Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 19 (1), 193-205

NMDA Receptor-Mediated Control of Presynaptic Calcium and Neurotransmitter Release

Affiliations

NMDA Receptor-Mediated Control of Presynaptic Calcium and Neurotransmitter Release

A J Cochilla et al. J Neurosci.

Abstract

Before action potential-evoked Ca2+ transients, basal presynaptic Ca2+ concentration may profoundly affect the amplitude of subsequent neurotransmitter release. Reticulospinal axons of the lamprey spinal cord receive glutamatergic synaptic input. We have investigated the effect of this input on presynaptic Ca2+ concentrations and evoked release of neurotransmitter. Paired recordings were made between reticulospinal axons and the neurons that make axo-axonic synapses onto those axons. Both excitatory and inhibitory paired-cell responses were recorded in the axons. Excitatory synaptic inputs were blocked by the AMPA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 10 microM) and by the NMDA receptor antagonist 2-amino-5-phosphonopentanoate (AP-5; 50 microM). Application of NMDA evoked an increase in presynaptic Ca2+ in reticulospinal axons. Extracellular stimulation evoked Ca2+ transients in axons when applied either directly over the axon or lateral to the axons. Transients evoked by the two types of stimulation differed in magnitude and sensitivity to AP-5. Simultaneous microelectrode recordings from the axons during Ca2+ imaging revealed that stimulation of synaptic inputs directed to the axons evoked Ca2+ entry. By the use of paired-cell recordings between reticulospinal axons and their postsynaptic targets, NMDA receptor activation was shown to enhance evoked release of transmitter from the axons that received axoaxonic inputs. When the synaptic input to the axon was stimulated before eliciting an action potential in the axon, transmitter release from the axon was enhanced. We conclude that NMDA receptor-mediated input to reticulospinal axons increases basal Ca2+ within the axons and that this Ca2+ is sufficient to enhance release from the axons.

