcium channels in frog motor nerve endings. J. Phvsiol.
(London). 482: 511-520.
Smith, S.J., and Augustine, G.J. (1988) Calcium ions, active zones and
synaptic transmitter release. TINS. 11:458-464.


14
a syringe, which was used to draw the nerve and bathing solution into
the tapered end. A silver wire (0.005 0.01 inch diameter) was
inserted through the wall of the tubing and placed within approximately
5 mm of the tapered end of the tubing. A second silver wire was
wrapped around each electrode shaft to within 5 mm of the tip to serve
as a ground electrode.
Short stimulus pulses (0.01-0.06 msec) were applied to the oculomo
tor nerve through a photoelectric stimulus isolation unit (Grass
Instruments, Quincy, MA) and the stimulus amplitude was adjusted until
it was clearly suprathreshold. The postganglionic responses were
amplified with a Grass P-5 series pre-amplifier and displayed on a Tek
tronix 5113 dual beam storage oscilloscope (Beaverton, OR). In most
experiments the response consisted of both an electrical and a chemical
component (see Figure 2-1), although in some ganglia from younger
embryos there was not a distinct peak for the electrotonically mediated
component. The amplitude of the chemically mediated component is a
function of the number of postsynaptic cells brought to threshold by
chemical neurotransmission (Martin and Pilar, 1963a). Thus, changes in
the amplitude of the chemically mediated component of the ganglionic
response represent changes in the number of postsynaptic cells acti
vated by orthodromic stimulation (Landmesser and Pilar, 1972; Poage and
Zengel, 1993). The main advantage of extracellular recording is the
fact that the postganglionic response represents the averaged activity
of the entire ganglion. Averaged ganglionic responses show consider
ably less variability than intracellular responses, thus decreasing the
number of trials needed to obtain a reliable result.


60
Figure 4-3. Effect of a short train of orthodromic stimuli on the
presynaptic action potential. Each stimulus generates an action poten
tial in the presynaptic nerve terminal. Stimulation rate for this
example is 20/sec. [Caz+] = 2.5 mM, [Mg2+] = 4 mM.


44
question remains: What subcellular mechanisms are conserved at the
synapse that produce these processes at different synapses in a variety
of species?
Since quantal content is known to be affected by manipulations of
Ca2+ buffering and entry of Ca2+ into the nerve terminal, speculations
on the mechanism of stimulation-induced increases in neurotransmitter
release focus on the role of intracellular Ca2+ in transmitter release
(e.g. Charlton et al., 1982; Katz and Miledi, 1967, 1968; Zengel et
al., 1993a,b). No single theory has been successful in accounting for
all of the observed stimulation-induced changes in synaptic transmitter
release. Proposed mechanisms include an increased entry of Ca2+ or an
accumulation of Ca2+ in the presynaptic nerve terminal (Erulkar and
Rahamimoff, 1978; Katz and Miledi, 1968; Miledi and Thies, 1971; Rosen
thal, 1969; Weinreich, 1971) and nerve terminal voltage changes or pro
cesses associated with these voltage changes (e.g. Martin and Pilar,
1964c; Bittner and Baxter, 1991). The ciliary ganglion offers a versa
tile system in which to study these possibilities through the use of
many techniques, including presynaptic intracellular recording (Dryer
and Chiappinelli, 1983; Martin and Pilar, 1964c; Yawo, 1990), patch
clamp recording of Ca2+ currents (Stanley, 1989), and Ca2+-imaging
using fluorescent dyes. Chapter 4 will describe the effects of repeti
tive stimulation on presynaptic potentials under conditions conducive
to facilitation of transmitter release.


80
duration of the presynaptic depolarization recruits a greater percent
age of voltage-dependent Ca2+ channels in the calyciform nerve terminal
and that the resulting increase in Ca2+ influx underlies facilitation
of transmitter release. The time constant of activation for Ca2+
currents in the calyx is 0.9 1.6 msec at 23 C (Stanley and Goping,
1991) and calcium currents recorded using whole-cell patch-clamp tech
niques do not reach a peak level for several milliseconds following
sustained depolarization (Stanley, 1989). The calyx action potential
only produces a depolarization sufficient to activate Ca2+ channels for
a very short time (1 to 2 msec, Martin and Pilar, 1963a; personal
observations), which would not activate the full complement of presyn
aptic Ca2+ channels. This conclusion is not without precedent. Earlier
studies have also suggested that increased Ca2+ influx by consecutive
stimuli might cause facilitation (e.g. Stinnakre and Tauc, 1973).
Several other areas of study bear mentioning when discussing the
effects of repetitive stimulation on neurotransmitter release. It has
been reported in the frog and crayfish neuromuscular junctions and
other preparations that the relative contribution of the various compo
nents of stimulation-induced changes in release may vary across cells
within an experimental preparation (Collins et al., 1984; Fadoga and
Brookhart, 1962; Frank, 1973; Meriney and Grinnell, 1991). One use of
these systems would be to identify the biochemical or electrophysiolog-
ical differences between, for example, facilitating and non
facilitating synapses.
Another promising area of research on exocytosis involves recently
discovered proteins in the presynaptic nerve terminal (reviewed in Ben
nett and Scheller, 1993 and in De Camilli, 1993) that appear to be


68
potential duration was increased and AHP amplitude was dramatically
decreased, whereas at longer intervals the effects on AHP amplitude and
action potential duration were less pronounced. The time course of the
effects on AHP amplitude are more clearly seen in Figure 4-9. Like
facilitation of transmitter release, the effects of a single condition
ing impulse on AHP amplitude were greatest at short interstimulus
intervals (less than 150 msec) and less pronounced at longer intervals
(Figure 4-9A). Figure 4-9B plots the same data as in (A), expressed as
percent inhibition of AHP amplitude. In three experiments where 5 or
more intervals were tested, the decay of the effect on AHP amplitude,
plotted as percent AHP inhibition, could be described by a dual expo
nential decay with time constants of 64+20 msec and 1219+292 msec. In
experiments where fewer than 5 intervals were applied, the shortest
interstimulus intervals (<200 msec) consistently produced the greatest
effects on the presynaptic action potential, similar to the results
presented in Figure 4-9.
Discussion
The goal of this study was to record from the presynaptic nerve
terminal of the chick ciliary ganglion during repetitive stimulation
and to investigate the relationship between the presynaptic action
potential and transmitter release. It is shown that repetitive stimu
lation, under conditions that produce facilitation of transmitter
release, gives rise to changes in the presynaptic action potential that
parallel facilitation. This effect is greatest during the first few
hundred milliseconds following a single conditioning stimulus, a time
when the facilitation phase of increased release is prevalent. Stimu
lation-induced changes in presynaptic action potential waveform appear


role for these previously described Ca^+ currents in initiation of
transmitter release in the chick ciliary ganglion.
It is shown that four distinct components contribute to increased
ganglionic efficacy following repetitive stimulation. Their time con
stants of decay (60 milliseconds, 400 milliseconds, 30 seconds and 200
seconds) are similar to those that describe the decay of the four com
ponents of stimulation-induced increases in transmitter release
described in other preparations (first and second components of facili
tation, augmentation and potentiation). The components described here
are also similar to the above-mentioned processes in their sensitivi
ties to Sr^+ and Ba^+. It is concluded that the components of stimu
lus-induced increases in release that have been described in other syn
aptic preparations are present in the chick ciliary ganglion.
Repetitive stimulation under conditions conducive to facilitation
of transmitter release causes an increase in the duration of the pre-
synaptic action potential and a decrease in the amplitude of its after
hyperpolarization (AHP). The time course of the effects of repetitive
stimulation on the presynaptic action potential parallels the time
course of facilitation in this preparation. It appears likely that the
observed effects on action potential duration result from a decrease in
K+ conductance.
Results are discussed in terms of possible mechanisms underlying
the individual components of stimulation-induced increases of release.
v


34
To test this, experiments were conducted to examine the effects of
conditioning impulse number on the two components of facilitation, aug
mentation, and potentiation. The open circles in Figure 2-3A plot the
decay of V(t) following a 5 impulse conditioning train. At all time
points tested, the increase in V(t) was much greater following the 5
impulse train (open circles) than following a single conditioning
impulse (filled circles). This increase in V(t) could be attributed to
an increase in the magnitudes of the second (lines through open circles
in Figure 2-3B) and first components of facilitation (lines through
open circles in Figure 2-3C). There was little or no effect of condi
tioning impulse number on the time constants of decay of the two compo
nents of facilitation. Similar results were observed in 2 additional
experiments in which 1, 2 and 5 impulse conditioning trains were
applied. Thus, as with the rabbit sympathetic ganglion (Zengel et al.,
1980) and frog neuromuscular junction (Zengel and Magleby, 1982),
increasing the number of conditioning impulses results in an increase
in the magnitudes of the two components of facilitation, while having
little effect on their time constants of decay.
I also examined the effect of conditioning impulse number on the
magnitudes of the slower decaying components. Figure 2-7 summarizes
the results of 6 experiments in which conditioning trains of various
duration were applied. In each of these experiments I observed an
increase in the magnitude of potentiation when the number of condition
ing impulses was increased (Figure 2-7A). In the 3 experiments in
which I was able to obtain reliable estimates of augmentation, a simi
lar effect was observed (Figure 2-7B). In some experiments, particu
larly with higher Ca2+ concentrations (> 3 mM) or longer stimulation


9
I also found that there are changes in the presynaptic action
potential during repetitive stimulation that have a similar time course
to facilitation, but a definitive role for changes in the presynaptic
action potential was not demonstrated. Possible mechanisms that could
underlie stimulation-induced changes in the action potential are dis
cussed.


57
Post-synaptic ciliary neuron
Pre-synaptic nerve terminal
Figure 4-1. Examples of electrophysiological responses of pre- and
postsynaptic cells of the embryonic ciliary ganglion to paired pulse
stimulation. A postsynaptic ciliary neuron is identified by the pres
ence of an excitatory postsynaptic potential (EPSP) in response to
orthodromic stimulation (upper record). Hyperpolarizing current was
injected through the recording electrode to keep the EPSP below thresh
old for action potential generation. [Ca^+] = 2.5 mM, [Mg^+J = 4 mM. A
presynaptic nerve terminal responds to orthodromic stimulation with an
action potential (lower record). [Ca^+J = 3 mM, [Mg^+j = 2 mM.


55
Martin and Pilar (1964a-c) employed intracellular recording from
the pre- and postsynaptic cells of the chick ciliary ganglion to inves
tigate mechanisms of stimulation-induced increases in release. They
reported no significant change in the presynaptic nerve terminal action
potential under conditions where increased release was observed. How
ever, those studies compared the amplitudes of individual action poten
tials, and it is possible that subtle changes in the action potential
amplitude and/or duration would not have been detected by this type of
analysis. It was therefore of interest to perform experiments of a more
quantitative nature to investigate the relationship between changes in
the presynaptic action potential and the individual processes of stimu
lation-induced increases in release described an Chapter 2.
Methods
Standard intracellular recording techniques were employed
(described in detail in Chapter 2, Methods). Cells were identified as
calyciform nerve endings or ciliary neurons using the electrophysiolog-
ical criteria of Martin and Pilar (1963a, 1964a). Briefly, stimulation
of the oculomotor nerve produced an action potential in both calyces
and ciliary neurons. When spike initiation was blocked by injection of
hyperpolarizing current, the ciliary neurons exhibited an underlying
excitatory postsynaptic potential (EPSP) (Figure 4-1, upper record).
Nerve terminals showed no EPSP following orthodromic stimulation when
hyperpolarizing current was injected. The only remaining response was
an attenuated action potential, which was presumably conducted electro-
tonically from outside the region blocked by hyperpolarization (Martin
and Pilar, 1963a). Calyces responded to antidromic stimulation with a


85
Fox, A.P., Nowycky, M.C. and Tsien, R.W. (1987b) Single-channel record
ings of three types of calcium channels on chick sensory neurons.
J. Physiol. (London). 394:173-200.
Frank. E. (1973) Matching of facilitation at the neuromuscular junction
of the lobster. J. Physiol. (London!. 233:635-658.
Greengard, P., Valtorta, F., Czernik, A.J., and Benfenati, F. (1993)
Synaptic vesicle phosphoproteins and regulation of synaptic func
tion. Science. 259:780-785.
Hess, A. (1965) Developmental changes in the structure of the synapse
on the myelinated cell bodies of the chicken ciliary ganglion. jL.
Cell Biol.. 25:1-19.
Hochner, B., Klein, M., Schacher, S., and Kandel, E.R. (1986) Action
potential duration and the modulation of transmitter release from
the sensory neurons of Aplysia in presynaptic facilitation and
behavioral sensitization. PNAS (USA). 83:8410-8414.
Hochner, B., Pamas, H., and Pamas, I. (1989) Membrane depolarization
evokes neurotransmitter release in the absence of calcium entry.
Nature. 342:433-435.
Hodgkin, A.L., and Huxley, A.F. (1939) Action potentials recorded from
inside a nerve fibre. Nature 144:710-711.
Hubbard, J.I. (1963) Repetitive stimulation at the neuromuscular junc
tion, and the mobilization of transmitter. J. Physiol. (London),
169:641-662.
Hubbard, J.I., and Schmidt, R.F. (1963) An electrophysiological inves
tigation of mammalian motor nerve terminals J. Physiol. (London),
166:145-165.
Katz, B., and Miledi, R. (1965) The measurement of synaptic delay, and
the time course of acetylcholine release at the neuromuscular
junction. Proc. Rov. Acad. B161:483-495
Katz, B., and Miledi, R. (1967) The timing of calcium action during
neuromuscular facilitation. J. Physiol. (London). 189:535-544.
Katz, B., and Miledi, R. (1968) The role of calcium in neuromuscular
facilitation. J. Physiol. (London). 195:481-92.
Klein, M., Camardo, J., and Kandel E.R. (1982) Serotonin modulates a
specific potassium current in sensory neurons that show presynaptic
facilitation in Aplysia. PNAS (USA). 79: 5713-5717.
Kuno, M. (1964) Mechanism of facilitation and depression of the excit
atory synaptic potential in spinal motoneurones. J, Physiol. (Lon
don). 175:100-112.


91
Vergara, C., and Latorra, R. (1983) Kinetics of Ca2+-activated K+
channels from rabbit muscle incorporated into planar bilayers.
Evidence for Ca2+ and Ba2+ blockade. J. Gen. Physiol.. 82:543-568.
Weinreich, D. (1971) Ionic mechanism of post-tetanic potentiation at
the neuromuscular junction of the frog. J. Physiol. (London!
212:431-446.
Yawo, H. (1990) Voltage-activated calcium currents in presynaptic nerve
terminals of the chicken ciliary ganglion. J. Physiol. (London).
428:199-213.
Yawo, H., and Chuhma, N. (1994) Omega-conotoxin-sensitive and
-resistant transmitter release from the chick ciliary presynaptic
terminal. J. Physiol. (London). 477:437-448.
Yawo, H., and Momiyama, A. (1993) Re-evaluation of calcium currents in
pre-and postsynaptic neurones of the chick ciliary ganglion. vL
Physiol. (London). 460:153-172.
Yoshikami, D., Bagobaldo, Z., and Olivera, B.M. (1989) The inhibitory
effects of omega conotoxin on calcium channels and synapses. Ann.
NY Acad. Sci.. 560:230-248.
Younkin, S.G. (1974) An analysis of the role of calcium in facilitation
at the frog neuromuscular junction. J. Physiol. (London). 237:1-14.
Zengel, J.E., Lee, D.T., Sosa, M.A. and Mosier, D.R. (1993a) Effects of
calcium channel blockers on stimulation-induced increases in trans
mitter release at the frog neuromuscular junction. Synapse,
15:251-262.
Zengel, J.E., and Magleby, K.L. (1977) Transmitter release during
repetitive stimulation: selective changes produced by Sr2+ and
Ba2+. Science, 197:67-69.
Zengel, J.E., and Magleby, K.L. (1980) Differential effects of Ba, Sr,
and Ca on stimulation-induced changes in transmitter release at the
frog neuromuscular junction. J. Gen. Physiol.. 76:175-211.
Zengel, J.E., and Magleby, K.L. (1981) Changes in miniature end-plate
potential frequency during repetitive stimulation in the presence
of Ca2+, Ba2+, and Sr2+ at the frog neuromuscular junction. jL
Gen. Physiol.. 77:503-529.
Zengel, J.E., and Magleby, K.L. (1982) Augmentation and facilitation of
transmitter release: a quantitative description at the frog neuro
muscular junction. J. Gen. Physiol.. 80:583-611.


CHAPTER 5
CONCLUSIONS
As recently as 100 years ago, energetic debate surrounded hypoth
eses concerning the functional organization of the nervous system. The
proponents of cell theory suggested that nerves were independent units,
a proposal that went against the prevalent theory that all nerves were
a single continuous structure, part of a syncytium, interconnected by
protoplasmic bridges (see Nicholls et al., 1992, p. 185). Later, it
was accepted that nerves were discontinuous, separated by small gaps
across which information was passed through unknown methods. Otto
Loewi, in 1921, performed a simple and convincing series of experiments
showing that stimulation of the vagus nerve acts to slow heart rate by
releasing a diffusible substance (acetylcholine). Dale and others soon
established the role of acetylcholine as a neurotransmitter at the
neuromuscular junction and in autonomic ganglia (e.g. Dale and Feld-
berg, 1936). These experiments led to the general acceptance of chemi
cal synaptic transmission between nerve cells and effector cells.
77


52
have been shown to completely block presynaptic Ca2+ currents in the
calyx (Stanley and Goping, 1991; Yawo and Momiyama, 1993). Application
of L-type Ca2+ channel blockers (verapamil and nifedipine) had little
or no effect on ganglionic transmission. These results suggest that
Ca2+ channels involved in release in this preparation are most similar
to N-type channels. Furthermore, it is reported here that the divalent
cations Cd2+, Co2+ and Ni2+ blocked ganglionic transmission with the
same order of potency as described in other synaptic preparations in
which N-type channels are involved in initiation of transmitter release
(Lentzner et al., 1992; Llinas et al., 1981; Penner and Dreyer, 1986).
The majority of the total presynaptic Ca2+ current in the calyci-
form nerve terminal is w-conotoxin sensitive (Stanley and Atrakchi,
1991; Yawo and Momiyama, 1993). An w-conotoxin insensitive component
to the calyx l£a has been reported (Stanley and Atrakchi, 1990; Yawo
and Momiyama, 1993). This component can support low levels of trans
mitter release when 4-aminopyridine is added to increase the presynap
tic action potential duration (Yawo and Chuhma, 1994). This w-cono
toxin insensitive component is blocked by 50 uM Cd2+. The role of such
a small Ca2+ current in release could not be assayed with the extracel
lular recording method used in this study. Under the conditions used
here, little if any transmitter release is supported by this w-cono
toxin insensitive current. The chemically mediated portion of the com
pound action potential is not present unless EPSP amplitude is large
enough to initiate an action potential in the postsynaptic cells in
the ganglion.


51
Figure 3-2. Time course of the effects of 2 concentrations of
w-conotoxin on ganglionic transmission. Either 1 uM (circles) or 2 jjM
w-conotoxin (triangles) was added to the Tyrode bathing solution at
time = 0. Each point represents the averaged amplitude of 4 to 16
consecutive responses and is normalized to the amplitude of the
response before toxin addition. The block of ganglionic transmission
was not reversed by 70 to 160 minutes of perfusion with toxin-free
Tyrode. [Ca2+] = 5 mM, [Mg2+] = 2 mM.


3
nerves and pierce the sclera to innervate the striated ciliary muscles
and constrictor muscles of the iris. The smaller choroid neurons are
innervated by multiple bouton-type synapses and project through three
to five choroid nerves to innervate the choroidal coat.
Synaptic contacts on both choroid and ciliary neurons are chemical
and cholinergic in nature. Calyx/ciliary neuron synapses also display
electrical coupling. The nature and size of the calical nerve terminal
and synapse provide a unique experimental opportunity to observe elec
trical activity in the nerve terminal of a vertebrate neuronal synapse.
Martin and Pilar (1963a) reported the unique nature of this synapse and
successfully recorded electrical activity in both pre-and postsynaptic
structures. In a series of elegant reports (Martin and Pilar, 1963a,b,
1964a-c), they described transmission in the calyx preparation, includ
ing properties of both electrical and chemical coupling. Since those
early descriptions, the chick ciliary ganglion has been used by several
laboratories to study neurotransmitter release and synaptic function
(Bennett and Ho, 1991; Dryer and Chiappinelli, 1985; Stanley, 1989;
Stanley and Goping, 1991; Yawo, 1990).
The chick ciliary ganglion is obviously a preparation of great
value to the neuroscientist. Due to the accessibility of the presynap-
tic element of the calyciform synapse in the chick ciliary ganglion, I
have chosen to use this preparation to study the relationship between
presynaptic electrical activity and transmitter release.
This dissertation describes: 1) stimulation-induced increases in
synaptic efficacy at the chick ciliary ganglion, 2) the effects of Ca^+
channel blocking agents on ganglionic efficacy at this synapse and 3)


72
Mechanisms for Action Potential Broadening During Repetitive Stimulation
It is not clear what electrical events bring about the increases in
action potential duration reported in this chapter. One possible mech
anism involves activation of Ca2+-dependent cation channels (Partridge
and Swandulla, 1988). Although these channels have been described in
chick sensory neurons, their activation range (over 1 pM; Razani-
Boroujerdi and Partridge, 1993) raises doubts as to their role in neur
onal function under more normal ionic conditions. A more likely mecha
nism for broadening the presynaptic spike is a stimulation-induced
inhibition of K+ current(s).
Figure 4-7 shows that the reversal potential for the calyx AHP lies
close to the predicted equilibrium potential for K+. In the ciliary
ganglion, several K+ currents have been observed in the presynaptic
nerve ending, including delayed rectifier (Dryer and Chiappinelli,
1985), Ca2+-activated (Fletcher and Chiappinelli, 1992b, 1993) and
inwardly rectifying K+ currents (Dryer and Chiappinelli, 1985; Fletcher
and Chiappinelli, 1992a).
There are several mechanisms through which repetitive stimulation
could affect K+ currents. One possibility is that the accumulation of
[Ca2+]i during stimulation (e.g. Charlton et al., 1982; Smith and
Augustine, 1988; Stinnakre and Tauc, 1973) could act to decrease K+
efflux. There have been reports of Ca2+-inhibited K+ channels in many
cell types (see Marty, 1989). For example, increases in intracellular
Ca2+ can decrease the opening probability of K+ channels in skeletal
muscle (Vergara and Latorra, 1983), and can inhibit BK K+ currents in
rat exocrine cells (Marty et al., 1984).


