Short-term synaptic depression and recovery at the mature mammalian endbulb of Held synapse in mice

Abstract

The endbulb of Held synapses between the auditory nerve fibers (ANF) and cochlear nucleus bushy neurons convey fine temporal information embedded in the incoming acoustic signal. The dynamics of synaptic depression and recovery is a key in regulating synaptic transmission at the endbulb synapse. We studied short-term synaptic depression and recovery in mature (P22-38) CBA mice with stimulation rates that were comparable to sound-driven activities recorded in vivo. Synaptic depression in mature mice is less severe ( approximately 40% at 100 Hz) than reported for immature animals and the depression is predominately due to depletion of releasable vesicles. Recovery from depression depends on the rate of activity and accumulation of intracellular Ca2+ at the presynaptic terminal. With a regular stimulus train at 100 Hz in 2 mM external [Ca2+], the recovery from depletion was slow (tauslow, approximately 2 s). In contrast, a fast (taufast, approximately 25 ms), Ca2+-dependent recovery followed by a slower recovery (tauslow, approximately 2 s) was seen when stimulus rates or external [Ca2+] increased. In normal [Ca2+], recovery from a 100-Hz Poisson-like train is rapid, suggesting that Poisson-like trains produce a higher internal [Ca2+] than regular trains. Moreover, the fast recovery was slowed by approximately twofold in the presence of calmidazolium, a Ca2+/calmodulin inhibitor. Our results suggest that endbulb synapses from high spontaneous firing rate auditory nerve fibers normally operate in a depressed state. The accelerated synaptic recovery during high rates of activity is likely to ensure that reliable synaptic transmission can be achieved at the endbulb synapse.

FIG. 1.

Activity rate and external [Ca²⁺] dependent synaptic depression. A: examples of a bushy neuron excitatory postsynaptic current (EPSC) response to trains of auditory nerve fiber stimulation at different rates. Traces are averages of 40 trials for each frequency; stimulus artifacts have been erased. B: evoked EPSCs normalized to the amplitude of the 1st EPSC at 100-, 200-, and 300-Hz shock stimulation in 2 mM external [Ca²⁺]. Superimposed lines are single-exponential fits of the data. C: normalized EPSC responses to 100-Hz stimulus trains at 1.5 and 2.5 mM external [Ca²⁺].

FIG. 2.

There is no relief of synaptic depression by 2 mM d-glutamylglycine (γ-DGG). A: evoked EPSC response is significantly reduced in the presence of 2 mM low-affinity, competitive AMPA receptor antagonist γ-DGG. Inset: sample traces of EPSCs recorded in the presence and absence of γ-DGG from a bushy neuron. B: γ-DGG has no effect on the normalized synaptic depression at 100- and 200-Hz stimulus rates. C: spontaneous miniature EPSCs before and immediately after 200-Hz shock stimulus trains. Notice the falling phase of EPSCs in postshock traces. Twenty-five individual traces were superimposed.

FIG. 3.

Recovery from depression depends on presynaptic firing rate. A: sample records of recovery from synaptic depression at different intervals after shock stimulus trains. Five trials for each recovery interval were recorded and EPSC responses were averaged. Synaptic depression traces were averaged from 30–40 trails. B: recovery from short-term depression after 100, 200, and 300 shock trains. Recovery time courses were fit with a bi-exponential function for 200- and 300-Hz trains, whereas a single-exponential function fit was used for 100-Hz shock trains. — and - - -, the fast and slow phase, respectively, derived from the bi-exponential fits. C: in a different set of cells, synaptic recovery was tested at a higher sampling rate within 100 ms immediately after the shock train at 100, 200, and 300 Hz. Data from 200- and 300-Hz recovery were fitted with a single-exponential function. Data from 100 Hz were fitted with a linear regression.

FIG. 4.

[Ca²⁺]_i dependence of rapid synaptic recovery. A: normalized synaptic depression under combinations of low (1.5 mM) or high (2.5 mM) external [Ca²⁺] and low (100 Hz) or high (200 Hz) stimulus rate. Error bars were omitted for 1.5 mM Ca²⁺/100-Hz and 2.5 mM Ca²⁺/200-Hz conditions in both A and B for clarity. Symbol legends in B apply to A. Inset: amplitudes of EPSC trains from varying external [Ca²⁺] and stimulus rates. B: recovery from synaptic depression under various combinations of external [Ca²⁺] and stimulus rates.

FIG. 5.

Effect of calmidazolium on rapid synaptic recovery. Group data of synaptic recovery from 200-Hz stimulus trains in the presence (n = 4) and absence (n = 7) of 20 uM calmidazolium, a membrane-permeable calmodulin inhibitor. *, t-test P values <0.05 for recovery levels at different time points.

FIG. 6.

Synaptic depression and recovery with 100-Hz regular and Poisson-like spike trains. A: sample EPSCs from 100-Hz Poisson-like spike trains. EPSCs from 100-Hz regular trains (faint traces) were superimposed in the 1st 100 ms expanded view to compare the rate of depression between 2 types of stimulus trains. B: the normalized mean EPSCs from both regular and Poisson-like trains at the end of 500-ms stimulation. C: synaptic recovery time course after the 100- and 200-Hz regular and Poisson-like trains.