Figures

Fig. 1.
Fig. 1.
Reticulospinal axons receive glutamatergic synaptic inputs that act on AMPA and NMDA receptors. Paired recordings were made between presynaptic neurons and postsynaptic axons. Postsynaptic axons were held under voltage clamp with a patch electrode, and presynaptic neurons were held under current clamp with a microelectrode. A, A spike induced in a presynaptic neuron by depolarizing current injection (lower trace) evoked an EPSC in the postsynaptic axon (upper trace). Application of CNQX (10 μm) reduced the amplitude of the postsynaptic axonal EPSC. The remaining component of the EPSC was abolished by the addition of AP-5 (100 μm; recording was in Ringer’s solution with no added Mg2+). B, Recordings from the same postsynaptic axon as in A and a different presynaptic neuron are shown. In this case, stimulation of the presynaptic neuron evoked an IPSC in the postsynaptic axon. C, In another postsynaptic axon, the spinal cord was stimulated to evoke a compound EPSC (the tissue was superfused with Ringer’s solution containing 2.6 mm MgCl2 and 10 μm bicuculline). Addition of CNQX (10 μm) reduced the amplitude of the response. D, TheCNQX-insensitive component of the EPSC was larger when recorded at −50 mV than when recorded at −70 mV. E, This component was reduced by the addition of AP-5 (50 μm) to the Ringer’s solution (recorded at a holding potential of −40 mV). F, Voltage–current plot of the peak amplitude of the evoked EPSC between holding potentials of −40 and −70 mV in normal Ringer’s solution is shown. Note that the amplitude of the synaptic response increases with decreasing holding potentials, indicating that Mg2+ block of the channel is being relieved at depolarized potentials.
Fig. 2.
Fig. 2.
NMDA evokes a Ca2+ transient. Axons were loaded with the Ca2+ indicator dye OGB-1. NMDA (500 μm) was washed into the bath in the presence of TTX (1 μm), causing an increase in fluorescence in the axons.A, Normalized fluorescence of one axon was recorded before, during, and after wash-in of NMDA.B, Normalized fluorescence of seven axons is shown before, during, and after NMDA wash-in.C, NMDA wash-in will depolarize the axons by ∼2–3 mV, which is not a substantial enough depolarization to allow Ca2+ entry through VOCCs. A representative axon recorded using a sharp microelectrode (recording made in 1 μm) TTX) is shown. Bath application ofNMDA led to a small depolarization (3 mV) in this example, followed by an afterhyperpolarization.
Fig. 3.
Fig. 3.
Ca2+ transients recorded in axons are sensitive to Mg2+ and AP-5.A, An axon was filled with OGB-1 dye, and we recorded the response to a train of stimuli (50 Hz stimulation for 0.5 sec) applied to the spinal cord. Left, The baseline fluorescence. Middle, The peak fluorescence.Right, The fluorescence 5 sec after stimulation (stim.). The field is 53 × 80 μm in size.B, The normalized brightness within the axon is graphed at 0.5 sec intervals. The stimulus was delivered at the time indicated by the horizontal bar. Responses in normal superfusate, in Mg2+-free superfusate, after addition ofAP-5 (50 μm) to the Mg2+-free superfusate, and after wash-out ofAP-5 are shown. The images in A were taken at the times indicated in the graph (pre,peak, and post). C, The normalized brightness for a group of axons is shown. The stimulus was delivered at the time indicated by the horizontal bar. Responses in normal Ringer’s solution (n = 12), in Mg2+-free Ringer’s solution (n= 12), after addition of AP-5 (50 μm;n = 6) to the Mg2+-free Ringer’s solution, and after wash-out of AP-5(n = 3) are shown. Graph data are mean ± SE. There is a significant difference in the peak brightness between responses measured in control versus Mg2+-free (p < 0.05) and in Mg2+-free versus AP-5(p < 0.02).
Fig. 4.
Fig. 4.
Ca2+ transients evoked by direct and indirect stimulation. A, Schematic of recording setup showing stimulating electrodes (stim) positioned to activate reticulospinal axons directly and indirectly. Imaging was performed from the region within the box. B, Images of axonal Ca2+ recorded before (a,d), during (b, e), and after (c, f) stimulation of the spinal cord. ac, The response to indirect stimulation. df, The response to direct stimulation. Each panel is 105 × 70 μm large. The inset shown above exploded from b shows a pseudocolor representation of the normalized fluorescence change during the stimulus. This data compares b with the prestimulus condition in a Ci, Normalized fluorescence from thebottom-most axon shown in the field inB. The response of the axon to indirect stimulation is shown for images recorded in normal Ringer’s solution and in the presence of AP-5 (50 μm).Cii, Normalized fluorescence from the same axon in response to direct stimulation. Fluorescence from images recorded in normal Ringer’s solution and in the presence of AP-5(50 μm) are shown. Di, In another preparation, an investigation of the response of a single axon to one indirect stimulus. A single line was repetitively scanned along a length of the axon at 500 Hz. The spinal cord adjacent to the axon was stimulated to evoke a Ca2+transient. Dii, The graphs were generated by measuring the fluorescence at each time point and normalizing this to the prestimulus level to show the rise in axonal Ca2+. The addition of AP-5 (50 μm) significantly reduced the amplitude of the evoked transient. The response recovered to control amplitude after washout ofAP-5.
Fig. 5.
Fig. 5.
Pooled data comparing Ca2+transients evoked by direct and indirect stimulation. A, Normalized fluorescence recorded in nine axons in response to indirect and direct stimulation. B, Normalized fluorescence from four axons in response to indirect stimulation in normal Ringer’s solution or in the presence of AP-5 (50 μm). C, Normalized fluorescence from the same four axons in response to direct stimulation recorded in normal Ringer’s solution or in the presence of AP-5 (50 μm). Note that the Ca2+ transient recorded in response to indirect stimulation is sensitive toAP-5, whereas the transient recorded in response to direct stimulation is not. Data are expressed as the mean ± SE. SEs for B and C are calculated relative to the control response.
Fig. 6.
Fig. 6.
Ca2+ transients evoked by presynaptic action potentials differ from those evoked by EPSPs.A, Schematic of recording setup showing a microelectrode recording from an axon and a stimulating electrode positioned to activate reticulospinal axons indirectly, i.e., synaptically. Imaging was performed from the region within the box.B, Axonal responses recorded with the microelectrode. The resting potential of the axon was approximately −70 mV. These recordings were made simultaneously to the fluorescence recordings inC and D. Stimulus artifacts have been removed for clarity. Bi, Action potentials recorded in the axon in response to intracellular current injection (2 msec current injections at 50 Hz for 1.0 sec). Bii, EPSPs recorded in the axon in response to extracellular stimulation adjacent to the axon (1 msec shocks at 50 Hz for 1.0 sec). C, Images of axonal Ca2+ recorded before (pre stim.) and during (peak) stimulation.Right, The normalized fluorescence change between thepre stim. and peak conditions. Top row, The response to intracellular stimulation seen inBi. Bottom row, The response to extracellular stimulation seen in Bii. The microelectrode recording is from the axon marked by thearrowheads. This is the only axon that shows a response to intracellular stimulation as is clear in the normalized image. Eachpanel is 158 × 106 μm in size.Di, Normalized fluorescence for the axon recorded by the microelectrode during extracellular stimulation. The axonal fluorescence in response to extracellular stimulation is shown for images recorded before and after addition of AP-5 (50 μm). Dii, The normalized fluorescence recorded from the same axon in response to intracellular stimulation shown for images recorded before and after addition ofAP-5 (50 μm).