46
mV) and are rapidly inactivated at negative holding potentials. N-type
channels show an intermediate voltage activation (-40 to -30 mV),
intermediate unitary conductances (about 13 pS) and sensitivity to
w-conotoxin GVIA. These calcium channels may also be characterized by
their sensitivity to inorganic cations. Cadmium (Cd2+) is by far the
most potent, blocking L- and N-type channels with a K about ten times more is needed to block T-type channels (Fox et al,
1987a,b). The more recently described P-type channel (named to honor
the Purkinje cell in which it was first described) shows little inacti
vation, has a voltage-dependence between those of N- and L-type chan
nels and is not blocked by w-conotoxin or by dihydropyridines, but is
blocked by a component of funnel-web spider venom, FTX or w-agatoxin
IVA (LIinas et al., 1989).
It has been shown that N-, L- and P-type channels can all contrib
ute to rapid Ca2+ influx associated with evoked transmitter release
(see Scott et al., 1991). It has also been demonstrated that multiple
types of Ca2+ channels can coexist in a single neuron (reviewed in
Miller, 1987), and that more than one channel type can be involved in
initiation of exocytosis from a single cell type (Artalejo et al.,
1994; Regehr and Mintz, 1994; Takahashi and Momiyama, 1993).
In the presynaptic calyciform nerve terminal of the chick ciliary
ganglion, several investigators have reported the presence of "N-like"
currents in the calical nerve terminal, yet no evidence for L- or
T-type currents (Stanley, 1991; Stanley & Atrakchi, 1990; Stanley and
Goping, 1991; Yawo and Momiyama, 1993). These investigators have sug
gested that the Ca2+ currents they described are acting to initiate
transmitter release in the chick ciliary ganglion.


90
Stanley, E.F. (1989) Calcium currents in a vertebrate presynaptic nerve
terminal: the chick ciliary ganglion calyx. Brain Res.,
505:341-345.
Stanley, E.F. (1991) Single calcium channels on a cholinergic presynap
tic nerve terminal. Neuron. 7:585-591.
Stanley, E.F. (1993) Single calcium channels and acetylcholine release
at a presynaptic nerve terminal. Neuron. 11: 1007-1011.
Stanley, E.F. and Atrakchi, A.H. (1990) Calcium currents recorded from
a vertebrate presynaptic nerve terminal are resistant to the dihy
dropyridine nifedipine. PNAS (USA). 87:9683-9687.
Stanley, E.F. and Goping, G. (1991) Characterization of a calcium cur
rent in a vertebrate cholinergic presynaptic nerve terminal, jh
Neurosci.. 11:985-993.
Stinnakre, J., and Tauc, L. (1973) Calcium influx in active Aplysia
neurons detected by injected aequorin. Nature. 293:115-117.
Stuenkel, E.L., and Nordmann, J.J. (1993) Sodium-evoked, calcium-
independent vasopressin release from rat isolated neurohypophysial
nerve endings. J. Physiol. (London). 468:357-78.
Sugita, S., Goldsmith, J.R., Baxter, D.A., amd Byrne, J.H. (1992)
Involvement of protein kinase C in serotonin-induced spike broaden
ing and synaptic facilitation in sensorimotor connections of Aply
sia. J. Neurophvsiol.. 68:643-651.
Suszkiw, J.B., O'Leary, M.E., Murawsky, M.M., and Wang, T. (1986) Pre
synaptic calcium channels in rat cortical synaptosomes: fast kinet
ics of phasic calcium influx, channel inactivation, and relation
ship to nitrendipine receptors. J. Neurosci.. 6:1349-1357.
Takahashi, T., and Momiyama, A. (1993) Different types of calcium chan
nels mediate central synaptic transmission. Nature 366:156-158.
Takeuchi, A., and Takeuchi, N. (1962) Electrical changes in pre- and
post-synaptic axons of the giant synapse of Loligo. J. Gen. Phy
siol.. 45:1181-1193.
Tarelli, F.T., Bossi, M., Fesce, R., Greengard, P., and Valtorta, F.
(1992) Synapsin 1 partially dissociates from synaptic vesicles dur
ing exocytosis induced by electrical stimulation. Neuron. 9:
1143-1153.
Thompson, S.H. (1977) Three pharmacologically distinct potassium chan
nels in molluscan neurones. J. Phvsiol. (London). 265:465-488.


49
[divalent cation] mM
Figure 3-1. Concentration-dependence of the effects of divalent
cations on postganglionic response amplitude. Values are expressed as
percent control compound action potential amplitude, with the exception
of the Mg2+ data, which is compared to the lowest [Mg2+] applied (2
mM). Points with standard error bars indicate data averaged from 2 to
4 experiments. [Ca2+] = 5 mM, [Mg2+] = 2 mM unless otherwise noted.


66
from 7 cells at interstimulus intervals between 25 and 2000 msec. Fig
ure 4-7A plots V(t) as a function of action potential duration. Figure
4-7B plots V(t) as a function of AHP amplitude. Although changes in
synaptic efficacy appear to correlate with both action potential dura
tion and AHP amplitude, the correlation with AHP amplitude was stron
ger, probably because of the greater reliability of measures of AHP
amplitude (see Note 1 at the end of this chapter). These results show
that stimulation-induced changes in the presynaptic action potential
and increases in transmitter release occur simultaneously in the chick
ciliary ganglion.
The effect of repetitive stimulation on the presynaptic action
potential is most pronounced in the first 150 msec of a 20/sec train
(Figure 4-5). The first component of facilitation (Fj), which has a
time constant of decay of about 60 msec in the chick ciliary ganglion
(see Chapter 2), seems to accumulate in a similar manner to that
described in the frog neuromuscular junction (Figure 2-3; Magleby and
Zengel, 1982). If this is the case, then the time course of accumula
tion of F¡ would be quite similar to the time course of the effects of
repetitive stimulation on the presynaptic action potential shown in
Figure 4-5. To determine whether changes in the presynaptic action
potential correlate with facilitation of transmitter release, a paired
pulse paradigm (like that used to describe facilitation in Chapter 2)
was employed to characterize changes in the presynaptic action poten
tial under conditions shown to produce facilitation.
Figure 4-8 shows the effect of a single conditioning stimulus on
the presynaptic action potential. In this experiment, paired pulses
were applied at seven different intervals. At shorter intervals, action


84
Dryer, S.E., and Chiappinelli, V.A. (1983) Kappa-bungarotoxin: an
intracellular study demonstrating blockade of neuronal nicotinic
receptors by a snake neurotoxin. Brain Res., 289:317-321.
Dryer, S.E., and Chiappinelli, V.A. (1985) Substance P depolarizes
nerve terminals in an autonomic ganglion. Brain Res.. 336:190-194.
Dudel, J., and Kuffler, S.W. (1961) The quantal nature of transmission
and spontaneous miniature potentials at the crayfish neuromuscular
junction. J. Physiol. (London). 155:514-529.
Dudel, J., Pamas, I., and Pamas, H. (1983) Neurotransmitter release
and its facilitation in crayfish, VI. Release determined by both
intracellular calcium and depolarization of the nerve terminal.
Pfluqers Archiv.. 339:1-10.
Edry-Schiller, J., and Rahamimoff, R. (1993) Activation and inactiva
tion of the bursting potassium channel from fused Torpedo synapto-
somes. J. Phvsiol. (London). 471:659-678.
Erulkar, S.D. (1983) The modulations of neurotransmitter release at
synaptic junctions. Rev. Biochem. Pharmacol.. 98:63-175.
Erulkar, S.D., and Rahamimoff, R. (1978) The role of calcium ions in
tetanic and post-tetanic increase in miniature end-plate potential
frequency. J. Phvsiol. (London). 287:501-511.
Fadoga, E., and Brookhart, J.M. (1962) Interactions of excitatory post-
synaptic potentials generated at different sites on the frog moto
neuron. J. Neurophysiol.. 25:790-804.
Feng, T.P. (1941) Studies on the neuromuscular junction. XXVI. The
changes of the end-plate potential during and after prolonged stim
ulation. Chin. J. Physiol.. 16:341-372.
Fletcher, G.H., and Chiappinelli, V.A. (1992a) An inward rectifier is
present in presynaptic nerve terminals in the chick ciliary gan
glion. Brain Res.. 575:103-112.
Fletcher, G.H., and Chiappinelli, V.A. (1992b) Spontaneous miniature
hyperpolarizations of presynaptic nerve terminals in the chick cil
iary ganglion. Brain Res.. 579:165-168.
Fletcher, G.H., and Chiappinel1i, V.A. (1993) The actions of the kappa
1 opioid agonist U-50,488 on presynaptic nerve terminals of the
chick ciliary ganglion. Neuroscience. 53:239-250.
Fox, A.P., Nowycky, M.C. and Tsien, R.W. (1987a) Kinetic and pharmaco
logical properties distinguishing three types of calcium currents
in chick sensory neurons. 0. Phvsiol. (London). 394:149-172.


83
Bittner, G.D., and Baxter, D.A. (1991) Synaptic plasticity at crayfish
neuromuscular junctions: facilitation and augmentation. Synapse.
7:235-243.
Bloedel, J., Gage, P.W., Llinas, R., and Quastel, D.M. (1966) Transmit
ter release at the squid giant synapse in the presence of tetrodo-
toxin. Nature, 212:49-50.
Bourque, C.W. (1991) Activity-dependent modulation of nerve terminal
excitation in a mammalian peptidergic system. TINS. 14:28-31.
Chang, C.C., Lin, S.-O., Hong, S.J., and Chiou, L.C. (1988) Neuromuscu
lar block by verapamil and diltiazem and inhibition of acetylcho
line release. Brain Res., 454:332-339.
Charlton, M., and Bittner, G.D. (1978a) Facilitation of transmitter
release at squid synapses. J. Gen. Physiol.. 72:471-486.
Charlton, M., and Bittner, G.D. (1978b) Presynaptic potentials and
facilitation of transmitter release at squid synapses. J. Gen.
Phvsiol.. 72:487-511.
Charlton, M.P., Smith, S.J., and Zucker, R.S. (1982) Role of presynap
tic calcium ions and channels in synaptic facilitation and depres
sion at the squid giant synapse. J. Phvsiol. (London), 323:
173-193.
Collins, W.F., Honig, M.G., and Mendell, L.M. (1984) Heterogeneity of
group la synapses on homonymous alpha-motoneurons as revealed by
high-frequency stimulation of la afferent fibers. J. Neurophysiol.,
52:980-993.
Crest, M., and Gola, M. (1993) Large conductance Ca2+-activated K+
channels are involved in both spike shaping and firing regulation
in Helix neurones. J. Phvsiol. (London). 465:265-287.
Curtis, D.R., and Eccles, J.C. (1960) Synaptic action during and after
repetitive stimulation. J. Phvsiol. (London). 150:374-398.
Dale, H.H., and Feldberg, W. (1936) The chemical transmission of secre
tory impulses to the sweat glands of the cat. J. Phvsiol. (London!,
82:121-128.
DeCamilli, P. (1993) Exocytosis goes with a SNAP. Nature. 364:387-389.
DeLorenzo, A.J. (1960) The fine structure of synapses in the ciliary
ganglion of the chick. J. Biophvs. Biochem. Cvtol.. 7:31-36.
del Castillo, J., and Katz, B. (1954) Statistical factors involved in
neuromuscular facilitation and depression. J. Physiol. (London!.
124:574-585.


25
Table 1. Time Constants of Decay of Components of Increased Transmitter Release
Rabbit
Frog NMjl Sympathetic Ganglion2
Rat
Hippocampus^
Chick
Ciliary Ganglion
Facilitation:
Fl
F2
60+3 msec
475+58 msec3
59+14 msec
388+97 msec3
present
63+4 msec
415+35 msec3
Augmentation
7.31.3 secb
7.21.0 secb
4.7 sec
30.72.3 secb
Potentiation
65+18 sec
88+25 sec
87 sec
205+24 sec
^Zengel & Magleby (1982); Magleby & Zengel (1976)
2Zengel et al. (1980)
^McNaughton (1982)
^magnitude and time constant increased in the presence of Sr2+
magnitude increased in the presence of Ba2+


39
The effect of Sr^+ on the second component of facilitation, which
could be reversed by washing the preparation with control Tyrode, was
seen in each of 6 experiments of this type. In the presence of Sr^+
the magnitude of increased from 0.50 + 0.09 to 0.92 + 0.21 (paired
t-test, P<0.05), whereas the time constant of decay of F2 increased
from 503 77 msec to 930 + 133 msec (paired t-test P<0.01). These
effects of Sr^+ are strikingly similar to the effects of Sr^+ at the
frog neuromuscular junction (compare Figure 2-8 to Figure 8 in Zengel
and Magleby, 1980).
Figure 2-9 shows the effect on augmentation and potentiation of
addition of small amounts of Ba^+. Notice the large increase in V(t)
during the first 90 sec of decay, the time during which augmentation is
decaying, in the presence of Ba^+ (A). This effect of Ba^+ on the
augmentation phase of decay is more clearly seen in Figure 2-9C, which
plots the decay of augmentation in the absence (filled triangles) and
presence of Ba^+ (open triangles). In contrast, Ba^+ had little or no
effect on potentiation (B).
Similar results were obtained in each of 6 experiments of this
type. In these experiments the magnitude of augmentation was signifi
cantly increased in the presence of 0.1 to 0.15 mM Ba^+ (1.10 + 0.26
vs. 2.87 + 0.89; paired t-test, P<0.05). In 4 other experiments, there
was an increase in V(t) during the first 80-100 sec of post
conditioning tests, but the data were too variable to obtain reliable
measures of augmentation and potentiation.
On the basis of these studies, it seems reasonable to conclude that
there are four components of stimulation-induced increases in transmit
ter release acting in the chick ciliary ganglion.


88
Marwitt, R., Pilar, G., and Weakly, J.N. (1971) Characterization of two
cell populations in avian ciliary ganglion. Brain Res.. 25:317-334.
Mastrogiacomo, A., Evans, C.J., and Gundersen, C.B. (1994) Antibody
peptides against a cysteine-string protein. J. Neurochem.
62:873-880.
McNaughton, B.L. (1982) Long-term enhancement and short-term potentia
tion in rat fascia dendata act through different mechanisms. jL
Phvsiol.(London). 324:249-262.
Meriney, S.D., and Grinnell, A.D. (1991) Endogenous adenosine modulates
stimulation-induced depression at the frog neuromuscular junction.
J. Physiol. (London). 443:441-455.
Miledi, R., and Slater, C.R. (1966) The action of calcium on neuronal
synapses in the squid. J. Physiol. (London) 198:473-498.
Miledi, R., and Thies, R. (1971) Tetanic and post-tetanic rise in
frequency of miniature end-plate potentials in low-calcium solu
tions. J. Physiol. (London) 212:245-257.
Miller, R.J. (1987) Multiple calcium channels and neuronal function.
Science. 235:46-51.
Mosier, D.R., and Zengel, J.E. (1994) Evoked transmitter release at the
frog neuromuscular junction in the presence of very low extracellu
lar Ca2+. Neurosci. Letters. 174:1-4.
Neher, E., and Augustine, G.J. (1990) Calcium gradients and buffers in
bovine chromaffin cells. J. Physiol. (London). 450:273-301.
Nicholls, J.G., Martin, A.R., and Wallace, B.G. (1992) From Neuron to
Brain, third edition, Sinauer Press, New York, USA.
Nowycky, M.C., Fox, A.P., and Tsien, R.W. (1985) Three types of neur
onal calcium channel with different calcium agonist sensitivity.
Nature. 316:440-443.
Nussinovitch, I., and Rahamimoff, R. (1988) Ionic basis of tetanic and
post-tetanic potentiation at a mammalian neuromuscular junction, >L
Physiol. (London) 396:435-455.
Pamas, H., Dudel, J., and Pamas, I. (1986) Neurotransmitter release
and its facilitation in crayfish VII. Another voltage-dependent
process beside calcium entry controls the time course of phasic
release. Pflug. Arch.. 406:131-137.
Partridge, L.D., and Swandulla, D. (1988) Single Ca-activated cation
channels in bursting neurons of Helix. Pfluo. Arch.. 410:627-631.


79
tion in the chick ciliary ganglion (Chapter 2). Many investigators
have suggested that separate underlying mechanisms bring about the four
components of increased release (e.g. Landau et al., 1973; Lev-Tov and
Rahamimoff, 1980; Magleby, 1973; Magleby and Zengel, 1982). It has
been suggested that the effects of repetitive stimulation are a
consequence of an increase in residual intracellular Ca2+ or a calcium
activated factor (Ca*) that causes a given presynaptic depolarization
to release an increased amount of transmitter (Katz and Miledi, 1965,
1968). This "residual calcium hypothesis" is the most widely accepted
theory to account for stimulation-induced increases in release. In its
simplest form, however, this model fails to adequately account for all
of the properties of stimulation-induced increased increases in release
(e.g. Zengel and Magleby, 1980, 1981, 1982; Bain and Quastel, 1992a).
Although Ca2+ ions appear to play an important role in the facilitation
phase (Katz and Miledi, 1968; Zengel et al., 1993a,b) and augmentation
phase (Erulkar and Rahamimoff, 1978; Magleby and Zengel, 1976; Zengel
et al., 1994) of increased release, the potentiation phase does not
appear to involve Ca2+; instead, it has been suggested that an accumu
lation of Na+ ions in the nerve terminal may be involved in this phase
of increased release (e.g. Birks and Cohen, 1968; Nussinovitch and
Rahamimoff, 1988). If the mechanisms underlying stimulation-induced
increases in release can be described, the release process itself will
be more completely understood.
Data presented in Chapter 4 of this dissertation show that facili
tation of transmitter release in the chick ciliary ganglion is accompa
nied by an increase in the duration of the presynaptic action poten
tial. One interpretation of this finding is that the increase in the


Figure 2-4. Facilitation of EPSP amplitude in ciliary neurons.
Inset: Example of averaged data showing paired pulse facilitation of
EPSP amplitude. [Ca2+] = 1.4 mM, [Mg2+] = 2 mM. (A) and (B) plot V(t)
as a function of time following a single conditioning stimulus. The
symbols connected by the dashed lines represent data from the same
cell. Data points represent the average of 5 to 15 identical trials.
The continuous lines describe the decay of the first and second compo
nents of facilitation (drawn using values for the magnitude and time
'racellular recordings in 4 mM
msec
msec


65
Figure 4-7. Correlation between facilitation of extracellularly
recorded compound action potential and changes in intracellularly
recorded presynaptic action potential. Each point represents the aver
age of 4 to 16 responses from a single cell. Data presented here are
from experiments done under low quantal conditions ([Ca2+] = 1.25 to
1.5 mM, [Mg2+] = 4 mM) where simultaneous measures of intracellularly
recorded presynaptic action potential and extracellularly recorded com
pound action potentials were obtained (V(t) calculated as described in
Chapter 2). Data are results of 8 different interstimulus intervals
applied to 7 different cells. (A): Relationship between extracel1ularly
recorded V(t) and changes in presynaptic action potential duration. A
regression line is drawn through the data (R= 0.62). (B): Relationship
between extracellularly recorded V(t) and presynaptic AHP amplitude. A
regression line is drawn through the data (R = 0.84).


67
Figure 4-8. Effects of paired pulse stimulation on the presynaptic
action potential. Traces represent the average of 2 to 6 identical
trials. Interstimulus intervals are indicated. [Ca2+] = 1.4 mM,
[Mg2+] = 4 mM.


A
23
12
>
O 5 impulses
1 impulse
0 1000 2000 3000 4000 5000
Time (msec)
0 200 400 600 8001000
Time (msec)
0 200
Time (msec)


53
It is generally accepted that transmitter release from vertebrate
peripheral synapses is initiated primarily by current flow through
N-type Ca2+ channels (reviewed in Miller, 1987). N-type channels, but
not channels of other types, have been shown to be physically connected
to proteins in the presynaptic terminal that act to regulate docking of
synaptic vesicles and priming of vesicles for release (reviewed in Ben
nett and Scheller, 1993). If these active zone proteins associate
exclusively with N-type Ca2+ channels, as has been suggested (Mastro-
giacomo et al., 1994), the limited role of other channel types to
exocytosis may be related to their distance from the release machinery.
In support of this, it has been reported (Stanley, 1993), that the
coupling of Ca2+ influx to acetylcholine release in the ciliary gan
glion appears to be conserved within a 3 pm2 membrane patch. Stanley
further suggested that Ca2+ influx through a single channel is suffi
cient to trigger quantal transmitter release, implying that Ca2+ influx
through N-type channels occurs extremely close to the transmitter
release mechanism in the chick ciliary ganglion.
In summary, I have characterized the effects of Ca2+ channel block
ing agents (divalent cations, dihydropyridine and phenyl alkyl amine
antagonists, and w-conotoxin GVIA) on ganglionic efficacy in the embry
onic chick ciliary ganglion. The results presented here further
strengthen the argument that the w-conotoxin sensitive Ca2+ currents
previously described in the presynaptic nerve terminal represent the
major source of Ca2+ elevation for initiation of transmitter release
from the calyx.


