| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departamento de Neurociencias, Centro de Investigación Médica Aplicada, Universidad de Navarra, Pamplona, Spain
Submitted 14 December 2007; accepted in final form 15 January 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; Chung et al. 2002; Cook et al. 2003; Markram et al. 1998; Nadim et al. 1999; Silberberg et al. 2004). Almost all synapses exhibit some form of short-term plasticity with some types of synapses mostly strengthening when used repetitively and others mostly weakening—i.e., short-term enhancement and short-term depression. However, the same synapses can exhibit both forms; most central synapses eventually depress when used extensively, but many of these first strengthen transiently at the beginning of a repetitive train (Zucker and Regehr 2002).
Short-term enhancement has been kinetically subdivided into several elemental processes, and at least some of these are mechanistically distinct from factors involved in depression. That is, some types of short-term depression are thought to be caused by depletion of a small, readily releasable pool (RRP) of primed vesicles in presynaptic terminals that are directly available to be triggered for release by action potentials (Zucker and Regehr 2002). In contrast, most elements of enhancement are driven by low levels of residual Ca2+ that accumulate presynaptically during action potential firing, and, at least one element, previously termed augmentation, is known to arise from the potentiated efficacy of the release machinery itself; i.e., augmented synapses are stronger because the probability of release per available vesicle within the RRP (PVES) is increased and not because of changes in the number of release-ready vesicles (Magleby and Zengel 1976b; Stevens and Wesseling 1999a; see also Rosenmund et al. 2002). Thus residual Ca2+ may continue to play a role in determining synaptic strength even after synapses have become depressed due to RRP depletion.
However, enhancement has typically been studied in isolation from depression with manipulations that limit coincident depletion of the RRP (Feng 1941; Hubbard 1963; Liley and North 1953; Magleby and Zengel 1976a), and the functional independence of enhancement from mechanisms of depression has never been tested thoroughly. Indeed it has been suggested that newly recruited readily releasable vesicles are primed for release gradually, at least at Calyx of Held synapses, which causes a reduction in PVES that contributes to depression (Wu and Borst 1999), and this additional mechanism might block, dampen, or simply subtract from, the enhancement of PVES that would otherwise be expected in the presence of residual Ca2+. While one report has shown that residual Ca2+ can enhance release parameters in the presence of masking depression at neuromuscular junctions (Kalkstein and Magleby 2004), two studies of excitatory hippocampal synapses (Brager et al. 2003; Klyachko and Stevens 2006) provided estimates of the timing of decay of enhancement under standard conditions that were up to fivefold slower than under low probability of release conditions (Fisher et al. 1997; Stevens and Wesseling 1999a) and also substantially slower than the clearance of residual Ca2+ (Brager et al. 2003). As the mismatch between the decay of enhancement and residual Ca2+ clearance would be a predicted consequence of convolving gradual vesicle priming with enhancement, these findings together suggest that gradual vesicle priming may play a substantial role in determining the timing of recovery from synaptic depression at a range of synapse types, from large Calyces of Held to small excitatory synapses of the hippocampus. An open question, then, is: what sort of impact do classically defined elements of short-term enhancement, such as augmentation, have at times when synapses are still recovering after the induction of short-term depression mechanisms that possibly include both RRP depletion and gradual/incomplete priming of the newly recruited vesicles that have become available?
We report here that residual Ca2+-dependent enhancement of release plays a substantial role during the first several seconds of recovery from depression at hippocampal synapses and, surprisingly, that the impact of gradual vesicle priming appears to be small or nonexistent during recovery intervals lasting 
1 s. The impact of enhancement during recovery intervals 1) can be large, as much as 4-fold in this study, 2) persists for several seconds, 3) tracks residual Ca2+, and 4) is likely due to the same residual Ca2+ dependent mechanism that underlies augmentation as defined under low probability of release conditions; the enhancement of PVES for newly recruited vesicles exhibits key kinetic and residual Ca2+ dependence features that fulfill the original criteria for augmentation (Delaney and Tank 1994; Magleby and Zengel 1976a; Zengel and Magleby 1982). These conclusions differ in some respects with previous reports, but differences can be attributed to differences in methods of analysis as will be further described in the DISCUSSION.
Thus our data show that low-level residual Ca2+ that is cleared slowly from presynaptic terminals following repetitive use seems to have a general, enhancing effect on neurotransmitter release that is mechanistically independent of the state of RRP depletion and replenishment, at least at excitatory synapses of the hippocampus. Phenomenologically, however, most of the enhancement at synapses that are recovering from depression is not observed as an increase in synaptic strength above baseline values because of masking by concurrent RRP depletion. Under standard conditions, residual Ca2+ instead causes a dynamic acceleration of the timing of the release events evoked during bursts of stimulation, which is manifest 1) as a quick rebound from depression in the probability of release (after single pulses) that outpaces the time course of RRP recovery, but also 2) as the faster onset of depression during a 2nd burst initiated within a few seconds of the 1st one, as expected because of the faster disbursement of an only partially full RRP. These observations have important implications for theoretical analyses of how dynamic synaptic behavior is used to support biological computation and also key technical implications concerning experiments designed for probing the cell biological mechanisms underlying short-term plasticity.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
. Slice preparation
Transverse slices (400 µm thick) were prepared from the hippocampi of 2- to 3-wk-old mice as described in detail previously (Wesseling and Lo 2002). Area CA3 was removed for all experiments. The extracellular recording solution contained (in mM) 120 NaCl, 1.25 NaH2PO4, 26 NaHCO3, 10 glucose, 3.5 KCl, picrotoxin (50 µM), and D-2-amino-5-phosphonovaleric acid D(–) APV (50 µM) and was continuously bubbled with a mixture of 95% O2-5% CO2 starting 20 min prior to addition of CaCl2 (usually 2.6 mM, except as indicated) and MgCl2 (usually 1.3 mM). Recording pipettes were filled with a solution containing (in mM) 130 Cs-gluconate, 5 CsCl, 5 NaCl, 2 MgCl2, 2 MgATP, 0.2 LiGTP, 1 EGTA, 0.2 CaCl2, and 10 HEPES, adjusted to a pH of 7.2, and an osmolarity of 290 mosM. Excitatory postsynaptic currents were evoked in CA1 pyramidal neurons with constant current pulses of <250 µA (almost always <75 µA) for 
100 µs via a monopolar silver/silver chloride electrode inserted into a glass pipette (tip diameter between 20 and 40 µm), filled with recording solution, and placed toward the CA3 end of the stratum radiatum.
Cell culture preparation
Isolated neurons were grown in cell culture essentially as described previously (Stevens and Wesseling 1998). Glass cover slips were coated with agarose (0.15%, type IIa) and allowed to dry before application of a particulate mist of substrate solution (1 mg/ml rat tail collagen, and 0.05 mg/ml poly-D-lysine) with an airbrush. Tissue from newborn mouse hippocampi was dissociated with papain (15 units/ml, Worthington) dissolved in minimum essential media (MEM, Gibco, no glutamine) for 30 min at 37°C and then washed with culture media containing 10% horse serum in MEM, penicillin (100 units/ml), streptomycin (100 µg/ml), sodium pyruvate (1 mM), and N-2 supplement (1x) from Gibco. Tissue was then triturated mechanically with a pipette. Cells were plated at 
3 x 104 cells (0.5 ml) per well in 24-well cell culture plates and were grown for 10 or 11 days in a 5% CO2 incubator (37°C) before use.
For cell culture experiments, the extracellular recording solution contained (in mM) 132 NaCl, 2 KCl, 10 glucose, 15 sorbitol, 10 HEPES, D(–)APV (50 µM), and CaCl2 and MgCl2 as indicated. Recording pipettes were filled with (in mM) 140 K-gluconate, 9 NaCl, 1 MgCl2, 2 MgATP, 0.2 LiGTP, 1 EGTA, 0.2 CaCl2, and 10 HEPES. Both solutions were adjusted to a pH of 7.2, and an osmolarity of 295–305 mosM. Uncontrolled action potentials were evoked presynapticaly by transiently depolarizing cell bodies as described in Bekkers and Stevens (1991). Osmotic shocks were induced with hypertonic solution, consisting of extracellular saline with 1.6 mM CaCl2, 1.6 mM MgCl2, and 500 mM sucrose, which was applied by picrospritzing from a glass pipette with a tip diameter of between 2 and 3 µm and was cleared rapidly with a vacuum pipette with a tip diameter of 10–50 µm.
