SUMMARY1. The hypothesis is advanced that the joint occurrence of unitary excitatory post-synaptic potentials (e.p.s.p.s) evoked in motoneurones by branches of common stem pre-synaptic fibres causes short-term synchronization of their discharge during the rising phases of the unitary e.p.s.p.s.2. This hypothesis was tested using the pre-and post-stimulus time (PPST) histogram to detect synchronized firing among groups of intercostal motoneurones discharging in response to their natural synaptic drives.3. Motor nerve action potentials were recorded monophasically from nerve filaments of the external intercostal muscles of anaesthetized, paralysed cats maintained on artificial ventilation.4. Computer methods were used to measure peak spike amplitude, spike interval and filament identification for simultaneous recordings from four filaments. The spike amplitude histograms were derived for each filament and groups of spikes were selected for analysis.5. With spikes of one group designated as 'stimuli' (occurring at zero time) and those of a second as 'response' the PPST histogram was computed with different time bin widths.6. With bin widths of 100 and 10 msec the central respiratory periodicity was apparent in the PPST histogram. With 1D0 msec bins the PPST histogram showed a narrow central peak extending to + 3 0 msec at its base. This 'short-term synchronization' supports the hypothesis of joint firing due to common presynaptic connectivity.7. It was shown that detection of short-term synchronization was critically dependent on a sufficient quantity of data but that provided a simple criterion of adequate counts per bin in the PPST histogram was met, short-term synchronization could be detected between intercostal motoneurones of the same and adjacent segments.

4. The peak depolarization lay between -1-0 and + 4-6 msec (mean + 0-7 msec) with respect to the trigger spikes and the rise times of its most prominent component ranged from 4 to 16 msec (mean 8-4 msec).5. The amplitudes of the a.c.e. potentials ranged from 6 to 104 gtV (mean 32 UV) when the trigger spikes were derived from a filament in the same segment as the relevant motoneurones, and from 3 to 42 1V (mean 19 uV) when the filament was two segments rostral to the motoneurone.6. Cells innervating the proximal region of the intercostal space gave larger a.c.e. potentials than those innervating more distal regions and also showed larger central respiratory drive potentials. 7. A.c.e. potentials were observed for either a or y spikes as triggers. The potentials were usually smaller for the y than for the a spikes, the mean ratio being about 0-6.

SUMMARY1. 4-Aminopyridine (4AP) and tetraethylammonium ions (TEA), which block voltage-dependent potassium channels in other nerve membranes, have been used to study nerve conduction in fibres of normal rat spinal roots and those demyelinated with diphtheria toxin. The pharmacological actions have been compared with those of temperature.2. Both TEA and 4AP increased the amplitude and duration of the monophasically recorded compound action potentials of non-myelinated fibres in normal rat dorsal roots. Enhancement of the action potential amplitude by 4AP was maximal near 1 mm, and was not readily reversed by washing. At concentrations up to 50 mm the action of TEA was weaker and reversible.3. In normal dorsal and ventral roots TEA (20 mm) and 4AP (5 mM) had only a mildly depressant action on the compound action potentials of myelinated fibres. Whereas the slight reduction in peak amplitude and increase in width was also found in single fibres treated with TEA, none was discerned in single fibres exposed to 4AP over a wide temperature range.4. It is concluded that voltage-dependent potassium channels occur in significant numbers in mammalian non-myelinated fibres, but not at nodes of Ranvier.5. Spinal roots previously treated with diphtheria toxin to cause demyelination were studied by longitudinal current analysis. Fibres affected by diphtheria toxin had a 1hte phase of outward current, either restricted to nodes or, in the case of continuous conduction, distributed along internodes, and this outward current was specifically blocked by 4AP.6. Both 4AP and TEA increased the temperature at which conduction block occurred in most single demyelinated fibres, so that in some cases fibres blockedatphysiological temperatures were enabled to conduct. 4AP was more potent than TEA, but less consistent in its effect.7. It is concluded that potassium channels are present at widened nodes and in internodal axolemma exposed by demyelination. Their presence enables TEA and 4AP to overcome conduction block in some demyelinated nerve fibres.