THE EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC TRANSMISSION
IN THE CHICK CILIARY GANGLION
By
Robert E. Poage
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
1995


THE EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC TRANSMISSION
IN THE CHICK CILIARY GANGLION
By
Robert E. Poage
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
1995

ACKNOWLEDGEMENTS
I would like to thank the many educators who have contributed to my
continued pursuit of higher education, most notably my advisor, Dr.
Janet Zengel. I have been very fortunate to have studied under and
learned from a scientist and teacher who is as dedicated and gifted as
any I have met. When I look back on years spent in graduate study I
will always be grateful for the opportunities I was given and the
patience that was always applied liberally. From my mentor I have
gathered many of the critical skills a scientist must have to be suc
cessful. From my friend I have seen how a scientist can pursue her
life's work with enthusiasm, integrity and a critical eye, turned as
closely inward as outward.
My graduate work at the University of Florida has been marked by
some excellent instruction, and opportunities to teach and to present
my work in front of peers. In retrospect, I am grateful for all of
these things.
My supervisory committee members, Dr. Janet Zengel, Dr. Stuart
Dryer, Dr. Peter Anderson, Dr. Philip Posner and Dr. Tom Vickroy, have
my gratitude for having confidence in my work when successes were slow
in coming. I thank them for their many suggestions and their support.
Finally, I would like to thank my parents, who always supported my
me, and my continuing education, and got me started in the field, and
my wife and family, who, whether they know it or not, are beautiful.

TABLE OF CONTENTS
ACKNOWLEDGEMENTS ii
ABSTRACT iv
CHAPTERS
1 INTRODUCTION 1
Chemical Synaptic Transmission 1
The Ciliary Ganglion of the Embryonic Chick 2
Stimulation-Induced Changes in Synaptic Efficacy 4
Voltage-Dependent Calcium Channels and Transmitter Release... 5
The Presynaptic Action Potential and Transmitter Release 7
Summary 8
2 EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC EFFICACY
IN THE CHICK CILIARY GANGLION 10
Methods 12
Results 18
Discussion 42
3 CHARACTERIZATION OF CA2+ CHANNELS INVOLVED IN SYNAPTIC
TRANSMISSION 45
Methods 47
Results 48
Discussion 50
4 EFFECT OF REPETITIVE STIMULATION ON THE PRESYNAPTIC
ACTION POTENTIAL 54
Methods 55
Results 56
Discussion 68
Notes 75
5 CONCLUSIONS 77
REFERENCES 82
BIOGRAPHICAL SKETCH
93

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
THE EFFECTS OF REPETITIVE STIMULATION ON
SYNAPTIC TRANSMISSION IN THE CHICK CILIARY GANGLION
By
Robert E. Poage
August 1995
Chair: Janet E. Zengel, Ph.D.
Major Department: Neuroscience
Under appropriate conditions, repetitive synaptic stimulation can
cause subsequent stimuli to release increased amounts of neurotrans
mitter. Repetitive stimulation can also change the shape of the pre-
synaptic action potential, but there are few experimental preparations
in which the relationship between these two phenomena may be studied
directly. The goals of this study were to use electrophysiological
recording techniques to investigate the role of Ca2+ channels in ini
tiation of release in the embryonic chick ciliary ganglion, to charac
terize stimulation-induced increases of synaptic efficacy at this
synapse, and to record electrical activity from the presynaptic element
of this synapse under conditions conducive to facilitation of release.
It is shown that treatments reported to block Ca2+ currents in
presynaptic terminals of this preparation cause concentration-dependent
blockade of synaptic transmission. These results are consistent with a
iv

role for these previously described Ca^+ currents in initiation of
transmitter release in the chick ciliary ganglion.
It is shown that four distinct components contribute to increased
ganglionic efficacy following repetitive stimulation. Their time con
stants of decay (60 milliseconds, 400 milliseconds, 30 seconds and 200
seconds) are similar to those that describe the decay of the four com
ponents of stimulation-induced increases in transmitter release
described in other preparations (first and second components of facili
tation, augmentation and potentiation). The components described here
are also similar to the above-mentioned processes in their sensitivi
ties to Sr^+ and Ba^+. It is concluded that the components of stimu
lus-induced increases in release that have been described in other syn
aptic preparations are present in the chick ciliary ganglion.
Repetitive stimulation under conditions conducive to facilitation
of transmitter release causes an increase in the duration of the pre-
synaptic action potential and a decrease in the amplitude of its after
hyperpolarization (AHP). The time course of the effects of repetitive
stimulation on the presynaptic action potential parallels the time
course of facilitation in this preparation. It appears likely that the
observed effects on action potential duration result from a decrease in
K+ conductance.
Results are discussed in terms of possible mechanisms underlying
the individual components of stimulation-induced increases of release.
v

CHAPTER 1
INTRODUCTION
Chemical Synaptic Transmission
Although the release of chemical neurotransmitter substances
mediates many forms of neuronal communication, the cellular mechanisms
underlying neurotransmitter release have yet to be identified. It has
been established that depolarization of the presynaptic nerve terminal
causes the activation of voltage-dependent Ca^+ channels and influx of
Ca^+ ions (Katz and Miledi, 1967). The resulting increase in Ca^+
concentration acts through an unknown mechanism to initiate release of
neurotransmitter. It is generally accepted that the resulting increase
in Ca^+ concentration causes synaptic vesicles within the nerve termi
nal to fuse with the nerve terminal membrane and spill their contents
(neurotransmitters) into the synaptic cleft via exocytosis. Diffusion
of the neurotransmitter across the cleft and its binding to specific
postsynaptic receptors results in the transmission of information,
either in the form of an increased ionic permeability or through the
action of second messenger systems. To understand neurotransmitter
release and its integral role in information processing and plasticity,
it will first be necessary to understand how the release process is
initiated and how it is modulated.
Several aspects of synaptic transmission make it difficult to
study. Exocytosis is an extremely rapid event. The time between nerve
l

2
terminal depolarization and release of transmitter is on the order of a
millisecond (Katz and Miledi, 1965). With very few exceptions the
nerve terminal is inaccessible to neurophysiological recording tech
niques, primarily because of the size of the presynaptic elements
involved. It would be of great interest to study many aspects of the
release process at a single synapse, but most experimental preparations
are not amenable to a wide range of available techniques.
The Ciliary Ganglion of the Embryonic Chick
The calyx nerve terminal in the ciliary ganglion of the embryonic
and posthatch chicken is one notable exception in as much as the struc
tures comprising the pre- and post-synaptic elements of the synapse are
large enough to be studied directly using standard electrophysiological
recording techniques. The accessibility of fertile eggs and the fact
that the cellular elements adapt well to cell culture enhance the value
of the preparation.
The ciliary ganglion is innervated by the third cranial nerve.
Preganglionic fibers originate in the approximately 2,000 cells of the
accessory motor nucleus, the avian analog of the mammalian Edinger-
Westphal nucleus. As the oculomotor nerve passes through the orbit, it
gives off branches to the muscles controlling eye movement. The
remaining fibers enter the ciliary ganglion, which is located behind
the eye lateral to the optic nerve. Fibers entering the ganglion form
two distinct types of synapses on two separate neuronal subpopulations.
The larger, more prevalent ciliary neurons receive single synaptic
inputs in the form of a "calyx" or cup-like nerve terminal (DeLorenzo,
1960). These nerve endings can cover 70% of the ciliary neuron's soma
(Hess, 1965). The axons of ciliary neurons exit in one or two ciliary

3
nerves and pierce the sclera to innervate the striated ciliary muscles
and constrictor muscles of the iris. The smaller choroid neurons are
innervated by multiple bouton-type synapses and project through three
to five choroid nerves to innervate the choroidal coat.
Synaptic contacts on both choroid and ciliary neurons are chemical
and cholinergic in nature. Calyx/ciliary neuron synapses also display
electrical coupling. The nature and size of the calical nerve terminal
and synapse provide a unique experimental opportunity to observe elec
trical activity in the nerve terminal of a vertebrate neuronal synapse.
Martin and Pilar (1963a) reported the unique nature of this synapse and
successfully recorded electrical activity in both pre-and postsynaptic
structures. In a series of elegant reports (Martin and Pilar, 1963a,b,
1964a-c), they described transmission in the calyx preparation, includ
ing properties of both electrical and chemical coupling. Since those
early descriptions, the chick ciliary ganglion has been used by several
laboratories to study neurotransmitter release and synaptic function
(Bennett and Ho, 1991; Dryer and Chiappinelli, 1985; Stanley, 1989;
Stanley and Goping, 1991; Yawo, 1990).
The chick ciliary ganglion is obviously a preparation of great
value to the neuroscientist. Due to the accessibility of the presynap-
tic element of the calyciform synapse in the chick ciliary ganglion, I
have chosen to use this preparation to study the relationship between
presynaptic electrical activity and transmitter release.
This dissertation describes: 1) stimulation-induced increases in
synaptic efficacy at the chick ciliary ganglion, 2) the effects of Ca^+
channel blocking agents on ganglionic efficacy at this synapse and 3)

4
the relationship between the presynaptic action potential and stimula
tion-induced increases in postsynaptic response amplitude.
Stimulation-Induced Changes in Synaptic Efficacy
Since the nervous system most often uses trains of electrical sig
nals to convey information, one important element of the study of
neurotransmission is to observe what happens to transmitter release
during and following repetitive stimulation. Many studies have shown
that the efficacy of synaptic transmission is affected by its prior
activity (c.f. Feng, 1941), but the cellular machinery involved is
almost as obscure now as it was fifty years ago.
Repetitive stimuli applied to a presynaptic axon under conditions
of low levels of release can lead to a progressive increase in the
amount of transmitter released by successive impulses (reviewed by
Zucker, 1989). Following stimulation, this increase in release decays
back to control levels with a time course that can range from millise
conds to minutes. Such stimulation-induced increases in release have
been studied most extensively at the frog neuromuscular junction, where
four distinct components have been described on the basis of their
kinetic and pharmacological properties. These components are the first
and second components of facilitation, which decay back to control lev
els of release with time constants of about 60 ms and 400 ms, respec
tively (Magleby, 1973; Mallart and Martin, 1967; Zengel and Magleby,
1982); augmentation, which decays with a time constant of approximately
7 seconds (Erulkar and Rahamimoff, 1978; Magleby and Zengel, 1976); and
potentiation, which decays with a time constant of tens of seconds to
minutes (Magleby and Zengel, 1975a,b; Rosenthal, 1969). Some or all of
these components of increased release have been observed in the rabbit

5
superior cervical ganglion (Zengel et al., 1980), in rat (Hubbard,
1963; Liley, 1956) and crayfish (Zucker, 1974) neuromuscular junctions,
in the squid giant synapse (Charlton and Bittner, 1978a), in cat spinal
cord (Curtis and Eccles, 1960; Kuno, 1964; Porter et al., 1970) and in
rat hippocampus (McNaughton, 1982).
I have characterized the kinetic and pharmacological properties of
stimulation-induced increases in synaptic efficacy in the chick ciliary
ganglion. These results, which are presented in Chapter 2, will show
that there are 4 components contributing to stimulation-induced
increases in synaptic efficacy and that these changes result from an
increase in chemical synaptic transmission.
Voltage-Dependent Calcium Channels and Transmitter Release
It has been proposed that accumulation of Ca2+ in the nerve termi
nal may be responsible for activity-dependent increases in neurotrans
mitter release (Katz and Miledi, 1968; Rosenthal, 1969; Weinreich,
1971). Results of experiments designed to test Katz and Miledi's
"residual Ca2+ hypothesis" support the involvement of Ca2+ as a
mediator of increased release, but suggest that Ca2+ must be acting at
several steps or sites in the release process to produce the observed
pattern of results (Landau et al., 1973; Zengel and Magleby, 1977,
1980; Zengel et al., 1993a,b, 1994). One element that will clearly
affect the concentration of Ca2+ present in the nerve terminal is Ca2+
influx through voltage-gated Ca2+ channels.
The ubiquitous role of Ca2+ in cellular function has led to a great
interest (and subsequently a large body of literature) in describing
and classifying voltage-activated Ca2+ channels according to their
kinetics and pharmacology. Neuronal Ca2+ channels are routinely

6
compared to the N-, L- and T-type channels described in chick dorsal
root ganglion cells (Fox et al., 1987a; Nowycky et al., 1985). T-type
channels produce transient membrane currents and have relatively low
conductances. L-type channels are noninactivating and produce long-
lasting currents. N-type channels are neither transient nor slowly
inactivating. A more recently described class of channels that is
activated by moderate depolarization appears to be most common in mam
malian neurons and has been designated "P-type" (Llinas et al., 1989).
In the chick ciliary ganglion, Ca2+ currents recorded under voltage
clamp from the calyx fail to meet the criteria for a single class of
Ca2+ channel, although they are most similar to the N-type group (Stan
ley, 1991; Stanley and Atrakchi, 1990). Calical Ca2+ currents are
insensitive to blockade by dihydropyridines and are blocked by w-cono-
toxin, consistent with an N-type classification, but they inactivate
slowly, if at all (Stanley and Goping, 1991). Stanley (1991) has
called these channels "Npj-"type (for neuronal-presynaptic terminal).
There is evidence that Ca2+ channels at other fast-transmitting syn
apses may have similar kinetic and pharmacological properties (Miller,
1987; Suszkiw et al., 1986; Yoshikami et al., 1989). These results,
taken with the finding that Ca2+ channels are located, possibly in
clusters, on the release face of the calical nerve terminal membrane
(Stanley, 1991), imply that the calical Ca2+ currents described above
may be acting in the release process. However, Ca2+ channels involved
in release at this synapse have not been fully characterized.
I have investigated the effects of Ca2+ channel blockers on initia
tion of evoked transmitter release in the chick ciliary ganglion. The
results of this study, presented in Chapter 3, contribute to a more

7
complete understanding of the Ca2+ channels present in the calyx and
the role these channels play in the initiation and regulation of
transmitter release.
The Presvnaptic Action Potential and Transmitter Release
The role of nerve terminal depolarization in triggering exocytosis
has been thought to be primarily through the activation of voltage-
activated Ca2+ channels and the subsequent rapid increase in local
intracellular Ca2+, although a direct role for depolarization in ini
tiating release has been proposed (Dudel et al., 1983; Hochner et al.,
1989; Sil insky et al., 1995). It stands to reason that the ability to
study the electrical activity of the presynaptic element of a synapse
is essential if the process of transmitter release is to be
well understood.
Since stimulation-induced changes in release are a general phenome
non seen at most synapses, it is of interest to discern the mechanism
or mechanisms acting to produce these effects on the release process.
In attempting to describe these underlying mechanisms, many investiga
tors have examined the role of the presynaptic action potential. If
the depolarization of the nerve terminal caused by the invasion of an
action potential is larger in amplitude or is prolonged (Hubbard and
Schmidt, 1963; Liley and North, 1953; Takeuchi and Takeuchi, 1962), the
resulting increase in activation of voltage-dependent Ca2+ channels
should lead to a larger influx of Ca2+ and increased release (e.g.
Augustine, 1990; Hochner et al., 1986).
In the ciliary ganglion, Martin and Pilar (1964c) found no gross
changes in the presynaptic action potential or resting membrane poten
tial under conditions conducive to what I will describe as facilitation

8
and potentiation. However, Martin and Pilar compared the amplitudes of
individual action potentials, and it is possible that subtle changes in
the action potential amplitude and/or duration would not have been
detected by this type of analysis. I have characterized the effects of
repetitive stimulation on the presynaptic action potential in the chick
ciliary ganglion. The results of these studies are presented in chap
ter 4 and possible contributions of changes in presynaptic electrical
activity to the processes of stimulation-induced increases in release
are discussed.
Summary
The major goal of this research project was to obtain a better
understanding of the mechanisms underlying stimulation-induced changes
in transmitter release. Experiments were performed using the ciliary
ganglion of the embryonic chicken.
I found that the chick ciliary ganglion responds to repetitive
stimulation with four components of increased release that are analo
gous to the first and second components of facilitation, augmentation
and potentiation, as have been described in other preparations. Both
kinetic (time constants of decay) and pharmacological (response to
Sr2+, Ba2+) properties of these components were used in their identifi
cation, and intracellular recording from postsynaptic cells verified
their presynaptic origin.
I found that the pharmacological characteristics of Ca2+ channels
coupled to evoked transmitter release are similar to the N-type channel
classification. This finding is in agreement with other investigators'
descriptions of presynaptic Ca2+-dependent currents in the chick
ciliary ganglion.

9
I also found that there are changes in the presynaptic action
potential during repetitive stimulation that have a similar time course
to facilitation, but a definitive role for changes in the presynaptic
action potential was not demonstrated. Possible mechanisms that could
underlie stimulation-induced changes in the action potential are dis
cussed.

CHAPTER 2
EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC EFFICACY
IN THE CHICK CILIARY GANGLION
For more than fifty years it has been known that synaptic efficacy
changes as a result of prior synaptic activity (for example, Feng,
1941). It has been well documented in a variety of synapses that these
changes arise from a change in the amount of neurotransmitter released
by each successive impulse during repetitive stimulation. At the frog
neuromuscular junction, where stimulation-induced changes in transmit
ter release have been studied most extensively, four components of
increased release have been identified on the basis of their kinetic
and pharmacological properties. These components are: the first and
second components of facilitation, which decay back to control levels
of release with time constants of approximately 60 and 400 msec
(Magleby, 1973; Mallart and Martin, 1967; Younkin, 1974; Zengel and
Magleby, 1982); augmentation, which decays with a time constant of
approximately 7 sec (Erulkar and Rahamimoff, 1978; Magleby and Zengel,
1976); and potentiation, which decays with a time constant of tens of
seconds to minutes (Magleby and Zengel, 1975a,b; Rosenthal, 1969).
Some or all of these components have been observed in the rabbit supe
rior cervical ganglion (Zengel et al., 1980), squid giant synapse
(Charlton and Bittner, 1978a), crayfish (Bittner and Baxter, 1991;
Zucker, 1974) and rat neuromuscular junction (Hubbard, 1963;
10

11
Liley, 1956; Nussinovitch and Rahamimoff, 1988), cat spinal cord (Curtis
and Eccles, 1960; Kuno, 1964; Porter, 1970) and rat hippocampus
(McNaughton, 1982).
The striking similarities in kinetics and pharmacological sensiti
vities of these components of increased release in different prepara
tions indicate that they may represent general phenomena that occur at
all synapses. However, the subcellular machinery subserving these
modulations of transmitter release has not yet been identified. The
chick ciliary ganglion is an ideal synapse to study modulation of
transmitter release because stimulation-induced increases in transmit
ter release have been reported (Martin and Pilar, 1964c), although they
have not been fully characterized, and because the preparation is amen
able to a variety of experimental techniques and methodologies.
The single most unusual property of the chick ciliary ganglion is
the development of a large "calyx"-type nerve terminal that can be
impaled by a microelectrode, providing a rare opportunity to observe
presynaptic electrical activity associated with transmitter release. As
a prelude to beginning an investigation of the nerve terminal electri
cal events associated with transmitter release, I began by characteriz
ing the changes in synaptic efficacy that occur during and following
repetitive stimulation in the chick ciliary ganglion. Intracellular
recording from postsynaptic ciliary neurons was employed to verify that
increases in synaptic efficacy result from an increase in transmitter
release. I report here that four components of stimulation-induced
increases in transmitter release are present in the chick ciliary gan
glion. These components have kinetic and pharmacological properties

12
similar to the components of increased transmitter release described
for other preparations.
Methods
Preparation and Solutions
Fertile White Leghorn chicken eggs (Poultry Science Unit, Univer
sity of Florida) were set in a forced draft rotating incubator (Peter-
sime model 1, Gettysberg, OH) kept at 37C, 70% humidity and candled on
days 4 to 10 to determine viability. Embryos were removed at embryonic
day 15-19 (stage 41-45) and sacrificed via decapitation. These ages
were chosen to coincide with maturation of the large "calyx" type syn
apse, before the synapse becomes primarily electrical in nature (Land-
messer & Pilar, 1972). The ciliary ganglia were dissected out under
intermittent washing with Tyrode solution (see below for composition).
Several dissection techniques were used. The most common approach was
to bisect the head and free 2 to 5 mm of the oculomotor nerve proximal
to the orbit. A lateral approach was then used to draw the eye aside,
liberate the ganglion from surrounding connective tissue, and dissect
free the ciliary nerve (3-10 mm) from both sides of the sclera.
A recording chamber was constructed entirely of Sylgard polymer
(Dow Corning, Midland, MI) poured into a small (about 6 cm diameter)
Petri dish. Two chambers of approximately equal volume (1.5 ml) were
connected by a 1 cm long passage through which solutions passed during
perfusion. Removing the bathing solution from a chamber physically
separate from the recording chamber minimized noise from surface vibra
tion. A ciliary ganglion, complete with the preganglionic (oculomotor)
and postganglionic (ciliary) nerve trunks, was pinned to the bottom of

13
the recording chamber using short lengths of very fine (0.1 mm
diameter) tungsten-iridium alloy wire (AlfaAESAR, Ward Hill, MA).
The recording chamber was held in place by small bits of clay and
surrounded by a plexiglass base to which perfusion apparati were
attached. The preparation was continuously perfused with an oxygenated
Tyrode solution (saline composition [in mM]: KC1 5; NaCl 150; CaCl2 1
to 5; MgCl2 2 to 12; glucose 10; HEPES 10; pH adjusted to 7.2-7.4) at a
rate of 1-2 ml/min (gravity driven). Fluid levels were kept constant
as saline was removed by suction through a bevelled hypodermic needle;
the level of the needle was adjusted to keep the preparation just below
the surface of the solution. In some experiments, Ba2+ (0.1-0.5 mM) or
Sr2+ (0.5-4.5 mM) was substituted for Ca2+ or added to the saline
solutions. In these experiments, the concentrations of Ca2+ and Ba2+
or Sr2+ were adjusted until the extracel 1 ularly-recorded postganglionic
response was approximately equal to the response in Ca2+-only Tyrode.
Salts for Tyrode solutions were purchased from Sigma Chemical (St.
Louis, MO). Solution changes were carried out between trials by chang
ing the source reservoir feeding the perfusion system. All experiments
were carried out at room temperature (20-23C).
Stimulating and Recording
Fluid suction electrodes (Dudel and Kuffler, 1961) mounted on
mechanical micromanipulators (Narishige, Japan) were used to draw up
the preganglionic (oculomotor) and postganglionic (ciliary) nerves.
These electrodes were made from PE-60 polyethylene tubing of 1.22 mm
outer diameter and 0.8 mm inner diameter (Becton, Dickinson & Co., Par-
sippancy, NJ). One end of the tubing was heated and pulled to an inner
diameter of 0.5 to 1.0 mm. The other end of the tube was connected to

14
a syringe, which was used to draw the nerve and bathing solution into
the tapered end. A silver wire (0.005 0.01 inch diameter) was
inserted through the wall of the tubing and placed within approximately
5 mm of the tapered end of the tubing. A second silver wire was
wrapped around each electrode shaft to within 5 mm of the tip to serve
as a ground electrode.
Short stimulus pulses (0.01-0.06 msec) were applied to the oculomo
tor nerve through a photoelectric stimulus isolation unit (Grass
Instruments, Quincy, MA) and the stimulus amplitude was adjusted until
it was clearly suprathreshold. The postganglionic responses were
amplified with a Grass P-5 series pre-amplifier and displayed on a Tek
tronix 5113 dual beam storage oscilloscope (Beaverton, OR). In most
experiments the response consisted of both an electrical and a chemical
component (see Figure 2-1), although in some ganglia from younger
embryos there was not a distinct peak for the electrotonically mediated
component. The amplitude of the chemically mediated component is a
function of the number of postsynaptic cells brought to threshold by
chemical neurotransmission (Martin and Pilar, 1963a). Thus, changes in
the amplitude of the chemically mediated component of the ganglionic
response represent changes in the number of postsynaptic cells acti
vated by orthodromic stimulation (Landmesser and Pilar, 1972; Poage and
Zengel, 1993). The main advantage of extracellular recording is the
fact that the postganglionic response represents the averaged activity
of the entire ganglion. Averaged ganglionic responses show consider
ably less variability than intracellular responses, thus decreasing the
number of trials needed to obtain a reliable result.

