SYNAPSE 6:91-100 (1990)
Phorbol Ester Enhances SvnaDtic Transmission at Crusticeai Neuromuscular Junctions E. GILAT AND B. HOCHNER The Otto Loewi Center for Cellular and Molecular Neurohiology and the Department of Neurohiology, The Hebrew University of Jerusalem, Israel
KEY WORDS
Phorbol ester, Synaptic release, Transmitter release, Prawn
ABSTRACT
Effects of phorbol ester (PE) (4P-phorbol-12,13-dibutyrate) on transmitter release were studied in the deep extensor neuromuscular system of the prawn, Macrobrachium rosenbergii. Our findings show that PE enhances transmitter release as indicated by an increase in the quanta1 content. PE had no post-synaptic effects. The increase in release is accompanied by a slight decline in twin pulse facilitation, suggesting a minor increase in Ca2+entry. The fact that the increase in Ca2+entry has a minor contribution to the PE effect is supported by the following observations: the duration of facilitation was not affected by PE, and 3,4-diaminopyridine (3,4-DAP),which by itself increased release, did not reduce the effect of PE. The time course of release was measured from synaptic delay histograms, upon which PE had no effect. This finding indicates that protein kinase C (PKC) is probably not involved in the rate limiting step of the process of secretion. The log/log plot of the initial part of the delay histogram is not affected by PE, suggesting a lack of effect on cooperativity of the release process. Increased release by loading the presynaptic terminal with Ca2+either by pretreatment with Ca2+ ionophore or by frequent stimulation prevented further increase in release by PE. We concludethat the main effect of PE is confined to stages of release that are secondary to the first elevation in presynaptic Ca2+.PKC in this system probably plays a role in long term modulation of release. and it can be activated in processes leading to presynaptic Ca2+accumulation.
INTRODUCTION Recently, it was sug ested that activation of protein kinase C (PKC), a a2+-and phospholipid-dependent enzyme, enhances synaptic transmission in various preparations-in the frog neuromuscular ‘unction (Haimann et al., 1987; Sha ira et al., 1987j in the sensory-motor s apse of plysia (Hochner et al., 1986a), and in rat hippocam us (Malenka et al., 1986b). Phorbol ester (PE) was a so implicated in the mechanism of long-term potentiation (Akers et al., 1986). The mechanisms underlying this increase in synaptic release are still unknown. For example, activation of PKC can increase Ca2+ conductance by modulating Ca2+channel activity in Aplysia bag cells (De Riemer et al., 1985) or can reduce potassium conductance (Baraban et al., 1985; Farley and Auerbach, 1986; Malenka et al., 1986a).This may lead to the broadening of action potentials and thereby increase release (Hochner et al., 1986131. In non-s.ynaptic systems there is evidence that PKC is directly involved in the secretion processes. Diacylglycerol and phorbol ester were reported to stimulate secretion without raising cytoplasmic free calcium in human platelets (Rink et al., 1983). PKC also stimulates Ca-dependent and Ca-independent release of neurotransmitter from PC12 cells (Pozzan et al., 1984). In the present study, we tested for the effect of
8
tg
@ 1990 WILEY-LISS,INC.
1
P
4P-phorbol-12,13-dibutyrate(PE) on transmitter release using a crustacean neuromuscular system that enables the detection of single quanta and direct depolarization of the terminal (Dudel, 1981). The results suggest that PE may increase transmitter release by acting in more than one way. Nevertheless, on the basis of indirect techniques, we conclude that the major effect of PE involves modulation of stages in the release process that occur after the entry of Ca2+ions but that do not affect the time course of release.
MATERIALS AND METHODS The L1 and L2bundles of the dee extensor abdominal medialis (DEAM) of the prawn acrobrachium rosenbergii were used (Miller et al., 1985). Excitatory axons
2
were stimulated with a small suction electrode. For intracellular recordin s, 3 M-KC1 microelectrodes (10-15 Ma)were used. ynaptic currents were recorded using “macropatch” electrodes (10-20 bm inner diameter) (Dudel, 1981). The macropatch technique is based on a patch-clamp amplifier which enables direct stimulation and recordings, under loose patch conditions. Stimulation arameters can be controlled to obtain different leve s of release. Unitary events can be initi-
5
P
Received November 27,1989; accepted in revised form February 14,1990. Address reprint requests to B. Hochner, The Otto Loewi Center for Cellular and Molecular Neurobiology, Institute of Life Science, The Hebrew University of Jerusalem, Israel
E. GILAT AND B. HOCHNER
92
ated and observed under stimulation conditions that lead to a low level of release. The macropatch electrodes were filled with a solution composed of imM) NaCl, 220; KCl, 5.4; CaCl,, 6 (unless otherwise indicated): MgC12, 2.5; Tris maleate, 10, which was also used as the bathing medium. The pH was adjusted t o 7.4 with NaOH. A constant flow (approximately one chamber volume, 3 ml per minute) was maintained. Bath temperature was controlled by a heat exchanger with an accuracy of k1"C. Recordings were averaged on a Nicolet (model 1174) signal averager. Traces were digitized (20 psecladdress) and stored on a video-tape (Neurocorder, model DR484). Quanta1 units, current amplitude, and delays were measured using home written software on an Olivetti 24 SP computer. In the case of nerve stimulation, the synaptic delay was measured from the negative peak of the excitatory nerve terminal action otential (ENTP) to the onset of the synaptic current. or direct depolarization, dela was measured from the beginning of the stimulus arti act. PE (4P-phorbol-12,13-dibutyrate,Sigma) was dissolved in DMSO. In all experiments, 1pM PE was used. Maximal effect was obtained at this concentration. Ionomycin (Calbiochem) was also dissolved in DMSO. To test the effect of an inactive phorbol, 4-or-phorbol (Sigma)was used. RESULTS PE increased the amplitude of the excitatory postsynaptic potential (EPSP). The time course of the PE effect on the EPSP amplitude is shown in Figure 1. Shortly after exposure to the PE, a gradual increase in EPSP amplitude was observed and reached a plateau, in this experiment in about 2 minutes. The maximal time to the onset of the response was 3 minutes. The plateau
P
P
I
level could be maintained for a long period (more than 20 minutes). Table I de icts a si ificant increase ( P < 0.005) in EPSP am itude in a 1 16 preparations tested. The increase in E SP amplitude ranges between 1.6 and 4 times. DMSO, a t the concentration used (0.01%), did not have an effect on EPSP amplitude. 4-a-Phorbo1, an inactive analog of PE, had no effect a t the same or even at higher concentrations (1 and 2 pM). This result suggests that the effect of PE was specific.
Y
Jf
Post-Synaptic effects of PE Post-synaptic membrane properties The input resistance of the post-synaptic muscle fiber was measured by applying constant current pulses via one intracellular electrode while voltage deflection was measured by a second intracellular electrode. PE had no effect on the input resistance (Fig. 2A), while showing a fourfold increase in EPSP amplitude. To test the effects of PE on the time constant of the muscle membrane, we examined the slowest part of the EPSP decay. This method is especially useful in crustacean muscle. The multiterminal pattern of innervation leads to isopotential voltage changes along the muscle fiber (Atwood, 1967). Figure 2B presents semilogarithmic plots of the decay phase of the EPSP in control and with PE. As can be seen, both time constants are similar, indicating the lack of an effect of PE on membrane capacitance and resistance. Post-synaptic current properties Although PE had no effect on the muscle membrane resistance, the increase in amplitude of the EPSP could have resulted from PE interactions with the post-synaptic receptors. This does not seem to be the case as the
I
I
1
1 4
7.2
> E
v
Q)
U
I .-
........
5.c
..
E" U
I
CL
PE
(I)
g
2.5
-
I
I
I
i
2
3
Time Fig. I. Phorbol-ester increases EPSP amplitude, as recorded with an intracellular electrode from the deep abdominal extensor muscle of the prawn. The nerve was stimulated at 1Hz. At the arrow, PE was applied to the bath to achieve a final concentration of 1 p M . An increase in
(min
f
EPSP amplitude started shortly after application, and a plateau was attained after 1.5 min. Arrowhead indicates a step-like increase in EPSP amplitude, due to possible recruitment of additional nerve terminals.
PHORBOL ESTER AND TRANSMITTER RELEASE
93
TABLE I. Summary of PE effects on synaptic transmission and its effect on twin pulse facilitation'
PE Muscle No.
