Postsynaptic Inhibition of Crayfish Tonic Flexor Motor Neurones By Escape Commands

The crayfish abdomen contains separate slow and fast neuromuscular systems that mediate posture and escape tailflips (Kennedy & Takeda, 1965 a,b). It was recently demonstrated that impulses in the medial and lateral giant axons, which trigger escape responses, also inhibit both spontaneous and evoked activity in the tonic, or slow, flexor motor neurones (Kuwada & Wine, 1979). The evidence for inhibition was based exclusively on extracellular recordings of the spontaneously active tonic flexor motor neurones. We have now used intracellular recordings from the somata and neuropilar processes of the tonic flexor motor neurones to clarify the nature of inhibition.

Our main findings are that synaptic potentials (from unknown sources) appear to underlie the spontaneous activity of the tonic flexor motor neurones, and that impulses in the giant axons cause large-amplitude, hyperpolarizing IPSPs in the tonic flexor motor neurones and EPSPs or spikes in the peripheral inhibitor. Of the six tonic flexor neurones in each half ganglion, we have recorded from f ₃, f₅, f₆ and possibly f₄ (the motor neurones are numbered 1−6 according to increasing size, with f5 being the peripheral inhibitor). The motor neurones were identified by the size of their extracellularly recorded axon spikes and by the positions of their somata (Wine, Mittenthal & Kennedy, 1974). Results are from six preparations.

General methods were the same as those used earlier (Kuwada & Wine, 1979). For intracellular recordings we used 3 M-KCI electrodes of 15−80 MΩ, inserted into desheathed ganglia of isolated nerve cords. The giant axons were stimulated via focal suction electrodes placed directly over them in sheathed connectives. Examples of tonic flexor motor neurone activity and inhibition of it by giant axon spikes are shown in Fig. 1.

The following points can be made.

1. In the totally isolated abdominal CNS (abdominal ganglia 1−6), the tonic flexor motor neurones receive copious synaptic potentials, which summate to fire the motor neurones (Fig. 1 A). We cannot yet say whether the motor neurones also have endogenous rhythms, but our results show that synaptic input from tonically active presynaptic elements plays an important role in maintaining the tonic flexor discharge.

2. The IPSPs produced by the giant axon impulses are prominent, hyperpolarizing potentials with latencies of approximately 4 ms and durations of up to 80 ms (Fig. 1 B-D). As much as 2 ms of the latency could be due to conduction delay in the giant axons. When multiple giant axon impulses are evoked at 100−200 Hz, the area of the IPSP as a function of a number of giant axon impulses increases with a slope slightly greater than one. These results are consistent with others (Kuwada & Wine, 1979) which suggest that the giant axons recruit a small population of interneurones which inhibit the tonic motor neurones.

3. In the only experiment in which we antidromically stimulated the fast flexor motor neurone axons, IPSPs were produced in an f6 soma (Fig. 1E). Antidromic stimulation of the fast flexor axons can also inhibit the motor giant (Wine, 1977 a), the phasic extensor motor neurones (Wine, 1977b), and the lateral giant interneurone (Wine, 1971). Thus the fast flexor motor neurones act in parallel with the giant axons to evoke widespread inhibition during both giant- and non-giant-mediated escape responses.

   Inhibition could also be evoked by stimulation of the 1st roots (Fig. 1F). This is to be expected since each 1st root contains the axon of the segmental giant which drives the fast flexor motor neurones strongly (Kramer, 1976).

4. Stimulation of sensory roots also evokes complex, variable, and relatively long-latency (to peak) EPSPs (Fig. 1 G). Evidence is provided elsewhere that the sensory pathways to the tonic flexors are shared, at least partially, with the lateral giant (Kuwada & Wine, 1979; R. A. Goldberg, C. L. Tillotson & J. J. Wine, in preparation). Nevertheless, a clear difference between the tonic and phasic systems is the much slower response time for the tonic reflexes. Whereas the peak response in the lateral giant is at 3−6 ms (Krasne, 1969), extracellular recordings of tonic flexor motor neurones show a peak response at about 50 ms (Kuwada & Wine, 1979), which correlates with the late component of the compound EPSP in the lateral giant (Krasne, 1969). Intracellular recordings from the tonic flexor motor neurones have so far shed little light on the mechanisms of slow reflexes. We do not yet know how sensory information reaches the tonic motor neurones, or the types of receptors, other than mechanoreceptors, that may contribute. Mixed inhibition and excitation can be evoked via sensory pathways (Fig. 1 H, I). Since IPSPs can sometimes be seen for 2nd-root stimulation, pathways other then the segmental giant must contribute. (The 2nd root contains the axons of extensor motor neurones and many sensory axons.)

5. Neuropile recordings from unidentified tonic flexor motor neurones show larger spikes (15−20 mV) than in the soma, but they are still non-overshooting (Fig. 2 A). Hyperpolarizing current pulses evoked spikes via rebound (Fig. 2B), but no clear evidence for rebound following IPSPs was seen in any of our recordings, although rebound following inhibition was noted in the extracellular work (Kuwada & Wine, 1979)

6. Finally, intracellular recordings in the soma of the peripheral inhibitor of the tonic flexors show that it, too, receives a tonic synaptic barrage that contributes to its tonic discharge rate (Fig. 2C). Spikes in the giant axons cause a long-lasting EPSP in f₅ (Fig. 2D); multiple giant spikes produce an augmented EPSP and fire f ₅ at variable latency (Figs. 2E, F). The pattern of giant-axon-induced excitation of f₅ is consistent with other evidence that it receives polysynaptic input via corollary discharge interneurones activated by giant axon spikes (Kuwada & Wine, 1979).

These findings clarify a number of points left open by the extracellular work, and correct an earlier misconception that the somata of the tonic flexors are electrically silent (Wine et al. 1974). The tonic flexors are similar to most crayfish neurones encountered so far, in that their somatic and dendritic membrane does not support overshooting impulses (Wine, 1975). Spikes in the soma are therefore quite small, but EPSPs and IPSPs are less attenuated.

Our current knowledge of the inhibition produced by escape commands can be summarized as follows: an impulse in a lateral or medial giant axon fires many interneurones and peripheral inhibitors to produce widespread inhibition. IPSPs have been recorded so far in 11 classes of elements. The flexion and extension portions of the escape system are inhibited at every level from afferents to muscles (references in Wine, 1977b), and the postural system is also inhibited at least at the levels of sensory neurones, motor neurones and muscles (this paper and Kuwada & Wine, 1979). Thus the giant axons inhibit thousands of elements in the central nervous system and the periphery to help co-ordinate the escape response.

We thank Lee Miller for criticism and Cecilia Bahlman and Jan Ruby for preparing the manuscript. The research was supported by National Science Foundation Grant BNS 78-14179 and by a National Institutes of Health Biomedical Research Support Grant. J. Y. Kuwada is a National Science Foundation Predoctoral Fellow; J. J. Wine is an Alfred P. Sloan Research Fellow.