15
When intracellular recording was used, several changes were made to
minimize vibration and excess connective tissue that could foul intra
cellular electrodes. The connective tissue capsule that adheres closely
to the ganglion was removed with fine forceps (DuMont #5, Fine Science
Tools, Belmont, CA). All intracellular experiments were performed on a
Kinetic Systems 9101-11 vibration isolation table (Roslindale, MA) with
the perfusion apparatus and a dissecting microscope mounted on a free
standing Faraday cage. A micromanipulator with a motorized advance
attachment (460XYZ micromanipulator, 860 series motorizer, Newport, RI)
held an intracellular recording probe connected to the balanced bridge
input of a Dagan 8500 intracellulular amplifier (Minneapolis, MN). The
output of the intracellular amplifier was sent to 2 channels of a Tek
tronix oscilloscope for AC and DC recording. Most experiments were
also recorded onto VCR tape through a PCM recording adapter (A.R. Vet
ter Company, Los Angeles, CA). The basic sampling rate was 88.2 kHz
and the channel rise time was 50 psec with 14 bit A/D resolution.
Microelectrodes were pulled on a horizontal pipette puller
(Brown/Flaming P-87, Sutter Instrument Co., Los Angeles, CA) using
glass capillary tubing (items # 1B100F-4, TW100F-4; 0.54 or 0.75 mm
i.d., 1 mm o.d., World Precision Instruments, Sarasota, FL) and filled
with 3 M KC1 (tip resistances, 25-100 megaOhms). The microelectrode
was placed above the ganglion under visual control and the micromanipu
lator was used to advance the electrode into the ganglion proper.
Direct visualization of individual cells was not necessary. Microelec
trode penetration of a cell was achieved by capacitance "ringing" or by
gently tapping the micromanipulator and was evident as a voltage

16
deflection of -45 to -80 mV. Intracellular recordings were usually of
short duration, with impalements usually lasting less than 15 minutes.
Cells were identified by electrophysiological means as previously
published (Dryer and Chiappinelli, 1985; Martin and Pilar, 1963a,
1964a,b). By injecting hyperpolarizing current through the recording
electrode, it was possible to render both electrotonic and chemical
potentials subthreshold (Martin & Pilar, 1963a; see Figure 2-4 inset) so
that the underlying postsynaptic potentials could be observed. Mem
brane responses were monitored from the balanced bridge outputs of the
intracellular amplifier. The bridge balance was adjusted before
recording and was verified by testing the bridge balance after the
microelectrode was removed from a cell.
Data Collection and Analysis
For paired pulse and 5 impulse experiments, a Grass Instruments S48
stimulator was used to generate the conditioning and testing stimuli.
Responses were averaged and their amplitudes measured using either a
Nicolet 1170 signal averager (Nicolet Instruments Co., Madison, WI) or
386-based data acquisition and analysis software (Axotape, Axon Instru
ments, Foster City, Ca). For experiments in which longer conditioning
trains were applied, a MINC-11 computer (Digital Equipment Corp., Marl
boro, MA) was often used to generate the stimulation patterns, measure
and store the postganglionic response amplitudes, and analyze the data
(Magleby and Zengel, 1976; Zengel and Magleby, 1982). Sufficient time
was allowed between trains to ensure that release had returned to pre
conditioning levels (8 to 20 minutes, depending on train duration and
stimulation rate).

17
Definition of Terms
Changes in response amplitude following conditioning simulation are
expressed as:
v(t) = (Vt/V0) 1 (2-1)
where V0 is the control (pre-conditioning) response amplitude and
is the amplitude of the response at time t following the conditioning
stimulation. For analysis of different components of stimulation-
induced changes in V(t), I have used the approach described by Zengel
and Magleby (1980, 1982). In brief, each component is defined as the
fractional change in response amplitude in the absence of other compo
nents. Since it is not always possible to measure one component in the
absence of others, the magnitudes and time constants of the individual
components are derived from the value of V(t) by assuming that these
components have distinct non-overlapping time constants of decay. The
slowest decaying component can be estimated by using data points col
lected after the more rapidly decaying components have decayed away.
Because the faster decaying components fall on top of the slower decay
ing ones, estimates of the contributions of these components can be
made only by assuming some relationship between the different compo
nents and using standard linear decay analyses. In this study I have
used a model shown to describe stimulation-induced changes in transmit
ter release at the frog neuromuscular junction (Magleby and Zengel,
1982; Zengel and Magleby, 1980, 1982). Basically, this model assumes
that there are four independent components of increased release which
interact according to the following equation:
Vt/V0 = (Fj + F2 + 1)(A + 1)(P + 1) (2-2)
where Fj and F2 are the first and second components of facilitation,

18
A is augmentation and P is potentiation. As reported in this chapter,
this model appeared to describe stimulation-induced changes in neuro
transmitter release in the chick ciliary ganglion.
Statistical Analysis
Control and experimental trials were recorded in each experiment.
The effect of the experimental treatment was compared to the response
of the same preparation under control conditions using t-test proce
dures. Statistical analysis was performed using the IBM PC version of
SigmaPlot 5.0 (Jandel Scientific). Averaged data are presented as mean
+ standard error.
Results
Description of Stimulation-Induced Increases in Synaptic Efficacy
A paired pulse paradigm was employed to examine the effects of a
single conditioning stimulus on the efficacy of ganglionic transmis
sion. Figure 2-1 presents averaged extracellular data from a single
preparation that was stimulated with pairs of impulses at an interstim
ulus interval of 50 msec. The chemically mediated component, which is a
function of the number of cells brought to threshold by chemical trans
mitter release, is increased following a single conditioning impulse
while the shock artifact and electrotonically mediated component of the
response were unaffected. Under conditions of low levels of release
many of the postsynaptic cells are below threshold for action potential
generation and are not contributing to either peak of the postgan
glionic response. Increases in transmitter release will bring some
previously unresponsive cells above threshold. These cells contribute
to the observed increase in compound action potential amplitude in
Figure 2-1.

19
C
Figure 2-1. Facilitation of compound action potential amplitude.
The oculomotor nerve was stimulated with a pair of pulses applied at an
interstimulus interval of 50 msec and the postganglionic response was
recorded extracellularly (see MATERIALS AND METHODS). In this record,
each response consisted of 3 upward deflections: a shock artifact (*)
and electrically mediated (e) and chemically mediated (c) components of
the postganglionic response. The trace represents the average of 8
consecutive trials. The temporal separation between the electrical and
chemical components of each response is due to the synaptic delay pre
sent for chemical neurotransmission. Note the large increase in the
amplitude of the chemical component of the response to the second stim
ulus. This facilitation was typically observed under low quantal con
ditions. [Ca]2+ = 1.5 mM, [Mg+] *= 2 mM.

20
In order to characterize the time course of the increase in synap
tic efficacy following a single conditioning impulse, testing impulses
were applied at intervals ranging from 25 msec to 5 sec. The presenta
tion of intervals was randomized during an experiment. The symbols in
Figure 2-2 plot V(t), the fractional increase in post-ganglionic
response amplitude (Equation (2-1)), as a function of interstimulus
interval. Each symbol represents data from a single Ca^+ and Mg^+
concentration. As illustrated in Figure 2-2, facilitation of the chem
ically mediated component was greatest when release was reduced by
decreasing Ca^+ (A) or increasing Mg^+ (B). Under these conditions,
the stimulation-induced increase in response amplitude was greatest at
short conditioning-testing intervals, and decreased as the interstimu
lus interval was increased.
Results from another preparation are presented in Figure 2-3. The
filled circles in Figure 2-3A plot V(t) as a function of time following
a single conditioning impulse. When these data are plotted on a semi-
logarithmic scale (filled circles, Figure 2-3B), it is obvious that the
decay cannot be described by a single exponential. The data are well
described by a dual exponential. The contribution of the slower of the
two components was estimated by fitting a regression line to the linear
portion of the curve (see figure legend). In this experiment the ini
tial magnitude of this slower component, given by the intercept of the
regression line at time 0, was 0.99 and its time constant of decay was
565 msec. To obtain an estimate of the more rapidly decaying compo
nent, the data were corrected for the contribution of the slower compo
nent using the model described in Equation 2-2 to determine the rela
tive contribution of each component (see figure legend). The filled

21
B
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
200 400 600 800 1000
Time (msec)
12 mM Mg
200 400 600 800 1000
Time (msec)
Figure 2-2. Effects of reducing extracellular Ca2+ or increasing
extracellular Mg2+ on facilitation of compound action potential ampli
tude. A: Plot of the decay of V(t), the fractional change in response
amplitude (Equation 2-1), as a function of time following a single con
ditioning impulse. The [Mg2+] was held constant at 2 mM while [Ca2+]
was varied from 1 to 3 mM. Conditioning-testing trials were applied
once every 20 sec. Data points represent the average of 8 to 16 iden
tical trials from a single preparation. B: In another preparation,
plot of the decay of V(t) as above, only with [Ca2+] held constant at 3
mM and [Mg2+] varied from 2 to 12 mM. Data points represent the aver
age of 8 to 24 identical trials.

Figure 2-3. Effect of 1 and 5 conditioning impulses on synaptic
efficacy. A: Plot of the decay of V(t) as a function of time follow
ing a single conditioning impulse (filled circles) and a train of 5
conditioning impulses applied at 20/sec (open circles). Single testing
impulses were applied at intervals of 25 to 5000 msec after the condi
tioning stimulation. Conditioning-testing trials were applied about
once every 30 sec. Data points represent the average of 16 trials from
a single preparation. [Ca2+]? = 1 mM, [Mg2+]g = 2 mM. B: Semilogarith-
mic plot of the decay of V(t) following one (filled circles) and 5
(open circles) conditioning impulses (same data as A). The lines rep
resent the exponential decay of the second component of facilitation
(F2) derived by fitting regression lines through the data points
between 200 and 2000 msec. C: Decay of the first component of facili
tation (Ft) following one (filled circles) and five conditioning
impulses (open circles). Values of Fj were obtained by subtracting off
from V(t) the contribution of the second component of facilitation
(Equation 2-2 in Methods). The lines represent the exponential decay
of Fj, derived by fitting a regression line through the data points.

A
23
12
>
O 5 impulses
1 impulse
0 1000 2000 3000 4000 5000
Time (msec)
0 200 400 600 8001000
Time (msec)
0 200
Time (msec)

24
circles in Figure 2-3C plot the estimated decay of the more rapidly
decaying component, which in this experiment had an initial magnitude
of 5.8 and a time constant of decay of 60 msec.
In 24 paired pulse experiments of this type, the average time con
stants describing the decay of V(t) were about 60 msec (n=14) and 400
msec (n=24; see Table 2-1). (In some experiments done under higher
quantal conditions (>1.5 mM Ca^+), there was an apparent depression of
ganglionic transmission at short interstimulus intervals, precluding
precise measures of the more rapidly decaying component.) These two
time constants of decay are very similar to the time constants pre
viously reported for the first and second components of facilitation at
the frog neuromuscular junction and the rabbit sympathetic ganglion
(Table 2-1).
Intracellular experiments were performed under the above conditions
to verify that facilitation of the ganglionic response results from an
increase in EPSP amplitude. As reported using extracellular recording,
paired pulse stimulation led to a facilitation of EPSP amplitude. Fig
ure 2-4 shows results from several ciliary neurons in two different
Ca^+ concentration ranges. The continuous lines describing the decay
of facilitation were drawn by taking the average values for the magni
tude and time constant of the two components of facilitation (using
results from extracellular data) and plotting their combined decay (see
figure legend). It appears that the estimates of facilitation provided
by extracellular records can provide a reliable estimate of the effects
of paired pulse stimulation on EPSP amplitude. These findings are
quite similar to the results of Martin and Pilar (1964c), who reported
a facilitation of EPSP amplitude in the chick ciliary ganglion that

25
Table 1. Time Constants of Decay of Components of Increased Transmitter Release
Rabbit
Frog NMjl Sympathetic Ganglion2
Rat
Hippocampus^
Chick
Ciliary Ganglion
Facilitation:
Fl
F2
60+3 msec
475+58 msec3
59+14 msec
388+97 msec3
present
63+4 msec
415+35 msec3
Augmentation
7.31.3 secb
7.21.0 secb
4.7 sec
30.72.3 secb
Potentiation
65+18 sec
88+25 sec
87 sec
205+24 sec
^Zengel & Magleby (1982); Magleby & Zengel (1976)
2Zengel et al. (1980)
^McNaughton (1982)
^magnitude and time constant increased in the presence of Sr2+
magnitude increased in the presence of Ba2+

Figure 2-4. Facilitation of EPSP amplitude in ciliary neurons.
Inset: Example of averaged data showing paired pulse facilitation of
EPSP amplitude. [Ca2+] = 1.4 mM, [Mg2+] = 2 mM. (A) and (B) plot V(t)
as a function of time following a single conditioning stimulus. The
symbols connected by the dashed lines represent data from the same
cell. Data points represent the average of 5 to 15 identical trials.
The continuous lines describe the decay of the first and second compo
nents of facilitation (drawn using values for the magnitude and time
'racellular recordings in 4 mM
msec
msec

o
100 200 300 400 500
Time (msec)

28
decayed with a time constant on the order of 65 msec. It was also
reported that the observed increase in EPSP amplitude could be
accounted for entirely by an increase in quantal content. The ionic
conditions and stimulation paradigms used in this earlier study are
nearly identical to the conditions of my own experiments. Therefore,
it seems reasonable to assume that the increases in synaptic efficacy
described here using paired pulse stimulation result from an increase
in quantal transmitter release.
To observe more slowly decaying increases in ganglionic response
amplitude, ganglia were conditioned using longer trains of stimuli
(200-1200 impulses at 10-50/sec). The filled circles in Figure 2-5A
plot the decay of V(t) of the chemically mediated component of the
action potential following a conditioning train of 800 impulses applied
at 20/sec. When the data are plotted on a semilogarithmic scale
against time following the end of the conditioning stimulation (filled
circles in Figure 2-5B), there appear to be two components of decay.
The contribution of the slower component was estimated by fitting a
regression line to the linear portion of the curve at times beyond 100
sec (see figure legend). In this experiment, the initial magnitude of
the slower component was 0.32 and its time constant of decay was 200
sec. Figure 2-5C plots the decay of the faster component, obtained by
correcting for the contribution of the slower component (see figure
legend). This component had an initial magnitude of 1.02 and a time
constant of decay of 12 sec. In experiments in which it was possible
to accurately measure the amplitude of both components of the postgan
glionic response, the amplitude of the electrically mediated peak never

Figure 2-5. Effect of trains of repetitive stimulation on synaptic
efficacy. A: Decay of V(t) as a function of time following a condi
tioning train of 800 impulses applied at 20/sec. Testing impulses were
applied at 2 sec intervals for 3 impulses, then every 10 sec. The
filled circles plot the fractional change in amplitude of the chemi
cally mediated peak of the compound action potential. The open circles
plot changes in the amplitude of the electrically mediated portion of
the compound action potential. Data averaged from 4 identical trials
from a single preparation. [Ca^+] =1.5 mM, [Mg^+] = 2 mM. Note that
the electrotonically mediated potential (open circles) is not increased
following repetitive stimulation. B: Semilogarithmic plot of the decay
of V(t) (same data as in A). The line represents the exponential decay
of potentiation, obtained by fitting a regression line through the data
points beyond 100 sec. C: Decay of augmentation, obtained after cor
recting for the contribution of potentiation using Equation 2-2. The
line represents the exponential decay of augmentation, derived by fit
ting a regression line through the data points.

30
Time (sec)
B
10
0.1
0.01
0.001
1/1 2
0 100 200 300 400 500
Time (sec)
0 100 200
Time (sec)
see

31
increased, and was occasionally depressed during and immediately
following tetanic stimulation (open circles in Figure 2-5A).
In 27 experiments of this type, the average values for the time
constants of decay of these two more slowly decaying phases were about
30 sec (n=17) and 200 sec (n=27). These time constants are similar to
those attributed to the processes of augmentation and potentiation,
respectively (Table 2-1).
In order to verify that the observed changes in postganglionic
response reflect a change in EPSP amplitude, intracellular recording
from ciliary neurons was used. Figure 2-6 shows the effect of 800 con
ditioning stimuli applied at 20/sec on the amplitude of the EPSP. The
amplitude of the control EPSP (before the conditioning stimulation) was
5 mV. Test pulses were applied at 10 second intervals following the
end of the train. Each test pulse applied less than 60 seconds after
the train produced a large EPSP that initiated an action potential (not
shown). EPSP amplitude declined over the next 6 minutes until the
response reached preconditioning levels.
In three experiments of this type, EPSP amplitude increased to a
maximum of 150% to 400% of control values during the test period. The
EPSP amplitude returned to a pre-conditioning level with a time con
stant of about 3 to 4 minutes, similar to that observed by Martin and
Pilar (1964c) under similar conditions. This time course of the decay
of EPSP amplitude following repetitive stimulation is also very similar
to the decay of the extracel1ularly recorded postganglionic response
(Figure 2-5; Table 2-1). The increases in EPSP amplitude reported by
Martin and Pilar were accompanied by an increase in the frequency, but

32
CONTROL
60 sec
90 sec
120 sec
5 mV
20 msec
250 sec
Figure 2-6. Effect of an 800 impulse train (20/sec) on EPSP ampli
tude. Intracellular recording of EPSPs from a ciliary neuron shows the
response to orthodromic stimulation prior to and at the times indicated
following the conditioning train. [Ca2+] = 5 mM, [Mg2+] = 4 mM. The
time course of potentiation in this experiment is very similar to
results reported by Martin and Pilar (1964c, Figure 5).

33
not the unit size, of spontaneous miniature EPSPs. This suggests that
the potentiation of the EPSP reported here is indeed due to an increase
in transmitter release, and not an effect on postsynaptic receptor
sensitivity (Martin and Pilar, 1964c).
The augmentation phase of increased release was difficult to
describe using intracellular recording. Two possible reasons for this
are: 1) the variability of EPSP amplitude under conditions of very low
quantal content may make observing the decay of augmentation, which is
described using only the first 5 to 10 test points (0 to 60 seconds
after conditioning), dependent on averaging large numbers of identical
trials, and 2) at higher quantal content, an apparent depression of
release immediately following tetanic stimulation may confound attempts
to observe effects of stimulation on increased EPSP amplitude during
the time when augmentation would be observed.
Effect of Number of Conditioning Impulses on the Components of
Increased Synaptic Efficacy
If the processes I have described arise from the same mechanisms
that produce stimulation-induced increases in release in other synaptic
preparations, then the growth and decay of the stimulation-induced
changes in synaptic efficacy in the chick ciliary ganglion should be
described by the "accumulation" models which have been shown to
describe release in other preparations (e.g. Magleby and Zengel,
1975a,b, 1982; Mallart and Martin, 1967; Younkin, 1974). According to
these models, each conditioning impulse adds an incremental increase to
each component of increased release. The components then decay with
their characteristic time constants between impulses. Increasing the
number of conditioning impulses would be expected to increase the mag
nitudes of the various components of increased ganglionic efficacy.

34
To test this, experiments were conducted to examine the effects of
conditioning impulse number on the two components of facilitation, aug
mentation, and potentiation. The open circles in Figure 2-3A plot the
decay of V(t) following a 5 impulse conditioning train. At all time
points tested, the increase in V(t) was much greater following the 5
impulse train (open circles) than following a single conditioning
impulse (filled circles). This increase in V(t) could be attributed to
an increase in the magnitudes of the second (lines through open circles
in Figure 2-3B) and first components of facilitation (lines through
open circles in Figure 2-3C). There was little or no effect of condi
tioning impulse number on the time constants of decay of the two compo
nents of facilitation. Similar results were observed in 2 additional
experiments in which 1, 2 and 5 impulse conditioning trains were
applied. Thus, as with the rabbit sympathetic ganglion (Zengel et al.,
1980) and frog neuromuscular junction (Zengel and Magleby, 1982),
increasing the number of conditioning impulses results in an increase
in the magnitudes of the two components of facilitation, while having
little effect on their time constants of decay.
I also examined the effect of conditioning impulse number on the
magnitudes of the slower decaying components. Figure 2-7 summarizes
the results of 6 experiments in which conditioning trains of various
duration were applied. In each of these experiments I observed an
increase in the magnitude of potentiation when the number of condition
ing impulses was increased (Figure 2-7A). In the 3 experiments in
which I was able to obtain reliable estimates of augmentation, a simi
lar effect was observed (Figure 2-7B). In some experiments, particu
larly with higher Ca2+ concentrations (> 3 mM) or longer stimulation

35
A B
Number of impulses Number of impulses
Figure 2-7. Effect of the number of conditioning impulses on the
magnitudes of potentiation (A) and augmentation (B). The magnitudes of
augmentation and potentiation were obtained as described for Figure
2-5. Results of 6 experiments in which the decay of V(t) was recorded
following conditioning stimulation at 20 impulses/sec. Lines connect
data from the same preparation.