Experimental design
In general, it was often possible to repeat several trials of each experiment on individual preparations. To allow the synapses to recover completely between trials, 4 min were allowed for rest before stimulation was initiated for each trial in the slice preparation (3 min for the near-minimal stimulation experiments), and 1 min in the culture preparation. For the experiments with a single experimental variable, the experimental and control trials were alternated. For time courses, the order of trials was shuffled.
Access resistance was monitored with –10-mV voltage steps before and after each trial, and data were only accepted if no change was detected during each trial, and also between trials for Figs. 2C and 6, A and B, and Supplementary Fig. S1.1 For Fig. 2C, experimental trials were interleaved with matched control trials where the stimulation frequency was maintained at 20 Hz throughout, and were repeated at least three times before adding DNQX; identical trials were averaged before further analysis.
|
|
Except as noted, synaptic responses were measured as the current integral. Rise times and peak current measurements gave similar results for the slice recordings (Wesseling and Lo 2002) but not always for the recordings obtained in cell culture because of activity dependent changes in asynchronous release in cell culture that do not seem to play a role in the slice preparation (Hagler and Goda 2001; Otsu et al. 2004).
RRP recovery time courses
RRP recovery values were estimated from paired trains of responses (e.g., diagram above Fig. 2A) in three steps. 1) An initial value was obtained by dividing the sum of the sizes of the 1st 60 responses during the 2nd train by the sum of the 1st 60 responses during the 1st train (in many cases the 1st and 2nd trains were longer than 60 pulses, but only the 1st 60 responses were needed for this analysis). The initial value is thought to be linearly related to the amount of RRP recovery occurring during the inter train interval, but would not be expected to equal 0 at times when RRP was empty, because of continued recruitment and release of new vesicles during stimulation (Wesseling and Lo 2002). Therefore, 2) the initial value for each recovery interval was adjusted by subtracting the initial value obtained from trials conducted with no rest interval between trains. 3) The adjusted value was subsequently normalized to vary from 0 to 1 by dividing by the range (i.e., 1 minus the initial value obtained for the no rest interval trials).
Near-minimal stimulation
Near-minimal stimulator settings were determined during low-frequency stimulation as the intensity needed to elicit successful synaptic transmission less than half of the time. To limit the likelihood of failures arising from axonal threshold fluctuations, stimulus settings were used only when it was possible to both increase and decrease the intensity by 2.5% without noticeably changing the probability of release (Allen and Stevens 1994; Raastad et al. 1992; see Fig. 1B of Wesseling and Lo 2002 for control experiments showing that axon firing threshold is stable with repetitive use).
Ca2+ imaging
Ca2+ indicator dye was prepared and loaded into slices essentially as described by (Wu and Saggau 1994). Dye (50 µg) was dissolved in 5 µl DMSO containing 20% Pluronic F-127 (Sigma), vortexed, diluted with 45 µl extracellular solution, which sometimes contained Alexa Fluor 488 hydrazide (100 µM, used for monitoring the loading procedure), and filtered at 0.22 µm. All dyes were from Molecular Probes. Dye solution was loaded into the tip of a glass pipette (tip diameter = 2.5–3 µM) and pressure injected into the s. radiatum with a picospritzer for 5–15 min; care was taken to use low injection pressures to avoid obvious damage to the slice. The stimulation pipette was placed near the injection site within the dye spot (radius: 75 µM), which was clearly visible (dark) with halogen light illumination immediately after injection (x5 objective). Subsequently, a x40 objective was aligned above a region of the s. radiatum 250 µM from the injection site and 100 µM from the edge of the original dye spot.
Optical recording was initiated after waiting 1 h for dye transport down the axons. Fluorescence changes were monitored at 5 or 1 Hz (indicated in figure legends) by switching on and off one of two discrete light-emitting diodes (LEDs; exposure length of 30 or 100 ms, LEDtronics, either pure blue BP280CWPB3K-3.6VF-050T, or ultraviolet BP200CUV750-250). Excitation and emission light was filtered with standard filter sets purchased from Omega Optical (either XF100-2 for Fluo4 or XF04-2 equipped with a 380 band-pass excitation filter). After passing through the objective and tube lens within the microscope frame (Olympus BX 51), emission light proceeded through an iris diaphragm, which was used to exclude light from areas outside of the s. radiatum (imaged area diameter was 150 µM); a plano-convex singlet lens (F = 50 mm) was then used to project this spatially restricted image onto a silicon avalanche photodiode (Perkin Elmer, C30902E).
At least three (Fluo 4) or 15 (Furaptra) trials were performed for each experiment and averaged before further analysis. Linear rundown in fluorescence signal due to photo-bleaching was nominally subtracted before analysis, but various combinations of neutral density filters were placed in the excitation path to ensure that this was negligible in all cases (see insets of Figs. 2D, 4C, and 5C). In all experiments, a permanent, activity dependent change of 0.5% was detected in steady-state fluorescence levels—possibly due to dye extrusion—but this was only readily noticeable in the Furaptra experiments where stimulation produced an extremely weak signal (see Fig. 2D). Additional experimental controls (not shown) showed that the amplitude but not the decay time course of the fluorescence signal depended on stimulation strength; i.e., this was expected because stronger stimuli elicit action potentials in more axons. No attempt was made to take advantage of the ratiometric properties of Furaptra for determining Ca2+ concentrations because this would require spatial heterogeneity in Ca2+ levels, and this condition was not met because stimulation would activate only a minor fraction of dye loaded terminals.
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
60 pulses, the response to the first pulse of the second train, initiated after a 2-s rest interval, was nearly as large as the response to the first pulse of the first train (Fig. 1Ai, inset). In contrast, the aggregate response during the entire second train was substantially smaller than the aggregate response during the first train (Fig. 1Ai) because of the more rapid induction of depression during the second train; it took 43 pulses during the first train to drive the synapses to within 10% of steady state but only 26 pulses during the second train (Fig. 1Aii).
|
10 trials per recording), while the total number of neurotransmission successes during the second train was significantly less than the total number during the first train (14.3 ± 1.0 vs. 25.6 ± 1.3, n = 47 trials). A 
2 test confirmed that the probability of release in response to the first pulse of the second train recovered significantly more during the 2-s rest interval than did recovery of the number of successes during the entire second train (P < 0.05). The mean amplitudes of successes after the second to sixth pulses of the first train were larger than the mean amplitude at later times during the first train and during the entire second train, suggesting that the number of successes during the first train was likely an underestimate of the number of quanta released owing to the simultaneous release of multiple quanta (see Fig. 1Bii); correction for the underestimate was not necessary for the 2 test, however, because the simultaneous release of multiple quanta was likely not a factor after the first pulse of the first train (see legend of Fig. 1Bii).
We interpret the incomplete recovery of the aggregate response, or total number of neurotransmission successes, during the second train as an indication that a RRP of presynaptic vesicles was emptied during the first train and only had enough time to recover partially during the 2-s rest interval. Evidence for this was published in an earlier report (Wesseling and Lo 2002), and includes key control experiments showing 1) that dynamic changes in synaptic strength are presynaptic in this preparation when long-term changes are prevented by blocking NMDA receptors, 2) that higher frequency stimulation, or higher extracellular Ca2+ levels, dramatically alter the timing of release events during the trains, but the only changes seen in the total amount of release for trains consisting of at least 60 pulses are predicted second order effects owing to recruitment of new vesicles, and 3) that the rate of transmitter release is no longer sensitive to the frequency of stimulation after the 1st 60 pulses as long as stimulation frequency is maintained at 20 Hz or above (see also Fig. 2C, below).
Indeed the incomplete recovery during the 2-s rest interval was expected from a series of earlier studies detailing a RRP recovery time course taking tens of seconds to run to completion when measured with independent techniques, including when using Ca2+-independent osmotic shocks instead of electrical stimulation to trigger release (Rosenmund and Stevens 1996; Stevens and Tsujimoto 1995; Stevens and Wesseling 1998; Wesseling and Lo 2002).
Thus for the still depleted synapses to release enough neurotransmitter to generate a fully recovered response to single pulses, PVES must have been increased because the probability of release per synapse had completely recovered even though there were fewer readily releasable vesicles (see Fig. 1, Ai and Bi). The increase in PVES would likely pertain to vesicles that had been recruited to the RRP during the previous 2 s because trains of 60 pulses (at 20 Hz or faster) are long enough to leave the RRP in a near-empty steady state (Wesseling and Lo 2002).
