Effects of intra-axonal injection of Ca2+ buffers on evoked release and on facilitation in the crayfish neuromuscular junction

Neurosclence Letters, 125 (1991) 215 218

215

© 1991 Elsevier Scientific Pubhshers Ireland Ltd 0304-3940/91/$ 03 50 ADONIS 030439409100189U N S L 07711

Effects of intra-axonal injection of Ca 2 + buffers on evoked release and on facilitation in the crayfish neuromuscular junction Binyamin Hochner, Hanna Parnas and Itzchak Parnas The Otto Loewt Center for Cellular and Molecular Neuroblology, The Hebrew Umversuy of Jerusalem, Jerusalem (Israel) (Received 25 January 1991, Revised version received 31 January 1991, Accepted 31 January 1991)

Key words

Ca 2+ buffer, Neuromuscular junction, Transmitter release, Synapse, Facilitation

Ca 2+ buffers were injected into the excitatory axon of the crayfish opener muscle The m a g m t u d e and time course of evoked release and of facilitation were measured_ E G T A (on-rate about 106 M-~s-~) had no effect on evoked release but reduced facdltatxon B A P T A and mtr-5, buffers with s x m d a r / ~ ' s but faster on-rates, reduced both evoked release and facdltaUon However, these buffers had no effect on the time course of evoked release These results show that fast Ca ~+ buffers reduce the Ca 2÷ transient assocmted with evoked release and also the level of residual Ca 2+ involved in facdltatlon However, Ca 2+ buffering is not the m e c h a m s m which controls the time course of release

In fast synapses, such as the giant synapse of the squid [14, 15] or m frog [11, 12] and crayfish [16] neuromuscular junctions, at room temperature, evoked release starts 0 . 2 ~ . 5 ms after the pulse and terminates within 1-2 ms [4]. The 'Ca z+ hypothesis' contends that this brief duration of evoked release is due to rapid removal of Ca 2+ from below the release sites [4, 6, 9, 19, 20]. Of the known Ca 2+ regulatory mechanisms, pumps and exchangers act too slowly to affect such fast removal of Ca 2+ [6], but Ca 2+ diffusion or Ca z+ buffering may be sufficiently rapid. On the other hand, twin pulse facihtatlon, a process known to depend on residual Ca 2+ [13], is slow enough to be affected by the Na/Ca exchange mechanism [3, 6]. In the present study we tested whether Ca z+ buffers with different on-rates affect the amount and time course of release as well as that of facilitation. According to the Ca 2+ hypothesis, buffers which bind Ca z+ faster than the 'release sites' would be expected to reduce evoked release and facilitation, while slower buffers would be expected to reduce only facilitation as they would affect only the level of residual Ca z+ Also, if Ca z+ buffering is an important mechanism In determining the time course of release, fast buffers should shorten the time course of evoked release. Experiments were performed on the first walking leg of small (4-4 5 cm) crayfish, Procambarus clarku, obtained from Atchafalaya Biological Supply, LouiCorrespondence I Parnas, The Otto Loewl Center for Cellular and Molecular Neuroblology, The Hebrew University of Jerusalem, Jerusalem 91904, Israel

siana. The opener muscle was dissected as described by Wojtowicz and Atwood [24] and Hochner et al. [10]. Intracellular injection of the buffer into the excltor axon was performed as in Hochner et al. [10] Synaptic potentials were recorded intracellularly using 2.5 M potassium acetate electrodes (6--10 MO), and single quanta events were recorded extracellularly with the macropatch techtuque [7] The time course of release was determined from synaptlc delay histograms [11, 12]. The nerve was stimulated with two pulses 20 ms apart. Repetition rate was 1-3.3 Hz For facilitation experiments, the repetition rate was 0.5-1 Hz. Facilitation was defined as the ratio between the averages of second and first responses, 100 sweeps were taken for averaging. When excitatory postsynaptic potentials (EPSPs) were smaller than 0.2 mV, 200 sweeps were averaged. At low release levels single quanta were counted. The preparations were perfused at a rate of 3 ml/min (bath volume 2 ml) with physiological solution containmg (in raM)' NaC1 220, KC1 5.4, CaClz 12, MgC12 2.5 and Tns maleate 10; pH 7.6. Bath temperature was controlled at 18-20°C. Ca 2+ buffers used were: BAPTA (l,2-bis (o-ammophenoxy) ethane N,N,N',N'-tetracetlc acid) [23], nitr-5 [1] and E G T A (ethylene glycol bis (amino ethyl ether) N,N,N'-N'-tetracetic acid) [23] The on-rate of E G T A is about 106 M - i s -~, and mtr-5 and BAPTA have faster on-rates [23] The 3 buffers have similar Kd'S of about 0.1 /~M. For buffer injection the microelectrode contained: BAPTA or nitr-5, 50 mM, 100 mM KCI, 35 m M CaC12, 0.5% fast green and 10 mM Tris-HCl. pH 7.4 For E G T A injection there were two

