ELECTROPHYSIOLOGICAL CHARACTERIZATION OF DIFFERENT TYPES OF NEURONS RECORDED INVIVO IN THE MOTOR CORTEX OF THE CAT .2. MEMBRANE PARAMETERS, ACTION-POTENTIALS, CURRENT-INDUCED VOLTAGE RESPONSES AND ELECTROTONIC STRUCTURES

1. Electrical properties of four functional classes [inactivating bursting (ib), noninactivating bursting (nib), fast spiking (fsp), and regular spiking (rsp)] of neurons in the motor cortex of conscious cats were studied with the use of intracellular voltage recording and single-electrode voltage-clamp (SEVC) techniques. Evaluations were made of action potentials and afterpotentials, current-voltage (I-V) relationships, and passive cable properties. Values of membrane potential (V(m)), input resistance (R(N)), membrane time constant (T0), and firing threshold (T50) were also measured. The data were used to extend the electrophysiological classifications of neurons described in the companion paper. 2. Average values of V(m) (from -63 to -66 mV), action-potential amplitudes (from 72 to 77 mV), and firing threshold (-54 mV) were not statistically different in different types of neurons. However, the magnitude of intracellularly injected depolarizing current required to induce spike discharge at 50% probability varied significantly (from 0.6 to 1.1 nA) among cell types. The mean R(N) and T0 measured at V(m) varied between 8.3 and 19.8 MOMEGA, and 7.2 and 15.1 ms, respectively, in the cell classes. 3. Action potentials were overshooting. Their mean duration at half amplitude varied from 0.25 to 0.73 ms among different cell types. Three types of action-potential configurations were distinguished. Type I action potentials found in nib and rsp neurons were relatively fast and had a depolarizing afterpotential (DAP) as well as fast and slow after hyperpolarizations (fAHPs, sAHPs). Type II action potentials found in ib and rsp cells had relatively slow rise and decay phases, DAPs, and sAHPs. Their fAHPs were small or absent. Type III action potentials were found exclusively in fsp cells, had very short durations, prominent fAHPs, but no sAHPs. 4. Steady-state I- V relationships were determined by measuring voltage responses to 0.2- to 1.0-nA hyperpolarizing, rectangular current pulses at different membrane potentials. Both R(N) and T0 exhibited nonlinear behavior over wide ranges of membrane potential; however, between -65 and -75 mV, the I-V relationships varied little, and they appeared constant in most cells. The steady-state values of R(N) increased with decreasing, and decreased with increasing the membrane potential in all but fsp cells. The I- V relationships were virtually linear in fsp neurons. 5. Transient I-V relationships were studied by measuring voltage responses to depolarizing and hyperpolarizing, rectangular current pulses of increasing amplitude from a preset membrane potential of -70 mV. The amplitudes of the resulting voltage deviations were measured at different time points and plotted as a function of injected current. In bursting and rsp cells, three rectifying components were found during depolarizing pulses: 1) fast, transient outward; 2) fast, persistent inward; and 3) slow outward rectification. Hyperpolarizing current pulses induced slow, inward, anomalous rectification in all but fsp cells. 6. The cable properties of 164 neurons were analyzed according to the passive cable model of Rall. The electrotonic length of the equivalent dendritic cylinder (L) and dendritic-to-somatic conductance ratio (p) were estimated with the use of voltage recording and voltage-clamp techniques. Values of L and p were estimated by four different procedures. The averaged values of L varied between 0.94 and 1. 17, and those of p between 2.83 and 3.51 among different cell types. The observations suggest that nonisopotential regions exist in most neurons of the motor cortex. If we assume a value of 1 muF/cm2 for the specific membrane capacitance, parameters of R(N), T0, and L were used to estimate the specific membrane resistance (R(m)), total cell capacitance and cell surface (C(cell)). The average values of R(m) (from 7,600 to 18,400 OMEGA . cm2) and C(cell) (from 0.48 to 2.07 muF x 10(5) mum2, calculated surface area) exhibited significant differences among different cell types. 7. Many electrical properties and patterns of firing activity in bursting, fast-spiking and regular-spiking cells in the motor cortex of cats were similar to those recorded in neocortical neurons in vitro. Detailed electrophysiological analyses, however, have disclosed consequent and significant differences in other properties. This study supported the classification (see companion paper) of at least two different types of bursting neurons in the motor cortex. In addition it supported the hypothesis that fast-spiking neurons are a separate class of interneurons with unique electrical properties.