Recruitment of N-Type Ca(2+) channels during LTP enhances low release efficacy of hippocampal CA1 perforant path synapses

Abstract

The entorhinal cortex provides both direct and indirect inputs to hippocampal CA1 neurons through the perforant path and Schaffer collateral synapses, respectively. Using both two-photon imaging of synaptic vesicle cycling and electrophysiological recordings, we found that the efficacy of transmitter release at perforant path synapses is lower than at Schaffer collateral inputs. This difference is due to the greater contribution to release by presynaptic N-type voltage-gated Ca(2+) channels at the Schaffer collateral than perforant path synapses. Induction of long-term potentiation that depends on activation of NMDA receptors and L-type voltage-gated Ca(2+) channels enhances the low efficacy of release at perforant path synapses by increasing the contribution of N-type channels to exocytosis. This represents a previously uncharacterized presynaptic mechanism for fine-tuning release properties of distinct classes of synapses onto a common postsynaptic neuron and for regulating synaptic function during long-term synaptic plasticity.

Figure 1. Two-photon imaging of FM 1-43 dye uptake and release at PP presynaptic terminals in acute hippocampal slices

(A) Experimental set-up and CA1 architecture. One electrode (labeled ‘1’) was used both for fEPSP recordings and FM 1-43 delivery and was placed in the inner half of stratum lacunosum moleculare (SLM). A stimulating electrode (labeled ‘2’), placed in SLM > 200 μm from the recording electrode, was used to elicit PP fEPSPs and to load and unload dye from PP presynaptic boutons. Dashed box shows approximate imaging field. In experiments on SC synapses we used a similar arrangement but placed the recording (FM 1-43) and stimulating electrodes in stratum radiatum (SR) (not shown). (B) Protocol for assaying FM 1-43 release. Terminals were loaded with dye by a 2 min period of 10 Hz stimulation of PP inputs in presence of D-AP5 (50 μM) (and 10 μM CNQX in some experiments). ADVASEP-7 (ADV-7) was then applied to remove extracellular FM 1-43. Next, PP inputs were stimulated at 1.5 Hz for 4 min and then at 10 Hz for 2 min to release dye from terminals (Unload). (C) Two-photon images of fluorescent puncta in SLM loaded with FM 1-43 before (-120 s and -60 s), during (60 s, 240 s and 360 s), and after (435 s) unloading stimulation. Unloading started at 0 s. During 240-360 s interval, terminals were maximally destained with 10 Hz stimulation. (D) Time course of normalized FM 1-43 fluorescence intensity of individual boutons during 1.5 Hz unloading stimulation from experiment in C. (E) Histogram of FM 1-43 destaining rates (1/τ) from individual terminals in SLM (N=218 boutons, 5 slices). See Experimental Procedures for details.

Figure 2. PP synapses exhibit slower kinetics of FM 1-43 release and larger paired-pulse facilitation compared to SC synapses

(A) Mean time course of FM 1-43 destaining for PP (black triangles) and SC (red squares) terminals. Error bars (SEM in all figures) are shown and are smaller than symbols for most points. (B) Cumulative probability histograms of FM 1-43 destaining rates (1/τ) from individual PP (black, N=133 boutons, 4 slices) and SC (red, N=276 boutons, 6 slices) terminals differ significantly (P=0.001, K-S test). (C) fEPSPs elicited by paired-pulse stimulation (50 ms ISI) of PP (black) and SC (red) inputs, recorded in SLM and SR, respectively. Top, average fEPSP from 3 consecutive sweeps. Bottom, average fEPSPs normalized to the peak of the first fEPSP of each pair to better show facilitation during second fEPSP. (D) Mean paired-pulse ratio (PPR = fEPSP2/fEPSP1) of PP (black triangles, n=9 slices) and SC (red squares, n=8 slices) inputs across a range of ISIs. Asterisks indicate significant differences (double asterisk, P<0.001; single asterisk, P<0.01; unpaired t-test).

