SYNAPSE 7:235-243 (1991)

Synaptic Plasticity at Crayfish Neuromuscular Junctions: Facilitation and Augmentation GEORGE D. BITTNER AM) DOUGLAS A. BAXTER Department of Zoology, College of Pharmacy, and Institute for Neurological Sciences, University of Texas, Austin, Texas 78712 (G.D.B.) and Department of Neurobiology and Anatomy, University of Texas Medical School, Houston, Texas 77225 (D.A.B.)

KEY WORDS

Facilitation, Post-tetanic potentiation, Synaptic plasticity, Neuromuscular transmission, Crayfish

ABSTRACT Simultaneous intracellular recordings from presynaptic nerve terminals and postsynaptic muscle fibers were used to investigate the extent to which changes in presynaptic voltage may contribute to short-term facilitation and augmentation of transmitter release at neuromuscular junctions of the crayfish (Procambarus sinulans) opener muscle. Presynaptic nerve terminals have an average resting membrane potential of about -80 mV, single action potentials have an average foot-to-peakamplitude of about 98 mV, and single action potentials are followed by a depolarizing after potential (DAP)of about 10 mV. During stimulus trains of 9-16 impulses at 100 Hz, amplitudes of excitatory postsynaptic potentials (EPSPs) continuously facilitate up to 100-fold. This dramatic facilitation is associated with only slight increases in the peak voltage and duration of APs for the first 2 4 pulses in such a stimulus train. Foot-to-peak total amplitude of APs usually decreases after the first pulse in a stimulus train. The data strongly suggest that short-term facilitation is not due to changes in the amplitude or duration of APs invading the presynaptic terminal. Upon cessation of a longer stimulus train, the presynaptic terminal exhibits a hyperpolarizing after potential (HAP)up to 16 mV in amplitude depending upon the frequency (10-100 Hz) and duration (1-10 sec) of the tetanic stimulation. This post-tetanic HAP decays with a time constant of 10-20 sec, which is approximately equal to the third time constant of decay in EPSP amplitude (augmentation) following tetanic stimulation. Hence, presynaptic voltage changes and/or processes associated with these voltage changes (e.g., accumulation of ions, changes in ionic conductances, etc.) may be partly responsible for augmentation of EPSP amplitudes.

INTRODUCTION In both vertebrates and invertebrates, transmitter release at nerve s apses is often dependent upon revious activity. T IS significance of such use-depensent changes in synaptic efficacy (i.e., synaptic plasticity) ranges from the fine control of muscle tension in some invertebrates (Bittner, 1968a, 1989) to more complex behaviors such as learning and memory in mammals (Kandel and Schwartz, 1985). Previous activity at a nerve terminal can alter subsequent release of transmitter for milliseconds to days at that same terminal (homosynaptic effects) or at an adjacent nerve terminal (heterosynaptic effects). An increased probability of homosynaptic transmitter release lasting several milliseconds to several seconds has been termed (short-term) facilitation (Del Castillo and Katz, 1954; Dudel and Kuffler, 1961; Mallart and Martin, 1967) and that lasting for about 10 sec has been termed au entation (Zengel et al., 1980). Voltage changes be ore or during an action potential (AP)may directly or indirectly increase the amount of transmitter released. Hence, voltage changes have been hypothe-

?i

Y

0 1991 WILEY-LISS,INC.

sized to cause facilitation andor other homosynaptic plasticities such as augmentation or post-tetanic potentiation (PTP) (Atwood, 1976; Dudel, 1971; Hubbard and Schmidt, 1963; Takeuchi and Takeuchi, 1962; Zucker, 1974a,b). Conversely, some of the calcium which enters the nerve terminal during an AP may remain to increase transmitter release to subsequent APs independent of small voltage changes. Such “residual calcium” has often been hypothesized to account for facilitation (Katz and Miledi, 1968; Zucker and Lara-Estrella, 1979) and other homosyna tic plasticities (Delaney et al., !ahamimoff, 1978; Kretz et al., 1989; Erulkar and l 1984). At crayfish neuromuscular junctions and preparations other than the squid giant synapse (which may only exhibit the homosynaptic plasticity of facilitation; Charlton and Bittner, 1978a,b), previous investigations of mechanisms for facilitation, augmentation, and PTP have been hampered by an inability to record intracellularly from the nerve terminals. The nature of the Received April 19,1990; accepted in revised form August 27,1990.

