Differential control of synaptic and ectopic vesicular release of glutamate - PubMed (original) (raw)
Differential control of synaptic and ectopic vesicular release of glutamate
Ko Matsui et al. J Neurosci. 2004.
Abstract
Exocytosis of synaptic vesicles occurs not only at synaptic active zones but also at ectopic sites. Ectopic exocytosis provides a direct and rapid mechanism for neurons to communicate with glia that does not rely on transmitter spillover from the synaptic cleft. In the cerebellar cortex the processes of Bergmann glia cells encase synapses between presynaptic climbing fiber varicosities and postsynaptic Purkinje cell spines and express both AMPA receptors and electrogenic glutamate transporters. AMPA receptors expressed by Purkinje cells and Bergmann glia cells are activated predominantly by synaptic and ectopic release, respectively, and therefore can be used to compare the properties of the two release mechanisms. We report that vesicular release differs at synaptic and ectopic sites in the magnitude of short-term plasticity and the proportions of Ca2+ channel subtypes that trigger glutamate release. High-affinity glutamate transporter-mediated currents in Bergmann glia cells follow the rules of synaptic release more closely than the rules of ectopic release, indicating that the majority of glutamate is released from conventional synapses. On the other hand, ectopic release produces high-concentration glutamate transients at Bergmann glia cell membranes that are necessary to activate low-affinity AMPA receptors rapidly. Ectopic release may provide a geographical cue to guide Bergmann glia cell membranes to surround active synapses and ensure efficient uptake of glutamate that diffuses out of the synaptic cleft.
Figures
Figure 1.
PPR recovery of CF-PC EPSCs and CF-BG responses in 2 m
m
Ca2+. A, Shown are pairs of CF-PC EPSCs evoked at interstimulus intervals (ISIs) indicated at the top of the traces. Paired stimuli were separated by 15-25 sec. The application of 2 m
m
γ-
d
-GG (thick line) decreased the amplitude of the first EPSC but reduced the second EPSC further, causing a decrease in the paired-pulse ratio (PPR = the amplitude of the second EPSC divided by the first EPSC).For ISIs shorter than 50 msec, the EPSCs overlapped. The second EPSC was isolated by subtracting the average of an unpaired EPSC (Vh = -10 mV at 32-35°C). B, Pairs of CF-BG responses were evoked at the same ISIs as above. NBQX (10 μ
m
) blocked AMPAR-mediated currents, isolating the STCs (thick line). Subtraction of the STCs from the control traces (thin line) yielded the AMPAR-mediated component of the evoked response. The STC could be abolished completely with the application of 50 μ
m
TBOA (dotted line). Other illustrated records are the result of subtracting traces recorded at the end of the experiment in NBQX and TBOA to reduce contamination by the stimulus artifact and the small but prolonged potassium component (Bergles and Jahr, 1997) (Vh = -65 mV). C, PPR of CF-PC EPSCs plotted as a function of ISI for control and γ-
d
-GG conditions. Each data point represents the data in 3-17 cells. Note that the _x_-axis is in log scale. Error bars in this and all other figures indicate SD. D, PPR of CF-BG responses plotted as a function of ISI for the AMPAR-mediated current and STC component. Each data point represents the data in three to seven cells.
Figure 2.
PPR recovery of CF-PC EPSCs and CF-BG responses in 0.5 m
m
Ca2+. A, Paired-pulse responses of CF-PC EPSCs in 0.5 m
m
Ca2+ and 2.8 m
m
Mg2+ (Vh = -10 mV). B, Paired-pulse responses of CF-BG responses in 0.5 m
m
Ca2+, 2.8 m
m
Mg2+, and 200 μ
m
CTZ (Vh = -65 mV). These responses are mediated predominantly by AMPARs, because there is very little contribution by the STC in this condition (Matsui and Jahr, 2003). C, PPR of CF-PC EPSCs plotted as a function of ISI (n = 3-20 cells). D, PPR of CF-BG responses plotted as a function of ISI (n = 4-35 cells).
Figure 3.
Responses of PCs and BGs to iontophoresis of glutamate. A, Responses of a PC and a BG to paired iontophoretic pulses (5 msec, -5 to approximately -50 nA, with the braking current of +1.2 to approximately +1.7 nA; 200 to ∼400 MΩ iontophoretic electrode) of glutamate (1
m
, pH 8.0) at an ISI of 150 msec (PC, Vh = -10 mV; BG, Vh = -65 mV). Extracellular solution contained 0.5 m
m
Ca2+, 2.8 m
m
Mg2+, and 200 μ
m
CTZ to match the condition of Figure 2. The iontophoresis electrodes were placed in the molecular layer at locations that produced large responses. Currents were blocked with 10 μ
m
NBQX in both cell types (data not shown). In the illustrated recordings, three intensities of pulses were applied for each cell. B, Summary of the PPR of the responses to iontophoresis pulses. The open bar represents the average PC PPR (n = 4), and the filled bar represents the average BG PPR (n = 5). A very small amount of facilitation was observed in both cells, which suggests that a small amount of “tip-warming effect” occurs at the iontophoresis electrode tip, causing a slightly larger amount of glutamate to be released on the second pulse.