Figure 3. Differences in presynaptic function between PP and SC synapses are due to a differential contribution of N-type Ca²⁺ channels to release

(A) Time course of mean normalized PP (black symbols, n=6 slices) and SC (red symbols, n=4 slices) fEPSP slope during block of P/Q-type channels with 1 μM ω-Aga IVA. (B) Time course of mean normalized PP (black symbols, n=14 slices) and SC (red symbols, n=15 slices) fEPSP slope during block of N-type channels with 1 μM ω-Ctx GVIA. (C) Mean time course of FM 1-43 destaining from PP and SC terminals under control conditions (PP: black filled triangles; SC: red filled circles) and with 1 μM ω-Ctx GVIA present during unloading (PP: blue open triangles; SC: orange open circles). (D) Cumulative probability histograms of FM 1-43 destaining rates (1/τ) under control conditions for PP (solid black line, n=116 boutons) and SC terminals (solid red line, n=102 boutons) and in presence of 1 μM ω-Ctx GVIA for PP (dotted blue line, n=86 boutons) and SC terminals (dotted orange line, n=114 boutons). (E) Mean rates of FM 1-43 destaining (1/τ) under the conditions in panels C, D. Asterisks, significant difference compared to PP control group (P<0.002, ANOVA with Tukey’s post hoc pairwise comparisons). (F) Mean fraction of nondestaining boutons per microscopic field of view under conditions of panels C, D. Differences between groups are not significant (P=0.68; ANOVA). Number of slices in each group is shown in data bars.

Figure 4. LTP induced at PP synapses with 200 Hz tetanization enhances the kinetics of FM 1-43 release

(A) Protocol showing periods of PP stimulation, FM 1-43 iontophoresis, and applications of ADVASEP-7 and D-AP5. (B) Time course of LTP induced with 200 Hz tetanus protocol (protocol lasted 50 s starting at time 0). Mean normalized fEPSP slope is plotted between Unload 1 and Load 2 periods in A (n=5 slices). (C) Mean FM 1-43 destaining time course measured ~40 min before (Unload 1, black squares; averaged from 176 boutons from 5 slices) and ~80-90 min after induction of LTP (Unload 2, red squares; averaged from 129 individual boutons from the same 5 slices). (D) Cumulative probability histograms of FM 1-43 destaining rates of individual boutons (1/τ) before (black trace, n = 176 boutons) and after (red trace, n = 129 boutons) induction of LTP differ significantly (P<0.001; K-S test). FM 1-43 data in C,D is from same experiments shown in B. (E) Time course of average FM 1-43 destaining in interleaved, non-tetanized control experiments (n=5 slices; Unload 1, open black squares; Unload 2, open red squares). (F) Cumulative probability histograms of FM 1-43 destaining rates in non-tetanized, control experiments showed no significant difference between Unload 1 (black trace, n=229 boutons) and Unload 2 (red trace, n=130 boutons; P=0.38; K-S test).

Figure 5. Data from individual slices show that LTP enhances the kinetics of FM 1-43 release without altering vesicle pool size or number of dye-loaded terminals

For all panels, small symbols show average data for boutons from individual slices (same microscopic field in Unload 1 and Unload 2); large symbols show mean data from all slices in each group. (A) Effect of 200 Hz tetanic stimulation on FM 1-43 destaining rates (1/τ). Filled squares, experiments where 200 Hz tetanus was applied after Unload 1. Open squares, interleaved control experiments where no tetanus was delivered. Significant differences were only seen between Unload 1 and Unload 2 destaining rates for tetanized slices (P<0.02; paired t-test). (B) Mean total fluorescence intensity per bouton, a measure of recycling vesicle pool size, with (filled symbols) or without (open symbols) 200 Hz tetanus after Load 1. Fluorescence intensity between Load 1 and Load 2 was not altered by tetanus (P=0.28; unpaired t-test). (C) The total number of boutons that load with dye with (filled symbols) or without (open symbols) delivery of tetanus after Load 1. Number of boutons loaded with dye was not altered by tetanus (P=0.91; unpaired t-test).

Figure 6. Induction of LTP with 200 Hz tetanization decreases paired-pulse facilitation

(A) PP fEPSPs elicited by paired-pulse stimulation with 50 ms ISI ~15 min before (PPR 1, black traces) and ~40 min after (PPR 2, red traces) induction of LTP. Top traces, average fEPSPs from 5 consecutive sweeps. Bottom traces, average fEPSPs normalized to the peak of the first fEPSP. (B) Paired-pulse ratios (50 ms ISI) for individual slices before (PPR 1) and after (PPR 2) induction of 200 Hz LTP (filled squares, n=14) or for interleaved, non-tetanized, control slices (open squares, n=8). Small symbols, average PPR of 3-5 consecutive sweeps from individual slices. Large symbols, mean of all slices in each group. Asterisk indicates significant difference (P<0.05) between PPR 2 and PPR 1 in tetanized slices (paired t-test). (C) Fractional change in PPR following tetanus (PPR2/PPR1) versus magnitude of LTP (~40 min post-tetanus) for tetanized slices. Dashed red line shows linear fit (P<0.01; R=-0.7, Pearson’s linear regression).