External Ca2+(mM)
augmentation of first EPSP (relative increaseI2
0.5 0.75 1 1 1 2 2 3
1
2 3 4
5 6 7 8 9 10
3 3 6 6 6 6
11 12 13 14 15 16
6 6
Twin pulse facilitation Control PE
2.66 2.85 1.64 3.33 4.00 2.53 2.00 2.11 2.47 1.69 1.95 2.59 2.06 1.78 1.87 1.81
1.12 1.21 1.00 1.33 1.20 1.33 1.33 1.12 1.19 1.17 0.92 0.96 1.06
1.15 1.25 0.98 1.17 1.23 1.28 1.33 1.02 1.15 1.12 0.97 0.92 1.03
0.95 1.10
0.93 1.04
;This table contains data from experiments conducted with the intracellular recordings Mean -t SD = 2.33 i- 0.65.
amplitude of unitary responses is not altered. Figure 3A depicts unitary responses before and after PE treatment. Workin at conditions of low quantal content enables clear etection of sin le uanta. Thirty unitary responses were measured in ivi ually and their peaks were aligned together to produce an average unitary response. Figure 3B shows the results of such a procedure, before and after PE treatment. As can be seen, no significant alteration in the average unitary response was observed. Paired t-test indicated no difference between the average unitary currents (n = 12, P > 0.1) (Fig. 3B). The decay time constant of the average unitary current was calculated from a semi-logarithmic plot. The difference in time constants was found to be small, 0.61 msec before and 0.69 msec after PE (paired t-test n = 5 , P > 0.5, Fig. 3 0 . Similar results were observed in two additional experiments. This further supports the conclusion that the increase in EPSP amplitude is not brought about by affecting a postsynaptic mechanism. Presynaptic effect of PE The lack of ost-synaptic effects of PE suggests that the effect of P is presynaptic. Quanta1 events can be discerned with the macropatch recording technique, as seen in Figure 3A. In six experiments, the increase in quantal content due to PE treatment increased by a 0.76 (mean r+ SD). This multiplication factor of 2.03 I result is similar to the increase in EPSP amplitude observed with the intracellular recording method 0.65, Table I). (2.33 I The presynaptic effects of PE could be the result of an increase in intracellular Ca2+ concentration available for release. Alternatively, PE could exert its effect a t sta es of the release process occurring after the entry of Ca'+. Direct methods to record changes in Ca2+concentration a t the small presynaptic terminal of the prawn are not available. We therefore had to resort to indirect experiments. The effect of PE on the time course of release Katz and Miledi (1965) used synaptic delay histograms as a measure for the time course of release. Delay
B
E
f 3
histograms were found to be independent of experimental conditions that change Ca2+entry or intracellular Ca2+concentration (Andreu and Barrett, 1980; Barrett and Stevens, 1972; Datyner and Gage, 1980; Matzner et al., 1988; Miller et al., 1985; Parnas et al., 1986a) but were affected when the slowest stage in the release process was altered (Matzner et al., 1988). The inset of Figure 4 shows an example of single quantum events recorded in response to twin pulse stimulation at 15°C using the direct stimulation techni ue in the presence of TTX (2 x M). Synaptic de ay histograms were constructed for the twin pulses. In this specific experiment, m, (quantal content of the first response) was 0.089 and m2was 0.145. Facilitation thus was 1.62. The delay histograms of the first and second pulses were the same. This enables pooling of the delays of the two pulses into one histogram. Figure 4 presents delay histograms of 460 responses for the control, and 928 responses for PE. Two thousand sweeps were analyzed under both control and PE conditions. No change in the three indicated parameters of the histogram (minimal dela , time to peak, and time course of decay) was observe (two tailed paired t-test, P > 0.5). Similar results were obtained in four experiments, including one experiment in which release was evoked by nerve stimulation. These experiments indicate that PE does not modify the kinetics of release, and hence it probably does not affect the rate-limiting step in the chain of events leading to release. PE does not affect the cooperativity of the release process Dodge and Rahamimoff (1967)showed that four Ca2+ ions are involved in the release of one quantum. An increase in release could have resulted from a decrease in the cooperativity (Parnas and Segel, 1982). It was therefore important to study whether the cooperativity is altered by PE. In a recent paper (Parnas et al., 1986b) it was shown that the lo /log plot of the rising phase of the delay histogram re lects the cooperativity of the release process. As was shown by Parnas et al. (1986b1, the slope ranges from 4 t o 8, depending on the strength of depolarization. We found the slope to be 6.25 ? 0.82
?
d
B
E. GILAT AND B. HOCHNER
94
A
PE
Control
techniques are required. Twin pulse facilitation was found to reflect residual calcium (Katz and Miledi, 1968). The magnitude and duration of facilitation were found to be useful tools to estimate entry and removal of Ca2+ (Parnas et al., 1982; Parnas and Segel, 1982, 1989). Facilitation expressed as the ratio between the averaged response of the second stimulus to that of the first was found to be rather low in this synapse. Table I shows that PE generally induced only very small changes in the de ree of short-term facilitation. PE reduced twin pulse acilitation from 1.13 t 0.14 (n = 15) to 1.10 i_ 0.13 (n = 15).This reduction was found not to be significant ( P < 0.1, paired t-test). However, minor but significant reduction (P < 0.025, paired t-test) in facilitation, from 1.11i 0.15 to 1.08 I 1.40 (n = lo), was observed in experiments conducted in Ca2+ concentration ranging from 2 to 6 mM. This result hints at the possibility that PE exerts part of its effect by elevating the entry of Ca”. However, this effect of PE although statistically significant, indicates that if Ca2’ entry is increased by PE, then the increase is small and cannot by itself be sufficient to account for the enhancement of release seen after PE application. Sup ort for this statement is given by the lack of effect o PE on the duration of facilitation (Fig. 5). One would ex ect a prolongation of duration of facilitation when Cap+ entry is increased owing to an increase in the time required for Ca2+ removal.
B
2mV
40ms
P
B > E
o - Conirol e - PE
1.6-
v
-
i
0 .-
z
a,
0.8-
PE alters excitability of the nerve terminal
c
Several observations support the ossibility that PE produced presynaptic excitability c anges. Repetitive c EPSPs elicited by one stimulus are often seen after PE P treatment. This phenomenon appears more frequently a in low Ca2+concentrations. The source of this activity is E most likely a repetitive train of presynaptic spikes. If ? O.I6this phenomenon is due to an excitability chan e, the T = 19.80rns increase in EPSP amplitude may be produced 7ly the recruitment of inactive terminals. However, this will not explain the increase in quantal content seen with direct terminal depolarization in the presence of TTX. 0.04 0 20 40 60 Nevertheless, in a few cases we observed a step-like change in the EPSPs shape that can be interpreted as time (msec) an activation by an action potential of a terminal that Fig. 2. PE does not change passive post-synaptic membrane proper- was not previously invaded (Fig. 1,arrowhead). To test the possibility that PE indeed increases the ties. A. Input resistance of the muscle is not affected by PE. No change in the voltage deflections generated by current pulses was observed excitability of the nerve terminal, it is possible to depoafter PE was introduced. Compare A1 and A2 in the control to A3 and larize the nerve ending directly and measure the threshA4 in the presence of PE. Note the fourfold increase in EPSP due to PE treatment. Compare A l , A2 in control conditions to A3, A4 in PE. B: PE old stimulus (Dudel et al., 1984). The relationship bedoes not change the membrane time constant as evaluated from the tween stimulus current and quantal content was semi-logarithmic plot of the decay of the EPSP. Control decay was measured (Fig. 6). The all or none behavior of this taken from EPSP shown in A l , and PE decay was taken from EPSP synapse indicated that this terminal is excitable (Dudel shown in A3. The curves were fitted using linear regression. et al., 1984).Apulse of -0.6 pAunder control conditions did not produce release (Fig. 5A), whereas upon the addition of PE, the same pulse produced release every in control conditions and 6.5 -C 0.99 in PE (n = 4, time (Fig. 5B). Note that the extrapolated stimulus P > 0.05, two tailed paired t-test). The absence of a current for 50% successes moved from -0.63 pA in significant effect of PE on the slope indicates that PE control conditions to -0.50 pA in the presence of PE does not influence the cooperativity of the release pro- (Fig. 5C). cess. The excitability changes induced by PE indicate that PE modifies presynaptic membrane properties. It is The effect of PE on the entry of Ca2+ possible that these changes lead to an increase in synIn the absence of a direct method for measuring entry aptic efficacy by increasing s ike amplitude and duraof Ca2+and its removal in the nerve terminal, indirect tion and by improvement o spike invasion into the 0
Q
a,
K
0.4 -
u
P
PHORBOL ESTER AND TRANSMITTER RELEASE
95
PE
Control
5rns
-toor
.----
I
I
1
0
a a
I
I
1
I
I
1
1
8
8
u t
c a L L
3
0 0 0
0
-701
0
L
I
I
1
1
1
I
I
0 0.40.8 1.2 1.6 2.0 2.4 2.8
I
-2
I
I
1
I
I
1
I
0 0.40.8 1.2 1.6 2.0 2.4 2.8
Time from peak (ms)
Time from peak (ms) Fig. 3. PE does not change unitary s aptic current properties. A Unitary synaptic currents as recorderwith the direct stimulation technique (only successes were chosen). B: 30 unitary responses of the control responses and “PE responses” were aligned according to their peaks. No significant difference in the amplitude and decay of the two
groups is apparent. The extrapolated broken lines starting from time zero represent the average rise times. C: Semi-logarithmic plot of the decays of control and PE traces taken from B. The average time constant was 0.61 msec before PE and 0.69 msec after PE.