36
trains, I observed an apparent depression of ganglionic transmission
immediately following the conditioning trains that made it difficult to
obtain reliable estimates of augmentation. There was no consistent
effect of increasing stimulus duration on the time constants of decay
of these processes. These results are similar to those observed at
other synapses (e.g. Magleby and Zengel, 1976; Zengel et al., 1980).
Pharmacological Characterization: Effects of Strontium and Barium
At the frog neuromuscular junction and the rabbit sympathetic gan
glion, the addition of certain divalent cations to the bathing solution
selectively affects individual components of stimulation-induced
increases in release. Barium increases the magnitude of the augmenta
tion phase and strontium increases the magnitude and time constant of
the second component of facilitation (Zengel and Magleby, 1977, 1980,
1981; Zengel et al., 1980). To further test the hypothesis that the
phenomena I describe here are analogous to the four components of stim
ulation-induced increases in release reported in other preparations, I
repeated the experiments described in Figures 2-3 and 2-5 in the pres
ence of these divalent cations.
Figure 2-8 illustrates the effect of Sr^+ on facilitation. In the
presence of Sr^* (open circles), V(t) was unchanged or slightly reduced
at short interstimulus intervals (less than 100 msec), but there was an
obvious enhancement of ganglionic efficacy at intervals of 300 to
2000 msec (A). This increase in V(t) could be attributed to an
increase in both the magnitude and time constant of decay of the second
component of facilitation (B). The first component of facilitation was
reduced in magnitude in the presence of Sr^+ (C).

Figure 2-8. Effect of Sr2+ on facilitation. A: Plot of the decay
of V(t) as a function of time following a single conditioning impulse
in 1.5 mM Ca2+ Tyrode (filled circles) and in Tyrode containing 1.0 mM
Ca2+ and 1.5 mM Sr2+ (open circles). Data points represent the average
of 32 trials from a single preparation. B: Semilogarithmic plot of the
decay of V(t) (same data as in A). The lines, obtained by fitting
regression lines through the data points between 300 and 2000 msec,
represent the exponential decay of the second component of facilitation
(F2). C: Decay of the first component of facilitation (Fj), obtained
after correcting for F2 as described earlier.

38
2.0
1.5
>
1.0
0.5
0.0
1.5 mM Ca2+
O 1 mM Ca2+/ 1.5 mM Sr2+
0 500 1000 1500 2000
Time (msec)
0 500 1000 1500 2000 0 500
Time (msec) Time (msec)

39
The effect of Sr^+ on the second component of facilitation, which
could be reversed by washing the preparation with control Tyrode, was
seen in each of 6 experiments of this type. In the presence of Sr^+
the magnitude of increased from 0.50 + 0.09 to 0.92 + 0.21 (paired
t-test, P<0.05), whereas the time constant of decay of F2 increased
from 503 77 msec to 930 + 133 msec (paired t-test P<0.01). These
effects of Sr^+ are strikingly similar to the effects of Sr^+ at the
frog neuromuscular junction (compare Figure 2-8 to Figure 8 in Zengel
and Magleby, 1980).
Figure 2-9 shows the effect on augmentation and potentiation of
addition of small amounts of Ba^+. Notice the large increase in V(t)
during the first 90 sec of decay, the time during which augmentation is
decaying, in the presence of Ba^+ (A). This effect of Ba^+ on the
augmentation phase of decay is more clearly seen in Figure 2-9C, which
plots the decay of augmentation in the absence (filled triangles) and
presence of Ba^+ (open triangles). In contrast, Ba^+ had little or no
effect on potentiation (B).
Similar results were obtained in each of 6 experiments of this
type. In these experiments the magnitude of augmentation was signifi
cantly increased in the presence of 0.1 to 0.15 mM Ba^+ (1.10 + 0.26
vs. 2.87 + 0.89; paired t-test, P<0.05). In 4 other experiments, there
was an increase in V(t) during the first 80-100 sec of post
conditioning tests, but the data were too variable to obtain reliable
measures of augmentation and potentiation.
On the basis of these studies, it seems reasonable to conclude that
there are four components of stimulation-induced increases in transmit
ter release acting in the chick ciliary ganglion.

Figure 2-9. Effect of Ba2+ on potentiation and augmentation. A:
Plot of the decay of V(t) following 800-impulse conditioning trains in
the absence (filled circles) and presence of Ba2+ (open circles).
Control, conditioning and testing impulses as in Figure 2-5. Data
averaged from 8 trials from a single preparation. B: Semilogarithmic
plot of the decay of V(t) (same data as in A). The lines, obtained by
fitting regression lines through the data points beyond 100 sec, repre
sent the exponential decay of potentiation. C: Decay of augmentation,
obtained after correcting for potentiation as described earlier.

41
8 r
7 -o
6 -
5
4
3
2 -
1 -
O

o
2 +
1 mM Ca
O a 0.75 mM Cq2+/0.15 mM Ba2+
0 100 200 300 400 500
Time (see)
Time (see)
Time (see)

42
Discussion
The aim of the experiments presented here was to fully characterize
the kinetic properties of stimulation-induced changes in synaptic effi
cacy in the embryonic chick ciliary ganglion, and to investigate the
sensitivity of these processes to the divalent cations Sr2+ and Ba2+.
The results indicate that there are four components of stimulation-
induced increases in ganglionic efficacy described by time constants of
about 60 msec, 400 msec, 30 sec and 200 sec (Table 2-1). In several
synaptic preparations, accumulation models describing stimulus-induced
increases in transmitter release have been successful in describing
increases in transmitter release (e.g. Magleby and Zengel, 1975b; Mal-
lart and Martin, 1967; Younkin, 1974). One basic attribute of these
models is that each conditioning impulse increments the mechanisms
underlying each of the components of increased release. The results
presented in Figures 2-3 and 2-7 show that the four components observed
in the present study accumulate as predicted by models describing
facilitation, augmentation and potentiation (e.g. Magleby and Zengel,
1982; Zengel and Magleby, 1982). Further identification of the second
component of facilitation and of augmentation was achieved by exploit
ing the pharmacological sensitivities of these processes to certain
divalent cations. The addition of Ba2+ to the bathing solution caused
an increase in the magnitude of augmentation, and partial or complete
substitution of Sr2+ for Ca2+ in the bathing solution resulted in an
increase in the magnitude and the time constant of the second component
of facilitation. Because of the similarity in kinetic and pharmacolog
ical properties to those described previously for other synapses, I
believe that the processes I have described here are analogous to the

43
first component of facilitation, the second component of facilitation,
augmentation and potentiation as described at the frog neuromuscular
junction and at other synapses.
In preparations where stimulation-induced increases in synaptic
efficacy have been studied extensively, most notably the frog neuromus
cular junction, the increases have been shown to result from an
increase in quantal release (del Castillo and Katz, 1954; Magleby and
Zengel, 1976). In one of the first electrophysiological studies using
the chick ciliary ganglion preparation, Martin and Pilar (1964c) showed
that paired pulse facilitation of EPSP amplitude occurs in the embry
onic chick ciliary ganglion, and that it is a result of increased quan
tal content which decays back to control levels with a time constant of
about 65 msec. Data reported here confirm the presence of fcilitatory
processes and further describe two individual components of facilita
tion with distinct kinetic and pharmacological properties. Martin and
Pilar (1964c) also reported, using intracellular recording from ciliary
neurons, the presence of a more slowly decaying potentiation of EPSP
amplitude, although its time course was not described in detail. The
experiments reported here describe the decay of ganglionic efficacy
after prolonged repetitive stimulation. A more rapidly decaying compo
nent is akin to augmentation, having very similar kinetic and pharmaco
logical properties. The more slowly decaying component is termed
potentiation.
It has now been demonstrated that there are four components con
tributing to stimulation-induced increases in synaptic efficacy that
appear to result from increased transmitter release. The unresolved

44
question remains: What subcellular mechanisms are conserved at the
synapse that produce these processes at different synapses in a variety
of species?
Since quantal content is known to be affected by manipulations of
Ca2+ buffering and entry of Ca2+ into the nerve terminal, speculations
on the mechanism of stimulation-induced increases in neurotransmitter
release focus on the role of intracellular Ca2+ in transmitter release
(e.g. Charlton et al., 1982; Katz and Miledi, 1967, 1968; Zengel et
al., 1993a,b). No single theory has been successful in accounting for
all of the observed stimulation-induced changes in synaptic transmitter
release. Proposed mechanisms include an increased entry of Ca2+ or an
accumulation of Ca2+ in the presynaptic nerve terminal (Erulkar and
Rahamimoff, 1978; Katz and Miledi, 1968; Miledi and Thies, 1971; Rosen
thal, 1969; Weinreich, 1971) and nerve terminal voltage changes or pro
cesses associated with these voltage changes (e.g. Martin and Pilar,
1964c; Bittner and Baxter, 1991). The ciliary ganglion offers a versa
tile system in which to study these possibilities through the use of
many techniques, including presynaptic intracellular recording (Dryer
and Chiappinelli, 1983; Martin and Pilar, 1964c; Yawo, 1990), patch
clamp recording of Ca2+ currents (Stanley, 1989), and Ca2+-imaging
using fluorescent dyes. Chapter 4 will describe the effects of repeti
tive stimulation on presynaptic potentials under conditions conducive
to facilitation of transmitter release.

CHAPTER 3
CHARACTERIZATION OF CALCIUM CHANNELS
INVOLVED IN SYNAPTIC TRANSMISSION
The relationship between Ca2+ channel subtypes and transmitter
release is not well understood. In most synapses evoked release of
transmitter is dependent upon a coordinated influx of Ca2+ ions that
elevates intracellular Ca2+ at some site in the presynaptic nerve
terminal (Katz and Miledi, 1967). Voltage-activated Ca2+ conductances
are most often responsible for this rapid increase in intracellular
Ca2+. For this reason and because Ca2+ channels are potential targets
for modulating the release process (reviewed in Scott et al., 1991), it
is of interest to determine which channel type(s) are acting to initi
ate release in the chick ciliary ganglion.
There are clear indications for the existence of at least 4 general
classifications of neuronal Ca2+ channels. Subtypes N, L and T have
been described in chick dorsal root ganglion cells (Fox et al., 1987a;
Nowycky et al., 1985). These channels have been identified by their
sensitivities to different classes of pharmacological agents and by
their kinetics of activation and inactivation. L-type channels are
characterized by large unitary conductances (about 25 pS), activation
voltages positive to -10 mV and sensitivity to Ca2+ channel blockers
(dihydropyridines (BAY K 8644, nifedipine, nimodipine) and phenylalky-
lamines (verapamil)). T-type channels have small unitary conductances
(near 8 pS), are activated with weak depolarization (positive to -70
45

46
mV) and are rapidly inactivated at negative holding potentials. N-type
channels show an intermediate voltage activation (-40 to -30 mV),
intermediate unitary conductances (about 13 pS) and sensitivity to
w-conotoxin GVIA. These calcium channels may also be characterized by
their sensitivity to inorganic cations. Cadmium (Cd2+) is by far the
most potent, blocking L- and N-type channels with a K about ten times more is needed to block T-type channels (Fox et al,
1987a,b). The more recently described P-type channel (named to honor
the Purkinje cell in which it was first described) shows little inacti
vation, has a voltage-dependence between those of N- and L-type chan
nels and is not blocked by w-conotoxin or by dihydropyridines, but is
blocked by a component of funnel-web spider venom, FTX or w-agatoxin
IVA (LIinas et al., 1989).
It has been shown that N-, L- and P-type channels can all contrib
ute to rapid Ca2+ influx associated with evoked transmitter release
(see Scott et al., 1991). It has also been demonstrated that multiple
types of Ca2+ channels can coexist in a single neuron (reviewed in
Miller, 1987), and that more than one channel type can be involved in
initiation of exocytosis from a single cell type (Artalejo et al.,
1994; Regehr and Mintz, 1994; Takahashi and Momiyama, 1993).
In the presynaptic calyciform nerve terminal of the chick ciliary
ganglion, several investigators have reported the presence of "N-like"
currents in the calical nerve terminal, yet no evidence for L- or
T-type currents (Stanley, 1991; Stanley & Atrakchi, 1990; Stanley and
Goping, 1991; Yawo and Momiyama, 1993). These investigators have sug
gested that the Ca2+ currents they described are acting to initiate
transmitter release in the chick ciliary ganglion.

47
Although Ca2+ channel classification schemes are useful in terms of
defining a point of reference (for comparison and discussion), there
appears to be such diversity in Ca2+ channel structure and function
that overlap between these subtypes (and ensuing subclassification) is
rendering these simple classifications insufficiently descriptive (as
discussed in recent reports: Bertolino and Llinas, 1992; Scott et al.,
1991; Stanley and Goping, 1991). It is, therefore, important to com
plement pharmacological classification studies with evaluations of the
functional properties of presynaptic Ca2+ channels. Experiments in
this chapter will address the role of presynaptic Ca2+ channels in
evoked transmitter release in the chick ciliary ganglion. To test the
contributions of different Ca2+ channel types to evoked transmitter
release, the pharmacological sensitivity of the release process to Ca2+
channel blocking agents is investigated.
Methods
Extracellular recording techniques were employed as described in
Chapter 2 (Methods), w-conotoxin (Bachem, Torrance, CA) was dissolved
in deionized H2O (stock concentration, 500 uM) and frozen in 30 jjI
aliquots (-20 C). Divalent cations were obtained as salts (Sigma, St.
Louis). All drugs were dissolved in Tyrode solution before being
applied. Due to high cost and limited availability, toxins were added
directly to small amounts of Tyrode and oxygenation was maintained by
bubbling O2 directly into the bath. Under control conditions, this
method of oxygenation had no effects on ganglionic transmission.
Orthodromic stimulation of the ciliary ganglion and subsequent
extracellular recording of the postganglionic compound action potential
were used to assay synaptic activity. The amplitude of the post-

48
ganglionic response has been shown to reflect changes in synaptic effi
cacy in this preparation (Landmesser and Pilar, 1972; Marwitt, Pilar
and Weakly, 1971; Poage and Zengel, 1993; Stanley and Goping, 1991).
Low frequency stimulation (0.1/sec or slower) was used to obtain con
trol values of postganglionic response amplitude and to test the
effects of drugs and divalent cations. Data are also presented from
experiments using more complex repetitive stimulation paradigms (see
Chapter 2). In these cases, sufficient time was allowed between trials
to allow the response to recover to pre-stimulus levels.
Results
The effects of Cd2+, Co2+, Ni2+ and Mg2+ on synaptic transmission
through the ciliary ganglion are shown in Figure 3-1. In separate
experiments, these ions were added to the bathing solution and their
effects on ganglionic transmission were observed. The addition of
these divalent cations led to a decrease in the amplitude of the chemi
cal component of the postganglionic compound action potential. Cd2+
was more than 2 orders of magnitude more potent in decreasing gan
glionic transmission than Co2+, Ni2+ and Mg2+ (Figure 3-1). This order
of potency of these divalent cations and the concentrations used to
impede ganglionic transmission are similar to those reported to block
Ca2+ currents in mouse neuromuscular junction (Penner and Dreyer,
1986), rat brain synaptosomes (Lentzner et al., 1992) and squid giant
synapse (Llinas et al., 1981) and synaptic transmission at the frog
neuromuscular junction (Zengel et al., 1993a). The effects of these
ions were reversible by washing with control Tyrode solution.
The effects of w-conotoxin GVIA were tested by adding small amounts
of stock solution to a static bath (see Methods). Application of 1 jjM

49
[divalent cation] mM
Figure 3-1. Concentration-dependence of the effects of divalent
cations on postganglionic response amplitude. Values are expressed as
percent control compound action potential amplitude, with the exception
of the Mg2+ data, which is compared to the lowest [Mg2+] applied (2
mM). Points with standard error bars indicate data averaged from 2 to
4 experiments. [Ca2+] = 5 mM, [Mg2+] = 2 mM unless otherwise noted.

50
w-conotoxin (Figure 3-2, circles) led to an irreversible decrease in
the amplitude of the chemically mediated portion of the compound action
potential. Higher concentrations of w-conotoxin acted more rapidly
(2 juM, triangles in Figure 3-2). Concentrations of 1 juM or greater
usually led to a complete block of the chemical component of the post
ganglionic response within 90 minutes (4 of 5 experiments).
Application of 5-10 juM verapamil, a phenyl alkyl amine Ca2+ channel
blocker, caused a very small decrease in the amplitude of the compound
action potential (4% to 7%) that was reversed by washing with control
Tyrode (n = 2 experiments). While it is possible that this effect is
due to a direct effect of verapamil on Ca2+ currents, it has been
reported that higher concentrations of verapamil (20 to 50 pM) can
block voltage-activated Na+ currents (Chang et al., 1988). Such an
effect would be expected to affect the amplitude of the compound
action potential.
In 2 experiments, 10 pM nifedipine, a dihydropyridine Ca2+ channel
blocker, had no effect on transmission through the ciliary ganglion.
Discussion
The results presented in this chapter are consistent with other
studies (e.g. Bennett and Ho, 1991; Stanley, 1989; Yawo and Momiyama,
1993) in which it was suggested that the "N-like" voltage-activated
Ca2+ currents described in the dissociated calyx preparation are
responsible for initiation of transmitter release in the chick ciliary
ganglion. Concentrations of w-conotoxin shown to have a maximal effect
on presynaptic Ca2+ currents (Stanley and Atrachki, 1991) blocked
chemically mediated synaptic transmission through the ciliary ganglion.
Similarly, Cd2+ blocked synaptic transmission at concentrations that

51
Figure 3-2. Time course of the effects of 2 concentrations of
w-conotoxin on ganglionic transmission. Either 1 uM (circles) or 2 jjM
w-conotoxin (triangles) was added to the Tyrode bathing solution at
time = 0. Each point represents the averaged amplitude of 4 to 16
consecutive responses and is normalized to the amplitude of the
response before toxin addition. The block of ganglionic transmission
was not reversed by 70 to 160 minutes of perfusion with toxin-free
Tyrode. [Ca2+] = 5 mM, [Mg2+] = 2 mM.

52
have been shown to completely block presynaptic Ca2+ currents in the
calyx (Stanley and Goping, 1991; Yawo and Momiyama, 1993). Application
of L-type Ca2+ channel blockers (verapamil and nifedipine) had little
or no effect on ganglionic transmission. These results suggest that
Ca2+ channels involved in release in this preparation are most similar
to N-type channels. Furthermore, it is reported here that the divalent
cations Cd2+, Co2+ and Ni2+ blocked ganglionic transmission with the
same order of potency as described in other synaptic preparations in
which N-type channels are involved in initiation of transmitter release
(Lentzner et al., 1992; Llinas et al., 1981; Penner and Dreyer, 1986).
The majority of the total presynaptic Ca2+ current in the calyci-
form nerve terminal is w-conotoxin sensitive (Stanley and Atrakchi,
1991; Yawo and Momiyama, 1993). An w-conotoxin insensitive component
to the calyx l£a has been reported (Stanley and Atrakchi, 1990; Yawo
and Momiyama, 1993). This component can support low levels of trans
mitter release when 4-aminopyridine is added to increase the presynap
tic action potential duration (Yawo and Chuhma, 1994). This w-cono
toxin insensitive component is blocked by 50 uM Cd2+. The role of such
a small Ca2+ current in release could not be assayed with the extracel
lular recording method used in this study. Under the conditions used
here, little if any transmitter release is supported by this w-cono
toxin insensitive current. The chemically mediated portion of the com
pound action potential is not present unless EPSP amplitude is large
enough to initiate an action potential in the postsynaptic cells in
the ganglion.

53
It is generally accepted that transmitter release from vertebrate
peripheral synapses is initiated primarily by current flow through
N-type Ca2+ channels (reviewed in Miller, 1987). N-type channels, but
not channels of other types, have been shown to be physically connected
to proteins in the presynaptic terminal that act to regulate docking of
synaptic vesicles and priming of vesicles for release (reviewed in Ben
nett and Scheller, 1993). If these active zone proteins associate
exclusively with N-type Ca2+ channels, as has been suggested (Mastro-
giacomo et al., 1994), the limited role of other channel types to
exocytosis may be related to their distance from the release machinery.
In support of this, it has been reported (Stanley, 1993), that the
coupling of Ca2+ influx to acetylcholine release in the ciliary gan
glion appears to be conserved within a 3 pm2 membrane patch. Stanley
further suggested that Ca2+ influx through a single channel is suffi
cient to trigger quantal transmitter release, implying that Ca2+ influx
through N-type channels occurs extremely close to the transmitter
release mechanism in the chick ciliary ganglion.
In summary, I have characterized the effects of Ca2+ channel block
ing agents (divalent cations, dihydropyridine and phenyl alkyl amine
antagonists, and w-conotoxin GVIA) on ganglionic efficacy in the embry
onic chick ciliary ganglion. The results presented here further
strengthen the argument that the w-conotoxin sensitive Ca2+ currents
previously described in the presynaptic nerve terminal represent the
major source of Ca2+ elevation for initiation of transmitter release
from the calyx.

