- Email: [email protected]
Spontaneous release of transmitter from the growth cones of Xenopus neurons in vitro: The influence of Ca2+ and Mg2+ ions
DEVELOPMENTAL
BIOLOGY
113, 373-380 (1986)
Spontaneous Release of Transmitter from the Growth Cones of Xenopus Neurons in Vitro: The Influence of Ca*+ and Mg*+ Ions STEVEN
H. YOUNG’
To determine whether spontaneous release of transmitter from the growth cones of neurons exhibits properties similar to the spontaneous release which occurs from the neurons at the neuromuscular junction, release of transmitter from the growth cones of XenolnL.ncs neurons in culture was monitored in salines containing varying calcium and magnesium concentrations. Release was monitored by use of an outside-out piece of muscle membrane attached to a patch clamp electrode. Spontaneous release of transmitter from the growth cones in standard saline (2 m M CaCl,, 1 m M M&l,) produces clusters of single-channel openings in the muscle membrane. Clusters are seen to consist of two types: a series of high-frequency channel openings, called “bursts,” and clusters of low-frequency channel openings called “singles.” The bursts were identified and examined for their possible relationship to MEPP-producing release, and the singles were identified and examined for their possible relationship to “leak” release of the neuromuscular junction. When the external saline contains high calcium (10 mA4 CaCl,, 1 m M MgC&) or high magnesium (2 m M CaCl,, 9 m M Mg&), the frequencies of both “bursts” and “singles” was greatly reduced. This reduction in release persists if the neurons are gro\vn in the high-calcium or high-magnesium solutions. When the saline is a low-calcium solution (0 m M CaClz, 3 m M ME&) the growth cones release transmitter at rates similar to those from standard saline. These results indicate that although the spontaneous release from the growth cone shares one characteristic with the leak release, neither the burst nor the singles release from the growth cones share exact relationship with either the MEPP producing release or the leak release. This suggests that further development of the mechanisms for spontaneous release of neurotransmitter occurs after nerve-muscle contact. (c I%%Academic Press. Inc INTRODUCTION
release change with development of the synapse? A comparison of the relations between spontaneous release observed at the mature neuromuscular junction, and the spontaneous release from the pre-muscle-contact growth cone presents an opportunity to study development of transmitter release and possibly to elucidate the mechanism(s) of the spontaneous release of transmitter. In Xenopus (Sun and Poo, 1985) and in mouse (Vsykocil et ah, 1983) neuromuscular junction, the amount of leak release is decreased by an increase in external calcium concentration, while the frequency of the spontaneous MEPPs increases. Since these responses are in opposite directions, an increase of external calcium concentration then can be used as a means to separate leak-like release from MEPP-like release. The experiments reported here were performed to determine whether spontaneous release of transmitter from the growth cone exhibits similar sensitivity to calcium ions as those of either the MEPP- or leak-producing release which occur at the neuromuscular junction. To address this question, spontaneous release from the growth cones of Xenopus neurons in culture was monitored in salines containing varying calcium and magnesium concentrations. In an earlier report Young and Poo (1983) had shown that spontaneous release of transmitter from the growth cones of Xenopa neurons occurs
The growth cones of isolated neurons in culture will release transmitter spontaneously (Young and Poo, 1983), and in response to electrical stimulation (Hume et al., 1983). The spontaneous release of transmitter is of particular interest when considering synaptogenesis of the neuromuscular junction because such release presumably occurs throughout the period of growth of the neurite toward the muscle cell, and it is natural to speculate that such release could reflect release of factors which might influence the formation and stability of the neuromuscular junction. Furthermore, two forms of spontaneous release of transmitter are seen at the mature neuromuscular junction: a packaged, or quanta1 form, in which release occurs in bursts and produces the miniature endplate potential (MEPP) in the muscle cell (de1 Castillo and Katz, 1954), and a low-level “leak” of transmitter from the nerve terminal (Katz and Miledi, 1977; Katz and Miledi, 1981). Does the growth cone already possess the characteristics of spontaneous transmitter release seen at the mature synapse? Or do the properties of spontaneous ’ Present address: Jerry Lewis Neuromuscular Research UCLA School of Medicine, Los Angeles, Calif. 90024.
Center,
373
0012-1606/86 $3.00 Copyright All rights
0 1986 by Academic Press. Inc. of reproduction in any form reserved
374
DEVELOPMENTALBIOLOGY
in a manner which produces clusters of single-channel openings in the “probe,” the transmitter measuring system. In the present report, the characteristics of these clusters are examined in greater detail. Clusters were seen to consist of two types: a series of high-frequency channel openings, called “bursts,” and clusters of lowfrequency channel openings called “singles.” The bursts are identified and examined for their possible relationship to the MEPP-producing release, and the singles are identified and examined for their possible relationship to leak release. Results show that neither the burst nor the singles release from the growth cones share exact relationship with either the MEPP-producing release or the leak release of the neuromuscular junction, and suggest that further development of the mechanism(s) of spontaneous release of neurotransmitter must occur after nervemuscle contact. METHODS
Solutions. The standard external saline contained (in m&f) 125 NaCl, 2 KCl, 2 CaQ, 1 MgQ, 5 Hepes, at pH 7.2. All experimental salines differed only in the amounts of CaCle and/or MgCl, present. The standard internal saline (pipette solution) contained (in mM) 82 KCl, 35 KOH, 1 CaClp, 11 EGTA, 1 MgCl,, 5 Hepes, pH 7.2. Disaggregating saline contained (in mM) 125 NaCl, 2 KCl, 0.1 EDTA, 1 unit/ml penicillin-streptomycin, pH 7.2. Culture medium contained 85% saline, 10% L-15 (GIBCO), 5% fetal calf serum (ZETA-D, AMF Biological and Diagnostic Products Company), and 1 unit/ml penicillin-streptomycin. With the exception of two experiments which investigated effects of growth in high-calcium- or high-magnesium-containing media, all cultures were grown in media using standard saline. Preparation of nerve-muscle cell culture. Preparation of nerve-muscle cultures follows the procedure of Poo et al. (1978). Briefly, fertilized eggs of Xenopus laevis were allowed to develop until stages 17-20 (Nieuwkoop and Faber, 1967). The neural tube, notocord, and myotomal regions were removed from the embryos, and placed in disaggregating solution for 30 min. Following tissue disaggregation, the cells were plated onto clean glass slides which contained culture medium. After 1 day in culture, nerve cells sent out neurites, and most muscle cells remained spherical or nearly spherical (see Fig. 1). At 2-4 days in culture, most surviving nerve cells had contacted muscle cells. Therefore, to study release of transmitter from growth cones of nerve cells which have not previously contacted muscle cells, and to standardize the age of nerve cells, all cultures used in this study were 1 day old. Detection of transmitter release. Methods for detection of transmitter release at very high sensitivity follow
VOLUME 113, 1986
FIG. 1. Phase contrast photomicrograph of a Xenoyl~s spinal cord neuron (cell body marked N) in culture. Spherical muscle cells (marked MJ are attached to the bottom of the glass slide. The small phase bright particles are yolk granules. On the right side of the photograph, a patch pipette (P) with a piece of muscle membrane at its tip is moved toward the tip of a neurite to monitor release of transmitter from the growth cone. Final position of the pipette will be 3-6 pm from the growth cone. Scale bar: 60 pm.
