We conducted a 10-center, double-blind trial to compare the efficacy and toxicity of four antiepileptic drugs in the treatment of partial and secondarily generalized tonic-clonic seizures in 622 adults. Patients were randomly assigned to treatment with carbamazepine, phenobarbital, phenytoin, or primidone and were followed for two years or until the drug failed to control seizures or caused unacceptable side effects. Overall treatment success was highest with carbamazepine or phenytoin, intermediate with phenobarbital, and lowest with primidone (P less than 0.002). Differences in failure rates of the drugs were explained primarily by the fact that primidone caused more intolerable acute toxic effects, such as nausea, vomiting, dizziness, and sedation. Decreased libido and impotence were more common in patients given primidone. Phenytoin caused more dysmorphic effects and hypersensitivity. Control of tonic-clonic seizures did not differ significantly with the various drugs. Carbamazepine provided complete control of partial seizures more often than primidone or phenobarbital (P less than 0.03). Overall, carbamazepine and phenytoin are recommended drugs of first choice for single-drug therapy of adults with partial or generalized tonic-clonic seizures or with both.

Neurons from the mammalian CNS have a noninactivating component of the tetrodotoxin-sensitive sodium current (INaP). Although its magnitude is < 1% of the transient sodium current, INaP has functional significance because it is activated about 10 mV negative to the transient sodium current, where few voltage-gated channels are activated and neuron input resistance is high. INaP adds to synaptic current, and evidence indicates that it is present in dendrites where relatively small depolarizations will activate INaP, thereby increasing effectiveness of distal depolarizing synaptic activity. The mechanism for INaP is not known. Research in striated muscle and neurons suggests a modal change in gating of conventional sodium channels, but it is also possible that INaP flows through a distinct subtype of noninactivating sodium channels. Modulation of INaP could have a significant effect on the transduction of synaptic currents by neurons.

Properties of the persistent sodium conductance and the calcium conductance of layer V neurons from cat sensorimotor cortex were examined in an in vitro slice preparation by use of a single microelectrode, somatic voltage clamp, current clamp, intra- and extracellular application of blocking agents, and extracellular ion substitution. The persistent sodium current (INaP) attained its steady level within 2-4 ms of a step change in voltage at every potential where it could be examined directly [to about 40 mV positive to resting potential (RP)]. Because of its fast onset INaP can be activated during a single excitatory postsynaptic potential (EPSP) and can influence the subsequent voltage time course and cell excitability. Application of a depolarizing holding potential greater than or equal to 20 mV positive to RP could inactivate spikes, thus allowing examination of INaP at voltages positive to spike threshold. At every potential where INaP was visible, it was mixed with a slow outward current. After depressing potassium currents with blocking agents, INaP could be observed during depolarizations to about 40 mV positive to RP where it is normally hidden by the larger outward currents. Indirect evidence suggests that INaP is present and large during prolonged depolarizations greater than 50 mV positive to RP. INaP was blocked by intracellular injection of the lidocaine derivative QX-314, as well as by extracellular tetrodotoxin (TTX). INaP was much more sensitive to QX-314 than was the height and rate of rise of the spike. This observation and the results in paragraph 3 above are best explained by separate INaP and spike sodium channels. After blockade of INaP and sodium spikes, Ca2+ spikes could be evoked only if potassium currents were first depressed. The Ca2+-dependent nature of the regenerative potentials was indicated by their disappearance when Co2+ or Mn2+ was substituted for Ca2+ in the perfusate and by the appearance of greatly enhanced potentials of similar form when Ba2+ was substituted for Ca2+. Ba2+ substitution greatly enhanced evoked and spontaneous synaptic potentials. Prolonged-plateau action potentials could be evoked in the presence of TTX and Ba2+. Ca2+ spike threshold was 30-40 mV positive to RP, which is significantly more positive than sodium spike threshold. Results of voltage clamp in the normal perfusate and in the presence of Ca2+-blockers or Ba2+ indicated that little or no Ca2+ conductance is activated in the voltage range 25 mV positive to RP where INaP is the dominant ionic current.(ABSTRACT TRUNCATED AT 400 WORDS)