CHAPTER 4
EFFECT OF REPETITIVE STIMULATION ON THE PRESYNAPTIC ACTION POTENTIAL
When considering possible mechanisms for stimulation-induced
increases in release, there are several points in the excitation-
secretion cascade that appear to be obvious candidates. One of these
is the presynaptic action potential. If the depolarization of the
nerve terminal caused by the invasion of an action potential is larger
in amplitude or is prolonged, this could result in an increased
recruitment of voltage-dependent Ca2+ channels and a greater influx of
Ca2+, leading to increased release.
Due to their small size, presynaptic elements of most synapses are
difficult to study in an intact synapse. However, there are certain
synaptic preparations that have extraordinarily large presynaptic ter
minals. In several of these preparations the relationship between
action potential parameters and transmitter release has been investi
gated. In the squid giant synapse, increasing the duration of the pre
synaptic action potential by pharmacological means leads to an increase
in the amplitude of the postsynaptic response (Augustine, 1990). In
sensory neurons of Aplvsia, presynaptic facilitation of release appears
to be mediated by a decrease in a specific K+ current that prolongs the
action potential, leading to an increase in Ca2+ influx (Klein et al.,
1982; Sugita et al., 1992). These results suggest that changes in the
presynaptic action potential may represent a viable mechanism for regu
lation of synaptic efficacy (see Augustine, 1990).
54

55
Martin and Pilar (1964a-c) employed intracellular recording from
the pre- and postsynaptic cells of the chick ciliary ganglion to inves
tigate mechanisms of stimulation-induced increases in release. They
reported no significant change in the presynaptic nerve terminal action
potential under conditions where increased release was observed. How
ever, those studies compared the amplitudes of individual action poten
tials, and it is possible that subtle changes in the action potential
amplitude and/or duration would not have been detected by this type of
analysis. It was therefore of interest to perform experiments of a more
quantitative nature to investigate the relationship between changes in
the presynaptic action potential and the individual processes of stimu
lation-induced increases in release described an Chapter 2.
Methods
Standard intracellular recording techniques were employed
(described in detail in Chapter 2, Methods). Cells were identified as
calyciform nerve endings or ciliary neurons using the electrophysiolog-
ical criteria of Martin and Pilar (1963a, 1964a). Briefly, stimulation
of the oculomotor nerve produced an action potential in both calyces
and ciliary neurons. When spike initiation was blocked by injection of
hyperpolarizing current, the ciliary neurons exhibited an underlying
excitatory postsynaptic potential (EPSP) (Figure 4-1, upper record).
Nerve terminals showed no EPSP following orthodromic stimulation when
hyperpolarizing current was injected. The only remaining response was
an attenuated action potential, which was presumably conducted electro-
tonically from outside the region blocked by hyperpolarization (Martin
and Pilar, 1963a). Calyces responded to antidromic stimulation with a

56
coupling potential, representing current flow through electrical syn
apses (Martin and Pilar, 1963a).
In paired pulse experiments, trials consisted of a conditioning
stimulus and a testing stimulus applied some time after the condition
ing stimulus. The conditioning stimulus serves as a control to which
test responses are compared. Typical responses to paired stimuli from
a ciliary neuron and calyx are presented in Figure 4-1. In experiments
in which trains of impulses were applied, changes in the response are
described relative to the first response of the train.
The parameters were defined as follows: action potential amplitude
was measured from resting membrane potential to the most depolarized
point of the action potential; time to action potential peak was meas
ured from the time of the last available baseline control point (the
last point at resting membrane potential before the stimulus artifact)
to the time of the action potential peak; action potential duration was
measured from the time of the last baseline point to the point where
the repolarizing action potential crossed the original resting poten
tial; afterhyperpolarization (AHP) amplitude was measured from the
resting membrane potential (prior to the action potential) to the most
hyperpolarized point during the AHP; AHP half-decay time was measured
by finding the peak hyperpolarization and determining the time for the
voltage to decay to half of that level.
Results
Effects of Repetitive Stimulation on Presvnaptic Potentials
As shown in Figure 4-1, both presynaptic and postsynaptic responses
were affected by paired pulse stimulation. At low levels of release,
paired pulse stimulation led to an increase in the amplitude of the

57
Post-synaptic ciliary neuron
Pre-synaptic nerve terminal
Figure 4-1. Examples of electrophysiological responses of pre- and
postsynaptic cells of the embryonic ciliary ganglion to paired pulse
stimulation. A postsynaptic ciliary neuron is identified by the pres
ence of an excitatory postsynaptic potential (EPSP) in response to
orthodromic stimulation (upper record). Hyperpolarizing current was
injected through the recording electrode to keep the EPSP below thresh
old for action potential generation. [Ca^+] = 2.5 mM, [Mg^+J = 4 mM. A
presynaptic nerve terminal responds to orthodromic stimulation with an
action potential (lower record). [Ca^+J = 3 mM, [Mg^+j = 2 mM.

58
second of a pair of EPSPs recorded from a postsynaptic ciliary ganglion
neuron (Figure 4-1, upper record; see Chapter 2 for description of
facilitation). Repetitive stimulation also caused an increase in pre-
synaptic action potential duration (control = 3.85 msec, test = 4.05
msec) and a decrease in AHP amplitude (control = 9 mV, test = 6.5 mV;
Figure 4-1, lower record). These changes are illustrated more clearly
in Figure 4-2 where the control and test responses from Figure 4-1
are superimposed.
During short trains of stimulation (10 impulses applied at 20/sec),
changes in the presynaptic action potential are apparent early in the
train and appear to reach a steady state (Figure 4-3). These changes
are more clearly seen in Figure 4-4, which plots averaged data from the
same cell as Figure 4-3. Repetitive stimulation caused an increase in
action potential duration (A) and decreases in both AHP amplitude (B)
and AHP half decay time (C). There were no consistent effects of
repetitive stimulation on action potential amplitude or time to peak
(data not shown).
Ionic Mechanisms Underlying Action Potential Repolarization and AHP
It is generally accepted that the repolarizing and afterhyperpolar
izing phases of the action potential are dependent on the activation of
K+ channels and a subsequent K+ efflux. To test the ionic basis of the
afterhyperpolarization, the membrane potential was altered by injecting
current through the recording electrode and the resulting changes in
AHP amplitude were observed. Figure 4-5 shows results from a single
cell in which the resting membrane potential was varied from -48 mV to
-95 mV. As the membrane potential approached the expected equilibrium
potential for K+ (-90 mV), the amplitude of the AHP was reduced to a

59
4 msec
10 mV
test
Figure 4-2. Nerve terminal action potentials evoked by paired pulse
stimulation. Responses are superimposed to illustrate changes in repo
larization phase. Same cell and interstimulus interval as Figure 4-1.
Single responses are shown (not averaged).

60
Figure 4-3. Effect of a short train of orthodromic stimuli on the
presynaptic action potential. Each stimulus generates an action poten
tial in the presynaptic nerve terminal. Stimulation rate for this
example is 20/sec. [Caz+] = 2.5 mM, [Mg2+] = 4 mM.

61
A
B
C
Figure 4-4 Parameters describing nerve terminal action potentials
during 10 impulse trains. The average of three identical trials from a
single cell (see Figure 4-3) are presented. Parameters were measured
as described in Methods. Changes in successive action potentials are
expressed as the percent of values obtained for the first response in
the train. (A): action potential duration. (B): AHP amplitude. (C):
AHP half-decay time. [Ca2+] = 2.5 mM, [Mg2+] = 4 mM.

62
Membrane potential (mV)
Figure 4-5. Effects of varying nerve terminal membrane potential on
afterhyperpolarization (AHP) amplitude. Data are from a single cell in
which the membrane potential was varied by injecting current through
the recording electrode. Line drawn through the data points by eye.
[Caz+] = 1.5 mM, [Mg2+] = 4 mM. Resting membrane potential = -52 mV.

63
point where it became unmeasurable. This finding is consistent with
a role of K+ currents in generation of the AHP in the presynaptic
nerve terminal.
Due to the similar mechanisms underlying action potential repolar
ization and AHP generation, it is not unreasonable to assume that a
single underlying change in the presynaptic terminal is affecting both
AHP amplitude and action potential duration. Figure 4-6 plots the
relationship between changes in action potential duration and changes
in AHP amplitude following paired pulse stimulation. As AHP amplitude
decreases, there is a clear increase in action potential duration.
The Presynaptic Action Potential and Facilitation
An ideal way to describe the correlation between changes in the
nerve terminal action potential and increases in transmitter release
would be to record simultaneously from pairs of pre- and postsynaptic
cells during repetitive stimulation. While there has been a report of
successful simultaneous penetration of both elements in the chick cil
iary ganglion, this procedure yielded only a few very brief recordings
(Yawo and Momiyama, 1993). Instead, the relationship between intracel-
lularly recorded changes in the presynaptic action potential and effi
cacy of release was investigated in this study using extracellular
records of postsynaptic compound action potentials. It has been shown
that changes in the compound action potential during repetitive stimu
lation accurately reflect changes in EPSP amplitude under the condi
tions used here (see Chapter 2).
Figure 4-7 plots the relationship between changes in intracellu-
larly recorded presynaptic action potential parameters and facilitation
of the postganglionic compound action potential. Data were collected

64
c
% control AHP amplitude
Figure 4-6. Correlation between measures of presynaptic action
potential duration and AHP amplitude. Paired pulse stimulation pro
duced an increase in action potential duration and a decrease in AHP
amplitude (see text). The effects of a single conditioning impulse on
action potential duration (ordinate) are compared to its effects on AHP
amplitude (abcissa). A regression line is drawn through the data,
which are results of 8 intervals applied in various combinations to 7
different cells.

65
Figure 4-7. Correlation between facilitation of extracellularly
recorded compound action potential and changes in intracellularly
recorded presynaptic action potential. Each point represents the aver
age of 4 to 16 responses from a single cell. Data presented here are
from experiments done under low quantal conditions ([Ca2+] = 1.25 to
1.5 mM, [Mg2+] = 4 mM) where simultaneous measures of intracellularly
recorded presynaptic action potential and extracellularly recorded com
pound action potentials were obtained (V(t) calculated as described in
Chapter 2). Data are results of 8 different interstimulus intervals
applied to 7 different cells. (A): Relationship between extracel1ularly
recorded V(t) and changes in presynaptic action potential duration. A
regression line is drawn through the data (R= 0.62). (B): Relationship
between extracellularly recorded V(t) and presynaptic AHP amplitude. A
regression line is drawn through the data (R = 0.84).

66
from 7 cells at interstimulus intervals between 25 and 2000 msec. Fig
ure 4-7A plots V(t) as a function of action potential duration. Figure
4-7B plots V(t) as a function of AHP amplitude. Although changes in
synaptic efficacy appear to correlate with both action potential dura
tion and AHP amplitude, the correlation with AHP amplitude was stron
ger, probably because of the greater reliability of measures of AHP
amplitude (see Note 1 at the end of this chapter). These results show
that stimulation-induced changes in the presynaptic action potential
and increases in transmitter release occur simultaneously in the chick
ciliary ganglion.
The effect of repetitive stimulation on the presynaptic action
potential is most pronounced in the first 150 msec of a 20/sec train
(Figure 4-5). The first component of facilitation (Fj), which has a
time constant of decay of about 60 msec in the chick ciliary ganglion
(see Chapter 2), seems to accumulate in a similar manner to that
described in the frog neuromuscular junction (Figure 2-3; Magleby and
Zengel, 1982). If this is the case, then the time course of accumula
tion of F¡ would be quite similar to the time course of the effects of
repetitive stimulation on the presynaptic action potential shown in
Figure 4-5. To determine whether changes in the presynaptic action
potential correlate with facilitation of transmitter release, a paired
pulse paradigm (like that used to describe facilitation in Chapter 2)
was employed to characterize changes in the presynaptic action poten
tial under conditions shown to produce facilitation.
Figure 4-8 shows the effect of a single conditioning stimulus on
the presynaptic action potential. In this experiment, paired pulses
were applied at seven different intervals. At shorter intervals, action

67
Figure 4-8. Effects of paired pulse stimulation on the presynaptic
action potential. Traces represent the average of 2 to 6 identical
trials. Interstimulus intervals are indicated. [Ca2+] = 1.4 mM,
[Mg2+] = 4 mM.

68
potential duration was increased and AHP amplitude was dramatically
decreased, whereas at longer intervals the effects on AHP amplitude and
action potential duration were less pronounced. The time course of the
effects on AHP amplitude are more clearly seen in Figure 4-9. Like
facilitation of transmitter release, the effects of a single condition
ing impulse on AHP amplitude were greatest at short interstimulus
intervals (less than 150 msec) and less pronounced at longer intervals
(Figure 4-9A). Figure 4-9B plots the same data as in (A), expressed as
percent inhibition of AHP amplitude. In three experiments where 5 or
more intervals were tested, the decay of the effect on AHP amplitude,
plotted as percent AHP inhibition, could be described by a dual expo
nential decay with time constants of 64+20 msec and 1219+292 msec. In
experiments where fewer than 5 intervals were applied, the shortest
interstimulus intervals (<200 msec) consistently produced the greatest
effects on the presynaptic action potential, similar to the results
presented in Figure 4-9.
Discussion
The goal of this study was to record from the presynaptic nerve
terminal of the chick ciliary ganglion during repetitive stimulation
and to investigate the relationship between the presynaptic action
potential and transmitter release. It is shown that repetitive stimu
lation, under conditions that produce facilitation of transmitter
release, gives rise to changes in the presynaptic action potential that
parallel facilitation. This effect is greatest during the first few
hundred milliseconds following a single conditioning stimulus, a time
when the facilitation phase of increased release is prevalent. Stimu
lation-induced changes in presynaptic action potential waveform appear

69
Interstimulus interval (msec)
Figure 4-9. Effect of a single conditioning impulse on presynaptic
action potential AHP amplitude. Data points represent averages of 4 to
10 identical trials. (A): Effects of paired pulse stimulation on AHP
amplitude. Data are expressed as percent of the control AHP amplitude.
(B): Same data as (A) expressed as percent maximal inhibition of AHP
amplitude and plotted on a semilogarithmic scale. A regression line
drawn through the linear portion of the data (points beyond 150 msec)
gives a line described by the equation Ts-|ow. The values of the slow
regression line at earlier points (<150 msec) were calculated and the
contributions of the slower regression were subtracted assuming an
additive relationship between the two decays. A regression line
through the resulting points (squares) is described by the equation
Tfast- tCa2+] = 1-2 mM, [Mg2+] = 4 mM.

70
to be a common phenomenon in neurons and other electrically excitable
cells (e.g. Bourque, 1991; Crest and Gola, 1993; Quattrocki et al.,
1994) and may represent a common mechanism for modulation of release
during and following repetitive stimulation.
The Action Potential
Before considering the implications of this study, a brief descrip
tion of the currents that comprise the action potential is in order.
Several overlapping currents comprise the action potential (Hodgkin and
Huxley, 1939). The depolarizing phase of the action potential results
from an increase in Na+ permeability and subsequent entry of Na+ ions
down a strong electrochemical gradient. The falling or repolarizing
phase of the action potential involves a decrease in Na+ permeability
(Na+ channel inactivation) and an increase in permeability to K+ due to
the opening of voltage-dependent K+ channels. The increase in K+
permeability can last for several milliseconds, so that in many cells
K+ efflux can hyperpolarize the membrane beyond the resting potential,
producing an afterhyperpolarization (AHP).
Repetitive Stimulation and the Action Potential
Martin and Pilar (1964c) looked at the effects of repetitive stimu
lation on the presynaptic action potential of the chick ciliary gan
glion. A paired pulse paradigm like the one used in the current study
was employed. The only effect they reported was a depression of action
potential amplitude at interstimulus intervals of 5 msec or less, (see
Figures 8 and 9, Martin and Pilar, 1964c); no measures of action poten
tial duration or AHP amplitude were reported. During stimulation at
higher frequencies than were applied here (50/sec for 20 sec), action
potential amplitude decreased (Figure 11, Martin and Pilar, 1964c).

71
Several factors may have contributed to differences between the
earlier results of Martin and Pilar and those presented here. When
Martin and Pilar performed their classic experiments describing elec
trical and chemical transmission through the ciliary ganglion, they
compared single action potentials. The results presented here are
taken from averaged data, which should make small changes in action
potential waveform more easily apparent. Also, in the 1960s, current
theories about the role of the action potential in facilitation seem to
have focussed primarily on changes in action potential amplitude (e.g.
Hubbard and Schmidt, 1963; Takeuchi and Takeuchi, 1962). For this rea
son, perhaps, the effects of repetitive stimulation on action potential
amplitude were more vigorously examined than other aspects of the
action potential.
In recent studies of a number of neuronal preparations, repetitive
stimulation has been shown to have effects on the action potential
waveform (reviewed in Nicholls et al., 1992). For example, Charlton and
Bittner (1978b) reported that repetitive stimulation of the squid giant
synapse can lead to an increase in the amplitude of the presynaptic
action potential, but only during the first few responses in a train.
The effects of repetitive stimulation on the duration of the action
potential were not reported, although their data appear to show a
decrease in AHP amplitude during tetanic stimulation (Figures 3 and 11,
Charlton and Bittner, 1978b). Similarly, at the crayfish neuromuscular
junction (Bittner and Baxter, 1991) it has been shown that repetitive
stimulation causes an increase in the duration and amplitude of the
presynaptic action potential for the first 3 or 4 impulses in a train.

72
Mechanisms for Action Potential Broadening During Repetitive Stimulation
It is not clear what electrical events bring about the increases in
action potential duration reported in this chapter. One possible mech
anism involves activation of Ca2+-dependent cation channels (Partridge
and Swandulla, 1988). Although these channels have been described in
chick sensory neurons, their activation range (over 1 pM; Razani-
Boroujerdi and Partridge, 1993) raises doubts as to their role in neur
onal function under more normal ionic conditions. A more likely mecha
nism for broadening the presynaptic spike is a stimulation-induced
inhibition of K+ current(s).
Figure 4-7 shows that the reversal potential for the calyx AHP lies
close to the predicted equilibrium potential for K+. In the ciliary
ganglion, several K+ currents have been observed in the presynaptic
nerve ending, including delayed rectifier (Dryer and Chiappinelli,
1985), Ca2+-activated (Fletcher and Chiappinelli, 1992b, 1993) and
inwardly rectifying K+ currents (Dryer and Chiappinelli, 1985; Fletcher
and Chiappinelli, 1992a).
There are several mechanisms through which repetitive stimulation
could affect K+ currents. One possibility is that the accumulation of
[Ca2+]i during stimulation (e.g. Charlton et al., 1982; Smith and
Augustine, 1988; Stinnakre and Tauc, 1973) could act to decrease K+
efflux. There have been reports of Ca2+-inhibited K+ channels in many
cell types (see Marty, 1989). For example, increases in intracellular
Ca2+ can decrease the opening probability of K+ channels in skeletal
muscle (Vergara and Latorra, 1983), and can inhibit BK K+ currents in
rat exocrine cells (Marty et al., 1984).

73
There are also reports that describe cumulative stimulation-induced
inactivation of K+ currents. In molluscan neurons, cumulative inacti
vation of a fast delayed rectifier K+ current contributes to stimula
tion-induced action potential broadening (Aldrich et al., 1979; Thomp
son, 1977). A delayed rectifier channel cloned from rat brain (Marom
and Levitan, 1994) also shows similar inactivation properties. Edry-
Schiller and Rahamimoff (1993), on the basis of data obtained from the
fused Torpedo synaptosome preparation, have proposed a "potassium
inactivation hypothesis" for frequency modulation of transmitter
release. They suggest that reactivation of slow K+ channels following
an action potential may be responsible for action potential broadening
and facilitation of release. Mallart (1985) proposed a similar role
for the inactivating Ca2+-activated K+ channel in the motor nerve
terminal of the mouse. In the chick ciliary ganglion, cumulative K+
current inactivation has been observed in the postsynaptic ciliary
neuron (S. Dryer, personal communication) and it is reasonable to
assume that similar K+ channels could be present in the presynaptic
nerve terminal.
Action Potential Duration: Effects on Release
Although the mechanisms underlying stimulation-induced increases in
action potential duration in the chick ciliary ganglion are not fully
understood, the fact remains that many studies have shown that changes
in presynaptic action potential duration can have large effects on syn
aptic transmitter release (e.g. Augustine, 1990; Hochner et al., 1986;
Robitailie and Charlton, 1992). In the squid giant synapse, applica
tion of the K+ channel blocker diaminopyridine had concentration-
dependent effects on action potential duration and transmitter release:

74
a ten percent increase in action potential duration resulted in a 60
percent increase in the amplitude of the postsynaptic current
(Augustine, 1990). Hochner and colleagues (1986) report similar
results in AdIvsia sensory ganglia: a 10% to 30% increase in presynap-
tic action potential duration was correlated with a 25% to 120%
increase in postsynaptic potential amplitude. A similar relationship
between presynaptic action potential duration and increased synaptic
efficacy can be seen in the chick ciliary ganglion (Figure 4-7).
Current Hypotheses: Stimulation-Induced Increases in Release
Some of the earliest theories concerning mechanisms of stimulation-
induced increases in transmitter release involved the presynaptic
action potential (Bloedel et al., 1966; Hubbard and Schmidt, 1963;
Miledi and Slater, 1966; Takeuchi and Takeuchi, 1962). Experimental
testing of this hypothesis has shown that changes in the presynaptic
action potential cannot account for all of the observed increases in
transmitter release during and following repetitive stimulation (Charl
ton and Bittner, 1978b; Zucker, 1974). However, these results should
not be interpreted as eliminating the possibility that changes in the
action potential play a role in mediating the effects of repetitive
stimulation on transmitter release. Several individual components,
which appear to arise from separate mechanisms, contribute to stimula
tion-induced increases in release. It has been suggested that differ
ent Ca^+-dependent processes are involved in initiation and modulation
of release (Bain and Quastel, 1992b; Zengel et al., 1993a,b). The
effects of repetitive stimulation on action potential duration probably
only represent one of many effects of repetitive stimulation.

75
Recent reports from this laboratory suggest that facilitation of
transmitter release is related to entry of Ca2+ through voltage-
activated channels (Zengel et al., 1993a,b), whereas augmentation
appears to be particularly sensitive to the intraterminal concentration
of Ca2+ (Zengel et al., 1994). Other mechanisms underlying stimula
tion-induced modulation of release (for example, changes in Ca2+ chan
nel kinetics [Lee, 1987], Ca2+ buffering [Llinas et al., 1992; Neher
and Augustine, 1990], phosphorylation of nerve terminal and vesicular
proteins [reviewed in Greengard et al., 1993]) may be contributing to
stimulation-induced increases in transmitter release in the ciliary
ganglion. Unfortunately, these processes are not readily observed by
electrophysiological means.
The "residual Ca2+" hypothesis (Katz and Miledi, 1968) proposes
that Ca2+ accumulates in the nerve terminal during repetitive stimula
tion, and that this cumulative increase in Ca2+, or some Ca2+-sensitive
process, contributes to stimulation-induced increases in release. The
findings reported here are consistent with a role for residual Ca2+ in
facilitation. Stimulation-induced broadening of the presynaptic action
potential should cause a significant increase in Ca2+ influx during
repetitive stimulation. The time course of stimulation-induced changes
in action potential duration and AHP amplitude indicate that, if these
changes are acting to bring about increases in neurotransmitter
release, they are most likely to be involved in facilitation, as
described in this and other preparations.
Notes
1 Because transmitter release has been shown to be a sensitive
function of action potential duration, describing the effects of repet-

76
itive stimulation on this parameter was a primary goal of this study.
Unfortunately, due in part to the length of the preganglionic nerve and
the rapid conduction of the action potential in this preparation, the
early rising phase of the presynaptic action potential was often tem
porally superimposed on the stimulus artifact (see Figure 4-2), making
precise measurement of action potential duration difficult. This tech
nical problem may have added an element of variability to measures of
action potential duration, even though changes were always expressed
relative to a paired control response. Due to the rapid time course of
the stimulus artifact, measures of AHP amplitude were not affected.

