SYNAPSE 7:244-251 (1991)

Synaptic Plasticity at Crayfish Neuromuscular Junctions: Presynaptic Inhibition DOUGLAS A. BAXTER AND GEORGE D. BITTNER De artment of Neurobiology and Anatomy, University of Texas Medical School, Houston, Texas 77225 6 . A . B . ) and Department of Zoology, College of Pharmacy, and lnstitute for Neurological Sciences, University of Texas, Austin, Texas 78712 (G.D.B.)

KEY WORDS

Presynaptic inhibition, Synaptic plasticity, Neuromuscular transmission, Crayfish

Intracellular recordings at sites electrotonically near terminals of the ABSTRACT opener excitor axon in the claw of crayfish (Procambarussimulans) show that stimulation of the inhibitor neuron produces hyperpolarizing or depolarizing presynaptic inhibitory potentials (PIPS). GABA applied anywhere along the length of the opener excitor or inhibitor axons also produces hyperpolarizing or depolarizing potentials. The amplitude of action potentials (APs)at recording sites near some excitor terminals is reduced by an average of 6 mV during presynaptic inhibition, which also reduces excitatory postsynaptic potentials (EPSPs) by 50-70%. The time course of AP reduction equals the time course of EPSP reduction and the amount of AP reduction is independent of the sign or amplitude of the PIPs. All these data are consistent with a hypothesis that a conductance increase produced by GABA in these presynaptic terminals of the excitor axon is responsible for presynaptic inhibition. However, the effect of presynaptic inhibition upon the accumulation of short-term facilitation of excitatory transmitter release is not the same in all muscle fibers. In some terminals, the accumulation of short-term facilitation during short, high-frequency trains of action potentials which are presynaptically inhibited often equals the accumulation of facilitation without inhibition. In other terminals, short-term facilitation accumulated during presynaptic inhibition often does not equal facilitation accumulated in the absence of presynaptic inhibition. These data suggest that some other factor which may contribute to presynaptic inhibition, such as a direct effect to decrease calcium currents, may also affect short-term facilitation in some terminals. INTRODUCTION Presynaptic inhibition is a widespread form of heterosyna tic plasticity in which a set of (inhibitory)nerve termina s secrete transmitter onto neighboring nerve terminals to produce a short-term (1 msec-1 sec) decrease in the number of transmitter uanta released by an action potential (AP)invading t e nerve endings. Presynaptic inhibition is common1 found in the central nervous s stem of vertebrates ( urke and Rudomin, 1977; Ecc es, 1964; Schmidt, 1971) and in the central and eri heral nervous systems of invertebrates (Atwoozanj Wojtowicz, 1986; Baxter and Bittner, 1980, 1981; Dudel and Kuffler, 1961; Kretz et al., 1986a,b; Shapiro et al., 1980). Although the phenomenon of presynaptic inhibition is well known, our understanding of the cellular mechanisms of presynaptic inhibition has been limited by a lack of information about the volta e changes occurring in the affected synaptic termina s. At least three mechanisms have been hypothesized to account for resyna tic inhibition. One mechanism proposes that t e inhi . itory neurotransmitter produces a voltage change which directly decreases transmitter

P

R

B

ry

release via a presynaptic inhibitory otential (PIP) in the excitor terminals (Eccles, 1964; fiicholls and Wallace, 1978; R all, 1978; Schmidt, 1971). Depolarizing PIPs could rejuce transmitter release by decreasing the foot-to- eak total amplitude of APs in the excitor terminals. dnversely, hyper olarizing PIPs could reduce transmitter release by ecreasing the peak voltage of APs invading the excitor terminals. A second mechanism proposes that the inhibitory neurotransmiter produces a conductance increase to one or more ionic species which could “shunt” AP currents and thereby reduce the am litude of presynaptic APs invading excitor terminals YBaxter and Bittner, 1981; Dudel, 1963, 1965; Nicoll and Alger, 1979). This reduction in AP amplitude would not depend upon the si er- or depolarizin ), am litude, or time course$?P:,?of but rather wou d ref ect the time course of the inhibitory

cp

K f

K

R

.g

0 1991 WILEY-LISS,INC.

Received April 19,1990;accepted in revised form August 27,1990.

