Calcium Sensitivity of Glutamate Release in a Calyx-Type Terminal
10.1126/science.289.5481.953
Crossref journal-article
American Association for the Advancement of Science (AAAS)
Science (221)
Abstract

Synaptic efficacy critically depends on the presynaptic intracellular calcium concentration ([Ca 2+ ] i ). We measured the calcium sensitivity of glutamate release in a rat auditory brainstem synapse by laser photolysis of caged calcium. A rise in [Ca 2+ ] i to 1 micromolar readily evoked release. An increase to >30 micromolar depleted the releasable vesicle pool in <0.5 millisecond. A comparison with action potential–evoked release suggested that a brief increase of [Ca 2+ ] i to ∼10 micromolar would be sufficient to reproduce the physiological release pattern. Thus, the calcium sensitivity of release at this synapse is high, and the distinction between phasic and delayed release is less pronounced than previously thought.

Bibliography

Bollmann, J. H., Sakmann, B., & Borst, J. G. G. (2000). Calcium Sensitivity of Glutamate Release in a Calyx-Type Terminal. Science, 289(5481), 953–957.

Authors 3
  1. Johann H. Bollmann (first)
  2. Bert Sakmann (additional)
  3. J. Gerard G. Borst (additional)
References 49 Referenced 411
  1. 10.1111/j.1749-6632.1991.tb36505.x
  2. 10.1016/S0896-6273(00)80171-3
  3. 10.1016/S0896-6273(00)80983-6
  4. 10.1016/S0143-4160(98)90060-4
  5. 10.1016/S0959-4388(99)80045-2
  6. 10.1016/S0896-6273(00)80789-8
  7. 10.1016/S0896-6273(01)80039-8
  8. 10.1523/JNEUROSCI.17-21-08137.1997
  9. Transverse brainstem slices were cut from 8- to 10-day-old Wistar rats with a vibratome. The extracellular recording solution contained (in mM) 125 NaCl 2.5 KCl 1 MgCl 2 2 CaCl 2 25 dextrose 1.25 NaH 2 PO 4 0.4 ascorbic acid 3 myo -inositol 2 sodium pyruvate and 25 NaHCO 3 (bubbled with 5% CO 2 95% O 2 ). Postsynaptic N -methyl- d -aspartate receptors were blocked by d -2-amino-5-phosphonovalerate (50 μM). Cyclothiazide (CTZ 100 μM) was added to minimize postsynaptic AMPA receptor desensitization. Simultaneous pre- and postsynaptic whole-cell recordings from giant synapses (40) were made with two Axopatch 200B amplifiers. Recordings were made at room temperature. The postsynaptic pipette solution contained (in mM) 125 cesium gluconate 20 CsCl 10 disodium phosphocreatine 4 MgATP 0.3 guanosine 5′-triphosphate (GTP) 10 Hepes 0.5 EGTA (pH 7.2). The presynaptic pipette solution contained (in mM) 90 potassium gluconate 20 KCl 30 Hepes 9 DM-nitrophen (DM-n Calbiochem CA) ∼8.6 CaCl 2 (pH 7.2). Either Mag-fura-2 or Fura-2FF was added as low-affinity Ca 2+ indicators (1 mM). We confirmed on each experimental day that the [Ca 2+ ] i of this solution was ∼100 nM using Fura-2 (0.2 mM). For [Ca 2+ ] i jumps to <6 μM and unless MgATP was included (20) the [DM-n] was lowered to 5 or 2.5 mM and the [CaCl 2 ] proportionately to reduce the rebinding of Ca 2+ to unphotolyzed DM-n. Postsynaptic pipettes had an open-tip resistance of 2 to 3 megohm. The uncompensated series resistance was 12 ± 1 megohm ( n = 43) electronically compensated to 90 to 95%. The recorded EPSCs were corrected off-line for the voltage error caused by the residual series resistance (41). Presynaptic pipettes (4 to 6 megohm) dialyzed the terminal for at least 4 min before a laser pulse was applied. All compound EPSCs and spontaneous quantal EPSCs were filtered at 5 kHz (4-pole Bessel filter) digitized stored and analyzed as in (28). Quantal release events after [Ca 2+ ] i jumps to <1.5 μM were filtered at 2 kHz (Fig. 2B) in some experiments after wash-out of CTZ.