The increase in PVES is opposite what would be expected if individual, newly recruited vesicles were primed for release gradually over time intervals lasting 2 s. But it is consistent with what would be expected if the residual Ca2+ remaining in synaptic terminals after the first train drove an increase in PVES for the new vesicles that were recruited to the RRP during the interval between trains. Indeed, Ca2+ imaging experiments indicated that a small amount of residual Ca2+ was still present in axons 2 s after stimulation (Fig. 1C). The increase in PVES would also explain why response size depressed more quickly during the second train because each pulse would elicit the release of a higher fraction of the available vesicles within the RRP, which would thus be expended sooner.
Relaxation: transient time course for increased PVES
To determine how long PVES remains increased above baseline during rest, the interval between trains was varied experimentally in a similar series of experiments, as diagrammed at the top of Fig. 2. The relative recovery exhibited by the response to the first pulse of the second train (Fig. 2A, circles) was calculated by dividing the size by the size of the response to the first pulse of the first train, while RRP recovery (Fig. 2A, triangles) was estimated as in Wesseling and Lo (2002); the procedure is also described in METHODS. The increase in PVES remaining after each rest interval was then quantified by dividing the recovery exhibited by the response to the first pulse of the second train by the RRP recovery (Fig. 2B, note that this calculation yields the relative change in PVES with respect to baseline, not the absolute value). The analysis showed that the increase in PVES was transient, returning to baseline with a time course that could be approximated with the same single-exponential function (
= 7 s) that also characterizes the decay of augmentation as defined previously (Magleby and Zengel 1976a; Stevens and Wesseling 1999a).
Confirmation of RRP exhaustion after 80 pulses
To confirm RRP exhaustion after 80 pulses under our standard conditions, additional experiments were conducted where the stimulation frequency was doubled to 40 Hz after 80 pulses at 20 Hz (Fig. 2C); additionally, identical matched trials were performed in 10 µM DNQX, which, in combination with the standard NMDA receptor blockers, nearly completely blocks glutamate receptor responses at Schaffer collateral synapses. The complete time-integrated synaptic response was then extracted from these data by binning the entire digitized traces (100-ms bins) after first subtracting the corresponding digitized traces recorded in DNQX. Under standard conditions, no change was detected in the time-integrated synaptic response [Fig. 2Ci, the complete response integral over the first 1 s after doubling the stimulation frequency was 99 ± 9% (n = 4) compared with matched controls where the stimulation frequency was maintained at 20 Hz throughout]. In contrast, a robust increase in the time-integrated synaptic response was detected in additional experiments conducted, as a positive control at a highly reduced extracellular Ca2+ level where PVES is lowered to the point where 20-Hz stimulation is no longer sufficient to empty the RRP [Fig. 2Cii, under highly reduced Ca2+, low PVES conditions, doubling the stimulation frequency increased the response integral to 168% ±5% (n = 3) of matched controls]. Together these experiments confirm that 80 pulses at 20 Hz leave the RRP in a nearly empty state under standard conditions because otherwise increasing the stimulation frequency would have increased the rate of neurotransmitter release and thus the aggregate postsynaptic response. Thus we reason that the increase in PVES apparent in Fig. 2, A and B, pertains to vesicles that were recruited to the RRP during the short rest intervals allowed between the paired trains of stimulation.
Residual Ca2+ clearance matches relaxation of PVES
As a first test of a working hypothesis that the transient increase in PVES is caused by the same residual Ca2+-dependent mechanism that underlies augmentation induced under other circumstances, fluorescence-imaging experiments were conducted to determine the time course of residual Ca2+ clearance from Schaffer collaterals after similar repetitive trains. In separate experiments, time courses were measured with either Fluo-4 or furaptra to control for possible artifacts associated with indicator dye saturation (Sabatini and Regehr 1998) or interference with the natural time course of Ca2+ clearance owing to Ca2+ buffering by the indicators; these dyes have more than a 100-fold difference in affinity for Ca2+ binding (Haugland 1992).
Small, but detectable, amounts of fluorescence remained 2 s after the end of stimulation (16.2 ± 0.9% of peak for Fluo4, 30.6 ± 14.7% for furaptra; not significantly different, n = 3 for both, see insets of Fig. 2D). Subsequent Ca2+ clearance time courses measured with the two dyes were indistinguishable (Fig. 2D) (see Brager et al. 2003 for a quantitatively similar measurement with different techniques) and matched the decay back to baseline in PVES (Fig. 2E) consistent with the working hypothesis that residual Ca2+ increases PVES of newly recruited vesicles, just as it does for vesicles that have been in the releasable state for minutes (Stevens and Wesseling 1999a).
No evidence for preferentially faster recruitment of vesicles with high intrinsic PVES
An alternate explanation for the faster recovery of the responses to single pulses would be that after being emptied, the RRP is preferentially restocked first with a few vesicles with intrinsically high PVES, followed later by a larger contingent with a lower PVES. However, previous studies of excitatory hippocampal synapses grown in cell culture showed that RRP recovery after partial depletion of only 15% proceeded with a time course that was similar to recovery after full RRP depletion, suggesting that the vesicles with the highest PVES are not restocked more quickly at hippocampal synapses (see Fig. 2 of Stevens and Wesseling 1999a). On the other hand, there is evidence for heterogeneity in the release properties among individual vesicles within the RRP at a large variety of synapse types, including excitatory hippocampal synapses (Hanse and Gustafsson 2001; Millar et al. 2005; Molder and Mennerick 2005; Neher and Zucker 1993). And several reports have suggested recently that the vesicles with a higher PVES can be restocked at a rate that is different from the average, at least at Calyx of Held synapses, although in those cases it was suggested that the high PVES vesicles were restocked more slowly than the average (Sakaba and Neher 2001a; Wadel et al. 2007).
It was thus worthwhile to test the synapses used in the present study for possible faster recruitment to the RRP of the vesicles with the highest intrinsic PVES. The experimental design was conceptually similar to the experiments documented in the preceding text, except the first train consisted of only three pulses (Fig. 3A). As in the preceding experiments, the second train consisted of 60 pulses, which were needed for complete RRP depletion. After the shortest rest intervals, the sum of the responses to the first 60 pulses during the second train was reduced by 18 ± 0.6% and subsequent recovery of this sum followed a time course that was similar, or slower, than the time course of full RRP recovery (Fig. 3B). As the three pulses in the first train would preferentially expend the vesicles with the highest PVES, these results indicate that high PVES vesicles are not replaced more quickly than the average, arguing against the alternate explanation (see DISCUSSION). Further experiments documented below show that synaptic strength transiently overshoots the baseline value under some circumstances (see Fig. 6C), and the overshoot would not be compatible with the alternate explanation either.
|
As episodes of extensive use slow the overall rate at which vesicles are recruited to the RRP (Stevens and Wesseling 1999b), an additional set of experiments (diagrammed above Fig. 4) was conducted to determine if this additional component of depression interferes with the increase in PVES. No interference was detected after stimulus trains of 600 pulses (Fig. 4A), even though RRP recovery was slower (compare Fig. 4A to Fig. 2A). The time courses of the decay back to baseline of the increase in PVES (Fig. 4B) and the clearance of residual Ca2+ (Fig. 4C) were both lengthened by about twofold but remained matched (Fig. 4D). In addition, three-way comparisons showed no difference in the relationship between intracellular Ca2+ and PVES after 80 and 600 pulses (Fig. 4D). This is a striking result because the electrophysiological subset of the data were extracted from recovery time courses that differed by >10-fold. (The bend in the theoretical curve in Fig. 4D is consistent with partial saturation of the underlying mechanism, see following text).
Faster relaxation near body temperature matches augmentation
One characteristic feature of the decay time course of augmentation is its temperature dependence, which was characterized by a Q10 of 3–4 (Delaney and Tank 1994; Magleby and Zengel 1976a). To determine if the decay time course for PVES at recovering synapses is equally sensitive to temperature changes, a series of experiments similar to Fig. 2 was conducted at 33–35°C. The mismatch between the time course of recovery of responses to single pulses and the time course of RRP recovery was even more dramatic at the elevated temperature (Fig. 5A), and PVES decayed to baseline more than threefold faster than at room temperature (Fig. 5B, time constant of 2 s; compare with Fig. 2B). The time course of residual Ca2+ clearance was faster as well (Fig. 5C) and remained matched to the decay of PVES (Fig. 5D).