216

changes 100 m M E G T A was used and the solution did not contain Ca 2+. An experiment demonstrating the time course of the reduction in the EPSP amphtude after buffer injection is shown in Fig 1. In this case the buffer used was nitr-5 (nitr-5 In the dark has buffering properties similar to those of BAPTA [1]) The axon action potential recorded w~th the nitr-5 electrode is shown in Fig. 1A (upper traces) The corresponding synaptlc potentials recorded from a muscle fiber are shown in Fig IA, lower traces. A gradual reduction in the amplitudes of both the first and second responses (elicited by twin impulses) is ewdent. In Fig 1B the amplitudes of the first response (filled circles) and of facihtatlon (open circles) are given as a percentage of the control before mtr-5 injection Approximately 40 min after injection began, the first EPSP of the pair was almost completely abolished. It was reduced from an initial average amplitude of 1 4 mV to 0 035 mV (This reduction did not result from possible toxic effects of the injected buffer, as repeated stimulanon led to recovery of the EPSP amplitude, probably due to saturation of the buffer with Ca 2+ [10].) At the same nine facilitation measured at 20 ms mtervals declined from 2 to 1 3 Similar results were obtained in 17

A

nitr-5,

epsp

0

II

and

facilitation

19

52

47mm

LLLLLL \

B

20ms

;~

I00(

,'x 80

\~

experiments where nitr-5 was reJected and 4 experiments with BAPTA For the 21 experiments the average steady state decline in EPSP amplitude was 87 % _+7.0 (_+ S D ) Results such as those shown m Fig 1 ln&cate that, with the experimental procedure employed, the injected nitr-5 or BAPTA reached the nerve terminals, and their buffering properties were effective in regulating the amount of evoked release and the level of short-term facilitation To test the effects of Ca 2+ buffering on the duration of facilitation, twin pulses were given at different intervals in the control and after nitr-5 injection. The results of such an experiment are given in Fig 2 In the control, the first EPSP of the pair was 0 13 inV. It declined to 0 03 mV after nltr-5 took effect, a reduction of 77% Facilitation was then measured at 20, 40 and 80 ms At all intervals, facilitation was reduced by the Ca 2+ buffer, and the duration of facilitation was shorter These results support the residual Ca 2+ hypothesis for faclhtatlon [13] The finding that intra-axonal injection of the fast Ca 2+ buffers (nltr-5 and BAPTA) reduced release and facllltanon shows that Ca 2+ is removed more rapidly than normal from below the release sites, and that the mjected buffer competes for Ca 2+ with the Ca 2+ binding molecule involved in the process of release If xt is assumed that entry of Ca 2+ and elevation of mtracellular Ca 2+ concentration to some critical level ~s the main trigger for lnlnatlon of release and that removal of Ca 2+ is the main reason for termination of release [4, 9, 20], tt would follow that a more rapid removal of Ca a+ [18] should cause release to stop sooner. We measured the time course of release by constructing synaptlc delay histograms [11, 12] before and after injection of nltr-5 Single quantum events were recorded w~th a macropatch electrode [7] and their delay from the nerve terminal potential measured The reset in Fig 3 shows the axon

o -focfllhon

-

2.5 t-0 ,6..

o

b

6o

q'~

40-

o -control

_~ 2.0

0

l

0

I

I0

t

I

t

20

1-

30

i

i

40

•-nilr-5

F~w J

50

rnln Fig 1 Effect of nltr-5 injection on EPSP amphtude and facllltanon A upper traces presynaptlc action potentials recorded with the mtr-5 inJecting electrode Lower traces - mtracellular recording of EPSPs from a muscle fiber The axon was stimulated with twin pulses 20 ms apart, at a rate of l Hz Each recording is an average of 100-200 sweeps Time after beginning of inJection is given at the top, B amphtude of the first of the two EPSPs (filled circles) and facilitation (open circles) as a percent of the control values m A

1.0

)