  10. 10.1073/pnas.85.17.6571
  11. 10.1007/s004240050444
  12. 10.1016/S0006-3495(96)79644-3
  13. UV pulses from a frequency-tripled yttrium-aluminum-garnet–Nd laser were coupled into the epifluorescence port of an Axioskop by means of a quartz light guide and combined with the light of a monochromator. An area of 30 μm by 30 μm in the specimen plane was homogeneously illuminated by the laser pulse. The pulse intensity was attenuated by insertion of neutral density filters. A photodiode was used to record the fluorescence from dialyzed terminals. The photocurrent was amplified with an Axopatch 200A amplifier (11) filtered and sampled at the same frequency as the whole-cell current recordings. UV illumination at 380/341 nm and 380/352 nm was used for dual-wavelength excitation of Mag-fura-2 and Fura-2FF respectively. Background luminescence following a UV laser pulse was recorded after each experiment and subtracted off-line. To calibrate the [Ca 2+ ] i measurement we determined ratiometric constants ( R min R max ) in presynaptic terminals in separate experiments [Mag-fura-2: R min = 0.358 ± 0.007 R max = 4.61 ± 0.52; Fura-2FF: R min = 0.734 ± 0.010 R max = 6.23 ± 0.27 ( n = 5 for each value)]. The [Ca 2+ ] i measurement critically depends on the in situ K d of the used Ca 2+ indicators. Reported values range from 23 to 100 μM for Mag-fura-2 and 5 to 35 μM for Fura-2FF. We determined the K d of the indicators in presynaptic terminals with a similar solution as in (9) but with the free [Ca 2+ ] i buffered to 32 and 14 μM respectively using 1 3-diamino-2-hydroxypropane-tetraacetic acid [25 mM K d = 81 μM (42)]. The K d 's of Mag-fura-2 and of Fura-2FF were 31 ± 3 μM ( n = 9) and 8.9 ± 0.6 μM ( n = 6) respectively. The photolysis rate by monochromator UV light was ∼0.07 s -1 which did not appreciably increase basal [Ca 2+ ] i before the flash. UV pulses in the absence of a presynaptic recording did not affect the size of the afferently evoked EPSCs ( n = 4).
  14. 10.1016/0143-4160(93)90079-L
  15. The kinetics of the photolysis reaction and the rebinding of Ca 2+ to unphotolyzed DM-n can lead to a rapidly decaying [Ca 2+ ] i spike (11 12 14). At a concentration of 1 mM the low-affinity dyes Mag-fura-2 and Fura-2FF act as potent buffers of μM concentrations of Ca 2+ and owing to their fast kinetics should mirror the [Ca 2+ ] i time course with delays of a few hundred μs. In the ∼200 μs after a laser pulse the fluorescence signal was unreliable because of luminescence in the optical path. We simulated the photolysis kinetics similar to a previous model (11). It reported a [Ca 2+ ] i overshoot of <100-μs duration that could exceed the measured [Ca 2+ ] i by <500% (for Fura-2FF assumed k on = 0.55 μM −1 ms −1 ) or <50% [for Mag-fura-2 k on = 0.75 μM −1 ms −1 (43)]. Adenosine 5′-triphosphate (ATP) and endogenous Ca 2+ buffers were not taken into account which should further dampen the initial [Ca 2+ ] i spike. The [Ca 2+ ] i spike is a function of for example the photolysis efficiency (1 to 14% in our experiments) and [Ca 2+ ] i buffer conditions (11). In our experiments release rates observed at similar measured [Ca 2+ ] i levels but with different buffer conditions and photolysis efficiencies were similar despite the varying amplitude of the modeled [Ca 2+ ] i spike under these conditions (44). A test of the sensitivity of the Ca 2+ sensor model (22) to the simulated [Ca 2+ ] i spikes revealed that simulated peak release rates were increased by 0.5 to 15% compared with the rates shown in Fig. 2C which were calculated with rectangular [Ca 2+ ] i steps. The rate of release exhibited a similar Ca 2+ sensitivity when estimated with slower flash lamp–evoked Ca 2+ uncaging (R. Schneggenburger and E. Neher Nature in press). In addition uncaging Ca 2+ in a ramplike fashion with the monochromator produced release rates that were predicted by the model (J. H. Bollmann unpublished observation).