As room temperature was 10° cooler (23–25°C), these results indicate that the decay time courses for PVES and residual Ca2+ both have Q10's of about 3 at recovering synapses, matching the temperature sensitivity of augmentation when studied in isolation. More generally, the time course of residual Ca2+ clearance matched the decay of PVES to baseline levels even when manipulated experimentally over a ninefold range [ = 2 s in Fig. 5, B and C vs. an 18 s best fitting single exponential (not shown) in Fig. 4, B and C], supporting the working hypothesis that residual Ca2+drives the transient increases in PVES.
Accretion: saturation of PVES increase
To explore further the relationship between the recovery from depression of responses to single pulses and augmentation, we next examined the dependence on the amount of stimulation. When studied in isolation, augmentation increases steadily with stimulating frequency (Stevens and Wesseling 1999a; Zengel and Magleby 1982), but, in contrast, the next series of experiments showed that the 60 pulses that are required to empty the RRP drive PVES to a near-maximal level.
Paired trains were used, similar to the preceding text. For the first set of experiments, the first train of each pair consisted of either 60 pulses at 20 Hz or 120 at 40 Hz; the duration of the first train (3 s) was kept constant for the two experimental conditions so that the amount of transmitter release would be equivalent (Wesseling and Lo 2002). After a 2-s rest interval, the response to the first pulse of the second train was similar for the two conditions (Fig. 6A, actually 12 ± 7.7% smaller after 40-Hz stimulation, but this difference was not statistically significant). Because the RRP recovered slightly less during the 2-s rest interval after 40-Hz stimulation (37 ± 5.2 vs. 43 ± 3.4%, also not significant), the amount of increase in PVES was indistinguishable (2.0 ± 0.3-fold vs. 2.0 ± 0.4). Thus although higher frequency stimulation was initially effective at releasing transmitter at a higher rate (Fig. 6A), no additional effect on PVES was detected when measured 2 s later.
These data indicate that PVES does not always increase linearly with more stimulation, specifically not after 60 pulses with a standard concentration of extracellular Ca2+ (2.6 mM). Although this behavior does not match that of augmentation when studied in isolation from depression, the increase in PVES above baseline that is apparent during recovery from depression could still reflect common underlying machinery that was already driven to a maximal level by the residual Ca2+ that accumulates during 60-pulse trains.
The possible presence of a maximal level for PVES was tested by varying the Ca2+ and Mg2+ concentrations in the extracellular recording solution. Raising Ca2+, or lowering Mg2+, increases the baseline synaptic response because more Ca2+ flows into the terminals with each action potential, thus increasing the baseline value of PVES (Augustine and Charlton 1986; Dodge and Rahamimoff 1967; Fernandez-Chacon et al. 2001). But, if PVES already approached a maximal level during the 60-pulse train under standard conditions, raising the Ca2+/ Mg2+ ratio would no longer be expected to have as much impact.
Raising the Ca2+/ Mg2+ ratio did indeed have a reduced impact on synaptic responses at times when PVES had already been increased (Fig. 6B). The response to the first pulse of the first train was 1.77 ± 0.094-fold larger at the higher Ca2+ concentration (4.5 mM, Fig. 6Bi). But the response to the first pulse of the second train was significantly less affected (P < 0.02); i.e., only 1.20 ± 0.17-fold larger in high Ca2+ (Fig. 6Bii). Because the RRP recovered to a similar extent during the 2-s inter-train interval under the two conditions (35 ± 2.2% at 4.5 mM Ca2+ vs. 37 ± 2.4%), these results indicate that raising the Ca2+/ Mg2+ ratio has a substantially reduced impact on PVES when PVES has already been increased by the first train.
Taken together, these results suggest that residual Ca2+ in Schaffer collateral terminals increases PVES for newly recruited vesicles to a near-maximal level by the Ca2+ admitted by 60-pulse trains under standard conditions. Although larger relative increases in synaptic strength have been reported for augmentation when studied in isolation, the baseline value of PVES in the earlier experiments was far from any maximum, typically having been reduced more than 100-fold prior to the start of the experiments by manipulating extracellular Ca2+ and Mg2+ levels (Dodge and Rahamimoff 1967; Magleby and Zengel 1976a).
Nevertheless, the theoretical curve in Fig. 4D, and the trend for the response to the first pulse of the second train to be somewhat larger at 4.5 mM Ca2+ versus at 2.5 mM Ca2+ (Fig. 6Bii) suggested that any underlying saturation of the release trigger may not have been complete after 60 pulses under the standard conditions used in the preceding text. Analysis of additional, similar experiments making up a larger data set confirmed this, having the resolution to show that the first response of the second train was 25% higher at the higher Ca2+ level, and that the induction of depression was correspondingly faster (Fig. 6Biii).
Synaptic strength overshoots in lower Ca2+
At the standard Ca2+ and Mg2+ concentrations used in the preceding text, responses to single pulses recovered to the baseline value more quickly than did the entire RRP but never surpassed the baseline by a substantial amount. If this is because of a limit on the maximum value of PVES, it should be possible to design an experiment where the responses to single pulses do overshoot baseline after the induction of depression by, for example, lowering the baseline value of PVES and thus increasing the available dynamic range.
A large amount of overshoot in the response to single pulses could indeed be achieved by lowering the Ca2+/Mg2+ ratio. After a 5-s rest interval in 1.3 mM Ca2+/ 2.6 mM Mg2+, the response to the first pulse of the second train was 2.13 ± 0.36-fold larger than the response to the first pulse of the first train (Fig. 6C, i and ii). Additional control experiments indicated that the first train was still sufficient to empty the RRP at the lower Ca2+ level (Supplementary Fig S1).
As noted earlier, the overshoot is predicted by the working model, but is incompatible with alternatives that would suppose that the mismatch between the time course of recovery of responses to individual pulses and the time course of RRP recovery arises instead from different replenishment rates of independent subdivisions of the RRP.
Cell culture measurements in the presaturating range
It was not feasible to measure the effect of fewer than 60 pulses on PVES for newly recruited vesicles in the slice preparation because the basic experimental design requires that the first train empty the RRP. To link the fast recovery of single pulse responses to previously defined augmentation, it was nevertheless important to determine if the increase in PVES requires action potentials and residual Ca2+ or if it is instead some intrinsic feature of newly recruited vesicles.
To circumvent the technical limitation, the RRP can be emptied in a Ca2+-independent fashion by using osmotic shocks instead of action potentials (Molder and Mennerick 2005; Rosenmund and Stevens 1996). Osmotic shocks are typically induced by flowing hypertonic solutions over synapses, and it was not feasible to temporally and spatially restrict solution exchanges to electrically activated synapses in the slice preparation as would be required. However, it is possible to grow isolated hippocampal neurons in primary culture where solution exchanges can be performed quickly and with more spatial precision (Bekkers and Stevens 1991; Rosenmund and Stevens 1996). Isolated neurons have the additional advantage that they form recurrent synaptic connections (autapses), making it uniquely possible to simultaneously elicit exocytosis with osmotic shocks and action potentials at the same afferent synapses (Rosenmund and Stevens 1996; Stevens and Wesseling 1998).
Pilot experiments showed that autapses retain the mismatch between the fast recovery of the responses to single pulses, and the slower RRP recovery time course (data not shown), permitting us to combine osmotic shocks with electrical stimulation to determine the effects of small numbers of presynaptic action potentials on PVES during RRP recovery. The experimental design is diagrammed at the top of Fig. 7. For each experiment, the RRPs were emptied twice in succession with pairs of osmotic shocks separated by 5-s rest intervals, and single-pulse responsiveness was monitored with "test" action potentials elicited 100 ms before each osmotic shock. Activity dependent enhancement of the single pulse response preceding the second osmotic shock (i.e., "test AP 2" in the diagram above Fig. 7) was achieved by evoking between 0 and 40 "activating" action potentials during the last 1 s of the first osmotic shock (Fig. 7, A and B). As some of the increase in the single pulse responses could have been due to an increase in the number of vesicles recruited to the RRP (Stevens and Wesseling 1998), the increase in RRP recovery was also estimated by comparing the response to the second osmotic shock to the response to the first osmotic shock but was substantially less (Fig. 7B). Increases in PVES (relative to baseline) were then extracted from these measurements by dividing the amount of increase in the response to test AP 2 by the amount of increase in RRP recovery (Fig. 7B). The analysis showed that PVES increased monotonically with the first 20 activating action potentials with a characteristic accretion that could be approximated by a single exponential with an exponential parameter of nine action potentials (Fig. 7B, see legend for equation).
|
The experiments summarized in Fig. 7, A and B, were conducted at a low Ca2+/Mg2+ ratio (1 mM/3 mM) to set the baseline value of PVES far from the maximum level, but the effect was also present after 2-s rest intervals with the standard concentrations of Ca2+ and Mg2+ used in the slice preparation (2.6 mM/1.3 mM). In these additional experiments, 40 activating action potentials enhanced the synaptic response to the second test action potential 4.1-fold, while increasing the amount of RRP recovery by only 1.5-fold, for an extrapolated 2.7-fold increase in PVES.