G.D. BITTNER AND D.A. BAXTER

236

presyna tic electrical events associated with these plasticities as been inferred from focal extracellular recordin s near the excitor terminals, and the interpretation o such extracellular recordin s has proven to be controversial. For example, Dude (1965, 1971) has suggested that A p s do not invade the terminal re on of the opener excitor axon, that APs are followed by yperpolarizing after potentials (HAPS),and that facilitation is due to a growth in the amplitude and time course of the resynaptic APs. Conversely, Zucker (1974a,b)conclu ed that APs do invade the terminal region (cf. Smith, 1983) and that presyna tic APs at these same neuromuscular junctions are fol owed by a de olarizing after potentials (DAF’). Conversely, APs mig t remain constant in size but increase in duration during repetitive stimulation (cf. Zucker and Lara-Estrella, 1979), which might account for facilitation. The introduction of techniques for impaling and recording intracellularly from the opener-excitor nerve terminals (Baxter and Bittner, 1981; Fuchs and Getting, 1980)have ermitted a more exact investigation of these issues. In t is paper, we describe pres and ostsynaptic membrane potentials recorde intracellufarly from crayfish opener-excitor nerve terminals and opener muscle fibers. We conclude that small changes do occur in peak voltage, total am litude, and duration of APs, but that these changes ave little effect on short-term facilitation. Conversely, a large HAP following the last AP in a tetanic stimulus train may be important for the development or expression of augmentation. METHODS Experiments were performed on the opener muscle located in the propodite segment of the claw or the first walking leg of juvenile crayfish, Procambarus simuZuns, 5-10 cm in carapace length. Animals were obtained from a local bait dealer, maintained in freshwater aquaria, and used within 3 weeks of their purchase. The isolated limbs were constantly perfused with van Harreveld’s h siological saline for crayfish (205 mM NaCI, 5.4 m h kCl, 14 mM CaCl,, 2.6 mM MgC12, 2.4 mM NaHC03, pH 7.4) at room temperature (19-21°C). The two nerve bundles containing the single excitor axon and the single inhibitor axon to the opener muscle were isolated in a more proximal limb segment (the meropodite) and placed over separate platinum wire hook electrodes to elicit APs. The closer muscle of the claw was dissected from the ropodite to provide access to the syna tic terminals o the opener nerve on the ventral su ace of the opener muscle. Anatomy and physiology of the crayfish opener excitor-inhibitor preparation A single opener inhibitor motor axon, which releases y-amino butyric acid (GABA),and a single excitor axon, which releases glutamate, are the sole innervation to most opener muscle fibers, except for a set of 3-10 very proximal muscle fibers that are also innervated by an inhibitor axon common to many claw muscles (Wiens, 1984). This latter “common inhibitor” axon and the fibers it innervates were not examined in this study. The opener excitor and inhibitor axons branch together on the ventral and dorsal surface of the opener muscle. The primar branches of the two axons form a Y-shaped pattern on t e ventral surface of the opener muscle from

B

K

P

a

CQ

P

K

K

i

T

bYR

P

P

f

rptic a

R

rp

which arise secondary and tertiary branches, all of which can be seen by using a dissection microscope with darkfield illumination. Scanning electron micro aphs reveal that small-diameter processes or “branch ettes” project from the secondary and tertiary branches and that these branchlettes make direct contact with the under1 ing ostsynaptic muscle fibers. Presumably, these ranc lettes are presynaptic release sites onto those muscle fibers directly beneath the secondar and tertiary nerve branches (see also Atwood and dojtowicz, 1986; Jahromi and Atwood, 1974). The opener inhibitor innervates each muscle fiber near sites of excitor innervation, thereby roviding postsynaptic inhibition to the muscle fiber. he opener inhibitor also makes syna ses on the o ener excitor terminals as well as on axona sites slight y (1-10 pm) more proximally where the opener excitor axon branches to innervate opener muscle fibers. These contacts presynaptically inhibit glutamate release from the excitor axon (Atwood and Wojtowicz, 1986; Bittner, 1989). Opener muscle fibers are of rather large diameter, measurin 200-500 pm in adult crayfish 10-20 cm in body lengt and 40-80 pm in juvenile crayfish 2-6 cm in body length (Bittner and Traut, 1978).Due to their large diameter and absence of a thick connective tissue sheath, opener muscle fibers are easily penetrated with 2-5 Mi2 microelectrodes. Space constants of opener fibers are usually several times muscle fiber length, muscle membrane responses are passive to excitor stimulation, and membrane conductances under o little change during de olarizations up to 7 mV hittner, 1968a,b). Hence, w en the opener excitor axon is stimulated to fire a single action potential, intracellular electrodes inserted near the midpoint of an opener muscle fiber record an EPSP which is the summed, a proximately e ual, contribution of 40-50 release sites (fSittner 1968a; ittner and Harrison, 1970) with relatively little spatial decay or alteration due to active or passive posts a tic membrane responses. The uantal content of tg EPSP can be obtained by Jviding its amplitude by the average spontaneous miniature EPSP (MEPSP) amplitude presumed to represent a single quantum (Baxter et al., 1985; Bittner and Harrison, 1970). Since the muscle fibers develop tension slowly in response to a step de olarization, one can record EPSPs from muscle fibers uring high-frequenc stimulation if the central tendon is immobilized. excitor synapses on a ven muscle fiber generally release transmitter in t e same fashion that can be characterized b their F, value, where F, is defined as the ratio of EP P amplitude with 10 Hz stimulation of the excitor axon to the EPSP amplitude with 1 Hz stimulation (Atwood and Bittner, 1971).Low F, (F < 3) synapses are usually located on the most distal and most proximal opener muscle fibers; high F synapses (F, > 3) are usually located on fibers in the central region of opener muscle (Atwood and Bittner, 1971; Bittner, 1968a; Bittner and Harrison, 1970; Bittner and Sewell, 1976). Hi h F, synapses were selected for this study because o f t eir greater synaptic plasticity (Bittner, 1989). Electrophysiological procedures Intracellular recordings from the synaptic terminals of the excitor motor axon were made wlth 40-80 MO microelectrodes filled with either 3 M KC1 or 2 M K

F

K

3.

B

f?

