LARGE, DEEP LAYER PYRAMID-PYRAMID SINGLE AXON EPSPS IN SLICES RAT MOTOR CORTEX DISPLAY PAIRED-PULSE AND FREQUENCY-DEPENDENT DEPRESSION, MEDIATED PRESYNAPTICALLY AND SELF-FACILITATION, MEDIATED POSTSYNAPTICALLY

1. In slices of rat sensorimotor cortex, dual intracellular recordings were obtained from 1,952 pairs of deep layer pyramidal neurons. Where action potentials in one neurone elicited excitatory postsynaptic potentials (EPSPs, n = 56) in the other, responses to different presynaptic firing rates and patterns and at different post-synaptic membrane potentials were recorded and on some occasions both neurons were filled with biocytin. 2. Slices were fixed, sectioned again at 60 mum, and incubated with Avidin horseradish peroxidase (HRP), which was then visualized using the 3,3'-diaminobenzidine tetrahydrochloride (DAB method. All neurones reported here that were identified histologically were pyramidal cells with their somata in the deep layers (V and VI). 3. One in 70 of the tests performed revealed a synaptic connection, 25 of which were studied in detail. Mean EPSP amplitude was 1.67 +/- 1.66 (SD) mV, with some single sweep events as large as 9 mV. For some of the smaller EPSPs the amplitude distributions contained a clear peak around 0 mV, the coefficient of variation (CV) was large, and paired pulse facilitation was apparent. EPSPs with large average amplitudes displayed no apparent failures of transmission, EPSP amplitudes were fairly evenly distributed around the mean, CVs were small, and paired pulse depression was apparent in 2.5 mM extracellular Ca2+. When single sweeps were selected according to the size of the first EPSP, large second EPSPs were found to follow small first EPSPs and small second EPSPs to follow large first EPSPs. Paired pulse effects appeared, in the majority of tests, to be due to a change in presynaptic release probability. 4. Two EPSPs were recorded in three different extracellular Ca2+ concentrations. In 1 mM Ca2+, the first EPSP of a short interval pair was small and paired pulse facilitation was apparent. In 5 mM Ca2+, first EPSPs were between 2.5 and 4 times larger than in 1 MM Ca2+ and paired pulse depression was apparent. In all Ca2+ concentrations however, averaged third and fourth EPSPs of brief bursts were of similar amplitudes and smaller than second EPSPs. If presynaptic inhibition does contribute to paired pulse effects here, it is not overcome by a combination of raised extracellular Ca2+ and repetitive presynaptic firing. 5. These EPSPs displayed a wide range of time courses. The mean 10-90% rise time was 2.49 +/- 1.08 ms, the mean width at half amplitude was 15.39 +/- 5.42 ms (n = 22), and the mean EPSP latency was 1.59 +/- 0.68 ms (n = 18). That EPSPs with a slower time course involved inputs distal to the recording site was also indicated by their relative insensitivity to injected current. No correlation between shape indexes and amplitude was found. 6. At a presynaptic firing rate of 1 Hz and postsynaptic membrane potentials more negative than -70 mV, EPSPs declined in amplitude. In the majority of tests, this decline could be described as due to a decrease in presynaptic release probability. When brief interval pairs of EPSPs were studied, the second EPSP declined in parallel with the first. The proportional rate of decline was not apparently affected by the extracellular Ca2+ concentration. The possibility that extracellular [Ca2+] may affect the size of the available pool of transmitter and thereby the number of available release sites is discussed. 7. When the postsynaptic neurone was depolarized, EPSPs increased in amplitude and duration, an increase that appeared to involve an increase in postsynaptic responsiveness. At postsynaptic membrane potentials of -70 mV, or less negative, 1-Hz activation resulted in self-facilitation. EPSPs increased further in both amplitude and duration. The standard deviation time course (SDTC) also increased and no longer paralleled the shape of the average EPSP. This change also appeared to be postsynaptic in origin. When larger EPSPs became facilitated, they readily elicited action potentials. These properties have been shown to depend on N-methyl-D-aspartate (NMDA) receptor activation in other cortical regions. On return to lower firing rates, or more negative membrane potentials, self-facilitation declined within a few minutes. 8. Large secure disynaptic EPSPs could be recorded between pairs of deep layer pyramidal neurones. That the large size of some of the deep layer connections and their ability to self-facilitate could lead to recruitment of an ever increasing population of pyramidal neurones, were it not for the relatively restricted target regions of their axons and presynaptically mediated, paired pulse and frequency-dependent depression, is discussed.