CHAPTER 5
CONCLUSIONS
As recently as 100 years ago, energetic debate surrounded hypoth
eses concerning the functional organization of the nervous system. The
proponents of cell theory suggested that nerves were independent units,
a proposal that went against the prevalent theory that all nerves were
a single continuous structure, part of a syncytium, interconnected by
protoplasmic bridges (see Nicholls et al., 1992, p. 185). Later, it
was accepted that nerves were discontinuous, separated by small gaps
across which information was passed through unknown methods. Otto
Loewi, in 1921, performed a simple and convincing series of experiments
showing that stimulation of the vagus nerve acts to slow heart rate by
releasing a diffusible substance (acetylcholine). Dale and others soon
established the role of acetylcholine as a neurotransmitter at the
neuromuscular junction and in autonomic ganglia (e.g. Dale and Feld-
berg, 1936). These experiments led to the general acceptance of chemi
cal synaptic transmission between nerve cells and effector cells.
77

78
In the 1960s, Bernard Katz and others showed that initiation of
transmitter release is dependent upon the presence of Ca2+ ions in the
extracellular medium (Katz and Miledi, 1965, 1967). These studies and
others led to the formulation of the "Ca2+ hypothesis", which proposed
that the entry of Ca2+ ions into the presynaptic terminal is an essen
tial step linking membrane depolarization to transmitter release.
Despite the wide acceptance of the Ca2+ hypothesis, recent reports
have suggested that other factors may be sufficient to initiate exocy-
tosis in the absence of extracellular Ca2+. For example, Stuenkel and
Nordmann (1993) report Na+-dependent neuropeptide release from the rat
neurohypophysis in the absence of a rise in intracellular Ca2+. Pamas
and co-workers have investigated the role of nerve terminal depolariza
tion in initiation of transmitter release at the frog neuromuscular
junction and have suggested that a depolarization-dependent factor pro
motes release in cooperation with intracellular Ca2+ (e.g. Hochner et
al., 1989; Pamas et al., 1986). Other investigators have also
reported depolarization-induced transmitter release that appears to
occur in the absence of Ca2+ influx (Mosier and Zengel, 1993; Sil insky
et al., 1995). These reports suggest that although intracellular Ca2+
is necessary for most forms of exocytosis, it is not the sole factor
acting to initiate transmitter release.
As described in previous chapters, four components of stimulation-
induced increases in synaptic transmitter release have been reported at
many synapses (e.g. Erulkar and Rahamimoff, 1978; Mallart and Martin,
1967; Rosenthal, 1969; Zengel and Magleby, 1982). All four of these
processes (two components of facilitation, augmentation and potentia
tion) are acting to increase EPSP amplitude during repetitive stimula-

79
tion in the chick ciliary ganglion (Chapter 2). Many investigators
have suggested that separate underlying mechanisms bring about the four
components of increased release (e.g. Landau et al., 1973; Lev-Tov and
Rahamimoff, 1980; Magleby, 1973; Magleby and Zengel, 1982). It has
been suggested that the effects of repetitive stimulation are a
consequence of an increase in residual intracellular Ca2+ or a calcium
activated factor (Ca*) that causes a given presynaptic depolarization
to release an increased amount of transmitter (Katz and Miledi, 1965,
1968). This "residual calcium hypothesis" is the most widely accepted
theory to account for stimulation-induced increases in release. In its
simplest form, however, this model fails to adequately account for all
of the properties of stimulation-induced increased increases in release
(e.g. Zengel and Magleby, 1980, 1981, 1982; Bain and Quastel, 1992a).
Although Ca2+ ions appear to play an important role in the facilitation
phase (Katz and Miledi, 1968; Zengel et al., 1993a,b) and augmentation
phase (Erulkar and Rahamimoff, 1978; Magleby and Zengel, 1976; Zengel
et al., 1994) of increased release, the potentiation phase does not
appear to involve Ca2+; instead, it has been suggested that an accumu
lation of Na+ ions in the nerve terminal may be involved in this phase
of increased release (e.g. Birks and Cohen, 1968; Nussinovitch and
Rahamimoff, 1988). If the mechanisms underlying stimulation-induced
increases in release can be described, the release process itself will
be more completely understood.
Data presented in Chapter 4 of this dissertation show that facili
tation of transmitter release in the chick ciliary ganglion is accompa
nied by an increase in the duration of the presynaptic action poten
tial. One interpretation of this finding is that the increase in the

80
duration of the presynaptic depolarization recruits a greater percent
age of voltage-dependent Ca2+ channels in the calyciform nerve terminal
and that the resulting increase in Ca2+ influx underlies facilitation
of transmitter release. The time constant of activation for Ca2+
currents in the calyx is 0.9 1.6 msec at 23 C (Stanley and Goping,
1991) and calcium currents recorded using whole-cell patch-clamp tech
niques do not reach a peak level for several milliseconds following
sustained depolarization (Stanley, 1989). The calyx action potential
only produces a depolarization sufficient to activate Ca2+ channels for
a very short time (1 to 2 msec, Martin and Pilar, 1963a; personal
observations), which would not activate the full complement of presyn
aptic Ca2+ channels. This conclusion is not without precedent. Earlier
studies have also suggested that increased Ca2+ influx by consecutive
stimuli might cause facilitation (e.g. Stinnakre and Tauc, 1973).
Several other areas of study bear mentioning when discussing the
effects of repetitive stimulation on neurotransmitter release. It has
been reported in the frog and crayfish neuromuscular junctions and
other preparations that the relative contribution of the various compo
nents of stimulation-induced changes in release may vary across cells
within an experimental preparation (Collins et al., 1984; Fadoga and
Brookhart, 1962; Frank, 1973; Meriney and Grinnell, 1991). One use of
these systems would be to identify the biochemical or electrophysiolog-
ical differences between, for example, facilitating and non
facilitating synapses.
Another promising area of research on exocytosis involves recently
discovered proteins in the presynaptic nerve terminal (reviewed in Ben
nett and Scheller, 1993 and in De Camilli, 1993) that appear to be

81
involved in synaptic vesicle trafficking, mobilization and exocytosis.
Studies have already begun to describe the role of these proteins in
release of neurotransmitter under control conditions (e.g. Lledo et
al., 1993; Pevsner et al., 1994) and following repetitive stimulation
(e.g. Tarelli et al., 1992). More recently, the use of antisense
nucleotides and transgenic mice has enabled researchers to study synap
tic transmission in synaptic preparations that have modified versions
of these proteins (e.g. Alder et al., 1992) or lack them entirely (e.g.
Rosahl et al., 1993).
The presynaptic nerve terminal is a highly specialized and complex
structure. It contains mechanisms for uptake, storage and synthesis of
transmitter substances, as well as voltage-sensitive proteins that can
alter ionic conditions in response to changes in membrane potential,
and ligand-activated complexes that can respond to cytoplasmic or
extracellular chemical signals. Consequently, there are many ways by
which modulation of the release process could occur, including
increases in intracellular Ca^+ and intracellular Na+, activation of
presynaptic ligand gated receptors, changes in the kinetics of voltage-
activated ion channels and changes in the presynaptic action potential.
Recent technological advances in many elements of neuroscience
research (cell culture, molecular biological, genetic and imaging tech
niques) are increasing the precision with which neurons can be studied.
With this enhanced ability to observe and describe neuronal physiology,
many long-standing questions will soon be answered. It would appear
that the Ca^+-dependent "trigger" mechanism that initiates the process
of neurotransmitter release will be characterized within half a century
of its postulation.

REFERENCES
Alder, J., Xie, Z.-P., Valtorta, F., Greengard, P., and Poo, M.-m.
(1992) Antibodies to synaptophysin interfere with transmitter
secretion at neuromuscular synapses. Neuron. 9:759-768.
Aldrich, R.W., Getting, P.A., and Thompson, S.H. (1979) Mechanism of
frequency-dependent broadening of molluscan neurone somata, jh
Physiol. (London). 291:531-544.
Artalejo, C.R., Adams, M.E. and Fox, A.P. (1994) Three types of calcium
channels trigger secretion with different efficacies in chromaffin
cells. Nature 367:72-76.
Augustine, G.J. (1990) Regulation of transmitter release at the squid
giant synapse by presynaptic delayed rectifier potassium current.
J. Physiol. (London). 431:343-364.
Augustine, G.J., Smith, S.J., and Zucker. R.S. (1982) Role of presynap
tic calcium ions and channels in synaptic facilitation and depres
sion at the squid giant synapse. J. Physiol. (London). 323:173-193.
Bain, A.I., and Quastel, D.M.J. (1992) Quantal transmitter release
mediated by strontium at the mouse neuromuscular junction. J. Phy
siol. (London). 450:63-87.
Bain, A.I., and Quastel, D.M.J. (1992) Multiplicative and additive
Ca2+-dependent components of facilitation at mouse endplates. jL
Physiol. (London). 455:383-405.
Bennett, M.K., and Scheller, R.H. (1993) The molecular machinery for
secretion is conserved from yeast to neurons. PNAS (USA1.
90:2559-2563.
Bennett, M.R., and Ho, S. (1991) Probabilistic secretion of quanta from
nerve terminals in avian ciliary ganglia modulated by adenosine, jh
Phvsiol.. 440:513-527.
Bertolino, M., and Llinas, R. (1992) The role of voltage activated and
receptor operated calcium channels in neuronal cells. Ann. Rev.
Pharmacol. Toxicol.. 32:399-421.
Birks, R.I., and Cohen, M.W. (1968) The influence of internal sodium on
the behaviour of motor nerve endings. Proc. Rov. Soc., 170:401-421.
82

83
Bittner, G.D., and Baxter, D.A. (1991) Synaptic plasticity at crayfish
neuromuscular junctions: facilitation and augmentation. Synapse.
7:235-243.
Bloedel, J., Gage, P.W., Llinas, R., and Quastel, D.M. (1966) Transmit
ter release at the squid giant synapse in the presence of tetrodo-
toxin. Nature, 212:49-50.
Bourque, C.W. (1991) Activity-dependent modulation of nerve terminal
excitation in a mammalian peptidergic system. TINS. 14:28-31.
Chang, C.C., Lin, S.-O., Hong, S.J., and Chiou, L.C. (1988) Neuromuscu
lar block by verapamil and diltiazem and inhibition of acetylcho
line release. Brain Res., 454:332-339.
Charlton, M., and Bittner, G.D. (1978a) Facilitation of transmitter
release at squid synapses. J. Gen. Physiol.. 72:471-486.
Charlton, M., and Bittner, G.D. (1978b) Presynaptic potentials and
facilitation of transmitter release at squid synapses. J. Gen.
Phvsiol.. 72:487-511.
Charlton, M.P., Smith, S.J., and Zucker, R.S. (1982) Role of presynap
tic calcium ions and channels in synaptic facilitation and depres
sion at the squid giant synapse. J. Phvsiol. (London), 323:
173-193.
Collins, W.F., Honig, M.G., and Mendell, L.M. (1984) Heterogeneity of
group la synapses on homonymous alpha-motoneurons as revealed by
high-frequency stimulation of la afferent fibers. J. Neurophysiol.,
52:980-993.
Crest, M., and Gola, M. (1993) Large conductance Ca2+-activated K+
channels are involved in both spike shaping and firing regulation
in Helix neurones. J. Phvsiol. (London). 465:265-287.
Curtis, D.R., and Eccles, J.C. (1960) Synaptic action during and after
repetitive stimulation. J. Phvsiol. (London). 150:374-398.
Dale, H.H., and Feldberg, W. (1936) The chemical transmission of secre
tory impulses to the sweat glands of the cat. J. Phvsiol. (London!,
82:121-128.
DeCamilli, P. (1993) Exocytosis goes with a SNAP. Nature. 364:387-389.
DeLorenzo, A.J. (1960) The fine structure of synapses in the ciliary
ganglion of the chick. J. Biophvs. Biochem. Cvtol.. 7:31-36.
del Castillo, J., and Katz, B. (1954) Statistical factors involved in
neuromuscular facilitation and depression. J. Physiol. (London!.
124:574-585.

84
Dryer, S.E., and Chiappinelli, V.A. (1983) Kappa-bungarotoxin: an
intracellular study demonstrating blockade of neuronal nicotinic
receptors by a snake neurotoxin. Brain Res., 289:317-321.
Dryer, S.E., and Chiappinelli, V.A. (1985) Substance P depolarizes
nerve terminals in an autonomic ganglion. Brain Res.. 336:190-194.
Dudel, J., and Kuffler, S.W. (1961) The quantal nature of transmission
and spontaneous miniature potentials at the crayfish neuromuscular
junction. J. Physiol. (London). 155:514-529.
Dudel, J., Pamas, I., and Pamas, H. (1983) Neurotransmitter release
and its facilitation in crayfish, VI. Release determined by both
intracellular calcium and depolarization of the nerve terminal.
Pfluqers Archiv.. 339:1-10.
Edry-Schiller, J., and Rahamimoff, R. (1993) Activation and inactiva
tion of the bursting potassium channel from fused Torpedo synapto-
somes. J. Phvsiol. (London). 471:659-678.
Erulkar, S.D. (1983) The modulations of neurotransmitter release at
synaptic junctions. Rev. Biochem. Pharmacol.. 98:63-175.
Erulkar, S.D., and Rahamimoff, R. (1978) The role of calcium ions in
tetanic and post-tetanic increase in miniature end-plate potential
frequency. J. Phvsiol. (London). 287:501-511.
Fadoga, E., and Brookhart, J.M. (1962) Interactions of excitatory post-
synaptic potentials generated at different sites on the frog moto
neuron. J. Neurophysiol.. 25:790-804.
Feng, T.P. (1941) Studies on the neuromuscular junction. XXVI. The
changes of the end-plate potential during and after prolonged stim
ulation. Chin. J. Physiol.. 16:341-372.
Fletcher, G.H., and Chiappinelli, V.A. (1992a) An inward rectifier is
present in presynaptic nerve terminals in the chick ciliary gan
glion. Brain Res.. 575:103-112.
Fletcher, G.H., and Chiappinelli, V.A. (1992b) Spontaneous miniature
hyperpolarizations of presynaptic nerve terminals in the chick cil
iary ganglion. Brain Res.. 579:165-168.
Fletcher, G.H., and Chiappinel1i, V.A. (1993) The actions of the kappa
1 opioid agonist U-50,488 on presynaptic nerve terminals of the
chick ciliary ganglion. Neuroscience. 53:239-250.
Fox, A.P., Nowycky, M.C. and Tsien, R.W. (1987a) Kinetic and pharmaco
logical properties distinguishing three types of calcium currents
in chick sensory neurons. 0. Phvsiol. (London). 394:149-172.

85
Fox, A.P., Nowycky, M.C. and Tsien, R.W. (1987b) Single-channel record
ings of three types of calcium channels on chick sensory neurons.
J. Physiol. (London). 394:173-200.
Frank. E. (1973) Matching of facilitation at the neuromuscular junction
of the lobster. J. Physiol. (London!. 233:635-658.
Greengard, P., Valtorta, F., Czernik, A.J., and Benfenati, F. (1993)
Synaptic vesicle phosphoproteins and regulation of synaptic func
tion. Science. 259:780-785.
Hess, A. (1965) Developmental changes in the structure of the synapse
on the myelinated cell bodies of the chicken ciliary ganglion. jL.
Cell Biol.. 25:1-19.
Hochner, B., Klein, M., Schacher, S., and Kandel, E.R. (1986) Action
potential duration and the modulation of transmitter release from
the sensory neurons of Aplysia in presynaptic facilitation and
behavioral sensitization. PNAS (USA). 83:8410-8414.
Hochner, B., Pamas, H., and Pamas, I. (1989) Membrane depolarization
evokes neurotransmitter release in the absence of calcium entry.
Nature. 342:433-435.
Hodgkin, A.L., and Huxley, A.F. (1939) Action potentials recorded from
inside a nerve fibre. Nature 144:710-711.
Hubbard, J.I. (1963) Repetitive stimulation at the neuromuscular junc
tion, and the mobilization of transmitter. J. Physiol. (London),
169:641-662.
Hubbard, J.I., and Schmidt, R.F. (1963) An electrophysiological inves
tigation of mammalian motor nerve terminals J. Physiol. (London),
166:145-165.
Katz, B., and Miledi, R. (1965) The measurement of synaptic delay, and
the time course of acetylcholine release at the neuromuscular
junction. Proc. Rov. Acad. B161:483-495
Katz, B., and Miledi, R. (1967) The timing of calcium action during
neuromuscular facilitation. J. Physiol. (London). 189:535-544.
Katz, B., and Miledi, R. (1968) The role of calcium in neuromuscular
facilitation. J. Physiol. (London). 195:481-92.
Klein, M., Camardo, J., and Kandel E.R. (1982) Serotonin modulates a
specific potassium current in sensory neurons that show presynaptic
facilitation in Aplysia. PNAS (USA). 79: 5713-5717.
Kuno, M. (1964) Mechanism of facilitation and depression of the excit
atory synaptic potential in spinal motoneurones. J, Physiol. (Lon
don). 175:100-112.

86
Landau, E.M., Smolinsky, A., and Lass, Y. (1973) Post-tetanic potentia
tion and facilitation do not share a common calcium-dependent mech
anism. Nature New Biology. 244:155-157.
Landmesser, L. and Pilar, G. (1972) The onset and development of trans
mission in the chick ciliary ganglion. J. Physiol. (London).
222:691-713.
Lee, K.S. (1987) Potentiation of the calcium-channel currents of
internally perfused mammalian heart cells by repetitive depolari
zation. PNAS (USA). 84:3941-3945.
Lentzner, A., Bykov, V., and Bartschat, D.K. (1992) Time-resolved
changes in intracellular calcium following depolarization of rat
brain synaptosomes. J. Physiol. (London). 450:613-628.
Lev-Tov, A., and Rahamimoff, R. (1980) A study of tetanic and post-
tetanic potentiation of miniature end-plate potentials at the frog
neuromuscular junction. J. Physiol. (London). 309:247-273.
Liley, A.W. (1956) The quantal components of the mammalian end-plate
potential. J. Physiol. (London). 133:571-587.
Liley, A.W., and North, K.A.K. (1953) An electrical investigation of
effects of repetitive stimulation on mammalian neuromuscular junc
tion. J^-NeunjfihYii, 16:509-527.
Lledo, P.-M., Vernier, P., Vincent, J.-D., Mason, W.T., and Zorec, R.
(1993) Inhibition of Rab3B expression attenuates calcium-dependent
exocytosis in rat anterior pituitary cells. Nature. 364:540-544.
Llinas, R., Steinberg, I.Z., and Walton, K. (1981) Presynaptic calcium
currents in squid giant synapse. Biophvs. J. 33:289-321.
Llinas, R., Sugimori, M., Lin, J.W., and Cherksey, B. (1989) Blocking
and isolation of a calcium channel from neurons in mammals and
cephalopods utilizing a toxin fraction (FTX) from funnel web spider
poison. PNAS (USA). 86:1689-1693.
Llinas, R., Sugimori, M., and Silver, R.B. (1992) Microdomains of high
calcium concentration in a presynaptic terminal. Science.
256:677-679.
Loewi, 0. (1921) Uber humorale Ubertragbarkeit der Herznervenwirkung.
Pflugers Archiv. 189:239-242.
Magleby, K.L. (1973) The effect of repetitive stimulation on facilita
tion of transmitter release at the frog neuromuscular junction. jL
Physiol. (London). 234:327-352.

87
Magleby, K.L., and Zengel, J.E. (1975a) A dual effect of repetitive
stimulation on post-synaptic potentiation of transmitter release at
the frog neuromuscular junction. J. Physiol. (London). 245:163-182.
Magleby, K.L., and Zengel, J.E. (1975b) A quantitative description of
tetanic and post-tetanic potentiation of transmitter release at the
frog neuromuscular junction. J. Physiol. (London! 245:183-208.
Magleby, K.L., and Zengel, J.E. (1976) Augmentation: a process that
acts to increase transmitter release at the frog neuromuscular
junction. J. Physiol. (London). 257:449-470.
Magleby, K.L., and Zengel, J.E. (1982) A quantitative description of
stimulation-induced changes in transmitter release at the frog
neuromuscular junction. J. Gen. Physiol.. 80:613-638.
Mallart, A. (1985) A calcium-activated potassium current in motor nerve
terminals of the mouse. J. Physiol.. 368:577-591.
Mallart, A., and Martin, A.R. (1967) An analysis of facilitation of
transmitter release at the neuromuscular junction of the frog. L
Physiol. (London). 193:679-694.
Marom, S., and Levitan, I.B. (1994) State-dependent inactivation of the
Kv3 potassium channel. Biophvs. J.. 67:579-589.
Martin, A.R., and Pilar, G. (1963a) Dual mode of synaptic transmission
in the avian ciliary ganglion. J. Physiol. (London). 168:443-463.
Martin, A.R., and Pilar, G. (1963b) Transmission through the ciliary
ganglion of the chick. J, Physiol. (London). 168:464-475.
Martin, A.R., and Pilar, G. (1964a) An analysis of electrical coupling
at synapses in the avian ciliary ganglion. J. Physiol. (London),
171:4545-475.
Martin, A.R., and Pilar, G. (1964b) Quantal components of the synaptic
potential in the ciliary ganglion of the chick. J. Physiol. (Lon
don). 175:1-16.
Martin, A.R., and Pilar, G. (1964c) Presynaptic and post-synaptic
events during post-tetanic potentiation and facilitation in the
avian ciliary ganglion. J. Physiol. (London). 175:17-30.
Marty, A. (1989) The physiological role of calcium-dependent channels
Trends Neurosci.. 12:420-424.
Marty, A., Tan, Y.P., and Trautmann, A. (1984) Three types of calcium-
dependent channels in rat lacrimal cells. J. Physiol. (London).
357:293-325.