B

a

d

237

FACILITATION AND AUGMENTATION IN CRAYFISH

citrate. A small hook was sli ped under the opener nerve at the Y branch or un er a seconda within 50-200 pm of a postsynaptic muscle fi er. This branch technique stabilized the two axons by a plying slight tension to the nerve and prevented musc e movements from interfering with the presynaptic recordings. The presynaptic recording electrode was oriented parallel to the nerve and a terminal branch was penetrated at an acute angle to the axonal surface. Identification of the penetrated terminal was confirmed by selectively stimulating the excitor and the inhibitor axons in the meropodite. After penetratin a nerve terminal, a second microelectrodewas placeLf into an adjacent postsynaptic muscle fiber to measure EPSPs produced by transmitter release from the penetrated terminal region. The presynaptic and ostsynaptic voltages were fed into ca acitance-neutrayized preamplifiers (WPI Inc., Model &-707) with a bandwidth of DC to 10 K Hz. The amplified signals were dis layed on an oscilloscope and entered into a computer o average transients (CAT).A calibration ulse was averaged with the data. The outut of the AT was recorded by using the XY plotter. !‘he CAT greatly improved the signal-to-noise ratio allowing the reliable measurement of small (20-500 pVi EPSPs. Intracellular recordings from the syna tic terminals were stable for several hours. Data was ta en only from preparations in which no damage occurred to the opener nerve or muscle fibers during dissection and subsequent enetration. The pre- and postsynaptic resting mem{ran, potentials, AP amplitudes, and MEPSP amplitudes were monitored throu hout an ex eriment. Damaged preparations had a l fecline in t ese potentials and/or a cloudy ap earance of muscle fibers. Several lines oIp evidence show that there is good electrotonic coupling between resynaptic recording sites in tertiar , secondary, or Ipbranches and sites of excitatory a n 2 inhibitory transmitter release. First, depolarizing and hyperpolarizin currents a plied through the presynaptic recording e ectrode modi y synaptic transmission evoked by an action potential, as reported by Wojtowicz and Atwood (1984). Second, synapses can be activated to release transmitter by relatively small (15-20 mV) depolarizing ulses (Wo’towicz and Atwood, 1984). Lar er pulses re ease muc! lmore neurotransmitter. Thir$ assively conducted “artificial” APs of 100 mV ampEtude and 1 msec duration generated in the presence of tetrodotoxin by a current passing electrode in the presynaptic terminal release 7040% of the transmitter released by natural1 occurring APs of 100 mV am litude and 1 msec uration (Sivaramakrishnan et a?, 1988). When 4-aminopyridine and tetraethylammonium are also added, such artificial APs release 9 5 9 8 % of the transmitter released by natural APs (Sivaramakrishnan et al., 1988). Finally, mathematical or circuit models of these nerve terminals suggest that membrane potential changes at the presynaptic recording sites are attenuated by 3050% at the res aptic release sites (Wojtowicz and Atwood, 198f). T g s , intracellular penetrations of these synaptic terminals record changes in presynaptic potentials which could be important in the regulation of transmitter release durin facilitation, presynaptic inhibition, augmentation P P, long-term potentiation, and other plastic phenomena at these synapses.

B

P

e

?J

P

K

K

P

5

P

RESULTS Recordings of nonfacilitated release A set of preliminary experiments revealed that the different regions of the presynaptic terminal accessible to intracellular recordings did not differ significantly in their electrophysiolo ‘cal pro erties. Opener excitor terminals were impa ed in eit er the primary (n=4), secondary (n=8), or tertiary (n=3) nerve branches and the excitor nerve was stimulated at 1 Hz, a frequenc which did not facilitate release from these high terminals when compared to 0,l Hz stimulation (Bittner and Segundo, 1989; Bittner and Sewell, 1976). Intracellular recordings revealed that each of the three regions of the presynaptic nerve terminal had similar membrane potentials and conducted action potentials of similar shape, duration, and after-potential. For these 15 excitor synaptic terminals, the avera e (iSD) resting membrane potential was -75 ? 7 mb. The average foot-to-peak total am litude of the AP when the excitor nerve was stimulate at 1Hz was 96 k 4 mV and the AP duration at one-half peak amplitude was about 0.75 msec. A single AP was followed by a DAP which persisted for 30-50 msec as previously noted by Wojtowicz and Atwood (1983). The average amplitude of the DAP at 2 msec after the onset of the AP was 10 -+ 2 mV and the DAP reversed olarity at a membrane potential of about -65 mV. T e characteristics of APs were not significantly different when recorded from primary secondary, or tertiary branches. Thus the data collected from the three different presynaptic recording sites were combined. In the present experiments, (n = 541, the average resting potential was significantly higher 4 (-86 mV), the avera e AP total amplitude was 101 I mV, the average DiP was 6 k 3 mV, and the DAP reversed its polarity at about - 75 mV. EPSPs at 0.1 Hz averaged 65 pV. For experiments described below, data were only taken from preparations having APs 2 90 mV in total am litude. Simultaneous postsynaptic recordings were ta en from nearby postsynaptic muscle fibers which had resting membrane potentials of -70 mV or eater. Each AP in the presyna tic terminal evoked an KPSP that averaged about 70 p when recorded from a postsynaptic muscle fiber.

