Spike-timing-dependent plasticity of neocortical excitatory synapses on inhibitory interneurons depends on target cell type

Abstract

Repetitive correlated spiking can induce long-term potentiation (LTP) and long-term depression (LTD) of many excitatory synapses on glutamatergic neurons, in a manner that depends on the timing of presynaptic and postsynaptic spiking. However, it is mostly unknown whether and how such spike-timing-dependent plasticity (STDP) operates at neocortical excitatory synapses on inhibitory interneurons, which have diverse physiological and morphological characteristics. In this study, we found that these synapses exhibit target-cell-dependent STDP. In layer 2/3 of the somatosensory cortex, the pyramidal cell (PC) forms divergent synapses on fast spiking (FS) and low-threshold spiking (LTS) interneurons that exhibit short-term synaptic depression and facilitation in response to high-frequency stimulation, respectively. At PC-LTS synapses, repetitive correlated spiking induced LTP or LTD, depending on the timing of presynaptic and postsynaptic spiking. However, regardless of the timing and frequency of spiking, correlated activity induced only LTD at PC-FS synapses. This target-cell-specific STDP was not caused by the difference in the short-term plasticity between these two types of synapses. Activation of postsynaptic NMDA subtype of glutamate receptors (NMDARs) was required for LTP induction at PC-LTS synapses, whereas activation of metabotropic glutamate receptors was required for LTD induction at both PC-LTS and PC-FS synapses. Additional analysis of synaptic currents suggests that LTP and LTD of PC-LTS synapses, but not LTD of PC-FS synapses, involves presynaptic modifications. Such dependence of both the induction and expression of STDP on the type of postsynaptic interneurons may contribute to differential processing and storage of information in cortical local circuits.

Figure 1.

Identification of synaptic connections between PC and LTS or FS interneurons in layer 2/3 of the somatosensory cortical slices. A, B, Firing property and morphology of LTS and FS interneurons. Reconstruction of a typical biocytin-filled LTS (A) or FS (B) cell in layer 2/3 in rat somatosensory cortex (soma and dendrites in black, and axons in gray). Insets, Example recordings depicting the firing characteristics of the corresponding cell in response to a step depolarizing current. The threshold current (I_T) for initiating APs was ∼40 and ∼220 pA for the LTS cell and the FS cell, respectively (top traces). At a higher current step (I_H, 500 pA) that induced high-frequency spiking, the LTS cell exhibited more apparent frequency adaptation than the FS cell (bottom traces) Calibration: 20 mV, 50 ms. C, D, Example recordings showing short-term facilitation of PC-LTS synapses (C) and depression at PC-FS synapses (D) when a train of APs was elicited in the presynaptic PC. Calibration: top traces, 2 mV, 100 ms; bottom traces, 40 mV, 100 ms. E, F, Example recordings showing reciprocal IPSPs recorded from the PCs when single APs were evoked in the LTS (E) or FS (F) cell. Bath application of the GABA_A receptor antagonist bicuculline (5 μm) reversibly abolished the IPSPs. Calibration: top, 40 mV, 25 ms; bottom, 2 mV, 25 ms.

Figure 2.

Distinct spike-timing-dependent LTP/LTD at PC-LTS and PC-FS synapses. A, B, Example recordings from PC-LTS cell pairs showing persistent increase (A) and decrease (B) in the EPSP amplitude after repetitive correlated presynaptic and postsynaptic spiking (at t = 0 min; arrow) at intervals of +8 ms (A; pre-post pairing) and −8 ms (B; post-pre pairing), respectively. Top traces, Example recordings of membrane potential changes in the presynaptic PC and postsynaptic LTS during spike pairing at 20 Hz; the postsynaptic spikes were truncated. Calibration: PC traces, 100 mV, 25 ms; LTS traces, 12.5 mV, 25 ms. Horizontal lines indicate the mean value (including failures) over the duration covered by the line. EPSP traces above depict averaged EPSP over the duration indicated by the black line, with failure included as 0 mV. Calibration: 20 ms. D, E, Similar to that in A, B, except that the recordings were from PC-FS cell pairs. Persistent synaptic depression was induced after spike pairing at either +8 ms or −8 ms intervals. C, F, Summary of results from all experiments similar to that described in A, B, D, E, including data using spiking intervals of +25 and −25 ms. The data points represent mean EPSP amplitude over a 3 min bin, normalized by the mean value before the induction. n, Number of cell pairs examined. For the induction at ±25 ms, no significant change in the EPSP amplitude was observed for all cases (p > 0.1, paired t test). Error bars indicate SEM.

Figure 3.