The techniques available in cell culture allowed a more direct test of the residual Ca2+ dependence of the transient increase above baseline in PVES at recovering synapses by using exogenous Ca2+ buffers. The increase that was evident after 2-s rest intervals under standard conditions was nearly completely abolished after 5 min exposure to 100 µM EGTA-AM (Fig. 7C, same neurons, P < 0.05), providing compelling evidence that residual Ca2+ drives the increase in PVES for newly recruited vesicles, just as it does to enhance synaptic strength at synapses where the RRP is full, and the release ready vesicles are known to be in the fully primed state.
Potentiated propensity for fusion of newly recruited vesicles
Even when applied rapidly, it takes several hundred milliseconds before osmotic shocks begin to elicit transmitter release, and the length of this delay is thought to reflect the propensity with which readily releasable vesicles can be triggered for exocytosis (Basu et al. 2007; Stevens and Wesseling 1999a). As a final check of the working hypothesis, it was important to determine if this delay is decreased by residual Ca2+ at times when the RRP consists only of newly recruited vesicles because residual Ca2+ is already known to drive a decrease in the delay when the RRP is full of vesicles that have been in the releasable state for minutes. Although designed and originally analyzed to answer a different question (Stevens and Wesseling 1998), experiments that test this have already been conducted in cell culture (Fig. 8A ). A more extensive analysis of the previous data set showed that, indeed, the residual Ca2+-dependent increase in PVES at recovering synapses is associated with a decrease in the delay before onset of the synaptic response to osmotic shocks (Fig. 8, B and C, left). The decrease was similar and has a similar time course of decay (Fig. 8D), to the decrease at synapses where the readily releasable vesicles had been recruited 1 min earlier (Stevens and Wesseling 1999a). Further, EGTA blocked the decrease in the delay indicating that the decrease depended on residual Ca2+ (Fig. 8, B and C, right).
|
Occlusion of paired-pulse facilitation
Under the low PVES conditions of previous experiments, augmentation has been shown to operate independently of other elements of enhancement such as paired-pulse facilitation (PPF) (Magleby and Zengel 1976a; Zengel and Magleby 1982). Despite the mechanistic independence, however, our results predict that PPF should be occluded by augmentation at times when PVES is already nearly maximal because PPF is also expressed as an increase in PVES. As occlusion could have important physiological implications, the electrophysiological data documented in Figs. 2 and 4 were reanalyzed to determine the extent of PPF during recovery intervals. The expression of PPF was indeed abolished for the first several seconds after stimulation, and this interference dissipated with a time course that was similar to the decay of PVES (Fig. 9, A and B). PPF was also transiently abolished in the near-body temperature experiments, but the baseline amount was too small (15%) to extract a clear recovery time course from the data set documented in Fig. 5. Additional experiments were thus conducted at a reduced extracellular Ca2+ level (1.6 mM instead of 2.6 mM), and the resulting recovery time course (Fig. 9C) matched the decay in PVES at the elevated temperature as well (compare with Fig. 5B). In this context, it is notable that robust PPF was present 2 s after the three-pulse trains used to partially deplete the RRP (see Fig. 3) in agreement with Fig. 7B, which indicates that three action potentials are not enough to drive PVES to the maximum level.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; Korn and Faber 1987; Zucker 1973). The positive components are known collectively as short-term enhancement. They are mostly driven by residual Ca2+ and include several kinetically distinct elements (Fisher et al. 1997). One element, previously termed augmentation, persists for several seconds owing to slow residual Ca2+ clearance from presynaptic terminals and is expressed as an increase in the fraction of the available vesicles within the RRP that is released following single action potentials; i.e., as an increase in PVES (Stevens and Wesseling 1999a). We show here that residual Ca2+ continues to increase PVES via the same or similar mechanism at excitatory hippocampal synapses at times when RRP replenishment is ongoing and the overall probability of release is depressed due to RRP depletion. 1) Single action potentials trigger the release of a larger fraction of the newly recruited readily releasable vesicles during recovery from depression than they do under baseline conditions (i.e., PVES is increased, Figs. 1 and 7B). 2) The increase in PVES requires residual Ca2+ (Fig. 7C). 3) The increase in PVES is transient, decaying away with a time course that matches the clearance of residual Ca2+ even when the time course is manipulated experimentally over a ninefold range (Figs. 2, 4, and 5). And 4), the increase in PVES occurs in concert with a decrease in the delay before onset of neurotransmitter release when the release is triggered with osmotic shocks instead of action potentials, suggesting that the energy barrier that normally prevents fusion is lowered (Fig. 8).
Alternative models that ascribe the increase in PVES during recovery from depression to a different type of mechanism than the one underlying augmentation could account for some of our observations, but would need to be complicated to account for all of them. For example, the nonmonotonic recovery of single pulse responses in Fig. 5A, the overshoot above baseline in single pulse responses in Fig. 6C, and the absence of a faster RRP recovery time course after partial depletion (see Fig. 3 of the present manuscript, and Fig. 2 of Stevens and Wesseling 1999a) would rule out alternatives whereby a highly releasable subset of the RRP is replenished first, followed later by reluctantly releasable vesicles.
The presence of residual Ca2+ does not necessarily cause synaptic strength to enhance above baseline, however, because residual Ca2+ levels are typically highest immediately after active episodes when the RRP tends to be most depleted, and by the time the RRP has fully recovered, residual Ca2+ clearance is complete. Residual Ca2+, instead, accelerates recovery from depression in the single pulse release probability so that it considerably outpaces RRP recovery (Fig. 10). Thus a fast rebound from depression often does not indicate that synapses are quickly restored to their initial functional state because the probability of release depresses more quickly during subsequent bursts of activity.
|
A slower decay for enhancement (Klyachko and Stevens 2006) and a delay between the clearance of residual Ca2+ and the decay of enhancement (Brager et al. 2003) have been reported previously for excitatory hippocampal synapses, but no such delay was apparent in our experiments. The discrepancies could be due to differences in the interpretation of data. One key difference is that Brager et al. and Klyachko and Stevens both judged that RRP replenishment was complete at times when depression was no longer evident in the responses to single pulses. However, a re-analysis of their data that takes account of the slower RRP replenishment time courses reported here and elsewhere (Rosenmund and Stevens 1996; Stevens and Tsujimoto 1995; Wesseling and Lo 2002) yield a faster decay for enhancement, which then does match the clearance of residual Ca2+, at least in the case of Brager et al., who also measured the residual Ca2+ clearance time course.
Thus some of our quantitative conclusions depend on our observation that the RRP takes longer to replenish than has been reported elsewhere. Our comparatively slow estimates of RRP replenishment rates are based on a variety of independent techniques, including pairs of osmotic shocks, which elicit transmitter release downstream of action potentials and Ca2+ influx (Rosenmund and Stevens 1996; Stevens and Tsujimoto 1995), pairs of 60-pulse trains of electrical stimulation (Wesseling and Lo 2002), and combinations of electrical stimulation and osmotic shocks (Figs. 7 and 8) (see also Stevens and Wesseling 1998, 1999b). All of these types of measurements yield a RRP size and recovery time course that 1) are consistent across experimental conditions, 2) do not change when the extracellular Ca2+ level is changed, even while synaptic responses to single pulses change dramatically, and 3) are similar or identical in cell culture and hippocampal slices (Rosenmund and Stevens 1996; Stevens and Williams 2007; Wesseling and Lo 2002).
On the other hand, faster estimates of RRP replenishment at comparable synapses have typically been derived from measurements of recovery from depression of the responses to single pulses or short trains of stimulation, and this measurement strategy can be compromised by residual Ca2+ dynamics. That is, a key prediction of the present study is that recovery time courses measured with short-train stimulation protocols (or single pulses) would be faster than the underlying recruitment of new vesicles if the short trains were not sufficient to empty the RRP because a larger fraction of the readily releasable vesicles would be released after brief rest intervals, before clearance of the residual Ca2+, than after longer rest intervals. Indeed we can confirm from our own data that recovery time courses measured with short trains are faster than time courses measured with our standard 60-pulse stimulation protocol, falling somewhere between the pairs of recovery time courses plotted in Figs. 2A, 4A, and 5A (not shown).
In this context, we make the perhaps counterintuitive prediction that estimates of the timing of RRP recovery that are based on short-train protocols would appear to be substantially slower at high extracellular Ca2+ levels, or under other conditions where baseline PVES is higher, because under high baseline PVES conditions, short trains would be more effective at emptying the RRP, which would avoid the confounding effect of residual Ca2+. And, indeed, this prediction seems to hold in several recent studies (Molder and Mennerick 2005; Schluter et al. 2006; Toonen et al. 2006).
Relation to a "preprimed pool"
Our study does not rule out the presence of an immediately releasable pool (IRP), or preprimed pool, that would be a subset of the RRP by our definition (Hanse and Gustafsson 2001; Sakaba and Neher 2001b). If present, such a pool would be replenished with vesicles either 1) from the greater RRP, i.e., the RRP by the present definition, or 2) from some other source. In either case, our main conclusions would remain valid for the following reasons. 1) If replenishment were to come from within the greater RRP, then our results could be reinterpreted as reflecting Ca2+ driven dynamic modulation of the size of the hypothetical IRP, but alternate mechanisms such as these would account equally well for synaptic enhancement defined under low probability of release conditions. 2) If the hypothetical IRP were replenished from some other source, the results in Fig. 3 indicate that recovery would proceed with a time course that is similar to that of the greater RRP. Thus, the calculations of increases in PVES described above would be similar, and the conclusions would be the same.
In either case, the observation that responses depress to a steady-state level more quickly during a second train started after a 2-s rest interval than during the preceding first train (Fig. 1A) suggest that residual Ca2+ enhances the release efficiency of all readily releasable vesicles. Indeed data from cell culture suggest that reluctantly releasable vesicles may exhibit even more enhancement than the highly releasable ones because during osmotic shocks, residual Ca2+ decreases the delay before release of the last to be released vesicles by more than the delay before release of the first ones (see Fig. 4 of Stevens and Wesseling 1999a).
Gradual versus stepwise vesicle priming: comparison to the Calyx of Held
Although it often takes tens of seconds or longer for spent vesicles to be restocked within the RRP, the present findings imply that individual vesicles make the final transition from not-at-all available to fully primed more quickly, at least at excitatory hippocampal synapses; i.e., in <1 s, but possibly faster. A study of Calyx of Held synapses has suggested that the underlying vesicle priming reaction is gradual because individual, newly recruited vesicles may be only reluctantly releasable, each taking several tens of seconds to achieve a fully primed state (Wu and Borst 1999; but see Wadel et al. 2007). Gradual priming does not appear to be due to a general molecular limitation of vesicle trafficking, however, as we find that readily releasable vesicles are more, rather than less, easily triggered to undergo exocytosis within 1 s after being recruited to the RRP at hippocampal synapses. While our data indicate that the heightened propensity for release of newly recruited vesicles is dependent on the residual Ca2+ that accumulates during the stimulation used to empty the RRP, the osmotic shock experiments summarized in Fig. 8 suggest that even in the absence of residual Ca2+, individual vesicles make a quick transition from completely unavailable to fully releasable. Definitively ruling out a small contribution of gradual vesicle priming was not feasible with the present techniques, however, because in addition to eliciting exocytosis, osmotic shocks also transiently reduce PVES (Supplementary Fig S2), possibly by inhibiting Ca2+ channels (Rosenmund and Stevens 1996).
The experimental protocols typically used to exhaust the RRP at the Calyx of Held elicit release much more rapidly than has been possible in our experiments at hippocampal synapses, and it is possible that it is this experimental difference that affects whether or not newly recruited vesicles prime for release gradually and over what time course, but, on the other hand, the physiological requirements for synaptic transmission at the Calyx of Held are exceptionally different from requirements at Schaffer collaterals (Hermann et al. 2007; Ranck 1973), and so the presence of basic differences in the mechanisms underlying short-term plasticity are not necessarily surprising.
In addition, another series of studies of the Calyx of Held has suggested that PVES may be partly determined by recruitment of Ca2+ channels to release sites, which is not complete until around 1 s after priming (Wadel et al. 2007). A similar mechanism would have gone undetected in the present study because rest intervals of <1 s were not tested. In any case, it would be technically difficult to detect the presence of such an event at hippocampal synapses with presently available techniques because the minimum amount of electrical stimulation required to empty the RRP already results in enough residual Ca2+ to drive PVES to a maximum value.
Mechanism
The molecular mechanism underlying augmentation remains to be elucidated. Facilitation, which is defined as a much more transient component of short-term enhancement, is thought to depend indirectly on residual Ca2+ due to partial saturation of endogenous Ca2+ buffers rather than directly via activation of a bona fide residual Ca2+ sensor (Felmy et al. 2003), at least in some cases (Zucker 2003). However, an analogous explanation for augmentation would not account for the decrease in the delay before onset of synaptic responses to osmotic shocks (e.g., Fig. 8) because osmotic shocks trigger neurotransmitter release by a mechanism that does not depend on the Ca2+ buffering capacity within the synaptic terminals (Rosenmund and Stevens 1996). Also, the Ca2+-dependent mechanism underlying augmentation would likely be distinct from the one that triggers exocytosis directly as the sensor is thought to bind Ca2+ with higher affinity and a slower dissociation rate (Kamiya and Zucker 1994), and it seems to function as a linear Ca2+ detector at the lowest Ca2+ concentrations, while the exocytic trigger likely requires the cooperative binding of multiple Ca2+ ions (Zucker and Regehr 2002). An intriguing role in controlling PVES at a variety of synapse types has recently been proposed for diacylglycerol-dependent activation of PKC and/or Munc 13–1 that might be related (Basu et al. 2007; Korogod et al. 2007; Wierda et al. 2007).
Saturating increases in PVES cause augmentation to occlude PPF
There is one notable difference between the generalized residual Ca2+ driven enhancement of the release machinery characterized here and augmentation by its original definition. Under standard conditions, it seems that PVES can be driven to a maximum level at excitatory hippocampal synapses, whereas under the low-Ca2+ conditions of the original experiments, no limit has been reported. The presence of a maximum level could have important implications for understanding the roles of other elements of short-term enhancement as well. For example, our data reveal that the expression of PPF is occluded for several seconds after moderate amounts of activity (Fig. 9) (see also Brager et al. 2003; McNaughton 1982). This occlusion is particularly striking, given that it tends to occur at synapses when they are depressed because of the contrast with the current concept that the magnitude of PPF tends to be inversely proportional to release probabilities at fully rested synapses (Dobrunz and Stevens 1997; Hanse and Gustafsson 2002; Manabe et al. 1993).
The cause for the maximal value for PVES is presently unclear but might be directly due to a limit in the modulatory range of some component of the release machinery or to a secondary effect of partial RRP depletion, perhaps because of depletion of a highly releasable subset of the RRP (i.e., an IRP, see preceding text). On one hand, a twofold increase under standard conditions seems to be near-maximal, which might suggest that the maximum value of PVES is only 10–15% because the starting value is only 5% (Wesseling and Lo 2002). On the other hand, it is likely that some of the individual vesicles within the RRP have a higher intrinsic PVES than others (Hanse and Gustafsson 2001; Molder and Mennerick 2005). If baseline PVES of these vesicles were much higher than the average, a twofold potentiation at the start of a second train could cause immediate depletion of this highly releasable subdivision of the RRP, which could explain the apparent maximum value for PVES of the population average seen here.