245

PRESYNAPTIC INHIBITION IN CRAYFISH

conductance increase. A third mechanism roposes that the inhibitory neurotransmitter directly b ocks voltagesensitive calcium channels in nerve terminals (Kretz et al., 1986a,b;Shapiro et al., 1980). Decreased calcium entry would reduce the amount of transmitter release by an AP invading the terminal. In this paper, we describe intracellular recordings from sites electrotonically near the synaptic terminals of the excitor motor neuron of the crayfish opener muscle during presynaptic inhibition. Stimulation paradigms that presyna tically inhibit the release of excitatory transmitter a so reduce the am litude of the excitator AP in some terminals. The re uction in exciduring presynaptic inhibition is independent tatory of the sign and amplitude of the PIP. In these terminals, the time course for the reduction in the excitatory spike amplitude is brief (3-6 msec) and is the same as the time course for the pres a tic inhibition of excitatory postsynaptic potentia s ( PSPs).Furthermore, the time course for the reduction in excitor AP am litudes and in EPSP amplitudes is much briefer than t e 20-50 msec duration of PIPs. Both the reduction in excitor AP amplitude and the EPSP amplitude durin presynaptic inhibition are blocked by picrotoxin and C - free saline. Thus, pres aptic inhibition at these synapses is best explained y a brief inhibitory conductance increase which "shunts" the amplitude of the excitor AP.Finally, we show that pres aptic inhibition can modulate the accumulation of s ort-term facilitation and that this modulation is not the same for all terminals. METHODS The dissection and experimental arran ement were essentially similar to those described in t e preceding paper (Bittner and Baxter, 1991). Intracellular recordings from the Y branch or from a secondary branch of the excitor axon on the ventral surface of the opener muscle were obtained in juvenile cra ish (Procambarus simulans). In some cases, intracel ular recordings were also made from the Y branch or secondary branches of the opener inhibitor motor axon on the ventral surface of the opener muscle. In other cases, intracellular recordings were made from the opener excitor or inhibitor axon in the meropodite segment, 1-3 cm roximal to the Y branch in the propodite segment. h e n presyna tic recordings were made from a Y branch or secon ary branch, postsynaptic potentials were often recorded simultaneously from nearby postsynaptic opener muscle fibers in the propodite. The opener excitor and inhibitor motor neurons were separated and stimulated independently in the meropodite. The relative latency between the stimulation of the excitor and inhibitor axons was adjusted so that EPSP amplitudes were maximally reduced. Since the inhibitor axon is usually larger than the excitor axon, inhibitor APs are conducted slight1 faster than excitor APs. Hence, the optimum interval etween excitor and inhibitor stimulation was usually 0-2 msec to maximize pres aptic inhibition, as previously reported (Atwood and ittner, 1971; Baxter and Bittner, 1981). Most of this reduction in EPSP amplitude was shown to result from presynaptic inhibition by inserting a second electrode into the muscle fiber in order to inject brief current pulses which gave membrane de olarizations of a time course and amplitude similar to PSPs. These artificial

f

P

B

d s

F2

R F

i?

x"

a

i"

B

EPSPs were then aired with inhibitory stimulation and the percent re uction of the artificial EPSPs was used as an index for the contribution of postsyna tic inhibition (Atwood and Bittner, 1971; Baxter and ittner, 1980). In 18 fibers tested, postsyna tic inhibition was found to contribute only about 4% o the observed 50-70% reduction in EPSP amplitudes. Addition of GABA, picrotoxin, etc., was accomplished by adding the drugs to the saline which continuously perfused the preparation. Addition or removal of drugs usually required the perfusion of 3-5 bath volumes (4-5 ml) and took 15-60 sec to complete. RESULTS Intracellular recordings from sites near excitor and inhibitor terminals Recordings from the Y branch or a secondary branch of an excitor axon are within 0.3-0.5 space constants of sites of excitor transmitter release (Baxter and Bittner, 1981; Bittner, 1989; Wojtowicz and Atwood, 1984).Presynaptic inhibitory sites are located on excitor release sites or roximal to those sites (Jahromi and Atwood, 1974). 1 s previously reported (Baxter and Bittner, 19811, simultaneous intracellular recordings of APs were recorded from the Y branch or a secondar branch of an o ener excitor motor neuron and EP Ps were recorde from a nearby postsynaptic muscle fiber. In the present study of 25 excitor axons recorded at the Y branch or at a secondary branch, the average resting potential was -74 t 7 mV and the average AP amplitude was 93 (? 4 SD) mV. An AP was followed by a depolarizing after potential (DAP) havin an average am litude of 1022 mV measured 2 msec a ter the onset o f t e AP. The APs recorded from inhibitor synaptic terminals were similar to those seen in the excitor terminals. For example, in Figure 1A (trace l),the peak voltage of the excitor AP was 21 mV and the total amplitude was 100 mV. The excitor AP was followed b a 12 mV DAP measured 2 msec after the onset of the In this same preparation, the presynaptic recording electrode was removed from the excitor neuron and was inserted into the inhibitor neuron. The eak voltage of and the total the inhibitor AP was also 21 mV (Fig. amplitude was 98 mV. The inhibitor AP was followed by a DAP that was 8 mV at 2 msec after the onset of the AP. For 20 inhibitor axons recorded at the Y branch or a secondary branch, the average resting potential was -75 2 6 mV, the average AP amplitude was 962 4 mV, and the average DAP was 9 ? 2 mV measured 2 msec after the onset of the AP. Stimulation of the inhibitor axon roduced both presynaptic inhibitory otentials (P s) in excitor Y branches or secondary73ranches (Fig. 1D) and inhibitory ostsynaptic otentials (IPSPs)in ostsynaptic muscle ibers (Fig. 1 ). No differences in IP amplitude, sign,

B

€f

F

2

a

B

K

d

18)

P

E

& 8

g

I?