1. The electrophysiological and pharmacological properties of slow afterpotentials in large layer V neurons from cat sensorimotor cortex were studied in an in vitro slice preparation using intracellular recording and single-microelectrode voltage clamp. These properties were used to assess the role of afterpotential mechanisms in prolonged excitability changes. 2. The mean duration of a slow afterhyperpolarization (sAHP) was 13.5 s following 100 spikes evoked at 100 Hz. Its time course was best described by two exponential components, which decayed with time constants of several hundred milliseconds (the early sAHP) and several seconds (the late sAHP). The amplitude of both the early and late components were sensitive to membrane potential and raised extracellular K+ concentration [( K+]o). 3. The early sAHP was reduced when divalent cations were substituted for Ca2+, whereas the late sAHP was unaffected. We conclude that a Ca2+-mediated K+ conductance is responsible for much of the early sAHP. In the presence of tetrodotoxin (TTX), 1-s voltage-clamp steps were used to evoke slow AHPs or outward ionic currents. These AHPs and currents were abolished in Ca2+-free perfusate, but they had a maximum duration of only a few seconds. Thus the slowest outward currents we could observe during voltage clamp in TTX were responsible only for the early sAHP. 4. The possible role of an electrogenic Na+-K+ pump in the late sAHP was examined by applying ouabain to the slice. Ouabain did not reduce selectively the late sAHP, and its effect was best explained by a decrease in intracellular K+ concentration and an increase in [K+]o. 5. Muscarinic and beta-adrenergic agonists reduced or abolished the entire (early and late) sAHP. Neither type of agonist affected the Ca2+-dependent, apamin-sensitive medium-duration afterhyperpolarization (35). We conclude that both the Ca2+-mediated K+ conductance underlying the early sAHP and the Ca2+-independent mechanisms underlying the late sAHP are sensitive to at least two classes of transmitter agonists. 6. We focused on the muscarinic effects. When concentrations greater than 5 microM were employed, the entire (early and late) sAHP was replaced by a slow afterdepolarization (sADP). Muscarine reduced the sAHP directly by reducing the underlying outward ionic currents and indirectly by causing the sADP. The sADP was Ca2+-mediated, since it was abolished by Ca2+-free perfusate but not by TTX. 7. The ionic currents underlying the sAHP and the sADP influenced excitability for seconds following evoked repetitive firing.(ABSTRACT TRUNCATED AT 400 WORDS)

The kinetic behavior of brain Na+ channels was studied in pyramidal cells from rat and cat sensorimotor cortex using either the thin slice preparation or acutely isolated neurons. Single-channel recordings were obtained in the cell-attached and inside-out configuration of the patch-clamp technique. Na+ channels had a conductance of about 16 pS. Patches always contained several Na+ channels, usually 4-12. In both preparations, long depolarizing pulses revealed two distinct patterns of late Na+ channel activity following transient openings. (1) Na+ channels displayed sporadic brief late openings sometimes clustered to "minibursts" of 10-40 msec. These events occurred at a low frequency, yielding open probability (NPo) values below 0.01 (mean = 0.0034). (2) In the second gating mode, an individual Na+ channel in the patch failed to inactivate and produced a burst of openings often lasting to the end of the pulse. This behavior was observed in about 1% of depolarizations. Shifts to the bursting mode were usually confined to a single 400 msec pulse, but rarely occurred during two or more consecutive pulses applied at 2 sec intervals. Sustained bursts did not require preceding transient openings to occur since they were also observed during slow depolarizing voltage ramps. The similar incidence of inactivation failures in cell-attached versus inside-out recordings suggests that the bursting mode is a property of the channel and/or adjacent membrane-bound structures. Calculations indicate that brief late openings and rare sustained bursts suffice to generate a small but significant whole-cell current. Since the Na+ channels mediating early, brief late, and sustained openings were identical in terms of their elementary electrical properties, we propose that the fast and the persistent Na+ currents of cortical pyramidal cells are generated by an electrophysiologically uniform population of Na+ channels that can individually switch between different gating modes.

SUMMARY1. Electrophysiological properties of cat motoneurones were measured using intracellular electrodes, after which Procion dye was injected iontophoretically into the neurone through the recording pipette.2. Histological procedures were chosen to minimize changes in neuronal morphology. Reconstructed motoneurones had more dendritic branches and larger surface areas than the Golgi-stained motoneurones of earlier reports.3. The sum of the 3/2 power of the dendritic diameters (the dendritic trunk parameter; Rall, 1959) of the reconstructed motoneurones was found to decrease with distance from the soma. Thus, the dendritic tree is not satisfactorily approximated by a non-tapering membrane cylinder.4. A computational technique was developed to allow calculation of the specific resistance (Rm) of the membrane using the measured value of the input resistance of the motoneurone and a more detailed approximation of the dendritic tree. These calculations indicate that the average resting value of dendritic Rm is at least 1800 Q cm2. The specific membrane capacity, calculated assuming uniform Rm, ranged between 2-3 ,tF/cm2.