Figure 7. Enhancement in the kinetics of FM 1-43 release and decrease in PPF following 200 Hz tetanization share same pharmacological profile as PP LTP

(A) Mean normalized PP fEPSP in response to 200 Hz tetanus (at 0 min, arrow) in absence or presence of pharmacological inhibitors. Red squares, ACSF alone (no inhibitors). Blue triangles, 50 μM D-AP5, applied during LTP induction protocol (black bar). Orange circles, 20 μM nitrendipine, present throughout experiment. Nitrendipine had no effect on basal synaptic transmission (data not shown). Gray diamonds, joint presence of 20 μM nitrendipine (throughout experiment) and 50 μM D-AP5 (black bar). Black triangles, ACSF alone without tetanus. Symbol key and color code apply to all panels. (B) Mean time course of FM 1-43 destaining ~80-90 min after 200 Hz tetanus in absence or presence of inhibitors as in A. Black triangles, no tetanus. (C) Cumulative probability histograms of FM 1-43 destaining rates (1/τ) showing effects of 200 Hz tetanus with or without pharmacological inhibitors. Legend gives number of boutons in each group. (D) Effect of pharmacological treatments on LTP magnitude measured as fEPSP slope ~40 min post-tetanus, expressed as percent of fEPSP slope during pretetanus baseline. No Tet group (black data bar) shows fEPSP measured at comparable times without tetanization. (E) Mean rates of FM 1-43 destaining (1/τ) measured ~80-90 min following 200 Hz tetanus, with or without indicated inhibitors. Destaining rate was measured at a comparable time in No Tet controls (black data bar). (F) Mean fractional change in PPR after LTP, PPR2 (after tetanus) divided by PPR1 (before tetanus), with or without inhibitors. (D-F) Numbers of slices in each group are shown in data bars. Single asterisks indicate significant difference compared to ACSF control group; double asterisks indicate significant difference between specified groups (P<0.05 in all comparisons; ANOVA with Tukey’s post hoc pairwise comparisons).

Figure 8. Increased contribution of N-type Ca²⁺ channels during expression of LTP accounts for decrease in PPF and enhancement of FM 1-43 release kinetics

(A) Time course of mean normalized fEPSP and (B) mean normalized PPR for slices receiving 200 Hz tetanus (at 0 min, arrow) without (Tet, red squares, n=6) or with application of 1 μM ω-Ctx GVIA (Tet, +Ctx, black circles, n=8). Toxin was applied ~38 min after tetanus (black bar). In control experiments, 1 μM ω-Ctx GVIA was applied without preceding tetanus (No Tet, +Ctx, open circles, n=6). For each time point in (B), asterisks indicate significant difference between starred group and other groups at that time point (P<0.01 at ~35 min; P<0.05 at ~65 min; ANOVA). (C) Mean FM 1-43 destaining time course ~80-90 min after tetanus in absence (Tet, red squares, n=291 boutons) or presence of 1 μM ω-Ctx GVIA (Tet, +Ctx, blue circles, n=146 boutons). Destaining data without tetanization (No Tet) obtained in absence (No Tet, black squares, n=152 boutons) or presence of 1 μM ω-Ctx GVIA (No Tet, +Ctx, gray circles, n=174 boutons). Slices were loaded with dye in absence of toxin. (D) Mean FM 1-43 destaining rates ~80-90 min post-tetanus. Number of slices in each group are shown in data bars. Asterisks indicate significant difference from tetanized group in absence of toxin (Tet) (P<0.01; ANOVA with Tukey’s post hoc pairwise comparisons). (E) Time course of mean normalized fEPSP with or without tetanus followed by successive applications of 1 μM ω-Aga IVA, high Ca²⁺ (6 mM), and 1 μM ω-Ctx GVIA. Solid circles, 200 Hz tetanus delivered at 0 min (Tet, n=4). Open circles, no tetanus given (No Tet, n=4).