P

R

3

cf

K

K

i:

Recordings from facilitated preparations

B

%

in APs was +16 mV. Second, the DAP following each A# summated and depolarized the synaptic terminal by 11 mV at the end of the high stimulus train. The

G.D. BITTNER AND D.A. RAXTER

238

A

A

B

Fig. 1. Simultaneous intracellular recordings ofexcitor APs from the presyna tic nerve terminals (A) and EPSPs from a nearby ostsynaptic muscye fiber (B) during a train of 11im ulses at 100 Hz.khe traces are CAT averages of 100 trains repeatejevery 5 sec. The pre- and postsyna tic recordings were separately displayed and plotted. In A, the solid Hne represents the origmal resting membrane potential (-80 mV) of the synaptic terminal, and the dashed line represents +16 mV. The restin potential of the muscle fiber was - 74 mV. Calibration bars: A = 50 1msec; B = 1 mV, 5 msec.

rnf,

increase in AP peak voltage and decrease in AP foot volta e interacted such that the total foot-to-peak amplitu es of the second and successive APs were all reduced by about 6 mV compared to the first AF' in the train. Similar changes in the pre- and postsynaptic potentials during 100 Hz stimulus trains of 5-20 impulses were seen in all preparations. For example, in the re aration shown in Figure 2, the amplitudes of the EPgPs ew progressively during a stimulus train until the EP& had increased by about 70-fold at the 11th ulse when compared t o the EPSP a t the first pulse. buring the same stimulus train, there were no continuous chan es in peak AP voltage, AP total amplitude, or DAP amp itude. During short, 100Hz stimulus trains in 15 such excitor terminals, the maximum increase in the peak voltage of the excitor AP was 6 2 4 mV; the total AP amplitude decreased by an average of 4 I 2 mV by the last pulse in the train as the summating DAPs depolarizin the synaptic terminals by an average of 10 2 3 mV. n all cases, changes in the potentials in the resynaptic terminals did not continue beyond the Fourth or fifth impulse of the stimulus train. In contrast, the amplitudes of the EPSPs grew continuously b an average of 45- 2 13-fold (maximum = 100-fold)as acilitation accumulated in the nerve terminals. We also searched for changes in excitor AP duration by graphical1 superim osing excitor APs during a brief 100 Hz stimu us train. uring the stimulus train shown in Figure 3A, the duration of the first excitor AP was 0.80 msec (measured at one-half total amplitude) and the duration of the ninth AP was 0.96 msec, an increase of 17%.The increases in AP duration usually reached a

f

7

K

P

9

6

Pulse Number Fig. 2. Chan es in intracellularly recorded AP amplitude (to anel) and EPSf' amplitude (bottom panel) from a secondary bran& %wing a train of 11 impulses given at 100 Hz. The secondary branch had a resting potential of -82 mV and the ostsynaptic muscle fiber had a resting otential of -76 mV. A Filid squares plot the total foot-to-peak voytages (left-hand scale) and filled circles plot the peak voltages (right-hand scale) in mV attained by successiveAPs invading the resynaptic terminals of the excitor axon. The summation of the D d a f t e r each APis equal to the differencebetween the two curves. No continuous changes in the pres aptic action potential were observed or other preparations (n > 30). B: during the stimulus train in Filled circles plot the EPSP amplitudes in pV produced by each successive impulse. The amplitudes of EPSPs were measured by using the method described by Bittner and Sewell (1976).The amplitudes of the EPSPs continuously increased throughout the stimulus train. No correlation was observed between changes in the presynaptic AP and short-term facilitation of the EPSP in this or other preparations.

tg

steady state by the fourth impulse as shown in Figure 3B, which superimposes all ten impulses of a 100 Hz stimulus train from a different reparation. In Figure 3B, the duration of the excitor A increased from 0.70 to 0.78 msec by the fourth im ulse in this train and did not increase thereafter. For a 1 synaptic terminals studied with 100 Hz stimulation, the excitor AP duration increased by an average of 18%by the fourth pulse and did not continue to increase beyond the fifth impulse in any preparation (n = 19).

3

f

239

FACILITATION AND AUGMENTATION IN CRAYFISH

A

A

1-

250 IO0I.-

Fig. 3. Su erposition of APs during at 100 Hz trains aligned so that the peaks opal1 APs coincide in time and voltage. A First and ninth (last)AP from a terminal with a -92 mV resting potential. Calibration bar = 20 mV, 0.1msec. B: First through tenth (last) AP from a terminal with a -88 mV resting potential. Calibration bar = 20 mV, 0.2 msec.

Finally, we examined changes in AP total amplitude or duration and EPSP amplitude when the excitor axon was stimulated with 15-20 pulses at lower (10-80 Hz) rates. In all cases (n = 15),changes in AP total amplitude or duration at any ven pulse in the train were no greater than (and usual y less than) that observed with 100 Hz stimulation. In general the lower the stimulus rate, the less dramatic were the changes in AP total amplitude or duration. At 10 Hz, changes in AP total amplitude or duration were often not detectable for the first 20 pulses. In all cases (including brief trains of 10 Hz stimulation), EPSP amplitudes monotonically increased during the entire stimulus train. In no case did the overall changes (often sli ht) in peak voltage, total foot-to-peak amplitude, or uration of APs correlate with the (often dramatic) growth of EPSP amplitude during short-term facilitation. Recordings from tetanized preparations We intracellularly recorded presynaptic potentials during tetanic stimulation which induced augmentation in these terminals. The excitor axon was stimulated for 1-10 sec at 1-100 Hz while A P s were simultaneously recorded from the presynaptic terminals and EPSPs recorded from postsynaptic muscle fibers (n = 9). A series of complex changes occurred in the presynaptic potentials during tetanizing stimulation. Fi re 4 shows presynaptic APs and ostsynaptic EPS s from an opener preparation which ad been stimulated with 1,10,20, or 30 Hz for 4 sec, by which time AP and EPSP amplitudes had reached a steady state in these high F, synapses (Bittner, 1968a; Bittner and Segundo, 1989).