1. Potassium conductances were studied in large layer V neurons using an in vitro slice preparation of cat sensorimotor cortex. The kinetics and pharmacological sensitivity of K+ currents were studied directly using single microelectrode voltage clamp and indirectly by evoking single or multiple spikes and recording the spike repolarization and subsequent afterhyperpolarizations (AHPs). 2. A fast-decaying afterhyperpolarization (fAHP) and a subsequent medium-duration afterhyperpolarization (mAHP) followed a single spike. The amplitude and duration of the mAHP increased when multiple spikes were evoked at a fast rate (e.g., 100 Hz), and a slower afterhyperpolarization (sAHP) appeared only after sustained repetitive firing. 3. All AHPs were reduced by membrane potential hyperpolarization and raised extracellular K+ concentration, suggesting they were caused by an increased K+ conductance. Only the mAHP and sAHP reversed at the estimated value of potassium equilibrium potential (-100 mV), whereas the mean reversal potential of the fAHP was nearly identical to the mean value of resting potential (-71 mV). 4. Mechanisms underlying spike repolarization, the fAHP, and the mAHP were investigated. Two rapidly activating outward currents, a fast-inactivating current and a slowly inactivating delayed rectifier, were detected by voltage clamp. Both currents were reduced rapidly by tetraethylammonium (TEA). The fast transient current was reduced slowly after divalent cations were substituted for Ca2+ (through a mechanism unrelated to blockade of Ca2+ channels), whereas the delayed rectifier was unaffected. 5. Spike duration was increased and the fAHP was abolished only by blocking agents that reduced the fast outward currents. Effects of extracellular and intracellular TEA were similar. Effects of TEA and Ca2+-free perfusate were additive and resembled the effects of intracellular Cs+. The addition of apamin, d-tubocurare, or Cd2+ was ineffective. We conclude that the two fast outward currents reflect pharmacologically and kinetically separate K+ conductances that are primarily responsible for spike repolarization and the fAHP. 6. Voltage-clamp studies revealed two additional outward currents, which were persistent and Ca2+-mediated. Each current activated and deactivated slowly, but the kinetics of one component were approximately 10 times slower than the other. The decay of these currents gave rise to AHPs resembling the mAHP and the early sAHP. 7. Neither the mAHP nor the sAHP was reduced by TEA. The mAHP was reduced when divalent cations were substituted for Ca2+ or when Cd2+, apamin, or d-tubocurare were added.(ABSTRACT TRUNCATED AT 400 WORDS)

The ionic mechanisms underlying anomalous rectification in large neurons from layer V of cat sensorimotor cortex were studied in an in vitro brain slice. The anomalous rectification was apparent as an increase of slope conductance during membrane hyperpolarization, and the development of anomalous rectification during a hyperpolarizing current pulse was signaled by a depolarizing sag of membrane potential toward resting potential (RP). Voltage-clamp analysis revealed the time- and voltage-dependent inward current (IAR) that produced anomalous rectification. IAR reversal potential (EAR) was estimated to be approximately -50 mV from extrapolation of linear, instantaneous, current-voltage relations. The conductance underlying IAR (GAR) had a sigmoidal steady-state activation characteristic. GAR increased with hyperpolarization from -55 to -105 mV with half-activation at approximately -82 mV. The time course of both GAR and IAR during a voltage step was described by two exponentials. The faster exponential had a time constant (tau F) of approximately 40 ms; the slow time constant (tau S) was approximately 300 ms. Neither tau F nor tau S changed with voltage in the range -60 mV to -110 mV. The fast component constituted approximately 80% of IAR at each potential. Both IAR and GAR increased in raised extracellular potassium [( K+]o) and EAR shifted positive, but the GAR activation curve did not shift along the voltage axis. Solutions containing an impermeable Na+ substitute caused an initial transient decrease in IAR followed by a slower increase of IAR. Brain slices bathed in Na+-substituted solution developed a gradual increase in [K+]o as measured with K+-sensitive microelectrodes. We conclude that GAR is permeable to both Na+ and K+, but the full contribution of Na+ was masked by the slow increase of [K+]o that occurred in Na+ substituted solutions. Chloride did not appear to contribute significantly to IAR since estimates of EAR were similar in neurons impaled with microelectrodes filled with potassium chloride or methylsulfate, whereas, ECl (estimated from reversal of a GABA-induced ionic current) was approximately 30 mV more positive with the KCl-filled microelectrodes. Extracellular Cs+ caused a reversible dose- and voltage-dependent reduction of GAR, whereas intracellular Cs+ was ineffective. The parameters measured during voltage clamp were used to formulate a quantitative empirical model of IAR.(ABSTRACT TRUNCATED AT 400 WORDS)