0

I

20

J

I

40 Interval

L

I

60 (ms)

t

1

80

i

I

I00

Fig 2 Nltr-5 reduces the magmtude and durahon of twin pulse faolltatlon Facilitation (epp2/eppl) at different intervals in controls (open circles) and after mtr-5 rejection (filled circles) For each interval, 100200 responses were averaged

217

o c

i~

4mY I

E Z

J

qo ms 20

n = 2000 control- 6 4 2 quanla mlr-5-

2 3 5 quanta

0 0

05

I0 ms

15

20

Fig 3 Synaptlc delay histograms in control and after mtr-5 rejection Insert samples of macropatch recordings of umtary releases Delay was measured as the time elapsing from the peak of the negative phase of the excitatory nerve terminal potential to the onset of the synaptlc current The presynaptlc action potentmls are seen in the lower trace A synaptlc delay histograms (bin size 0 1 ms) constructed from the responses to the first and second stlmuh, 2000 pairs of pulses 20 ms apart were employed at 3 3 Hz. The empty and filled bars depict the responses prior to and followmg rejection of nltr-5, respectively Without mtr-5, 642 releases were obtained for the two pulses The quantml content of the first stimulus (273 releases out of 2000 pulses) was 0 13 Following mtr-5 injection, the quantal content of the first pulse decreased to 0 05 (104 out of 2000 pulses), and 235 releases were obtained for the two pulses B each of the two h~stograms normahzed to ~ts total number of releases It is clear that despite the 62% reduction m the quantal content, the time course of release remained unaltered

action potential (lowest trace) and the smgle q u a n t a T w o t h o u s a n d t w m impulses 20 ms a p a r t (4000 stimuli) were given at a repetition rate of 3 3 Hz. The c o n t r o l yielded 645 q u a n t a , a n d after nltr-5 injection, for the same n u m b e r o f pulses, only 235 q u a n t a were obtained. I n Fig. 3A, the delay histograms of the c o n t r o l (empty bars) a n d after mtr-5 injection (black bars) are presented as actual n u m b e r of q u a n t a (for 0.1 ms bins). The normalized histograms are given in Fig. 3B. It is clear that the time course o f release was n o t altered, even t h o u g h release was reduced by 62%. Stmilar results were o b t a i n e d in three m o r e experiments W e also tested for the effects of E G T A , a calcium buffer wtth a slower on-rate. Fig 4 shows EPSPs after 3 consecutive action potentials 30 ms apart, for controls (upper trace) a n d 53 m i n after rejection of E G T A (middle trace) The b o t t o m trace is a superposltlon of these two traces. N o t e that the first response was n o t altered while factlitation o f the second a n d third responses was reduced (66 % a n d 54 % respectwely). The time course o f release was n o t affected by E G T A . T h u s a buffer with a slower o n - r a t e does n o t reduce evoked release. These results suggest that the 'release sttes' brad Ca 2+ faster

Fig 4 Effect of presynapUcallyrejected EGTA on release and facdltatlon Three pulses 30 ms apart were gwen Upper trace - control_ Middle trace 40 mln after EGTA rejection Lowest trace - superpos~t~on of the two traces Note that the first responses are the same whde facdltat~on of the second and third responses ~s reduced_ Repetition rate 0 5 Hz Average of 200 sweeps

t h a n E G T A These results also indicate that nttr-5 a n d B A P T A m a i n l y reduced the Ca 2+ t r a n s i e n t a n d n o t the restmg Ca 2+ c o n c e n t r a t i o n . This is because the 3 buffers have similar Kd'S b u t different on-rates, a n d E G T A had no effect o n the a m p h t u d e of the evoked release. The effects o n release a n d facilitation of Ca 2+ buffers have previously been studied by T a n a b e a n d K i j i m a [21, 22], Adler et al [2] a n d Duffy et al. [8] But these a u t h o r s dtd n o t measure the time course of release before a n d after application o f buffer. The finding that Ca 2+ buffers injected lntracellularly reduced the q u a n t a l c o n t e n t without affecting the time course of release implies that the n o r m a l Ca 2+ buffers, within the terminal, do n o t determ i n e the time course of n e u r o t r a n s m i t t e r release. This research was supported by the Goldle A n n a T r u s t F u n d (to I P.), by the D F G (to I P., H P. a n d J Dudel), a n d by a U n i t e d States-Israel BSF g r a n t (to H P.), I.P. is the Greenfield Professor of N e u r o b i o l o g y . W e t h a n k F. Bogot for p r e p a r i n g the m a n u s c r i p t for publicatton. 1 Adams, S R, Kao, J P Y, Grynklewlcz, G, Mmta, A and Tslen, R Y, Biologicallyuseful chelators that release Ca2+ upon dlummaUon, J Am Chem_Soc, 110 (1988) 3212 3220 2 Adler, E M, Augustine, G J, Duffy, S N and Charlton, M P, Modulation of neurotransm~tter release by mtracellular calcium chelators at the squid gaant synapse, Soc Neuroscl Abstr, 31 (1988) 5 3 Arechlga, H, Cannone, A, Parnas, H and Parnas, I, Blockage of synaptlc release by brief hyperpolanzlng pulses, J Physiol., 430 (1990) 119-133