I/

Y

cf

R

8-

0 2 '

Y

Fig. 4. Pres aptic APs (A) and posts aptic EPSPs (B)after 4 sec of tetanic stimugion with 1, 10, 20, andin30 Hz. The dashed line in A represents the original resting membrane potential of -86 mV in the resynaptic terminal. Calibration bars: A = 50 mV and 1 msec; 500 FV and 5 msec.

i;=

The dotted line in Figure 4A shows the ori 'nal resting potential of the resyna tic terminal. AP ter 4 sec of stimulation at eac of the our frequencies, the potential at the foot of each A€'became increasingly hyper olarized as stimulus frequency increased from 1to 0 Hz (Fi .4A).In other words, each AP was first followed b a DA!b and then b a HAP. For example, after 4 sec o 1, 10,20, and 30 z tetanic stimulation in Fi re 4A, the otential at the foot of each AP was hyperpo arized by 2, $4, and 6 mV, respectively. In contrast, during stimulation at higher tetanic frequencies from 60 to 100 Hz, the DAP phase often became increasingly larger such that at 100 Hz the foot of each AP began at a depolarized level (see Fig. 6 C6-Cs). Facilitation accumulated in the terminals during tetanic stimulation at 10-100 Hz as evidenced by an increased EPSP amplitude after 4 sec com ared to EPSP amplitudes at 1Hz (i.e., unfacilitated E SPs).As

R

d

F

P

T

d

P

240

G.D. BI’M’NER AND D.A. BAXTER

shown in Figure 4B for 1-30 Hz tetanic stimulation, the higher the tetanic frequency, the greater was the EPSP amplitude after 4 sec. In the preparation shown in Figure 4B and other pre arations (n = 8),MEPSP amlitudes were unchange$ during such stimulus trains. kence, the increase in EPSP amplitude is almost certainly due to increased quanta1 release rather than to conductance changes in the postsyna tic membrane (Baxter et al., 1985; Bittner, 1968a; Jittner and Segundo, 1989). The membrane potential at the foot of each AP was also affected by the duration of the tetanic stimulation at any frequenc 2 10 Hz. For example, as shown in Figure 5 for a 20 z tetanus lastin 10 sec, after 1sec the presynaptic membrane at the oot of each AP was depolarized from the original resting potential by 0.6 mV. After 3,7, and 10 sec of 20 Hz stimulation, resynaptic membrane potential at the foot of each was hyperpolarized by 1.5, 2.6, and 3.3 mV respective1 . Similar HAPS durin tetanic stimull of about 10 z have also been noted y Wojtowicz and Atwood (1985). Figure 6 illustrates some further com lexities among tetanus frequency, tetanus length, an membrane tential during and following tetanic stimulation in t IS and other (n = 6) presynaptic terminals. For example, when the opener-excitor was stimulated at frequencies from 10 to 50 Hz, the potential at the foot of each AP was always depolarized for the first few APs in the tetanus. As tetanic stimulation continued, the s p a tic terminals progressively hyperpolarized (Fig. 6A1- 4; C1-C5). Once the hyperpolarizing phase began a t any given tetanus frequency from 10 to 50 Hz, then the magnitude of the hyperpolarization increased as tetanic stimulation continued. The hi her the tetanus rate from 10 to 50 Hz, the later the yperpolarizing phase occurred during the tetanus. Presumably, the time of outset and the magnitude of the hyperpolarization varied durin a tetanus because of the interaction of summating D h s and develo ing hyperpolarization. In the terminal shown in I;;pigure 6, the hyper olarizing phase began within 0.5 sec at 10 Hz (Fig. 6 1, within 2 sec at 30 Hz (Fig. 6C3),and within 7 sec at 50kz (Fig. 6C5).When the opener-excitor axon was stimulated at 60-100 Hz, the

&

B

h

B

%

r-

B

x

fl

cp

presynaptic terminal remained depolarized throughout the stimulation and no hy er olarizin phase occurred ; C6-$). However, folduring the tetanus (Fi . 6% lowing cessation of t e high- requency tetanus, the summated DAPs quickly decayed and the membrane otential of the presynaptic terminal became hyper ofarized. Similar data were obtained from five ot er preparations. Prolonged stimulation at any fre uency above 10 Hz always produced a lon -1astin HA% following the ces’ following the tetanus sation of stimulation. hrs decayed with a time constant of about 18 sec in the terminal from which data is shown in Figure 6. At other terminals (n = 5), the time constant (7) of decay ranged from 10 t o 20 sec. Increasin the duration or frequency of stimulation increased the%A P, but did not chan e its deca T. After 10 sec of 100 Hz stimulation, the coul be as great as 16 mV (range = 8-16 mV). EPSP amplitude increased during 1-10 sec of tetanic stimulation at 10-100 Hz. Following cessation of such a tetanus, EPSP amplitude as measured by a test pulse decayed exponentially over several minutes (n = 11).In such cases, augmentation (A) at a test ulse (TP)given times after cessation of t e tetanus was calculate by at

