Differential Effect on Hippocampal Synaptic Facilitation by the Presynaptic Protein Mover

Abstract

Neurotransmitter release relies on an evolutionarily conserved presynaptic machinery. Nonetheless, some proteins occur in certain species and synapses, and are absent in others, indicating that they may have modulatory roles. How such proteins expand the power or versatility of the core release machinery is unclear. The presynaptic protein Mover/TPRGL/SVAP30 is heterogeneously expressed among synapses of the rodent brain, suggesting that it may add special functions to subtypes of presynaptic terminals. Mover is a synaptic vesicle-attached phosphoprotein that binds to Calmodulin and the active zone scaffolding protein Bassoon. Here we use a Mover knockout mouse line to investigate the role of Mover in the hippocampal mossy fiber (MF) to CA3 pyramidal cell synapse and Schaffer collateral to CA1. While Schaffer collateral synapses were unchanged by the knockout, the MFs showed strongly increased facilitation. The effect of Mover knockout in facilitation was both calcium- and age-dependent, having a stronger effect at higher calcium concentrations and in younger animals. Increasing cyclic adenosine monophosphate (cAMP) levels by forskolin equally potentiated both wildtype and knockout MF synapses, but occluded the increased facilitation observed in the knockout. These discoveries suggest that Mover has distinct roles at different synapses. At MF terminals, it acts to constrain the extent of presynaptic facilitation.

Keywords: CA3; hippocampus; mossy fiber; presynapse; short-term plasticity; synaptic transmission.

FIGURE 1

Global knockout of Mover. (A) Gene targeting strategy for Mover KO mice. (B) Results of the PCR used for genotyping. Primers P1, P2, and P3 shown in panel (A) were always used in the same reaction. When a WT and a KO allele were present, P1 and P3 produce a 697 bp band, P2 and P3 produce a 867 bp band (lane “Het”). When only WT alleles are present the primers produce only the 867 bp band (lane “WT”), when only KO alleles are present the primers produce only the 697 (lane “KO”). (C) Example of sequencing results for WT (top) and KO (bottom). Examples shown start from nucleotide 45 from sequencing result and show a part of intron 3 using the primer P2 for WT and the flox site followed by intron 3 in KO, showing the absence of exons 1–3. (D) Western blot of lysates from dissected hippocampi from WT (left) and KO mice (right) probed for β-Tubulin and Mover. (E) Immunofluorescence of WT (left) and KO (right) mouse brain sections stained for Mover and Synaptophysin.

FIGURE 2

Synaptic transmission is unchanged in SC in absence of Mover. (A) Diagram representing stimulation of Schaffer collaterals and extracellular recording at stratum radiatum of the CA1. (B) Example traces (left) and quantification (right) of fEPSP slopes versus fiber volley amplitude recorded at increasing stimulation strengths depicts an input–output relationship unchanged by the absence of Mover. (C) Paired-pulse ratios recorded from fEPSP amplitudes across varying inter-stimulus intervals show no difference between WT and KO. (D) Normalized responses to a 25 Hz train of stimulation showed unchanged facilitation in KO MF when compared to WT. Scale bars: vertical = 250 μV, horizontal = 20 ms. Error bars indicate SEM. WT n = 10–13; KO n = 12–18.

FIGURE 3

Increased paired-pulse ratio in MF in absence of Mover. (A) Diagram representing stimulation of mossy fibers and extracellular recording in the stratum lucidum of the CA3. (B) Example traces (left) and quantification (right) of fEPSP slopes versus fiber volley amplitude recorded at increasing stimulation strengths depicts an input–output relationship unchanged by the absence of Mover. (C) Paired-pulse ratios recorded from fEPSP amplitudes across varying inter-stimulus intervals reveal increased facilitation in KO when compared to WT. Scale bars: vertical = 250 μV, horizontal = 20 ms. Error bars indicate SEM. MF WT n = 11–18; KO n = 13–22. ^∗p < 0.05.

FIGURE 4

Absence of Mover does not interfere with miniature EPSC parameters in CA3 pyramidal cells. (A) Representative traces from WT (gray) and KO (red) CA3 pyramidal cells under presence of 1 μM TTX. (B) Representative miniature EPSC waveform from traces in A. (C) Amplitudes of miniature EPSC events were unchanged in their average amplitude (C₁) and in their cumulative probability (C₂). (D) Frequency of events was not changed by the absence of Mover. Miniature EPSC kinetics, namely the time constant of decay (E) and the 10–90 rise time (F), showed no difference between WT and KO. Error bars indicate SEM. WT n = 15; KO n = 11.