8

p,6

80 to 500 in different nerve axons. PIPs had a time constant of decay (7) equalling 20-50 msec and a rise time from foot-to-peak of 3-7 msec. The amplitude

D.A. BAXTER AND G.D. BITTNER

246

E!

0

20-

a 3

%/

O

A

A

D

I

J+

1 n

B

E

&=;

-3

0

Fig. 1. A, B: Simultaneous intracellular records of pres from secondary branch of the excitor nerve axon (A) and E%;%g: nearb osts aptic muscle fiber (B). The excitor APs and EPSPs were recor&$ u n g control conditions (traces I), during presynaptic inhibition (traces 2), and with 0.5 mM GABA added to the bath (traces 3). Stimulation of the inhibitor de olarized this excitor axon and reduced its AP peak voltage from 100m e to 93 mV. Addition of 0.5 mM GABA to the bath further depolarized this excitor axon and further reduced its AP eak voltage to 83 mV. Dotted line = -79 mV; solid line = 0 mV. Intracellular records of an AP from the secondar branch of the inhibitor motor neuron (C), the PIP from the Y b r a n d of the excitor motor neuron (D), and an IPSP from a postsynaptic muscle fiber (El. The resting potentials of the inhibitor axon, excitor axon, and muscle fibers were -77 mV, -79 mV, and -70 mV, respectively. Each trace in Figure 1A-E are computer averages of 100 repeated trials at 2 Hz. All recordings made from the same preparation Calibration lines: A, C = 10 mV, 1msec; B = 0.25 mV, 5 msec; D, E = 0.25 mV, 2 msec.

C-g

affected by presyna tic inhibition, in contrast to APs recorded from the Y ranch. Because the recording site at the Y branch of the excitor is about 50 to 500 m away from the presynaptic inhibitory contacts, the oiserved reduction in the peak volta e of the excitor AP during presynaptic inhibition proba ly underestimates the reduction in AP peak voltage which occurs at the sites of presynaptic inhibition. In fact, the potential change observed at the Y branch probably represents 40 to 70% of the total potential change that occurs at sites of resynaptic excitor transmitter release (Baxter and $ittner, 1981; Fuchs and Getting, 1980;Wojtowicz and Atwood, 1984). For example, the addition of GABA in su ra-maximal concentration (0.5 mM) to the bath per usate reduced the peak voltage of the excitor AP by 17 mV (trace 3 in Fig. 1A) (see also Baxter and Bittner, 1981). The reduction in excitor AP during bath application of GABA (see below and Fig. 2) may represent the full effect of presynaptic inhibition upon the excitor AP amplitude at excitor release sites (Baxter and Bittner, 1981).By this reason, the 7 mV reduction at the Y branch in the peak vo tage of the excitor AP by presynaptic inhibition in Fi re 1A would reflect about 41% of a possible 17 mV re uction of peak AP voltage at the excitor release sites. Effect of GABA and other drugs on excitor and inhibitor axons Although inhibitory syna ses are made on the excitor axon only at or near sites o excitor transmitter release (Atwood and Wojtowicz, 1986; Jahromi and Atwood,

g

(50300 pV), rise time (4-9 msec), and T (30-80 msec) of IPSPs also varied from muscle fiber to muscle fiber, as previously reported (Baxter and Bittner, 1981; Onodera and Takeuchi, 1979). In order to observe the effect of presynaptic inhibition upon APs in syna tic terminals of the excitor axon, we stimulated the in ibitor axon 8-12 times at 10 msec intervals so as to facilitate the release of inhibitory transmitter (Atwood and Bittner, 1971). The excitor axon was then stimulated so that the AP arrived at excitor syna tic terminals soon (1-2 msec) after the last impulse of t\e inhibitor train arrived at the inhibitor terminals. This stimulus paradigm produced maximal presynaptic inhibition of excitor transmitter release as evidenced b a 40-85% reduction in EPSP.amplitude (trace 2 in ig. 1B; Baxter and Bittner, 1980). In addition to reducing the EPSP amplitude, this stimulus aradigm also reduced the peak voltage of excitor AP From 100 mV to 93 mV (trace 2 in Fig. 1A). The same attern of inhibitory stimulation reduced artificial EP~ P by s an average of 440, su gesting that postsynaptic inhibition is relatively smafi in these opener muscle fibers, as reported in previous studies (Atwood and Bittner, 1971; Baxter and Bittner, 1980; Dude1 and Kuffler, 1961).With this stimulus paradigm, presyna tic inhibition reduced the eak voltage of the excitor b an avera e of 6 ? 2 m (n = 11). APs recorded from t e meropo ite segment of the excitor axon were not