a

J-

%f

K

5 &

aAp

B

K

varying,

ATP =

EPSPT~ - EPSPl EPSPl

where EPSPTp was the amplitude (mV) of the EPSP evoked by the test pulse and EPSPl was the EPSP am litude (mV)evoked by the first ulse in the tetanus. EPEP, was always insignificantly iifferent from EPSP amplitudes evoked by 0.1 Hz or 1 Hz stimulation in these high F, terminals (see Methods). Augmentation following a tetanus was assumed to decay in a series of ex onentials which began at t = O + according to the formu ation At = Al e-“r + A2 ePwT2 A ePwT3 as devised for decay or short-term facilitation aker a brief train of stimuli (Bittner and Sewell, 1976; Mallart and Martin, 1967). In this formulation, the first ~ T ~ represent ) shorttwo time constants of decay ( T and term facilitation and usually range from 10 to 1,000 msec. The third time constant of decay generally ran ed from 10 to 20 sec as reported for augmentation in ot er pre arations (Zengel et al., 1980). &gure 7 shows the decay of facilitation of EPSPs following a 50 Hz tetanus gwen for 2 sec. The tetanus stimulation was repeated once every 8-10 min for about 3 hours without any significant change in the amplitude of EPSPs at the beginnin (45 pV) or the end (2,300 pV) of the tetanus train. Fo lowing each tetanic burst, a single test pulse was gwen at 5-800 msec thereafter followed by 3-5 test pulses at least 2 sec apart begmning 2-14 sec after the end of the tetanic burst. As shown in Figure 7, T~ at this synapse was about 25 msec, 72 was about 200 msec, and 73 was about 16 sec. The synapse shown in Figure 7 showed three distinct components of decay in the first 30 sec. In six other reparations, all showed evidence for components of ecay during the first 30 sec with T~ ranging from 11t o 20 sec. Values of and T~ at all synapses ranged from 18to 45 msec and 130 to 450 msec res ectively. These values are similar to those reported or r and 72 for decay of short-term facilitation followingbrief(1-20 pulse) stimulus train at

P

a

P

1

Fig. 5. APs (clipped) and after-potentials recorded from a resynaptic terminal durin 15 sec of tetanic stimulation a t 20 Hz. &ch trace shown is an indivifual trace recorded at the time indicated (0,1,3,7,10 sec) after beginning the stimulus train. The dashed line represents the original resting membrane potential of -78 mV. The HAP is slow to develop and its magnitude increases continuously throughout the stimulus train. Calibration bars: 12 mV, 2 msec.

2

P

FACILITATION AND AUGMENTATION IN CRAYFISH *I

--i

A2

241

A4

A3

20 nz , 2 s e ~

*L T+

100Hz.1 sec

Cl

Fig. 6. Tracings of membrane potentials that develop in a resynap tic terminal during and after tetanic stimulation at 1G100 for 1-8 sec. For example, tracing B, is taken from the original record of B, photo aphed from an oscilloscope screen in which peak A€' voltage is well offthe top of the photographic record during this 8 sec train at 100

8z

230 ; -20

Hz. The upward arrow in each trace A C marks the onset of tetanic stimulation; the downward arrow mar& t i e cessation of stimulation. The dashed line re resents the normal resting potential of -81 rnV for this terminal. Caligration bars = 2 mV, 2 msec.

v s e c

-

02 4t I

1

2

I

1

4

I

I

1

1

8 Post - train Timdrec) 6

Fig. 7. Au entation of EPSP amplitude vs. time as defined by Equation 1fogwing 2 sec of 50 Hz stimulation. Each data point from 5 to 800 msec represents a single test pulse; each data point from 2 to 4 sec represents the average of 2-5 test pulses. Augmentation values

I

I

10

I

I

12

t

4

plotted on a logarithmic scale; time values plotted on linear scales with the up er scale in msec (open circles) and the lower scale in sec (filled circlesf Values for T , T~ calculated by mathematically subtracting and/or T~ (Bittner an3 Sewell, 1976; Mallart and Martin, 1967).