20 8 u
-? c 10 a,
>
w
0
I
2
3
4
5
Delay ( r n s e c ) Fig. 4. PE does not modify the time course of phasic release. Delay histograms of release for direct depolarizing pulses (see inset), before and after PE, are superimposed. Delay was measured from the onset of stimulation (arrows)to the onset of the responses (arrowheads). 2,000
swee s were taken before and after PE was applied. The results from the 8rst and second pulses were pooled together (see text). 460 responses under control conditions and 928 responses in the presence of PE. Temperature, 15°C.
E. GILAT AND B. HOCHNER
96 1.30
s
I
-
1.25 1.20.c c 0 1.15.= 1.10V 0 1.05LL
I
I
I
-
0 0
0
0
0
-
a
8
0
-’
8-
-____--___----
1.00 0
I50
I00
50
200
Interval ( m s e c ) Fig. 5. PE has no significant effect on the duration offacilitation. In three experiments in which the Ca2 ’ concentration was 3 mM, facilitation was measured, employing different intervals between the twin pulses before (closed circles) and after (open circles) PE was applied.
(A)
Control
I\
.
0.6yA .___
I00
--
-8
U
80
u
60 m Q)
cn
0 0
40
3
cT>
20
0
r 1200pA
Stimulus (PA )
lOms Fig. 6. PE reduces the threshold for Stimulation for evoked release using the macropatch technique. A: No evoked release was observed under control condition upon a p lication ofa -0.6 PApulse. B: The same current in the presence ofPE results in evoked release. C: The rate ofsuccesses versus stimulus amplitu!e under control conditions and in the presence of PE is plotted. PE shifted the graph to the left, indicating a decrease in the threshold for stimulation.
terminal. 3,4 Diaminopyridine (3,4 DAP) (100 FM) produced a marked increase in transmitter release, but did not occlude the PE effect. After pre-treatment with 3,4 DAP, PE still increased release twofold (1.99 i 0.32, n = 4).This augmentation is similar to the one found
without 3,4 DAP treatment (2.33 2 0.65, Table I). This finding supports the notion that the major part of the increase in release produced by PE is by mechanisms not associated with the invasion of the action potential into inactive terminals.
PHORBOL ESTER AND TRANSMITTER RELEASE
The interaction of PE with Ca2+-dependent processes of release PE activates the enzyme C-kinase, an enzyme which is a Ca2+-,diacylglycerol-,and phospholipid-dependent protein kinase. The C-kinase can be activated syner 'stically by both Ca2+and PE, since PE mimics diacylg ycerol action (May et al., 1985; Nishizuka, 1986; Wolf et al., 1985). Hence, it should be possible to occlude this effect, a t least artially, by an increase of the intracellular level of 8a2+. A Ca2+ ion0 hore can serve as a means to increase internal ca' concentration. Figure 7A shows the effect of 0.3 pM ionomycin on the EPSP, which increased by a factor of 4. Addition of PE led to a further small increase in EPSP amplitude (10%). In other experiments, in which the ionomycin effect was more prominent, PE caused depression (Fi 7C). The average increase in EPSP amplitude caused y ionomytin was 4.8 ? 1.8-fold(n = 5).With the addition of 1pM PE, the EPSP remains virtually the same. To rule out the possibility that the lack of a PE effect is due to saturation of release induced by ionomycin, we show that synaptic release, after ionomycin and PE, can still be increased by an increase in external Ca2+concentration (Fig. 7A4J. The occlusion of the PE effect b ionomycin suggests that the increase in the intrace lular Ca2+ Drobablv activated PKC. To resolve whether ionomycin ixerts i& effect solely via C-kinase as the mediator between intra-
P
%
iv
97
cellular Ca2+ concentration and modulation of release, the application of ionomycin and PE was reversed (i.e., PE was applied before ionomycin). Figure 7B shows that PE, when applied alone, induced a marked increase in release, but now addition of ionomycin further increased release, indicating that part of the activity of the ionomycin is not mediated by PKC. Another way to increase intracellular Ca2+ concentration is by repetitive stimulation of the presynaptic terminal. Figure 8A,B shows the effect of stimulation frequency on EPSP amplitude. PE doubled the amplitude of the EPSP at 0.4 Hz, and, unlike control conditions in which frequent stimulation increased the level of release, frequency of stimulation does not affect the level of release in the presence of PE. The effect of PE on frequency modulation of release in five synapses is summarized in Figure 8C.
DISCUSSION PE enhances release of neurotransmitter in the DEAM of the prawn. As PE activates PKC it is reasonable to assume that this increase in release is mediated by PKC activation. The specificity of PE as a PKC activator is supported, in this study, by the lack of effect of the inactive phorbol analog 4-a-phorbol. The effects of PE were limited to the presynaptic terminal. There was no change in the input resistance of the post-synaptic cell, and the amplitude and shape of
8) PE I
Ionomycin I
A 40ms
1
1
I 2
I 3
I 4
I
I
5
6
I
Relative increase in EPSP ( Ionomycin Fig. 7. Ionornycin occludes the effect of PE. A: Ionornycin increases the EPSP amplitude and occludes the effect of PE: 1 )Control. 2) After ionomycin (0.3 pM),a n increase in EPSP amplitude is apparent. 3) PE has a minor effect (10%)when added after pretreatment with ionornycin. 4)In the presence ofPE and ionornycin, addition of Ca2' leads to a further increase in EPSP amplitude. Calibration pulse = 1rnV, 2 msec. B: PE does not occlude the whole effect of ionomycin. PE increases the
EPSP, and a further addition of ionornycin (1 FMj leads to a n additional increase in EPSP a m litude. C: The ordinate represents the additional PE increase of the EKSP, following ionomycin. The abscissa presents the relative increase in EPSP amplitude due to ionornycin. The effect of PE following ionornycin is reduced when the ionomycin effect is increased.
E. GILAT AND B. HOCHNER
98
I .5 U W
In 0
2 1.4
Q)
PE
.-
Q.
In
Q.
Q)
u 1.3 Q)
V
3
i
“ I
0
2
I
Frequency
( Hz )
3
0
c .-
u” 1.2 C Q)
3
0-
2
LL
I.
Control
PE
Fig. 8. PE modifies the dependence of transmission efficacy on the frequency of transmission. A EPSP amplitude is increased when the frequency of stimulation is increased (as indicated by the numbers at the bottom of each record). PE increases the EPSP and reduces the effect of frequency of stimulation. Averages were taken when the EPSP amplitude attained a steady state. B: Graphic demonstration of the records presented in A. The first EPSP in each sweep was measured
and plotted versus frequency of stimulation. To avoid possible muscle contraction, the frequency of stimulation, in the case ofthe control, was not increased to more than 2 Hz. C: Summary ofthe frequency effect on EPSP amplitude in control (n = 4)and in the presence of PE (n = 5). The effect of stimulation frequency is expressed as the relative increase in EPSP due to a change in frequency from 0.5 Hz to 2 Hz. Standard error bars are presented.
unitary post synaptic currents remained the same after PE treatment. Haimann et al. (1987) and Shapira et al. (1987) also reported a presynaptic effect of phorbol ester. A lack of an effect of phorbol ester on the postsynaptic membrane of the frog neuromuscular junction was reported by Shapira et al. (1987) while Haimann et al. (1987) suggested some post-synaptic effect. An increase in quanta1 content can be induced by an elevation in the resting level of Ca2+or an increase in Ca2+entry. The latter can result from direct modulation of presynaptic Ca2+channels or from an indirect mechanism such as enhanced invasion of the action potential into the presynaptic terminals, and broadening of the action potential (Hochner et al., 1986b). Our finding that PE alters excitability indicates that such a modulation is possible. However, the presence of a similar augmentation of the EPSP by PE alone, and by PE following augmentation of the EPSP by 3,4 DAP, indicates that enhanced invasion is unlikely to be the main mechanism by which PE exerts its effect. Moreover, in loose patch clamp experiments in the presence of TTX,
which blocks sodium excitability, PE still enhances release. This excludes effects of PE on presynaptic conductances other than Ca2+(Hochner et al., 1986b). To test whether Ca2+ entry is increased by PE we measured twin pulse facilitation. Katz and Miledi (1968) provided evidence for the hyeothesis that facilitation depends upon residual Ca” left from the first impulse. Thus facilitation (amplitude and duration) can be used as an indicator for Ca2+ entry. We observed a minor decrease in facilitation a t Ca2 concentration between 2 and 6 mM. This slight decrease suggests that PE increases Ca2+ entry. Such an increase, if present, must be very small as it did not prolong the duration of facilitation (Parnas and Segel, 1989). A further action of PE can involve processes that are secondary to the increase in Ca2+concentration. PE did not affect the kinetics of release as evaluated from unitary release delay histograms that express the robability of release with time after each change stimulus ( atz and Miledi, 1965). We can conclude that
PHORBOL ESTER AND TRANSMITTER RELEASE
PE does not affect the rate-limitingstep in the release, a step suggested to be the vesicle interaction and fusion with the releasing site (Parnas et al., 1986a). It seems that PKC is not directly involved in the vesicle interaction with the release site. PE could reduce cooperativity of the release process and thereby enhance release. Cooperativity, as calculated from the loghog plot of the initial part of the delay histogram (Parnas et al., 1986b), was not altered by PE. Haimann et al. (1987) suggested that phorbol ester changes the Ca2+cooperativity of release in the neuromuscular junction of the frog. They based their suggestion on the effect of phorbol ester on the relationship between release and extracellular Ca2+ concentration. However, as stated by Van der Hoot (1988) and Parnas et al. (1986b1,an apparent decline in cooperativity may be obtained if intracellular Ca2+content increases. We suggest that PKC can partici ate in the modulation of release. It is well establis ed that PKC is a Ca2+-de endent enzyme (Nishizuka, 1986), so that processes t at lead to Ca2+ accumulation ma modulate release by activating PKC. Indeed, we o served an occlusion ofPE effect due to pretreatment with the Ca2+ ionojhore. Similarly, when loading the terminal with Ca2 b hi h-frequency stimulation, the effect of PE was re uce . This suggests a ossible physiological role of PKC in the activity-depen ent modulation of release that was established for crustacean neuromuscular junctions (Atwood and Wojtowitz, 1986). While we cannot arrive a t a clear conclusion regarding the mechanism of activity of PE, based on techniques currently available for the neuromuscular junction, we can exclude certain possibilities. PE does not reduce cooperativity of release, nor does it intervene with the rate-limiting step of release. PE has very little effect on Ca2+entry, and this effect cannot account for its major activity. The results favor a predominant effect of PE on certain stages in the release process that are secondary to the increase in internal Ca2+ concentration due to presynaptic depolarization. This effect can enhance the affinity of the release machinery to internal Ca2+ concentration andfor the availability of transmitter vesicles and the activity of releasing sites. These may involve phosphorylation of synapsin-like molecules which have been sug ested to modulate the availability of synaptic vesicles linas et al., 1985). ACKNOWLEDGMENTS This research was supported by DFG g a n t No. SFB220A6, Binational grant No. 2653-1-86, and the Goldie-Anna Charitable Trust Endowment Fund. We wish to thank Drs. Hanna Parnas and Itzchak Parnas for their continued support and advice, Drs. Y. Yarom and L. Segel for their careful reading and commentary, and Henry Matzner for fruitful discussions. We would also like to thank Frances Bogot for typing and editing this manuscript.