  16. To derive peak release rates from compound EPSCs we simulated the time course of the release rate with a function of the form f( t ) = A {1 + erf[ k 1 ( t − t 0 )]} exp[− k 2 ( t − t 0 )] where erf is the error function. The convolution of this time course with the average quantal EPSC was fitted to the compound EPSCs in the interval from the onset to the peak of the EPSC by adjusting A t 0 k 1 and k 2 . This method takes the finite rise time (130 μs) of the average quantal EPSC into account.
  17. 10.1073/pnas.91.12.5267
  18. 10.1016/S0896-6273(00)80522-X
  19. 10.1016/S0006-3495(97)78792-7
  20. Because DM-nitrophen binds Mg 2+ well [Ca 2+ ] i jumps to >15 μM were done in the absence of Mg 2+ . In this case ATP-dependent steps were not functional. However in about half of the experiments with a [Ca 2+ ] i jump between 0.5 and 15 μM the presynaptic solution contained Na 2 ATP (10 mM) GTP (0.3 mM) and MgCl 2 (3 mM) and therefore release in these experiments was not limited by the absence of MgATP (45). The Ca 2+ sensitivity was similar in these experiments and they were included in Fig. 2 C and D. Despite the presence of MgATP EPSCs in response to a second [Ca 2+ ] i jump were often reduced in size. Thus most of the compound release data in Fig. 2 C and D are from the first laser pulse after terminal dialysis. In separate experiments ( n = 5) we verified that the releasable pool size did not change appreciably after terminal dialysis by comparing cumulative EPSC amplitudes in response to trains of afferent stimuli in the intact and dialyzed terminal.
  21. The resting [Ca 2+ ] i before the flash affected the EPSC size and kinetics. In experiments in which the pre-flash Fura-2FF ratio exceeded R min by more than 1 SD (corresponding to a basal [Ca 2+ ] i ≥ 300 nM) the normalized EPSC amplitude was reduced by ∼50% and the rise time was increased by ∼20% compared with experiments in which the basal [Ca 2+ ] i was lower. These experiments were excluded from further analysis. Apart from partial pool depletion possible mechanisms include adaptation of the Ca 2+ sensor (31) and an increase in the fraction of “reluctant” vesicles (7).
  22. We used a sequential model to describe the binding of Ca 2+ to the Ca 2+ sensor (X): X ⇄β5α[Ca] XCa1 ⇄2β4α[Ca] XCa2 ⇄3β3α[Ca] XCa3 ⇄4β2α[Ca] XCa4 ⇄5βα[Ca] XCa5 ⇄δγ XCa5*The Ca 2+ binding (α) and dissociation (β) rate constants were 0.3 μM −1 ms −1 and 3 ms −1 respectively ( K d = β/α = 10 μM). The Ca 2+ sensor was modeled to undergo a Ca 2+ -independent isomerization step similar to some existing Ca 2+ -binding proteins [e.g. (46)]: A fully occupied sensor reversibly switches to a release-promoting state (XCa 5 * ) with isomerization rate constants γ (30 ms −1 ) and δ (8 ms −1 ). Fusion of a vesicle was modeled in a second scheme in which the fusion rate constant is scaled by the fraction of Ca 2+ sensors that are in the release-promoting state: V→ρXCa5*(t) F. ρ is the maximal fusion rate constant (40 ms −1 ) V the releasable vesicle and F the fused vesicle. The parameter XCa 5 * ( t ) represents the fraction of Ca 2+ sensors promoting release at time t. As an important aspect of this model the time course of the probability of release after an action potential depended little on the amount of Ca 2+ influx (29) because it was largely controlled by the Ca 2+ -independent constants γ and δ. The differential equations derived from this reaction scheme were solved numerically (Mathematica 3.0). To obtain the time course of the EPSC computed release rates were convolved with the time course of the measured quantal EPSC in the presence of CTZ ( n = 10 cells). The time course of the averaged quantal EPSC could be approximated with a rise time of 130 μs and a biexponential decay with time constants of 2.8 ms (54%) and 7.5 ms. Ca 2+ currents during an action potential were simulated with a Hodgkin-Huxley model (47). Resting [Ca 2+ ] i in intact terminals was assumed to be 50 nM (19). It was assumed that the [Ca 2+ ] i transients had the same time course as the Ca 2+ current and were identical at all release sites. The rationale behind these assumptions is our previous observation that many Ca 2+ channels contribute to the release of most vesicles during an action potential at the calyx of Held (28). This indicates that at most release sites [Ca 2+ ] i cannot be expected to change faster than the presynaptic Ca 2+ currents.