242

G.D. BI'M'NER AND D.A. BAXTER

10-100 Hz (Bittner and Segundo, 1989; Bittner and Sewell, 1976).Values for T3 are similar to those reported at rabbit su erior cervical synapses Thus the evelopment and decay of tetanic stimulation appeared t o correlate with augmentation of transmitter release at crayfish opener excitor synapses (see also Wojtowicz and Atwood, 1985). DISCUSSION Changes in presynaptic voltage during repetitive stimulation have often been ro osed to account for the homos aptic plasticities of aci itation, augmentation, and &in various preparations. For example, several aspects of AP voltage (level of hy er- or depolarization voltage, total footimmediately before an AP,peak to-peak AP amplitude, AP duration) can all affect the amount of transmitter released by a sin le pulse. Hence changes in AP volta es in a train of APs ave often been suspected to cause acilitation in crayfish (Atwood and Wojtowicz, 1986; Dudel, 1965, 1971; Zucker, 1974a,b), squid (Takeuchi and Takeuchi, 1962), and vertebrate neuromuscular (Hubbard and Schmidt, 1963)synapses. Extracellular recordings from crayfish opener-excitor nerve terminals have been interpreted to sug est that APs did not invade the terminal region an% that a growth in the am litude and a rolongation of presynaptic HAPs coul be responsib e for facilitation in a train of APs (Dudel, 1965, 1971). Conversely, extracellular records from these same terminals have been interpreted t o suggest that APs did invade the terminal re 'on (cf. Smith, 1983; Zucker, 1974a,b). 8 u r intracellular recordings from secondary and tertiary branches of the o ener excitor axon resolve most of these controversies. first, APs are indeed conducted without decrement at least into tertiary branches which are within 0.5 space constants of sites of excitatory transmitter release in hysiological salines and within 0.1 space constants w en tetrodotoxin, tetraethylammonium, and 4-amino pyridine are used to block sodium and otassium conductances (Sivaramakrishnan et al., 19885. Second, small but significant changes in peak volta e, foot-to-peak total amplitude, and duration of APs o occur durin brief (9-20 pulses) stimulus trains which produce muc facilitation of transmitter release. However, these small and noncontinuous chan es in AP voltage can in no wa account for all of t e large, continuous increase in PSP amplitude during the brief stimulus train. In this regard, crayfish opener synapses are similar to the on1 other preparation (the squid ant synapse) in whic simultaneous recordings have een made of APs and EPSPs during homosynaptic facilitation (Charlton and Bittner, 1978a,b). As in the squid synapse, changes in AP volta e during a stimulus train may influence transmitter re ease during facilitation (e.g., pulses 5-7 in Fig. 21, but such voltage changes er se are not responsible for the facilitation. Rather, Facilitation must result from some other process, such as residual calcium (Katz, 1966).

CQ

I? -P.

ib'

5

B

B

P

R

f

a

f

a

E K

k

pulses (Dudel, 1971).

a

(1984, 1985) have shown that long-lastin (30-100 msec), artificially induced depolarization oft e openerexcitor synaptic terminals can enhance evoked transmitter release. If DAPs, HAPs, or some other otential change of the synaptic terminals is responsib e for the accumulation of facilitation during repetitive tetanic stimulation, then such voltage changes should correlate with increases of transmitter release following the tetany (i.e., the voltage chan e would be a primary mechanism for augmentation). e have shown that au tation havin a decay time constant of 10-20 sec ( ig. 7) is associate with a large HAP having a decay time constant of 10-20 sec (Fig. 7). We have yet to determine if the H A P directly produces augmentation or if it is an indirect effect of some other membrane activit such as an electrogenic sodium pump (Atwood and djtowicz, 1986; Wojtowicz and Atwood, 1984). Whatever their underlying molecular mechanisms in these terminals, facilitation and augmentation are almost certainly partly responsible for the exquisite sensitivity of transmitter release to the frequenc and duration of tetanic stimuli (Bittner an egundo, 1989). Over 20 years ago, Bernard Katz (1966) su gested that crustacean neuromuscular systems wou d provide an excellent system with which to study various synaptic plasticities found in the mammalian CNS. Indeed, this prediction has been amply fulfilled as the opener neuromuscular preparation has now been shown to exhibit facilitation, augmentation, lon -term otentiation, presynaptic inhibition, and modu ation ( ittner, 1989).

P

6

f

rn-

dr!Ptern, k

k- i

ACKNOWLEDGMENTS This study was supported in part by NIAAA grant #AA07746 to G.D.B. REFERENCES Atwood, H.L. (1976) Organization and synaptic physiology of custacean neuromuscular systems. Prog. Neurobiol., 7:291-391. Atwood, H.L., and Bittner, G.D. (1971)Matchin excitatory and inhibitory inputs to crustacean muscle fibers. J. deurophysiol., 34:157170. Atwood, H.L., and Wojtowicz, J.M. (1986) Short-term and long-term plasticity and ph siological differentiation of crustacean motor synapses. Int. Rev. deurobiol., 28:275-362. Baxter, D.A., and Bittner, G.D. (1981) Intraceellular recordings from crustacean motor axons during presynaptic inhibition. Brain Res., 223:422-438. Baxter, D.A., Bittner, G.D., and Brown, T.H. (1985) Quanta1 mechanisms of long-term synaptic potentiation. Proc. Natl. Acad. Sci. USA, 82:597%5982. Bittner, G.D. (1968a)Differentiation ofnerve terminals in the crayfish opener muscle and its functional significance. J . Gen. Physiol., 511731-758. Bittner, G.D. (1968b) The differentiation of crayfish muscle fibers during development. J. Exp. Zool., 167:439456. Bittner, G.D. (1989) Syna tic plasticity at the crayfish opener neuromuscular preparation. Neurobiol., 29:386-408. Bittner, G.D., and Harrison, J. (1970) A reconsideration of the Poisson hypothesis for transmitter release a t the crayfish neuromuscular junction. J. Physiol. (Lond.),206:l-23. Bittner, G.D., and Segundo, J.P. (1989) Effects of stimulus timing on transmitter release and post-syna tic membrane potential at crayfish neurornuscular junctions. J. &mp. Physiol., 165:371-382. Bittner, G.D. and Sewell, L. (1976) Faci itation at crayfish neuromuscular junctions. J. Comp. Physiol., 109:287-308. Bittner, G.D., and Traut, D.L. (1978) Growth of crustacean muscles: Constancy of fiber number and sarcomere number. J. Comp. Physiol., 124277-285. Charlton, M.P., and Bittner, G.D. (1978aj Facilitation of transmitter release at squid synapses. J. Gen. Physiol., 72:471486.