K
K
d v
B
&
REFERENCES Akers, R.F., Lovinger, D.M., Colley, P.A., Linden, D.J., and Routtenberg, A. (1986) Translocation of protein kinase C activity may mediate hippocampal long-term potentiation. Science, 231:587-589. Andreu, R., and Barett, E.F. (1980) Calcium dependence of evoked transmitter release at very low quantal contents a t the frog neuromuscular junction. J. Physiol. (London),308:79-97. Atwood, H.L. (1967) Crustacean neuromuscular mechanism. Am. ZOO^. , 7:527-551.
99
Atwood, H.L., and Wojtowicz, J.M. (1986) Short-term and long-term plasticity and physiolog~aldifferentiation of crustacean motor synapses. Int. Rev. Neurobiol., 28:275-362. Baraban, J.M., Snyder, S.H., and Alger, B.E. (1985) Protein kinase C regulates ionic conductance in hi pocampal pyramidal neurons Electrophysiological effects of phor!ol ester. Proc. Natl. Acad. Sci USA, 82:2438-2542. Barrett, E.F., and Stevens, C.F. (1972) The kinetics of transmitter release at the frog neuromuscular junction. J. Physiol. (London) 227:691-708. Datyner, N.B., and Gage, P.W. (1980) Phasic secretion ofacetylcholine at a mammalian neuromuscular junction. J. Physiol., 303:299-314. De Riemer, S.A., Stron , J A , Albert, K.A., Greengard, P., and Kaczmarek, L.K. (1985) inhancement of calcium current in Aplysia neurons by phorbol ester and protein kinase C. Nature, 313:313-316. Dodge, F.A., and Rahamimoff, R. (1967) Cooperative action of calcium ions in transmitter release at the neuromuscular junction. J. Physiol., 372:191-205. Dudel, J. (1981) The effect of reduced calcium on quantal unit current release at the crayfish neuromuscular junction. Pflugers Arch., 391:3540. Dudel, J., Parnas, I., Cohen, I., and Franke, C. (1984) Excitability and depolarization-Release characteristics of excitatory nerve terminals in a tail muscle of spiny lobster. Pflugers Arch., 402:293-296. Farley, J., and Aurbach, S. (1986) Reduction of K‘ currents and enhancement of Ca2 current in Hermissenda photoreceptors by phorbol ester and protein kinase C. Nature, 319:220-223. Haimann, C., Meldolesi, J.,and Ceccarelli,B. (1987)The phorbol ester, 12-0-tetradecanoyl-phorbol-13-acetate, enhances the evoked quanta release of acetylcholine at the frog neuromuscular junction. Pflugers Arch., 408:27-31. Hochner, B., Braha, O., Klein, M., and Kandel, E.R. (1986a) Distinct processes in presynaptic facilitation contribute to sensitization and dishabituation in aplysia: Possible involvement of C kinase in dishabituation. Neurosci. Abstr., 2:1340. Hochner, B., Klein, M., Schacher, S., and Kandel, E.R. (1986b)Actionpotential duration and modulation of transmitter release from the sensory neurons of Aplysia in presynaptic facilitation and behaviora! sensitization. Proc. Natl. Acad. Sci. USA, 8323410-8414, Katz, B., and Miledi, R. (1965)The measurement of synaptic delay and the time course of Ach release at the neuromuscular junction. Proc. R. SOC. B., 161:483495. Katz, B., and Miledi, R. (1968) The role of calcium in neuromuscular facilitation. J. Physiol. (London), 195:481492, Llinas, R., McGuiness, T.L., Leonard, C.S., Sugmori, M., and Greengard, P. (1985) Intraterminal injection of synapsin I or calciud calmodulin-dependent protein kinase I1 alters neurotransmitter release a t the squid giant synapse. Proc. Natl. Acad. Sci. USA, 8213035-3039. Malenka, R.C., Madison, D.V., Andrade, R., and Nicoll, R.A. (1986a) Phorbol esters mimic some cholinergic actions in hippocampal pyramidal neurons. J. Neurosci., 6:475480. Malenka, R.C., Madison, D.V., and Nicoll, R.A. (1986b) Potentiation of synaptic transmission in the hippocampus by phorbol esters. Nature, 3211175-177. Matzner, H., Parnas, H., and Parnas, I. (1988) Presynaptic effects of d-tubocurarine on neurotransmltter release at the neuromuscular junction of the frog. J. Physiol., 398:109-121. May, W.S. Jr., Sahyoun, N., Wolf, M., and Cuatrecasas, P. (1985)Role of intracellular calcium mobilization in the regulation of protein kinase C-mediated membrance processes. Nature, 317:549-551. Miller, M.W., Parnas, H., and Parnas, I. (1985)Dopaminer ‘c modulaPhyslol., tion of neuromuscular transmission in the prawn. 363~363-375. Nishizuka, Y. (1986) Studies and perspectives of protein kinase C. Science, 233:305-312. Parnas, H., Dudel, J., and Parnas, I. (1982) Neurotransmitter release and its facilitation in crayfish. I. Saturation kinetics of release and of entry and removal of Ca”. Pflugers Arch., 393:l-14. Parnas, H., Dudel, J., and Parnas, I. (1986a)Neurotransmitter release and its facilitation in crayfish. VII. Depolarization, no Ca inflow determines the time course of phasic release. Pflugers Arch., 406:121-130. Parnas, H., Parnas, I., and Segel, L.A. (1986b) A new method for determining cooperativity in neurotransmitter release. J. Theor. Biol., 119:481499. Parnas, H., and Segel, L.A. (1982) Ways to discern the presynaptic effect of drugs on neurotransmitter release. J. Theor. Biol., 94:923-941. Parnas, H., and Segel, L.A. (1989) Facilitation as a tool.to study the entry of calcium and the mechanism of neurotransmitter release. Prog. Neurobiol., 3 2 - 9 . +
?
100
E. GILAT AND B. HOCHNER
Pozzan, T., Gatti, G., Dozio, N., Vicentini, L.M., andMeldolesi, J . (1984) Ca2 ' dependent and independent release of neurotransmitters from PC 12 cells: A role for protein kinase C activation? J. Cell Biol., 99:628-638. Rink, T.J., Sanchez, A., and Hallam, T.J. (1983) Diacylglycerol and phorbol ester stimulate secretion without raising cytoplasmic free calcium in human platelets. Nature, 305:317-319. Shapira, R., Silberberg, S.A., Ginsburg, S., and Rahamimoff, R. (1987) Activation of protein kinase C augments evoked transmitter release. Nature, 325:58-60.