Figure 4.
PPR recovery of PF-PC EPSCs and PF-BG responses. A, Paired PF-PC EPSCs in 2 m
m
Ca2+ and 1.3 m
m
Mg2+ (Vh = -30 mV). B, Paired PF-BG responses in 2 m
m
Ca2+, 1.3 m
m
Mg2+, and 200 μ
m
CTZ (Vh = -65 mV). PF-BG responses were recorded in the presence of CTZ to block desensitization, augment the AMPAR response, and minimize the proportion of the PF-BG current caused by glutamate transport (Dzubay and Jahr, 1999). Therefore, in the presence of CTZ, most of the PF-BG response is mediated by AMPARs and could be blocked by 10 μ
m
NBQX (data not shown). C, PPR of PF-PC EPSCs versus ISI (n = 3-5 cells). D, PPR of PF-BG responses versus ISI (n = 3-9 cells).
Figure 5.
Glutamate transporters do not cause a large paired-pulse depression in CF-BG responses. A, CF-BG responses to paired-pulse stimuli with an ISI of 150 msec in 0.5 m
m
Ca2+, 2.8 m
m
Mg2+, and 200 μ
m
CTZ (Vh = -65 mV). The application of 30 μ
m
TBOA (thick line) increased both the first and the second response but had no effect on the PPR. B, Summary of the PPR of CF-BG response before (open bar) and after the application of TBOA (filled bar; n = 4).
Figure 6.
PPR difference between PCs and BGs persists in paired recordings. A, Paired recording from a PC (Vh = -10 mV) and a BG (Vh = -65 mV) in 2 m
m
Ca2+, 1.3 m
m
Mg2+, and no CTZ. Paired stimuli (timing shown by arrowheads) were given at an interval of 150 msec; stimulus strength was set near threshold so that some failures of CF stimulation occurred. The CF-evoked responses either succeeded or failed at the same time in the two cells, indicating that the two cells share the same CF. B, Responses to paired-pulse stimulation from the same cells as in A, with the extracellular solution changed to 0.5 m
m
Ca2+, 2.8 m
m
Mg2+, and 200 μ
m
CTZ. Notice that the PPR difference in the two cells persists at this condition. C, Summary of the PPR at an ISI of 150 msec in paired recordings in the same condition as in B. Open bar is for CF-PC EPSC, and filled bar is for CF-BG responses (n = 7 cell pairs; paired t test, *p < 0.001).
Figure 7.
CF-PC EPSCs and CF-BG responses are affected differentially by subtype-specific Ca2+ channel blockers. A, CF-PC EPSCs were recorded in 0.5 m
m
Ca2+ and 2.8 m
m
Mg2+ (Vh = -10 mV), and CF-BG responses were recorded in the same divalents with 200 μ
m
CTZ (Vh = -65 mV) in the absence and presence of the P/Q-type Ca2+ channel blocker ω-agatoxin IVA (200 n
m
). B, Summary of the ratio of the response amplitudes in the presence and absence of ω-agatoxin IVA in A. CF-PC EPSC (n = 5); CF-BG response (n = 5). C, CF-PC EPSCs and CF-BG responses were recorded in the same conditions as in A in the absence and presence of the N-type Ca2+ channel blocker ω-conotoxin GVIA (1 μ
m
). D, Summary of the ratio of the response amplitude in the presence and absence of ω-conotoxin GVIA in C. CF-PC EPSC (n = 4); CF-BG response (n = 4); unpaired t test, *p = 0.002. E, ω-Conotoxin GVIA (1 μ
m
) partially blocked the CF-PC EPSC but mainly blocked the initial fast component of the CF-BG response, whereas the late slow component was blocked less (gray lines in BG recordings are double exponential fits). CF-PC EPSCs were recorded in 2 m
m
Ca2+, 1.3 m
m
Mg2+, and 2 m
m
γ-
d
-GG (Vh = -10 mV); CF-BG responses were recorded with the same divalents without γ-
d
-GG (Vh = -65 mV). F, Summary of the ratio of response amplitudes in the presence and absence of ω-conotoxin GVIA in E. CF-PC EPSC (n = 5); CF-BG response (n = 5); ANOVA and Fisher's PLSD post hoc test, significantly different for all combinations; *p < 0.004.
References
- Bergles DE, Jahr CE (1997) Synaptic activation of glutamate transporters in hippocampal astrocytes. Neuron 19: 1297-1308. -PubMed
- Beutner D, Voets T, Neher E, Moser T (2001) Calcium dependence of exocytosis and endocytosis at the cochlear inner hair cell afferent synapse. Neuron 29: 681-690. -PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Miscellaneous