Antidromic discharges in dorsal roots of decerebrate cats

Abstract

In a previous companion paper [Brain Res. 846 (1999) 87–105] we have shown that the dorsal root activity of a decerebrate cat is composed of both orthodromic and antidromic discharges as determined by spike triggered averaging (STA). Furthermore we have shown that, during fictive locomotion in decerebrate and paralyzed cats, antidromic discharges peak in different parts of locomotion cycle but mainly in the flexion phase. In the present study, we have recorded unit potentials from dorsal rootlets during treadmill locomotion in order to understand better the role of movement-related feedback in the generation of antidromic potentials. The unitary activity of 92 antidromically discharging units was recorded in proximal stumps of cut dorsal roots, and that of 20 such units was recorded in uncut roots using two bipolar Ag/AgCl electrodes in both cases. The activity of 80% (74/92) units in cut filaments and of 70% (14/20) units in uncut ones was phasewise related to stepping movements. The peaks of activity of different units occurred during different phases of the step cycle both in cut and uncut filaments. In most cases, the peak of activity was superimposed upon a background of sustained discharge. After blocking the orthodromic flow in a filament (local anesthesia or distal section), the antidromic discharges continued to peak during the same phase but the rate of the discharges increased. We conclude that movement-related afferent feedback significantly modulates the antidromic discharges in dorsal roots during treadmill locomotion. We suggest that these antidromic discharges have a role in controlling afferent feedback during movement.

Introduction

It is generally assumed that activity recorded in a dorsal root or in a dorsal root ganglion reflects the orthodromic activation of primary afferents in the periphery and that the characteristics of the discharge represent the parameters of stimulation of the receptive field of the afferents. This must indeed be true when afferents are recorded in the distal stump of a cut dorsal root in continuity with the periphery but separated from the spinal cord. In an uncut dorsal root, however, there are instances where spiking activity might ginate from the proximal processes of the afferents in the spinal cord and be conducted antidromically towards the periphery. Such dorsal root discharges have been observed to occur spontaneously in the dorsal roots [27], [28], [40], [41], [42]. Dorsal root discharges have also been described as dorsal root reflexes when occurring on dorsal root potentials evoked by stimulating a peripheral nerve or another dorsal root [1], [2], [6], [7], [16]. More recently such antidromic afferent discharges have also been observed in peripheral nerves in the course of joint inflammation [30], [31].

We have previously shown, during fictive locomotion occurring spontaneously in paralyzed decerebrate cats or in spinal cats after intravenous injection of DOPA, that dorsal root discharges could be rhythmically superimposed on dorsal root potentials during either the flexion or the extension phases [9], [10], [11], [12], [13]. Intracellular recording of various primary afferents of cutaneous [21] or muscle origin [22]demonstrated that some Group I and II afferents discharge rhythmically during fictive locomotion in the absence of any overt movement in paralyzed cats. In the companion paper to the present study [4], we have confirmed that the antidromic discharges can be recorded in intact filaments during fictive locomotion. Using spike triggered averaging (STA) we have also shown that both orthodromic and antidromic spikes coexist in the same filaments [4].

This finding suggested that antidromic discharges, if present during natural behavior, could modulate the orthodromic discharges of the afferents. Indeed, work on the possible effects of the antidromic discharge on orthodromic discharges demonstrated that a single shock to the distal stump of the root inhibited orthodromic discharges sometimes for hundreds of milliseconds [20]. It could also be hypothesized, following the work of others [23] that the activation of ascending collaterals by antidromic spikes could reach the brain stem. Similarly, antidromic activity generated in some collaterals could be conducted to other collaterals and activate other neurons in the same or in different spinal cord segments thus establishing a network of interactions. Locomotion appears to be an excellent preparation to look at such interactions since both antidromic and orthodromic discharges can be recorded (see also [14]).

In the present paper we have recorded, in decerebrate non-paralyzed cats walking on the treadmill, antidromic discharges in cut or in intact dorsal roots (determining direction of propagation by STA) in order to elucidate the role of movement-related feedback in their generation. In addition, we investigated interactions between orthodromic discharges in the same filament by observing the changes in the antidromic firing of single units as a result of a removal of the orthodromic discharges of the filament by section or lidocaine blockade distally to the recording site. Some of the results have been reported briefly [3], [5], [33], [34].