A delayed response enhancement during hippocampal presynaptic plasticity in mice

Abstract

High frequency afferent stimulation of chemical synapses often induces short-term increases in synaptic efficacy, due to increased release probability and/or increased supply of readily releasable synaptic vesicles. This may be followed by synaptic depression, often caused by vesicle depletion. We here describe an additional, novel type of delayed and transient response enhancement phase which occurred during prolonged stimulation at 5-20 Hz frequency of excitatory glutamatergic synapses in slices from the adult mouse CA1 hippocampal region. This second enhancement phase, which was most clearly defined at physiological temperatures and essentially absent at 24 degrees C, was dependent on the presence of F-actin filaments and synapsins I and/or II, and could not be ascribed to changes in presynaptic action potentials, inhibitory neurotransmission or glutamate receptor desensitization. Time course studies showed that the delayed response phase interrupted the synaptic decay 3-4 s after stimulus train initiation and continued, when examined at 5-10 Hz frequencies, for approximately 75 stimuli before decay. The novel response enhancement, probably deriving from a restricted pool of synaptic vesicles, may allow maintenance of synaptic efficacy during prolonged periods of excitatory synaptic activity.

Figure 1

A delayed response enhancement (DRE) at 20 Hz stimulation in hippocampal excitatory CA3-to-CA1 synapses A, normalized fEPSP slope measurements during afferent stimulation in stratum radiatum at 0.1 Hz, followed by 20 Hz for 60 s (indicated by black bar, note difference in time scales), and reversal to 0.1 Hz, obtained in an experiment performed at 29°C during NMDA receptor blockade (see Methods). The inset shows superimposed synaptic responses at the stimulation times indicated by arrows and numbers. B, normalized and pooled fEPSP slope measurements during the initial 30 s of 20 Hz stimulation (boxed area in A). Red circles indicate experimental results obtained at 29°C (n = 22); blue indicates a simulated second-order exponential decay based on the normalized and pooled fEPSP slope measurements from response number 20 to response number 1180; green indicates a similar decay based on measurements from stimulus number 20–40 and 1160–1180 (see Methods). a, the time point of the maximum magnitude of the initial frequency facilitation; b, time to the transition point between the initial frequency facilitation and the DRE; c, time needed to reach the peak of the DRE. The inset graph shows a comparison of normalized and pooled intracellular EPSP slope measurements from CA1 pyramidal cell recordings (blue circles, n = 17) and fEPSP data (red circles, n = 31), both obtained during 20 Hz stimulation at 29°C in slices from mice at postnatal day 36–39. The sweeps represent superimposed intracellular responses at the stimulation times indicated by arrows and numbers. Vertical bars indicate s.e.m.

Figure 2

The DRE depends on the presence of presynaptic synapsin molecules A, effects of the absence of synapsin I/II on fEPSP slope measurements in the CA3-to-CA1 synapse during 20 Hz stimulation. Experiments were performed on slices from synapsin I/II double knock-out mice (blue circles, n = 18). Red trace indicates a simulated second order exponential decay based on the normalized and pooled fEPSP slopes from stimulus number 20–40 and 1160–1180 (see Methods). The inset shows superimposed synaptic responses at the stimulation times indicated by arrows and numbers. Vertical bars indicate s.e.m.B, normalized and pooled presynaptic volley amplitudes as a function of time during 20 Hz stimulation in wild-type (red circles, n = 17) and synapsin DKO (blue circles, n = 13) mice. The prevolley amplitudes were binned in groups of two, and the experiments were done either in the presence of 50 μm APV and 10 μm CNQX, or in the presence of 5 mm kynurenic acid. The inset shows superimposed prevolley responses in slices from wild-type (left) or DKO mice (right) at the stimulation times indicated by arrows and numbers. Vertical bars indicate s.e.m.