those of Young and Poo (1983) and Hume et al. (1983). The central element of the method is the use of the excised membrane patch technique (Horn and Patlak, 1980), which can be used to resolve the opening of single ionic channels. When a piece of membrane containing receptor-ionic channels is formed across the tip of a glass pipette (~1 pm diam) with the outer surface of the membrane facing the bath, the pipette now becomes a very sensitive probe for molecules which activate the membrane receptor channels. For example, the membranes of 1 day-in-culture Xenopus muscle cells contain acetylcholine (ACh) receptors at a high enough density that ACh receptor channel openings can be detected at ACh concentrations as low as 30 nM. This concentration corresponds to about 20 molecules/pm”, and illustrates the high sensitivity of the method. In addition, because of the small size of the pipette tip, the probe can be manipulated into position very close to the growth cone membrane (within 3-5 pm). To form an outside-out patch, the fire polished tip of the glass pipette (which contains internal solution) is placed against a muscle cell membrane. Slight suction applied to the top of the pipette produces a very tight membrane-glass seal (>lO GQ). Further suction ruptures the small area of membrane under the pipette, and allows access to the cell interior, where the muscle cell resting potential is measured. Finally, when the pipette is pulled away from the muscle cell, a small piece of muscle membrane will form across the tip of the pipette, with the outer surface of the membrane facing the bath. Figure 1 shows such a
STEVEN
H. YOITNG
Trunsmitter
“transmitter probe” approaching the growth cone of a neurite. Experimental protocol. Glass slides containing l-dayold nerve and muscle cells were removed from a humidifying cabinet, and the culture medium was very slowly (150 pl/min at 250 yl total volume) exchanged for experimental saline. The solution change was performed slowly to minimize the possibility of induced transmitter release due to bulk flow of fluid across the growth cone. As a control for possible flow-induced effects, transmitter release has been observed in cultures before the exchange of culture medium for experimental saline, indicating that Ilow of solution is not the sole cause of release under experimental conditions. Furthermore, to minimize any possible effect of fluid tlow on release, recordings were never started earlier than 4 min after solution flow had stopped. The outside-out patch of muscle membrane was formed on the tip of the pipette, and the patch membrane potential was set to the value of the muscle cell resting potential (-60 to -90 mV). Then baseline (BATH) recordings were made of the activity of the ionic channels in the patch when the probe was at least 100 pm away from any portion of any neurite. Records in the bath were taken for 2-5 min, and then the probe was positioned to within 3-6 pm of the tip of a growth cone to monitor release from that area. Only nerve cells which clearly were not in contact with any muscle cell were used in this study (see Fig. 1). Analysis. Records of current from open ionic channels were recorded on tape by an FM tape recorder and played back at slower speed onto an oscillographic recorder (Gould 2200) for analysis. The previous study (Young and Poo, 1983) has shown that most of the single-channel openings are due to ACh receptor channels, although other channel types are present. In the present study, no separation of channel types before analysis has been performed. Two types of channel opening patterns, which reflect transmitter release patterns, are observed: a series of single-channel openings occurring at low frequency, termed “singles,” and a high-frequency cluster of channel openings termed “bursts.” A burst is defined as a series of channel openings lasting longer than 1 set, with no more than 2 set between openings. Burst frequency is measured directly from records of current obtained near the growth cone. No corrections to the measured frequencies are performed, since bursting is not observed in BATH recording. However, singles openings can be observed in BATH records. These openings could possibly result from at least three sources: (1) residual transmitter in the saline, (2) openings of non-ligand bound channels (Brehm et ab, 1984), or (3) from residual membrane tension in the
Rrletrse bg Growth
375
Cones
patch, occurring as a result of patch excision from the muscle, producing pressure-activated channels (Guharay and Sachs, 1984). The measured frequency of singles openings obtained near the growth cones were corrected by subtraction of singles frequencies present in BATH recordings. Singles frequencies can be highly variable, and the correction procedure can produce corrected singles frequencies from the growth cone which are negative. Average burst and corrected singles frequencies in experimental salines were compared to records obtained in the standard external saline using Student’s t test. RESULTS
Spontaneous
Release of Transmitter
in Standard
Saline
Typical appearance of ionic channel openings produced by spontaneous release of transmitter from two separate growth cones in standard saline are shown in Fig. 2. All growth cones which showed release of transmitter exhibited such release by causing groupings or clusters of channel openings. The beginning of the current record taken near growth cone 1 (GC 1) shows two bursts of channel openings (marked B) followed at a later time by two small clusters of singles (marked S) flanking a small burst (middle trace). The recording taken near growth cone 2 (GC 2) shows a long-lasting series of singles. The definition of bursts and singles is described in Methods. Whether a growth cone showed bursts or sin-
BATH -c*
$LIOS BATH -Gc~
FIG. 2. Spontaneous release of transmitter from two different growth cones in standard external saline as measured by the current in the patch pipette. Top traces: The current record marked BATH was taken with the pipette at least 100 pm away from any neuron. When the pipette was moved to within 3-6 pm of growth cone 1 (GC l), two bursts (marked B) of channel openings were recorded. Middle trace: Continuation of the record of GC 1. A small burst is flanked by two clusters of singles (marked S). Bottom traces: Bath recording from a different membrane patch, and current record from a second growth cone (GC 2) which showed a long duration cluster of singles (for definition of burst and singles see Methods).
376
DEVELOPMENTAL
u+vI=I 2 -8
I,, 4
SINGLES
, 28
20
12
FREQUENCY
BIOLOGY
(per
mini
FI(;. 3. No correlation exists between burst and singles frequencies. The average burst frequencies are plotted against the average singles frequencies from 13 growth cones in standard saline
gles clusters (or both) the frequency of occurrence of such events was not highly regular. Some growth cones showed predominantly bursting release, some showed both bursts and singles (GC 1, Fig. 2), and some showed predominantly singles release (GC 2, Fig. 2). It is of interest to know if bursts and singles frequencies are correlated, to help to determine if these different release patterns reflect different release processes. If bursts and singles are expressions of very similar processes, it might be expected that growth cones showing high rates of bursting would also show high singles frequencies. Figure 3 is a plot of average burst frequency vs average singles frequency from each of 13
VOLUME
113. 1986
growth cones recorded in standard saline. The figure shows that there is no correlation between bursts and singles frequencies. The grand average burst and singles frequencies from all 13 growth cones in standard saline are shown in Fig. 4. The left panel of Fig. 4 shows average frequencies (with standard error) of bursts (labeled GC) and singles. Burst and singles openings were detected in 7 out of 13 (54%)growth cones, a fraction similar to that found in the previous report (51%, Young and Poo, 1983). Note the large standard error of the singles frequency. Corrected singles frequencies varied over a large range, under all experimental conditions. The large standard errors which resulted decreased the levels of significance of difference between average frequencies of singles in standard saline and each of the experimental salines (see below), even when the fraction of growth cones exhibiting such release was reduced. The Eflect of Elevated External Calcium on Spontaneous Release of Transmitter Transmitter release from growth cones exposed to increased external calcium ion concentrations was monitored, and the results are plotted in Fig. 4. The center panel shows that average burst and singles frequencies were reduced when the external saline contained elevated external calcium concentrations (10 mM CaCl,). Only 2 out of 10 growth cones showed any bursts, and the average burst frequency was significantly decreased (P < 0.02) compared to standard saline. The average
2 mM Ca++
10 mM Ca++
2 mM Ca++
1 mM Mg++
1 mM Mg++
9 mM Mg++
7
P10.2 P10.02
BATH
BATH
GC
BURSTS
SINGLES
BURSTS
SINGLES
BURSTS
GC
SINGLES
FIG. 4. Average burst and singles frequencies in salines with high concentrations of calcium or magnesium were reduced below values in standard saline. Concentration of other ions remained unchanged. The top of each chart is labeled with the appropriate calcium and magnesium concentrations. The stippled bars illustrate average burst frequencies (with standard error) in the BATH and near the growth cone (GC). The open bars illustrate average corrected singles frequency (with standard error). The left axis of each chart is marked for bursts frequency, and the right axis is marked for singles frequency. The fractions of growth cones showing release are printed above each bar. Left panel: Results from standard saline. Center and right panels: Results from two experimental solutions. Each frequency was compared to the corresponding frequency in standard saline. Levels of significance are printed above each bar (see Methods).
377 corrected singles frequency was reduced, but a large standard error prevents a greatly significant difference from that of standard saline (P < 0.2). However, the fraction of growth cones showing release (2110) was reduced below that observed in standard saline (P < 0.04). The Effect of Elevated External Magnesium Spontaneous Release of Transmitter
on
To determine whether the reduction of spontaneous release is due specifically to calcium ions, a second set of experiments were performed in which CaClz was maintained at 2 mM, but MgClz was increased to 9 mM. In this manner, both experimental solutions contained 11 mM of divalent ions, with the calcium concentration of the second solution remaining equal to that of the standard saline. The results of these experiments are plotted in the right panel of Fig. 4. Both burst and corrected singles frequencies are reduced. Only 1 out of 7 growth cones showed bursts, and the resulting average burst frequency is lower than that in standard saline (P < 0.02). Average corrected singles frequencies were reduced, but a large standard error prevented a greatly significant difference from that in standard saline. The fraction of growth cones showing singles (28%)was reduced from that in the standard saline. Spontaneous Release of Transmitter Calcium Concentration
at Low External
To acertain whether release continues at low calcium concentrations, nerve cells were bathed in external saline that did not contain CaC&, and with MgClp increased to 3 mM (in order that total divalents remained at 3 mM, the same value as in standard saline). The experimental design is not intended to test for an absolute requirement of spontaneous release upon calcium, but to test if release is affected at calcium levels far below the l-2 mM which is present in most physiological salines. Results from one growth cone are shown in Fig. 5, and illustrate that spontaneous transmitter release continues to occur at low levels of calcium. The outside-out patches were far
BATH -r
.d 10
mg
“-M”
FIN:. 5. Spontaneous release of transmitter from a growth cone in low-calcium solution. Current records from the BATH and near the growth cone (GC) of a neuron bathed in saline without calcium (0 m M CaC12, 3 m M M&l,).
0 mM Ca++ 3 mM Mg++
.L E
P20.4
3
k a &
3/4 2
I5 2 w [r LL
’
l&?&lid
P10.95 3/4 1 / 3 ii$+1/i!; /iiii/i/ ai! ,, ,x1.: @##j BATH GC
BURSTS
SINGLES
FIN;. 6. Frequencies of spontaneous release of transmitter from growth cones in low-calcium solution are not significantly different from those in standard saline. Chart markings as in Fig. 4.
less stable in the low-calcium saline, tending to break easily. Combined data from 4 growth cones using 3 patches which remained stable for at least 4 min are shown in Fig. 6. One of the patches showed a single burst in the bath. Bursting release was observed at 3 of 4 growth cones tested, and the average burst frequency was not significantly different from standard saline (P < 0.95). Singles were seen in 3 out of 4 growth cones tested. Spontaneous Release of Transmitter upon Continuous Exposure to High-Calcium or -Magnesium Solutions It is possible that whatever process(es) are involved in the spontaneous release of transmitter from the growth cones, they can “adapt” in time to maintained increased calcium or magnesium levels, and eventually show release rates similar to those obtained in standard saline. To test this possibility, nerve cells were grown in culture media containing elevated calcium or magnesium levels, and then placed in recording salines with the same ionic concentrations. The left panel of Fig. ‘7 shows results from experiments in which culture medium containing 10 mM CaClz and 1 mM MgClz was exchanged for saline also containing 10 mM CaClz and 1 mM MgC12 before the start of recording. The panel shows that both bursts and singles frequencies remain below those in standard saline-the spontaneous release does not adapt (after 1 day in culture) to maintained elevated calcium levels. The right panel of Fig. 7 shows results from experiments in which the cultures were maintained in medium containing 2 mM CaClz and 9 mM MgC12. Again, the frequencies and fractions of growth cones showing release remain below those recorded in stan-
378
DEVELOPMENTAL BIOLOGY .E E
z f
mM
Ca++
2
mM
Cat+
1 mM
Mg++
9
mM
Mgtf
10
r4
2l
BURSTS
2i
SINGLES
r4
BURSTS
SINGLES
FIG. ‘7. Growth cones maintained in high-calcium or high-magnesium solutions do not “adapt” and show rates of spontaneous release of transmitter similar to those obtained in standard saline. Neurons were cultured in high-calciumor high-magnesium-containing media for 1 day before recording spontaneous release. Chart markings as in Fig. 4.
dard saline. The spontaneous release cannot maintained elevated magnesium levels.
adapt
to
DISCUSSION
In the presence of 2 mMcalcium, growth cones of Xenopus spinal cord neurons kept in culture can spontaneously release neurotransmitter. The release is not steady, but occurs in a manner which produces clusters of channel openings in the outside-out muscle membrane containing patch pipette. These clusters can be separated into two classes: a series of high-frequency channel openings, called bursts, and clusters of low-frequency openings called singles. The average frequencies of both bursts and singles are reduced when the external saline contains high calcium (10 mM). The reduction in spontaneous release from the growth cone is not specific to the calcium ion-an increase in magnesium ion concentration (9 mM) will also reduce spontaneous release. The appearance of bursts and singles is not dependent upon the presence of millimolar levels of calcium. When MgCl, is substituted for CaC12 in the standard saline, transmitter continues to be released from the growth cones, with average burst and singles frequencies not significantly different from those in standard saline. Burst Release from the Growth Cone Is Not Identical MEPP Producing Release at the Synapse
to
The burst openings are caused by the immediate release of a concentrated amount of transmitter near the patch pipette. This is the type of behavior which might be expected to occur with the release of the contents of a synaptic vesicle filled with neurotransmitter, and in
VOLUME 113, 1986
fact, the growth cones of Xenops spinal cord neurons are known to contain clear vesicles of 50-nm diam (Weldon and Cohen 1979). In previous reports (Young and Poo, 1983; Young, 1984) as well as the present one, some bursts were seen to consist of “staircase” openings. These are rapid initial openings of multiple channels followed by slower closure to baseline. This pattern would be expected if the patch pipette was positioned directly adjacent to a region of the growth cone membrane where the contents of a transmitter containing vesicle was released. At this level of examination, the burst release from the growth cone has the characteristics associated with the MEPP producing vesicular release of the neuromuscular junction. But how close is this apparent relationship? In mouse (Vyskocil et ah, 1983) and in Xenopus (Sun and Poo, 1985) neuromuscular junction, elevated external calcium levels increase MEPP frequency. The present study has shown that an increase in extracellular calcium or extracellular magnesium will decrease the spontaneous release of neurotransmitter from Xenopus growth cones (Fig. 4). Both bursts and singles frequencies are decreased. In this respect, the burst release does not behave like release which produces MEPPs at an established synapse. Singles Release from the Growth Cone Is Not Identical to the Leak Release from the Synapse The singles clusters are caused by a low-concentration (and sometimes long-lasting) release of transmitter. In addition, the frequency of singles clusters is greatly reduced in salines with increased calcium or magnesium concentrations. Transmitter leak at the neuromuscular junction is greatly reduced at high calcium concentrations (Vyskocil et al., 1983; Sun and Poo, 1985), and in this respect singles release is similar to leak release. At present there is no information on whether blockage of leak release can be accomplished by an increase of magnesium ions, preventing further detailed comparison of the two types of spontaneous release. However, the singles release is not identical to leak release. In Xenopus neuromuscular junction, reduction of calcium levels below 2 mM results in further increase of the amount of leak release, but in low-calcium saline, singles release from the growth cone occurs at about the same rate as in 2 mM calcium saline (Fig. 6). Singles and Burst Releases Are Not Correlated What might be the cause of singles clusters? One simple interpretation would be that the singles are caused by the same mechanism as the bursts, and that they are
low level (low frequency of openings) because the release point is far away from the growth cone. This would cause the concentration profile of the transmitter to become diffuse by the time it reaches the patch pipette, and produce series of low-frequency single-channel openings. However, a study of the distribution of release sites along the neurites shows that release is confined to the region of the growth cone (Young and Poo, 1983), ruling out this possibility. Another possibility is that the singles result from release of partially filled vesicles, and still another is that singles represent a periodic release of cytoplasmic transmitter (Marchbanks, 1968; Taut 1982), perhaps as a result of membrane addition during neurite growth. Whatever the causes, if both the bursts and the singles are the result of the same release process (filled and partially filled vesicles as only one example) the frequencies of release of both types should be correlated. In this instance, the expectation would be that high rates of burst release from a growth cone should be accompanied by high rates of singles. However, as shown in Fig. 3, no such correlation exists between bursts and singles, suggesting that they do not result from identical processes. Frequencies of Bursts and Singles Neurite Grmuth, Rate
Are Not Related to
It is possible that the transmitter release from the growth cone is incidental to neurite growth. Perhaps during neurite growth, (1) vesicles containing transmitter are fused with the membrane and release transmitter causing burst behavior, and (2) small amounts of cytoplasmic transmitter can also be released during membrane incorporation, causing the singles behavior. This model for spontaneous release predicts that under increased external calcium levels, the decrease in the rate of spontaneous release of transmitter should be reflected in decreased growth rate of the neurons. However, Pate1 and Poo (1982) showed that increasing external calcium from 0.4 to 10 mM does not change the rate of neurite growth of Xenopus neurons. In addition, Bixby and Spitzer (1984) report that Xenopus neurons in calcium-free media show increased rates of growth. In the present set of experiments decreasing the calcium levels does not increase the rates of release. Thus, the changes in the release rate observed here are not likely to be linked to changes in the growth rate of the neurite. SUMMARY
Results from the present study indicate that spontaneous release of transmitter from growth cones of isolated neurons, although occurring in clusters, does not
resemble the process which produces spontaneous MEPPs at the neuromuscular junction. Occasional release by transmitter-containing vesicles is not ruled out, in fact the presence of “staircase” events in 2 mM calcium is highly suggestive of such an event. The release from the growth cone is decreased at increased levels of external calcium-behavior which more closely resembles that of transmitter leak at the neuromuscular junction. However, the release from the growth cone is not identical to leak release, since exposure to low-calcium solution does not further increase the rate of release. Taken together, the results described above suggest that the growth cone does not possess the mature mechanism(s) for MEPP-producing release or leak releasethese components of spontaneous release must develop more fully after nerve-muscle contact. Finally, an interesting speculation is that the spontaneous release of neurotransmitters and possibly other substances from the growth cone performs a necessary function in synaptogenesis-perhaps in guidance of the nerve to the muscle and/or stabilization of the nervemuscle contact. In saline with 10 mM calcium, spontaneous release of transmitter is reduced to very low levels. This inhibition of release persists even after 1 day in culture (Fig. 7). Do synapses still form? Henderson et al. (1984) have shown that in the presence of 10 mM calcium, Xelzopus spinal cord neurons will form functional synaptic contacts. Since spontaneous release from the growth cone is greatly decreased in the 10 mM calcium saline, it is unlikely that release at the levels seen from growth cones in 2 mM calcium is necessary for normal synaptogenesis. Dr. I. Chow gave valuable criticism on the manuscript. This work was supported by a grant from the Muscular Dgstrophy Association of America, and National Science Foundation Grant BNS 85-1079.
REFERENCES B~suu, J. L., and QPITZER, N. C. (1984). Early differentiation of vertehrate spinal neurons in the absence of voltage dependent Ca2+ and Na’ influx. &I’. Biol. 106, 89-96. BKE~W, P., K~ILLBERC;, R., and MOODY-CORBETT, F. (1984). Properties of nonjunctional acetylcholine receptor channels on innervated muscle of Xwop~s lurcis. J. Physiol. 350, 631-648. r)EI, CASTIIU, J., and KATZ, B. (1954). Quanta1 components of the endplate potential. J. Ph?/.siol. 124, 560-573. GUHARAY, F., and SACHS, F. (1984). Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle. J. Phy,siol. 352, 685-701. HENDERSON, L. P., SMITEI, M. A., and SPITZEK, N. C. (1984). The formation of neuromuscular contacts between embryonic cells of Xw0~12~slnevis differentiating in culture in the absence of calcium. Sot. Neurosci. Abstr. 10, 924. HORN, R., and PATLAK, J. (1980). Single channel currents from excised
380
DEVELOPMENTAL BIOLOGY
patches of muscle membrane. Proc. Natl. Acad. Sci. USA 77,69306934. HUME, R. I., ROLE, L. W., and FISCHBACH, G. D. (1983). Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes. Nature 305,632-634. KATZ, B., and MILEDI, R. (1977). Transmitter leakage from motor nerve endings. Proc. R. Sot. Loruh Ser. B 196, 59-72. KATZ, B., and MILEDI, R. (1981). Does the motor nerve impulse evoke ‘non-quanta1 transmitter release? Proc. R. Sot. London Ser. B 212, 131-137. MARCHBANKS, R. M. (1968). Exchangeability of radioactive acetylcholine with the bound acetylcholine of synaptosomes and synaptic vesicles. Biochem. J. 106, 87-95. NIEUWKOOP, P. D., and FABER, J. (1967). “Normal Table of Xenoms km-is (Daudin),” 2nd ed. North-Holland, Amsterdam. PATEL, N., and Poo, M.-M. (1982). Orientation of neurite growth by extracellular electric fields. J. Neurosci. 2,483-496. POO, M.-M., POO, W.-J. H., and LAM, J. (1978). Lateral electrophoresis
VOLUME 113, 1986
and diffusion of concanavalin A receptors in the membrane of embryonic muscle cell. J. Cell Biol. 76, 483-501. SUN, Y.-A., and POO, M.-M. (1985). Non-quanta1 release of acetylcholine at a developing neuromuscular synapse in culture. J. Neurosci. 5, 634-642. TALJC, L. (1982). Nonvesicular release of neurotransmitter. Physiol. Rev. 62, 857-893. VYSKOCIL, F., NIKOLSKY, E., and EDWARDS, C. (1983). An analysis of the mechanisms underlying the non-quanta1 release of acetylcholine at the mouse neuromuscular junction. Neuroscience 9,429-435. WELDON, P. R., and COHEN, M. W. (1979). Development of synaptic ultrastructure at neuromuscular contacts in an amphibian cell culture system. J. Neurocytol. 8, 238-259. YOUNG, S. H., and POO, M.-M. (1983). Spontaneous release of transmitter from growth cones of embryonic neurons. Nature 305, 634-637. YOUNG, S. H. (1984). Release of acetylcholine from growing growth cones of cultured Xenyms neurons: The influence of external calcium. Sot. N~~rosci. Abstr. 10, 581.
BIOLOGY
113, 373-380 (1986)
Spontaneous Release of Transmitter from the Growth Cones of Xenopus Neurons in Vitro: The Influence of Ca*+ and Mg*+ Ions STEVEN
H. YOUNG’
To determine whether spontaneous release of transmitter from the growth cones of neurons exhibits properties similar to the spontaneous release which occurs from the neurons at the neuromuscular junction, release of transmitter from the growth cones of XenolnL.ncs neurons in culture was monitored in salines containing varying calcium and magnesium concentrations. Release was monitored by use of an outside-out piece of muscle membrane attached to a patch clamp electrode. Spontaneous release of transmitter from the growth cones in standard saline (2 m M CaCl,, 1 m M M&l,) produces clusters of single-channel openings in the muscle membrane. Clusters are seen to consist of two types: a series of high-frequency channel openings, called “bursts,” and clusters of low-frequency channel openings called “singles.” The bursts were identified and examined for their possible relationship to MEPP-producing release, and the singles were identified and examined for their possible relationship to “leak” release of the neuromuscular junction. When the external saline contains high calcium (10 mA4 CaCl,, 1 m M MgC&) or high magnesium (2 m M CaCl,, 9 m M Mg&), the frequencies of both “bursts” and “singles” was greatly reduced. This reduction in release persists if the neurons are gro\vn in the high-calcium or high-magnesium solutions. When the saline is a low-calcium solution (0 m M CaClz, 3 m M ME&) the growth cones release transmitter at rates similar to those from standard saline. These results indicate that although the spontaneous release from the growth cone shares one characteristic with the leak release, neither the burst nor the singles release from the growth cones share exact relationship with either the MEPP producing release or the leak release. This suggests that further development of the mechanisms for spontaneous release of neurotransmitter occurs after nerve-muscle contact. (c I%%Academic Press. Inc INTRODUCTION
release change with development of the synapse? A comparison of the relations between spontaneous release observed at the mature neuromuscular junction, and the spontaneous release from the pre-muscle-contact growth cone presents an opportunity to study development of transmitter release and possibly to elucidate the mechanism(s) of the spontaneous release of transmitter. In Xenopus (Sun and Poo, 1985) and in mouse (Vsykocil et ah, 1983) neuromuscular junction, the amount of leak release is decreased by an increase in external calcium concentration, while the frequency of the spontaneous MEPPs increases. Since these responses are in opposite directions, an increase of external calcium concentration then can be used as a means to separate leak-like release from MEPP-like release. The experiments reported here were performed to determine whether spontaneous release of transmitter from the growth cone exhibits similar sensitivity to calcium ions as those of either the MEPP- or leak-producing release which occur at the neuromuscular junction. To address this question, spontaneous release from the growth cones of Xenopus neurons in culture was monitored in salines containing varying calcium and magnesium concentrations. In an earlier report Young and Poo (1983) had shown that spontaneous release of transmitter from the growth cones of Xenopa neurons occurs
The growth cones of isolated neurons in culture will release transmitter spontaneously (Young and Poo, 1983), and in response to electrical stimulation (Hume et al., 1983). The spontaneous release of transmitter is of particular interest when considering synaptogenesis of the neuromuscular junction because such release presumably occurs throughout the period of growth of the neurite toward the muscle cell, and it is natural to speculate that such release could reflect release of factors which might influence the formation and stability of the neuromuscular junction. Furthermore, two forms of spontaneous release of transmitter are seen at the mature neuromuscular junction: a packaged, or quanta1 form, in which release occurs in bursts and produces the miniature endplate potential (MEPP) in the muscle cell (de1 Castillo and Katz, 1954), and a low-level “leak” of transmitter from the nerve terminal (Katz and Miledi, 1977; Katz and Miledi, 1981). Does the growth cone already possess the characteristics of spontaneous transmitter release seen at the mature synapse? Or do the properties of spontaneous ’ Present address: Jerry Lewis Neuromuscular Research UCLA School of Medicine, Los Angeles, Calif. 90024.
Center,
373
0012-1606/86 $3.00 Copyright All rights
0 1986 by Academic Press. Inc. of reproduction in any form reserved
374
DEVELOPMENTALBIOLOGY
in a manner which produces clusters of single-channel openings in the “probe,” the transmitter measuring system. In the present report, the characteristics of these clusters are examined in greater detail. Clusters were seen to consist of two types: a series of high-frequency channel openings, called “bursts,” and clusters of lowfrequency channel openings called “singles.” The bursts are identified and examined for their possible relationship to the MEPP-producing release, and the singles are identified and examined for their possible relationship to leak release. Results show that neither the burst nor the singles release from the growth cones share exact relationship with either the MEPP-producing release or the leak release of the neuromuscular junction, and suggest that further development of the mechanism(s) of spontaneous release of neurotransmitter must occur after nervemuscle contact. METHODS
Solutions. The standard external saline contained (in m&f) 125 NaCl, 2 KCl, 2 CaQ, 1 MgQ, 5 Hepes, at pH 7.2. All experimental salines differed only in the amounts of CaCle and/or MgCl, present. The standard internal saline (pipette solution) contained (in mM) 82 KCl, 35 KOH, 1 CaClp, 11 EGTA, 1 MgCl,, 5 Hepes, pH 7.2. Disaggregating saline contained (in mM) 125 NaCl, 2 KCl, 0.1 EDTA, 1 unit/ml penicillin-streptomycin, pH 7.2. Culture medium contained 85% saline, 10% L-15 (GIBCO), 5% fetal calf serum (ZETA-D, AMF Biological and Diagnostic Products Company), and 1 unit/ml penicillin-streptomycin. With the exception of two experiments which investigated effects of growth in high-calcium- or high-magnesium-containing media, all cultures were grown in media using standard saline. Preparation of nerve-muscle cell culture. Preparation of nerve-muscle cultures follows the procedure of Poo et al. (1978). Briefly, fertilized eggs of Xenopus laevis were allowed to develop until stages 17-20 (Nieuwkoop and Faber, 1967). The neural tube, notocord, and myotomal regions were removed from the embryos, and placed in disaggregating solution for 30 min. Following tissue disaggregation, the cells were plated onto clean glass slides which contained culture medium. After 1 day in culture, nerve cells sent out neurites, and most muscle cells remained spherical or nearly spherical (see Fig. 1). At 2-4 days in culture, most surviving nerve cells had contacted muscle cells. Therefore, to study release of transmitter from growth cones of nerve cells which have not previously contacted muscle cells, and to standardize the age of nerve cells, all cultures used in this study were 1 day old. Detection of transmitter release. Methods for detection of transmitter release at very high sensitivity follow
VOLUME 113, 1986
FIG. 1. Phase contrast photomicrograph of a Xenoyl~s spinal cord neuron (cell body marked N) in culture. Spherical muscle cells (marked MJ are attached to the bottom of the glass slide. The small phase bright particles are yolk granules. On the right side of the photograph, a patch pipette (P) with a piece of muscle membrane at its tip is moved toward the tip of a neurite to monitor release of transmitter from the growth cone. Final position of the pipette will be 3-6 pm from the growth cone. Scale bar: 60 pm.
those of Young and Poo (1983) and Hume et al. (1983). The central element of the method is the use of the excised membrane patch technique (Horn and Patlak, 1980), which can be used to resolve the opening of single ionic channels. When a piece of membrane containing receptor-ionic channels is formed across the tip of a glass pipette (~1 pm diam) with the outer surface of the membrane facing the bath, the pipette now becomes a very sensitive probe for molecules which activate the membrane receptor channels. For example, the membranes of 1 day-in-culture Xenopus muscle cells contain acetylcholine (ACh) receptors at a high enough density that ACh receptor channel openings can be detected at ACh concentrations as low as 30 nM. This concentration corresponds to about 20 molecules/pm”, and illustrates the high sensitivity of the method. In addition, because of the small size of the pipette tip, the probe can be manipulated into position very close to the growth cone membrane (within 3-5 pm). To form an outside-out patch, the fire polished tip of the glass pipette (which contains internal solution) is placed against a muscle cell membrane. Slight suction applied to the top of the pipette produces a very tight membrane-glass seal (>lO GQ). Further suction ruptures the small area of membrane under the pipette, and allows access to the cell interior, where the muscle cell resting potential is measured. Finally, when the pipette is pulled away from the muscle cell, a small piece of muscle membrane will form across the tip of the pipette, with the outer surface of the membrane facing the bath. Figure 1 shows such a
STEVEN
H. YOITNG
Trunsmitter
“transmitter probe” approaching the growth cone of a neurite. Experimental protocol. Glass slides containing l-dayold nerve and muscle cells were removed from a humidifying cabinet, and the culture medium was very slowly (150 pl/min at 250 yl total volume) exchanged for experimental saline. The solution change was performed slowly to minimize the possibility of induced transmitter release due to bulk flow of fluid across the growth cone. As a control for possible flow-induced effects, transmitter release has been observed in cultures before the exchange of culture medium for experimental saline, indicating that Ilow of solution is not the sole cause of release under experimental conditions. Furthermore, to minimize any possible effect of fluid tlow on release, recordings were never started earlier than 4 min after solution flow had stopped. The outside-out patch of muscle membrane was formed on the tip of the pipette, and the patch membrane potential was set to the value of the muscle cell resting potential (-60 to -90 mV). Then baseline (BATH) recordings were made of the activity of the ionic channels in the patch when the probe was at least 100 pm away from any portion of any neurite. Records in the bath were taken for 2-5 min, and then the probe was positioned to within 3-6 pm of the tip of a growth cone to monitor release from that area. Only nerve cells which clearly were not in contact with any muscle cell were used in this study (see Fig. 1). Analysis. Records of current from open ionic channels were recorded on tape by an FM tape recorder and played back at slower speed onto an oscillographic recorder (Gould 2200) for analysis. The previous study (Young and Poo, 1983) has shown that most of the single-channel openings are due to ACh receptor channels, although other channel types are present. In the present study, no separation of channel types before analysis has been performed. Two types of channel opening patterns, which reflect transmitter release patterns, are observed: a series of single-channel openings occurring at low frequency, termed “singles,” and a high-frequency cluster of channel openings termed “bursts.” A burst is defined as a series of channel openings lasting longer than 1 set, with no more than 2 set between openings. Burst frequency is measured directly from records of current obtained near the growth cone. No corrections to the measured frequencies are performed, since bursting is not observed in BATH recording. However, singles openings can be observed in BATH records. These openings could possibly result from at least three sources: (1) residual transmitter in the saline, (2) openings of non-ligand bound channels (Brehm et ab, 1984), or (3) from residual membrane tension in the
Rrletrse bg Growth
375
Cones
patch, occurring as a result of patch excision from the muscle, producing pressure-activated channels (Guharay and Sachs, 1984). The measured frequency of singles openings obtained near the growth cones were corrected by subtraction of singles frequencies present in BATH recordings. Singles frequencies can be highly variable, and the correction procedure can produce corrected singles frequencies from the growth cone which are negative. Average burst and corrected singles frequencies in experimental salines were compared to records obtained in the standard external saline using Student’s t test. RESULTS
Spontaneous
Release of Transmitter
in Standard
Saline
Typical appearance of ionic channel openings produced by spontaneous release of transmitter from two separate growth cones in standard saline are shown in Fig. 2. All growth cones which showed release of transmitter exhibited such release by causing groupings or clusters of channel openings. The beginning of the current record taken near growth cone 1 (GC 1) shows two bursts of channel openings (marked B) followed at a later time by two small clusters of singles (marked S) flanking a small burst (middle trace). The recording taken near growth cone 2 (GC 2) shows a long-lasting series of singles. The definition of bursts and singles is described in Methods. Whether a growth cone showed bursts or sin-
BATH -c*
$LIOS BATH -Gc~
FIG. 2. Spontaneous release of transmitter from two different growth cones in standard external saline as measured by the current in the patch pipette. Top traces: The current record marked BATH was taken with the pipette at least 100 pm away from any neuron. When the pipette was moved to within 3-6 pm of growth cone 1 (GC l), two bursts (marked B) of channel openings were recorded. Middle trace: Continuation of the record of GC 1. A small burst is flanked by two clusters of singles (marked S). Bottom traces: Bath recording from a different membrane patch, and current record from a second growth cone (GC 2) which showed a long duration cluster of singles (for definition of burst and singles see Methods).
376
DEVELOPMENTAL
u+vI=I 2 -8
I,, 4
SINGLES
, 28
20
12
FREQUENCY
BIOLOGY
(per
mini
FI(;. 3. No correlation exists between burst and singles frequencies. The average burst frequencies are plotted against the average singles frequencies from 13 growth cones in standard saline
gles clusters (or both) the frequency of occurrence of such events was not highly regular. Some growth cones showed predominantly bursting release, some showed both bursts and singles (GC 1, Fig. 2), and some showed predominantly singles release (GC 2, Fig. 2). It is of interest to know if bursts and singles frequencies are correlated, to help to determine if these different release patterns reflect different release processes. If bursts and singles are expressions of very similar processes, it might be expected that growth cones showing high rates of bursting would also show high singles frequencies. Figure 3 is a plot of average burst frequency vs average singles frequency from each of 13
VOLUME
113. 1986
growth cones recorded in standard saline. The figure shows that there is no correlation between bursts and singles frequencies. The grand average burst and singles frequencies from all 13 growth cones in standard saline are shown in Fig. 4. The left panel of Fig. 4 shows average frequencies (with standard error) of bursts (labeled GC) and singles. Burst and singles openings were detected in 7 out of 13 (54%)growth cones, a fraction similar to that found in the previous report (51%, Young and Poo, 1983). Note the large standard error of the singles frequency. Corrected singles frequencies varied over a large range, under all experimental conditions. The large standard errors which resulted decreased the levels of significance of difference between average frequencies of singles in standard saline and each of the experimental salines (see below), even when the fraction of growth cones exhibiting such release was reduced. The Eflect of Elevated External Calcium on Spontaneous Release of Transmitter Transmitter release from growth cones exposed to increased external calcium ion concentrations was monitored, and the results are plotted in Fig. 4. The center panel shows that average burst and singles frequencies were reduced when the external saline contained elevated external calcium concentrations (10 mM CaCl,). Only 2 out of 10 growth cones showed any bursts, and the average burst frequency was significantly decreased (P < 0.02) compared to standard saline. The average
2 mM Ca++
10 mM Ca++
2 mM Ca++
1 mM Mg++
1 mM Mg++
9 mM Mg++
7
P10.2 P10.02
BATH
BATH
GC
BURSTS
SINGLES
BURSTS
SINGLES
BURSTS
GC
SINGLES
FIG. 4. Average burst and singles frequencies in salines with high concentrations of calcium or magnesium were reduced below values in standard saline. Concentration of other ions remained unchanged. The top of each chart is labeled with the appropriate calcium and magnesium concentrations. The stippled bars illustrate average burst frequencies (with standard error) in the BATH and near the growth cone (GC). The open bars illustrate average corrected singles frequency (with standard error). The left axis of each chart is marked for bursts frequency, and the right axis is marked for singles frequency. The fractions of growth cones showing release are printed above each bar. Left panel: Results from standard saline. Center and right panels: Results from two experimental solutions. Each frequency was compared to the corresponding frequency in standard saline. Levels of significance are printed above each bar (see Methods).
377 corrected singles frequency was reduced, but a large standard error prevents a greatly significant difference from that of standard saline (P < 0.2). However, the fraction of growth cones showing release (2110) was reduced below that observed in standard saline (P < 0.04). The Effect of Elevated External Magnesium Spontaneous Release of Transmitter
on
To determine whether the reduction of spontaneous release is due specifically to calcium ions, a second set of experiments were performed in which CaClz was maintained at 2 mM, but MgClz was increased to 9 mM. In this manner, both experimental solutions contained 11 mM of divalent ions, with the calcium concentration of the second solution remaining equal to that of the standard saline. The results of these experiments are plotted in the right panel of Fig. 4. Both burst and corrected singles frequencies are reduced. Only 1 out of 7 growth cones showed bursts, and the resulting average burst frequency is lower than that in standard saline (P < 0.02). Average corrected singles frequencies were reduced, but a large standard error prevented a greatly significant difference from that in standard saline. The fraction of growth cones showing singles (28%)was reduced from that in the standard saline. Spontaneous Release of Transmitter Calcium Concentration
at Low External
To acertain whether release continues at low calcium concentrations, nerve cells were bathed in external saline that did not contain CaC&, and with MgClp increased to 3 mM (in order that total divalents remained at 3 mM, the same value as in standard saline). The experimental design is not intended to test for an absolute requirement of spontaneous release upon calcium, but to test if release is affected at calcium levels far below the l-2 mM which is present in most physiological salines. Results from one growth cone are shown in Fig. 5, and illustrate that spontaneous transmitter release continues to occur at low levels of calcium. The outside-out patches were far
BATH -r
.d 10
mg
“-M”
FIN:. 5. Spontaneous release of transmitter from a growth cone in low-calcium solution. Current records from the BATH and near the growth cone (GC) of a neuron bathed in saline without calcium (0 m M CaC12, 3 m M M&l,).
0 mM Ca++ 3 mM Mg++
.L E
P20.4
3
k a &
3/4 2
I5 2 w [r LL
’
l&?&lid
P10.95 3/4 1 / 3 ii$+1/i!; /iiii/i/ ai! ,, ,x1.: @##j BATH GC
BURSTS
SINGLES
FIN;. 6. Frequencies of spontaneous release of transmitter from growth cones in low-calcium solution are not significantly different from those in standard saline. Chart markings as in Fig. 4.
less stable in the low-calcium saline, tending to break easily. Combined data from 4 growth cones using 3 patches which remained stable for at least 4 min are shown in Fig. 6. One of the patches showed a single burst in the bath. Bursting release was observed at 3 of 4 growth cones tested, and the average burst frequency was not significantly different from standard saline (P < 0.95). Singles were seen in 3 out of 4 growth cones tested. Spontaneous Release of Transmitter upon Continuous Exposure to High-Calcium or -Magnesium Solutions It is possible that whatever process(es) are involved in the spontaneous release of transmitter from the growth cones, they can “adapt” in time to maintained increased calcium or magnesium levels, and eventually show release rates similar to those obtained in standard saline. To test this possibility, nerve cells were grown in culture media containing elevated calcium or magnesium levels, and then placed in recording salines with the same ionic concentrations. The left panel of Fig. ‘7 shows results from experiments in which culture medium containing 10 mM CaClz and 1 mM MgClz was exchanged for saline also containing 10 mM CaClz and 1 mM MgC12 before the start of recording. The panel shows that both bursts and singles frequencies remain below those in standard saline-the spontaneous release does not adapt (after 1 day in culture) to maintained elevated calcium levels. The right panel of Fig. 7 shows results from experiments in which the cultures were maintained in medium containing 2 mM CaClz and 9 mM MgC12. Again, the frequencies and fractions of growth cones showing release remain below those recorded in stan-
378
DEVELOPMENTAL BIOLOGY .E E
z f
mM
Ca++
2
mM
Cat+
1 mM
Mg++
9
mM
Mgtf
10
r4
2l
BURSTS
2i
SINGLES
r4
BURSTS
SINGLES
FIG. ‘7. Growth cones maintained in high-calcium or high-magnesium solutions do not “adapt” and show rates of spontaneous release of transmitter similar to those obtained in standard saline. Neurons were cultured in high-calciumor high-magnesium-containing media for 1 day before recording spontaneous release. Chart markings as in Fig. 4.
dard saline. The spontaneous release cannot maintained elevated magnesium levels.
adapt
to
DISCUSSION
In the presence of 2 mMcalcium, growth cones of Xenopus spinal cord neurons kept in culture can spontaneously release neurotransmitter. The release is not steady, but occurs in a manner which produces clusters of channel openings in the outside-out muscle membrane containing patch pipette. These clusters can be separated into two classes: a series of high-frequency channel openings, called bursts, and clusters of low-frequency openings called singles. The average frequencies of both bursts and singles are reduced when the external saline contains high calcium (10 mM). The reduction in spontaneous release from the growth cone is not specific to the calcium ion-an increase in magnesium ion concentration (9 mM) will also reduce spontaneous release. The appearance of bursts and singles is not dependent upon the presence of millimolar levels of calcium. When MgCl, is substituted for CaC12 in the standard saline, transmitter continues to be released from the growth cones, with average burst and singles frequencies not significantly different from those in standard saline. Burst Release from the Growth Cone Is Not Identical MEPP Producing Release at the Synapse
to
The burst openings are caused by the immediate release of a concentrated amount of transmitter near the patch pipette. This is the type of behavior which might be expected to occur with the release of the contents of a synaptic vesicle filled with neurotransmitter, and in
VOLUME 113, 1986
fact, the growth cones of Xenops spinal cord neurons are known to contain clear vesicles of 50-nm diam (Weldon and Cohen 1979). In previous reports (Young and Poo, 1983; Young, 1984) as well as the present one, some bursts were seen to consist of “staircase” openings. These are rapid initial openings of multiple channels followed by slower closure to baseline. This pattern would be expected if the patch pipette was positioned directly adjacent to a region of the growth cone membrane where the contents of a transmitter containing vesicle was released. At this level of examination, the burst release from the growth cone has the characteristics associated with the MEPP producing vesicular release of the neuromuscular junction. But how close is this apparent relationship? In mouse (Vyskocil et ah, 1983) and in Xenopus (Sun and Poo, 1985) neuromuscular junction, elevated external calcium levels increase MEPP frequency. The present study has shown that an increase in extracellular calcium or extracellular magnesium will decrease the spontaneous release of neurotransmitter from Xenopus growth cones (Fig. 4). Both bursts and singles frequencies are decreased. In this respect, the burst release does not behave like release which produces MEPPs at an established synapse. Singles Release from the Growth Cone Is Not Identical to the Leak Release from the Synapse The singles clusters are caused by a low-concentration (and sometimes long-lasting) release of transmitter. In addition, the frequency of singles clusters is greatly reduced in salines with increased calcium or magnesium concentrations. Transmitter leak at the neuromuscular junction is greatly reduced at high calcium concentrations (Vyskocil et al., 1983; Sun and Poo, 1985), and in this respect singles release is similar to leak release. At present there is no information on whether blockage of leak release can be accomplished by an increase of magnesium ions, preventing further detailed comparison of the two types of spontaneous release. However, the singles release is not identical to leak release. In Xenopus neuromuscular junction, reduction of calcium levels below 2 mM results in further increase of the amount of leak release, but in low-calcium saline, singles release from the growth cone occurs at about the same rate as in 2 mM calcium saline (Fig. 6). Singles and Burst Releases Are Not Correlated What might be the cause of singles clusters? One simple interpretation would be that the singles are caused by the same mechanism as the bursts, and that they are
low level (low frequency of openings) because the release point is far away from the growth cone. This would cause the concentration profile of the transmitter to become diffuse by the time it reaches the patch pipette, and produce series of low-frequency single-channel openings. However, a study of the distribution of release sites along the neurites shows that release is confined to the region of the growth cone (Young and Poo, 1983), ruling out this possibility. Another possibility is that the singles result from release of partially filled vesicles, and still another is that singles represent a periodic release of cytoplasmic transmitter (Marchbanks, 1968; Taut 1982), perhaps as a result of membrane addition during neurite growth. Whatever the causes, if both the bursts and the singles are the result of the same release process (filled and partially filled vesicles as only one example) the frequencies of release of both types should be correlated. In this instance, the expectation would be that high rates of burst release from a growth cone should be accompanied by high rates of singles. However, as shown in Fig. 3, no such correlation exists between bursts and singles, suggesting that they do not result from identical processes. Frequencies of Bursts and Singles Neurite Grmuth, Rate
Are Not Related to
It is possible that the transmitter release from the growth cone is incidental to neurite growth. Perhaps during neurite growth, (1) vesicles containing transmitter are fused with the membrane and release transmitter causing burst behavior, and (2) small amounts of cytoplasmic transmitter can also be released during membrane incorporation, causing the singles behavior. This model for spontaneous release predicts that under increased external calcium levels, the decrease in the rate of spontaneous release of transmitter should be reflected in decreased growth rate of the neurons. However, Pate1 and Poo (1982) showed that increasing external calcium from 0.4 to 10 mM does not change the rate of neurite growth of Xenopus neurons. In addition, Bixby and Spitzer (1984) report that Xenopus neurons in calcium-free media show increased rates of growth. In the present set of experiments decreasing the calcium levels does not increase the rates of release. Thus, the changes in the release rate observed here are not likely to be linked to changes in the growth rate of the neurite. SUMMARY
Results from the present study indicate that spontaneous release of transmitter from growth cones of isolated neurons, although occurring in clusters, does not
resemble the process which produces spontaneous MEPPs at the neuromuscular junction. Occasional release by transmitter-containing vesicles is not ruled out, in fact the presence of “staircase” events in 2 mM calcium is highly suggestive of such an event. The release from the growth cone is decreased at increased levels of external calcium-behavior which more closely resembles that of transmitter leak at the neuromuscular junction. However, the release from the growth cone is not identical to leak release, since exposure to low-calcium solution does not further increase the rate of release. Taken together, the results described above suggest that the growth cone does not possess the mature mechanism(s) for MEPP-producing release or leak releasethese components of spontaneous release must develop more fully after nerve-muscle contact. Finally, an interesting speculation is that the spontaneous release of neurotransmitters and possibly other substances from the growth cone performs a necessary function in synaptogenesis-perhaps in guidance of the nerve to the muscle and/or stabilization of the nervemuscle contact. In saline with 10 mM calcium, spontaneous release of transmitter is reduced to very low levels. This inhibition of release persists even after 1 day in culture (Fig. 7). Do synapses still form? Henderson et al. (1984) have shown that in the presence of 10 mM calcium, Xelzopus spinal cord neurons will form functional synaptic contacts. Since spontaneous release from the growth cone is greatly decreased in the 10 mM calcium saline, it is unlikely that release at the levels seen from growth cones in 2 mM calcium is necessary for normal synaptogenesis. Dr. I. Chow gave valuable criticism on the manuscript. This work was supported by a grant from the Muscular Dgstrophy Association of America, and National Science Foundation Grant BNS 85-1079.
REFERENCES B~suu, J. L., and QPITZER, N. C. (1984). Early differentiation of vertehrate spinal neurons in the absence of voltage dependent Ca2+ and Na’ influx. &I’. Biol. 106, 89-96. BKE~W, P., K~ILLBERC;, R., and MOODY-CORBETT, F. (1984). Properties of nonjunctional acetylcholine receptor channels on innervated muscle of Xwop~s lurcis. J. Physiol. 350, 631-648. r)EI, CASTIIU, J., and KATZ, B. (1954). Quanta1 components of the endplate potential. J. Ph?/.siol. 124, 560-573. GUHARAY, F., and SACHS, F. (1984). Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle. J. Phy,siol. 352, 685-701. HENDERSON, L. P., SMITEI, M. A., and SPITZEK, N. C. (1984). The formation of neuromuscular contacts between embryonic cells of Xw0~12~slnevis differentiating in culture in the absence of calcium. Sot. Neurosci. Abstr. 10, 924. HORN, R., and PATLAK, J. (1980). Single channel currents from excised
380
DEVELOPMENTAL BIOLOGY
patches of muscle membrane. Proc. Natl. Acad. Sci. USA 77,69306934. HUME, R. I., ROLE, L. W., and FISCHBACH, G. D. (1983). Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes. Nature 305,632-634. KATZ, B., and MILEDI, R. (1977). Transmitter leakage from motor nerve endings. Proc. R. Sot. Loruh Ser. B 196, 59-72. KATZ, B., and MILEDI, R. (1981). Does the motor nerve impulse evoke ‘non-quanta1 transmitter release? Proc. R. Sot. London Ser. B 212, 131-137. MARCHBANKS, R. M. (1968). Exchangeability of radioactive acetylcholine with the bound acetylcholine of synaptosomes and synaptic vesicles. Biochem. J. 106, 87-95. NIEUWKOOP, P. D., and FABER, J. (1967). “Normal Table of Xenoms km-is (Daudin),” 2nd ed. North-Holland, Amsterdam. PATEL, N., and Poo, M.-M. (1982). Orientation of neurite growth by extracellular electric fields. J. Neurosci. 2,483-496. POO, M.-M., POO, W.-J. H., and LAM, J. (1978). Lateral electrophoresis
VOLUME 113, 1986
and diffusion of concanavalin A receptors in the membrane of embryonic muscle cell. J. Cell Biol. 76, 483-501. SUN, Y.-A., and POO, M.-M. (1985). Non-quanta1 release of acetylcholine at a developing neuromuscular synapse in culture. J. Neurosci. 5, 634-642. TALJC, L. (1982). Nonvesicular release of neurotransmitter. Physiol. Rev. 62, 857-893. VYSKOCIL, F., NIKOLSKY, E., and EDWARDS, C. (1983). An analysis of the mechanisms underlying the non-quanta1 release of acetylcholine at the mouse neuromuscular junction. Neuroscience 9,429-435. WELDON, P. R., and COHEN, M. W. (1979). Development of synaptic ultrastructure at neuromuscular contacts in an amphibian cell culture system. J. Neurocytol. 8, 238-259. YOUNG, S. H., and POO, M.-M. (1983). Spontaneous release of transmitter from growth cones of embryonic neurons. Nature 305, 634-637. YOUNG, S. H. (1984). Release of acetylcholine from growing growth cones of cultured Xenyms neurons: The influence of external calcium. Sot. N~~rosci. Abstr. 10, 581.