K

P

i

i

8

As,

%

P

inK

r

P

247

PRESYNAFTIC INHIBITION IN CRAYFISH

19741, the excitor axon was sensitive to GABA alon its entire length. Figure 2 shows that GABA reducecfAP amplitude in the excitor axon near the Y branch (Fig. 2A), in the excitor axon in the mero odite limb segment (1-3 cm proximal to the Y branch) TFig. 2B), and in the inhibitor axon near the Y branch (Fig. 2C). The doseresponse curve to GABA was similar in shape and sensitivity at all three recordin sites, reaching a maximum effect at about 0.5 mM G h A (Fig. 2). The threshold concentration for GABA effects (2 x 10-5M) was also the same at all three recording sites. Similar data were obtained from the stretcher inhibitor in the meropodite segment. The ability of 0.5 mM GABA to reduce AP am litude was stable for over 2 hours in the stretcher inhifitor axon (n = 2), in the opener inhibitor near the Y branch (n = 4),in the excitor axon in the meropodite (n = 51, and at the Y branch or secondary branches (n = 5).For 25 excitor axons recorded from the Y branch or a secondary branch havin an AP amplitude of 93 t 4 mV and a DAPof 10 i 2 m%, 0.5 mM GABA reduced the AP by 15 i 7 mV and the DAP by 6 t 2 mV. GABA (0.5 mM) reduced the AP recorded from the excitor in the mero odite (n = 8) by 16 i 4 mV and the DAP by 7 2 2 mV. 8ABA also reduced the AP amplitude recorded from the inhibitor Y branch (n = 9) b 16 i: 6 mV and the DAP by 6 i 2 mV. The adition of 2.5 x 10-5M icrotoxin or C1- free salines reversibly blocked the e ects of GABA to decrease AP total am litude at all three recording sites. Picrotoxin or C1- ree salines also reversibly eliminated recordin s of PIPs in the excitor axon and the recording of I SPs in the postsyna tic muscle fiber. The adition of 0.5 mM GABA to the bath perfusate had similar effects on the restin membrane potential of the excitor axon at the Y branc or the meropodite, as well as the inhibitor axon at the Y branch or anywhere else in the propodite segment. For all three recording sites, the equilibrium potential for 0.5 mM GABA was about -69 mV (also see Baxter and Bittner, 1981). The effect of GABA to hy erpolarize or depolarize the membrane depended on t e relationship between the GABA reversal otential (- 69 mV) and the resting membrane potentiaf The sign and am litude of PIPS in individual excitor Y branches or secon ary branches also de ended on the relationship between this GABA reversa potential and the resting membrane potential. Perfusion with C1- free saline or picrotoxin blocked both the reduction in excitor AP amplitude and the reduction of excitor transmitter release which otherwise occurred during inhibitor stimulation at an optimum interval for presynaptic inhibition of EPSPs (see Methods).

r4?

P

8

a

izing PIPs ran ed from +120 to t340 pV and reduced the excitor spi e amplitude by an average of 6.0 t 1.1 mV. In the remaining six inhibition was associated with which ranged from -80 to -380 izing PIPs reduced the excitor age of 5.9 i 1.0 mV. Hence, the ability of presynaptic inhibition to reduce excitor AP amplitude was not affected by the sign or the amplitude of PIPs. Pres aptic only inhibition reduced the amplitude of excitor when those APs were recorded from Y branches or secondary branches near the synaptic terminal region; action potentials recorded from the meropodite segment of the excitor axon where no inhibitor synapses have been reported were not effected by presynaptic inhibition. As a further test, the si of two of the six hyperpolarizing PIPs was reversed ( ig. 3, arrows) by direct mtracellular injection of GI- from KC1-filled microelectrodes by passing 5-10 mV hyperpolarizing ulses. No change in resting membrane otential was oYIserved following this C1- in'ection. owever, after injectin C1- for 30-60 sec, t e hyperpolarizing PIPs became epolarizing PIPs. The hyperpolarizing PIPs origmally reduced the excitor AP am litude by 7.0 and 6.0 mV, res ectively; following Cf- injection, the de olarizing IPS reduced the excitor AP by 5.8 and 6.0 m , respectively. Thus, a complete reversal of sign and amplitude had little or no effect upon the reduction of excitor AP am litude during resynaptic inhibition. investigate t e time course of presynaptic inhibition, a single excitor AP was ven at varyin times after ten inhibitor pulses at 100 E Fz. The electro es stimulat-