Figure 3

The DRE depends on an intact F-actin network A, fEPSP slope measurements from CA3-to-CA1 synapses during 20 Hz stimulation in slices treated with cytochalasin B (20 μm, n = 14, blue circles) or jasplakinolide (1 μm, n = 19, green circles) compared to experiments without these compounds (n = 22, red circles). Coloured horizontal bars along the abscissa, corresponding to the colours of the respective manipulations, indicate P < 0.05 when compared to the standard situation. Vertical bars indicate s.e.m. The inset shows simulated exponential decays based on the normalized and pooled fEPSP slope measurements from response number 30 to response number 1180 in slices from wild-type mice exposed to cytochalasin B (blue curved line) compared to slices from DKO mice in normal solution (black curved line; see Fig. 2A). Their respective time constants of decay are indicated. Ba, the time point of the maximum magnitude of the initial frequency facilitation (as indicated with arrow in A and with a colour code as in A); b, time to the transition point between the initial frequency facilitation and the DRE; c, time needed to reach the peak of the DRE. Vertical bars indicate s.e.m. *P < 0.05 when compared to the control situation.

Figure 4

Modulation of the DRE in excitatory CA3-to-CA1 synapses by temperature and [Ca²⁺]_o A, normalized and pooled experiments (20 Hz, 60 s, 2 mm CaCl₂) as described in Figure 1. Results were obtained at 37°C (blue circles, n = 24), 29°C (red circles, n = 22) and 24°C (green circles, n = 24). Coloured horizontal bars along the abscissa, corresponding to the colours of the respective curves, indicate P < 0.05 when compared to fEPSP slope values at 29°C. Ba, the time point of the maximum magnitude of the initial frequency facilitation (as indicated with arrow and with a colour code as in A); b, time to the transition point between the initial frequency facilitation and the DRE; c, time needed to reach the peak of the DRE. Bars indicate s.e.m. *P < 0.05 when compared to the control situation. C, experiments (20 Hz, 60 s, 29°C) done after equilibration to different extracellular concentrations of [Ca²⁺]_o. The results are presented relative to fEPSP slope values at 2 mm CaCl₂. Coloured horizontal bars along the abscissa corresponding to the respective curves indicate P < 0.05 when compared to fEPSP slope values at 2 mm CaCl₂. Green circles, 1 mm (n = 12), red circles, 2 mm (n = 22), and blue circles, 4 mm CaCl₂ (n = 11), respectively. D, as B, but the coloured bars and analytical results correspond to the curves in different [Ca²⁺]_o as shown in C. Bars indicate s.e.m. *P < 0.05 when compared to the results at 2 mm CaCl₂. E, subtraction of the values obtained at 1 mm (green circles) and 4 mm CaCl₂ (blue circles) from those obtained at 2 mm CaCl_2.

Figure 5

Frequency modulation of the DRE in excitatory CA3-to-CA1 synapses A, normalized and pooled experiments (29°C, 2 mm CaCl₂) as described in Fig. 1B. Results were obtained at a stimulation frequency of 20 Hz (red circles, n = 22), 10 Hz (blue circles, n = 20) and 5 Hz (green circles, n = 12). Coloured horizontal bars along the abscissa corresponding to the respective curves indicate P < 0.05 when compared to 20 Hz stimulation. Ba, the time point of the maximum magnitude of the initial frequency facilitation (as indicated with arrow and with a colour code as in A); b, time to the transition point between the initial frequency facilitation and the DRE; c, time needed to reach the peak of the DRE. Vertical bars indicate s.e.m. *P < 0.05 when compared to the control situation. C and D same as in A and B, but presented as a function of stimulus number. Bars indicate s.e.m. *P < 0.05 when compared to the results at 20 Hz.

Figure 6

Neither GABA-mediated receptor modulations nor AMPA receptor desensitization interfere with the shape of the DRE A, normalized and pooled fEPSP slope measurements during 20 Hz stimulation in the presence of the GABA_B receptor antagonist saclofen (50 μm, n = 19, blue circles) and the GABA_B receptor agonist baclofen (50 μm, n = 16, green circles). The results are presented relative to fEPSP slope values without GABA_B receptor interference (n = 22, red circles). Coloured horizontal bars along the abscissa correspond to the respective manipulations and indicate P < 0.05 when compared to the standard situation. Vertical bars indicate s.e.m.Ba, the time point of the maximum magnitude of the initial frequency facilitation (as indicated with arrow and with a colour code as in A); b, time to the transition point between the initial frequency facilitation and the DRE; c, time needed to reach the peak of the DRE. Bars indicate s.e.m. *P < 0.05 when compared to the control situation. C and D, as A and B, but experiments performed on slices treated with bicuculline (10 μm, n = 21, blue circles. E and F, as A and B, but experiments performed on slices treated with cyclothiazide (100 μm, n = 18, blue circles).