Van der Kloot, W. (1988) The kinetics of uantal releases during endplate currents at the frog neuromuscujar junction. J. Physiol. (London), 402:605-626. Wolf, M., LeVine, H. 111, Stratford, M. Jr., Cuatrecasas, P., and Sahyoun, N. (1985)A model for intracellular translocation of protein kinase C involving synergism between Ca2 and phorbol esters. Nature, 317546-549. +
Phorbol Ester Enhances SvnaDtic Transmission at Crusticeai Neuromuscular Junctions E. GILAT AND B. HOCHNER The Otto Loewi Center for Cellular and Molecular Neurohiology and the Department of Neurohiology, The Hebrew University of Jerusalem, Israel
KEY WORDS
Phorbol ester, Synaptic release, Transmitter release, Prawn
ABSTRACT
Effects of phorbol ester (PE) (4P-phorbol-12,13-dibutyrate) on transmitter release were studied in the deep extensor neuromuscular system of the prawn, Macrobrachium rosenbergii. Our findings show that PE enhances transmitter release as indicated by an increase in the quanta1 content. PE had no post-synaptic effects. The increase in release is accompanied by a slight decline in twin pulse facilitation, suggesting a minor increase in Ca2+entry. The fact that the increase in Ca2+entry has a minor contribution to the PE effect is supported by the following observations: the duration of facilitation was not affected by PE, and 3,4-diaminopyridine (3,4-DAP),which by itself increased release, did not reduce the effect of PE. The time course of release was measured from synaptic delay histograms, upon which PE had no effect. This finding indicates that protein kinase C (PKC) is probably not involved in the rate limiting step of the process of secretion. The log/log plot of the initial part of the delay histogram is not affected by PE, suggesting a lack of effect on cooperativity of the release process. Increased release by loading the presynaptic terminal with Ca2+either by pretreatment with Ca2+ ionophore or by frequent stimulation prevented further increase in release by PE. We concludethat the main effect of PE is confined to stages of release that are secondary to the first elevation in presynaptic Ca2+.PKC in this system probably plays a role in long term modulation of release. and it can be activated in processes leading to presynaptic Ca2+accumulation.
INTRODUCTION Recently, it was sug ested that activation of protein kinase C (PKC), a a2+-and phospholipid-dependent enzyme, enhances synaptic transmission in various preparations-in the frog neuromuscular ‘unction (Haimann et al., 1987; Sha ira et al., 1987j in the sensory-motor s apse of plysia (Hochner et al., 1986a), and in rat hippocam us (Malenka et al., 1986b). Phorbol ester (PE) was a so implicated in the mechanism of long-term potentiation (Akers et al., 1986). The mechanisms underlying this increase in synaptic release are still unknown. For example, activation of PKC can increase Ca2+ conductance by modulating Ca2+channel activity in Aplysia bag cells (De Riemer et al., 1985) or can reduce potassium conductance (Baraban et al., 1985; Farley and Auerbach, 1986; Malenka et al., 1986a).This may lead to the broadening of action potentials and thereby increase release (Hochner et al., 1986131. In non-s.ynaptic systems there is evidence that PKC is directly involved in the secretion processes. Diacylglycerol and phorbol ester were reported to stimulate secretion without raising cytoplasmic free calcium in human platelets (Rink et al., 1983). PKC also stimulates Ca-dependent and Ca-independent release of neurotransmitter from PC12 cells (Pozzan et al., 1984). In the present study, we tested for the effect of
8
tg
@ 1990 WILEY-LISS,INC.
1
P
4P-phorbol-12,13-dibutyrate(PE) on transmitter release using a crustacean neuromuscular system that enables the detection of single quanta and direct depolarization of the terminal (Dudel, 1981). The results suggest that PE may increase transmitter release by acting in more than one way. Nevertheless, on the basis of indirect techniques, we conclude that the major effect of PE involves modulation of stages in the release process that occur after the entry of Ca2+ions but that do not affect the time course of release.
MATERIALS AND METHODS The L1 and L2bundles of the dee extensor abdominal medialis (DEAM) of the prawn acrobrachium rosenbergii were used (Miller et al., 1985). Excitatory axons
2
were stimulated with a small suction electrode. For intracellular recordin s, 3 M-KC1 microelectrodes (10-15 Ma)were used. ynaptic currents were recorded using “macropatch” electrodes (10-20 bm inner diameter) (Dudel, 1981). The macropatch technique is based on a patch-clamp amplifier which enables direct stimulation and recordings, under loose patch conditions. Stimulation arameters can be controlled to obtain different leve s of release. Unitary events can be initi-
5
P
Received November 27,1989; accepted in revised form February 14,1990. Address reprint requests to B. Hochner, The Otto Loewi Center for Cellular and Molecular Neurobiology, Institute of Life Science, The Hebrew University of Jerusalem, Israel
E. GILAT AND B. HOCHNER
92
ated and observed under stimulation conditions that lead to a low level of release. The macropatch electrodes were filled with a solution composed of imM) NaCl, 220; KCl, 5.4; CaCl,, 6 (unless otherwise indicated): MgC12, 2.5; Tris maleate, 10, which was also used as the bathing medium. The pH was adjusted t o 7.4 with NaOH. A constant flow (approximately one chamber volume, 3 ml per minute) was maintained. Bath temperature was controlled by a heat exchanger with an accuracy of k1"C. Recordings were averaged on a Nicolet (model 1174) signal averager. Traces were digitized (20 psecladdress) and stored on a video-tape (Neurocorder, model DR484). Quanta1 units, current amplitude, and delays were measured using home written software on an Olivetti 24 SP computer. In the case of nerve stimulation, the synaptic delay was measured from the negative peak of the excitatory nerve terminal action otential (ENTP) to the onset of the synaptic current. or direct depolarization, dela was measured from the beginning of the stimulus arti act. PE (4P-phorbol-12,13-dibutyrate,Sigma) was dissolved in DMSO. In all experiments, 1pM PE was used. Maximal effect was obtained at this concentration. Ionomycin (Calbiochem) was also dissolved in DMSO. To test the effect of an inactive phorbol, 4-or-phorbol (Sigma)was used. RESULTS PE increased the amplitude of the excitatory postsynaptic potential (EPSP). The time course of the PE effect on the EPSP amplitude is shown in Figure 1. Shortly after exposure to the PE, a gradual increase in EPSP amplitude was observed and reached a plateau, in this experiment in about 2 minutes. The maximal time to the onset of the response was 3 minutes. The plateau
P
P
I
level could be maintained for a long period (more than 20 minutes). Table I de icts a si ificant increase ( P < 0.005) in EPSP am itude in a 1 16 preparations tested. The increase in E SP amplitude ranges between 1.6 and 4 times. DMSO, a t the concentration used (0.01%), did not have an effect on EPSP amplitude. 4-a-Phorbo1, an inactive analog of PE, had no effect a t the same or even at higher concentrations (1 and 2 pM). This result suggests that the effect of PE was specific.
Y
Jf
Post-Synaptic effects of PE Post-synaptic membrane properties The input resistance of the post-synaptic muscle fiber was measured by applying constant current pulses via one intracellular electrode while voltage deflection was measured by a second intracellular electrode. PE had no effect on the input resistance (Fig. 2A), while showing a fourfold increase in EPSP amplitude. To test the effects of PE on the time constant of the muscle membrane, we examined the slowest part of the EPSP decay. This method is especially useful in crustacean muscle. The multiterminal pattern of innervation leads to isopotential voltage changes along the muscle fiber (Atwood, 1967). Figure 2B presents semilogarithmic plots of the decay phase of the EPSP in control and with PE. As can be seen, both time constants are similar, indicating the lack of an effect of PE on membrane capacitance and resistance. Post-synaptic current properties Although PE had no effect on the muscle membrane resistance, the increase in amplitude of the EPSP could have resulted from PE interactions with the post-synaptic receptors. This does not seem to be the case as the
I
I
1
1 4
7.2
> E
v
Q)
U
I .-
........
5.c
..
E" U
I
CL
PE
(I)
g
2.5
-
I
I
I
i
2
3
Time Fig. I. Phorbol-ester increases EPSP amplitude, as recorded with an intracellular electrode from the deep abdominal extensor muscle of the prawn. The nerve was stimulated at 1Hz. At the arrow, PE was applied to the bath to achieve a final concentration of 1 p M . An increase in
(min
f
EPSP amplitude started shortly after application, and a plateau was attained after 1.5 min. Arrowhead indicates a step-like increase in EPSP amplitude, due to possible recruitment of additional nerve terminals.
PHORBOL ESTER AND TRANSMITTER RELEASE
93
TABLE I. Summary of PE effects on synaptic transmission and its effect on twin pulse facilitation'
PE Muscle No.