&

P

If

h

%

rp

%

&I

K

cf

1

E

B

f

Mechanism of presynaptic inhibition To investigate the possibility that changes in membrane potential per se might produce presynaptic inhibition, we compared the ability of hy er- and depolarizin PIPs to produce presyna tic inhi ition (Fig, 3). The in ibitory axon was stimu ated by short, highfrequency trains (8-12 pulses at 100 Hz) and the excitor axon was stimulated after the last inhibitory pulse at an optimum interval to maximize presynaptic inhibition (0-2 msec, see Methods). Pres aptic inhibition was associated with depolarizing PI s in five of 11 recordings from Y branches or secondary branches near excitor synaptic terminals. The amplitude of these depolar-

a

7

F

fl

I -I

I

0

-400

Fig. 3. Reduction of the total amplitude (mV ordinate) of a single excitor AP recorded in the Y or secondary branch vs. the sign (h erpolarizing = negative; depolarizing = positive) and amplitude (pvabscissa) of the potential change at the last PIP in a train of %12 PIPs at 100 Hz. The excitor AP was given at an optimum latenc followin the last inhibitor pulse to produce maximum reduction in A$ amplituse with respect to the excitor AP amplitude a t 1 Hz excitor stimulation without any inhibitor stimulation. Closed s bols: PIPs recorded from excitor terminals. 0 en symbols: hyperpoEzing PIPS converted to depolarizing PIPSby 81- iontophoresis(see text).

D.A. BAXTER AND G.D. BITTNER

248

A

> E a

Y

m

loot

95

0 c

P

-

s

J -3 -2 -I 0 I 2 3 4 h k r v o i Between Excitotory ond Inhibitory Stimulus (msec)

0

-4

@

Q)

a

n

a

Fig. 4. Reduction of total AP amplitude (mV) recorded in the Y or secondary branches of the excitor axon (left-handed ordinate, solid line) and percent reduction in EPSP amplitude recorded in a subsynaptic muscle fiber (right-hand ordinate, dotted line) vs. the interval between excitor axon stimulation and inhibitor axon stimulation in the meropodite. For this graph, a “zerolatency” interval was defined as the optimum interval for maximum presynaptlc reduction of the excitor AP. Each data point represents the average of four preparations; standard deviations from the mean shown for AP amplitudes.

ing the excitor and inhibitor axons in the meropodite were placed at a proximately equal distances (3-5 cm) from the recor ing site in the excitor Y branch or secondary branch. A single excitor AP was delivered at various relative latencies with respect to the last inhibitor AP (see Methods). For each synapse, the timinF for optimum presyna tic inhibition was plotted as ‘zero relative latency.” figure 4 de icts the average reduction at four synapses in AP an EPSP amplitude during resynaptic inhibition relative to zero relative latenc !he time course for the reduction in excitor AP ampi: tude (3-5 msec) was much shorter than the duration of the voltage associated with a PIP (20-50 msec) recorded

2

s

(- 300 and

Reduction of excitor APs and EPSPs during presynaptic inhibition To investigate the effects of presynaptic inhibition on the eak voltage and total am litude of excitor APs, the inhititor axon was stimulate alone a t 100 Hz for 4-8 pulses before pairing 6-12 inhibitor ulses with 6-12 excitor ulses at 100 Hz. The interva between excitor APs an inhibitor APs was adjusted to product maximal reduction of the first excitor AP and associated EPSP recorded from a nearb postsynaptic muscle fiber (see Methods and Fig. 4). Tge reductions in excitor APs and EPSPs at each pulse during resynaptic inhibition were compared with APs and EPs+ s produced b stimulating the excitor alone in brief (8-14 pulse) 100 z trains. typical data from over 40 the excitor was stimulated alone

a

B

Y

b

amplitude (filled squares) decreased at the second AP.EPSP amplitude (filled circles) continuously increased during the entire train (see receding pa er (Bittner and Baxter, 1991) for further %xussion). 8urin presynaptic inhibition (0 en symbols, dotted lines) t e total amplitude and pea! voltage

a

n v)

n W

Y J*t!

I

P

2

4

6

8

I

,

1

,

,

0

Pulse Number Fig. 5. Plot of AP amplitudes (mV) recorded in an excitor secondary branch (A) and EPSP amplitudes (pV)recorded in a nearby postsynaptic muscle fiber (B)durin 100 Hz stimulus trains. In A, squares plot foot-to-peakAP total ampbtudes (left-hand scale) and circles plot peak voltages attained by APs (right-hand scale). Solid symbols in A and B plot AP or EPSP values for a train of ten excitor APs at 100 Hz. Open symbols plot excitor AP or EPSP values when the first seven excitor APs were presynaptically inhibited (see inset). Six inhibitor APs preceded the first excitor AP,the sixth occurring at the optimum interval to maximize presynaptic inhibition of the first excitor AP.The first seven excitor APs were presynaptically inhibited by the last seven inhibitor APs (inhibitor pulses 612). The last three excitor APs (excitor pulses 8-10) were .ven without being paired with an inhibitor pulse (open symbols with sfishes). In other words, release of transmitter from the excitor axon was presynapticdly inhibited for excitor pulses 1-7 and was not presynaptically inhibited for excitor pulses

%lo.

ofAPs and EPSP amplitudes were consistent1 reduced. For example, AP total am litude and peak vo tage were reduced b 12-20 mV in igure 5, de ending on pulse number; E! PSP amplitude was reduce by 50-70% at the each of the first seven pulses. In Figure 6 at the sixth pulse, AP peak voltage was reduced by 8 mV and total amplitude was reduced by 6 mV while EPSP amplitude was reduced by about 75%. In Fi re 6, presynaptic . In other preparainhibition also eliminated the DA F tions, reductions in AP am litudes ranged from 5 to 21 mV and reductions in EPSb amplitude ranged from 40 to 85%.