88
Marwitt, R., Pilar, G., and Weakly, J.N. (1971) Characterization of two
cell populations in avian ciliary ganglion. Brain Res.. 25:317-334.
Mastrogiacomo, A., Evans, C.J., and Gundersen, C.B. (1994) Antibody
peptides against a cysteine-string protein. J. Neurochem.
62:873-880.
McNaughton, B.L. (1982) Long-term enhancement and short-term potentia
tion in rat fascia dendata act through different mechanisms. jL
Phvsiol.(London). 324:249-262.
Meriney, S.D., and Grinnell, A.D. (1991) Endogenous adenosine modulates
stimulation-induced depression at the frog neuromuscular junction.
J. Physiol. (London). 443:441-455.
Miledi, R., and Slater, C.R. (1966) The action of calcium on neuronal
synapses in the squid. J. Physiol. (London) 198:473-498.
Miledi, R., and Thies, R. (1971) Tetanic and post-tetanic rise in
frequency of miniature end-plate potentials in low-calcium solu
tions. J. Physiol. (London) 212:245-257.
Miller, R.J. (1987) Multiple calcium channels and neuronal function.
Science. 235:46-51.
Mosier, D.R., and Zengel, J.E. (1994) Evoked transmitter release at the
frog neuromuscular junction in the presence of very low extracellu
lar Ca2+. Neurosci. Letters. 174:1-4.
Neher, E., and Augustine, G.J. (1990) Calcium gradients and buffers in
bovine chromaffin cells. J. Physiol. (London). 450:273-301.
Nicholls, J.G., Martin, A.R., and Wallace, B.G. (1992) From Neuron to
Brain, third edition, Sinauer Press, New York, USA.
Nowycky, M.C., Fox, A.P., and Tsien, R.W. (1985) Three types of neur
onal calcium channel with different calcium agonist sensitivity.
Nature. 316:440-443.
Nussinovitch, I., and Rahamimoff, R. (1988) Ionic basis of tetanic and
post-tetanic potentiation at a mammalian neuromuscular junction, >L
Physiol. (London) 396:435-455.
Pamas, H., Dudel, J., and Pamas, I. (1986) Neurotransmitter release
and its facilitation in crayfish VII. Another voltage-dependent
process beside calcium entry controls the time course of phasic
release. Pflug. Arch.. 406:131-137.
Partridge, L.D., and Swandulla, D. (1988) Single Ca-activated cation
channels in bursting neurons of Helix. Pfluo. Arch.. 410:627-631.

89
Penner, R., and Dreyer, F. (1986) Two different presynaptic calcium
currents in mouse motor nerve terminals. Pfluo. Arch.. 406:190-197.
Pevsner, J., Hsu, S.-c., Braun, J., Calakos, N., Ting, A.E., Bennett,
M.K., and Scheller, R.H. (1994) Specificity and regulation of a
synaptic vesicle docking complex. Neuron. 13:353-361.
Poage, R.E., and Zengel, J.E. (1993) Kinetic and pharmacological exami
nation of stimulation-induced increases in synaptic efficacy in the
chick ciliary ganglion. Synapse. 14:81-89.
Porter, R. (1970) Early facilitation at corticomotoneuronal synapses.
J. Physiol. (London). 207:733-745.
Quattrocki, E.A., Marshall, J., and Kaczmarek, L.K. (1994) A Shab
potassium channel contributes to action potential broadening in
peptidergic neurons. Neuron. 12:73-86.
Razani-Boroujerdi, S., and Partridge, L.D. (1993) Activation and modu
lation of calcium-activated non-selective cation channels from
embryonic chick sensory neurons. Brain Res.. 623:195-200.
Regehr, W.G., and Mintz, I.M. (1994) Participation of multiple calcium
channels types in transmission at single climbing fiber to Purkinje
cell synapses. Neuron. 12:605-613.
Robitaille, R., and Charlton, M. (1992) Presynaptic calcium signals and
transmitter release are modulated by calcium activated potassium
channels. J. Neurosci.. 12:297-305.
Rosahl, T.W., Geppert, M., Spillane, D., Herz, J., Hammer, R.E., Mal-
enka, R.C., and Sudhof, T.C. (1993) Short-term synaptic plasticity
is altered in mice lacking synapsin 1. Cell. 75:661-670.
Rosenthal, J. (1969) Post-tetanic potentiation at the neuromuscular
junction of the frog. J. Physiol (London). 203:121-133.
Scheller, M.K., and Bennett, M.K. (1994) Molecular correlates of synap
tic vesicle docking and fusion. Curr. Qpin. Neurobiol.. 4:324-9.
Scott, R.H., Pearson, H.A., and Dolphin, A.C. (1991) Aspects of verte
brate neuronal voltage-activated calcium currents and their regula
tion. Prog. Neurobiol. 36:485-520.
Silinsky, E.M., Watanabe, M., Redman, R.S., Qui, R., Hirsh, J.K., Hunt,
J.M., Solsona, C.S., Alford, S., and MacDonald, R.C. (1995) Neuro
transmitter release evoked by nerve impulses without calcium entry
through calcium channels in frog motor nerve endings. J. Phvsiol.
(London). 482: 511-520.
Smith, S.J., and Augustine, G.J. (1988) Calcium ions, active zones and
synaptic transmitter release. TINS. 11:458-464.

90
Stanley, E.F. (1989) Calcium currents in a vertebrate presynaptic nerve
terminal: the chick ciliary ganglion calyx. Brain Res.,
505:341-345.
Stanley, E.F. (1991) Single calcium channels on a cholinergic presynap
tic nerve terminal. Neuron. 7:585-591.
Stanley, E.F. (1993) Single calcium channels and acetylcholine release
at a presynaptic nerve terminal. Neuron. 11: 1007-1011.
Stanley, E.F. and Atrakchi, A.H. (1990) Calcium currents recorded from
a vertebrate presynaptic nerve terminal are resistant to the dihy
dropyridine nifedipine. PNAS (USA). 87:9683-9687.
Stanley, E.F. and Goping, G. (1991) Characterization of a calcium cur
rent in a vertebrate cholinergic presynaptic nerve terminal, jh
Neurosci.. 11:985-993.
Stinnakre, J., and Tauc, L. (1973) Calcium influx in active Aplysia
neurons detected by injected aequorin. Nature. 293:115-117.
Stuenkel, E.L., and Nordmann, J.J. (1993) Sodium-evoked, calcium-
independent vasopressin release from rat isolated neurohypophysial
nerve endings. J. Physiol. (London). 468:357-78.
Sugita, S., Goldsmith, J.R., Baxter, D.A., amd Byrne, J.H. (1992)
Involvement of protein kinase C in serotonin-induced spike broaden
ing and synaptic facilitation in sensorimotor connections of Aply
sia. J. Neurophvsiol.. 68:643-651.
Suszkiw, J.B., O'Leary, M.E., Murawsky, M.M., and Wang, T. (1986) Pre
synaptic calcium channels in rat cortical synaptosomes: fast kinet
ics of phasic calcium influx, channel inactivation, and relation
ship to nitrendipine receptors. J. Neurosci.. 6:1349-1357.
Takahashi, T., and Momiyama, A. (1993) Different types of calcium chan
nels mediate central synaptic transmission. Nature 366:156-158.
Takeuchi, A., and Takeuchi, N. (1962) Electrical changes in pre- and
post-synaptic axons of the giant synapse of Loligo. J. Gen. Phy
siol.. 45:1181-1193.
Tarelli, F.T., Bossi, M., Fesce, R., Greengard, P., and Valtorta, F.
(1992) Synapsin 1 partially dissociates from synaptic vesicles dur
ing exocytosis induced by electrical stimulation. Neuron. 9:
1143-1153.
Thompson, S.H. (1977) Three pharmacologically distinct potassium chan
nels in molluscan neurones. J. Phvsiol. (London). 265:465-488.

91
Vergara, C., and Latorra, R. (1983) Kinetics of Ca2+-activated K+
channels from rabbit muscle incorporated into planar bilayers.
Evidence for Ca2+ and Ba2+ blockade. J. Gen. Physiol.. 82:543-568.
Weinreich, D. (1971) Ionic mechanism of post-tetanic potentiation at
the neuromuscular junction of the frog. J. Physiol. (London!
212:431-446.
Yawo, H. (1990) Voltage-activated calcium currents in presynaptic nerve
terminals of the chicken ciliary ganglion. J. Physiol. (London).
428:199-213.
Yawo, H., and Chuhma, N. (1994) Omega-conotoxin-sensitive and
-resistant transmitter release from the chick ciliary presynaptic
terminal. J. Physiol. (London). 477:437-448.
Yawo, H., and Momiyama, A. (1993) Re-evaluation of calcium currents in
pre-and postsynaptic neurones of the chick ciliary ganglion. vL
Physiol. (London). 460:153-172.
Yoshikami, D., Bagobaldo, Z., and Olivera, B.M. (1989) The inhibitory
effects of omega conotoxin on calcium channels and synapses. Ann.
NY Acad. Sci.. 560:230-248.
Younkin, S.G. (1974) An analysis of the role of calcium in facilitation
at the frog neuromuscular junction. J. Physiol. (London). 237:1-14.
Zengel, J.E., Lee, D.T., Sosa, M.A. and Mosier, D.R. (1993a) Effects of
calcium channel blockers on stimulation-induced increases in trans
mitter release at the frog neuromuscular junction. Synapse,
15:251-262.
Zengel, J.E., and Magleby, K.L. (1977) Transmitter release during
repetitive stimulation: selective changes produced by Sr2+ and
Ba2+. Science, 197:67-69.
Zengel, J.E., and Magleby, K.L. (1980) Differential effects of Ba, Sr,
and Ca on stimulation-induced changes in transmitter release at the
frog neuromuscular junction. J. Gen. Physiol.. 76:175-211.
Zengel, J.E., and Magleby, K.L. (1981) Changes in miniature end-plate
potential frequency during repetitive stimulation in the presence
of Ca2+, Ba2+, and Sr2+ at the frog neuromuscular junction. jL
Gen. Physiol.. 77:503-529.
Zengel, J.E., and Magleby, K.L. (1982) Augmentation and facilitation of
transmitter release: a quantitative description at the frog neuro
muscular junction. J. Gen. Physiol.. 80:583-611.

92
Zengel, J.E., Magleby, K.L., Horn, J.P., McAfee, D.A., and Yarowski,
P.J. (1980) Facilitation, augmentation and potentiation of synaptic
transmission at the superior cervical ganglion of the rabbit. vL
Gen. Physiol.. 76:213-231.
Zengel, J.E., Sosa, M.A., and Poage, R.E. (1993b) Omega-conotoxin
reduces facilitation of transmitter release at the frog neuromuscu
lar junction. Brain Res.. 611:25-30.
Zengel, J.E., Sosa, M.A., Poage, R.E., and Mosier, D.R. (1994) Role of
intracellular Ca2+ in stimulation-induced increases in transmitter
release at the frog neuromuscular junction. J. Gen. Physiol. 104:
337-355.
Zucker, R.S. (1974) Characteristics of crayfish neuromuscular facili
tation and their calcium dependence. J. Physiol. (London),
241:91-110.
Zucker, R.S. (1989) Short-term synaptic plasticity. Ann. Rev. Neuro-
sci.. 12:13-31.

BIOGRAPHICAL SKETCH
Robert Eliot was born, a month early, to James David Poage and
Scottie Johnston Poage on August 6, 1964, in Winston-Salem, North Caro
lina. The second of three boys, Bob moved with the family to Baltimore
and then to Jacksonville and Orange Park, Florida, where he spent his
formative years. Dr. Jim Poage received his doctoral degree from the
University of Florida in counselor education, and has spent the better
part of his life working in rehabilitation of drug- and alcohol-
dependent persons. Scottie Poage-Hotchkiss has a successful practice
marriage counseling in Yuma, AZ.
His parents' careers interested young Mr. Poage in the field of
psychology, in which he pursued a bachelor's degree from the University
of Florida in 1987. Rather than continue in that vein, Bob decided
that understanding the underlying causes of mental illness and nervous
system dysfunction represented the future of psychology. He began
working in the field of neuroscience with primates, then for a while at
the cellular and biochemical level, before finally settling in for a
long stay as an electrophysiologist under the guidance of Janet Zengel.
Upon completing his dissertation, Bob will be working with Dr. Steve
Meriney at the University of Pittsburgh, where it is said to snow on
occasion. He currently lives in Gainesville, Florida, with his wife,
Heidi, his children, Shea Elizabeth and Scott Eliot, and pets too
numerous and ill-behaved to mention. In his spare time, he used to
like the outdoors, but now he just stares blankly at the TV, or the
wall, or his kids.
93

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Janet E. Zengel, C
Associate Professor
ir
of
Neuroscience
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Peter A.V. Anderson
Professor of Neuroscience
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
kjo -Met
Thomas W. Vickroy
Associate Professor of
oscience
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Stuart E. Dryer
Associate Professor of Biological Science
Florida State University

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
PhiTip Posnerv
Professor of Physiology
This dissertation was submitted to the Graduate Faculty of the
College of Medicine and to the Graduate School and was accepted as
partial fulfillment of the requirements for the degree of Doctor of
Philosophy.
August, 1995
K
lege of Medicine
Dean, Graduate School

UNIVERSITY OF FLORIDA
3 1262 08667 709 2



36
trains, I observed an apparent depression of ganglionic transmission
immediately following the conditioning trains that made it difficult to
obtain reliable estimates of augmentation. There was no consistent
effect of increasing stimulus duration on the time constants of decay
of these processes. These results are similar to those observed at
other synapses (e.g. Magleby and Zengel, 1976; Zengel et al., 1980).
Pharmacological Characterization: Effects of Strontium and Barium
At the frog neuromuscular junction and the rabbit sympathetic gan
glion, the addition of certain divalent cations to the bathing solution
selectively affects individual components of stimulation-induced
increases in release. Barium increases the magnitude of the augmenta
tion phase and strontium increases the magnitude and time constant of
the second component of facilitation (Zengel and Magleby, 1977, 1980,
1981; Zengel et al., 1980). To further test the hypothesis that the
phenomena I describe here are analogous to the four components of stim
ulation-induced increases in release reported in other preparations, I
repeated the experiments described in Figures 2-3 and 2-5 in the pres
ence of these divalent cations.
Figure 2-8 illustrates the effect of Sr^+ on facilitation. In the
presence of Sr^* (open circles), V(t) was unchanged or slightly reduced
at short interstimulus intervals (less than 100 msec), but there was an
obvious enhancement of ganglionic efficacy at intervals of 300 to
2000 msec (A). This increase in V(t) could be attributed to an
increase in both the magnitude and time constant of decay of the second
component of facilitation (B). The first component of facilitation was
reduced in magnitude in the presence of Sr^+ (C).


REFERENCES
Alder, J., Xie, Z.-P., Valtorta, F., Greengard, P., and Poo, M.-m.
(1992) Antibodies to synaptophysin interfere with transmitter
secretion at neuromuscular synapses. Neuron. 9:759-768.
Aldrich, R.W., Getting, P.A., and Thompson, S.H. (1979) Mechanism of
frequency-dependent broadening of molluscan neurone somata, jh
Physiol. (London). 291:531-544.
Artalejo, C.R., Adams, M.E. and Fox, A.P. (1994) Three types of calcium
channels trigger secretion with different efficacies in chromaffin
cells. Nature 367:72-76.
Augustine, G.J. (1990) Regulation of transmitter release at the squid
giant synapse by presynaptic delayed rectifier potassium current.
J. Physiol. (London). 431:343-364.
Augustine, G.J., Smith, S.J., and Zucker. R.S. (1982) Role of presynap
tic calcium ions and channels in synaptic facilitation and depres
sion at the squid giant synapse. J. Physiol. (London). 323:173-193.
Bain, A.I., and Quastel, D.M.J. (1992) Quantal transmitter release
mediated by strontium at the mouse neuromuscular junction. J. Phy
siol. (London). 450:63-87.
Bain, A.I., and Quastel, D.M.J. (1992) Multiplicative and additive
Ca2+-dependent components of facilitation at mouse endplates. jL
Physiol. (London). 455:383-405.
Bennett, M.K., and Scheller, R.H. (1993) The molecular machinery for
secretion is conserved from yeast to neurons. PNAS (USA1.
90:2559-2563.
Bennett, M.R., and Ho, S. (1991) Probabilistic secretion of quanta from
nerve terminals in avian ciliary ganglia modulated by adenosine, jh
Phvsiol.. 440:513-527.
Bertolino, M., and Llinas, R. (1992) The role of voltage activated and
receptor operated calcium channels in neuronal cells. Ann. Rev.
Pharmacol. Toxicol.. 32:399-421.
Birks, R.I., and Cohen, M.W. (1968) The influence of internal sodium on
the behaviour of motor nerve endings. Proc. Rov. Soc., 170:401-421.
82


8
and potentiation. However, Martin and Pilar compared the amplitudes of
individual action potentials, and it is possible that subtle changes in
the action potential amplitude and/or duration would not have been
detected by this type of analysis. I have characterized the effects of
repetitive stimulation on the presynaptic action potential in the chick
ciliary ganglion. The results of these studies are presented in chap
ter 4 and possible contributions of changes in presynaptic electrical
activity to the processes of stimulation-induced increases in release
are discussed.
Summary
The major goal of this research project was to obtain a better
understanding of the mechanisms underlying stimulation-induced changes
in transmitter release. Experiments were performed using the ciliary
ganglion of the embryonic chicken.
I found that the chick ciliary ganglion responds to repetitive
stimulation with four components of increased release that are analo
gous to the first and second components of facilitation, augmentation
and potentiation, as have been described in other preparations. Both
kinetic (time constants of decay) and pharmacological (response to
Sr2+, Ba2+) properties of these components were used in their identifi
cation, and intracellular recording from postsynaptic cells verified
their presynaptic origin.
I found that the pharmacological characteristics of Ca2+ channels
coupled to evoked transmitter release are similar to the N-type channel
classification. This finding is in agreement with other investigators'
descriptions of presynaptic Ca2+-dependent currents in the chick
ciliary ganglion.


TABLE OF CONTENTS
ACKNOWLEDGEMENTS ii
ABSTRACT iv
CHAPTERS
1 INTRODUCTION 1
Chemical Synaptic Transmission 1
The Ciliary Ganglion of the Embryonic Chick 2
Stimulation-Induced Changes in Synaptic Efficacy 4
Voltage-Dependent Calcium Channels and Transmitter Release... 5
The Presynaptic Action Potential and Transmitter Release 7
Summary 8
2 EFFECTS OF REPETITIVE STIMULATION ON SYNAPTIC EFFICACY
IN THE CHICK CILIARY GANGLION 10
Methods 12
Results 18
Discussion 42
3 CHARACTERIZATION OF CA2+ CHANNELS INVOLVED IN SYNAPTIC
TRANSMISSION 45
Methods 47
Results 48
Discussion 50
4 EFFECT OF REPETITIVE STIMULATION ON THE PRESYNAPTIC
ACTION POTENTIAL 54
Methods 55
Results 56
Discussion 68
Notes 75
5 CONCLUSIONS 77
REFERENCES 82
BIOGRAPHICAL SKETCH
93


92
Zengel, J.E., Magleby, K.L., Horn, J.P., McAfee, D.A., and Yarowski,
P.J. (1980) Facilitation, augmentation and potentiation of synaptic
transmission at the superior cervical ganglion of the rabbit. vL
Gen. Physiol.. 76:213-231.
Zengel, J.E., Sosa, M.A., and Poage, R.E. (1993b) Omega-conotoxin
reduces facilitation of transmitter release at the frog neuromuscu
lar junction. Brain Res.. 611:25-30.
Zengel, J.E., Sosa, M.A., Poage, R.E., and Mosier, D.R. (1994) Role of
intracellular Ca2+ in stimulation-induced increases in transmitter
release at the frog neuromuscular junction. J. Gen. Physiol. 104:
337-355.
Zucker, R.S. (1974) Characteristics of crayfish neuromuscular facili
tation and their calcium dependence. J. Physiol. (London),
241:91-110.
Zucker, R.S. (1989) Short-term synaptic plasticity. Ann. Rev. Neuro-
sci.. 12:13-31.


UNIVERSITY OF FLORIDA
3 1262 08667 709 2


43
first component of facilitation, the second component of facilitation,
augmentation and potentiation as described at the frog neuromuscular
junction and at other synapses.
In preparations where stimulation-induced increases in synaptic
efficacy have been studied extensively, most notably the frog neuromus
cular junction, the increases have been shown to result from an
increase in quantal release (del Castillo and Katz, 1954; Magleby and
Zengel, 1976). In one of the first electrophysiological studies using
the chick ciliary ganglion preparation, Martin and Pilar (1964c) showed
that paired pulse facilitation of EPSP amplitude occurs in the embry
onic chick ciliary ganglion, and that it is a result of increased quan
tal content which decays back to control levels with a time constant of
about 65 msec. Data reported here confirm the presence of fcilitatory
processes and further describe two individual components of facilita
tion with distinct kinetic and pharmacological properties. Martin and
Pilar (1964c) also reported, using intracellular recording from ciliary
neurons, the presence of a more slowly decaying potentiation of EPSP
amplitude, although its time course was not described in detail. The
experiments reported here describe the decay of ganglionic efficacy
after prolonged repetitive stimulation. A more rapidly decaying compo
nent is akin to augmentation, having very similar kinetic and pharmaco
logical properties. The more slowly decaying component is termed
potentiation.
It has now been demonstrated that there are four components con
tributing to stimulation-induced increases in synaptic efficacy that
appear to result from increased transmitter release. The unresolved