External Ca2+(mM)
augmentation of first EPSP (relative increaseI2
0.5 0.75 1 1 1 2 2 3
1
2 3 4
5 6 7 8 9 10
3 3 6 6 6 6
11 12 13 14 15 16
6 6
Twin pulse facilitation Control PE
2.66 2.85 1.64 3.33 4.00 2.53 2.00 2.11 2.47 1.69 1.95 2.59 2.06 1.78 1.87 1.81
1.12 1.21 1.00 1.33 1.20 1.33 1.33 1.12 1.19 1.17 0.92 0.96 1.06
1.15 1.25 0.98 1.17 1.23 1.28 1.33 1.02 1.15 1.12 0.97 0.92 1.03
0.95 1.10
0.93 1.04
;This table contains data from experiments conducted with the intracellular recordings Mean -t SD = 2.33 i- 0.65.
amplitude of unitary responses is not altered. Figure 3A depicts unitary responses before and after PE treatment. Workin at conditions of low quantal content enables clear etection of sin le uanta. Thirty unitary responses were measured in ivi ually and their peaks were aligned together to produce an average unitary response. Figure 3B shows the results of such a procedure, before and after PE treatment. As can be seen, no significant alteration in the average unitary response was observed. Paired t-test indicated no difference between the average unitary currents (n = 12, P > 0.1) (Fig. 3B). The decay time constant of the average unitary current was calculated from a semi-logarithmic plot. The difference in time constants was found to be small, 0.61 msec before and 0.69 msec after PE (paired t-test n = 5 , P > 0.5, Fig. 3 0 . Similar results were observed in two additional experiments. This further supports the conclusion that the increase in EPSP amplitude is not brought about by affecting a postsynaptic mechanism. Presynaptic effect of PE The lack of ost-synaptic effects of PE suggests that the effect of P is presynaptic. Quanta1 events can be discerned with the macropatch recording technique, as seen in Figure 3A. In six experiments, the increase in quantal content due to PE treatment increased by a 0.76 (mean r+ SD). This multiplication factor of 2.03 I result is similar to the increase in EPSP amplitude observed with the intracellular recording method 0.65, Table I). (2.33 I The presynaptic effects of PE could be the result of an increase in intracellular Ca2+ concentration available for release. Alternatively, PE could exert its effect a t sta es of the release process occurring after the entry of Ca'+. Direct methods to record changes in Ca2+concentration a t the small presynaptic terminal of the prawn are not available. We therefore had to resort to indirect experiments. The effect of PE on the time course of release Katz and Miledi (1965) used synaptic delay histograms as a measure for the time course of release. Delay
B
E
f 3
histograms were found to be independent of experimental conditions that change Ca2+entry or intracellular Ca2+concentration (Andreu and Barrett, 1980; Barrett and Stevens, 1972; Datyner and Gage, 1980; Matzner et al., 1988; Miller et al., 1985; Parnas et al., 1986a) but were affected when the slowest stage in the release process was altered (Matzner et al., 1988). The inset of Figure 4 shows an example of single quantum events recorded in response to twin pulse stimulation at 15°C using the direct stimulation techni ue in the presence of TTX (2 x M). Synaptic de ay histograms were constructed for the twin pulses. In this specific experiment, m, (quantal content of the first response) was 0.089 and m2was 0.145. Facilitation thus was 1.62. The delay histograms of the first and second pulses were the same. This enables pooling of the delays of the two pulses into one histogram. Figure 4 presents delay histograms of 460 responses for the control, and 928 responses for PE. Two thousand sweeps were analyzed under both control and PE conditions. No change in the three indicated parameters of the histogram (minimal dela , time to peak, and time course of decay) was observe (two tailed paired t-test, P > 0.5). Similar results were obtained in four experiments, including one experiment in which release was evoked by nerve stimulation. These experiments indicate that PE does not modify the kinetics of release, and hence it probably does not affect the rate-limiting step in the chain of events leading to release. PE does not affect the cooperativity of the release process Dodge and Rahamimoff (1967)showed that four Ca2+ ions are involved in the release of one quantum. An increase in release could have resulted from a decrease in the cooperativity (Parnas and Segel, 1982). It was therefore important to study whether the cooperativity is altered by PE. In a recent paper (Parnas et al., 1986b) it was shown that the lo /log plot of the rising phase of the delay histogram re lects the cooperativity of the release process. As was shown by Parnas et al. (1986b1, the slope ranges from 4 t o 8, depending on the strength of depolarization. We found the slope to be 6.25 ? 0.82
?
d
B
E. GILAT AND B. HOCHNER
94
A
PE
Control
techniques are required. Twin pulse facilitation was found to reflect residual calcium (Katz and Miledi, 1968). The magnitude and duration of facilitation were found to be useful tools to estimate entry and removal of Ca2+ (Parnas et al., 1982; Parnas and Segel, 1982, 1989). Facilitation expressed as the ratio between the averaged response of the second stimulus to that of the first was found to be rather low in this synapse. Table I shows that PE generally induced only very small changes in the de ree of short-term facilitation. PE reduced twin pulse acilitation from 1.13 t 0.14 (n = 15) to 1.10 i_ 0.13 (n = 15).This reduction was found not to be significant ( P < 0.1, paired t-test). However, minor but significant reduction (P < 0.025, paired t-test) in facilitation, from 1.11i 0.15 to 1.08 I 1.40 (n = lo), was observed in experiments conducted in Ca2+ concentration ranging from 2 to 6 mM. This result hints at the possibility that PE exerts part of its effect by elevating the entry of Ca”. However, this effect of PE although statistically significant, indicates that if Ca2’ entry is increased by PE, then the increase is small and cannot by itself be sufficient to account for the enhancement of release seen after PE application. Sup ort for this statement is given by the lack of effect o PE on the duration of facilitation (Fig. 5). One would ex ect a prolongation of duration of facilitation when Cap+ entry is increased owing to an increase in the time required for Ca2+ removal.
B
2mV
40ms
P
B > E
o - Conirol e - PE
1.6-
v
-
i
0 .-
z
a,
0.8-
PE alters excitability of the nerve terminal
c
Several observations support the ossibility that PE produced presynaptic excitability c anges. Repetitive c EPSPs elicited by one stimulus are often seen after PE P treatment. This phenomenon appears more frequently a in low Ca2+concentrations. The source of this activity is E most likely a repetitive train of presynaptic spikes. If ? O.I6this phenomenon is due to an excitability chan e, the T = 19.80rns increase in EPSP amplitude may be produced 7ly the recruitment of inactive terminals. However, this will not explain the increase in quantal content seen with direct terminal depolarization in the presence of TTX. 0.04 0 20 40 60 Nevertheless, in a few cases we observed a step-like change in the EPSPs shape that can be interpreted as time (msec) an activation by an action potential of a terminal that Fig. 2. PE does not change passive post-synaptic membrane proper- was not previously invaded (Fig. 1,arrowhead). To test the possibility that PE indeed increases the ties. A. Input resistance of the muscle is not affected by PE. No change in the voltage deflections generated by current pulses was observed excitability of the nerve terminal, it is possible to depoafter PE was introduced. Compare A1 and A2 in the control to A3 and larize the nerve ending directly and measure the threshA4 in the presence of PE. Note the fourfold increase in EPSP due to PE treatment. Compare A l , A2 in control conditions to A3, A4 in PE. B: PE old stimulus (Dudel et al., 1984). The relationship bedoes not change the membrane time constant as evaluated from the tween stimulus current and quantal content was semi-logarithmic plot of the decay of the EPSP. Control decay was measured (Fig. 6). The all or none behavior of this taken from EPSP shown in A l , and PE decay was taken from EPSP synapse indicated that this terminal is excitable (Dudel shown in A3. The curves were fitted using linear regression. et al., 1984).Apulse of -0.6 pAunder control conditions did not produce release (Fig. 5A), whereas upon the addition of PE, the same pulse produced release every in control conditions and 6.5 -C 0.99 in PE (n = 4, time (Fig. 5B). Note that the extrapolated stimulus P > 0.05, two tailed paired t-test). The absence of a current for 50% successes moved from -0.63 pA in significant effect of PE on the slope indicates that PE control conditions to -0.50 pA in the presence of PE does not influence the cooperativity of the release pro- (Fig. 5C). cess. The excitability changes induced by PE indicate that PE modifies presynaptic membrane properties. It is The effect of PE on the entry of Ca2+ possible that these changes lead to an increase in synIn the absence of a direct method for measuring entry aptic efficacy by increasing s ike amplitude and duraof Ca2+and its removal in the nerve terminal, indirect tion and by improvement o spike invasion into the 0
Q
a,
K
0.4 -
u
P
PHORBOL ESTER AND TRANSMITTER RELEASE
95
PE
Control
5rns
-toor
.----
I
I
1
0
a a
I
I
1
I
I
1
1
8
8
u t
c a L L
3
0 0 0
0
-701
0
L
I
I
1
1
1
I
I
0 0.40.8 1.2 1.6 2.0 2.4 2.8
I
-2
I
I
1
I
I
1
I
0 0.40.8 1.2 1.6 2.0 2.4 2.8
Time from peak (ms)
Time from peak (ms) Fig. 3. PE does not change unitary s aptic current properties. A Unitary synaptic currents as recorderwith the direct stimulation technique (only successes were chosen). B: 30 unitary responses of the control responses and “PE responses” were aligned according to their peaks. No significant difference in the amplitude and decay of the two
groups is apparent. The extrapolated broken lines starting from time zero represent the average rise times. C: Semi-logarithmic plot of the decays of control and PE traces taken from B. The average time constant was 0.61 msec before PE and 0.69 msec after PE.
