Evidence for altered NMDA receptor function as a basis for metaplasticity in visual cortex

Abstract

Sensory deprivation alters the properties of synaptic plasticity induced in the superficial layers of the visual cortex, facilitating long-term potentiation and reducing long-term depression (LTD) across a range of stimulation frequencies. Available data are compatible with either a downregulation of the mechanisms of LTD or an upregulation of NMDA receptor function in the visual cortex of dark-reared animals. Here, we provide evidence for enhanced NMDA receptor function by showing that deprivation produces a horizontal shift in the frequency-response function, decreasing LTD in response to 1 Hz stimulation, but increasing LTD in response to 0.5 Hz stimulation. In addition, we show that the effects of dark-rearing on the frequency dependence of LTD can be reversed acutely by partial NMDA receptor blockade. Finally, we show that an in vivo manipulation that rapidly downregulates NMDA receptor function in the visual cortex, brief light exposure, also rapidly reverses the effect of dark-rearing on LTD.

Figure 1.

Alternative models for how visual experience can modify the frequency-response function measured in layers 2/3 and evoked by layer 4 stimulation in the slice preparation. A, In a vertical shift of the frequency-response function, LTD is difficult to induce with low frequencies of stimulation in the cortex of dark-reared animals (filled symbols) as compared with the cortex of normal, light-reared animals (open symbols). B, In a horizontal shift, substantial LTD can be induced in dark-reared cortex at very low frequencies of stimulation. Data points are adapted from Kirkwood et al. (1996).

Figure 2.

Dark-rearing exerts a horizontal shift in the frequency-response function; the magnitude of LTD is greater in dark-reared cortex than normal cortex with 0.5 Hz stimulation, whereas the converse is true with 1 Hz stimulation. A, Representative experiments (A₁, A₂) and summary graph (A₃) demonstrating the effects of a 30 min, 0.5 Hz stimulation on FP amplitude in dark-reared cortex (A₁) and light-reared cortex (A₂). B, Representative experiments (B₁, B₂) and summary graph (B₃) demonstrating the consequence of a 15 min, 1 Hz stimulation. DR, Dark-reared; LR, light-reared. Vertical scale bar, 0.4 mV (A₁, A₂, B₂) or 0.2 mV (B₁).

Figure 3.

Partial NMDA receptor blockade in deprived cortical slices alters the properties of synaptic depression. A, Representative experiments (A₁, A₂) and summary graph (A₃) demonstrating the effects of a 30 min, 0.5 Hz stimulation on FP amplitude in dark-reared cortex (A₁) and dark-reared cortex treated with 1 μm APV(A₂). B, Representative experiments (B₁, B₂) and summary graph (B₃) demonstrating the consequence of a 15 min, 1 Hz stimulation. DR, Dark-reared; DR+APV, dark-reared + 1 μm APV. Vertical scale bar, 0.4 mV.

Figure 4.

Brief LE (2 hr) modifies the induction of LTD with low-frequency stimulations. A, Representative experiments (A₁, A₂) and summary graph (A₃) demonstrating the effects of a 30 min, 0.5 Hz stimulation on FP amplitude in dark-reared cortex (A₁) and cortex from dark-reared animals given brief LE (A₂). B, Representative experiments (B₁, B₂) and summary graph (B₃) demonstrating the consequence of a 15 min, 1 Hz stimulation. DR, Dark-reared; LE, light-exposed. Vertical scale bar = 0.4 mV (A₂, B₁, B₂) or 0.2 mV (A1).

Figure 5.

Summary demonstrating that the light-reared LTD phenotype can be mimicked either by partially blocking NMDA receptor function in dark-reared visual cortex or by briefly exposing dark-reared animals to light (a procedure known to shorten NMDA receptor currents). A, The magnitude of synaptic depression induced by 0.5 or 1 Hz stimulation in dark-reared cortex can be reproduced across experiments. B, Summary graphs demonstrating the average magnitude (±SEM) of synaptic depression induced by 0.5 or 1 Hz stimulation in dark-reared (DR) cortex, APV-treated DR cortex (DR + APV), light-exposed DR cortex (DR + LE), and light-reared (LR) cortex. Dark-reared values represent averaged data across all three experiments.