  23. 10.1016/0896-6273(93)90274-U
  24. 10.1016/S0006-3495(94)80744-1
  25. 10.1113/jphysiol.1996.sp021682
  26. 10.1038/371513a0
  27. 10.1073/pnas.91.26.12765
  28. 10.1038/383431a0
  29. 10.1016/S0006-3495(92)81872-6
  30. J. H. Bollmann B. Sakmann J. G. G. Borst data not shown.
  31. 10.1016/S0896-6273(00)80182-8
  32. 10.1113/jphysiol.1990.sp018150
  33. 10.1152/jn.1999.81.2.634
  34. 10.1523/JNEUROSCI.19-15-06257.1999
  35. 10.1523/JNEUROSCI.16-17-05312.1996
  36. 10.1111/j.1469-7793.1998.135by.x
  37. 10.1016/0896-6273(93)90076-4
  38. 10.1523/JNEUROSCI.19-18-07846.1999
  39. 10.1016/S0166-2236(98)01293-4
  40. 10.1113/jphysiol.1995.sp021095
  41. 10.1016/S0165-0270(98)00140-X
  42. 10.1016/0896-6273(93)90238-M
  43. 10.1016/S0143-4160(97)90064-6
  44. Supplemental Web material is available at Science Online at www.sciencemag.org/feature/data/1052156.shl.
  45. 10.1085/jgp.111.2.225
  46. 10.1002/pro.5560071123
  47. 10.1111/j.1469-7793.1998.143bx.x
  48. Simulated delays were measured between the [Ca 2+ ] i jump and the time when the integral of the release rate equaled one vesicle. Delays predicted by the model were about 250 μs faster than the measured delays. Ca 2+ uncaging is too fast (10 to 20 μs) to contribute substantially to this additional delay (11 12). Therefore it probably originates from processes downstream of Ca 2+ binding including vesicle fusion glutamate diffusion and activation of AMPA receptors.
  49. We thank C. J. Meinrenken A. D. G. de Roos and C. C. H. Petersen for comments on an earlier version of the manuscript; A. Roth for advice on simulations; and F. Helmchen for help with early flash photolysis experiments. J.G.G.B. was supported by a Pionier program of Netherlands Organization for Scientific Research (NWO).
Dates
Type When
Created 23 years, 1 month ago (July 27, 2002, 5:40 a.m.)
Deposited 1 year, 7 months ago (Jan. 13, 2024, 4:19 a.m.)
Indexed 1 month, 4 weeks ago (July 6, 2025, 1:44 p.m.)
Issued 25 years ago (Aug. 11, 2000)
Published 25 years ago (Aug. 11, 2000)
Published Print 25 years ago (Aug. 11, 2000)
Funders 0

None

@article{Bollmann_2000, title={Calcium Sensitivity of Glutamate Release in a Calyx-Type Terminal}, volume={289}, ISSN={1095-9203}, url={http://dx.doi.org/10.1126/science.289.5481.953}, DOI={10.1126/science.289.5481.953}, number={5481}, journal={Science}, publisher={American Association for the Advancement of Science (AAAS)}, author={Bollmann, Johann H. and Sakmann, Bert and Borst, J. Gerard G.}, year={2000}, month=aug, pages={953–957} }