31

243

FACILITATION AND AUGMENTATION IN CRAYFISH Charlton, M.P., and Bittner, G.D. (1978b) Effect of chan es in resyn aptic potentials on facilitation in squid synapses. J. hen. P\ysiol., 72:487-511. Delaney, K.R,, Zucker, R.S., and Tank, D.W. (1989) Calcium in motor nerve terminals associated with Dosttetanic Potentiation. J. Neurosci., 9:3558-3567. Del Castillo, J., and Katz, B. (1954) Statistical factors involved in neuromuscular facilitation and depression. J. Physiol. (Lond.), 124:57b585 - .- . . .-. Dudel, J. (1965) Potential changes in the crayfish motor nerve terminal during repetitive stimulation, Pflugers Arch. Ges. Physiol., 282:323337. Dudel, J. (1971) The effect of polarizing current on action potentials and transmitter release in crayfish motor nerve terminals, Pflugers Arch. Ges. Ph siol ,324 227 248 Dudel, J., and & f l e r , S.W. (196lj Mechanisms of facilitation a t the crayfish neuromuscular junction, J. Physiol. (Lond.), 155:530-542. Erulkar, S.D., and Rahamimoff, R. (1978) The role of calcium ions in tetanic and posttetanic increase of miniature and plate potential frequency, J. Physiol. (Lond.), 278:501-511. Fuchs, P.A., and Getting, P.A. (1980) Ionic basis of presynaptic inhibitory potentials a t crayfish claw opener. J. Neurophysiol., 43:15471557. Hubbard, J.I., and Schmidt, R.F. (1963) An electrophysiological investigation of mammalian motor nerve terminals. J. Physiol. (Lond.), 1661145-165. Jahromi, S.S.,and Atwood, H.L. (1974) Three-dimensional ultrastructure of the crayfish neuromuscular apparatus. J. Cell. Biol., 63599613. Kandel, E.R., and Schwartz, J.H. (1985) Principals of Neuroscience, 2nd ed. Elsevier, New York, 979 p Katz, B. (1966) Nerve Muscle and $&apse. McGraw-Hill, New York. Katz, B., and Miledi, R. (1968) The role of calcium in neuromuscular facilitation. J. Ph siol (Lond.), 195:481-492. Kretz, R., Shapiro, &, Connor, J.,and Kandel, E.R. (1984) Posttetanic potentiation, presynaptic inhibition, and the modulation of the free

Ca2' in the presynaptic terminals. Exp. Brain Res., Suppl. 9:240283. Mallart, A., and Martin, A.R. (1967) An analysis of facilitation of transmitter release at the neuromuscular junction of the frog. J. Physiol. (Gond.),193:447-466. Sivaramakrishnan, S., Bittner, G.D., and Brodwick, M.S. (1988) Charybdotoxin-sensitive calcium-activated otassium conductance at the crayfish neuromuscular junction. Soc.Geurosci. Abstr., 14:69. Smith, D.O. (1983) Variable activation of synaptic release sites at the neuromuscular junction. Ex Neurol., 80:520-528. Takeuchi, A., and Takeuchi, (1962) Electrical changes in re and post-synaptic axons of the giant synapse of Loligo. J. Gen. Fhysiol., 45~1181-1193. Wiens, T.J. (1984) Triple innervation of the crayfish opener muscle: The astacuran common inhibitor. J. Neurobiol., 16:183-191. Wojtowicz, J.M., and Atwood, H.L. (1983)Maintained depolarization of synaptic terminal facilitates evoked transmitter release at a crayfish neuromuscular junction. J. Neurobiol., 4:385390. Wojtowicz, J.M., and Atwood, H.L. (1984) Presynaptic membrane potential and transmitter release a t the crayfish neuromuscular junction. J. Neuro hysiol., 52:99-113. Wojtowicz, J.M., an$ Atwood, H.L. (1985) Correlation of pres naptic and postsynaptic events during establishment of long-term Lcilitation at crayfish neuromuscular junction. J. Neurophysiol., 54:220230. Zengel, J.E., Magleby, K.K., Horn, J.P., McAfee, D.A., and Yarowsky, P.J. (1980)Facilitation, augmentation, and potentiation of syna tic transmission at the superior cervical ganglion of the rabbit. J. &en. Physiol., 76:213-231. Zucker, R.S. (1974a) Crayfish neuromuscular facilitation activated b constant res naptic action potentials and depolarizing pulses. Phvsiol. CEondI,. 241:69-89. ZuckGr, R.S. (1974b) Excitability changes in crayfish motor neurone terminals. J. Physiol. (Lond.), 214:lll-126. Zucker, R.A., and Lara-Estrella, L.O. (1979) Is synaptic facilitation caused by presynaptic spike broadening? Nature, 278:57-59.

8:

3: