Static input-output relations in the spinal recurrent inhibitory pathway

Modulation of motoneuron firing by recurrent inhibition in the adult rat in vivo

Recent reports show that synaptic inhibition can modulate postsynaptic spike timing without having strong effects on firing rate. Thus synaptic inhibition can achieve multiplicity in neural circuit operation through variable modulation of postsynaptic firing rate vs. timing. We tested this possibility for recurrent inhibition (RI) of spinal motoneurons. In in vivo electrophysiological studies of adult Wistar rats anesthetized by isoflurane, we examined repetitive firing of individual lumbosacral motoneurons recorded in current clamp and modulated by synchronous antidromic electrical stimulation of multiple motor axons and their centrally projecting collateral branches. Antidromic stimulation produced recurrent inhibitory postsynaptic potentials (RIPSPs) having properties similar to those detailed in the cat. Although synchronous RI produced marked short-term modulation of motoneuron spike timing and instantaneous firing rate, there was little or no suppression of average firing rate. The bias in firing modulation of timing over average rate was observed even for high-frequency RI stimulation (100 Hz), perhaps because of the brevity of RIPSPs, which were more than twofold shorter during motoneuron firing compared with rest. These findings demonstrate that RI in the mammalian spinal cord has the capacity to support and not impede heightened motor pool activity, possibly during rapid, forceful movements.

Recurrent inhibition of wrist extensor motoneurones: a single unit study on a deafferented patient

In order to document the effects of recurrent inhibition on the firing times of human alpha-motoneurones during natural motor behaviour, a case study was performed on a deafferented patient. The fact that this subject had completely lost the large-diameter sensory afferents provided us with a unique opportunity of selectively stimulating the motor axons in the nerves. The tonic activity of single motor units (n = 21) was recorded in the extensor carpi radialis muscles while applying randomly timed antidromic electrical stimuli to the radial nerve. The peristimulus time histogram analysis showed the presence of biphasic inhibitory effects, including an early, short-lasting component followed by a longer-lasting component occurring 20-40 ms later. The interspike interval (ISI) during which the stimulation occurred was generally lengthened as compared to the previous ISIs. The stimulation was most effective when delivered early (20-30 ms) after a spike. It was also effective, although less so, when delivered at the end of the ISI (70-100 ms after a spike). The lengthening effect sometimes extended over one or two of the subsequent ISIs. The lengthening effect of the motor axon stimulation was followed by an excitatory-like effect, which took the form of a shortening that affected up to five ISIs after the stimulation. The biphasic inhibitory effects and the subsequent facilitatory effects are discussed in terms of the dual nature of the synaptic processes involved in the recurrent inhibitory network, the postactivation facilitation/depression processes and the mutual inhibition occurring between Renshaw cells.

Decorrelating actions of Renshaw interneurons on the firing of spinal motoneurons within a motor nucleus: a simulation study

A simulation of spinal motoneurons and Renshaw cells was constructed to examine possible functions of recurrent inhibition. Recurrent inhibitory feedback via Renshaw cells is known to be weak. In our model, consistent with this, motoneuron firing was only reduced by a few pulses per second. Our initial hypothesis was that Renshaw cells would suppress synchronous firings of motoneurons caused by shared, dynamic inputs. Each motoneuron received an identical pattern of noise in its input. Synchrony coefficients were defined as the average motoneuron population firing relative to the activity of selected reference motoneurons; positive coefficients resulted if the motoneuron population was particularly active at the same time the reference motoneuron was active. With or without recurrent inhibition, the motoneuron pools tended to show little if any synchronization. Recurrent inhibition was expected to reduce the synchrony even further. Instead, it reduced the variance of the synchrony coefficients, without a comparable effect on the average. This suggests-surprisingly-that both positive and negative correlations between motoneurons are suppressed by recurrent inhibition. In short, recurrent inhibition may operate as a negative feedback mechanism to decorrelate motoneurons linked by common inputs. A consequence of this decorrelation is the suppression of spectral activity that apparently arises from correlated motoneuron firings due to common excitatory drive. Without recurrent inhibition, the power spectrum of the total motoneuron pool firings showed a peak at a frequency corresponding to the largest measured firing rates of motoneurons in the pool. Recurrent inhibition either reduced or abolished this peak, presumably by minimizing the likelihood of correlated firing among pool elements. Renshaw cells may act to diminish physiological tremor, by removing oscillatory components from aggregate motoneuron activity. Recurrent inhibition also improved coherence between the aggregate motoneuron output and the common drive, at frequencies above the frequency of the "synchronous" peak. Sensitivity analyses demonstrated that the spectral effect became stronger as the duration of inhibitory synaptic conductance was shortened with either the magnitude or the spatial extent of the inhibitory conductances increased to maintain constant net inhibition. Overall, Renshaw inhibition appears to be a powerful way to adjust the dynamic behavior of a neuron population with minimal impact on its static gain.