20 8 u
-? c 10 a,
>
w
0
I
2
3
4
5
Delay ( r n s e c ) Fig. 4. PE does not modify the time course of phasic release. Delay histograms of release for direct depolarizing pulses (see inset), before and after PE, are superimposed. Delay was measured from the onset of stimulation (arrows)to the onset of the responses (arrowheads). 2,000
swee s were taken before and after PE was applied. The results from the 8rst and second pulses were pooled together (see text). 460 responses under control conditions and 928 responses in the presence of PE. Temperature, 15°C.
E. GILAT AND B. HOCHNER
96 1.30
s
I
-
1.25 1.20.c c 0 1.15.= 1.10V 0 1.05LL
I
I
I
-
0 0
0
0
0
-
a
8
0
-’
8-
-____--___----
1.00 0
I50
I00
50
200
Interval ( m s e c ) Fig. 5. PE has no significant effect on the duration offacilitation. In three experiments in which the Ca2 ’ concentration was 3 mM, facilitation was measured, employing different intervals between the twin pulses before (closed circles) and after (open circles) PE was applied.
(A)
Control
I\
.
0.6yA .___
I00
--
-8
U
80
u
60 m Q)
cn
0 0
40
3
cT>
20
0
r 1200pA
Stimulus (PA )
lOms Fig. 6. PE reduces the threshold for Stimulation for evoked release using the macropatch technique. A: No evoked release was observed under control condition upon a p lication ofa -0.6 PApulse. B: The same current in the presence ofPE results in evoked release. C: The rate ofsuccesses versus stimulus amplitu!e under control conditions and in the presence of PE is plotted. PE shifted the graph to the left, indicating a decrease in the threshold for stimulation.
terminal. 3,4 Diaminopyridine (3,4 DAP) (100 FM) produced a marked increase in transmitter release, but did not occlude the PE effect. After pre-treatment with 3,4 DAP, PE still increased release twofold (1.99 i 0.32, n = 4).This augmentation is similar to the one found
without 3,4 DAP treatment (2.33 2 0.65, Table I). This finding supports the notion that the major part of the increase in release produced by PE is by mechanisms not associated with the invasion of the action potential into inactive terminals.
PHORBOL ESTER AND TRANSMITTER RELEASE
The interaction of PE with Ca2+-dependent processes of release PE activates the enzyme C-kinase, an enzyme which is a Ca2+-,diacylglycerol-,and phospholipid-dependent protein kinase. The C-kinase can be activated syner 'stically by both Ca2+and PE, since PE mimics diacylg ycerol action (May et al., 1985; Nishizuka, 1986; Wolf et al., 1985). Hence, it should be possible to occlude this effect, a t least artially, by an increase of the intracellular level of 8a2+. A Ca2+ ion0 hore can serve as a means to increase internal ca' concentration. Figure 7A shows the effect of 0.3 pM ionomycin on the EPSP, which increased by a factor of 4. Addition of PE led to a further small increase in EPSP amplitude (10%). In other experiments, in which the ionomycin effect was more prominent, PE caused depression (Fi 7C). The average increase in EPSP amplitude caused y ionomytin was 4.8 ? 1.8-fold(n = 5).With the addition of 1pM PE, the EPSP remains virtually the same. To rule out the possibility that the lack of a PE effect is due to saturation of release induced by ionomycin, we show that synaptic release, after ionomycin and PE, can still be increased by an increase in external Ca2+concentration (Fig. 7A4J. The occlusion of the PE effect b ionomycin suggests that the increase in the intrace lular Ca2+ Drobablv activated PKC. To resolve whether ionomycin ixerts i& effect solely via C-kinase as the mediator between intra-
P
%
iv
97
cellular Ca2+ concentration and modulation of release, the application of ionomycin and PE was reversed (i.e., PE was applied before ionomycin). Figure 7B shows that PE, when applied alone, induced a marked increase in release, but now addition of ionomycin further increased release, indicating that part of the activity of the ionomycin is not mediated by PKC. Another way to increase intracellular Ca2+ concentration is by repetitive stimulation of the presynaptic terminal. Figure 8A,B shows the effect of stimulation frequency on EPSP amplitude. PE doubled the amplitude of the EPSP at 0.4 Hz, and, unlike control conditions in which frequent stimulation increased the level of release, frequency of stimulation does not affect the level of release in the presence of PE. The effect of PE on frequency modulation of release in five synapses is summarized in Figure 8C.
DISCUSSION PE enhances release of neurotransmitter in the DEAM of the prawn. As PE activates PKC it is reasonable to assume that this increase in release is mediated by PKC activation. The specificity of PE as a PKC activator is supported, in this study, by the lack of effect of the inactive phorbol analog 4-a-phorbol. The effects of PE were limited to the presynaptic terminal. There was no change in the input resistance of the post-synaptic cell, and the amplitude and shape of
8) PE I
Ionomycin I
A 40ms
1
1
I 2
I 3
I 4
I
I
5
6
I
Relative increase in EPSP ( Ionomycin Fig. 7. Ionornycin occludes the effect of PE. A: Ionornycin increases the EPSP amplitude and occludes the effect of PE: 1 )Control. 2) After ionomycin (0.3 pM),a n increase in EPSP amplitude is apparent. 3) PE has a minor effect (10%)when added after pretreatment with ionornycin. 4)In the presence ofPE and ionornycin, addition of Ca2' leads to a further increase in EPSP amplitude. Calibration pulse = 1rnV, 2 msec. B: PE does not occlude the whole effect of ionomycin. PE increases the
EPSP, and a further addition of ionornycin (1 FMj leads to a n additional increase in EPSP a m litude. C: The ordinate represents the additional PE increase of the EKSP, following ionomycin. The abscissa presents the relative increase in EPSP amplitude due to ionornycin. The effect of PE following ionornycin is reduced when the ionomycin effect is increased.
E. GILAT AND B. HOCHNER
98
I .5 U W
In 0
2 1.4
Q)
PE
.-
Q.
In
Q.
Q)
u 1.3 Q)
V
3
i
“ I
0
2
I
Frequency
( Hz )
3
0
c .-
u” 1.2 C Q)
3
0-
2
LL
I.
Control
PE
Fig. 8. PE modifies the dependence of transmission efficacy on the frequency of transmission. A EPSP amplitude is increased when the frequency of stimulation is increased (as indicated by the numbers at the bottom of each record). PE increases the EPSP and reduces the effect of frequency of stimulation. Averages were taken when the EPSP amplitude attained a steady state. B: Graphic demonstration of the records presented in A. The first EPSP in each sweep was measured
and plotted versus frequency of stimulation. To avoid possible muscle contraction, the frequency of stimulation, in the case ofthe control, was not increased to more than 2 Hz. C: Summary ofthe frequency effect on EPSP amplitude in control (n = 4)and in the presence of PE (n = 5). The effect of stimulation frequency is expressed as the relative increase in EPSP due to a change in frequency from 0.5 Hz to 2 Hz. Standard error bars are presented.
unitary post synaptic currents remained the same after PE treatment. Haimann et al. (1987) and Shapira et al. (1987) also reported a presynaptic effect of phorbol ester. A lack of an effect of phorbol ester on the postsynaptic membrane of the frog neuromuscular junction was reported by Shapira et al. (1987) while Haimann et al. (1987) suggested some post-synaptic effect. An increase in quanta1 content can be induced by an elevation in the resting level of Ca2+or an increase in Ca2+entry. The latter can result from direct modulation of presynaptic Ca2+channels or from an indirect mechanism such as enhanced invasion of the action potential into the presynaptic terminals, and broadening of the action potential (Hochner et al., 1986b). Our finding that PE alters excitability indicates that such a modulation is possible. However, the presence of a similar augmentation of the EPSP by PE alone, and by PE following augmentation of the EPSP by 3,4 DAP, indicates that enhanced invasion is unlikely to be the main mechanism by which PE exerts its effect. Moreover, in loose patch clamp experiments in the presence of TTX,
which blocks sodium excitability, PE still enhances release. This excludes effects of PE on presynaptic conductances other than Ca2+(Hochner et al., 1986b). To test whether Ca2+ entry is increased by PE we measured twin pulse facilitation. Katz and Miledi (1968) provided evidence for the hyeothesis that facilitation depends upon residual Ca” left from the first impulse. Thus facilitation (amplitude and duration) can be used as an indicator for Ca2+ entry. We observed a minor decrease in facilitation a t Ca2 concentration between 2 and 6 mM. This slight decrease suggests that PE increases Ca2+ entry. Such an increase, if present, must be very small as it did not prolong the duration of facilitation (Parnas and Segel, 1989). A further action of PE can involve processes that are secondary to the increase in Ca2+concentration. PE did not affect the kinetics of release as evaluated from unitary release delay histograms that express the robability of release with time after each change stimulus ( atz and Miledi, 1965). We can conclude that
PHORBOL ESTER AND TRANSMITTER RELEASE
PE does not affect the rate-limitingstep in the release, a step suggested to be the vesicle interaction and fusion with the releasing site (Parnas et al., 1986a). It seems that PKC is not directly involved in the vesicle interaction with the release site. PE could reduce cooperativity of the release process and thereby enhance release. Cooperativity, as calculated from the loghog plot of the initial part of the delay histogram (Parnas et al., 1986b), was not altered by PE. Haimann et al. (1987) suggested that phorbol ester changes the Ca2+cooperativity of release in the neuromuscular junction of the frog. They based their suggestion on the effect of phorbol ester on the relationship between release and extracellular Ca2+ concentration. However, as stated by Van der Hoot (1988) and Parnas et al. (1986b1,an apparent decline in cooperativity may be obtained if intracellular Ca2+content increases. We suggest that PKC can partici ate in the modulation of release. It is well establis ed that PKC is a Ca2+-de endent enzyme (Nishizuka, 1986), so that processes t at lead to Ca2+ accumulation ma modulate release by activating PKC. Indeed, we o served an occlusion ofPE effect due to pretreatment with the Ca2+ ionojhore. Similarly, when loading the terminal with Ca2 b hi h-frequency stimulation, the effect of PE was re uce . This suggests a ossible physiological role of PKC in the activity-depen ent modulation of release that was established for crustacean neuromuscular junctions (Atwood and Wojtowitz, 1986). While we cannot arrive a t a clear conclusion regarding the mechanism of activity of PE, based on techniques currently available for the neuromuscular junction, we can exclude certain possibilities. PE does not reduce cooperativity of release, nor does it intervene with the rate-limiting step of release. PE has very little effect on Ca2+entry, and this effect cannot account for its major activity. The results favor a predominant effect of PE on certain stages in the release process that are secondary to the increase in internal Ca2+ concentration due to presynaptic depolarization. This effect can enhance the affinity of the release machinery to internal Ca2+ concentration andfor the availability of transmitter vesicles and the activity of releasing sites. These may involve phosphorylation of synapsin-like molecules which have been sug ested to modulate the availability of synaptic vesicles linas et al., 1985). ACKNOWLEDGMENTS This research was supported by DFG g a n t No. SFB220A6, Binational grant No. 2653-1-86, and the Goldie-Anna Charitable Trust Endowment Fund. We wish to thank Drs. Hanna Parnas and Itzchak Parnas for their continued support and advice, Drs. Y. Yarom and L. Segel for their careful reading and commentary, and Henry Matzner for fruitful discussions. We would also like to thank Frances Bogot for typing and editing this manuscript.