FIGURE 5

Whole-cell recordings confirm increased paired-pulse ratio in MF-CA3 pyramidal cells in Mover KO. (A) Example traces and quantification of single evoked NMDA EPSCs had similar amplitude between WT and KO. (B) Paired-pulse ratios recorded from NMDA EPSC amplitudes across varying inter-stimulus intervals show increased ratios in KO pyramidal cells. Scale bars: vertical = 100 pA, horizontal = 100 ms. Error bars indicate SEM. WT n = 9; KO n = 10. ^∗p < 0.05.

FIGURE 6

Increased facilitation in MF in the absence of Mover. (A) Example traces and normalized responses to 25 Hz trains of stimulation showed increased facilitation in KO MF when compared to WT. (B) Normalized responses to stimuli delivered at 0.05, 0.2, and 0.5 Hz reveal increased facilitation in KO MF. Responses were normalized to the amplitude of the first fEPSP. (C) Paired-pulse ratio with an inter-stimulus interval of 40 ms using the same stimulation frequencies as above (0.05, 0.2, and 0.5 Hz, see text). Scale bars: vertical = 500 μV, horizontal = 20 ms. Error bars indicate SEM. SC WT n = 10–13; SC KO n = 12–18; MF WT n = 18; KO n = 22. ^∗p < 0.05; ^∗∗∗p < 0.001.

FIGURE 7

Increased facilitation in KO is age- and calcium-dependent. MF of 8-week old animals KO (KO adult) has stronger facilitation than WT (WT adult) only in high extracellular calcium concentration. (A) Increasing calcium concentration leads to similar baseline responses in WT and KO. Responses were normalized to fEPSP amplitudes at 3.5 mM Ca²⁺. (B–D) Short-term plasticity at different extracellular calcium concentrations. (B₁) Normalized responses to stimuli delivered at 0.05, 0.2, and 0.5 Hz at 1.25 mM extracellular calcium. (B₂) Normalized responses to a 25 Hz train of stimulation at 1.25 mM extracellular calcium. (C₁) Normalized responses to stimuli delivered at 0.05, 0.2, and 0.5 Hz at 2.5 mM extracellular calcium, in four different conditions: 8-week old WT (adult), 8-week old KO (adult), 3-week old WT (WT young), and 3-week old KO (KO young). The last two conditions are the same dataset as presented in Figures 6A,B. (C₂) Normalized responses to a 25 Hz train of stimulation at 2.5 mM extracellular calcium. (D₁) Normalized responses to stimuli delivered at 0.05, 0.2, and 0.5 Hz at 3.5 mM extracellular calcium. (D₂) Normalized responses to a 25 Hz train of stimulation at 3.5 mM extracellular calcium. (B₁,C₁,D₁) Each dot represents the average response to five consecutive stimuli. (B–D) Responses were normalized to the amplitude of the first fEPSP. (Insets) Representative traces from WT (black) and KO (red) hippocampi. Scale bars: vertical = 1 mV, horizontal = 10 ms. Error bars indicate SEM. WT adult n = 13; KO adult n = 13. ^∗p < 0.05; ^∗∗∗p < 0.001.

FIGURE 8

Forskolin potentiation occlude KO boost in facilitation. (A) Normalized MF fEPSP amplitudes during the time course of experiment in which forskolin (fsk, 10 μM) is applied for 10 min and frequency of stimulation is changed from 0.05 to 0.2 and further to 0.5 Hz. Each data point corresponds to the average response to five consecutive stimuli. (Inset) Representative traces from WT (black) and KO (red) hippocampi. (B) Normalized MF responses to a 25 Hz train of stimulation in four different conditions: WT without forskolin application (WT-NoFsk), KO without forskolin application (KO-NoFsk), WT after forskolin potentiation (WT-Fsk), and KO after forskolin potentiation (KO-Fsk). The first two conditions are the same dataset as present in Figure 6B_. Responses were normalized to the amplitude of the first response. Scale bars: vertical = 200 μV, horizontal = 10 ms; WT n = 10; KO n = 10. Error bars indicate SEM. n.s.: not significant; ^∗∗∗p < 0.001.