%

2

r“

-22 >

-20 g

.2

a

1

1

1

1

1

1

I

I

O

f?

2

l

0 E aloneI Istirnulolion IllII,I E 0 E* 1 slirnulofion I t I I l l II It II II II II l l l

B I000-

500 -

/?

*-PI I# 2

@

i

t

I

/ I

I

I

4 6 8 Pulse Number

Fig. 6. Plot of AP amplitudes (A) recorded in the excitor Y branch and EPSP amplitudes (pV) recorded in a nearby postsyna tic muscle fiber (B) during 100 Hz stimulus trains. In A, squares plot got-to-peak AP amplitudes (left-hand scale) and circles plot peak voltages attained by APs (right-hand scale). Solid symbols in A and B plot AP or EPSP values for a train of nine excitor APs a t 100 Hz. Open symbols plot excitor AP or EPSP values when the first six excitor APs were P r e s Faptically inhibited (see inset). Seven inhibitor APs preceded the irst excitor AP,the seventh occurring a t the optimum interval to maximize presynaptic inhibition of the first excitor AP.The first six excitor APs were presyna tically inhibited by the last six inhibitor APs (inhibitor pulses 7-12). T i e last three excitor APs (excitor pulses 7-9) were given without being paired with an inhibitor pulse (open symbols with slashes). In other words, release of transmitter from the excitor axon was presyna tically inhibited for excitor pulses 1-6 and was not DresvnaDticallviniibited for excitor Dulses 7-9.

The effect of presyna tic inhibition upon short-term acilitation An AP which invades the excitor synaptic terminals is followed by a eriod lasting several hundred milliseconds during w ich a second AP releases more transmitter quanta than the first AP (Bittner and Sewell, 1976; Zucker, 1974). This phenomenon is called (short-term) facilitation. In the preceding aper (Bittner and Baxter, 1991), we described intracel ular recordings near excitor synaptic terminals during repetitive stimulation of the excitor axon and accumulation of facilitation in excitor terminals. Although the accumulation of facili-

P

K

P

E

Y presi?

1

I

tation is associated with small increases in AP total amplitude and duration, it is doubtful that these changes are responsible for the increase in transmitter release. To investigate the effect ofexcitor AP am litude on the accumulation of facilitation, we compared PSPs produced by a noninhibited train of excitor APs (Figs. 5, 6, filled symbols, solid lines) with EPSPs produced by noninhibited APs (Figs. 5,6; slashed o en symbols,solid lines) that follow a set of inhibited A3s (Figs. 5,6, open symbols, dotted lines). As described in the previous section, excitor APs and EPSPs are reduced in amplitude during presynaptic inhibition compared to noninhibited (control) excitor APs and EPSPs. Immediately followingthe release from presynaptic inhibition (Fig. 5: arrow), the amplitude of the next excitor AP and amplitude of the next EPSP (slashed symbols) often abruptly increased to the same values observed for the corresponding pulse in the noninhibited train. Such a response was observed in 7/10 preparations from which recordings were made from secondar branches and nearby postsynaptic muscle fibers on t i e ventral surface of the muscle. In 2/10 preparations recorded from the ventral surface, the recover of the excitor AP and the EPSP amplitude to levels ottained in noninhibited trains was slower than the recovery observed for the fibers cited above. For example, in Fi aptic inhibition reduced excitor AP amp itudes at t e sixth pulse in a 100 Hz train b about 6 mV and EPSP amplitudes by about 67%. I? owever, AP and EPSP amplitudes at the seventh-ninth pulses following the release from inhibition (Fie. 6: arrow) were reduced comDared to the seventhni&h AP and EPSP amplitudes'in the control train. Lastly, in one preparation, the excitor AP amplitude returned to control values immediately following its release from inhibition but EPSP amplitudes remained dem-essed. All these data taken together suggest that th; modulation of facilitation may Got be thgiame for terminals on different opener muscle fibers.

DISCUSSION Transmitter release is a function ofAP amplitude and duration as well as the level of polarization in the syna tic terminal immediately before the AP.For examp e, constant or brief depolarizations of squid @ant synapses or vertebrate neuromuscular syna ses depress evoked transmitter release (Charlton andPBittner, 1978; Hubbard and Willis, 1968; Takeuchi and Takeuchi, 1962). Conversely, a hy erpolarizin otential may reduce the peak voltage o an AP an ence reduce transmitter release (Nicholls and Wallace, 1978; Ryall, 1978). Thus, a change in the level of terminal polarization has been roposed as the mechanism under1 'n resynaptic infibition (Eccles, 1964; Nicholls and $af face, 1978; Ryall, 1978; Schmidt, 1971). Since hyper- and depolarizing PIPSreduce excitor AP amplitude to the same extent in many terminals (Fig. 31, it is unlikely that chan es in terminal polarization account for presynaptic in ibition in the crayfish opener excitor. This conclusion is in agreement with the observation that a change in the sign of a PIP bas very little effect on the reduction in AP amplitude associated with the PIP in any terminal examined. A reduction in excitor spike amplitude might, however, be roduced by or other a presynaptic conductance increase to