SUMMARY1. A new method is described for recording external longitudinal currents from single undissected nerve fibres in rat ventral roots. The method permits identification of the sites of fifteen or more successive nodes of Ranvier in a given single fibre and the measurement of internodal conduction times between them.2. Average internodal conduction time for normal ventral root fibres of internodal length between 0*75 and 1-45 mm is 19-7 + 4.6 (S.D.) ,usec at 370 C. Internodal conduction time appeared to show a minimum for fibres of internodal length 1 0 mm.3. Ventral roots were demyelinated by focal application of diphtheria toxin. Although conduction is markedly slowed in demyelinated fibres, sites of inward membrane current remain spatially separated indicating that conduction remains saltatory to the point of conduction block rather than becoming continuous as in unmyelinated fibres.4. Slowing of conduction appears to be due to changes in the passive electrical properties of the internodal myelin. Evidence is presented suggesting that there is an increase in internodal capacitance and a decrease in internodal transverse resistance at internodes of demyelinated fibres; such changes would have the effect of delaying excitation at the nodes. The changes in passive electrical properties, which appear to be primarily in the vicinity of the nodes, would be consistent with the pathological changes observed in demyelinated fibres. 6. As in normal fibres, nodes of demyelinated fibres generate less current when excited by the second of two closely spaced impulses. This results in an increased internodal conduction time for the second impulse and, at a critically short interstimulus interval, conduction block of the second impulse.7. The increased refractory period of transmission of internodes with increased internodal conduction times is a consequence of the decreased ability of such internodes to sustain propagation in the face of small decreases in nodal current.8. During tetanic stimulation, increases in internodal conduction time are associated with corresponding decreases in nodal current generated by the node proximal to the internode in question.9. It is suggested that changes in the magnitude of the nodal current during repetitive activity are due to changes in transmembrane concentration gradients of sodium, the increased internodal conduction time and eventual conduction block during tetanic stimulation being caused by intracellular sodium accumulation.10. Intracellular sodium accumulation is also offered as the explanation for the post-tetanic depression seen in demyelinated fibres.

No abstract

No abstract

We have previously demonstrated that a drug which prolongs action potentials can, like a reduction in temperature, overcome conduction failure in demyelinated nerve fibres. Although the particular substance then used, a scorpion venom, was not a suitable therapeutic agent, we suggested that other drugs, with similar but milder effects on the action potential, might be effective in the symptomatic treatment of multiple sclerosis. We now report some encouraging results obtained with 4-amino-pyridine (4AP), a substance which blocks the voltage-dependent potassium current in squid giant axons. The use of a potassium-blocking agent to prolong action potentials may seem surprising because we previously found tetraethylammonium chloride (TEA), another potassium blocker, ineffective on normal rat myelinated fibres, and two recent voltage-clamp studies have confirmed that mammalian nodes have few, if any, potassium channels. On the other hand, 4AP strongly potentiates transmitter release from the unmyelinated terminals of rat motor nerves, and the possibility arose that demyelinated axon membrane, which can conduct impulses continuously like an unmyelinated fibre, might further resemble its unmyelinated terminals by responding to 4AP. In testing this hypothesis, we have found that both TEA and 4AP prolong action potentials of demyelinated and unmyelinated fibres, and both facilitate conduction in fibres blocked by demyelination. 4AP is effective at lower concentrations, and is the more promising for clinical use, as it has already been used with beneficial effects in the treatment of Eaton-Lambert syndrome and myasthenia gravis.

SUMMARY1. Simultaneous recordings were made of the discharges of inspiratory bulbospinal neurones and phrenic or external intercostal z-motoneurones in the anaesthetized cat.The connexions between these neurones were studied by the construction of cross-correlation histograms from their discharges.2. Peaks observed in the cross-correlation histograms were divided into three groups on the basis of their time courses: narrow, medium-width and high-frequency oscillations (h.f.o.). Narrow peaks were defined as having half-widths less than 1 1 ms and medium-width peaks as having half-widths greater than this, while h.f.o. was characterized by periodic waves in the range 60-120 Hz.3. H.f.o. peaks were interpreted as being derived from the well known periodic synchronization of medullary inspiratory neurones in this frequency range.4. The time courses and latencies ofthe medium-width peaks could be quantitatively explained by a simple model representing excitation of the motoneurones by bulbospinal neurones whose discharges showed synchronization within + 1 ms of the reference spike, together with temporal dispersion in bulbospinal axons having a distribution of conduction velocities given by the measurements of this study. Such an explanation was essential for some of the medium-width peaks, whose latencies were short compared to the conduction times to the spinal cord for their own axons, but for other medium-width peaks oligosynaptic excitation of the motoneurones from the identified bulbospinal neurones was another possible explanation.5. The narrow peaks were of appropriate durations for monosynaptic connexions and were all at appropriate latencies (0-62-4 ms after the calculated arrival time of the bulbospinal impulse in the segment concerned).6. It is concluded from the observations of narrow peaks that monosynaptic excitation exists between inspiratory bulbospinal neurones and both phrenic and external intercostal motoneurones. However, because of the existence of presynaptic synchronization, as shown by the presence of the medium-width peaks, such a conclusion is predicated upon being able to discriminate against such an effect. The model showed that this restriction applies just as much to the measurements of excitatory post-synaptic potentials (e.p.s.p.s) by spike-triggered averaging as it does 2 P HY 368 J. G. McF. DAVIES, P. A. KIRKWOOD AND T. A. SEARS to cross-correlation measurements. We suggest that the discrimination against presynaptic synchronization here was possible only because the long conduction distance created temporal dispersion in the synchronized presynaptic impulses.