Meta short-term plasticity
Current theories emphasize the importance of both the initial probability of release and the rate of depression for filtering and transforming information encoded within bursts of action potentials (Chance et al. 1998; Fuhrmann et al. 2002; Natschlager et al. 2001; Tsodyks and Markram 1997). In fact, long-term plasticity of these parameters has been described (Abbott and Nelson 2000; Markram and Tsodyks 1996); the terminology used was redistribution of synaptic efficacy. This report shows how these two parameters can also be controlled dynamically and differentially over the several seconds it takes for synapses to recover during rest intervals that routinely separate bursts of action potentials in the Schaffer collaterals of awake and behaving animals (O'Keefe and Dostrovsky 1971; Ranck 1973). Residual Ca2+ causes the initial probability of release to recover from depression considerably more quickly than the RRP, but the same synapses then depress more rapidly during subsequent bursts that are initiated after short rest intervals. In this context, classically defined augmentation could be considered to function as a medium-term modulator of the short-term synaptic plasticity exhibited during bursts—i.e., meta short-term plasticity—and would be the second type of residual Ca2+-dependent mechanism known to operate at the synaptic level to counter short-term depression, contrasting importantly with mechanisms that accelerate vesicle recruitment and thus hasten the recovery of all parameters to full functionality (Dittman and Regehr 1998; Sakaba and Neher 2001a; Stevens and Wesseling 1998; Wang and Kaczmarek 1998).
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
1 The online version of this article contains supplemental data. 
Address for reprint requests and other correspondence: J. F. Wesseling, Dept. de Neurociencias, CIMA, Universidad de Navarra, Pio XII, 55, 31008 Pamplona, Spain (E-mail: jwesseling{at}unav.es)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Abbott LF, Varela JA, Sen K, Nelson SB. Synaptic depression and cortical gain control. Science 275: 220–224, 1997.[CrossRef][Web of Science][Medline]
Allen C, Stevens CF. An evaluation of causes for unreliability of synaptic transmission. Proc Natl Acad Sci USA 91: 10380–10383, 1994.
Augustine GJ, Charlton MP. Calcium dependence of presynaptic calcium current and post-synaptic response at the squid giant synapse. J Physiol 381: 619–640, 1986.
Basu J, Betz A, Brose N, Rosenmund C. Munc13-1 C1 domain activation lowers the energy barrier for synaptic vesicle fusion. J Neurosci 27: 1200–1210, 2007.
Bekkers JM, Stevens CF. Excitatory and inhibitory autaptic currents in isolated hippocampal neurons maintained in cell culture. Proc Natl Acad Sci USA 88: 7834–7838, 1991.
Brager DH, Cai X, Thompson SM. Activity-dependent activation of presynaptic protein kinase C mediates post-tetanic potentiation. Nat Neurosci 6: 551–552, 2003.[CrossRef][Web of Science][Medline]
Chance FS, Nelson SB, Abbott LF. Synaptic depression and the temporal response characteristics of V1 cells. J Neurosci 18: 4785–4799, 1998.
Chung S, Li X, Nelson SB. Short-term depression at thalamocortical synapses contributes to rapid adaptation of cortical sensory responses in vivo. Neuron 34: 437–446, 2002.[CrossRef][Web of Science][Medline]
Cook DL, Schwindt PC, Grande LA, Spain WJ. Synaptic depression in the localization of sound. Nature 421: 66–70, 2003.[CrossRef][Medline]
Delaney KR, Tank DW. A quantitative measurement of the dependence of short-term synaptic enhancement on presynaptic residual calcium. J Neurosci 14: 5885–5902, 1994.[Abstract]
Dittman JS, Regehr WG. Calcium dependence and recovery kinetics of presynaptic depression at the climbing fiber to Purkinje cell synapse. J Neurosci 18: 6147–6162, 1998.
Dobrunz LE, Stevens CF. Heterogeneity of release probability, facilitation, and depletion at central synapses. Neuron 18: 995–1008, 1997.[CrossRef][Web of Science][Medline]
Dodge FA Jr, Rahamimoff R. Co-operative action of calcium ions in transmitter release at the neuromuscular junction. J Physiol 193: 419–432, 1967.
Felmy F, Neher E, Schneggenburger R. Probing the intracellular calcium sensitivity of transmitter release during synaptic facilitation. Neuron 37: 801–811, 2003.[CrossRef][Web of Science][Medline]
Feng TP. Studies on the neuromuscular junction XXVI, the changes of the end-plate potential during and after prolonged stimulation. Chin J Physiol 16: 341–372, 1941.
Fernandez-Chacon R, Konigstorfer A, Gerber SH, Garcia J, Matos MF, Stevens CF, Brose N, Rizo J, Rosenmund C, Sudhof TC. Synaptotagmin I functions as a calcium regulator of release probability. Nature 410: 41–49, 2001.[CrossRef][Medline]
Fisher SA, Fischer TM, Carew TJ. Multiple overlapping processes underlying short-term synaptic enhancement. Trends Neurosci 20: 170–177, 1997.[CrossRef][Web of Science][Medline]
Frakenhaeuser B, Hodgkin AL. The action of calcium on the electrical properties of squid axons. J Physiol 137: 218–244, 1957.
Fuhrmann G, Segev I, Markram H, Tsodyks M. Coding of temporal information by activity-dependent synapses. J Neurophysiol 87: 140–148, 2002.
Hagler DJ Jr, Goda Y. Properties of synchronous and asynchronous release during pulse train depression in cultured hippocampal neurons. J Neurophysiol 85: 2324–2334, 2001.
Hanse E, Gustafsson B. Vesicle release probability and pre-primed pool at glutamatergic synapses in area CA1 of the rat neonatal hippocampus. J Physiol 531: 481–493, 2001.
Hanse E, Gustafsson B. Release dependence to a paired stimulus at a synaptic release site with a small variable pool of immediately releasable vesicles. J Neurosci 22: 4381–4387, 2002.
Haugland RP. Handbook of Fluorescent Probes and Research Chemicals. Eugene, OR: Molecular Probes, 1992.
Hermann J, Pecka M, von Gersdorff H, Grothe B, Klug A. Synaptic transmission at the calyx of Held under in vivo like activity levels. J Neurophysiol 98: 807–820, 2007.
Hubbard JI. Repetitive stimulation at the mammalian neuromuscular junction, and the mobilization of transmitter. J Physiol 169: 641–662, 1963.
Kalkstein JM, Magleby KL. Augmentation increases vesicular release probability in the presence of masking depression at the frog neuromuscular junction. J Neurosci 24: 11391–11403, 2004.
Kamiya H, Zucker RS. Residual Ca2+ and short-term synaptic plasticity. Nature 371: 603–606, 1994.[CrossRef][Medline]
Klyachko VA, Stevens CF. Temperature-dependent shift of balance among the components of short-term plasticity in hippocampal synapses. J Neurosci 26: 6945–6957, 2006.
Korn H, Faber DS. Regulation and significance of probabilistic release mechanism at central synapses. In: Synaptic Function, edited by Edelman GM, Gall WE, Cowan WM. New York: Wiley, 1987, p. 57–108.
Korogod N, Lou X, Schneggenburger R. Posttetanic potentiation critically depends on an enhanced Ca (2+) sensitivity of vesicle fusion mediated by presynaptic PKC. Proc Natl Acad Sci USA 104: 15923–15928, 2007.
Liley AW, North KA. An electrical investigation of effects of repetitive stimulation on mammalian neuromuscular junction. J Neurophysiol 16: 509–527, 1953.
Magleby KL, Zengel JE. Augmentation: a process that acts to increase transmitter release at the frog neuromuscular junction. J Physiol 257: 449–470, 1976a.
Magleby KL, Zengel JE. Long term changes in augmentation, potentiation, and depression of transmitter release as a function of repeated synaptic activity at the frog neuromuscular junction. J Physiol 257: 471–494, 1976b.
Manabe T, Wyllie DJ, Perkel DJ, Nicoll RA. Modulation of synaptic transmission and long-term potentiation: effects on paired pulse facilitation and EPSC variance in the CA1 region of the hippocampus. J Neurophysiol 70: 1451–1459, 1993.
Markram H, Tsodyks M. Redistribution of synaptic efficacy between neocortical pyramidal neurons. Nature 382: 807–810, 1996.[CrossRef][Medline]
Markram H, Wang Y, Tsodyks M. Differential signaling via the same axon of neocortical pyramidal neurons. Proc Natl Acad Sci USA 95: 5323–5328, 1998.
McNaughton BL. Long-term synaptic enhancement and short-term potentiation in rat fascia dentata act through different mechanisms. J Physiol 324: 249–262, 1982.
Millar AG, Zucker RS, Ellis-Davies GC, Charlton MP, Atwood HL. Calcium sensitivity of neurotransmitter release differs at phasic and tonic synapses. J Neurosci 25: 3113–3125, 2005.
Moulder KL, Mennerick S. Reluctant vesicles contribute to the total readily releasable pool in glutamatergic hippocampal neurons. J Neurosci 25: 3842–3850, 2005.