On the role of recurrent inhibitory feedback in motor control

This article reviews presumed roles of recurrent inhibition in motor control, that have been proposed over the past five decades. The discussion is structured in an order of increasing complexity. It starts out with the simplest and earliest circuit, that is recurrent self-inhibition of skeleto-motoneurons, and related functions. It soon becomes clear that in order to understand recurrent inhibition, we must look beyond the simple self-inhibitory CNS circuit. First, recurrent inhibition must be seen in the context of other neural circuits. Second, some quantitative features appear to be correlated with features of the neuromusculo-skeletal periphery. Third, the aspect of lateral inhibition between different members of a motoneuron pool as well as between different motoneuron pools points to the essential multiple input-multiple output structure of recurrent inhibition that again can be understood only by correlating it with features of the neuromusculo-skeletal periphery. Another extension results from the discovery that recurrent inhibition affects not only skeleto-motoneurons, but also gamma-motoneurons, Ia inhibitory interneurons mediating reciprocal inhibition between antagonist motoneurons, other Renshaw cells and cells of origin of the ventral spinocerebellar tract (VSCT). Then the view broadens again, investigating the potential role that recurrent inhibition plays in two far-ranging theories of motor control, the inverse-dynamics approach and the equilibrium-point hypothesis. Finally, the present author tries to formulate, in broad strokes, a personal functional interpretation of recurrent inhibition. All the functional considerations, right or wrong, should yield ideas for new experiments, and this then is the last objective of this review.

Shaping static elbow torque-angle relationships by spinal cord circuits: a theoretical study

Static torque-angle relationships (invariant characteristics) as measured by Feldman [Feldman A. G. (1980) Neuroscience 5, 81-90] at the human elbow joint for constant descending excitatory drive have a monotonic convex shape determining joint angle-dependent stiffness. In contrast, for constant activation of elbow flexors, the torque increases, peaks and decreases again with increasing angle because of related moment arm alterations [Hasan Z. and Enoka R. M. (1985) Expl Brain Res. 59, 441-450]. Conversion of such constant-excitation torque-angle shapes into an invariant characteristic might result from action of the stretch reflex which adds excitation with increasing joint angle. To test whether a simple linear model of the stretch reflex could convert constant excitation torque-angle relationships into invariant characteristics, the following assumptions were made. (1) Muscle fibre length increases linearly with joint angle. (2) Reflex muscle excitation (electromyogram) is linearly related to muscle (fibre) length. With these assumptions, invariant characteristic shape cannot be derived from constant excitation torque-angle relationships because it would be sigmoid at low and nearly straight at large joint angles, whilst real flexor invariant characteristics are more convex at large than small angles. It is suggested that recurrent inhibition via Renshaw cells contributes to bend the invariant characteristics into their right shape. Renshaw cells show a nonlinear saturating dependence on motor axon input rate and amount of excitation, i.e. number of active axon collateral synapses. These relationships can contribute to shape motoneuron output so as to yield convex invariant characteristics. Whilst it is not quite clear whether the gain of recurrent inhibition from and to skeleto-motoneurons is high enough to co-determine the invariant characteristic shape significantly, recurrent inhibition of Ia inhibitory interneurons mediating reciprocal inhibition between antagonists is supposed to be quite strong and may influence joint stiffness by interacting with reciprocal inhibition. The arguments presented here extend those of Feldman and co-workers concerning the role of recurrent inhibition and in addition provide a possible explanation for the functional role of mutual inhibition between Renshaw cells. Together with reflex feedback, recurrent inhibition thus contributes to fine-regulate force output and joint stiffness. To account for this cooperation and to make another step towards a general theory of spinal cord circuits, major traits of a new concept are briefly outlined.