218 4 Augustine, G J , Charlton, M P and Smith, S J, Calcmm action m synaptlc transmitter release, Annu Rev Neuroscl, 10 (1987) 633~93 5 Blttner, G D and Sewell, V L , Facilitation of crayfish neuromuscularjunctlons, J Comp Physlol, 109 (1976) 287 308 6 Blausteln, M P , Calcmm transport and buffenng m neurons, Trends Neuroscl, 11 (1988) 438443 7 Dudel, J , The effect of reduced calcium on quantal unit current and release at the crayfish neuromuscular junction, Pflugers Arch, 391 (1981) 3540 8 Duffy, S N , Wlnslow, J L and Charlton, M P , Effect of increased presynaptlc calcium buffering capacity on short-term facd~tatlon, Soc Neuroscl Abstr 15 (1) (1989) 475 9 Fogelson, A L and Zucker, R S, Presynaptlc calcmm &ffUSlOn from various arrays of single channels lmphcatlons for transmitter release and synaptlc facilitation, Blophys J , 48 (1985) 1003-1017 10 Hochner, B, Parnas, H and Parnas, I , Membrane depolarization evokes neurotransmltter release in the absence of calcium entry, Nature, 342 (1989) 433435 11 Katz, B and MlledJ, R , The measurement of synaptlc delay, and the t~me course of acetylchohne release at the neuromuscular junction, Proc Roy Soc Lond B, 161 (1965a)656~70 12 Katz, B and Mile&, R , The effect of temperature on the synaptlc delay at the neuromuscular Junction, J Physlol, 181 (1965b) 656--670 13 Katz, B and Miledl, R . The role ofcalcmm m neuromuscular faclhtatlon, J Physlol, 195 (1968) 481492 14 Lllnfis, R , Steinberg, I Z and Walton, K . Presynaptlc calcium currents m squid giant synapse, Biophys J , 33 (1981) 289-322 15 Lhnfis, R , Steinberg, I Z and Walton, K , Relationship between presynaptlc calcium current and post synaptlc potential in squid giant synapse, Biophys J , 33 (1981) 323-352

16 Parnas, H , Dudel, J and Parnas, I, Neurotransmltter release and its facilitation in crayfish VII Another voltage dependent process besides Ca entry controls the time course of phasic release, Pflugers Arch, 406 (1986) 121 130 17 Parnas, I , Parnas, H and Dudel, J , Neurotransmltter release and its faclhtatlon in crayfish II Duration of faclhtatlon and removal processes of calcium from the terminal Pflugers Arch, 393 (1982) 232-236 18 Sala, F and Hern/mdez-Cruz, A , Calcium diffusion modeling in a spherical neuron Relevance of buffering properties, Blophys J , 57 (1990) 313 324 19 Simon, S M and Lhnas, R R , Compartmentahzatlon of the submembrane calcium activity during calcium reflux and Its significance in transmitter release, Blophys J , 48 (1985) 485498 20 Smith, S J and Augustme, G J , Calcium Ions, active zones and synaptlc transmitter release, Trends Neuroscl, 11 (1988) 458464 21 Tanabe, N and Ktglma, H , Transmitter release at the frog end plate loaded with a Ca2+-chelator, Bapta, hypertomcJty and erythrosin B augment the release independently of internal Ca 2+, Neuroscl Lett, 92 (1988) 52-57 22 Tanabe, N and K~glma, H , Both augmentation and potentiation occur independently of internal Ca 2+ at the frog neuromuscular junction, Neuroscl Lett, 99 (1989) 147 152 23 Tslen, R Y , New calcium ln&cators and buffers with high selectivity against magnesium and protons design, synthesis, and properties of pro to type structures, Biochemistry, 19 (1980) 2396-2404 24 Wo.ltOWlCZ,J M and Atwood, H L , Presynaptlc membrane potential and transmitter release at the crayfish neuromuscular junction, J Neurophyslol, 52 (1984) 99 113