Fig. 7.
Fig. 7.
Activation of synaptic input to reticulospinal axon facilitates transmission from reticulospinal axon to motoneuron. A paired recording was made between a presynaptic reticulospinal axon and a postsynaptic motoneuron. The axon was held under current clamp with a microelectrode, and the motoneuron was held under voltage clamp with a patch electrode. A, Schematic to describe the recording arrangement. The prestimulus was provided via an extracellular stimulating electrode placed lateral to the recorded pair.B, The firing of an action potential by an axon in response to a brief (2 msec) depolarizing current injection (top). This evoked an EPSC in the postsynaptic motoneuron (bottom). C, Recordings from the same pair of neurons shown in B. Before current was injected to elicit an action potential in the axon, the spinal cord was stimulated to elicit an EPSP in the axon (prestim). This stimulation also elicited EPSCs in the postsynaptic neuron. After the prestimulus, current was injected into the axon to elicit an action potential and a resulting EPSC in the postsynaptic neuron. D, Enlargement of the postsynaptic response to the presynaptic action potential before and after the application of a prestimulus. The response is clearly enhanced by the prestimulus. All traces are averages of four sequential responses.
Fig. 8.
Fig. 8.
Activation of synaptic input to reticulospinal axon facilitates transmission from reticulospinal axon to motoneuron. A, A paired recording made between a presynaptic reticulospinal axon and a postsynaptic motoneuron (as in Fig. 7). The axon was held under current clamp with a microelectrode, and the motoneuron was held under voltage clamp with a patch electrode.Ai, Action potential evoked in the axon in response to current injection (top). The motoneuron responds to the presynaptic action potential with a mixed chemical and electrical EPSC (bottom). Aii, Enlargement of the responses of the postsynaptic neuron to the presynaptic action potentials. Individual traces are shown ingray, and the average of the four tracesis shown in black. There is a failure rate of 63% for the chemical component. Bi, Recordings from the same pair of neurons shown in A. Before current was injected to elicit an action potential in the axon (intracellular stimulus), the spinal cord was stimulated to elicit an EPSP in the axon (extracellular prestimulus). This stimulation also elicited EPSCs in the postsynaptic neuron. After the prestimulus, current was injected into the axon to elicit an action potential in the axon and a resulting EPSC in the postsynaptic neuron as described for Figure 7. Bii, Enlargement of the responses of the postsynaptic neuron to the presynaptic action potentials. The failure rate of the EPSCs elicited after the prestimulus is now 38%. Ci, The same protocol described in Bi performed in the presence ofAP-5 (50 μm). Cii, Enlargement of the responses of the postsynaptic neuron to the presynaptic action potentials. The failure rate of the EPSCs elicited after the prestimulus in the presence of AP-5 is increased to 54%.
Fig. 9.
Fig. 9.
Amplitude histograms of paired synaptic responses between a reticulospinal axon and a postsynaptic neuron. This data are from the pair of neurons illustrated in Figure 8. A, The EPSC amplitudes recorded after stimulation of only the presynaptic reticulospinal neuron. Left, A histogram of all event amplitudes. Those events marked in white were considered failures. Events marked gray were amplitudes of individual electrical events, in which the amplitude histogram showed a similar distribution to the recording noise. The blackevents were identified as chemical events. No event was included (either electrical or chemical) if there was no preceding electrical event. Right, Comparison of the rate of failures with the existence of a chemical response regardless of the amplitude of the chemical response. B, The EPSC amplitudes recorded on stimulation of the reticulospinal axon 100 msec after the application of a prepulse to the spinal cord. The panels are as described for A. The application of a prepulse markedly reduced the number of recorded failures. Note that the mean amplitude of the electrical component was slightly reduced. This effect reflects the synaptic drive from the prepulse on the postsynaptic neuron.C, EPSC amplitudes recorded similarly toB but after the addition of AP-5 (100 μm).
Fig. 10.
Fig. 10.
Model explaining the mechanism of presynaptic NMDA receptor enhancement of transmitter release. A, Schematic demonstrating the relationship between reticulospinal neurons and their postsynaptic targets. The electrodes shown represent the approximate placement of electrodes used to fill the neurons for the image in C. B, The axoaxonic input to the reticulospinal neuron activating ionotropic glutamate receptors, including NMDA receptors, on the presynaptic reticulospinal axon (blue). This causes a wide diffusion of Ca2+ in the axon (red). The extent of the Ca2+ signal can be >500 μm along the axon (as demonstrated in Fig. 6). This Ca2+ transient will alter the basal Ca2+ concentration at numerous synaptic terminals. In the schematic, the region of elevated calcium encompasses glutamatergic vesicles and Ca2+channels, and there is an enhancement of transmission onto postsynaptic neurons. C, Three-dimensional reconstruction of a synaptically coupled reticulospinal axon (dark vertical structure) and postsynaptic premotor interneuron. The pair was identified electrophysiologically using paired-microelectrode recording between the two cells as shown in the schematic in A. Both cells were filled with fluorescent dye (Lucifer yellow) by pressure ejection through the microelectrodes. Images were obtained using a confocal microscope. The boxed region of thetop is enlarged and rotated 90° on the bottom left, to view along the long axis of the axon. On thebottom right, two individual synaptic contacts (boxed region of the bottom left) are shown rotated at high magnification for improved visualization. This image demonstrates the relationship between the structure of the axons and the location of the en passant synaptic terminals. Each axon makes many overlapping synaptic contacts similar to the contact shown. Thus, an axoaxonically evoked Ca2+ rise that diffuses hundreds of micrometers along the axon will affect Ca2+ concentrations at numerous presynaptic terminals in the reticulospinal axon.

Similar articles

See all similar articles

Cited by 19 articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

Feedback