P

f!

ft

a

d-

250

D.A. BAXTER AND G.D. BITTNER

ACKNOWLEDGMENTS ion(s) whose combined equilibrium (reversal) potential is near the restin membrane otential for the synaptic Su ported in part by NIAAA grant #€A07746 to terminal. The in ibitory con uctance increase would G . D . ~ . act to “shunt” the excitor AP amplitude or block the com lete invasion of the excitor terminal regon by the REFERENCES APK ahromi and Atwood, 1974).Either effect of presynaptic inhibition would reduce transmitter output. Atwood, H.L., and Bittner, G.D. (1971) Matching of excitatory and If conductance increases are the underlying mechainhibitory inputs to crustacean muscle fibers. J. Neurophysiol., 34:157-1 70. nism for presynaptic inhibition in some opener excitor H.L., and Wojtowicz, J.M. (1986) Short-term and long-term terminals, then the time course for the reduction in Atwood, plasticity and physiological differentiation of crustacean motor synexcitor AP amplitude should be briefer than the 20-50 apses. Int. Rev. Neurobiol., 28:275-362. msec duration of the PIP in those terminals. The time Baxter, D.A., and Bittner, G.D. (1980) The normal accumulation of facilitation during presynaptic inhibition. Brain Res. 189:535-539. course of the postsynaptic inhibitory conductance inD.A., and Bittner, G.D. (1981) Intracellular recordings from crease produced by this same inhibitor axon in muscle Baxter, crustacean motor axons during presynaptic inhibition. Brain Res., fibers is about 5-10 msec (Onodera an Takeuchi, 1976, 223:422428. 1979) whereas the duration of the IPSP is about 30-80 Bittner, G.D. (1989) Synaptic plasticity at the crayfish opener neuromuscular preparation. J . Neurobiol., 29:386408. msec. Since the duration of a PIP is very similar to the G.D., and Baxter, D.A. (1991) Synaptic plasticity a t crayfish duration of an IPSP, one might expect that the time Bittner, neuromuscular junctions: facilitation and augmentation. Synapse, course of the inhibitory conductance increase durin a 7:235-243. PIP would also be 5-10 msec. In fact, the reduction of fP Bittner, G.D., and Sewell, L. (1976) Facilitation a t crayfish neuromuscular junctions, J. Comp. Physiol., 109:287-308. amplitude and EPSP amplitude in some terminals has a J.A., and Bittner, G.D. (1989) Presynaptic inhibition in the time course of 5-10 msec (Fig. 4).All these data are in Blundon, crayfish opener neuromuscular junction including effects of ethanol. agreement with the hypothesis that presynaptic inhibiSOC.Neurosci. Abstr., 15:191. tion in some of these opener excitor terminals is due to a Brown, D.A., and Marsh, S. (1978)Axonal GABA receptors in mammalian peripheral nerve trunks. Brain Res., 156:187-191. conductance increase, probably to C1-, due to the action R.E., and Rudomin, P. (1977) Spinal neurons and synapses. In: of GABA receptors. Such GABA receptors ap ear to be Burke, Handbook of Physiology, Vol. 1, Part 2. Williams and Wilkins, distributed over much of the membrane o f t e opener Baltimore, pp. 877-944. excitor since the entire axon is sensitive to GABA at all Charlton, M.L., and Bittner, G.D. (1978)Effect ofchanges in presyna tic potentials on facilitation in squid synapses. J . Gen. Physiof; points examined in the crayfish claw (Fig. 2). Hence, the 72:487-511. inhibitor axon would also appear to have GABA recep- Dudel, J. (1963) Presynaptic inhibition of the excitator nerve termitors distributed over its entire peripheral axon as renals in the neuromuscular junction of the crayfish. Jflugers Arch. ported for mammalian peripheral nerve axons (Brown Ges. Physiol., 277:537-557. and Marsh, 1978). Hence, GABA-releasin synapses Dudel, J . (1965) The mechanism of resynaptic inhibition at the crayfish neuromuscular junction. fflugers Arch. Ges. Physiol., made anywhere on either axon-and per aps other 284:66-80. peri heral and CNS axons as well-could affect AP Dudel, J., and Kuffler, S.W. (1961) Presyna tic inhibition at the am itude andor calcium channels (see below). crayfish neuromuscular junction. J. Physiol. Lond.), 155:543-562. ! & relative contribution of different mechanisms of Eccles, J.C. (1964) The Physiology of Synapses. Academic Press, New York. presynaptic inhibition may differ at different synapses P.A., and Getting, P.A. (1980) Ionic basis of presynaptic inhibior different opener muscle fibers. For example, the Fuchs, tory potentials at crayfish claw opener. J . Neurophysiol., 43:1547effect of presynaptic inhibition on the growth of facilita1557. tion is not the same on all muscle fibers (compare Figs. 5 Hubbard, J.I., and Willis, W.D. (1968) The effect of depolarization of motor nerve terminals u on the release of transmitter by nerve and 6). It is possible that GABA has a second effect on impulses. J. Physiol. (Lon&, 194:381407. resynaptic terminals of the opener excitor axon to Jahromi, S.S., and Atwood, H.L. (1974) Three-dimensional ultrastruc$ecrease the conductance of calcium channels. This ture of the crayfish neuromuscular apparatus. J. Cell Biol., 63:599613. mechanism would be rather similar to that pro osed for R., Sha iro, E., and Kandel, E.R. (1986a)Presynaptic inhibition A Zysia (Kretz et al., 1986a,b; Shapiro et a?., 1980) Kretz, # a n identified presynaptic neuron. I. Physiological mechanisms. J. w ere the neurotransmitter (histamine) decreases cale u r o p h p l , ,55:113:129. cium conductance and increases potassium conductance Kretz, R., S apiro, E ,Bailey, C.H., Chen, M., and Kandel, E.R. (198613) Presynaptic inhibition produced by a n identified inhibitory neuron. during presynaptic inhibition. In bullfrogs, presynaptic 11. Presynaptic conductance changes caused by histamine. J . Neuroinhibition in the spinal cord is also associated with two physiol., 55:131-146. kinds of GABA rece tors, one of which decreases AP Nicholls, J., and Wallace, B.G. (1978)Modulation oftransmission at an amplitude and one o which does not (Peng and Frank, inhibit0 synapse in the central nervous system of the leech. J. 1989a,b).We (Blundon and Bittner, 1989)have recently Physiol.%ond. 1,281:157-1 70. resented some preliminary evidence that there are at Nicoll, R.A., and Al er, B E (1979) Presyna tic inhibition Transmitter and ionic mecganisms: Int. Rev. Neuroiiol., 21:217-258. feast two pharmacolo ically distinct GABA receptors in Onodera, K., and Takeuchi, A. (1976) Inhibitory postsynaptic current the terminal re ‘on o the excitor axon which resemble in volta e clamped crayfish muscle. Nature, 263353-154. GABAAand GA%A receptors in vertebrates (Peng and Onodera, and Takeuchi, A. (1979) An anal sis of the inhibitor ost-syna tic current in the voltage-clampe8 crayfish muscle. ? Frank, 1989a,b).G h A , receptors enerall produce an bhysiol. (Eond.),286:265-283. increase conductance to C1- and re uce pea AP voltage Peng, Y., and Frank, E. (1989a) Activation of GABA, rece tors cause whereas GABABrece tors enerally produce a decrease presynaptic inhibition a t synapses between muscle spindye afferents and motoneurons in the spinalcord ofbullfrogs. J . Neurosci., 9:1502in conductance to Cap+ anfhave little effect on peak AP 1515. voltage. Thus, the effect of GABA in a given o ener Y., and Frank, E. (1989b)Activation of GABAAreceptors causes terminal may be determined by the relative num er of Peng, presynaptic and postsynaptic inhibition a t syna ses bet,ween muscle GABA receptors which produce either a C1- conducr d l e afferents and motoneurons in the spina? cord of bullfrogs. J. tance increase or a Ca+ conductance decrease. eurosci, 9:1516-1522.

