1. The goal of this study was to compare the relative effectiveness of intrinsic inhibitory synaptic inputs in different layers of the cat motor cortex. Postsynaptic potentials (PSPs) were evoked in neurons located in the superficial (layer II-III) or deep layers (layer V) by local extracellular stimulation in vitro. Electrophysiological properties and intracellular filling indicated that the recorded neurons were pyramidal cells. 2. The shape and time course of the evoked PSPs differed. Layer II-III cells showed stereotyped triphasic PSPs consisting of a fast excitatory PSP (fEPSP) and a fast and slow inhibitory PSP (fIPSP and sIPSP, respectively). PSPs in layer V cells, in contrast, were much more variable, mainly depolarizing at resting membrane potential, and lacked a hyperpolarizing IPSP in 84% of neurons tested at rest. 3. Blockade of glutaminergic neurotransmission with 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and D-2-amino-5-phosphonovaleric acid (AP5) revealed that fIPSPs could be evoked in all layer V cells by local stimulation of the superficial or deep layers, even in those that showed small or no IPSPs in control perfusate. Small (<1 mV) isolated sIPSPs were evoked in only one-fifth of layer V cells when the deep layers were stimulated, and in about one-half of the layer V cells when the superficial layers were stimulated. In layer II-III cells, stimulation of the superficial layers always resulted in fIPSP-sIPSP combinations. No IPSPs could be evoked in layer II-III neurons by stimulating the deep layers after glutaminergic blockade. Selective blockade of gamma-aminobutyric acid-A (GABA(A)) or GABA(B) receptor-mediated neurotransmission showed that in both cell types fIPSPs were due to GABA(A) receptor stimulation, whereas sIPSPs were mediated by GABA(B) receptors. 4. Isolated fIPSPs were recorded in perfusate containing CNQX, AP5, and the GABA(B) antagonist CGP 35348. The rise and decay times of the fIPSPs in layer II-III cells were significantly longer than those in layer V cells. Rise and decay times normalized for differences in membrane time constant were not significantly different, however, suggesting that the intrinsic membrane properties of the postsynaptic membrane account for the difference in time course of the fIPSPs in these two cell types. 5. Selective blockade of the inward rectifier current I-h with extracellular Cs+ showed that this conductance functions to shorten and attenuate fIPSPs in layer V cells. In contrast, I-h is absent or small in layer II-III cells, and, consequently, Cs+ had little or no effect on the fIPSPs evoked in these cells. In layer V cells that had sIPSPs, extracellular Cs+ enhanced sIPSP amplitude, but these sIPSPs were still small compared with those in layer II-III cells. These results indicate that I-h contributes to the relatively weak and short-lived IPSPs recorded in layer V cells. In contrast, the presence of sIPSPs depends more strongly on differences in inhibitory circuitry or properties of GABAB synapses between deep and superficial layers of the cortex. 6. When trains of IPSPs were evoked by repetitive extracellular stimulation after blockade of glutaminergic neurotransmission, our experiments showed that the amount of temporal summation and ''accumulation'' of hyperpolarizing inhibition depended both on the presence of GABA(B)-mediated sIPSPs and the duration of the fIPSP. The small or absent sIPSP and the short fIPSP in layer V cells makes temporal summation of inhibitory synaptic input less effective. 7. When GABA release was enhanced by the convulsant drug 4-aminopyridine (with excitation blocked), spontaneous and evoked biphasic IPSPs resembling fast and slow IPSPs were observed in all layer V cells tested. This indicates that GABA(B)-mediated IPSPs require a larger release of GABA to be visible in layer V compared with layer II-III neurons. 8. Evoked single fIPSPs were able to veto action potentials generated in response to intracellular stimulation, whereas single sIPSPs were more suited to modulate firing rate but were able to do so over longer time periods. 9. In conclusion, our studies show that pyramidal neurons in the superficial layers of the cat motor cortex are controlled by a powerful local inhibitory circuit that operates through both GABA(A)- and GABA(B)-mediated IPSPs. This dual-component inhibitory input tends to hyperpolarize these cells away from firing threshold during bursts of synaptic input. In comparison, intrinsic membrane properties and weak or absent sIPSPs in layer V pyramidal cells appear to be designed to rapidly bring these neurons to firing threshold during repeated synaptic input. These results suggest that there are laminar-specific inhibitory circuits operating in the cat motor cortex.