K
K
d v
B
&
REFERENCES Akers, R.F., Lovinger, D.M., Colley, P.A., Linden, D.J., and Routtenberg, A. (1986) Translocation of protein kinase C activity may mediate hippocampal long-term potentiation. Science, 231:587-589. Andreu, R., and Barett, E.F. (1980) Calcium dependence of evoked transmitter release at very low quantal contents a t the frog neuromuscular junction. J. Physiol. (London),308:79-97. Atwood, H.L. (1967) Crustacean neuromuscular mechanism. Am. ZOO^. , 7:527-551.
99
Atwood, H.L., and Wojtowicz, J.M. (1986) Short-term and long-term plasticity and physiolog~aldifferentiation of crustacean motor synapses. Int. Rev. Neurobiol., 28:275-362. Baraban, J.M., Snyder, S.H., and Alger, B.E. (1985) Protein kinase C regulates ionic conductance in hi pocampal pyramidal neurons Electrophysiological effects of phor!ol ester. Proc. Natl. Acad. Sci USA, 82:2438-2542. Barrett, E.F., and Stevens, C.F. (1972) The kinetics of transmitter release at the frog neuromuscular junction. J. Physiol. (London) 227:691-708. Datyner, N.B., and Gage, P.W. (1980) Phasic secretion ofacetylcholine at a mammalian neuromuscular junction. J. Physiol., 303:299-314. De Riemer, S.A., Stron , J A , Albert, K.A., Greengard, P., and Kaczmarek, L.K. (1985) inhancement of calcium current in Aplysia neurons by phorbol ester and protein kinase C. Nature, 313:313-316. Dodge, F.A., and Rahamimoff, R. (1967) Cooperative action of calcium ions in transmitter release at the neuromuscular junction. J. Physiol., 372:191-205. Dudel, J. (1981) The effect of reduced calcium on quantal unit current release at the crayfish neuromuscular junction. Pflugers Arch., 391:3540. Dudel, J., Parnas, I., Cohen, I., and Franke, C. (1984) Excitability and depolarization-Release characteristics of excitatory nerve terminals in a tail muscle of spiny lobster. Pflugers Arch., 402:293-296. Farley, J., and Aurbach, S. (1986) Reduction of K‘ currents and enhancement of Ca2 current in Hermissenda photoreceptors by phorbol ester and protein kinase C. Nature, 319:220-223. Haimann, C., Meldolesi, J.,and Ceccarelli,B. (1987)The phorbol ester, 12-0-tetradecanoyl-phorbol-13-acetate, enhances the evoked quanta release of acetylcholine at the frog neuromuscular junction. Pflugers Arch., 408:27-31. Hochner, B., Braha, O., Klein, M., and Kandel, E.R. (1986a) Distinct processes in presynaptic facilitation contribute to sensitization and dishabituation in aplysia: Possible involvement of C kinase in dishabituation. Neurosci. Abstr., 2:1340. Hochner, B., Klein, M., Schacher, S., and Kandel, E.R. (1986b)Actionpotential duration and modulation of transmitter release from the sensory neurons of Aplysia in presynaptic facilitation and behaviora! sensitization. Proc. Natl. Acad. Sci. USA, 8323410-8414, Katz, B., and Miledi, R. (1965)The measurement of synaptic delay and the time course of Ach release at the neuromuscular junction. Proc. R. SOC. B., 161:483495. Katz, B., and Miledi, R. (1968) The role of calcium in neuromuscular facilitation. J. Physiol. (London), 195:481492, Llinas, R., McGuiness, T.L., Leonard, C.S., Sugmori, M., and Greengard, P. (1985) Intraterminal injection of synapsin I or calciud calmodulin-dependent protein kinase I1 alters neurotransmitter release a t the squid giant synapse. Proc. Natl. Acad. Sci. USA, 8213035-3039. Malenka, R.C., Madison, D.V., Andrade, R., and Nicoll, R.A. (1986a) Phorbol esters mimic some cholinergic actions in hippocampal pyramidal neurons. J. Neurosci., 6:475480. Malenka, R.C., Madison, D.V., and Nicoll, R.A. (1986b) Potentiation of synaptic transmission in the hippocampus by phorbol esters. Nature, 3211175-177. Matzner, H., Parnas, H., and Parnas, I. (1988) Presynaptic effects of d-tubocurarine on neurotransmltter release at the neuromuscular junction of the frog. J. Physiol., 398:109-121. May, W.S. Jr., Sahyoun, N., Wolf, M., and Cuatrecasas, P. (1985)Role of intracellular calcium mobilization in the regulation of protein kinase C-mediated membrance processes. Nature, 317:549-551. Miller, M.W., Parnas, H., and Parnas, I. (1985)Dopaminer ‘c modulaPhyslol., tion of neuromuscular transmission in the prawn. 363~363-375. Nishizuka, Y. (1986) Studies and perspectives of protein kinase C. Science, 233:305-312. Parnas, H., Dudel, J., and Parnas, I. (1982) Neurotransmitter release and its facilitation in crayfish. I. Saturation kinetics of release and of entry and removal of Ca”. Pflugers Arch., 393:l-14. Parnas, H., Dudel, J., and Parnas, I. (1986a)Neurotransmitter release and its facilitation in crayfish. VII. Depolarization, no Ca inflow determines the time course of phasic release. Pflugers Arch., 406:121-130. Parnas, H., Parnas, I., and Segel, L.A. (1986b) A new method for determining cooperativity in neurotransmitter release. J. Theor. Biol., 119:481499. Parnas, H., and Segel, L.A. (1982) Ways to discern the presynaptic effect of drugs on neurotransmitter release. J. Theor. Biol., 94:923-941. Parnas, H., and Segel, L.A. (1989) Facilitation as a tool.to study the entry of calcium and the mechanism of neurotransmitter release. Prog. Neurobiol., 3 2 - 9 . +
?
100
E. GILAT AND B. HOCHNER
Pozzan, T., Gatti, G., Dozio, N., Vicentini, L.M., andMeldolesi, J . (1984) Ca2 ' dependent and independent release of neurotransmitters from PC 12 cells: A role for protein kinase C activation? J. Cell Biol., 99:628-638. Rink, T.J., Sanchez, A., and Hallam, T.J. (1983) Diacylglycerol and phorbol ester stimulate secretion without raising cytoplasmic free calcium in human platelets. Nature, 305:317-319. Shapira, R., Silberberg, S.A., Ginsburg, S., and Rahamimoff, R. (1987) Activation of protein kinase C augments evoked transmitter release. Nature, 325:58-60.
Van der Kloot, W. (1988) The kinetics of uantal releases during endplate currents at the frog neuromuscujar junction. J. Physiol. (London), 402:605-626. Wolf, M., LeVine, H. 111, Stratford, M. Jr., Cuatrecasas, P., and Sahyoun, N. (1985)A model for intracellular translocation of protein kinase C involving synergism between Ca2 and phorbol esters. Nature, 317546-549. +