Nadim F, Manor Y, Kopell N, Marder E. Synaptic depression creates a switch that controls the frequency of an oscillatory circuit. Proc Natl Acad Sci USA 96: 8206–8211, 1999.
Natschlager T, Maass W, Zador A. Efficient temporal processing with biologically realistic dynamic synapses. Network 12: 75–87, 2001.[Web of Science][Medline]
Neher E, Zucker RS. Multiple calcium-dependent processes related to secretion in bovine chromaffin cells. Neuron 10: 21–30, 1993.[CrossRef][Web of Science][Medline]
O'Keefe J, Dostrovsky J. The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat. Brain Res 34: 171–175, 1971.[CrossRef][Web of Science][Medline]
Otsu Y, Shahrezaei V, Li B, Raymond LA, Delaney KR, Murphy TH. Competition between phasic and asynchronous release for recovered synaptic vesicles at developing hippocampal autaptic synapses. J Neurosci 24: 420–433, 2004.
Raastad M, Storm JF, Andersen P. Putative single quantum and single fiber excitatory postsynaptic currents show similar amplitude range and variability in rat hippocampal slices. Eur J Neurosci 4: 113–117, 1992.[CrossRef][Web of Science][Medline]
Ranck JB Jr. Studies on single neurons in dorsal hippocampal formation and septum in unrestrained rats. I. Behavioral correlates and firing repertoires. Exp Neurol 41: 462–531, 1973.[CrossRef][Web of Science]
Rosenmund C, Sigler A, Augustin I, Reim K, Brose N, Rhee JS. Differential control of vesicle priming and short-term plasticity by Munc13 isoforms. Neuron 33: 411–424, 2002.[CrossRef][Web of Science][Medline]
Rosenmund C, Stevens CF. Definition of the readily releasable pool of vesicles at hippocampal synapses. Neuron 16: 1197–1207, 1996.[CrossRef][Web of Science][Medline]
Sabatini BL, Regehr WG. Optical measurement of presynaptic calcium currents. Biophys J 74: 1549–1563, 1998.[Web of Science][Medline]
Sakaba T, Neher E. Calmodulin mediates rapid recruitment of fast-releasing synaptic vesicles at a calyx-type synapse. Neuron 32: 1119–1131, 2001a.[CrossRef][Web of Science][Medline]
Sakaba T, Neher E. Quantitative relationship between transmitter release and calcium current at the calyx of held synapse. J Neurosci 21: 462–476, 2001b.
Schluter OM, Basu J, Sudhof TC, Rosenmund C. Rab3 superprimes synaptic vesicles for release: implications for short-term synaptic plasticity. J Neurosci 26: 1239–1246, 2006.
Silberberg G, Wu C, Markram H. Synaptic dynamics control the timing of neuronal excitation in the activated neocortical microcircuit. J Physiol 556: 19–27, 2004.
Stevens CF, Tsujimoto T. Estimates for the pool size of releasable quanta at a single central synapse and for the time required to refill the pool. Proc Natl Acad Sci USA 92: 846–849, 1995.
Stevens CF, Wesseling JF. Activity-dependent modulation of the rate at which synaptic vesicles become available to undergo exocytosis. Neuron 21: 415–424, 1998.[Medline]
Stevens CF, Wesseling JF. Augmentation is a potentiation of the exocytotic process. Neuron 22: 139–146, 1999a.[CrossRef][Web of Science][Medline]
Stevens CF, Wesseling JF. Identification of a novel process limiting the rate of synaptic vesicle cycling at hippocampal synapses. Neuron 24: 1017–1028, 1999b.[CrossRef][Web of Science][Medline]
Stevens CF, Williams JH. Discharge of the readily releasable pool with action potentials at hippocampal synapses. J Neurophysiol 98: 3221–3229, 2007.
Toonen RF, Wierda K, Sons MS, de Wit H, Cornelisse LN, Brussaard A, Plomp JJ, Verhage M. Munc18-1 expression levels control synapse recovery by regulating readily releasable pool size. Proc Natl Acad Sci USA 103: 18332–18337, Epub 12006 Nov 18316, 2006.
Tsodyks MV, Markram H. The neural code between neocortical pyramidal neurons depends on neurotransmitter release probability. Proc Natl Acad Sci USA 94: 719–723, 1997.
Wadel K, Neher E, Sakaba T. The Coupling between synaptic vesicles and Ca(2+) channels determines fast neurotransmitter release. Neuron 53: 563–575, 2007.[CrossRef][Web of Science][Medline]
Wang LY, Kaczmarek LK. High-frequency firing helps replenish the readily releasable pool of synaptic vesicles. Nature 394: 384–388, 1998.[CrossRef][Medline]
Wesseling JF, Lo DC. Limit on the role of activity in controlling the release-ready supply of synaptic vesicles. J Neurosci 22: 9708–9720, 2002.
Wierda KD, Toonen RF, de Wit H, Brussaard AB, Verhage M. Interdependence of PKC-dependent and PKC-independent pathways for presynaptic plasticity. Neuron 54: 275–290, 2007.[CrossRef][Web of Science][Medline]
Wu LG, Borst JG. The reduced release probability of releasable vesicles during recovery from short-term synaptic depression. Neuron 23: 821–832, 1999.[CrossRef][Web of Science][Medline]
Wu LG, Saggau P. Presynaptic calcium is increased during normal synaptic transmission and paired-pulse facilitation, but not in long-term potentiation in area CA1 of hippocampus. J Neurosci 14: 645–654, 1994.[Abstract]
Zengel JE, Magleby KL. Augmentation and facilitation of transmitter release. A quantitative description at the frog neuromuscular junction. J Gen Physiol 80: 583–611, 1982.
Zucker RS. Changes in the statistics of transmitter release during facilitation. J Physiol 229: 787–810, 1973.
Zucker RS. NCS-1 stirs somnolent synapses. Nat Neurosci 6: 1006–1008, 2003.[CrossRef][Web of Science][Medline]
Zucker RS, Regehr WG. Short-term synaptic plasticity. Annu Rev Physiol 64: 355–405, 2002.[CrossRef][Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  E. Garcia-Perez, D. C. Lo, and J. F. Wesseling Kinetic Isolation of a Slowly Recovering Component of Short-Term Depression During Exhaustive Use at Excitatory Hippocampal Synapses J Neurophysiol, August 1, 2008; 100(2): 781 - 795. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
, measured as the size of the response to the 1st pulse of the 2nd train divided by the size of the response to the 1st pulse of the 1st train) and the RRP (
, measured as described in METHODS). Dashed line, single-exponential function with 
F/F vs. 20 s. ii: single wavelength Furaptra measurements (excitation was 
(t) starts off at 0.4 /s and decays exponentially 3.7-fold with a 2-s time constant. Note the nonmonotonic recovery of responses to single pulses. The gray boxes indicate measurements after the same rest interval, which are compared in B. B: decay vs. time (i.e., rest interval) of PVES calculated by dividing single pulse response recovery values by RRP recovery values (both displayed in A). The dashed line is an exponential function with a time constant of 2 s. C: clearance vs. time of residual Ca2+ after 60 pulses (40-Hz stimulation, 33–35°C, Fluo-4, detection at 5 Hz, n = 3 preparations). The white line is a scaled version of the same exponential function as in B. Data points are normalized by the response size after the 1st 1 s of rest. Inset: example, scale is 20% 
) by the relative increase in RRP recovery (
; n = 8 trials from 4 neurons). To accomplish this, the relative increase in the response to test AP 2 was calculated by first normalizing the size of the response to test AP 2 by the size of the response to test AP 1 from the same trial, and then this quantity was renormalized by the same quantity from other trials, in the same autapse, where activating APs were omitted during the 1st osmotic shock. Likewise, relative increases in RRP recovery were calculated by normalizing the size of the response to the 2nd osmotic shock by the size of the response to the 1st osmotic shock from the same trial and then renormalizing by the results of other trials where activating APs were omitted. - - -, y = [(1 – e–x/9)*1.5] +1. C: enhancement of PVES is blocked by EGTA. These experiments were similar to the ones documented in A and B except extracellular Ca2+/Mg2+ was 2.6 mM/1.3 mM, the interval between osmotic shocks was 2 s, and only trials with 40 and 0 activating APs were conducted (paired experiments on 4 neurons, * = P < 0.05). Fractional recovery was calculated as the size of the response to the test AP 2 divided by the size of the response to test AP 1.