Interaction of recurrent inhibitory and muscle spindle afferent feedback during muscle fatigue

Mammalian skeletal motor units have differing properties including their different susceptibility to fatigue. The question discussed in this paper is whether and to what extent proprioceptive feedback via muscle spindles can contribute to shape the firing patterns of motor units so as to minimize their loss of force during fatiguing contraction. The firing of a skeleto-motoneuron dispatches signals which are fed back to the same and homonymous as well as synergistic motoneuron. Two feedback pathways are of concern here: one via the related muscle unit and muscle spindle afferents (proprioceptive path), and one via recurrent motor axon collaterals and Renshaw cells (recurrent inhibitory path). It is suggested that the contraction of a motor unit or a small group of adjacent ones is signalled to the homonymous alpha-motoneurons via proprioceptive afferents, the signal being filtered and enhanced by spinal recurrent inhibition. This is effected by timed correlation of the signals which are propagated through the two feedback loops. The effects of the correlation can be strengthened by (i) topographical order of the feedback connections, (ii) heterosynaptic interaction, and (iii) tendencies towards synchronous discharge between motoneurons. These mechanisms render the possibility more likely that information about the unfused contractions of a muscle unit (or a small group of them), mediated via proprioceptive afferents, play a role in shaping the precise discharge pattern of the innervating motoneuron(s). These mechanisms could be used to optimize the force output of large fatiguing motor units during long activation, during which their activation rates normally decrease (adapt) over time. Our results show that during adapting motoneuron firing Renshaw cells and muscle spindle afferents may show discharge patterns which at least in part are in keeping with such an hypothesis.

Higher-order non-linear phenomena in Renshaw cell responses to random motor axon stimulation

Renshaw cell responses to random motor axon stimulation exhibit second-order non-linearities in that they depend on the occurrence of a preceding stimulus, although these non-linearities are not strong enough to significantly depress the coherence. However, higher-order non-linearities have not been checked for so far. This is carried out here. Lumbosacral Renshaw cells were recorded with micropipettes in anaesthetized cats. Their responses to random (pseudo-Poisson) stimulation of motor axons in peripheral nerves or ventral roots were quantified by calculating peristimulus-time histograms of various sorts, conventional and conditional. Conventional peristimulus-time histograms were computed with respect to all the stimuli in a train. Conditional peristimulus-time histograms were calculated with respect to "test" stimuli which were sorted out (by computer) from the original stimulus train when they were preceded by "conditional" stimuli at average intervals of delta 1 or delta 2 or both. These conditioned responses were compared with those to be expected from hypothetical linear superposition. Renshaw cell responses showed small third-order non-linearities to pairs of conditioning stimuli at small intervals (up to some tens of milliseconds before the test stimuli). These third-order effects were smaller than each of the second-order non-linearities elicited by any of the single-conditioning stimuli. Also, further higher-order non-linearities were apparent, but of little average significance. Hence, the non-linearities in Renshaw cell responses to random inputs are essentially of second-order and fairly small.

Inhibitory connections of ipsilateral semicircular canal afferents onto Renshaw cells in the lumbar spinal cord of the cat

1. In intercollicularly decerebrate cats, the excitability of lumbar spinal Renshaw cells (tested by single shocks to ventral roots and deafferented muscle nerves) decreased for 600-1000 ms after conditioning electrical stimulation of ipsilateral semicircular canal nerves. 2. Conditioning stimulation of posterior canal afferents and combined stimulation of anterior and lateral canal afferents were equally effective in causing inhibition of Renshaw cells. No significant differences were observed for Renshaw cells excitable from hind-limb flexor or extensor nerves. 3. Inhibition appeared when one to five stimuli were applied to the canal afferents and arrived at the spinal segmental level 11-15 ms after the onset of conditioning stimulation. 4. Evidence is adduced to suggest that the inhibitory effects on Renshaw cells following stimulation of semicircular canal afferents were mediated directly, i.e. they were not caused by alterations of motoneurone activity. 5. Excitation of Renshaw cells due to stimulation of the canal afferents was rarely observed; it could not be excluded that it was secondary to motoneurone discharges. 6. It is suggested that vestibular inhibition of Renshaw cells ensure a high gain of hind-limb alpha-motoneurones during postural adjustments following a massive disturbance of body equilibrium.

Dynamic properties of Renshaw cells: equivalence of responses to step changes in recruitment and discharge frequency of motor axons

In decerebrate cats, the dynamic responses of Renshaw cells to step changes in input were determined separately both for changes in the number of alpha-axons excited and for changes in the frequency at which they were stimulated. Together, these two input variables to the Renshaw cells describe the level of activity in the motor output from the spinal cord. In either case, the dynamic responses of the interneurons depend only on their static activity before and after an input step occurs, but are otherwise indistinguishable. This favors the interpretation that the two input variables are equivalent under dynamic conditions, i.e., Renshaw cells respond to total motor output.