a

l

2

R

a

e

x

P

f

8,

f

E

1

+

PRESYNAPTIC INHIBITION IN CRAYFISH Ryall, R.W. (1978) Presynaptic inhibition. Trends Neurosci., 3:16P 166.

Schmidt, R.F. (1971) Pres aptic inhibition in the vertebrate central siol 63-3&101. nervous system. Ergeb. Sharro, E., Castellucci, U . 8 , anh Kandel, E.R. (1980) Presynaptic in ibition in Aplysia involves a decrease in the Ca++current of the resynaptic neuron. Proc. Natl. Acad. Sci. USA, 77:1185-1189. TaReuchi, A,, and Takeuchi, N. (1962) Electrical changes in pre- and

$?I

251

post-synaptic axons of the giant synapse of Lolzgo. 3. Gen. Physiol.,

45:1181-1193.

Wojtowicz, J.M., and Atwood, H.L. (1984) Prespaptic membrane potential and transmitter release a t the crayfish neuromuscular junction. J. Neuro hysiol., 52:99-113. Zucker, R.S. (1974) trayfish neuromuscular facilitation activated b constant res naptic action potentials and depolarizing pulses. Physiol. (Eondl), 241:69-89.

?

Synaptic plasticity at crayfish neuromuscular junctions: presynaptic inhibition.

Intracellular recordings at sites electronically near terminals of the opener excitor axon in the claw of crayfish (Procambarus simulans) show that st...
939KB Sizes 0 Downloads 0 Views