Triplet recordings showed target-cell-specific STDP of divergent synapses made by the same PC on LTS and FS cells. A, An image of a representative triplet (i.e., one PC innervating one LTS cell and one FS cell, depicted by the dashed lines) in the layer 2/3 of the somatosensory cortical slice. White traces shown near the recorded cells depict a train of APs evoked in the PC by current injection and corresponding EPSPs evoked in the LTS or FS cell. B, Example recordings from the triplet shown in A: LTP (194% of the baseline, for data beyond 15 min after induction; top) at PC-LTS synapses and LTD (69% of the baseline; bottom) at PC-FS synapses, induced by the same repetitive correlated pre-post (+8 ms) spike pairing (arrow). Horizontal lines indicate the mean value (including failures) over the duration covered by the line. Traces above depict averaged EPSPs over the duration indicated by the line. Calibration: 20 ms. C, Results from five triplet experiments similar to that shown in B (PC-LTS synapses, 160 ± 13%; PC-LTS synapses, 76 ± 3%, relative to the baseline before induction). Data from the same triplet are connected by the line.

Figure 4.

Target-cell-specific STDP is independent of short-term synaptic plasticity. A, Examples of EPSPs evoked by the a train of five presynaptic spikes at 20 and 1 Hz at PC-LTS and PC-FS synapses. Traces were the average of 60 EPSPs. Calibration: 1 mV, 25 ms. B, Changes in the EPSP amplitude elicited by five spikes. Summary of results from all experiments shown in A. The amplitude of EPSPs was normalized by that of the first EPSP. Filled symbols, Spiking at 20 Hz; open symbols, spiking at 1 Hz. Error bars depict SEM. C, Time course of the change in EPSP amplitude before and after the 1 Hz pre-post spike pairing (+8 ms, 60 pulses) at PC-LTS (square) and PC-FS (circle) synapses. The amplitude of EPSPs was normalized by the mean value before spike pairing. n, Number of cell pairs examined.

Figure 5.

Spike-timing-dependent modification of PC-FS synapses under various frequency and timing of correlated spiking. A, Dependence on the spiking frequency at the same interval of correlated pre-post spiking (+8 ms). B, Dependence on the time interval of presynaptic and postsynaptic spiking at the same frequency (1 Hz). Each point represents result from one cell pair. Curves, Single-exponential fit with y = y_o + A × e^−x/τ [y, normalized EPSP values; x, timing intervals; for pre-post pairing, A = −0.46, τ = 39.9 ms (n = 18); for post-pre pairing, A = −0.56, τ = 39.1 ms (n = 16)]. Error bars depict SEM.

Figure 6.

Distinct roles of NMDARs and mGluRs in spike-timing-dependent LTP/LTD of PC-LTS and PC-FS synapses. A, B, Left, Cumulative percentage plot for the distribution of the mean EPSP amplitude 10–60 min after correlated spiking (at 20 Hz) at an interval of +8 ms at PC-LTS synapses, normalized by the mean amplitude before spiking, for experiments in the bath presence of various inhibitors (AP5, AP5 plus MCPG, AP5 plus AM251, MCPG, or AM251) or after intracellular loading of MK801 in LTS cells. Right, Time course of the average change in EPSP amplitude (over a 3 min bin in the baseline and over a 5 min bin after the pairing) before and after the +8 ms pre-post spike pairing (at 20 Hz) for the cells shown on the left. C, Similar to A except that −8 ms post-pre spike pairing (at 20 Hz) was used at PC-LTS synapses. D, Similar to C except that the experiments were performed at PC-FS synapses. Error bars indicate SEM.

Figure 7.

Distinct expression of STDP at PC-LTS and PC-FS synapses. A, Analysis of CV of evoked EPSPs showed changes at PC-LTS synapses, but not at PC-FS synapses after correlated spiking. CV = (σ/m), where σ and m are the SD and the mean of EPSP amplitudes. The values of σ and m measured for a duration of 10 min immediately before and 20∼30 min after correlated spiking were used to calculate CV_bef and CV_aft, respectively. Each pair of connected circles represents mean CV_bef and CV_aft obtained from individual experiments shown in Figure 2, C and F, and Figure 6A. Unmasked LTD, LTD found at PC-LTS synapses in experiments with +8 ms pairing in the presence of MK801 and AP5. A significant difference between CV_bef and CV_aft was observed at PC-LTS synapses (LTP of +8 ms pairing, p = 0.0013, n = 21; LTD of −8 ms pairing, p = 0.0065, n = 12; unmasked LTD, p = 0.0013, n = 18; paired t test), but not at PC-FS synapses (±8 ms pairing, p = 0.33, n = 41, paired t test). The histogram depicts the ratio of CV, determined as CV_bef/CV_aft. B, Change in the transmission failure rates at PC-LTS synapses accompanying LTP and LTD induction from experiments shown in Figure 2C, as well as unmasked LTD from experiments shown in Figure 6A. The failure rates were measured for 10 min before correlated spiking and for the duration from 10 min after spiking to end of the experiment, respectively. Data from the same cell pair were connected by a line in the left panels. Histograms depict the average values (±SEM) of normalized failure rates (to mean value before the induction) for the data set shown on the left. C, Analysis of PPR of evoked EPSPs showed bidirectional changes at PC-LTS synapses, but no change at PC-FS synapses after correlated spiking. The interval between paired pulses was 100 ms. The value in PPR for individual cell pairs was measured in the same manner as in A. The histogram depicts the ratio of PPR, determined as PPR_aft/PPR_bef.